The Oxford Handbook of International Tax Law 9780192897688, 0192897683

International Tax Law is at a turning point. Increased tax transparency, the tackling of Base Erosion and Profit Shiftin

138 85 95MB

English Pages 1185 Year 2023

Table of contents :
Cover
The Oxford Handbook of International Tax Law
Copyright
Contents
Table of Cases
Table of Legislation
List of Abbreviations
List of Contributors
Part I History and Scope of International Tax Law
1. The History of International Tax Law
2. From the ‘1920s Compromise’ to the ‘2020s Compromise’
3. Sources of Law and Legal Methods in International Tax Law
4. Jurisdictional Underpinnings of International Taxation
5. International Tax Law and Customary International Law
6. International Tax Law and its Influence on National Tax Systems
7. International Tax Law and Personal Nexus
8. International Tax Law and Low- and Middle-Income Countries
9. International Tax Law: Status Quo, Trends, and Perspectives
Part II Relationship Between International Tax Law and Other Legal and Social Spheres
10. International Tax Law and Private International Law
11. International Tax Law and Public International Law
12. International Tax Law and Corporate Law
13. International Tax Law and International Trade Law
14. International Tax Law and Economic Analysis of Law
15. International Tax Law and Language
16. Comparative Tax Law
Part III Selected Issues on Tax Treaties and International Tax Law
17. Qualification Conflicts and Tax Treaties
18. Triangular Cases and Tax Treaties
19. The Future of Avoiding Double Taxation
20. Charities in Tax Conventions
21. Exchange of Information and Tax Treaties
22. Beneficial Ownership and Tax Treaties
23. The Principal Purpose Test under Tax Treaty Law
24. Tax Treaties and Human Rights Law
25. Taxation of International Partnerships
26. Regional Double Tax Treaty Models
27. Unilateralism, Bilateralism, and Multilateralism in International Tax Law
28. Agents in International Tax Treaties
Part IV Legal Aspects of International Transfer Pricing
29. The Role of Article 9 OECD MC
30. OECD Transfer Pricing Guidelines and International Tax Law
31. Corresponding Adjustments
32. Transfer Pricing versus Formulary Apportionment
Part V The Europeanization of International Tax Law
33. The Role of the ECJ in the Development of International Tax Law
34. Tax Treaties and EU Fundamental Freedoms
35. State Aid and International Taxation
36. International Tax Law and the EEA/EFTA
37. Twenty-First Century Tax Challenges of the EU Candidate Countries
38. European Anti-Tax-Avoidance Regimes
39. Alternative Dispute Resolution in the European Union
Part VI Selected Issues of Cross-Border Indirect Taxation
40. Aspects of Cross-Border VAT: The EU Approach and Evolving Trends
41. Taxation of Imports
42. ‘White Supplies’ and Double Taxation in Cross-Border VAT Law
43. A Comparison Between EU VAT Law and the OECD International VAT/GST Guidelines
Part VII Recent International Tax Trends in Major Economies and Regions
44. The US Perspective on International Tax Law
45. The Chinese Perspective on International Tax Law
46. The Indian Perspective on International Tax Law
47. The Brazilian Perspective on International Tax Law
48. The German Perspective on International Tax Law
49. The Perspective of the EAC on International Tax Law
50. The Japanese Perspective on International Tax Law
Part VIII Emerging Issues and the Future of International Tax Law
51. The Emerging Consensus on Value Creation: Theory and Practice
52. The Allocation of Taxing Rights under Pillar One of the OECD Proposal
53. Effective Minimum Taxation under Pillar Two of the OECD Proposal (‘GloBE’)
54. Challenges of the Emerging International Tax Consensus for Low- and Middle-Income Countries
55. Global Tax Governance
56. The Future of Labour Taxation and the ‘Rise of the Robots’
57. Digitalization and the Future of VAT in the European Union
Index

Recommend Papers

The Oxford Handbook of International Tax Law 9780192652348, 9780192897688, 0192652346

International Tax Law is at a turning point. Increased tax transparency, the tackling of Base Erosion and Profit Shiftin

112 85 3MB Read more

The Oxford Handbook of International Environmental Law 9780192589033, 9780198849155, 0192589032

The second edition of this leading reference work provides a comprehensive discussion of the dynamic and important field

104 40 4MB Read more

The Oxford Handbook of International Criminal Law 9780192558893, 9780198825203, 0192558897

In the past twenty years, international criminal law has become one of the main areas of international legal scholarship

99 85 2MB Read more

The Oxford Handbook of Jurisdiction in International Law 019878614X, 9780198786146

The Oxford Handbook of Jurisdiction in International Law provides an authoritative and comprehensive analysis of the con

627 24 5MB Read more

The Oxford Handbook of International Investment Law 9780199231386, 0199231389

This work offers a comprehensive account of the current state and likely future developments of international investment

100 42 6MB Read more

The Oxford Handbook of International Trade Law (2e) 9780192868381, 0192868381

The Oxford Handbook of International Trade Law offers extensive analysis and critique on the principles of modern intern

112 28 62MB Read more

The Oxford Handbook of International Refugee Law 9780198848639, 0198848633

This Handbook draws together leading and emerging scholars to provide a comprehensive critical analysis of international

118 32 10MB Read more

The Oxford Handbook of International Climate Change Law 9780199684601, 019968460X

Climate change presents one of the greatest challenges of our time, and has become one of the defining issues of the twe

113 110 7MB Read more

The Oxford Handbook of International Human Rights Law 9780199640133, 0199640130

The Oxford Handbook of International Human Rights Law provides an authoritative and original overview of one of the key

116 32 10MB Read more

The Oxford Handbook of Jurisdiction in International Law 9780198786146, 019878614X

The Oxford Handbook of Jurisdiction in International Law provides an authoritative and comprehensive analysis of the con

121 38 4MB Read more

The Oxford Handbook of International Tax Law
9780192897688, 0192897683

Author / Uploaded
Florian Haase

0 0 0
Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up

File loading please wait...

Citation preview

T h e Ox f o r d H a n d b o o k o f

I N T E R NAT IONA L TAX L AW

The Oxford Handbook of

INTERNATIONAL TAX LAW Edited by

FLORIAN HAASE and GEORG KOFLER

Great Clarendon Street, Oxford, ox2 6dp, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © The Several Contributors 2023 The moral rights of the authors have been asserted First Edition published in 2023 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Public sector information reproduced under Open Government Licence v3.0 (http://www.nationalarchives.gov.uk/doc/open-government-licence/open-government-licence.htm) Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2023933677 ISBN 978–0–19–289768–8 DOI: 10.1093/oxfordhb/9780192897688.001.0001 Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

Preface

It is with great pleasure and humility that we present The Oxford Handbook of International Tax Law. During the past roughly forty months, it has been a truly international endeavour with sixty-two authors from twenty-eight countries in all five continents, and we are grateful for this once-in-a-lifetime experience. All authors are distinguished scholars and experts, each of them contributing a unique and passionate perspective on a specific topic chosen by each author from a broader range of overarching fields that build the architecture of international tax law. There could not, indeed, be a better time for this Handbook. The past decade has seen numerous developments in international tax law that might lead the way to an entirely new era. Increased tax transparency, tackling ‘Base Erosion and Profit Shifting’, reconstruction of the network of bilateral tax treaties, renewed discussion about a fair and efficient allocation of taxing rights between States in a global, digitalized economy, the push for minimum corporate taxation, and attempts at good tax governance are some expressions of this shift. This new era also demonstrates the increased influence of international ‘standard setters’ such as the OECD, the UN, and the EU. Each of these developments alone has the potential to be ‘disruptive’ to the traditional world of international tax law, but together they have the potential to reshape the international tax system even in a sustainable way. Against this background, this Handbook aims to provide a truly international and thorough review of the current status of international tax law, which on the one hand addresses the central issues in this field and on the other hand evaluates the perspectives and directions international tax law might and should take in the light of current developments. Hence, it is time to take stock of the underlying ideas dominating the current system, revisit extrinsic influences, and identify the major issues and trends in tax treaty law, transfer pricing, indirect tax law, and EU tax law. However, it also needs to be recognized that an international tax system has domestic foundations, which requires a comprehensive view on international tax trends in major economies and regions. This provides a sound basis for identifying the emerging issues and the potential future of international tax law. The Handbook therefore reflects a unique perspective from the international compromise of the 1920s, when the contours of the current international tax regime emerged, to the rising foundations of a potential compromise of the 2020s. The Handbook would not have been possible without the passion, determination, and commitment of many. First and foremost, we wish to thank all authors for being the most important part of this project. We are grateful to the reviewers for their

vi Preface initial approval of the concept of this project and for their valuable input in shaping and improving the Handbook. Work on the manuscripts has been finalized in mid-2022. Last, but certainly not least, this book would not have been possible without the engagement and support of Oxford University Press and its staff, namely Victoria Harris, Charlotte Holloway, Lisa Butts, Jamie Berezin, John Smallman, and Brian Stone. Thank you! Florian Haase and Georg Kofler Hamburg and Vienna Winter 2022/2023

Contents

Table of Cases Table of Legislation List of Abbreviations List of Contributors

xiii xxxi lxi lxvii

PA RT I H I STORY A N D S C OP E OF I N T E R NAT IONA L TAX L AW 1. The History of International Tax Law Marilyne Sadowsky

3

2. From the ‘1920s Compromise’ to the ‘2020s Compromise’ Roberto Bernales Soriano

21

3. Sources of Law and Legal Methods in International Tax Law Rainer Prokisch

45

4. Jurisdictional Underpinnings of International Taxation H. David Rosenbloom and Fadi Shaheen

65

5. International Tax Law and Customary International Law Elizabeth Gil García

73

6. International Tax Law and its Influence on National Tax Systems Craig Elliffe

87

7. International Tax Law and Personal Nexus Michael Dirkis

107

8. International Tax Law and Low-and Middle-Income Countries Akhilesh Ranjan

123

9. International Tax Law: Status Quo, Trends, and Perspectives Reuven S. Avi-Yonah

143

viii Contents

PA RT I I R E L AT ION SH I P B E T W E E N I N T E R NAT IONA L TAX L AW A N D OT H E R L E G A L A N D S O C IA L SP H E R E S 10. International Tax Law and Private International Law Polina Kouraleva-Cazals

159

11. International Tax Law and Public International Law Christiana HJI Panayi and Katerina Perrou

177

12. International Tax Law and Corporate Law Marcos André Vinhas Catão and Verônica Souza

197

13. International Tax Law and International Trade Law Servaas Van Thiel

213

14. International Tax Law and Economic Analysis of Law Werner Haslehner

231

15. International Tax Law and Language Florian Haase

249

16. Comparative Tax Law Marco Barassi

265

PA RT I I I SE L E C T E D I S SU E S ON TAX T R E AT I E S A N D I N T E R NAT IONA L TAX L AW 17. Qualification Conflicts and Tax Treaties Gianluigi Bizioli

285

18. Triangular Cases and Tax Treaties Paolo Arginelli

299

19. The Future of Avoiding Double Taxation Martin Berglund

319

20. Charities in Tax Conventions Sigrid Hemels

335

21. Exchange of Information and Tax Treaties Xavier Oberson

353

Contents ix

22. Beneficial Ownership and Tax Treaties Dietmar Gosch and Nadia Altenburg

373

23. The Principal Purpose Test under Tax Treaty Law Robert J. Danon

391

24. Tax Treaties and Human Rights Law Philip Baker

417

25. Taxation of International Partnerships Ton Stevens

433

26. Regional Double Tax Treaty Models Craig West

455

27. Unilateralism, Bilateralism, and Multilateralism in International Tax Law Miranda Stewart 28. Agents in International Tax Treaties Sunita Jogarajan

473 497

PA RT I V L E G A L A SP E C T S OF I N T E R NAT IONA L T R A N SF E R P R IC I N G 29. The Role of Article 9 OECD MC Miguel Teixeira de Abreu

517

30. OECD Transfer Pricing Guidelines and International Tax Law Yuri Matsubara and Clémence Garcia

535

31. Corresponding Adjustments Matthias Hofacker

555

32. Transfer Pricing versus Formulary Apportionment Georgios Matsos

571

PA RT V T H E E U ROP E A N I Z AT ION OF I N T E R NAT IONA L TAX L AW 33. The Role of the ECJ in the Development of International Tax Law Adrian Cloer

591

x Contents

34. Tax Treaties and EU Fundamental Freedoms Marjaana Helminen

609

35. State Aid and International Taxation Patricia Lampreave Márquez

629

36. International Tax Law and the EEA/EFTA Patrick Knörzer

645

37. Twenty-First Century Tax Challenges of the EU Candidate Countries Savina Mihaylova-Goleminova

663

38. European Anti-Tax-Avoidance Regimes Paloma Schwarz Martínez

679

39. Alternative Dispute Resolution in the European Union Isabelle Richelle

697

PA RT V I SE L E C T E D I S SU E S OF C RO S S - B OR DE R I N DI R E C T TAX AT ION 40. Aspects of Cross-Border VAT: The EU Approach and Evolving Trends Roberto Scalia

731

41. Taxation of Imports Thomas Bieber

747

42. ‘White Supplies’ and Double Taxation in Cross-Border VAT Law Heidi Friedrich-Vache

765

43. A Comparison Between EU VAT Law and the OECD International VAT/GST Guidelines Eleonor Kristoffersson

785

PA RT V I I R E C E N T I N T E R NAT IONA L TAX T R E N D S I N M AJ OR E C ON OM I E S A N D R E G ION S 44. The US Perspective on International Tax Law Kimberly A. Clausing

799

Contents xi

45. The Chinese Perspective on International Tax Law Bristar Mingxing Cao

821

46. The Indian Perspective on International Tax Law Kuntal Dave

841

47. The Brazilian Perspective on International Tax Law Fernando Souza de Man

863

48. The German Perspective on International Tax Law Gerhard Kraft

879

49. The Perspective of the EAC on International Tax Law Afton Titus

899

50. The Japanese Perspective on International Tax Law Masao Yoshimura

917

PA RT V I I I E M E RG I N G I S SU E S A N D T H E F U T U R E OF I N T E R NAT IONA L TAX L AW 51. The Emerging Consensus on Value Creation: Theory and Practice Allison Christians

937

52. The Allocation of Taxing Rights under Pillar One of the OECD Proposal Aitor Navarro

951

53. Effective Minimum Taxation under Pillar Two of the OECD Proposal (‘GloBE’) Joachim Englisch

969

54. Challenges of the Emerging International Tax Consensus for Low-and Middle-Income Countries Natalia Quiñones

991

55. Global Tax Governance Irma Mosquera Valderrama

1007

56. The Future of Labour Taxation and the ‘Rise of the Robots’ Georg Kofler

1025

xii Contents

57. Digitalization and the Future of VAT in the European Union Michael Tumpel

1041

Index

1059

Table of Cases

UNITED KINGDOM

Anson v Commissioners for Her Majesty’s Revenue and Customs [2015] UKSC 44�� 101n.92 Clark (Inspector of Taxes) v Oceanic Contractors Inc. [1983] 2 AC 130 (HL)�� 104n.105 DSG Retail Ltd & others v HMRC TC00001 [2009] UKFTT 31 (TC) ��183n.29 Government of India v Taylor [1955] AC 491�� 166n.39 Henderson v Henderson [1965] All ER 179�� 115n.34 Inco Europe Ltd. and Others v. First Choice Distribution (A Firm) and Others, [2000] UKHL 15�� 847n.19 Indofood International Finance Ltd. v JP Morgan Chase Bank [2006] EWCA Civ 158��381n.51, 382n.60 Pastoral Fin. Ass’n [1914] AC 1088 ��942–43n.14 Stock v Frank Jones (Tipton) Ltd., [1978] 1 W.L.R�� 847n.18 Udny v Udny (1869) LR I Sc & Div 441��115n.35 EUROPEAN UNION CASES European Commission Decision

European Commission Decision 1999/779/EC of 3 February 1999 [1999] OJ L305/27��649n.18 European Commission Decision of 21 October 2015, Case SA.38374, on state aid SA.38374 (2014/C ex 2014/NN) implemented by the Netherlands to Starbucks, No. 2017/502/UE [2017] OJ L83�� 639, 639n.43, 641 European Commission Decision of 30 August 2016 on state aid, SA.38373 (2014/C), No. 2017/1283 [2017] OJ L187��641n.49 Court of Justice of the European Union

Case C-6/64 Flaminio Costa v ENEL�� 667n.13, 884n.3 Case C-45/64 Commission of the European Communities v. Italian Republic�� 659n.64 Case C-28/67 Firma Molkerei-Zentrale Westfalen/Lippe GmbH ��659n.58 Case C-29/69 Stauder��259n.40, 611n.16 Case C-8/74 Dassonville �� 620n.67 Case C-41/74 Yvonne van Duyn�� 787n.10 Case C-43/75 Defrenne ��596n.26 Case C-11/77 Patrick��596n.26 Case C-120/78 Cassis de Dijon �� 620n.66, 620n.67

xiv Table of Cases Case C-15/81 Gaston Schul Douane-expediteur BV�� 658n.57 Case C-258/81 Metallurgiki Halyps ��596n.26 Case C-283/81 CILFIT �� 259n.41, 888n.14 Case C-42/83 Dansk Denkavit��763n.52, 763n.53 Case C-270/83 Commission v. French Republic��228n.77, 610n.8, 612n.20, 623n.83, 624n.91 Case C-47/84 Staatssecretaris van Financiën �� 658n.57 Case C-168/84 Gunter Berkholz��735, 735n.29, 735n.30, 736, 736n.39, 736n.48, 737, 743n.100, 745n.108 Case C-248/84 Germany v Commission�� 648n.13 Case C-249/84 Profant�� 773n.25 Case C-127/86 Ledoux�� 773n.25 Case C-286/86 Deserbais�� 610n.7, 610n.11 Case C-81/87 Daily Mail��613n.25, 613n.26, 614n.31 Case C-102/87 France v. European Commission ��631n.11, 650n.20 Case C-142/87 Belgium v. European Commission��631n.11, 650n.21 Case C-235/87 Matteucci�� 610n.7, 610n.11 Case C-50/88 Kühne��775n.31 Case C-175/88 Biehl �� 613n.24 Joined Cases C-6/90 and C-9/90 Andrea Francovich and Danila Bonifaci & Others��885n.12 Case C-41/90 Klaus Höfner and Fritz Elser��631n.10 Case C-204/90 Bachmann��621–22, 621n.72 Case C-300/90 Commission v. Belgium �� 621n.72 Case C-193/91 Mohsche��775n.31 Case C-330/91 The Queen v. Inland Revenue Commissioners�� 228n.78, 621n.73 Opinion 1/92��647n.7 Case C-24/92 Corbiau �� 722n.147, 722n.148 Case C-68/92 Commission v. French Republic�� 773n.25 Joined Cases C-278/92 and C-280/92 Spain v. European Commission�� 631n.11, 649n.19 Case C-62/93 BP Soupergaz�� 755n.26 Case C-279/93 Schumacker�� 610n.9, 613n.23, 613n.24, 614n.29, 620n.65, 648n.11 Joined Cases Roders (C-367/93–377/93); Matteucci (235/87); Kortmann (32/80); and Gottardo (C-55/00) �� 616n.41, 617n.49 Case C-415/93 Bosman��613n.25 Case C-55/94 Gebhard��613n.25, 613n.27, 620n.67 Case C-80/94 Wielockx�� 613n.23, 622n.74, 622n.77 Case C-107/94 Asscher��613n.23, 614n.31, 621n.73 Case C-215/94 Mohr��766n.2 Case C-231/94 Faaborg-Gelting Linien A/S and Finanzamt Flensburg �� 736n.39, 736n.41 Case C-241/94 France v. European Commission ��632n.17, 632n.21

Table of Cases xv Case C-327/94 Dudda�� 732n.7 Case C-190/95 ARO Lease BV and Inspecteur van de Belastingdienst Grote Ondernemingen, Amsterdam (1997)��735n.30, 736n.42, 736n.46, 738–40, 738n.58 Case C-250/95 Futura�� 614n.29, 620n.67 Cases C-260/95 Commissioners of Customs and Excise v. DFDS A/S�� 735n.30, 738–40, 738n.57 Case C-296/95 EMU Tabac et al.�� 259n.41, 259n.42 Case C-337/95 Parfums Christian Dior �� 723n.149 Case C-118/96 Safir��613n.25, 613n.26 Case C-264/96 ICI �� 614n.31, 623n.80, 624n.91, 624n.93, 624n.94, 625n.102 Case C-336/96 Gilly �� 226n.64, 419n.7, 609n.1, 615n.33, 615n.34, 615n.36, 615n.37, 617n.51, 617n.52 Case C-390/96 Lease Plan Luxembourg SA �� 736n.43, 736n.46, 736n.48 Case C-48/97 Kuwait Petroleum��1052n.32 Case C-212/97 Centros�� 600n.43, 623n.80, 623n.86, 657n.50, 680n.8 Case C-294/97 Eurowings�� 623n.80, 623n.81, 623n.83, 624n.91 Case C-307/97 Compagnie de Saint-Gobain, Zweigniederlassung Deutschland��228n.74, 610n.9, 610n.10, 612n.20, 615n.33, 615n.36 Case C-311/97 Royal Bank of Scotland��228n.75, 613n.23 Case C-391/97 Gschwind��613n.23 Case C-35/98 Verkooijen��620n.67, 622n.74 Case C-55/98 Vestergaard ��614n.29 Case C-200/98 X AB and Y AB��614n.31 Case C-251/98 C. Baars��228n.74, 613n.26, 614n.31, 622n.74 Joined Cases C-397/98 and C-410/98 Metallgesellschaft��624n.91 Case C-415/98 Bakcsi��775n.31 Joined Cases C-446-469/98, C-471-472/98 and C-475-476/98, Open Skies�� 610n.5, 612n.22 Case C-110/99 Emsland Stärke��602n.59, 624n.94 Case C-143/99 Adria-Wien Pipeline GmbH and Wietersdorfer & Peggauer Zementwerke GmbH��630n.7 Case C-55/00 Compare to Gottardo�� 612n.22, 616n.45 Case C-136/00 Danner��621n.73, 623n.82 Case C-208/00 Überseering��614n.31 Case C-257/00 Givane et al.��259n.42 Case C-324/00 Lankhorst-Hohorst�� 607n.85, 620n.67, 623n.80, 624n.90, 624n.91, 624n.94 Case C-436/00 X and Y �� 620n.67, 622n.76, 624n.91 Case C-152/01 Kyocera�� 259n.41 Case C-167/01 Inspire Art ��623n.80 Case C-168/01 Bosal Holding �� 622n.74 Case C-185/01 Auto Lease Holland BV��776n.36

xvi Table of Cases Case C-209/01 Schilling��614n.31 Case C-364/01 Barbier��623n.80 Case C-422/01 Skandia and Ramstedt��621n.73, 623n.82, 623n.83 Case C-452/01 Ospelt und Schlössle Weissenberg �� 647n.4, 653n.31 Case C-9/02 de Lasteyrie du Saillant ��208n.58, 620n.67, 624n.94 Case C-224/02 Pusa��614n.31 Case C-255/02 Halifax and Others�� 602n.58, 623n.85, 624n.94, 669n.20 Case C-315/02 Lenz��623n.81, 624n.90 Case C-319/02 Manninen�� 622n.74, 654n.32 Case C-32/03 Fini H�� 669n.20 Case C-88/03 Portuguese Republic v. Commission of the European Communities�� 630n.6, 631n.14 Joined Cases C-182/03 and C-217/03 Kingdom of Belgium v. European Commission�� 638n.41 Case C-188/03 Junk��259n.40 Case C-253/03 CLT-UFA SA��228n.76 Case C-268/03 De Baeck��614n.31 Case C-376/03 D��612n.21, 616n.42 Case C-446/03 Marks & Spencer �� 620n.67, 621n.68, 624n.92, 624n.94, 647n.3 Case C-513/03 van Hilten��615n.36 Cases C-23/04, C-24/04, and C-25/04 Sfakianakis�� 594n.16 Case C-196/04 Cadbury Schweppes and Cadbury Schweppes Overseas�� 603n.64, 605–6, 605n.78, 620n.67, 623n.81, 624n.94, 624n.95, 624n.96, 625n.100, 625n.102, 680n.8 Case C-210/04 Ministero dell’Economia e delle Finanze, Agenzia delle Entrate��737n.54, 737n.56, 740n.74, 741, 741n.77 Case C-265/04 Bouanich��610n.7, 615n.36 Case C-290/04 Scorpio ��615n.36 Case C-374/04 Test Claimants in Class IV of the ACT Group Litigation ��615n.36, 616n.43, 617n.52, 626n.106 Case C-386/04 Stauffer�� 338, 338n.12 Joined Cases C-439/04 and C-440/04 Kittel and Recolta Recycling �� 669n.20 Case C-470/04 N�� 615n.34, 615n.37, 615n.38, 617n.51 Case C-471/04 Keller Holding�� 650n.24 Case C-513/04 Kerckhaert and Morres�� 615n.33, 615n.34, 615n.37, 617n.51 Case C-520/04 Turpeinen��614n.31 Case C-524/04 Thin Capitalization Group Litigation Order��615n.34, 615n.37, 617n.51, 620n.67, 624n.94, 625n.101 Case C-101/05 Skatteverket�� 655n.41 Case C-111/05 Aktiebolaget NN �� 774n.27 Case C-167/05 Commission v. Sweden�� 659n.61 Case C-170/05 Denkavit�� 654, 654n.35

Table of Cases xvii Case C-231/05 Oy AA��615n.34, 615n.36, 620n.67, 621n.68 Case C-298/05 Columbus Container Services BVBA & Co.��325n.11, 450n.64, 605n.79, 617n.50, 618n.56, 618n.57 Case C-321/05 Kofoed�� 600n.43, 691n.73 Case C-379/05 Amurta�� 654, 654n.33 Case C-383/05 Talotta��613n.23, 613n.24 Case C-432/05 Unibet��595n.17 Case C-451/05 ELISA�� 655n.42 Case C-73/06 Planzer Luxembourg Sàrl ��735n.31, 735n.33, 736n.36 Case C-194/06 Orange European Smallcap Fund��616n.44, 617n.50 Case C-210/06 Cartesio�� 614n.31, 657n.51 Case C-293/06 Deutsche Shell��615n.35 Case C-414/06 Lidl��620n.67, 621n.68 Case C-487/06 British Aggregates �� 629n.1 Case C-43/07 Arens-Sikken��615n.36 Case C-157/07 Krankenheim Ruhesitz am Wannsee-Seniorenheimstatt ��320n.2, 613n.27, 615n.35, 615n.36, 620n.67 Case C-291/07 Kollektivavtalsstiftelsen TRR Trygghetsrådet�� 732n.7 Case C-303/07 Aberdeen��654n.34 Case C-318/07 Persche��774n.29 Case C-328/07 Persche��338, 338n.13 Case C-349/07 Sopropé�� 724n.156 Case C-521/07 Commission v Netherlands��656n.43 Case C-540/07 Commission v. Italy�� 654n.37 Case C-67/08 Block �� 617n.50 Joined Cases C-78-80/08 Paint Graphos and Others�� 631n.12 Case C-96/08 CIBA�� 617n.50 Case C-128/08 Damseaux �� 617n.50 Case C-182/08 Glaxo Wellcome �� 621n.68, 629n.1 Case C-230/08 Dansk Transport�� 753n.15 Case C-242/08 Swiss Re Germany Holding�� 773n.25 Case C-250/08 Commission v. Belgium�� 622n.75 Cases C-261/08 and C-348/08 Zurita García and Choque Cabrera �� 260n.44 Case C-311/08 SGI�� 621n.68 Case C-337/08 X Holding�� 621n.68 Joined Cases C-436/08 and C-437/08, Haribo��617n.52, 618n.58 Case C-511/08 Handelsgesellschaft Heinrich Heine�� 260n.47 Case С-2/09 Regionalna Mitnicheska Direktsia –Plovdiv�� 674 Case C-72/09 Établissements Rimbaud SA�� 655–56, 655n.39 Joined Cases C-106/09 P and C-107/09 P European Commission (C-106/09 P) and Kingdom of Spain (C-107/09 P)�� 632n.20, 649n.15, 685n.32

xviii Table of Cases Case C-155/09 Commission v. Greece ��613n.24, 647n.9 Case C-196/09 Miles�� 722n.144, 723n.149 Case C-253/09 Commission v. Hungary �� 622n.75 Case C-284/09 Commission v. Germany��654n.38 Case C-203/10 Direktsia “Obzhalvane i upravlenie na izpalnenieto” –Varna�� 674 Case C-218/10 ADV Allround Vermittlungs AG, in liquidation�� 732n.7 Case C-240/10 Schulz-Delzers ��618n.53, 619n.63 Case C-318/10 SIAT�� 625n.103 Case C-371/10 National Grid Indus ��208n.57, 621n.71, 657n.52 Case C-400/10-PPU J.McB�� 724n.159 Case C-412/10 Homawoo�� 260n.47 Case C-414/10 Véleclair�� 760n.45 Case C-417/10 3M Italia�� 686n.34 Case C-567/10 Inter-Environnement Bruxelles et al.�� 260n.45 Case C-617/10 Fransson�� 724n.158 Joined Cases C-621/10 and C-129/11 Balkan and Sea Properties ADSITS and Provadinvest OOD�� 674 Case C-35/11 Test Claimants in the FII Group Litigation 2�� 619n.59 Case C-48/11 A Oy �� 656n.44 Case C-80/11 and C-142/11 Mahagében and Dávid�� 669n.20 Case C-118/11 Eon Aset Menidjmunt OOD�� 674 Case C-234/11 TETS Haskovo AD �� 674 Case C-243/11 RVS Levensverzekeringen NV��660n.66 Case C-261/11 Commission v. Denmark �� 656n.46 Case C‑284/11 EMS-Bulgaria Transport OOD�� 674 Case C-285/11 Bonik EOOD �� 674 Joined Cases C-318/11 and C-319/11 Daimler AG, Widex A/S v. Skatteverket��738n.57, 738n.58 Case C-388/11 Le Crédit Lyonnais�� 732n.9 Case C-418/11 Texdata Software �� 724n.158 Case C-425/11 Ettwein ��612n.18 Case C-549/11 Direktor na Direktsia ‘Obzhalvane i upravlenie na izpalnenieto’ —grad Burgas pri Tsentralno upravlenie na Natsionalnata agentsia za prihodite ��674, 1052n.29 Case C-550/11 PIGI —Pavleta Dimova ET �� 674 Case C-642/11 Stroy trans EOOD�� 674 Case C-47/12 Kronos ��617n.52 Case C-78/12 ‘Evita-K’ EOOD�� 674 Case C-79/12 SC Mora IPR ��763n.52, 763n.55 Case C-138/12 Rusedespred OOD�� 674 Case C-142/12 Hristomir Marinov�� 674 Case C-164/12 DMC��208n.57, 692n.83 Joined Cases C-204-208/12 Essent Belgium �� 647n.9

Table of Cases xix Case C-276/12 Sabou �� 371n.110, 724n.156 Case C-282/12 Itelcar ��C34P60 n.103 Case C-283/12 Serebryannay vek�� 1052n.29 Case C-298/12 Confédération paysanne �� 260n.48 Case C-303/12 Imfeld and Garcet�� 609n.4 Case C-480/12 X�� 752n.10 Case C-605/12 Welmory sp. z o.o.��732n.7, 734n.28, 735n.31, 736n.44, 738–40, 738n.59, 738n.61 Case C-7/13 Skandia��740–42, 740n.72, 740n.74, 740–41n.75, 741n.77, 741n.79, 741n.84, 742n.86, 743 Case C-18/13 Maks Pen EOOD�� 669n.20, 674 Case C-107/13 FIRIN OOD�� 674 Case C-377/13 Ascendi Beiras Litoral e Alta ��699n.10, 723n.149 Case C-489/13 Verest and Gerards��618n.53, 619n.63 Case C-492/13 Traum EOOD �� 674 Case C-589/13 Familienprivatstiftung Eisenstadt��621n.69 Case C-15/14 P MOL Magyar Olaj-és Gázipari Nyrt��636n.29 Case C-76/14 Mihai Manea�� 674 Case C-111/14 GST –Sarviz AG Germania�� 674 Case C-112/14 Commission v. United Kingdom��625n.101 Case C-187/14 DSV Road�� 760, 760n.46, 761 Joined Cases C-226/14 and C-228/14 Eurogate Distribution��753n.14 Case C-241/14 Bukovansky��615n.34 Case C-264/14 Skatteverket v. David Hedqvist�� 745n.109 Case C-388/14 Timac Agro Deutschland�� 621n.70, 621n.71, 622n.78 Case C-413/14 P Intel v. Commission�� 164n.25 Case C-419/14 WebMindLicenses��774n.28 Case C-24/15 Josef Plöckl�� 759n.41 Case C-48/15 NN��614n.29 Case C-176/15 Riskin and Timmermans ��617n.46 Case C-284/15 Achmea��700n.17, 701n.18, 701n.19, 701n.21, 701n.24, 721n.141, 721n.142, 722n.145 Case C-432/15 Baštová ��1051n.26 Case C-478/15 Radgen��612n.18 Case C-571/15 Wallenborn Transports��753n.14 Case C-580/15 Van der Weegen and Pot��614n.30 Case C-648/15 Austria v. Germany��174n.75, 187n.53, 257n.34, 264n.57, 607n.83, 720n.135, 725n.162 Case C-682/15, Berlioz Investment Fund SA�� 371n.110, 427n.30, 427n.31 Case C-6/16 Eqiom and Enka�� 624n.95, 625n.104, 680n.5 Case C-14/16 Euro Park Service ��625n.104, 680n.5, 680n.6 Case C-20/16 Bechtel�� 619n.63 Case C-21/16 Euro Tyre BV�� 759n.41

xx Table of Cases Case C-115/16 N Luxembourg�� 600n.44, 600n.48, 600n.49, 600n.50, 601n.51, 601n.52, 601n.53, 601n.54, 602n.63 Joined Cases C-115/16 N Luxembourg 1, C-118/16 X Denmark, C-119/16 C Denmark I and C-299/16 Z Denmark�� 379n.35, 600, 623n.87, 623n.88, 625n.99, 625n.101, 625n.105, 680n.7 Case C-132/16, Direktor na Direktsia “Obzhalvane i danachno-osiguritelna praktika” –Sofia�� 674 Case C-154/16 Latvijas dzelzceļš��753n.14 Case C-203/16 P Heitkamp BauHolding v. Commission�� 631n.13 Case C-298/16, Ispas�� 724n.156 Case C-355/16 Picart��612n.18 Case C-359/16 Altun��600n.49 Case C-504/16 Deister Holding�� 680n.6 Case C-544/16 Marcandi�� 775n.30 Case C-552/16 “Wind Inovation 1” EOOD�� 674 Case C-553/16 ‘TTL’ EOOD�� 674 Case C-34/17 Donnollan �� 725n.161 Case C-58-581/17 Martin Wächtler�� 612n.18, 894n.23 Case C-108/17 Enteco Baltic ��758n.40 Case C-135/17 X-GmbH��605–6, 605n.77, 624n.98 Case C-410/17 A Oy�� 1052n.29 Case C-440/17 GS��604n.66 Case C-528/17 Milan Božičevič Ježovnik��759n.42 Case C-531/17 Vetsch �� 759n.43 Case C-602/17 Sauvage and Lejeune��609n.4, 615n.35 Case C-647/17 Srf konsulterna AB�� 769n.12 Case C-26/18 Federal Express�� 753n.14, 753n.17, 754 Case C-74/18, A Ltd��660n.66 Case C-189/18 Glencore�� 724n.156 Case C-242/18 ‘UniCredit Leasing’ EAD�� 674 Case C-276/18 KrakVet Marek Batko sp.k. ��773n.26 Case C-323/18 Tesco-Global Áruházak �� 601n.55 Case C-401/18 Herst, s.r.o.�� 766–67n.3 Case C-446/18 Agrobet ��724n.157 Case C-482/18 Google Ireland Ltd �� 164n.29, 620n.67 Case C-547/18 Dong Yang Electronics sp. z o.o. �� 732n.7, 735n.30, 738–40, 738n.60, 739–31nn.63–5, 769n.14 Case C-565/18 Société Générale SA �� 164n.23, 164n.24, 164n.28 Case C-43/19 Vodafone Portugal �� 1052n.28 Joined Cases C-168/19 and C-169/19 HB and IC��615n.34 Joined Case C-245/19 and C-246/19 State of Luxembourg�� 371n.110, 427n.32 Case C-337/19 P Magnetrol International and Belgium v. European Commission�� 637n.33

Table of Cases xxi Case C-337/19 P Magnetrol International and Belgium v. European Commission and France v. European Commission �� 637n.34 Case C-403/19 Société Générale��702n.27 Case C-562/19 Commission v. Poland ��162n.20, 175n.81, 668n.17 Case C-562/19 P European Commission v Republic of Poland��668n.17 Case C-621/19 Weindel Logistik Service �� 761n.48 Case C-638/19 P Commission v. European Food SA and others��701n.24 Case C-741/19 Republic of Moldova v. Komstroy LLC�� 701n.22 Case C-812/19 Danske Bank A/S, Danmark, Sverige Filial ��740–42, 740n.73, 741n.78, 741n.80, 742n.85, 743 Case C-931/19 Titanium Ltd v. Finanzamt Österreich �� 736n.45, 740n.70, 745 Case C-4/20 “ALTI” OOD�� 674 Case C-7/20 VS��754n.20 Case C-109/20 Republiken Polen��701n.24 Case C-257/20 ‘Viva Telecom Bulgaria’ EOOD�� 674 Case C-333/20 Berlin Chemie A. Menarini SRL��738–40, 739n.67, 739n.69, 740n.71, 744n.103, 745n.108 Case C-489/20 UB ��755n.25, 760n.47 Case C-1/21, Direktor na Direktsia “Obzhalvane i danachno-osiguritelna praktika”�� 674 European Court of Human Rights

Bendenoun v France App. No. 12547/86�� 423n.18 Burden v United Kingdom, App. No. 13378/05��278n.84 Darby v Sweden App. No. 11581/85 ��428n.34 Di Belmonte v Italy App. No. 72638/01��430n.37 Ferrazzini v Italy App. No. 44759/98��423n.17 FS v Germany App. No. 30128/96��420n.11, 425 Gall v Hungary App. No. 49570/11 ��430n.36 GSB v Switzerland App. No. 28601/11�� 420–21n.13 H v Sweden App. No. 12670/87�� 420n.9 Hanzmann v Austria App. No. 12560/86�� 419n.7 HM v Germany App. No. 62512/00�� 423n.19 Jussila v Finland App. No. 73053/01 �� 423n.22 Kaira v Finland App. No. 27109/95��430n.36 Lindkvist v Denmark App. No. 25737/94��420n.10 NKM v Hungary App. No. 66529/11��430n.36 Othymia Investments BV v Netherlands App. No. 75292/10�� 420–21n.13 Remy v Germany App. No. 70826/01�� 423n.19 RSz v Hungary App. No. 41838/11��430n.36 Wasa Liv v Sweden App. No. 13013/87��430n.36

xxii Table of Cases European Free Trade Association Court

Case E-2/97 Mag Instrument Inc. v. California Trading Co. Norway��647n.7 Case E-3/98 Herbert Rainford-Towning��647n.8 Case E-6/98 Norway v. ESA��646n.1 Case E-1/01 Einarsson ��658n.56 Case E-1/01 Einarsson ��646n.1 Case E-1/03 ESA v. Iceland��646n.1, 659n.63 Case E-1/04 Fokus Bank�� 647n.3, 653–54 Case E-2/06 EFTA Surveillance Authority v. Norway�� 647n.6 Case E-6/07 HOB vín ehf and Faxaflóahafnir sf��659n.60, 659n.62 Case E-7/07 Seabrokers v. Staten v/Skattedirektoratet��652n.28 Joined Cases E-4/10, E-6/10 and E-7/10 Liechtenstein and Others��649n.15, 650n.22 Joined Cases E-17/10 and E-6/11, Principality of Liechtenstein and VTM Fundmanagement AG��650n.22 Case E-15/11 Arcade Drilling�� 658n.53 Case E-16/11 ESA v. Iceland��647n.10 Joined Cases E-3/13 and E-20/13 Fred. Olsen and Others v. Norwegian State��652n.29 Case E-14/13 The EFTA Surveillance Authority v. Iceland�� 658n.55 Case E-15/16 Yara International ASA v. Norwegian Government��651n.25 Case E-3/21 PRA Group Europe AS v. Staten v/Skatteetaten�� 651n.27 European Free Trade Association Court Surveillance Authority

Decision No. 149/99/COL of 30 June 1999�� 649n.14, 649n.16, 650n.23 Decision No. 3/17/COL of 18 January 2017��649n.17 General Court

Joined Cases T-755/15 and T-759/15 Grand Duchy of Luxembourg and Fiat Chrysler Finance Europe v. European Commission�� 640n.47 Joined Cases T-760/15 and T-636/16 The Netherlands v. European Commission��639n.44, 684n.28 Joined Cases C-116/16 T Denmark and C-117/16 Y Denmark��378n.31, 379n.34, 600n.44, 600n.48, 602n.63, 623n.87, 623n.88, 625n.99, 625n.101, 625n.105, 680n.7 Joined Cases T-131/16 and 263/16 Magnetrol International and Belgium v. European Commission�� 636n.31 Joined Cases T-778/16 and T-892/16 Ireland/Commission��671–72, 684n.28 Joint Cases T-778/16 and T-892/16 Ireland v. European Commission and Apple Sales International/Apple Operations Europe v. European Commission��641n.48, 671–72, 684n.28 Joined Cases T-778/16 and T-892/16 Ireland and Others v Commission (Case C-465/20 P, 2021/C 35/33 ��641n.51

Table of Cases xxiii T-892/16 Apple Sales International and Apple Operations Europe v Commission �� 671–72 INTERNATIONAL CASES International Court of Justice (ICJ)

Asylum (Colombia v Peru), Judgment of 20 November 1950, ICJ Reports 1950, 276�� 76n.15, 77n.26 Case Concerning the Military and Paramilitary Activities in and against Nicaragua (Nicaragua v United States of America), ICJ Reports 1986, 14��69n.19 North Sea Continental Shelf Cases (Federal Republic of Germany v Denmark; Federal Republic of Germany v Netherlands), ICJ Reports 1969, 3�� 69n.17, 69n.19, 70n.20, 70n.21, 70n.23, 77, 77n.24, 77n.30, 92n.36 Pulau Ligitan and Pulau Sipadan (Indonesia v Malaysia), ICJ Reports 2002, 645��55n.46 International Centre for Settlement of Investment Disputes (ICSID)

Burlington Resources Inc. v Republic of Ecuador (2012), ICSID Case No. ARB/08/5 �� 192n.77, 193n.78, 193n.79, 193n.81 Cargill, Incorporated v United Mexican States (2009), ICSID Case No. ARB(AF)/05/2��193n.83 International Covenant on Civil and Political Rights (ICCPP)

Lederbauer v Austria, CCPR/C/90/D/1454/2006, decision of 11 September 2007�� 423n.21 Permanent Court of Arbitration (PCA)

Cairn (PCA Case No. 2016-07)��414n.122 Permanent Court of International Justice (PCIJ)

SS ‘Lotus’ (France v Turkey), 1927 PCIJ, Ser. A, No. 10 (7 September)�� 66, 66n.4, 67n.5, 166n.37 United Nations Commission on International Trade Law (UNCITRAL)

EnCana Corporation v Republic of Ecuador (2006), LCIA Case No. UN3481, UNCITRAL�� 193n.84 World Trade Organization (WTO)

1952: Italy ship plates�� 221n.36

xxiv Table of Cases 1958: Italy agricultural machinery�� 221n.36 1976: Income Tax Practices Maintained by Belgium (L/442-23S/127), France (L/442-23S/114), and Netherlands (L/442-23S/137)��223n.51 1976: US Tax Legislation (DISC) Panel (L/4422 BISD 23S/98) �� 223n.50 1982: BISD 28S/114 �� 223n.52 1987: United States—Tax Reform Legislation for Small Passenger Aircraft (L/6153)��221n.37 1987: US Taxes on petroleum (L/6175 GATT DOC BISD 34S/136�� 218n.24 1988: US Customs fee��215n.10 1992: US Tuna I�� 219–20n.31 1994: US Tuna II �� 219–20n.31 1994: US—Taxes on Automobiles��217n.23 1996: DS2 US Gasoline point 6.37�� 219n.30 1997: DS27 EC bananas��225n.60 1997: DS27 EC Bananas AB Report point 220��225–26n.62 1998–2000: DS 139, 142 Canada Automotive Industry��222n.42 1998: DS 56 Argentina Textiles and Apparel ��215n.10 1998: DS54 Indonesia car tax��218n.25 1998: DS56 Argentina Textiles and Apparel��215n.8 1998: DS58 US Shrimp Turtle AB points 50 and 141�� 219n.30 1999: 2000: DS87, 110 Chile tax on alcohol��218n.25 1999: 2005: DS308 Mexico tax on soft drinks ��218n.25 1999: 2010: DS371 Thai tax on cigarettes��218n.25 1999: DS70 Canada Export of civilian aircraft��222n.44 1999: DS75, 84 Korea tax on alcohol��218n.25 2000: DS139 Canada Automotive, AB Report point 155��225–26n.62 2001–2: DS 108 US Foreign Sales Corporations, Panel and Appellate Body�� 223n.50 2001: DS58—US Shrimp Turtle�� 219–20n.31 2001: ETI��221 2001: DS 155 Argentina—exports of bovine hides and import of finished leather �� 220n.35 2001: DS108 US Foreign Sales Corporations��221n.38 2001: DS135 EC Asbestos point 115 �� 219n.30 2002: DS108 US Foreign Sales Corporations�� 222n.43 2004: DS246 EC Tariff Preferences�� 215n.9 2005: DS269/286 EU customs classification of chicken cuts��215n.8 2005: DS302 Dominican Republic Import and sale of Cigarettes��215n.8 2008: DS360 India Duties on Imports from the US��215n.8 2009–10: DS70, DS71 Canada Export Civil Aircraft and ongoing DS522 Canada Trade in Commercial Aircraft��224–25n.58 2010: DS375/377 EU classification of IT products��215n.8 2011–20: DS316, 317, 353 US Large Civil Aircraft, and 2010–19: DS316 EC Large Civil Aircraft ��224–25n.58

Table of Cases xxv 2015: DS453 Argentina, Panel Report�� 225–26n.62, 227n.69 2019: DS510 US Renewable Energy��222n.40 2019: DS510 US Renewable Energy; 2017–18: DS472 Brazil Taxation�� 222n.39 DS537: Canada Measures Governing the Sale of Wine�� 218n.26 DS563: US Renewable Energy��222n.40 DS608: Russian Federation—Measures Concerning the Exportation of Wood Products ��215n.14 NATIONAL CASES Australia

Commissioner of Taxation v SNF (Australia) Pty [2011] FCAFC 74 ��184n.35 Harding v Commissioner of Taxation [2018] FCA 837�� 114n.31 Lamesa Holding BV v Federal Commissioner of Taxation [1997] FCA 134, 35 ATR 239��395n.19 Lamesa Case Federal Court of Australia, 20 August 1997, 36 ATR 589 (1997), 77 FCR 597�� 406–7n.83, 415n.127 Bulgaria

Interpretative Decision No. 3, Bulgarian Supreme Administrative Court, 6 June 2008�� 675n.37 Canada

A Holding ApS v Federal Tax Administration, 8 ITLR 536�� 396–97, 397n.26, 397n.30 Alta Energy Luxembourg SARL v R, 2020 FCA 43, 22 ITLR 509 ��392–93n.6, 403n.61, 403n.64, 410n.99, 411n.101 Alta Energy Luxembourg SARL v R, 21 ITLR 219��392–93n.6, 395n.19, 400n.49, 403n.61, 403n.62 Canada Trustco Mortgage Co. v Canada, 2005 SCC 54, [2005] 2 SCR 601 �� 403n.69 Canada v Alta Energy Luxembourg SARL [2021] SCC 49, 24 ITLR 346��392–93n.6, 393, 398, 403n.65, 411n.102, 411n.103, 412n.109, 412n.111, 413–14, 414n.124, 416 Garron and another v R; Re Garron Family Trust; Garron v R; St Michael Trust Corp. v R; Re Fundy Settlement; Dunin v R; St Michael Trust Corp. v R; Re Summersby Settlement, 12 ITLR 79�� 410–11n.100 MIL (Investments) SA v Canada, 9 ITLR 29, Tax Court of Canada and Federal Court of Appeal, 9 ITLR 1111�� 397n.28 Prévost Car Inc. v Canada, 2008 TCC 231, 2004-4226 (IT) G�� 382n.53, 382n.55 R v Crown Forest Industries Ltd (1995) DTC 5,389��101n.93 R v Prevost Car Inc. [2009] FCA 57, [2010] 2 FCR 65��101–2n.96

xxvi Table of Cases Thomson v Minister of National Revenue [1946] DTC 812�� 113n.27 Velcro Canada Inc. v Canada, 2012 TCC 57, 2007-1806 (IT) G�� 382n.54, 382n.59 Verdannet, Re, 20 ITLR 832, 862�� 395n.16, 396n.20, 396n.21, 400n.50 Vodafone International Holdings BV v Union of India and another, 14 ITLR 431��396–97, 397n.28, 397n.29, 399n.47 Yanko-Weiss Holdings 1 (1996) Ltd v Holon Assessing Office, 10 ITLR 524�� 396–97, 397n.27, 397n.32 Denmark

Cook Case, The SKM 2011,485 LSR��384n.76 HHU Case, The SKM 2011.57 LSR�� 384n.74 ISS Case, The SKM 2010.268, previously SKM 2011.121�� 385n.78 France

Artémis, 24 November 2014, no. 363556�� 170n.58 France v Zimmer Ltd., French Conseil d’État, 31 March 2010, Nos. 304715, 308525�� 504n.35 French Conseil d’État, 13 November 2018, No 52.949��694n.90 Min. v LVMH Moët Hennessy Louis Vuitton, (French Supreme Administrative Court), 13 April 2018, no. 398271��170n.57, 172n.71 Nouvelle-Calédonie v Sté Allianz-Vie, subss. 8e and 3e, 1 June 2005, no. 259617 ��169n.50 Nouvelle-Calédonie v SA Eagle Star Vie, no. 259618: Dr. fisc. 2006, no. 7, comm. 177��169n.50 Sté Zimmer, 31 March 2010, nos 304715 and 308525��169n.49 Valueclick International Ltd, 11 December 2020, no. 420174��169n.51 Germany

BVerfG, 17 January 1957, 1 BvL 4/54, BVerfGE 6, p. 55, ��278–79 BFH 27 February 1991, BStBl. II, pp. 444–448��449n.57 BFH 26 February 1992, BStBl. II, pp. 937–940��449n.57 BFH 30 August 1995, BStBl. 1996 II, p. 563��449n.57 BFH 23 October 1996, BStBl. 1997 II, p. 313 ��449n.57 BFH 18. May 2021, I R 4/17�� 548n.75 BFH 21 October 2009, I R 114/08 �� 618n.56 BFH 22 February 2001, V R 26/00 �� 776n.35 India

AAR-Bid Services Division (Mauritius), Ltd [2020] (2) TMI 1183 �� 842n.1 Aberdeen Institutional Commingled Funds LLC [2019] 104 taxmann.com 63 (Bombay)�� 848n.35

Table of Cases xxvii A. P. Moller TS-555-ITAT-2013 �� 845n.14 Ashapura Minichem Ltd 131 TTJ 291 ��848n.34 Asia Satellite Telecommunication Co. Ltd 332 ITR 340�� 847n.26, 847n.27 Azadi Bachao Andolan [2002] 125 Taxman 826 (SC)��845n.13 Azadi Bachao Andolan 263 ITR 706 (SC)��847n.20, 847n.24 Director Of Income Tax v M/S Nokia Networks Oy, (2012) 358 ITR 259 (Del)�� 856 DIT v Morgan Stanley and Co. Inc. [2007] 292 ITR 416 (SC)�� 741, 856n.44 Engineering Analysis Centre of Excellence (P.) Ltd, 125 taxmann.com 42 (SC)��861n.51 Ensco Maritime Ltd 91 ITD 459 ��847n.22 Formula One World Championship Ltd. v Commissioner Of Income Tax, International Taxation, Delhi & Anr, Civil Appeal No. 3849/2017�� 856 FRS Hotel Group Lux Sarl 404 ITR 676 AAR��847n.28 Hindalco Industries 94 ITD 242��847n.24 Hira Mills Ltd 14 ITR 417�� 843n.5 Kotak Securities 67 taxmann.com 356 (SC)�� 860n.48 Linklaters & Paines 65 SOT 266 �� 845n.14 Linklaters LLP 40 SOT 51�� 845n.14 Madyanchal Vidyut Vitran Nigam Ltd 88 taxmann.com 664 (Allahabad)�� 860n.49 Maersk Line UK Ltd 68 taxmann.com 173�� 845n.14 MasterCard Asia Pacific Pte Ltd. v Assistant Director of Income Tax, Authority for Advance Rulings (Income-tax) (AAR), New Delhi, 6 June 2018, No. 1573/2014�� 856 New Skies Satellite NV 382 ITR 114��847n.30 NHAI OMP 633/2012 dt. 8 July 2014��848n.34 Norasia Lines Malta Ltd 279 ITR 268�� 848n.33 P & O Nedlloyd Ltd 369 ITR 282�� 845n.14 Patni Computer System Ltd 84 CCH 116��847n.26 PILCOM 116 taxmann.com 394 (SC)�� 842n.1 Qualcomm International Inc. 93 taxmann 80�� 843n.5 Rajeev Sureshbhai Gajwani 129 ITD 145�� 848n.31 R. D. Agarwal and Co. 56 ITR 20 SC�� 843n.5 Reliance Communications Ltd 52 CCH 292 ��847n.23 Sanofi Pasteur Holding SA 354 ITR 316 ��847n.29 Set Satellite Singapore Ltd 106 ITD 175��848n.34 Siemens Gamesa Renewable Power Pvt. Ltd WP No. 19436, WMP No. 22874 of 2018, High Court of Madras�� 861n.52 Sky Cell 119 Taxman 496 (Madras)�� 860n.50 SNC Lavalin International Inc. 332 ITR 314��847n.25 Steria (India) Ltd 386 ITR 390��848n.32, 860 Taj TV Ltd 161 ITD 339��847n.30 Technip SA v SMS Holding (P.) Ltd [2005] 60 SCL 249 (SC) ��848n.36 Union Texas Petroleum Corp v Critchley (1988) STC 69�� 847n.21

xxviii Table of Cases Vegetable Products Ltd 88 ITR 192 (SC)��847n.23 Visakhapatnam Port Trust 144 ITR 146�� 847n.17 Volkswagen Finance 115 taxmann.com 386��842n.2 Italy

Italian Corte costituzionale, 15 July 1976, No. 179/1976��278–79 Kenya

Unilever Kenya Limited v The Commissioner of Income Tax [2005] eKLR Netherlands

Hoge Raad, 6 June 1994, Market Marker Case, BNB 1994/217�� 383n.64 Hoge Raad, 21 March 2008, No. 43050, NL:HR:2008:BA8179��368–69 Hoge Raad, 4 December 2009, 07/10383, NL:HR:2009:BF0938�� 340n.26 New Zealand

CIR v JFP Energy Inc. [1990] 3 NZLR 536 (CA) ��102n.99 Thiel v FCT (1990) 90 ATC 4717��101n.93 United Dominions Trust Ltd v CIR (1973) 1 NZTC 61,028��101n.93 Norway

Dell Products vs. Staten v/Skatt øst., HR-2011-2245-A��102n.99 Spain

Spanish Tribunal Constitucional, 20 February 1989, No. 45/1989, ES:TC:1989:45 ��278–79 Switzerland

FAC, A-7342/2008 and A-7426/2008 (5 March 2009) �� 357n.20 FAC, A-6053/2010, ASA 79 (2010/2011)��384n.69 FAC, A-6537/2010 (7 March 2012)��384n.70 FAC, A-1246/2011 (23 July 2012) �� 384n.71 FAC, A-5694/2017 (29 June 2018)��368n.87 FAC, A-3785/2018 (25 May 2020) ��368n.87 Federal Tax Appeals Commission of Switzerland, VBP 65.86 (28 Feb. 2001) ��384n.69 FSC, 2C_690/2015 (15 March 2016)��357n.17 FSC, 2C_954/2015 (13 February 2017)��365n.68 FSC, 2C_241/2016 (7 April 2017)�� 356n.14 FSC, 2C_1162/2016 (4 October 2017) ��356n.15 FSC, 2C_648/2017 (17 July 2018)��365n.71

Table of Cases xxix FSC, 2C_695/2017 (29 October 2018)�� 358n.26 FSC, 2C_819/2017 (2 August 2018)��365n.71 FSC, 2C_344/2018 (4 Feb. 2020)��384n.69 FSC, 2C_354/2018 (20 Apr. 2020)��384n.69 FSC, 2C_615/2018 (26 March 2019)��368n.87 FSC, 2C_616/2018 (9 July 2019)��368n.87 FSC, 2C_653/2018, ATF 146 II 150 (26 July 2016)�� 358n.27, 359n.28, 359n.29, 359n.30 FSC, 2C_764/2018 (7 June 2019)��356n.12 FSC, 2C_1053/2018 (22 July 2019)��356n.12 FSC, 2C_1037/2019 (27 August 2020)�� 370n.102 FSC, 2C_537/2019 (13 July 2020)��362n.48 FSC, ATF 141 II 436 (24 September 2015)��369n.100 FSC, ATF 142 II 161 (24 September 2015)��356n.9, 356n.10, 360n.34, 360n.35, 365n.67, 369n.97 FSC, ATF 142 II 69 (1 March 2016)��357n.16 FSC, ATF 142 II 218 (5 April 2016)�� 356n.11 FSC, ATF 143 II 136 (12 September 2016)�� 357n.22 FSC, ATF 143 II 185 (13 February 2017)��357n.18 FSC, ATF 143 II 202 (16 February 2017)��365n.68, 365n.69 FSC, ATF 143 II 224 (17 March 2017)��365n.70 FSC, ATF 143 II 628 (1 September 2017)�� 358n.24 FSC, ATF 144 II 130 (3 November 2017)��356n.13 FSC, ATF 144 II 206 (16 April 2018)��366n.72, 366n.76, 369n.100 FSC, ATF 144 II 29 (18 December 2017)��360n.37, 362n.49 FSC, ATF 145 II 112 (1 February 2019)��356n.12 United States

Aiken Industries, Inc. v Commissioner of Internal Revenue, United States Tax Court, 5 August 1971, 292–69, 56 TC 925�� 382n.62 Altera Corp. v Commissioner (Altera I), 145 TC (2015)�� 547n.67, 547n.68 Altera Corp. v Commissioner (Altera II-1), No. 16-70497 (2018)�� 547n.64, 547n.68, 548n.69 Altera Corp. v Commissioner (Altera II-2), Nos 6-70496 and 70497��548n.70 Attorney General of Canada v R. J. Reynolds Tobacco Holdings Inc., 268 F. 3d 103 (2d Cir. 2001), cert. denied 537 US 1000, 123 S. Ct. 513, 154 L. Ed 2d. 392 (2002) �� 166n.39 Coca-Cola Co. & Sub. v Commissioner, 149, TC No. 21 (2017)�� 548n.71 Coca-Cola v Commissioner, TC Docket No. 31183-15 �� 548n.72 Container Corp., 463 US�� 574n.14, 575n.17 Del Commercial Properties Inc. v Commissioner of Internal Revenue, United States Tax Court, 20 December 1999, 1887-98, TC Memo 441�� 382n.62, 383n.63

xxx Table of Cases Exxon Corp. v Commissioner, 66 TC Memo (CCH) 1707, TC Memo (RIA) 93,616 (1993)��167n.40, 168n.46 McCullock v Maryland, 4 Wheaton 316, 429 (1819) ��109n.11 Moorman Manufacturing Co. v Bair, 437 US 267 (1978)��574n.11, 581n.43 Opinion of Coca-Cola Co. & Sub. v Commissioner, 115 TC No. 10, Docket No. 31183-15 (Nov 2020)�� 548n.72 Order of Coca-Cola v Commissioner, Docket No. 31183-15 (TC 2021)�� 548n.73 Procter & Gamble Co. v Commissioner, 95 TC 323, 325 (1990), aff ’d, 961 F. 2d 1255 (6th Cir. 1992) ��168n.46 United States v Aluminum Co. of America, 148 F. 2d 416 (2d Cir. 1945)�� 164n.25

Table of Legislation

INTERNATIONAL INSTRUMENTS Conventions & Treaties

Act Concerning the Conditions of Accession of the Republic of Bulgaria and the Adjustments to the Treaties on which the European Union ��675 Art 19��675 Art 53��675 African Charter on Human and People’s Rights �� 190n.67 Agreement among the Governments of the Member States of the Caribbean Community for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, Profits or Gains and Capital Gains and for the Encouragement of Regional Trade and Investment (1994) �� 457 Agreement Creating the European Economic Area (EEA Agreement) �� 611, 611–12n.17, 645, 646, 647, 648, 650, 653, 654–55, 656–57, 660–61 Art 4��647–48, 652 Art 14�� 658–59 Art 14(2)�� 659 Art 15 ��658, 659 Art 28��647–48 Arts 31–34 ��652 Art 31 �� 611–12n.17, 650, 651, 652

Art 34��647–48 Art 37��647–48 Art 40�� 652–54 Art 49�� 648n.12 Art 59�� 648n.12 Arts 61–64�� C36P13 n.12 Art 61�� 650 Art 61(1)��648, 649n.16 Art 105�� 647 Art 107�� 648 Art 126�� 645 Annex XII��652 Annex XV�� 648n.12 Protocols 26–27 �� 648n.12 Agreement on Subsidies and Countervailing Duties �� 213 Art 3��216 Agreement on Trade-Related Investment Measures (1994) ��221 American Convention on Human Rights ��190n.67, 418, 423n.20 Andean Community Income and Capital Model Tax Treaty (1971)�� 456t, 458, 459, 460, 461, 462, 463, 464, 465, 466, 468–69, 470–7 1 ASEAN Model Double Taxation Convention signed by the members of the Association of South East Asian Nations 1987 (ASEAN) ��45–46, 456t, 458, 459, 460, 461, 462, 463, 464, 465, 466, 467, 468–69 Art 1��466 Art 2��466

xxxii Table of Legislation Art 20��464 ATAF Income Model Convention 2016 (revised in 2019)�� 456t, 458, 459–60, 461–62, 463, 464, 465, 467, 468, 470 Austria–Germany tax treaty (1954) ��419 Austria–Netherlands Estate, Inheritance and Gift Tax Treaty (2001) Art 10(2) ��350 Berne Convention for the Protection of Literary and Artistic Works (9 September 1886) Art 31(1)(c)�� 262n.52 Canada–Luxembourg DTC Art 1�� 402–3, 410, 411, 412–13 Art 4��402–3, 410, 411, 412 Art 4(1)�� 412–13 Art 10(2)(a)�� 407–8, 409, 414–15 Art 13 �� 402–3 Art 13(4)�� 407–8, 409, 410, 411, 414–15 Art 13(5)��402–3, 411, 413 Charter of the United Nations Art 2(1)��189 Art 2(3)��186 Art 13(1)(a)�� 91n.30 Art 33��187 Art 33(1)��186 Art 681��481n.44 COMESA Model Convention ��456t, 458, 459–60, 462, 463, 469–70, 471 Convention between the Kingdom of Sweden and the Federal Republic of for the Avoidance of Double Taxation with respect to the Taxes on Income and Capital as well as on Inheritances and Gifts and Concerning Mutual

Administrative Assistance in Tax Matters (1992) Art 41(5)��188n.57 Convention between the Nordic Countries for the Avoidance of Double Taxation with respect to Taxes on Income and on Capital (1996)�� 457, 485 Convention between the Republic of Austria and the Federal Republic of Germany for the Avoidance of Double Taxation with respect to Taxes on Income and Capital Art 25(5)��174n.75, 187n.53 Convention for the Unification of Certain Rules for International Carriage by Air (Montreal Convention), 28 May 1999 ��171n.62 Convention on Direct Taxes��23 Convention on Mutual Administrative Assistance in Tax Matters (CMAA)��52–53, 354, 486–87 Art 6��486–87 Convention on the Elimination of Double Taxation in Connection with the Adjustment of Profits of Associated Enterprises— Final Act—Joint Declarations— Unilateral Declarations (90/436/EEC) [1990] OJ L225, 10–24��52, 523n.24, 552–53, 567, 567n.70, 568, 606, 606n.80, 705–6, 706n.53, 707–10, 713, 718, 726–27 Art 4�� 568, 706

Table of Legislation xxxiii Art 4(2) ��708–9 Art 8(1)�� 709n.71 Art 18��606 Art 12(1)��708n.65 Convention sur les bâtiments hospitaliers, The Hague, 21 December 1904 �� 19n.110 Denmark–Switzerland tax treaty 1973��420 Draft Convention on the Allocation of Business Income (1933)��518, 519 Art 5�� 518, 518n.5, 519 Double Taxation Convention��522, 527, 528, 530, 573 Art 9(16) �� 526n.36 Dutch–Canadian Treaty Art 3(2)��382–83 EAC Model Taxation Convention (2012)�� 456t, 458, 459–60, 461–62, 463, 464–65, 466, 467, 468, 469–70 Art 8��464 Energy Charter Treaty (ECT) ��701 European Convention on the Peaceful Settlement of Disputes (29 April 1957) Ch I��188n.57 Ch II��188n.57 Ch IV��188n.57 France and Belgium Concerned the Exchange of Information on Deeds Presented at Registration (1843)�� 17–18 France–Austria Inheritance and Gift Tax Treaty 1993 Art 10(2) ��350 France–Italy Inheritance and Gift Tax Treaty 1990

Art 17(2)��350 France–Luxembourg DTC �� 395–96 France–Portugal Inheritance and Gift Tax Treaty 1994��350 France–Sweden Inheritance and Gift Tax Treaty 1994 Art 11(2)��350 Art 28(2) ��359 French–Swiss DTT Protocol, art. IX para. 2�� 359n.31 General Agreement on Tariffs and Trade (GATT)��54, 213, 214, 747–48, 763 Art I��214, 216, 748 Art II ��214 Art II:1(b)��214n.5 Art III�� 217–18, 219, 219–20n.31, 220–21, 747–48 Art III(1) ��216 Art III(2)��216, 217, 747–48, 749, 757 Art III(4)��221–22, 747–48 Art V ��748 Art VII��748 Art VIII��214 Art X(1)��748 Art XVI��216–17, 220–21, 222–23 Art XVII ��221–22 Art XX�� 218–19, 219–20n.31 Art XX(b) ��219 Art XX(d) �� 226–27 Art XX(g) ��219 Art XXVIII��214 Art XXIV(4)–(10)��748 General Agreement on Trade in Services (GATS) �� 213 Art I��225 Art I:3 (b)��225n.60 Art II ��225, 226 Art XIV��226n.63 Art XIV(d) �� 226–27 Art XIV(e)�� 226 Art XVI��225

xxxiv Table of Legislation Art XVII ��225, 227, 228 Art XIX ��225 Art XXII(3)��227n.70 Art XXIII��227n.70 Art XXVIII��225 Art XXVIII(b)�� 225, 225n.60 Art XXVIII(o)��225n.61 Geneva Convention Protocol I�� 82n.61 Germany–Austria DTA (2000) Art 11 ��607 Art 25(5)��607 German Imperial Double Taxation Law (1870) �� 180n.14 German–Australian Double Taxation Treaty (1972) Protocol��376–77n.27 German–Italian Double Taxation Treaty (1989)��376–77n.27 German Model Tax Convention�� 251n.18 German–Norwegian Double Taxation Treaty (1992) Protocol��376–77n.27 German–Swedish Double Taxation Treaty Art 43(3)��376–77n.27 German–USA Double Taxation Treaty (1992) Protocol��376–77n.27 GloBE Model Rules ��971–72, 982, 986, 988–89 Art 2.4.1�� 989n.89 Art 2.6 ��987n.82 Art 3.1.2 �� 983n.59 Art 3.2.4�� 984n.66 Art 4.1.2(d) �� 984n.66 Art 4.4 �� 981n.55, 984n.65 Art 5.1��986n.74 Art 5.2.2�� 985n.70, 986n.74 Art 10.1��983n.59, 983n.60 Havana Charter for an International Trade Organisation �� 214n.4

ILADT Multilateral Model Convention for Latin America for the Avoidance of Double Taxation (2012) ��456t, 457, 467, 468–69 s. I, para. 1 �� 186n.47 ILC, Second Report on Identification of Customary International Law, 66th session, A/CN.4/672 (2014) �� 77n.25 Income, Capital, Inheritance and Gift Tax Treaty between Germany and Sweden (1992) Art 28(2) ��350 India–China DTAA Art 5(3)�� 859 Inheritance Tax Treaty (1959) Art 14(2)��18n.102 International Covenant on Civil and Political Rights (1976)��418, 418n.3, 418n.4, 419n.6, 421, 421n.15, 423, 424, 427, 429n.35 Art 6��421 Art 8��421 Art 13 ��421 Art 14��421 Art 17�� 424 Art 26�� 428 Ireland–Netherlands DTC�� 340–41 Israel–Belgium DTC ��396–97 Israel–Sweden Death Duties Treaty (1962) Article IX�� 351 Italy–Greece Inheritance Tax Treaty (1964) Art 10��350 Italy–Israel Estate and Inheritance Tax Treaty (1968)�� 351 Art VI�� 351

Table of Legislation xxxv League of Nations, Double Taxation and Tax Evasion. General Meeting of Government Experts on Double Taxation and Tax Evasion (Oct. 1928)��322 League of Nations Report on Double Taxation (1923)��237–38 London Model Tax Convention (1946) ��25, 518n.6, 703–4 Art IV�� 519n.8 Art VI��214, 1039n.88 Luxembourg Convention �� 411 Manila Declaration on the Peaceful Settlement of International Disputes (1982)�� 186–87 s. I, para. 1 �� 186n.47 para. 5 �� 186n.47 para. 11�� 186n.47 Mexico Model Tax Convention (1943)��25, 519n.8, 703–4 Art VII�� 1039n.88 Multilateral Competent Authority Agreement (MCAA)��354 Multilateral Convention on Mutual Administrative Assistance and Exchange of Information (MAC)�� 128–29, 655 Multilateral Convention on Mutual Administrative Assistance in Tax Matters (MCMAA)�� 109–10 Multilateral Convention to Implement Tax Treaty Related Measures to Prevent BEPS (MLI), signed on 7 June 2017 (entered into force 1 July 2018) �� 41, 62, 71, 99, 180n.15, 236, 489–90, 510, 510n.58, 512n.65, 622 Pt VI��422–23, 704–5, 846 Art 1�� 489–90

Art 2(2) �� 62 Art 3�� 845, 846 Art 3(1)�� 445 Arts 4–15 �� 845 Art 4(1)�� 98n.73 Art 5�� 845, 846 Art 6�� 846 Art 7�� 846 Art 7(6)–(13)�� 626n.107 Art 7(1)�� 19, 626n.109 Art 8�� 846 Art 8(1)��374 Art 9�� 846 Art 9(4) ��138–39 Art 11 ��522n.21 Art 12��114–15, 511–12, 846, 858 Art 13 ��845, 846, 858–59 Art 14�� 846, 858 Art 15 �� 114–15, 511–12, 846 Art 16��705n.49, 845 Art 17��705n.49, 845 Arts 18–26�� 846 Art 30��71 Nordic Convention between Denmark, Finland, Iceland, Norway, and Sweden (1983)�� 48n.12 OECD Model Convention for the Avoidance of Double Taxation with Respect to Taxes on Estates and Inheritances and on Gifts (1982)��336, 348 Art 10��349, 350 Art 11 ��348–49 OECD Model Tax Convention on Income and on Capital 2017 (1963, 1977, 2010, 2014) �� 45–46, 55–56, 87n.2, 98–99, 98n.74, 101n.94, 126, 132–34, 185, 286, 301, 328, 334, 336, 455, 459, 462, 464, 480–81, 501, 502–3, 504–5, 510, 512, 514, 539,

xxxvi Table of Legislation 552–53, 559, 563–64, 567, 568, 592–93, 615, 617, 622, 626–27, 703–4, 703n.37, 777, 843, 863–64, 865n.6, 871–73, 876–77, 918, 921, 988, 992–93, 1010–11, 1014–15 Preamble�� 148, 394–95, 404n.73, 413–14, 414n.119, 416 Art 1��289, 303, 312, 360, 437–38, 458, 865n.6, 872 Art 1(1)��339, 437–38 Art 1(2)�� 172n.72, 234–35, 445–47, 870 Art 1(3)�� 446–47, 522, 523, 526, 527, 988 Art 2�� 360, 458, 744n.104, 788n.18 Art 2(2) �� 458 Art 2(4) �� 458 Art 3�� 462, 521 Art 3(1)�� 251, 256, 292, 341, 459, 521, 559 Art 3(1)(a) ��210, 438, 521n.16 Art 3(1)(b)��202, 210, 438, 521n.17 Art 3(2)�� 56–57, 58, 62, 172n.66, 207, 252, 253–54, 256–57, 287–89, 290, 291–93, 296, 376–77, 384, 447–49, 459, 521, 522, 525n.31 Art 3(12)�� 522, 527 Art 4��336n.4, 339–40, 360, 377, 392–93, 394, 398, 402–3, 404–5, 410, 413, 414n.121, 416, 441–42, 459–60 Art 4(1)��289, 301, 310–11, 316, 340–41, 348–49, 414n.121, 439–40 Art 4(2) ��309, 316 Art 4(2)(c)��113 Art 4(2)(d)��113, 424 Art 4(3) ��309, 310–11, 315, 316, 317, 348–49, 870 Art 5�� 29–30, 251–52, 256, 292, 460, 461, 502–3, 507, 509, 521n.18, 846, 858–59, 870 Art 5(3)��405, 460

Art 5(5)��502–3, 507, 509, 510 Art 5(6) �� 501, 510 Art 5(8) ��521n.18, 525, 684 Art 6–12�� 303, 311n.15 Arts 6–22��404–5 Art 6�� 172–73, 304–5, 307, 308, 460 Art 6(2) �� 172–73, 251–52, 257, 261, 291, 307, 307n.10, 308 Art 7�� 304–5, 306, 308, 310, 314, 316–17, 447–49, 460, 461, 519, 520n.14, 559, 693–94, 708–9, 872, 988 Art 7(1)�� 556, 565–66 Art 7(2) �� 520n.14, 526–27 Art 7(4) �� 306, 308, 447–49 Art 8��304–5, 460, 461, 464–65 Art 9��234–35, 461–62, 517, 518, 519, 519n.9, 520, 520n.14, 521, 521n.17, 521n.18, 522, 524–25, 525n.33, 526, 526n.36, 527, 528, 530, 531, 532, 533, 555–56, 557, 559, 561, 567n.62, 568–69, 988 Art 9(1)��517, 519–20, 520n.14, 521, 523, 524–25, 525n.31, 526–27, 528–30, 531–33, 539, 555–56, 557, 558–59, 561, 563–64, 567, 568, 569 Art 9(2) �� 319, 328–30, 461, 522, 523, 525n.33, 526, 532, 555–56, 557, 558–59, 560–61, 562, 563–65, 566–67, 568–69 Art 9(3) �� 519–20 Art 9(4) ��522 Art 9(6)��523 Arts 10–12 �� 394–95 Art 10�� 202, 207–8, 303, 304, 313, 314, 316, 317, 322, 326–27, 336n.5, 407–8, 462 Art 10(2) ��303, 311, 322, 374, 404n.74 Art 10(2)(a)�� 405–6, 407–8 Art 10(3)�� 252, 287–88, 291 Art 10(4) �� 304 Art 10(5)��312, 313, 314, 315, 316

Table of Legislation xxxvii Art 11 �� 304, 313, 314, 316–17, 322, 326–27, 336n.6, 341, 462–63 Art 11(1)�� 447–49 Art 11(2)�� 313, 374, 447–49 Art 11(5)�� 313, 314, 447–49 Art 12�� 304, 313, 314, 316–17, 336n.7, 341, 463, 872–73, 872n.43 Art 12(1)��374 Art 12A�� 129, 873–74 Art 13 �� 207, 304–5, 306, 402–3, 464 Art 13(1)�� 306, 307 Art 13(2)��306–8 Art 13(5)�� 306–7, 392–93, 394–95, 399, 405n.75, 414–15, 894 Art 14�� 464, 873–74, 873n.50 Art 15 �� 102n.99, 464–65, 1038–40 Art 16�� 465 Art 17�� 465 Art 18�� 465 Art 19�� 189–90, 465 Art 20�� 465 Art 21��251, 304–5, 306, 308, 466 Art 21(1)��307, 308, 414–15 Art 21(2)�� 308 Art 22��466 Art 23��233n.11, 305, 306, 308, 313, 314, 442–43, 446, 447–49, 466 Art 23A�� 47, 58, 234n.14, 286–87, 290, 304, 305n.5, 319, 320, 322, 326, 327, 329, 330, 404–5, 558, 560, 565–66 Art 23A(1) ��323, 327 Art 23A(2)�� 326–27 Art 23A(3)��322–23 Art 23A(4)��323, 325–26, 327, 442–43 Art 23B�� 47, 58, 234n.14, 286–87, 304, 319, 321, 322–23, 329, 330, 404–5, 558, 560, 565–66 Art 23B(1) ��290, 323 Art 24�� 228–29, 305, 342–43, 428, 466, 988

Art 24(3) �� 304–6, 308, 342–43, 744 Art 25�� 57, 61–62, 191–92, 258, 422n.16, 447–49, 466–67, 521, 523, 558, 559, 563, 565, 566, 567, 777–78 Art 25(1)�� 187n.50, 551, 555–56, 566, 606, 703n.36 Art 25(2)�� 555–56, 606, 703n.36, 703n.38 Art 25(3)��187n.50, 566 Art 25(5)��187, 422–23, 524, 526, 552, 562, 568, 606 C702 Art 25(5)(b)��552 Art 26�� 191–92, 353, 360, 363, 366, 421, 424, 425n.23, 426–27, 467, 468, 1014–15 Art 26(1)��355–56, 362, 363, 364, 428 Art 26 (2)�� 361, 362, 364, 428 Art 26 (3)�� 364, 367–68, 369, 428 Art 26(3)(a)��366–67 Art 26(3)(b)��366–67 Art 26(3)(c)��367–69 Art 26(4)�� 428 Art 26(5) ��369–70, 428 Art 27�� 431, 467, 468 Art 28��189, 467 Art 29��342, 388, 467 Art 29(1)–(7)��342n.36, 404–5, 626n.107 Art 29(2) �� 342 Art 29(2)(e)(i)��344n.39 Art 29(8) �� 304, 467 Art 29(8)(b)�� 304n.3 Art 29(8)(c)��304n.4 Art 29(9) ��60n.67, 342n.36, 391, 394, 399, 401–3, 404–5, 406, 407–8, 413, 415, 416n.130, 467, 626n.109 Art 30�� 467 Art 31 �� 468 Art 32�� 468

xxxviii Table of Legislation OECD Model Tax Convention on Income and on Capital: Condensed Version 2017�� 89n.14, 202, 510n.58, 517, 517n.1, 767n.5 Art 10(12.1) �� 375n.16, 375n.19, 376n.26 Art 10(12.4)�� 376n.23 Art 11(9)�� 375n.16, 375n.19, 376n.26 Art 12(4)�� 375n.16, 375n.19, 376n.26 Art 12(4.3)��375n.19, 376n.23, 376n.26 Art 10(12.4)�� 376n.23 Art 11(9)�� 375n.16, 375n.19, 376n.26 Art 11(102)�� 376n.23 Art 12(4)�� 375n.16, 375n.19, 376n.26 Art 12(4.3)�� 376n.23 OECD Model Tax Convention on Income and on Capital: Condensed Version 2008�� 704 Art 25.5��704n.41 Protocol to the Switzerland– Netherlands DTC Art IV�� 340–41 SADC Model Tax Agreement on Income 2011 (revised 2013) ��456t, 458, 459–60, 461–62, 463, 465, 466, 467, 468, 469–70, 471 Art 10(6) ��462n.7 Statute of the International Court of Justice��49–50, 186n.45 Art 38�� 49–50, 73, 74–75, 85, 90, 179 Art 38(1)�� 68–69, 180, 183, 393, 413–14 Art 38(1)(b)�� 69n.17, 74–75, 92n.32 Statute of the International Criminal Court �� 166–67 Treaty establishing the European Atomic Energy Community Art 225��258 Treaty of Accession of the Republic of Bulgaria to the European Union ��675 Art 2��675

Treaty of Friendship, Commerce and Navigation between the United States of America and the Federal Republic of Germany (1955)�� 54n.41 Treaty of Rome (1957)��705–6 UNIDROIT Convention on international financial leasing, Ottawa (28 May 1988)��171n.62 United Nations, Department of Economic & Social Affairs, Model Double Taxation Convention between Developed and Developing Countries, 2021 update (New York: UN, 2021)�� 45–46n.2 United Nations, Model Double Taxation Convention between Developed and Developing Countries (2017)��25, 55–56, 97–98, 98n.75, 113, 114–15, 126, 132–33, 288–89, 290, 291, 292, 336, 404, 455, 459, 460, 461, 463, 464, 465, 481, 502–3, 512, 992–93, 1014–15 Art 1(3)�� 532n.62 Art 5(3)(b)��871–72 Art 5�� 502–3 Art 5(5)�� 502–3, 512 Art 5(5)(b)�� 502–3 Art 5(7)�� 512 Art 6(2) �� 287–88 Art 7�� 132–33, 134, 875–76 Art 7(1)(c)�� 463 Art 9�� 532n.62 Art 9(1)�� 532n.62 Art 9(2) �� 532n.62 Art 9(3) �� 532n.62 Art 12�� 829–30, 872, 875–76

Table of Legislation xxxix Art 12A��98n.69, 134–35 , 463, 870–7 1, 874, 875–76, 900–1, 908, 909–10, 915 Art 12B�� 38, 98n.69, 829–30, 900–1, 908, 909–10, 997–98, 1004 Art 13(4)��138–39 Art 13(5)��138–39 Art 13(6)��139, 394–95, 414–15 Art 15 �� 1038–40 Art 20�� 872 Art 24�� 344 United Nations, Report of the International Law Commission, 68th session, A/7 1/10 (2016) �� 75n.13, 76n.19, 76n.21, 77n.27, 77n.31, 80n.46, 82n.60, 83n.64, 94n.45 United Nations, Text of the Draft Conclusions as Adopted by the Drafting Committee on Second Reading, 70th session, A/CN. 4/L.908 (2018) �� 75n.13 Uruguay Round 1987–1994��214 US–Canada Income and Capital DTC (1980) Art 21(1)�� 346 Art 21(5)�� 346 Art 21(6)�� 347 Art 21(7)�� 347 US–Denmark Inheritance and Gift Tax Treaty (1983) Art 9�� 351–52 US Estate and Gift Model Convention (1980)��336 Art 8(3) �� 351 Art 10(2)(c)�� 351–52 US–France Inheritance and Gift Tax Treaty (1978) Art 10�� 351–52 US–Germany DTC (1989) Art 27�� 346

US–Germany Inheritance and Gift Tax Treaty (1980) Art 10�� 351–52 US–German Treaty (1989/2008) Art 10(2) ��385 US–Israel DTC 1975 Art 15A�� 347 US–Mexico DTC 1992 Art 22(1)�� 346 Art 22(2) �� 348 Art 22(3)�� 348 Art 22(4) ��C20P43 US Model Income Tax Convention (1981) Art 9(3) ��522n.19 US Model Income Tax Convention (1996) Art 1(6)�� 444, 445–46 Art 9(1)�� 522n.20 US Model Income Tax Convention (2016) Art 3(1)(l)(iv)�� 345–46 Art 4(2) ��345, 346 Art 22(2)(e)�� 346 US–Netherlands DTC (1992) Art 4(1)�� 340 Art 35�� 340 Art 36�� 340, 346 US–Sweden Inheritance and Gift Tax Treaty (1983) Art 8�� 351–52 Universal Declaration on Human Rights (1948) ��190–91 Art 10��698n.3 Vienna Convention on Consular Relations (1963)��189 Vienna Convention on Diplomatic Relations (1961)�� 74–75, 189 Vienna Convention on the Law of Treaties 23 May 1969, 1155 UNTS 331 (VCLT)��50, 184, 290, 317, 412, 527

xl Table of Legislation Art 2��180 Art 4��184n.35 Art 24�� 364–65 Art 26��70n.24, 398n.38, 412–13, 412n.108 Art 27��527 Art 30�� 378n.30 Art 31 �� 184, 364–65, 392n.3, 393, 394–95, 396–98, 404–5, 414n.120, 521n.15 Arts 31–32 �� 59 Arts 31–33��55, 184, 291, 296 Arts 31ff�� 286, 394 Art 31(1)�� 55–56, 60, 101n.90, 185, 254, 293, 393n.7, 394–95 Art 31(2)�� 56, 59n.65, 394–95 Art 31(2)(a) ��59–60 Art 31(2)(b)�� 46n.3, 59 Art 31(3)�� 56 Art 31(3)(a) ��57 Art 31(3)(c) �� 393, 413–14, 414n.122, 416 Art 31(4)�� 55–56, 57, 60 Art 32�� 56, 184, 521n.15 Art 33��61 Art 33(1)��254 Art 33(4)�� 61, 254 Art 34��378 Art 46�� 68n.12 Art 53�� 69n.15 Art 64�� 69n.15 World Trade Organization (WTO) rules�� 54 WTO, 2016, Doha Ministerial Declaration, WT/MIN(01)/ DEC/1 of 20 November 2001, para. 16 �� 214n.6 OECD, ‘2010 Report on the Attribution of Profits to Permanent Establishments’ (CTPA 2010)�� 511n.61

OECD, ‘Action Plan on Base Erosion and Profit Shifting’ (19 July 2013)��506n.46, 679n.1, 953n.16 OECD, ‘Additional Guidance on the Attribution of Profits to a Permanent Establishment’ (2018)�� 511n.63 OECD, Addressing the Tax Challenges of the tw Digital Economy, Action 1—2015 Final Report, OECD/G20 Base Erosion and Profit Shifting Project (Paris: OECD Publishing, 2015) ��30n.50, 33n.68, 777n.41, 864n.5, 867n.22, 910n.93, 952n.22, 993n.8, 1041n.3, 1042n.6 OECD, Addressing the Tax Challenges of the Digitalization of the Economy— Public Consultation Document (2019) �� 864n.5, 955n.28, 973n.15 OECD, Aligning Transfer Pricing Outcomes with Value Creation, Actions 8–10— 2015 Final Reports: OECD/ G20, Base Erosion and Profit Shifting Project (2015)��927n.30 OECD, ‘BEPS Action 7 Additional Guidance on the Attribution of Profits to Permanent Establishments’ (2016) ��511n.60, 511n.62 OECD, ‘BEPS Action 14, 2015 Final Report: Making Dispute Resolution

Table of Legislation xli Mechanisms More Effective’ (2015) �� 258, 564n.49 OECD, Cover Statement by the Inclusive Framework on the Reports on the Blueprints of Pillar One and Pillar Two as Approved by the Inclusive Framework at Its Meeting on 8–9 October (Paris: OECD Publishing, 2020)�� 34n.74 OECD, ‘Designing Effective Controlled Foreign Company Rules, Action 3—2015 Final Report: OECD/G20, Base Erosion and Profit Shifting Project’ (2015)��927n.31, 928n.35, 975n.22 OECD, Electronic Commerce: Taxation Framework Conditions (Paris: OECD Publishing, 1998)��744n.106, 953n.14 OECD, Explanatory Statement, Base Erosion and Profit Shifting Project, 2015 Final Reports (2015) �� 969n.1, 975n.22 OECD, ‘Global Anti-Base Erosion Proposal (‘GloBE’)—Pillar Two, Public Consultation Document’ (2019)�� 970n.6 OECD, Going Digital: Shaping Policies, Improving Lives (Paris: OECD Publishing, 2019)�� 978n.41 OECD, ‘Guidance on the transfer pricing implications of the Covid-19 pandemic’ (2020)�� 536n.5 OECD, Harmful Tax Competition: An Emerging Global Issue (Paris: OECD Publishing, 1998) �� 29n.45, 32n.61

OECD, Improving the Efficiency of Dispute Resolution Mechanisms, Action Item 14—Final Report 2015�� 566n.57 OECD, ‘Inclusive Framework on BPES. Progress Report July 2016–June 2017’ (2017)�� 1016n.42 OECD, ‘International Compliance Assurance Programme, Pilot Handbook 2.0’ (2019) �� 989n.91 OECD, ‘International VAT/GST Guidelines’ (2017) �� 732n.8, 768n.10, 777n.37, 789, 790, 792–94, 795–96, 962n.77 Ch 1�� 789 Ch 2 �� 789–90 Ch 3 �� 789 OECD, ‘Jurisdictions participating in the convention on mutual administrative assistance in tax matters’ (2020)��1014n.33 OECD, Limiting Base Erosion Involving Interest Deductions and Other Financial Payments, Action 4–2015 Final Report: OECD/G20 Base Erosion and Profit Shifting Project (2015)��200, 929n.37 OECD, Neutralising the Effects of Hybrids Mismatch Arrangements, Action 2— 2015 Final Report (Paris: OECD Publishing, 2015) ��201n.20 OECD, OECD Employment Outlook 2017 (Paris: OECD Publishing, 2017)�� 1029n.22 OECD, OECD Employment Outlook 2019: The Future

xlii Table of Legislation of Work (Paris: OECD Publishing, 2019)��1026n.7, 1026n.9, 1027n.10, 1027n.15, 1031n.36, 1032n.40, 1032n.42 OECD, ‘OECD/G20 Inclusive Framework on BEPS: Progress Report July 2019– July 2020’ (2020) ��1053–54n.35 OECD, Policy Responses to New Forms of Work Paris: (OECD Publishing, 2019)��1039n.83 OECD, Preventing the Granting of Treaty Benefits in Inappropriate Circumstances, Action 6–2015 Final Report (Paris: OECD Publishing, 2015) ��342n.35, 408n.89, 684n.26 OECD, Programme of Work to Develop a Consensus Solution to the Tax Challenges Arising from the Digitalisation of the Economy (Paris: OECD Publishing, 2019)�� 34n.72, 34n.75, 954n.27, 961n.70, 970n.5 OECD, Progress Report on Amount A of Pillar One, Two Pillar Solution to the Tax Challenges of the Digitalization of the Economy, OECD/G20 Base Erosion and Profit Shifting Project (Paris: OECD Publishing, 2022)��46n.8, 171n.63, 513–14n.74 OECD, ‘Proposed Changes to Commentaries in the OECD Model Tax Convention on Article 9

and Related Article’ (2021)�� 519n.9 OECD, ‘Public Consultation Document, Addressing the Tax Challenges of the Digitalisation of the Economy’ (13 February–6 March 2019)�� 34n.71 OECD, ‘Public Consultation Document on Pillar One—Amount A: Draft Model Rules for Tax Base Determinations’ (published 18 February 2022)�� 583n.51, 974n.20 OECD, Receita Federal do Brasil, Transfer Pricing In Brazil: Towards Convergence with the OECD Standard (Paris: OECD, 2019) �� 866n.15 OECD, Report to G20 Development Working Group on the Impact of BEPS in Low Income Countries (Part 1) (Paris: OECD Publishing, 2014)�� 128n.4 OECD, Secretary-General Tax Report to G20 Finance Ministers and Central Bank Governors. Italy October (2021)��956n.38, 956n.45, 957n.49, 960n.64, 960n.65 OECD, Statement by the OECD/ G20 Inclusive Framework on BEPS on the Two-Pillar Approach to Address the Tax Challenges Arising from the Digitalisation of the Economy—January 2020, OECD/G20 Inclusive Framework on BEPS

Table of Legislation xliii (Paris: OECD Publishing, 2020)��34n.73, 35n.79 OECD, ‘Tax Challenges Arising from Digitalisation— Economic Impact Assessment: Inclusive Framework on BEPS’, OECD/G20 Base Erosion and Profit Shifting Project (2020)��962n.80, 980n.50 OECD, Tax Challenges Arising from Digitalisation— Interim Report 2018: Inclusive Framework on BEPS (Paris: OECD Publishing, 2018)�� 33n.70, 955n.28, 994n.12 OECD, Tax Challenges Arising from the Digitalisation of the Economy—Global Anti-Base Erosion Model Rules (Pillar Two) (Paris: OECD Publishing, 2021) ��46n.9, 171n.64, 971n.9 OECD, Taxation and Electronic Commerce: Implementing the Ottawa Taxation Framework Conditions (Paris: OECD Publishing, 2001)�� 744n.106, 789n.21 OECD, Taxation and Philanthropy, OECD Tax Policy Studies no. 27 (Paris: OECD Publishing, 2020)��335, 337, 338 OECD, Taxing Wages 2020 (Paris: OECD Publishing, 2020)��1025n.2 OECD, Taxing Wages 2021 (Paris: OECD Publishing, 2021) ��1025n.2 OECD, The Application of the OECD Model Tax Convention to Partnerships

(Paris: OECD Publishing, 1998) �� 322n.6 OECD, The Application of the OECD Model Tax Convention to Partnerships, Issues in International Taxation, vol. 6 (Paris: OCDE Publishing, 1999)��172n.72 OECD, The Next Production Revolution: Implications for Governments and Business (Paris: OECD Publishing, 2017)�� 1032n.40, 1032n.41 OECD, Transfer Pricing and Multinational Enterprises: Three Taxation Iissues (Paris: OECD Publishing, 1984) ��703–4n.39, 704n.40 OECD, ‘Transfer Pricing, Corresponding Adjustment and the Mutual Agreement Procedure’ (1982)��567n.68 OECD, Transfer Pricing Guidance on Financial Transactions: Inclusive Framework on BEPS (2020)��536n.5, 543n.45 OECD, Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations (Paris: OECD Publishing, 2017) �� 46n.4, 637n.35, 796n.57, 976n.26 OECD, Transfer Pricing Guidelines for Multinational Enterprises and Tax Administration (2022) �� 205n.40, 535n.2, 578n.35

xliv Table of Legislation OECD, Transfer Pricing Documentation and Country‑by‑Country Reporting, Action 13— 2015 Final Report (2015)�� 964n.90 OECD, ‘Updated guidance on tax treaties and the impact of the Covid-19 pandemic’ (21 Jan. 2021)��1039n.85 OECD/Council of Europe, The Multilateral Convention on Mutual Administrative Assistance in Tax Matters: Amended by the 2010 Protocol (Paris: OECD Publishing, 2011)�� 52n.30 OECD/G20 Base Erosion and Profit Shifting Project, ‘Addressing the Tax Challenges of the Digitalisation of the Economy—Policy Note’ (2019) ��970n.4 OECD/G20 Base Erosion and Profit Shifting Project, ‘Country-by-Country Reporting—Compilation of Peer Review Reports (Phase 1): Inclusive Framework on BEPS: Action 13’ (2018)�� 1019n.60, 1019n.61 OECD/G20 Base Erosion and Profit Shifting Project, Preventing the Artificial Avoidance of Permanent Establishment Status, Action 7—2015 Final Report ( Paris: OECD Publishing, 2015)�� 509n.55, 684n.26, 953n.8, 1053–54n.35

OECD/G20 Base Erosion and Profit Shifting Project, Tax Challenges Arising from Digitalisation–Interim Report 2018: Inclusive Framework on BEPS (Paris: OECD Publishing, 2018) ��1039n.87, 1053–54n.35, 1054n.36 OECD/G20 Base Erosion and Profit Shifting Project, Tax Challenges Arising from Digitalisation— Report on Pillar One Blueprint: Inclusive Framework on BEPS, OECD/G20 Base Erosion and Profit Shifting Project (Paris: OECD Publishing, 2020) �� 173n.73, 280n.88, 331–34, 513n.69, 572n.4, 1054n.37 OECD/G20 Base Erosion and Profit Shifting Project, ‘Tax Challenges Arising from Digitalisation —Report on Pillar Two Blueprint: Inclusive Framework on BEPS’ (2020)�� 34n.76, 46n.6, 132n.8, 513n.72, 825n.5, 832n.11, 864n.5, 913n.121, 952n.3, 955n.30, 955n.34, 957n.51, 960n.64, 961n.67, 962n.80, 963n.82, 964, 966n.97, 970n.7, 975–76, 984n.64, 986, 986n.80, 987, 988, 988n.85, 989n.90, 996n.15 OECD/G20 Inclusive Framework on BEPS, ‘Addressing the Tax Challenges of the Digitalisation of the Economy—Policy Note’ (2019) ��995n.13

Table of Legislation xlv OECD/G20 Inclusive Framework on BEPS, ‘Statement by the OECD/G20 Inclusive Framework on BEPS on the Two-Pillar Approach to Address the Tax Challenges Arising from the Digitalisation of the Economy’ (Jan. 2020)�� 1054n.39 OECD/G20 Inclusive Framework on BEPS, ‘Statement of 136 member jurisdictions on a Two-Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy’ (8 Oct. 2021)�� 971n.8 OECD/G20, Statement on a Two- Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy (8 Oct. 2021) �� 36n.80, 46n.7, 50n.18, 153, 513–14n.74, 910n.94, 976n.31 EUROPEAN UNION

Charter of Fundamental Rights of the European Union [2016] OJ C202/389�� 418, 427, 673–74, 673n.32 Art 7��673n.33 Art 8��673n.33 Art 11 ��673n.33 Art 16��673n.33 Art 17��673n.33 Art 21(1)�� 262 Art 22�� 262 Art 41�� 673–74, 673n.33, 724n.155 Art 41(4) ��260n.43 Art 42�� 673n.33, 724n.155 Art 45��673n.33 Art 47��52, 673n.33, 722, 725

Art 48��673n.33 Art 49��673n.33 Art 50��673n.33 Art 51��724n.158 Art 51(1)�� 673–74 EU Treaty Consolidated version [2016] OJ C202/1, 15–45��50–51, 665–66 European Convention for the Protection of Human Rights and Fundamental Freedoms (1950) (ECHR)�� 54, 190–91, 418–19, 418n.4, 419n.6, 421, 424, 425, 426–27, 427n.29, 428 Art 5��420 Art 6��191–92, 420, 422, 423–24, 425n.23, 427, 427n.29 Art 8��191–92, 420, 424, 425–27 Art 8(2) �� 420, 425 Art 13 ��426–27, 426n.27, 426n.28, 427n.29 Art 14��419, 428, 428n.33, 429 Protocol 1 Art 1��419, 420, 421, 428, 429, 429n.35, 430, 431 Protocol 4 Art 2��421, 431 European Declaration of Human Rights Art 6��698n.3 Treaty Establishing the European Community (EC Treaty)�� 609, 609n.3, 610n.12, 768 Art 18��666 Art 25��665–66 Arts 39–42��666 Arts 43–48��666 Arts 49–55��666 Arts 56–60��666 Art 90��665–66 Art 93�� 665–66, 771

xlvi Table of Legislation Art 94��666 Art 95��666 Art 249��675 Art 293��609, 767 Art 308��666 Treaty on European Union (TEU)��593–94, 609, 609n.2, 629n.2, 663–64 Art 3(3)�� 594, 594n.14, 725 Art 4��611, 664–65 Art 4(3) �� 674 Art 4(3)(2)��595 Arts 4–5�� 787n.7 Art 5��664–65 Art 5(3)�� 687 Art 5(4) �� 687 Art 6(1)��663–64 Art 19�� 599 Art 19(1)�� 594 Art 20��971–72 Art 49��663–64 Art 50�� 596–97 Art 50(3) �� 596–97 Art 55(1)�� 258–59 Art 55(2)�� 258–59 Arts 90–93��665–66 Arts 268–280��665–66 Treaty of Lisbon �� 609n.2, 609n.3, 700 Treaty on the Functioning of the European Union (TFEU)�� 50–51, 593–94, 609, 609n.3, 611, 611–12n.17, 612, 613–15, 616, 617, 618, 619–20, 627, 629, 634, 640, 647, 650, 653, 665–66, 679–80, 725n.162, 747, 763, 768, 771, 884 Art 3��664–65 Art 3(3)�� 593–94 Arts 3–6��611n.15 Art 4�� 664–65, 772–73, 787 Art 4(2) �� 593–94 Art 4(2)(a)�� 609n.4, 665n.5 Art 4(3) �� 593–94

Art 5(1)�� 593n.9 Art 6��664–65 Art 18�� 613n.26 Art 18(1)��647–48 Art 21��611, 611–12n.17, 666 Art 24�� 750 Art 24(4)��260n.43 Art 26��725 Art 26(2) �� 884n.5 Art 28��611–12n.17 Arts 28–32��884n.6 Art 29�� 750 Art 30��665–66 Art 31 ��611–12n.17 Art 34�� 611 Art 36��611–12n.17 Art 40��611–12n.17 Art 45�� 611, 613 Arts 45–48�� 666, 884n.7 Art 49��605–6, 611, 613, 656–57, 971–72 Arts 49–55�� 666, 884n.8 Art 54��605–6 Art 56�� 611 Arts 56–62�� 666, 884n.9 Art 63�� 605–6, 611, 653 Arts 63–65��50–51 Arts 63–66��666, 884n.10 Art 65�� 50–51, 613 Art 65(1)(a)��653 Art 65(1)(b)�� C36P34 Art 75��666 Art 107��627n.113, 641, 685 Art 107(1)�� 629, 641, 643, 644, 671–72, 684–85 Art 107(3)(b)��633 Arts 107–109�� 51, 629 Art 108(1)�� 630 Art 109��280n.85 Art 110 �� 658, 665–66, 749 Art 110(1)��748, 749 Art 110(2)�� 749 Arts 110–113��665–66

Table of Legislation xlvii Art 111�� 658 Art 113�� 597–98, 665–66, 749, 771, 779, 787, 884n.4 Art 114 �� 598, 666 Art 114(2)�� 597–98 Art 115��51–52, 666, 687 Art 116 �� 598n.32 Art 116(2)�� 598 Art 117 �� 598 Art 258��604–5, 776 Arts 251–281�� 599 Art 267��594, 595, 596, 699n.10, 700, 772, 776 Art 267(3) �� 887–88, Art 273�� 607, 707 Art 288��787n.9 Art 291��779n.48 Arts 310–325 ��665–66, 672–73 Art 311��669 Art 325��669 Art 344��700 Art 351 �� 610n.12, 693–94, 695 Art 352�� 664–65, 666 Art 398�� 779 Directive

Directive 67/227/EEC of 11 April 1967 on the harmonisation of legislation of Member States concerning turnover taxes [1967] OJ 71/1301�� 731n.2, 731n.4, 771, 785n.2, 1044n.12 Preamble 1��731n.4 Preamble 3��731n.4 Art 2�� 1044 Directive 67/228/EEC of 11 April 1967 on the harmonisation of legislation of Member States concerning turnover taxes—Structure and procedures for application

of the common system of value added tax [1967] OJ 71/1303�� 731n.2, 733n.14, 785–86, 785n.2, 1045, 1045n.13 Directive 77/388 Sixth VAT Directive��675, 731n.3, 732, 734, 771–72, 1045, 1047 Preamble 4��731n.6 Art 9(1)��733 Art 9(2)(e)��733 Art 19��675 Art 24�� 676 Art 26��733 Art 28(3)(b)�� 676 Art 28(4) �� 676 Annex III��675 Annex F(17)�� 676 Directive 77/799/EEC of 19 December 1977 concerning mutual assistance by the competent authorities of the member states in the field of direct taxation [1977] OJ L336 �� 52n.31, 371n.110, 420n.12 Directive 88/361/EEC of 24 June 1988 for the implementation of Article 67 of the Treaty [1988] OJ L178/5�� 652n.30 Art 1(1)��652 Directive 90/434/EEC of 23 July 1990 on the common system of taxation applicable to mergers, divisions, transfers of assets and exchanges of shares concerning companies of different Member States [1990] OJ L225/1��451n.66, 885

xlviii Table of Legislation Art 4(2) ��451 Art 11(1)(a)�� 680n.4 Directive 90/435/EEC of 23 July 1990 on the common system of taxation applicable in the case of parent companies and subsidiaries of different member states [1990] OJ L225/6 �� 51, 203n.35 Art 1(2)��600, 680n.3 Annex I ��652 Directive 91/680/EEC of 16 December 1991 supplementing the common system of value added tax and amending Directive 77/388/EEC with a view to the abolition of fiscal frontiers [1992] OJ L272/72 �� 1046n.14 Directive 92/12/EEC Art 8�� 677 Directive 92/79/EEC Art 2(1)�� 677 Directive 2002/38/EC of 7 May 2002 amending and amending temporarily Directive 77/388/EEC as regards the value added tax arrangements applicable to radio and television broadcasting services and certain electronically supplied services [2002] OJ L128/41��484n.60, 1046n.15 Directive 2003/48/EC of 3 June 2003 on taxation of savings income in the form of interest payments [2003] OJ L157, 38��53n.33, 602, 602n.60 Directive 2003/49/EC of 3 June 2003 on a common system

of taxation applicable to interest and royalty payments made between associated companies of different member states as amended, consolidated text [2013] OJ L141, 30�� 51, 378, 451n.69, 600, 600n.47, 646, 885 Art 1�� 378, 677–78 Art 5(1)��680n.3 Art 5(2)��600, 680n.4 Directive 2003/96/EC Art 7�� 677 Art 9�� 677 Art 10�� 677 Directive 2003/123/EC of 22 December 2003 [2004] OJ L7/41 ��203n.35, 451n.68 Directive 2006/79/EC of 5 October 2006 on the exemption from taxes of imports of small consignments of goods of a non-commercial character from third countries�� 757, 757n.34 Directive 2006/112/EC of 28 November 2006 on the common system of value-added tax [2006] OJ L347/1��484, 731, 747, 749–63, 749n.2, 766–67, 771–73, 776, 787, 788, 789, 790, 795, 1047, 1047n.16, 1049–50, 1056n.46 Preamble��767n.9 Recital 5��773 Recital 7��773 Recital 8��773, 774 Recital 17��774 Recital 62��774 Art 2�� 788 Art 2(1)(d)�� 749, 750, 752,

Table of Legislation xlix 755, 756 Art 14(1)�� 750 Art 14(a)��1049–50 Art 16��1052 Art 17a�� 744 Art 24��790n.23, 791n.24 Art 26(1)��1052 Art 30�� 749n.3 Art 30(1)��750, 752 Art 31ff�� 770 Arts 31–39 ��731n.5 Art 34��1049–50 Art 36(2) �� 777n.38 Art 38�� 768, 769 Art 39�� 768 Art 43ff�� 770 Art 43–59c��731n.5 Art 44�� 733n.19, 734n.20, 769, 789n.19, 791–92 Art 45��733n.19, 734n.20, 792 Art 47�� 791, 795 Art 53��791 Art 59a�� 777 Art 59c��1049–50 Art 60��749n.3, 751, 752, 754–55 Art 61��749n.3, 750, 751, 752 Art 61(1)��751, 752 Art 70�� 749n.3, 752, 755 Art 71�� 749n.3 Art 71(1)(1)�� 752, 753 Art 71(1)(2) ��753 Art 71(2)��753 Art 75�� 1053 Art 80�� 756–57 Art 85�� 755, 756–57 Art 85–89�� 749n.3 Art 86(1)(b)�� 756 Art 87�� 756 Art 91�� 749n.3 Art 93��743n.93, 749n.3 Art 94�� 749n.3 Art 132(1)(f)��743

Art 138�� 1055 Art 138(1)�� 758–59 Art 138(2)(c) �� 758–59 Art 143(1)��757 Art 143(1)(a) ��757 Art 143(1)(b) ��757 Art 143(1)(d)�� 758–59 Art 143(1)(f)–(i) �� 758n.37 Art 143(1)(j)–(l)�� 758n.38 Art 143(2)��758 Art 143(b)��758 Art 143–45�� 749n.3 Art 156�� 751, 762 Art 157 �� 762 Art 158 �� 762 Art 160�� 762 Art 161 �� 762 Art 168��760 Art 168(e)�� 759, 760 Art 168 lit. e�� 749n.3 Art 178(e)��759–60 Art 178 lit. e�� 749n.3 Art 192a��734n.22, 742 Art 201��749n.3, 759, 761–62 Art 202�� 761–62 Art 205�� 762 Art 211 ��749n.3, 763 Art 211(1)�� 762 Art 211(2)�� 762–63 Art 242a��1050n.21 Art 250�� 762–63 Art 260�� 749n.3, 762 Art 262�� 1055 Art 274–277�� 749n.3 Art 398�� 779 Art 408–410 �� 749n.3 Directive 2007/74/EC of 20 December 2007 on the exemption from value- added tax and excise duty of goods imported by persons travelling from third countries�� 757, 757n.32

l Table of Legislation Directive 2008/8/EC of 12 February 2008 amending Directive 2006/112/EC as regards the place of supply of services [2008] OJ L44/11 �� 733n.17, 789, 1047n.17 Directive 2009/132/EC of 19 October 2009 determining the scope of Article 143(b) and (c) of Directive 2006/ 112/EC in regards to exemption from value- added tax on the final importation of certain goods��208, 757, 757n.33, 758 Directive 2009/133/EC of 19 October 2009 on the common system of taxation applicable to mergers, divisions, partial divisions, transfers of assets and exchanges of shares concerning companies of different member states and to the transfer of the registered office of an SE or SCE between member states as amended, consolidated text [2013] OJ L141, 30 ��51, 208, 646 Art 15(1)(a) �� 680n.4 Directive 2009/138/EC Solvency II Directive Art 157��660 Directive 2010/24/EU of 16 March 2010 concerning mutual assistance for the recovery of claims relating to taxes, duties and other measures�� 646, 655–57, 725 Directive 2011/16/EU of 15 February 2011 on administrative cooperation

in the field of taxation and repealing Directive 77/799/ EEC [2011] OJ L64, 1–12 ��53n.32, 355, 672, 672n.28, 682n.16, 1050, 1050n.23, 1050n.24 Art 8.a.1��711n.84 Art 12�� 707n.61 Directive 2011/96/EU of 30 November 2011 on the common system of taxation applicable in the case of parent companies and subsidiaries of different member states, amended through Directive 2014/ 86/EU of the Council of 8 July 2014, consolidated text [2015] OJ L96, 1�� 51, 600, 600n.46, 602, 602n.61, 646, 653, 681, 694–95, 885 Art 1(2) �� 680n.3, 681 Art 1(3)��681 Art 1(4) ��600 Art 4(1)(a)��681 Directive 2014/86/EU of 8 July 2014 amending Directive 2011/ 96/EU on the common system of taxation applicable in the case of parent companies and subsidiaries of different Member States [2014] OJ L219/40 �� 681n.10 Directive 2014/107/EU of 9 December 2014 amending Directive 2011/16/EU as regards mandatory automatic exchange of information in the field of taxation [2014] OJ L359, 1–29 ��53n.35, 682n.16

Table of Legislation li Directive (EU) 2015/849 Money Laundering Directive��53–54 Directive (EU) 2015/121 of 27 January 2015 amending Directive 2011/96/EU on the common system of taxation applicable in the case of parent companies and subsidiaries of different Member States [2015] OJ L21/1�� 203n.35, 600n.46, 681n.11 Directive (EU) 2015/2376 of 8 December 2015 amending Directive 2011/16/EU as regards mandatory automatic exchange of information in the field of taxation [2015] OJ L332, 1–10�� 53n.36, 682n.16 Directive (EU) 2016/881 of 25 May 2016 amending Directive 2011/16/EU as regards mandatory automatic exchange of information in the field of taxation [2016] OJ L146, 8–21 �� 53n.37, 682n.16 Directive (EU) 2016/1164 of 12 July 2016 laying down rules against tax avoidance practices that directly affect the functioning of the internal market [2016] OJ L193, 1–14 (ATAD)�� 51–52, 146n.11, 200n.15, 263, 280n.86, 581, 594, 594n.11, 598–99, 598n.38, 602, 603–7, 679, 679n.2, 681–82, 681n.13, 685–87, 688–89, 690–95, 690n.67, 690n.68, 691nn.70–72, 707, 712, 884–85, 884n.11 Recital 2��687n.39 Recital 3��687n.39 Recital 6��689n.57 Recital 10�� 688n.51

Recital 11 ��687n.45 Recital 12�� 605n.75, 688n.46 Recital 16�� 687n.39, 687n.44 Art 1��687n.42 Art 2(9) ��452 Art 3�� 692 Art 4��681–82, 687, 689, 691–92, 692n.79 Art 4(1)��689n.59 Art 4(3)(b)��689n.58, 689n.60, 692n.85 Art 4(4) ��689n.62 Art 4(5) ��689n.63 Art 4(7) ��689n.61, 692–93 Art 5��208n.59, 681–82, 687, 688–89, 694 Art 5(1)�� 688n.52 Art 5(2)�� 688n.53 Art 5(3)�� 688n.54, 692n.84 Art 5(5)�� 689n.55, 691 Art 6�� 602n.62, 603–4, 681–82, 687, 689–90, 692 Art 6(1)�� 690n.65 Arts 7–8��325n.10 Art 7�� 325, 605–6, 681–82, 687–88 Art 7(1)��693–94 Art 7(1)(a)��688n.47 Art 7(1)(b)�� 688n.48 Art 7(2)(a)�� 688n.49 Art 7(2)(a)(vi)��263 Art 7(2)(b)��688n.50 Art 8�� 681–82, 687–88 Art 9��452, 681–82, 687, 690 Art 9a�� 452, 681–82, 687, 687n.43, 690, 690n.67 Art 9a(1)��452 Art 9b�� 681–82, 687, 690 Art 9(1)��452 Art 9(2) ��452 Art 9(2)(b)��693n.89 Art 11 �� 687n.41, 689n.64 Art 11(6)�� 691n.77 Art 23�� 712n.89

lii Table of Legislation Directive (EU) 2016/2258 of 6 December 2016 amending Directive 2011/16/EU as regards access to anti-money-laundering information by tax authorities [2016] OJ L342, 1–3 ��53n.38, 682n.16 Directive (EU) 2017/952 of 29 May 2017 amending Directive (EU) 2016/1164 as regards hybrid mismatches with third countries [2017] OJ L144, 1–11�� 51–52, 151n.19, 201n.21, 280n.86, 452, 581, 679n.2, 681–82, 681n.13, 686, 690n.68, 884n.11, 885 Art 2�� 687n.41 Art 4��598–99 Art 6��598–99 Art 9(4) �� 690n.69 Directive (EU) 2017/1371 of the European Parliament and of the Council of 5 July 2017 on the fight against fraud to the Union’s financial interests by means of criminal law [2017] OJ L198/29/�� 670 Art 2(2) �� 670 Art 3(2)�� 670 Directive (EU) 2017/1852 of 10 October 2017 on tax dispute resolution mechanisms in the European Union [2017] OJ L265, 1–14 �� 52n.27, 187n.54, 523n.24, 552–53, 594n.12, 606n.81, 672, 672n.29, 702, 707, 707n.63, 712–26, 712n.87, 778 Recital 2��713n.92 Recital 6�� 187–88n.55, 713n.93 Recital 7��713n.93 Recital 9�� 723–24

Art 1�� 713n.91, 723n.151 Art 2.1�� 713n.90 Art 2(2) ��719 Art 3��713n.95 Art 3.1�� 714n.97, 714n.99 Art 3.3f��714 Art 4.1(1) �� 715n.105 Art 4.3 ��715n.107 Art 5.2�� 714n.100 Art 5.3��714n.101 Art 5.4 �� 718n.125 Art 6��713n.92, 718–19, 718n.129 Art 6.1��715n.108 Art 6.1(1) �� 715n.102 Art 6.1(2)�� 715n.103 Art 6.1(3)��716n.110 Art 6.2(3)�� 716n.114 Art 6.3 �� 716n.114 Art 8�� 716n.109, 716n.111, 718 Art 9�� 716n.112 Art 10��174n.79 Art 10.2�� 720n.136 Art 11.3�� 726n.165 Art 13.1��717n.120 Art 14.2��716n.115 Art 14.3��716n.116, 717n.124 Art 15.1�� 716n.117 Art 15.2�� 716n.118, 717n.122 Art 15.3�� 717n.123 Art 16��719n.132, 721 Art 16.4 �� 719n.133 Art 16.6 ��718–19, 718n.127, 725 Art 16.7��713n.92, 719n.131, 725 Art 18.2�� 726n.168 Art 18.3(2)�� 726n.169 Art 19.3�� 726n.166 Art 20�� 726n.167 Art 21�� 726n.164 Art 23�� 712n.89 Directive (EU) 2017/2455 of 5 December 2017 amending Directive 2006/112/EC and Directive 2009/132/EC as

Table of Legislation liii regards certain value added tax obligations for supplies of services and distance sales of goods [2017] OJ L348/7��758, 1048n.18, 1049–50, 1049n.20 Directive (EU) 2018/822 of 25 May 2018 amending Directive 2011/16/EU as regards mandatory automatic exchange of information in the field of taxation in relation to reportable crossborder arrangements [2018] OJ L139, 1–13��53n.39, 594, 594n.13, 682n.16 Directive (EU) 2018/1695 of 6 November 2018 amending Directive 2006/112/EC on the common system of value added tax as regards the period of application of the optional reverse charge mechanism in relation to supplies of certain goods and services susceptible to fraud and of the Quick Reaction Mechanism against VAT fraud [2018] OJ L282/5 ��1056n.44 Directive (EU) 2018/2057 of 20 December 2018 amending Directive 2006/112/EC on the common system of value added tax as regards the temporary application of a generalised reverse charge mechanism in relation to supplies of goods and services above a certain threshold [2018] OJ L329/3 �� 1056n.45

Directive (EU) 2019/1995 of 21 November 2019 amending Directive 2006/112/EC as regards provisions relating to distance sales of goods and certain domestic supplies of goods [2019] OJ L310/1�� 1048n.18 Directive (EU) 2020/284 of 18 February 2020 amending Directive 2006/112/EC as regards introducing certain requirements for payment service providers [2020] OJ L62/7 �� 1050n.22 Directive (EU) 2021/514 of 22 March 2021 amending Directive 2011/16/EU on administrative cooperation in the field of taxation [2021] OJ L104/1 (DAC 7)�� 682n.16 Decisions

Decision 1999/779/EC of 3 February 1999 [1999] OJ L305/27�� 649n.18 Decision 2003/757/EC of 17 February 2003, on the aid scheme implemented by Belgium for coordination centres established in Belgium [2003] OJ L282�� 638n.39 Decision 2004/77/EC of 24 June 2003, on the aid scheme implemented by Belgium— Tax ruling system for United States foreign sales corporations [2004] OJ L23�� 638n.38

liv Table of Legislation Decision 2007/75/EC of 22 Dececber 2006 setting up an expert group on transfer pricing [2007] OJ L32/189��706n.59 Art 2�� 706n.60 Decision (EU) 2016/1699 of 11 January 2016 on the excess profit exemption State aid scheme S.A.37667 (2015/C) (ex 2015/NN) implemented by Belgium (notified under document C(2015) 9837), OJ L260/61�� 636n.32, 684–85nn.28–29 Decision 2017/502/EU of 21 October 2015, Case SA.38374, on state aid SA.38374 (2014/C ex 2014/ NN) implemented by the Netherlands to Starbucks [2017] OJ L83 �� 639n.43 Decision 2017/1283 of 30 August 2016 on state aid, SA. 38373 (2014/C) [2017] OJ L187 ��641n.49 Decision (EU) 2018/859 of 4 October 2017 on State aid SA.38944 (2014/C) (ex 2014/NN) implemented by Luxembourg to Amazon (notified under document C(2017) 6740) [2018] OJ L153/1��684n.28 Decision (EU) 2020/1109 of 20 July 2020 amending Directives (EU) 2017/2455 and (EU) 2019/1995 as regards the dates of transposition and application in response to the COVID-19 pandemic [2020] OJ L244/3�� 1048n.18

Decision (EU, Euratom) 2020/2053 of 14 December 2020 on the system of own resources of the European Union and repealing Decision 2014/ 335/EU, Euratom [2020] OJ L424/1 �� 667–68, 667n.14, 669, 670n.22 Art2(1)(b) ��669n.21 Regulations

Regulation No. 1 determining the languages to be used by the European Economic Community of 15 April 1958 [1958] OJ 385/58�� 258–59 Regulation (EEC, Euratom) No. 1553/89��669n.21 Regulation (EC, Euratom) No 2988/95 of 18 December 1995 on the protection of the European Communities financial interests [1995] OJ L312/1��670, 673n.31 Regulation (EC) No. 883/2004 of the European Parliament and of the Council of 29 April 2004 on the coordination of social security systems [2019] OJ L186, 21�� 52n.28 Regulation (EC) No. 593/2008 of the European Parliament and of the Council of 17 June 2008 on the law applicable to contractual obligations (Rome I) Art 3��161n.9 Art 9�� 151n.11 Regulation (EC) No. 987/2009 of the European Parliament

Table of Legislation lv and of the Council of 16th September 2009 laying down the procedure for implementing Regulation (EC) No. 883/2004 on the coordination of social security system��781n.54 Regulation (EC) 1186/2009 of 16 November 2009 setting up a Community system of reliefs from customs duty �� 750, 750n.9, 757, 757n.35 Regulation (EU) No. 904/2010 of 7 October 2010 on administrative cooperation and combating fraud in the field of value added tax [2020] OJ L268/1 ��773, 773n.24, 775, 776, 781–82, 784, 1055, 1055n.42 Art 1��773 Art 7��774 Art 8��773 Art 13 ��774 Arts 28–30��774 Regulation (EU) No. 282/2011 of 15 March 2011 laying down implementing measures for Directive 2006/112/EC on the common system of value added tax [2011] OJ L77/1 ��734, 734n.28, 737, 742, 769n.13, 791–92, 791n.26, 1049n.19 Preamble 14��734n.24 Art 9a�� 745–46, 1049 Art 10��735–36 Art 10(1)�� 736n.36 Art 10(2) ��735–36 Art 10(3)�� 736n.37 Art 11 ��742 Art 11(1)��737n.50, 739–40

Art 11(2)�� 737n.49 Art 21�� 791–92 Art 22��739 Art 22(1)��737n.52, 791n.28 Art 22(2) ��737n.52, 791n.29, 791n.30, 795n.53 Art 24a�� 792 Art 24b�� 792 Art 24d�� 792 Art 24f�� 792 Art 52��759–60 Art 53(1)�� 742n.89 Art 53(2)��742n.89, 742n.90 Regulation 517/2013 [2013] OJ L158/1 Art. 1�� 258n.37 Regulation (EU) No. 952/2013 of the European Parliament and of the Council of 9 October 2013 laying down the Union Customs Code�� 750, 757, 760 Art 70�� 756–57 Art 70(1)��755–56 Art 70(3) �� 756 Art 70(3)(d)�� 755–56, 757 Art 71(1)(c) ��755–56 Art 74�� 756 Art 74(2) ��755 Art 74(3)��755 Art 77(3)�� 761–62 Art 77(3)(2)�� 761–62 Art 79��760 Art 79(3) �� 762 Art 79(b)�� 762 Art 79(4) �� 762 Art 82��760 Art 87(4) ��754–55 Art 105(4)�� 756–57 Art 116ff��761 Art 120��761 Art 120(1)��761 Art 124��760

lvi Table of Legislation Art 124(1)(e) �� 755, 760n.47 Art 124(1)(h)��760 Art 124(1)(k)��760 Art 201��751, 754–55 Art 201(2)�� 750 Art 201(3)�� 750 Art 210��754–55 Art 210ff�� 751 Art 226ff��752 Regulation (EU) 2015/1589 of 13 July 2015 laying down detailed rules for the application of Article 108 of the Treaty on the Functioning of the European Union (Text with EEA relevance) [2015] OJ L248 ��630n.4 Art 1(d) �� 636–37, 636n.30 Regulation (EU) 2015/2446 of 28 July 2015 supplementing Regulation (EU) No. 952/ 2013 of the European Parliament and of the Council as regards to detailed rules concerning certain provisions of the Union Customs Code�� 750n.7 Regulation (EU) 2015/2447 of 24 November 2015 laying down detailed rules for implementing certain provisions of Regulation (EU) No. 952/2013 of the European Parliament and of the Council laying down the Union Customs Code�� 750, 750n.8 Art 127�� 755–56, 757 Art 134�� 755–56, 757 Art 136(4)��755–56 Regulation (EU) 2016/679 of the European Parliament and

of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC��426n.26, 830–31 Regulation (EU, Euratom) 2018/ 1046 of the European Parliament and of the Council of 18 July 2018 on the financial rules applicable to the general budget of the Union, amending Regulations (EU) No 1296/2013, (EU) No 1301/2013, (EU) No 1303/2013, (EU) No 1304/ 2013, (EU) No 1309/2013, (EU) No 1316/2013, (EU) No 223/2014, (EU) No 283/ 2014, and Decision No 541/ 2014/EU and repealing Regulation (EU, Euratom) No 966/2012 [2018] OJ L193/1��670, 670n.23 Art 33�� 673–74 Regulation (EU) 2018/1541 of 2 October 2018 amending Regulations (EU) No. 904/ 2010 and (EU) 2017/2454 as regards measures to strengthen administrative cooperation in the field of value added tax [2018] OJ L259/1 �� 1056n.43 Regulation (EU) 2019/52 of 24 April 2019 laying down standard Rules of Functioning for the

Table of Legislation lvii Advisory Commission or Alternative Dispute Resolution Commission and a standard form for the communication of information concerning publicity of the final decision in accordance with Council Directive (EU) 2017/1852 [2019] OJ L110/26 �� 717n.119, 726n.165 Regulation (EU) 2020/194 of 12 February 2020 laying down detailed rules for the application of Council Regulation (EU) No. 904/2010 as regards the special schemes for taxable persons supplying services to non-taxable persons, making distance sales of goods and certain domestic supplies of goods [2020] OJ L40/114 ��1050n.21 Regulation (EU, Euratom) 2020/ 2092 of the European Parliament and of the Council of 16 December 2020 on a general regime of conditionality for the protection of the Union budget [2020] OJ L433I/1��664n.2 Regulation (EU) 2021/1529 of the European Parliament and of the Council of 15 September 2021 establishing the Instrument for Pre-Accession assistance (IPA III) [2021] OJ L330/1��678n.39

NATIONAL INSTRUMENTS Australia

Customs Law Implementing Act Art 73��761 Domicile Act 1982 (Cth)��116n.36 Income Tax Assessment Act 1936 (Cth) s. 6(1) ��114n.28 s. 6(1)(a)(i)��117n.47 s. 6(1)(a)(ii)��116n.38, 117n.47 s. 6(1)(a)(iii) �� 117n.45 Treasury Laws Amendment (2021 Measures No. 2) Act No. 110, 2021 (Cth), Sch. 2 �� 488n.81 Belgium

Code des impôts sur les revenus 1992 art. 2�� 116n.38, 117n.43, 117n.46 art 228 §§ 1–2�� 121n.59 Code judiciaire, s. 36��115n.33 Brazil

Decree 9.482/2019�� 875n.59 Law 9.249/25 Art 25��865n.6 Law 4.131/1962 Art 12�� 867 Art 42�� 867n.21 Law 6.404/1976��209–10 Law 2.303/1986 Art 7��865n.6 Law 9.249/1995 Art. 10 �� 203n.34 Art. 22 �� 206n.49 Art 25(5)�� 207n.51, 865n.7 Law 9.430/1996�� 866n.13 Law 9.779/1999 Art 8��868n.23 Law 10.833/2003 Art 26�� 206n.50 Law 12.249/2010 ��866n.14 Law 12.973/2014 Art 78��865n.8

lviii Table of Legislation Law 4506/1964�� 866n.12 Law-Decree 5.844/43 Art 27��865n.6 Art 97(a) �� 867n.18 National Tax Code �� 868 Art 98��868n.25 Art 116 �� 865 Ordinance 436/1958 �� 866n.12 Provisional Measure 2.159-70/2001 Art 3�� 867n.18 Provisional Measure 685/2015 Art 7��866n.16 Tax Code�� 868 Bulgaria

Constitution of the Republic of Bulgaria Art 4��663–64 Canada

Bankruptcy and Insolvency Act, RSC 1985 s 2��942–43n.14 Income Tax Act RSC, c.1 1985 (5th Supp.) �� 113n.27 s 52.1(3)(c)��942–43n.14 s 52.1(3)(d)��942–43n.14 s 69(2) �� 546n.61 s 250(1)(a) �� 116n.41 s 250(1)(c) ��117n.44 s 250(1)(d.1)��117n.44 s 250(1)(f) ��117n.44 s 250(1)(g) ��117n.44 Income Tax Conventions Interpretation Act of 20 December 1984��47n.11 China

EIT Regulations, art. 17(4)��203n.30

France

Code général des impôts Art 4B�� 115n.33, 117n.44 General Tax Code Art 752�� 165n.30 Germany

Basic Law Art 3�� 278–79 Code of Civil Procedure Art 293�� 169n.54 Constitution �� 278–79 Art 1��886–87 Art 2��886–87 Art 3(1)�� 887 Art 23��886–87 Art 24��886–87 Art 59��886–87 Art 95�� 888 Art 101(1)��887–88, 889 Art 101(2)�� 887 Fiscal Code (Abgabenordnung) Art 2�� 563–64 § 8�� 115n.33, 116n.38 § 9��116n.38 Foreign Tax Act (AStG)�� 47n.11, 882–83, 893 s 1 �� 887 s 6�� 894 ss 7–14 ��893n.21 s 15�� 895 s 8�� 893 Foreign Transaction Tax Act s 1(1)�� 887 General Tax Code (GTC) s 42�� 890, 890n.16, 891 s 42(1)��890 s 42(2)��890 s 50d(3)�� 388n.93 s 53 �� 889–90

Table of Legislation lix Income Tax Act—EStG s 4K�� 896 s 17 �� 894 s 50d(3)�� 604, 893 Inheritance and Gift Tax Act s 1 �� 886 s 13a�� 886 s 13b �� 886 s 19 �� 886

Law No. 228 laying down provisions for the establishment of the annual and multiannual State budget (Stability Law 2013) Art 1(491)��163n.22 Art 1(492)��163n.22 Art 1(494) ��163n.22

India

Kenya

Black Money (Undisclosed Foreign Income and Assets) and Imposition of Tax Act 2015��853 Finance Act 2012��853 Finance Act 2021��860 Income-tax Act, 1961 �� 407n.85, 842–44, 848–49, 851, 852 Ch XIX B�� 848n.37 Ch XX ��848n.38 s 2(29A)�� 845n.16 s 3��842–43, 842n.4 s 5�� 842–43 s 9�� 843 s 90�� 844 s 90(2) �� 844, 850 s 90(2A)�� 850 s 92CB �� 848n.37 s 92CC�� 848n.37 s 94B��861 s 144C��848n.38 s 195 ��861

Finance Act 2018 s 32(b)(i)��906n.60 Income Tax Act 1973��906n.58 s 3(ba)��906n.58 s 12E��906n.58 Sch 3(12)��906n.58

Italy

Codice Civile Art 43(1)�� 115n.33, 116n.38, 117n.43 Constitution (art. 53) �� 278–79 Law 218/1995 Art 14�� 169n.54 Art 15 �� 169n.54

Luxembourg

Income Tax Code Art 164ter��692–93 Art 168bis(8)��693n.88 Art 413(2)–(8)��699n.11 New Zealand

Income Tax Act 2007 s YD1(6)��116n.37, 116n.40 s YD1(7)��117n.44 s YD 4�� 121n.61 Land and Income Tax Assessment Act 1891 Sch C (1)�� 120n.57 Tanzania

Electronic and Postal Communications (Online) Content Regulations 2018�� 906n.62

lx Table of Legislation Income Tax Act 2019 �� 904, 904n.41 s 1 �� 904n.42 Uganda

Excise Duty (Amendment) Act 2018 s 4(5)��906n.61 Sch 2��906n.61 Income Tax Act 1997��904 s 3(2)(h)��904n.43 United Kingdom

Finance Act 2008 ��165n.34 Finance Act 2013 s 218 ��114n.29, 116, 118n.51 Sch 45�� 114n.29, 118n.51 paras 6–16��116n.42 Income Tax Act 2007 s 992��209n.65 Income Tax Act 2010 s 1121��209n.65 United States

26 CFR § 1.861–8(f)�� 121n.61 CFR § 1.1–1(c) �� 112n.19 Currency Act 1764 �� 10–11 Foreign Account Tax Compliance Act (FATCA) �� 53, 354, 420–21n.13 Internal Revenue Code, 26 USC § 1��537, 112n.19 § 61�� 276n.71 § 428�� 546

§ 482��535n.3 § 501(c)��345 §§ 951–964�� 324n.8 § 951A�� 325n.9 § 954(c)(6)�� 152–53, 152n.25 § 7701(b)(3)��116n.39 Quartering Act 1765�� 10–11 Restatement (Fourth) of the Foreign Relations Law of the United States Part IV��65–66 § 407��67n.8 Restatement (Second) of the Foreign Relations Law of the United States § 187(2)(b)�� 161n.10 Restatement (Third) of the Foreign Relations Law of the United States § 102�� 68n.14 § 102(2)�� 69n.17, 92n.31 § 102(3)��69–70 § 102, Comment b�� 69n.18 Revenue Act 1921�� 518 § 240(d)��518n.3 Revenue Act 1928 �� 518 § 45�� 518n.4 Stamp Act 1765 �� 10–11 Sugar Act 1764�� 10–11 Tax Act 2017��799–811, 799n.1 Tax Reform Act 1986�� 799–800, 801–2 Townshend Act 1767�� 10–11 Treasury Regulations § 301.7701-2(b)(1)–(7).��209n.63

Abbreviations

AB

Advisory Body (WTO)

AC

Arbitration Convention

ADRC

Alternative Dispute Resolution Commission

ADS

Automated Digital Services

AEOI

Automatic Exchange of Information

AG

Advocate General

ALP

Arm’s-Length Principle

ALS

Arm’s-Length Standard

AMT

Alternative Minimum Tax

AOA

Authorized OECD Approach

APA

Advance Pricing Agreement/Arrangement

APC

Administrative Procedure Code (Bulgaria)

ASEAN

Association of South-East Asian Nations

ATAD

Anti-Tax Avoidance Directive (EU)

ATAF

African Tax Administration Forum

ATAP

Anti-Tax Avoidance Package (EU)

B2B

Business-to-Business

B2C

Business-to-Consumer

BEAT

Base Erosion and Anti-Abuse Tax (USA)

BEPS

Base Erosion and Profit Shifting

BP

Benefits Principle

BRICS

Brazil, Russia, India, China, and South Africa

BTA

Border Tax Adjustments

CAMT

Corporate Alternative Minimum Tax (USA)

CbCR

Country-by-Country Reports/Reporting

CBDT

Central Board of Direct Taxes (India)

CCA

Cost Contribution Arrangement

CCCTB

Common Consolidated Corporate Tax Base (EU)

CE

Constituent Entity

lxii Abbreviations CEN

Capital Export Neutrality

CFB

Consumer-Facing Business

CFC

Controlled Foreign Corporation/Company

CIL

Customary International Law

CIN

Capital Import Neutrality

CIT

Corporate Income Tax

CITL

Consensus International Tax Law

CJEU

Court of Justice of the European Union

CMAA

Convention on Mutual Administrative Assistance in Tax Matters

COMESA

Common Market for Eastern and Southern Africa

CON

Capital Ownership Neutrality

CPM

Cost-Plus Method

CRS

Common Reporting Standard

CSA

Cost-Sharing Arrangement

CTA

Covered Tax Agreement

CUP

Comparable Uncontrolled Price

DAC

EU Directive on administrative cooperation in the field of taxation

DAPE

Dependent Agent PE

DBCFT

Destination-Based Cash Flow Tax

DD

Double Deduction

D/NI

Deduction Without Inclusion

DST

Digital Services Tax

DTA

Double Tax Arrangement

DTAA

Double Taxation Avoidance Agreement

DTABR

Deemed Tax Added Back Rule

DTC

Double Taxation Convention

DTT

Double Taxation Treaty

EAC

East African Community

EBITDA

Earnings before Interest, Taxes, Depreciation, and Amortization

ECHR

European Convention on Human Rights

ECJ

European Court of Justice

ECLI

European Case Law Identifier

ECnHR

European Commission of Human Rights

ECT

Energy Charter Treaty

ECtHR

European Court of Human Rights

Abbreviations lxiii EBITDA

Earnings before interest, taxes, depreciation and amortization

EL

Equalization Levy (India)

EOI

Exchange of Information

ETR

Effective Tax Rate

EUT

EU Treaty

FACTI

Financial Accountability, Transparency and Integrity

FATCA

Foreign Account Tax Compliance Act (USA)

FDII

Foreign-Derived Intangible Income

FE

Fixed Establishment

FFI

Foreign Financial Institution

FTA

Forum on Tax Administration

FTS

Fees for Technical Services

G20

Group of Twenty

G24

Group of Twenty-Four

G7

Group of Seven

GAAP

Generally Accepted Accounting Principle

GAAR

General Anti-Avoidance Rule

GATS

General Agreement on Trade in Services

GATT

General Agreement on Tariffs and Trade

GDP

Gross Domestic Product

GG

Grundgesetz (German Constitution or ‘Basic Law’)

GILTI

Global Intangible Low-Taxed Income (USA)

GloBE

Global Anti-Base Erosion

GST

Goods and Services Tax

GTC

General Tax Code (Germany)

HMA

Hybrid Mismatch Arrangement

HO

Head Office

HTVI

Hard-to-Value Intangible

IAS

International Accounting Standard

ICAP

International Compliance Assurance Program

ICC

International Chamber of Commerce

ICCR

International Covenant on Civil and Political Rights

ICJ

International Court of Justice

IF

Inclusive Framework

IFRS

International Financial Reporting Standards

lxiv Abbreviations IGA

Intergovernmental Agreement

IIR

Income Inclusion Rule

ILADT

Instituto Latinoamericano de Derecho Tributario (Latin American Institute of Tax Law)

ILC

International Law Commission

IP

Intellectual Property

I&R

Interest and Royalties

IRS

Internal Revenue Service (USA)

ITL

International Tax Law

ITR

International Tax Regime

JCT

Joint Committee on Taxation (USA)

JTPF

Joint Transfer Pricing Forum

LMICs

Low-and Middle-Income Countries

LOB

Limitation on Benefits

MAATM

Multilateral Agreement on Administrative Cooperation in Tax Matters

MAC

Multilateral Convention on Mutual Administrative Assistance and Exchange of Information

MAP

Mutual Agreement Procedure

MC

Model Convention

MCAA

Convention on Mutual Administrative Assistance in Tax Matters

METI

Ministry of Economy, Trade, and Industry (Japan)

MF

Master File

MFN

Most-Favoured-Nation

MLI

Multiple Location Entity

MLI

Multilateral Instrument

MNC

Multinational Corporation

MNE

Multinational Enterprise

NGO

Non-Governmental Organization

NN

National Neutrality

NR

Non-Resident

NTA

National Tax Agency (Japan)

OECD

Organisation for Economic Co-operation and Development

OECD EIGMC

OECD Model Convention for the avoidance of double taxation with respect to taxes on estates and inheritances and on gifts

OECD MC/Model

OECD Model Tax Convention on Income and on Capital

Abbreviations lxv OEEC

Organisation for European Economic Co-operation

PBT

Profit Before Taxes

PCT

Platform for Collaboration on Tax

PE

Permanent Establishment

PIL

Private International Law

PIT

Personal Income Tax

POEM

Place of Effective Management

POS

Place of Supply

PoW

Programme of Work

PPM

Process and Production Method

PPT

Primary/Principal Purpose Test

P–S

Parent–Subsidiary

QBAI

Qualified Business Asset Investment (USA)

QDMTT

Qualified Minimum Domestic Minimum Top-Up Tax

RP

Resale Price

SAAR

Specific Anti-Avoidance Rule

SAARC

South Asian Association for Regional Cooperation

SADC

Southern African Development Community

SBIE

Substance-Based Income Exclusion

SCM

Agreement on Subsidies and Countervailing Duties

SDGs

Sustainable Development Goals

SEP

Significant Economic Presence

SLOB

Simplified Limitation of Benefits

SME

Small and Medium-Sized Enterprise

STP

Single-Tax Principle

STTR

Subject to Tax Rule

TCJA

Tax Cuts and Jobs Act (USA)

TEU

Treaty on European Union

TFEU

Treaty on the Functioning of the European Union

TIEA

Tax Information Exchange Agreement

TNM

Transactional Net Margin

TPM

Transactional Profit Split

TRC

Tax Residency Certificate

TRIMs

Agreement on Trade-Related Investment Measures

TSSPC

Tax and Social Security Procedure Code (Bulgaria)

lxvi Abbreviations UCC

Union Customs Code

UC-IA

Union Customs Code, Commission Implementing Regulation

UN

United Nations

UN MC/ Model

United Nations Model Double Taxation Convention between Developed and Developing Countries

UPE

Ultimate Parent Entity

US EGMC

US Estate and Gift Model Convention

US MC/Model

US Model Income Tax Convention

UTPR

Undertaxed Payments Rule, Undertaxed Profit Rule

VAT

Value Added Tax

VCLT

Vienna Convention on the Law of Treaties

WBG

World Bank Group

WHT

Withholding Tax

WTO

World Trade Organization

List of Contributors

Miguel Teixeira de Abreu, Invited Professor, Universidade Católica de Lisboa—Global School of Law LLM, Portugal. Nadia Altenburg, Partner, Flick Gocke Schaumburg, Germany. Paolo Arginelli, Professor of Tax Law, Department of Law, Università Cattolica del Sacro Cuore, Italy. Reuven S. Avi-Yonah, Irwin I. Cohn Professor of Law, University of Michigan, United States. Philip Baker KC, Visiting Professor, Oxford University, UK. Marco Barassi, Associate Professor of Tax Law, Department of Law, University of Bergamo, Italy. Martin Berglund, Associate Professor of Fiscal Law, Department of Law, Uppsala University, Sweden. Roberto Bernales Soriano, Professor of Tax Law, Public Law Department. Attorney at Law, University of Deusto, Spain. Thomas Bieber, Professor of Tax Law, Institute of Fiscal Law, Tax Law and Tax Policy, Johannes Kepler University Linz, Austria. Gianluigi Bizioli, Professor of Tax Law and International and EU Tax Law at the Department of Law, University of Bergamo, Italy. Bristar Mingxing Cao, Director, China International Tax Center, Central University of Finance & Economics, China. Marcos André Vinhas Catão, Director, ABDF/IFA Brazil, Brazil. Allison Christians, Full Professor and H. Heward Stikeman Chair in Tax Law, Faculty of Law, McGill University, Canada. Kimberly A. Clausing, Eric M. Zolt Professor of Tax Law and Policy, University of California at Los Angeles (UCLA), United States. Adrian Cloer, Professor for Business Taxation, International Tax Law and Audit, EBS Universität for Business and Law, Germany.

lxviii List of Contributors Robert J. Danon, Full Professor of Swiss and International Tax Law, University of Lausanne, Switzerland. Kuntal Dave, Proprietor— Nanubhai Desai & Co., The Institute of Chartered Accountants of India, India. Michael Dirkis, Professor of Taxation Law, University of Sydney Law School, Australia. Craig Elliffe, Professor of Taxation Law, Faculty of Law, University of Auckland, New Zealand. Joachim Englisch, Professor of Public Law and Tax Law, Institute for Tax Law, University of Münster, Germany. Heidi Friedrich-Vache, Lecturer Ludwig-Maximilians-University Munich, Certified Tax Advisor and Partner (Global Head of VAT Services) at Rödl & Partner, Germany. Heidi Friedrich-Vache, University, Germany.

Lecturer

Ludwig- Maximilians

Ludwig- Maximilians-

Clémence Garcia, Professor of Accounting, Department of International Social Sciences, Gakushuin University, Japan. Elizabeth Gil García, Lecturer and Tax Law Department, University of Alicante, Spain. Dietmar Gosch, Honorary Professor, Christian-Albrechts-Universität zu Kiel, Lawyer, Tax Advisor, Partner WTS GmbH, Presiding Judge at the Federal Supreme Tax Court (ret.), Germany. Florian Haase, Professor for German, European and International Tax Law and Tax Lawyer, IU International University, Bad Honnef/Rödl & Partner, Hamburg, Germany. Werner Haslehner, ATOZ Chair for European and International Tax Law, Department of Law, University of Luxembourg, Luxembourg. Marjaana Helminen, Professor of Comparative Tax Law, University of Helsinki, Finland. Sigrid Hemels, Professor of Tax Law, Erasmus School of Law, Erasmus University Rotterdam, Netherlands. Matthias Hofacker, Lawyer/ Tax Advisor, SOE Hofacker Rechtsanwaltsgesellschaft mbH, Germany Sunita Jogarajan, Professor, Melbourne Law School, University of Melbourne, Australia. Patrick Knörzer, Deputy Head, International Division, Fiscal Authority, Principality of Liechtenstein.

List of Contributors lxix Georg Kofler, Professor of International Tax Law, Vienna University of Economics and Business (WU Vienna), Austria. Polina Kouraleva-Cazals, Professor, University Savoie Mont Blanc, France. Gerhard Kraft, University Professor of Business Taxation, Department of Law and Economics, Martin-Luther-University, Germany. Eleonor Kristoffersson, Professor of Law, School of Behavioral, Social and Legal Sciences, Örebro University, Sweden. Fernando Souza de Man, Senior Tax Manager, Bolt, Netherlands. Patricia Lampreave Márquez, Lawyer. Accredited Tax Professor. Academic Expert of the EU Parliament. Former Senior Policy Officer on Fiscal State Aid, EU Commission, Stock Market Studies Institute. Georgios Matsos, Attorney-at-Law, Matsos & Associates Law Office, Greece. Yuri Matsubara, Professor of Tax Law, School of Commerce, Meiji University, Japan. Savina Mihaylova-Goleminova, Associate Professor, Faculty of Economics and Business Administration, Sofia University ‘St Kliment Ohridski’, Bulgaria. Aitor Navarro, Senior Research Fellow, Max Planck Institute for Tax Law and Public Finance, Germany. Xavier Oberson, Professor of Swiss and International Tax Law, University of Geneva, Switzerland. Christiana HJI Panayi, Professor in Tax Law, Queen Mary University of London, UK. Katerina Perrou, Postdoctoral Fellow, Law School, National and Kapodistrian University of Athens, Greece. Rainer Prokisch, Professor of International Tax Law, Faculty of Law, Maastricht University, Netherlands. Natalia Quiñones, Partner, Quiñones Cruz Abogados, Colombia. Akhilesh Ranjan, Advisor Tax Policy, Price Waterhouse & Co LLP, India. Isabelle Richelle, Professor of Tax Law, HEC Business School; Co-Chair of the Tax Institute of the University of Liège, University of Liège, Belgium. H. David Rosenbloom, Director, International Tax Program, New York University School of Law and Member, Capkin & Drysdale, Chartered, Washington, DC. Marilyne Sadowsky, Associate Professor, Sorbonne Tax & Public Finance Department, Sorbonne Law School—University Paris 1 Panthéon-Sorbonne, France. Roberto Scalia, Professor of Tax Law, University of Bergamo, Italy.

lxx List of Contributors Paloma Schwarz Martínez, Postdoctoral Researcher, University of Luxembourg, Luxembourg. Fadi Shaheen, Professor of Law and Professor Charles Davenport Scholar, Rutgers Law School, United States. Verônica Souza, International Tax Partner, Gaia Gaede Silva Advogados—Av. Pres. Juscelino Kubitschek, Brazil. Ton Stevens, Professor of Corporation Tax, Tilburg University, Netherlands. Miranda Stewart, Professor, Melbourne Law School, University of Melbourne, Australia. Servaas Van Thiel, Emeritus Professor International and European Law at the Free University Brussels (VUB), Free University Brussels (VUB), Belgium. Afton Titus, Associate Professor, Commercial Law Department, University of Cape Town, South Africa. Michael Tumpel, Professor of Business Administration, Head of the Institute of Tax Management of JKU Linz and Dean of JKU Business School, Johannes Kepler University Linz, Austria. Irma Mosquera Valderrama, Professor Tax Governance, Jean Monnet Chair Holder EU Tax Governance, Leiden Law School, Leiden University, Netherlands. Craig West, Associate Professor, Commerce Faculty, University of Cape Town, South Africa. Masao Yoshimura, Professor of Graduate School of Law, Hitotsubashi University, Japan.

Pa rt I

H I STORY A N D S C OP E OF I N T E R NAT IONA L TAX L AW

Chapter 1

The History of Internationa l Tax L aw Marilyne Sadowsky

1.1 Introduction ‘Say to Yasmah-Addu: thus speaks Aplahanda, your brother. This transport is mine; so do not bother my servitors with a tax story.’1 This was a letter given by Aplahanda (King of Karkamiš) to an Assyrian merchant carrying royal merchandise, to deliver it to Yasmah-Addu (King of Mari). It is the first evidence in Mesopotamian history of tax cooperation between kings, showing that taxation was already an obstacle to the free movement of goods and persons who traded and travelled between different kingdoms. This tax, called miksum, was applicable to goods in transit, whether international or local, and not to goods used by individuals for their own needs.2 At that time, there were royal incomes—‘the wealth of kings’—divided into several categories including five types of revenue from conquest, diplomacy, suzerainty, monarchy, and commercial activities.3 This tax was imposed ‘on harvests, the proceeds of the sale of harvests and of agricultural speculation, on herds, on goods brought in and transshipped’,4 as monarchy income.5 The reference to the notion of ‘speculation’ is reminiscent of the modern concept of capital gains and, indeed, today’s entrepreneurs were the ‘agrarian entrepreneurs’

1

J. M. Durand, ed., Documents épistolaires du palais de Mari, vol. III, LAPO 18 (Paris: Les éditions du Cerf, 2000), 25, no. 861. 2 Ibid., 26. 3 J. M. Sasson, ‘Chapter 1: Kingship’, in From the Mari Archives: An Anthology of Old Babylonian Letters (Winona Lake, IN: Eisenbrauns, 2015), 38ff. 4 M. D. Ellis, ‘Taxation in Ancient Mesopotamia: The History of the Term miksu’, Journal of Cuneiform Studies 26/4 (1974), 228. 5 Sasson, From the Mari Archives, 60.

4 Marilyne Sadowsky of the past.6 In other cases, the Assyrians also used taxation to control foreign trade by levying taxes to prohibit the export of certain goods by other kings to particular destinations,7 thus recalling that taxation is the result of a link between different international authorities. Even in ancient times, one can find examples of agreements linking sovereign authority with the right to levy taxes. During the Second Punic War between Carthage and Rome, some nobles from the Roman city of Tarentum helped the Carthaginian general Hannibal to capture their city in exchange for an agreement giving them complete freedom to retain their own laws, not to pay taxes, and not to have a Carthaginian garrison.8 Indeed, states did not hesitate to double taxes to fund the war effort—such as the Roman Republic with its double tributum9—or to broaden their citizenship in order to increase tax revenue.10 The same idea was used by Alexander the Great when he sent one of his Macedonian officers to take the satrapal capital of Dascyleion in order to instruct the local population to continue to pay their current Persian taxes and, on the contrary, to grant a remission to the city of Zelia because of a claim to Greek ancestry.11 The oldest study on the relief of double taxation dates back to the Middle Ages, when local taxes were established in French and Italian cities. One by one, glossators, canonists, and theologians began to find solutions to the problem. Did a woman from Carcassonne (a French city in the south of France) who had married in a neighbouring village have to pay a wealth tax (collecta) in both places? For the jurists of the time, the answer was no, because it was a case of double taxation. On the other hand, for Bishop Guillaume of Cuneo and the glossator Bartholomé the answer was in the affirmative because if the woman lived in the village but owned property in the town, she must be taxed in both places because of her ownership of the property and her personal obligation to pay the general wealth tax.12 These historical accounts reveal two things. First, and contrary to prevailing thought, international tax law is not a law born in the 1920s. International tax law has

6 G. Chambon, ‘Fiscal Regime and Management of Resources by the “King’s Household” in Mari during the Old Babylonian Period’, in J. Mynářová and S. Alivernini, eds, Economic Complexity in the Ancient Near East—Management of Resources and Taxation (Third–Second Millennium BC) (Prague: Czech Institute of Egyptology, 2020), 255. 7 A. Altman, ‘Tracing the Earliest Recorded Concepts of International Law—The Ancient Near East (2500–330 BCE)’, in R. Lesaffer, ed., Legal History Library: Studies in the History of International Law, vol. 8/4, (Leiden: Brill, Nijhoff, 2012), 180. 8 E. Owens, ‘The Second Punic War, 220– 202 BC’, in M. Whitby and H. Sidebottom, eds, The Encyclopedia of Ancient Battles, vol. II (Chichester: John Wiley & Sons, 2017), 743; J. S. Reid, ‘Problems of the Second Punic Wars: III, Rome and Her Italian Allies’, Journal of Roman Studies 5 (1915), 98. 9 J. Tan, ‘The Roman Republic’, in A. Monson and W. Scheidel, eds, Fiscal Regimes and the Political Economy of Premodern States (Cambridge: Cambridge University Press, 2015), 283. 10 G. F. Bransbourg, ‘Fiscalité et enjeux de pouvoir dans le monde romain’, PhD thesis (2010), 9. 11 I. Worthington, ‘Campaigns of Alexander the Great, 336– 323 bc’, in Whitby and Sidebottom, Encyclopedia of Ancient Battles, 20. 12 E. R. A. Seligman, ‘La double imposition et la coopération fiscale internationale’, in Recueil des Cours de l’Académie de Droit international 1927 V, vol. 20, (Paris: Librairie Hachette, 1929), 488ff.

The History of International Tax Law 5 always existed. Tax law and states are inseparable, as it is unlikely that a state—from its premodern form to its postmodern one—would exist without an easily assessable and accessible basic element that could serve as the foundation for appropriation and that could easily be taxed, such as grain.13 Secondly, the concepts we know today in international tax law are not new. They have gradually evolved over time and across countries to take account of the different functions of taxation. International tax law is rooted in economic, political, social, cultural, and philosophical history. Thus, the purposes of levies differ according to the ideals promoted in the country at the time of the levy.14 In practice, material evidence of very early tax history is difficult to gather because the reading and interpretation of its historical source (a stone stele) depends on Assyriologists, which is how the letters of Mari were deciphered.15 That history will be explored in this chapter as a quest for truth in order to understand what international taxation16 was in the past and what it is today. In fact, this is the primary function of archives, which are consulted ‘to try to understand the current position when seeking policy and drafting solutions to current issues’.17 Modernity challenges the original foundations of international tax law and raises the question: what is international tax law? It is possible to answer this question by showing that international taxation did not come into being in the 1920s, at the time of the League of Nations, but rather is the result of a historical construction from ancient times. In fact, this historical approach reflects the meaning of international tax law as discussed in modern tax literature. Many countries distinguish between two categories of rules: domestic tax rules, including foreign elements, and international tax rules applicable to tax matters.18 For some contemporary authors, the concept of ‘transnational’ income give an overview of these two categories.19 In effect, domestic tax rules reflect a permanence over time of national tax systems with foreign elements but without any harmonization. On the other hand, international tax rules reflect the idea of a mutation, aiming to determine whether the repetition of old rules, practices, and procedures has given way to certain common mandatory rules applicable to each country in matters of taxation. In 1955, the question was asked: are there some common mandatory rules in tax matters applicable to each country? Maxime Chrétien replied that a ‘common international tax law’,

13

J. C. Scott, Against the Grain—A Deep History of the Earliest States (New Haven, CT: Yale University Press, 2017), 22. 14 J. Snape, ‘The Sinews of the State: Historical Justifications for Taxes and Tax Law’, in M. Bhandri, ed., Philosophical Foundations of Tax Law (Oxford: Oxford University Press, 2017), 9–33. 15 I would like to thank M. Mickaël Guichard, Director of Studies in History and Philology of Mesopotamia at the Ecole pratique des Hautes études, for his help on this part of the research. 16 The term ‘international taxation’ will be used here in the same sense as ‘international tax law’. 17 R. J. Vann, ‘Writing Tax Treaty History’, Sydney Law School Research Paper no. 10/ 19 (March 2011), 13. 18 Examples can be given from the German, Anglo-American, and Italian doctrines. See A. Kallergis, La compétence fiscale (Paris: Editions Dalloz, coll. Nouvelle Bibliothèque de Thèses, 2018), 4ff. 19 E. Reimer, ‘Transnationales Steuerrecht’, in C. Möllers, ed., Internationales Verwaltungsrecht—eine Analyse anhand von Referenzgebieten (Tübingen: Mohr Siebeck, 2007), 181–182.

6 Marilyne Sadowsky understood as containing certain rules binding on all states, was still in its infancy.20 The passage of time is always perceived, as it was for Thucydides in the past, as an oscillation between ‘always’ and ‘change’.21 Both sides of the same coin can be found in the historical nature of jus gentium. In Roman law, jus gentium was applicable to non-Latin subjects of Rome (peregrini); foreigners were excluded from the jus civile reserved for Roman citizens22 and what today we call domestic law. Then, in the sixteenth century, a shift towards public international law was possible thanks to the transformation of jus gentium into jus inter gentes, a set of legal rules governing relations between states23 and what we today call international law. Thus, international tax law naturally follows this evolution from ancient to modern history, from miksum to the multilateral instrument (MLI), thus revealing its permanence (discussed in Section 1.2) and its mutation (Section 1.3).

1.2 Permanence of International Tax Law Two main reasons can explain the permanence of international tax law. First, throughout history it has always been necessary to link the tax authority—the state, or any other form of political power—and the wealth produced by land or activities, traded or distributed within or outside territorial boundaries. The contribution to public expenses by the taxation of a person or matter through a taxable nexus had a very ancient birth (see Section 1.2.1). Secondly, the rhythm of history teaches us that modernity is often the repetition of old facts. Modern international tax law is necessarily the result of ancient international tax law. Like Thucydides in his time, historians maintain that since human nature does not change, the same types of events can recur at any time.24 In 1905, Santayana make this a modern aphorism: ‘Those who cannot remember the past are condemned to repeat it.’25 In other words, the past teaches us lessons ‘for present and future action’.26 Thus, in international tax law there is a series of phenomena

20

M. Chrétien, A la recherche du droit international fiscal commun (Paris: Recueil Sirey, 1955), 242. P. Vidal- Naquet, ‘Temps des dieux et temps des hommes— Essai sur quelques aspects de l’expérience temporelle chez les Grecs’, Revue de l’histoire de religions 157/1 (1960), 69. 22 A. Truyol v Serra, ‘Théorie du droit international public—Cours général’, in Recueil des Cours 1981 IV—Collected Courses of the Hague Academy of International Law, vol. 173 (Dordrecht: Martinus Nijhoff, 1992), 33. 23 Ibid., 36. 24 J. de Romilly, ‘Thucydide et l’idée de progrès, Annali della Scuola Normale Superiore di Pisa’, Lettere, Storia e Filosofia, Serie II, 35/3/4 (1966), 176–177. 25 G. Santayana, ‘Chapter XII: Flux and Constancy in Human Nature’, in The Life of Reason or the Phases of Human Progress, vol. 1 (New York: Charles Scribner’s Sons, 1906), 284. 26 G. W. Trompf, The Idea of Historical Recurrence in Western Thought: From Antiquity to the Reformation (Berkeley, CA: University of California Press, 1979), 3. 21

The History of International Tax Law 7 that constantly renew themselves in an immutable order, what are here called historic recurrences (Section 1.2.2).

1.2.1 A Very Ancient Birth During the age of antiquity, certain tax rules were applied to foreigners, showing that tax liability—the nexus—has changed over the time. The analysis of the Wilbour Papyrus (1147 bce) shows the importance of taxes in ancient Egypt, the role of scribes—untaxed because of their writing work—and the application of taxes to foreigners.27 The scribe was the main accountant and tax collector who recorded on papyrus details of the amounts collected as taxes and who indicated how those amounts were disposed of on arrival in Thebes. The famous scribe of the Dhutmose Necropolis received sacks of corn from ‘foreigners’ and collected taxes in different places (cities, temples, boats) with the help of ‘janitors’ or ‘door-keepers’ able to use such coercion as might be required.28 At that time, and this is also true for the even older Mari Kingdom period, a foreigner seems to have been a person not necessarily linked to the land producing the harvest, but who brought sacks of corn to be taxed at the place where the land was located. There was a disconnection between people and wealth, essentially because the income was the income of the ‘King’s Household’ which included the royal family, servants, and vassals.29 The owner of the land was less important than the person who granted the right to cultivate it and who would benefit from the harvest.30 A study of the Rosetta Stone (196 bce) reveals the will to fight against fiscal pressure by granting to the population, the army, and the temples of Egypt certain tax remissions or reductions. Profits for the army are not specified on the stone, unlike those offered to the temples, as sources of income, with a reduction of the ‘artaba-tax’ for cultivated land and of the ‘ceramion-tax’ for vineyards, and to the population.31 These privileges created a permanent decrease in royal revenues and caused discontent among the population. Egypt’s economic degradation was accelerated by difficulties in tax collection and the accumulation of arrears. The Sarapeum papyri provides an illustration of the system of compulsion in the fiscal domain and of the transmission of fiscal pressure from the tax-farmers responsible for tax collection to the taxpayers.32 It is interesting to note that Ptolemies permitted ‘a wide breach to be made in their system of state control’ by

27 A.

H. Gardiner, ‘Ramesside Texts Relating to the Taxation and Transport of Corn’, Journal of Egyptian Archaeology 27 (Dec. 1941), 19–73. 28 Ibid., 20ff. 29 Chambon, ‘Fiscal Regime’, 252. 30 Ibid., 254. 31 M. Rostovtzeff, The Social and Economic History of the Hellenistic World, vol. 2 (Oxford: Oxford University Press, 1941), 713ff. 32 Ibid., 724–725.

8 Marilyne Sadowsky granting more economic freedom mainly to native farmers and to a lesser extent to foreign mercenaries.33 Finally, during the reign of Augustus and his Julio-Claudian successors,34 fiscal distinctions were made between some groups of citizens, from the most favoured to the least favoured: (1) Roman citizens (the dominant elite) benefited from exemption from the payment of poll tax and other capitation taxes levied on males between the ages of 14 and 62; (2) citizens from Greek cities (Alexandria, Naukratis in the Delta, and Ptolemaisin Upper Egypt) benefited from exemption from poll tax and the privilege of continuing to use Greek civic institutions; (3) those with ‘metropolite citizenship’ were less privileged and paid capitation taxes at a half-rate; (4) and members of the ‘foreign’ politeumata were taxed at the highest rate, as Egyptians peasants.35 One of the oldest and most important politeumata was that of the Jews,36 as well as other groups including Phrygians, Cretan, Cilician, Boeotian, or Idumaean politeumata. Even today, the nature and composition of politeumata is discussed in the historical literature from the perspective of its characteristics—ethnic, military, religious, or immigration. For some authors, the recognition of this group as a single political organization ensures that the Ptolemaic kingdom remains an attractive country of residence.37 A distinction is made between citizenship and the status of foreigners, independently of the place of residence. On the basis of these early historical examples of taxation, a population is taxed on its products or labour and the revenue is attributed to a person—a king, a pharaoh, or an emperor—who exercises coercive power over a territory. The tax burden depends on the category of the population. In addition, the distinction between natives, citizens, and foreigners leads to different taxation and the nexus has evolved from a personal link without any reference to a city or a political entity, to a political one. At that time, the status of being a foreigner implied a fiscal constraint without specific rights. Gradually, and in order to facilitate trade relations, minimum rights were granted to foreigners by Roman law, particularly in the area of contracts.38 During the Middle Ages, the first trade treaties were created to guarantee some privileges between the citizens of

33

Ibid., 728ff. A. E. Hanson, ‘Egyptians, Greeks, Romans, Arabes, and Ioudaioi in the First Century A.D. Tax Archive from Philadelphia: P. Mich. inv. 880 recto and P. Princ. III 152 revised’, in J. Johnson, ed., Life in a Multi-Cultural Society: Egypt from Cambyses to Constantine and Beyond (Chicago: The Oriental Institute, 1992), 133–145. 35 Ibid., 133–134. 36 S. Honigman, ‘ “Politeumata” and Ethnicity in Ptolemaic and Roman Egypt’, Ancient Society 33 (2003), 61–102; M. Avi-Yonah, ‘Historical Geography of Palestine’, in S. Safrai and M. Stern, eds, The Jewish People in the First Century, vol. 1 (Assen: Van Gorcum & Comp. BV, 1974), 109. 37 P. Stränger, ‘The Politeuma in the Hellenistic World (Third to First Century B.C.): A Form of Organisation to Integrate Minorities’, in J. Dahlvik, W. Sievers, and C. Reinprecht, eds, Migration und Integration—wissenschaftliche Perspektiven aus Österreich, Jahrbuch 2/2013 (Göttingen: V & R Unipress, 2014), 64. 38 E. Cuq, Manuel des institutions juridiques des romains, 2nd ed. (Paris: Librairie Plon, LGDJ, 1928), 14. 34

The History of International Tax Law 9 Genoa and Narbonne, notably to prevent tax abuses.39 According to medieval law, the status of foreigners depended on various criteria—ethnic, political, economic, and religious. Its definition was based on the distinction between residence and domicile, far from the concept of origin derived from Roman law, and the children of foreigners were considered as citizens thanks to the jus soli.40 During the Ancien Régime, in European countries (France, England, Spain, and Hungary) an important source of revenue for the king was given by the ‘droit d’aubaine’; that is, the right of the king to seize the estates of foreigners who died in the kingdom without French heirs by birth.41 Even though the philosophers and economists of the Enlightenment denounced such a ‘barbarian law’, it was not definitively abolished until the nineteenth century. This development coincides with the emergence of the concept of nationality and jus sanguinis in France and the complete assimilation of foreigners and nationals in terms of private rights. At the same time, Georg von Schanz found the concept of economic allegiance42 and its consequences related to non-discriminatory treatment, personal or real subordination, and source or residence taxation. In his view, economic criteria must determine where a taxpayer must pay income taxes, so that anyone who benefits from public services will also contribute to their financing.43 His preference was for taxation of income at source rather than taxation in the country of residence. Under the impetus of Seligman,44 the League of Nations clearly advocated this system in its report of 1923, setting out four fundamental considerations of wealth: acquisition (place of origin), location (situs), enforceability of rights (place of enforcement of legal rights), and consumption (residence or domicile).45 Thus, the needs of states are equally as important as the duties of taxpayers, and the elimination of double taxation is the foundation of the 1920 compromise.

1.2.2 Historic Recurrences Fiscal sociology teaches us that the particular form of a tax system lies in economic development, which brings democracy.46 European economic history describes four 39 B. Nolde, ‘Droit et techniques des Traités de Commerce’, in Recueil des Cours 1924 II, vol. 3 (Paris: Librairie Hachette, 1925), 300. 40 L. Mayali, ‘Étranger’, in C. Gauvard, A. de Libera, and M. Zink, eds, Dictionnaire du Moyen-Âge (Paris: PUF, 2002), 499–500. 41 P. Sahlins, ‘Sur la citoyenneté et le droit d’aubaine à l’époque modern—Réponse à Simona Cerutti’, Annales Histoire, Sciences Sociales 63/2 (2008), 385–386. 42 G. V. Schanz, ‘Zur Frage der Steuerpflicht’, FinanzArchiv/Public Finance Analysis 9 (1892), 365–438. 43 Ibid., 372. 44 Seligman rejected the ‘socialistic theory of progression’ defended by Adolph Wagner, which favoured residence taxation on a worldwide tax base. On the history of this, see L. Cavelti, International Tax Cooperation: The Sovereignty Conflict between the Residence and the Source Country (Bern: Stämpfli Verlag AG, 2016), SSW Band 26, 18ff. 45 G. Bruins et al., Report on Double Taxation, League of Nations Economic and Financial Commission, Doc. E.F.S.73.F.19 (Geneva: League of Nations, 1923), 23. 46 I. W. Martin, A. K. Mehrotra, and M. Prasa, eds, ‘The Thunder of History: The Origins and Development of the New Fiscal Sociology’, in The New Fiscal Sociology—Taxation in Comparative and Historical Perspectives (Cambridge: Cambridge University Press, 2009), 7ff.

10 Marilyne Sadowsky long economic ‘waves’ of price revolutions: the later medieval, the Renaissance, the Enlightenment, and the Victorian era.47 These waves have some similarities48 with US economic history:49 prosperity, political disorder, inflation, fiscal stresses, increasing inequality regarding wealth and income, social disruption, revolution, wars, and equilibrium. In fact, taxation is an important part of the rhythm of history, and we can find in international tax law some historic recurrences showing that tax history repeats itself. First, most revolutions were caused by fiscal problems. During the Renaissance, this was the case in Europe for: Iberia (1640) with Catalonia and Portugal due to the increase of revenue by Spanish ministers; France (1648–1654) with the Fronde due to increased fiscal pressure to finance the Thirty Years War, but also Palermo (1647), England, Italy (1647), and Denmark (1660).50 The same events were repeated in these European countries and the USA during the Enlightenment because of growing inequality. Indeed, new taxes created to maintain the solvency of states ‘fell heavily on those who were least able to bear them’.51 Whilst the rich were exempt from taxes or could reduce their tax obligations by paying a lump sum, in the countryside the peasants began to revolt. In the eighteenth century, the British colonies of North America, as with the French during the Revolution, revolted against an arbitrary and unfair tax system. The idea was not to abolish taxes but to give them legitimacy through the consent of the people.52 An important step towards the American Revolution was the Stamp Act (1765) passed by the British government to pay off debts caused by the end of the Seven Years War. This tax on all printed materials for commercial and legal use triggered a strong protest in the colonies and raised the question of taxation without representation; especially since the measure was accompanied by other colonial taxes enacted almost at the same time: the Sugar Act and the Currency Act in 1764, the Quartering Act in 1765, and the Townshend Act in 1767.53 Refusing to pay a tax on tea, the ‘Sons of liberty’ led by Samuel Adams and John Hancock dumped 342 chests of tea, imported by the British East India Company, into the harbour. The Boston Tea Party caused a fall of 90% in the importation of tea from Britain with a consequent decline in the receipt of duties54 which was later followed by the American Revolution. Nowadays, taxation continues to cause popular discontent. In France, the ‘red caps’ movement against the eco tax on

47

D. Hackett Fischer, The Great Wave: Price Revolutions and the Rhythm of History (Oxford: Oxford University Press, 1996), 5. 48 Ibid., 237ff. 49 See, e.g., the changes that occurred in the nineteenth century alone: R. E. Gallman, ‘Economic Growth and Structural Change in the Long Nineteenth Century’, in L. Engerman and R. E. Gallman, eds, The Cambridge Economic History of the United States, vol. 2 (Cambridge: Cambridge University Press, 2000), 1–55. 50 Hackett Fischer, Great Wave, 97–100. 51 Ibid., 138ff. 52 N. Delalande, Les batailles de l’impôt—Consentement et résistances de 1789 à nos jours (Paris: Seuil, 2014), 23. 53 A. Rabushka, Taxation in Colonial America (Princeton, NJ: Princeton University Press, 2010), 749ff. 54 Ibid., 758.

The History of International Tax Law 11 heavy goods vehicles (2013) and the ‘yellow vests’ protests against carbon tax (2018) were directly linked to tax justice and the acceptability of taxes by the taxpayer. Secondly, the history of the state is closely linked to war and raises the question of specific wartime tax rules. The consent of the taxpayer depends on ‘those aspects of a nation’s history that determine the collective willingness to sacrifice’.55 Throughout history, wars have caused economic upheaval and a large part of public spending is needed to fund them. Times of war are characterized by the impossibility for a state to effectively control its economic activities and the need for constant political decision making, both of which contribute to the perpetuation of taxes. Wars are a great consumer of revenue thus calling for taxation, with three resulting problems to resolve: finding the money to fund the cost of the war, taxation during the war, and dealing with the postwar debt crisis. For example, in ancient Egypt, extraordinary levies on temple wealth were created to finance the wars of Saite pharaohs from 664 to 525 bce.56 In the Middle Ages, to support the armies in Gascony and Flanders, Edward I used various means such as annual direct taxes on income, cash subsidies, export duties on merchants (‘maltote’), and bank loans.57 The same occurred in later periods, as in the example of France and Great Britain during the Napoleonic Wars where the fiscal means to support the war depended on the state.58 The period of the First World War and Second World War also reveals its specificities. In 1915, a new ‘war profits tax’ was proposed in both Denmark and Sweden aimed at the enormous profits made through exports. Gradually59 many countries introduced such a tax: in 1915 in Italy, Germany, and Austria; in 1916 in Russia, Canada, the Netherlands, France, Spain, New Zealand, Switzerland, and the USA; and in 1917 in Australia. A war profits tax is a business tax imposing a ‘normal standard’ above which the excess is measured and taxed. Originally, this meant a ‘tax on excess profits levied during the war’ to be extended to ‘excess profits’, in which the ability of the taxpayer was measured by the privileges enjoyed.60 Historically, the excess profits tax has been the most important type of war wealth.61 The idea of a tax levied on something abnormal, with the intent of giving the public a share in the gains of profiteering as something transitory or undesirable, is interesting.62 Indeed, this type of fiscal war rule aims

55 N. Feldman and J. Slemrod, ‘War and Taxation: When Does Patriotism Overcome the Free-Rider Impulse?’, in Martin, Mehrotra, and Prasa, New Fiscal Sociology, 138. 56 M. Jursa and J. C. Moreno García, ‘The Ancient Near East and Egypt’, in Monson and Scheidel, Fiscal Regimes and the Political Economy of Premodern States, 141. 57 E. Miller, ‘War, Taxation and the English Economy in the Late Thirteenth and Early Fourteenth Centuries’, in J. M. Winter, ed., War and Economic Development: Essays in Memory of David Joslin (Cambridge: Cambridge University Press, 1975), 11ff. 58 M. D. Bordo and E. N. White, ‘A Tale of Two Currencies: British and French Finance During the Napoleonic Wars’, Journal of Economic History 51/2 (1991), 303–316. 59 J. C. Stamp, ‘The Taxation of Excess Profits Abroad’, Economic Journal 27/105 (1917), 28. 60 E. R. A. Seligman, ‘The War Revenue Act’, Political Science Quarterly 33/1 (1918), 26–27. 61 J. R. Hicks, U. K. Hicks, and L. Rostas, The Taxation of War Wealth (Oxford: Oxford University Press, 1942), 29. 62 C. C. Plehn, ‘War Profits and Excess Profit Taxes’, American Economic Review 10/2 (1920), 283.

12 Marilyne Sadowsky to counter a temporary economic crisis. Consequently, it would be possible to apply a similar taxation regime to the situation caused by the Covid-19 pandemic. We are facing a temporary economic crisis which offers privileges to some companies benefiting from the abnormal situation. Some authors have already theorized this ‘modern excess profits tax’ for companies that have benefited from the pandemic.63 This logic is also found in Pillar One market jurisdictions, as amount A provides for taxation of the fraction of a ‘residual’ profit, considered after crossing a break-even point, and calculated at the group level. Finally, the implementation of taxation is sometimes the result of a phenomenon of mimicry in which the behaviour of one state is reproduced, with some of its tax experiences or practices being copied by others. In comparative tax law, the influence of foreign tax systems on the construction of national tax systems can be observed through different techniques of transplantation or rejection. In early states, the subsistence bases bear a ‘remarkable resemblance’ to one another in Mesopotamia, Egypt, the Indus Valley, and Yellow River where grain units were used as a basis for the taxation of trade and tribute, including labour. Grains are ‘visible, divisible, assessable, storable, transportable, and rationable’, so that they can easily serve as a basis for taxation for states.64 Later, with the advent of coins, the idea remained of taxing land, trade, and labour. The roots of income tax as an annual income tax can be found in the form of ‘Danegeld’ from 1012 onwards. This was an English land tax assessed according to a fiscal unit called a hide; that is, a household characterized by an area of agricultural land sufficient to support a peasant family.65 Consequently, the introduction in 1799 by the British Prime Minister William Pitt of a ‘modern’ income tax is unsurprising and was ‘influenced by the tax laws of other European countries and, perhaps, its colonies’.66 Along with Italy,67 France was one of the leading countries in Europe. Indeed, in the Middle Ages, France introduced the taille,68 a tax levied on the peasants of the king’s domain to give them seigneurial protection which was extended to a royal taille to buy back military service.69 Under the Ancien Régime, the tax evolved into two separate taxes: (1) a personal income tax in the north of France (taille personnelle) taking into account all of a person’s

63 R. S. Avi-Yonah, ‘Covid-19 and US Tax Policy: What Needs to Change?’, Intertax 48/8/9 (2020), 790–793. 64 Scott, Against the Grain, 129ff. 65 B. A. Abraham, ‘ “Danegeld”—From Danish Tribute to English Land Tax: The Evolution of Danegeld from 991 to 1086’, in J. Tiley, ed., Studies in the History of Tax Law, vol. 6 (Oxford: Oxford University Press, 2013), 277. Originally, it was a payment made by English kings to the Viking invaders as the price for peace. 66 P. Harris, Income Tax in Common Law Jurisdictions— From the Origins to 1820, vol. 1 (Cambridge: Cambridge University Press, 2006), 1. 67 E. R. A. Seligman, The Income Tax: A Study of the History, Theory, and Practice of Income Taxation at Home and Abroad, 2nd ed. (New York: Macmillan, 1914), 338ff. 68 In practice, this was a wooden stick with notches to provide numerical evidence of value and was used for credit payments. 69 J. Favier, ‘Taille’, in Gauvard, de Libera, and Zink, Dictionnaire du Moyen-Âge, 1365–1366.

The History of International Tax Law 13 activities and their situation (number of children, charges, age); (2) a real tax on land in the south of France (taille réelle) taking into account the cadastral value of properties.70 Later, Louis XIV enacted the dixième, a general income tax of 10% on all incomes throughout the country. Towards the end of the nineteenth or beginning of the twentieth century, most countries adopted a ‘modern’ income tax with a few exceptions.71 Other example of the behaviour of one state being reproduced in other states are the concepts of permanent establishment (PE) and residence and the influence of the court jurisdiction and European thinking on British legislation.72 Modern examples of this mimicry effect can be found with digital tax.

1.3 Mutation of International Tax Law International tax law is a story of change. There are two reasons for this. The first is that international tax law is the reflection of one point in time. Thus, the core international tax principles have been shaped by time and the meaning of well-known concepts or even words has evolved to adapt to the times (Section 1.3.1). The second is again linked to the rhythm of history. The cyclical vision of the Hellenistic time must be rejected here because there is no determinism in the evolution of states’ relationships,73 therefore it is necessary to go beyond that vision to make progress (Section 1.2.2).

1.3.1 Adapting to the Times The history of US international taxation has been divided into four ages: benefits (1918– 1960); neutrality (1961–1980); competition (1981–1997); and cooperation (since 1998).74 Each of these periods is characterized by a predominant element and a back and forth between source-based and residence-based taxation giving some continuity to US fiscal policy: the right to tax (age 1), capital export neutrality (age 2), preservation of the competitiveness of the US economy (age 3), and the coordination of residence and source taxation to prevent double taxation and double non-taxation (age 4).75 There are two interesting arguments in this respect. The first is that international tax policy has 70

M. Touzery, ‘Taille’, in L. Bély, ed., Dictionnaire de l’Ancien Régime (Paris: PUF, 2018), 1200. V. Thuronyi, K. Brooks, and B. Kolozs, Comparative Tax Law, 2nd ed. (Dordrecht: Wolters Kluwer, 2016), 209. 72 J. F. Avery Jones, ‘Jurisdiction to Tax Companies: The Influences of the Jurisdiction of the Courts and of European Thinking’, in J. Tiley, ed., Studies in the History of Tax Law, vol. 4 (Oxford: Hart Publishing, 2010), 163–212. 73 R. Aron, Paix et guerre entre les nations (Domont: Calmann-Lévy, 2004), 145ff. 74 R. S. Avi-Yonah, ‘All of a Piece Throughout: The Four Ages of U.S. International Taxation’, Virginia Tax Review 25/2 (2005), 313–338. 75 Ibid., 315–317. 71

14 Marilyne Sadowsky different objectives at different times. Our current concerns are not those of the past, such as global warming or the consequences of the digital economy. The second is the construction of international tax law as a reaction to historical realities and adaptations to economic, political, social, and technical developments. At different times, international tax law has reflected the difficulties encountered by states in the establishment and allocation of taxes. The fundamental principles of international tax law have followed the changing concepts of person and wealth. According to Seligman,76 the evolution of wealth takes place in four stages: (1) apparition (production/creation); (2) property (producer/ slaver); (3) exchange (transactions); and (4) consumption. The concept of wealth has gradually evolved to take account of the person. In the past, it was possible to own a person, and tax incentives were given for the import of ‘good’ slaves.77 Later, during the Sierra Leone Colony and Protectorate, opponents of the hut tax pointed out that ‘the prosperity of West African Colonies largely depends upon the amount of capital and energy which they, the Merchants, put into them’.78 Thanks to these revolts, the individual considered as an object of wealth becomes a subject of wealth, from colonization to civilization. Thus, it was possible to make the link between nationality, domicile, residence of the individual, and place of ownership, the source of income or the possession of income, payment of profits, enjoyment, and expenditure of income.79 These links and the concepts they connect evolved over time. Economic progress, globalization, and new technologies have changed the way taxpayers live. In the past, there was no need to distinguish different places of production, ownership, or consumption because taxation was carried out at the location of the source of wealth; that is, in the place where the person was born, lived, and worked, as the words ‘place, site and residence’ were synonymous.80 Nowadays, individuals and companies are located in different places, assets are intangibles, and the wealth is moved from property to different markets in different states where individuals and companies actively participate in economic growth. The challenge is to distinguish the individual or the group from the wealth and to establish where the income is taxed. Thus, economic and political developments led to changes in certain concepts, from citizenship to nationality,81 from tax sovereignty to tax jurisdictions, from states to market states, from foreigners to consumers, from bilateral to multilateral relations, or from the allocation of income to the fight against tax evasion. In fact, it can be interpreted as a paradigm shift between the moment of the design of international tax rules and their application in a modern context, as a choice made by policymakers or legislators at a specific point

76

Seligman, ‘Double imposition et la coopération fiscale internationale’, 477ff. K. Outterson, ‘Slave Taxes’, in J. Tiley, ed., Studies in the History of Tax Law, vol. 1 (Oxford: Oxford University Press, 2004), 271–272. 78 E. D. Morel, The Sierra Leone Hut-Tax Disturbances (Liverpool: John Richardson & Sons, 1899), 21. 79 Seligman, ‘Double et la coopération fiscale internationale’, 487. 80 Ibid., 482ff. 81 Ibid., 466. For Seligman, nationality is the result of capitalism. 77

The History of International Tax Law 15 in time. This temporality has been well demonstrated with respect to the way international income is taxed,82 and there is a distinction to make depending on the nature of the concept—legal, economic, or both. Indeed, a legal concept such as the source of income which is not ‘a well-defined economic idea’,83 may evolve in a way that does not necessarily follow economic developments in relation to a legal concept built on an economic concept, such as the economic standard of the arm’s-length principle. The concepts of international tax law were born in cultural and economic contexts that no longer exist today. Originally, the PE was a non-tax law concept, created in 1845 under the Prussian general regulation on enterprises (‘Betriebsstätte’) representing the total space used for conducting a business activity.84 Then a similar concept of fixed trade was used (‘Stehendes Gewerbe’) by reference to a Prussian trade tax introduced in 1810 to designate trade activities that did not fall under the concept of itinerant trade, although the term was not defined in the Prussian–Saxony double tax convention of 1869.85 Swiss inter-cantonal case law also used the concept of a branch in the sense of private or commercial law from 1875 onwards.86 The concept of PE was finally taken up in tax law in 188587 and used in a more modern sense from 1891 with the idea of subordination of other business undertakings—branches or places for purchasing.88 Subsequently, the concept of PE was adopted in tax treaties and has evolved since 1899 within the history of tax treaties characterized by a shift from source-state to residence- state taxation and is now looking for a new basis since the 2017 base erosion and profit shifting (BEPS) changes.89 The emergence of PE as a tax concept coincides with the second industrial revolution, where capital shifted from circulating/working capital to investment/fixed capital and the factors of production were relatively immobile. At that time, with the exception of specific rules applicable to international shipping, physical presence was the principle, and only rarely gave rise to controversy between the source and the residence states. Nowadays, factors of production as well as labour are mobile, and some companies can create wealth without a physical presence in a state. The digital economy has transformed this traditional pattern of a single source of income into a

82 Michael

J. Graetz, ‘The David R. Tillinghast Lecture Taxing International Income: Inadequate Principles, Outdated Concepts, and Unsatisfactory Policies’, Tax Law Review 54/3 (2001), 261–336. 83 H. J. Ault and D. F. Bradford, ‘Taxing International Income: An Analysis of the U.S. System and Its Economic Premises’, in A. Razin and J. Slemrod, eds, Taxation in the Global Economy (Chicago: University of Chicago Press, 1990), 26. 84 A. A. Skaar, Permanent Establishment: Erosion of a Tax Treaty Principle, 2nd ed. (Alphen aan den Rijn: Wolters Kluwer, 2020), 76. 85 M. Castelon, International Taxation of Income from Services Under Double Taxation Conventions: Development, Practice and Policy (Alphen aan den Rijn: Wolters Kluwer, 2016), 34ff. 86 J. F. Avery Jones, ‘Jurisdiction to Tax Companies: The Influences of the Jurisdiction of the Courts and of European Thinking’, in J. Tiley, ed., Studies in the History of Tax Law, vol. 4 (Oxford: Oxford University Press, 2010), 204 fn. 199. 87 J. Avery Jones et al., ‘The Origins of Concepts and Expressions Used in the OECD Model and their Adoption by States’, British Tax Review 6 (2006), 722. 88 Skaar, Permanent Establishment, 78. 89 Ibid., 105–106.

16 Marilyne Sadowsky plural source of income, and the use of the Heckscher–Ohlin theory of trade has been proposed to resolve these new problems.90 Rather than the location of the taxpayer, the location of the consumer would better serve as a basis for taxation rights. This change also affects the links between states and their international relations in the field of taxation. Even in the earliest international treaties, there were tax clauses regulating different aspects of trade.91 For many years, each state freely chose which taxes it levied without regard to the actions of other states. A state’s tax freedom can only be curtailed if it so decides with respect to another contracting party, by expressing its will through a treaty. The increase in the volume of cross-border business and changes in the structure and form of such businesses affect the interaction of states, and consequently their international tax relations.92 Certain fundamental concepts of international tax law originated from the history of these international relations. For example, the most- favoured-nation clause predates the very concept of trade treaties and derives from European international law in the Middle Ages. The clause developed from the privileges granted by medieval rulers to foreign traders, reflecting the idea of parity of rights with other foreign states. The earliest example dates back to 1055, when Henry III granted the city of Mantua all the customs privileges granted ‘by any city whatsoever’.93 Two other examples are recorded. The privileges of the people of Marseille obtained from the King of Jerusalem, Guy de Lusignan, in 1190, gave them the right to export the same goods as other nations, and the same privileges as the people of Montpellier. Emperor Frederick II, in 1226, also granted the people of Marseille the same rights as the citizens of Pisa and Genoa.94 At this stage, the aim was to stimulate trade by granting protection in general terms without any reference to the treatment of the state’s own nationals.95 Since the fifteenth century, the principle of non-discrimination has been shaped by trade treaties and, from the sixteenth century, nationality clauses were included.96 The clause gradually moved from the most-favoured clause to a non-discrimination clause, and finally amounted to treatment ‘on the same footing as nationals’.97 Thus, the development of international relations through the conclusion of international agreements and the will of states to develop and strengthen their economic relations led to changes in the

90

Ibid., 69ff, particularly 72 fn. 12. Lesaffer, ‘Treaties within the History of International Law’, in M. Bowman and D. Kritsiotis, eds, Conceptual and Contextual Perspectives on the Modern Law of Treaties (Cambridge: Cambridge University Press, 2018), 73. 92 D. Ring, ‘International Tax Relations: Theory and Implications’, Tax Law Review 60/ 2 (2007), 83–154. 93 B. Nolde, ‘La clause de la nation la plus favorisée et les tarifs préférentiels’, in Recueil des Cours 1932 I, vol. 39 (Paris: Librairie du Recueil Sirey, 1932), 25. 94 Ibid. 95 N. Bammens, ‘The Principle of Non-Discrimination in International and European Tax Law’, IBFD Doctoral Series, vol. 24 (2012), 33–4. 96 K. Van Raad, Nondiscrimination in International Tax Law (Deventer: Kluwer Law and Taxation, 1986), 34. 97 Bammens, ‘Principle of Non-Discrimination in International and European Tax Law’, 34. 91 R.

The History of International Tax Law 17 protection of taxpayers. Some efforts have been made by states to design and implement international tax policies to address historic realities.

1.3.2 Making Progress The question is always the same: where and how should income be taxed? Depending on the times, the answers will be different. Today, international tax law is in crisis because it has lost its traditional landmarks, making it impossible to answer the question. To evolve, we have two possibilities: to ignore history, with the subsequent risk of repeating the same effects, or to take history into account in order to move forwards, to progress. Much progress has been made, particularly in conventional fields, thanks to the diplomatic compromise of 1920. After the First World War, the free movement of capital was established as one of the conditions for global economic reconstruction. That is why double taxation and tax evasion were at the heart of the recommendations made by the representatives of the International Economic Conference in Genoa in 1922. Tax treaties appeared as a technique for normalizing relations: vertically between the state and a natural or legal person that depends on it politically, economically, and socially; and horizontally between states to normalize their relations in order to effect cooperation. Initially, the aim was mainly to combat double taxation and to exchange information. The early bilateral tax treaties date back to the nineteenth century and are based on the principle of reciprocity in order to reduce double taxation.98 Nevertheless, as mentioned in the introduction, the phenomenon of double taxation is not new. Unlike Bishop Guillaume of Cuneo, Petrus Antibolus in 1350 reduced the distinction between general and local taxes, focused on domicile, and proposed the taxation of personal property instead of its location when the location was also the birthplace of the person.99 In 1916, in the context of the extension of the general tax on capital to small German principalities in the sixteenth and seventeenth centuries, Thomae Maullii first distinguished the two key concepts of the situs of property and the domicile of its owner.100 In the eighteenth century, the issue remained predominant due to general property and inheritance taxes. The problems encountered were in relation to property owned by foreigners and the solutions differed from country to country: a distinction was made between, one the one hand, immovable property subjected to the lex loci rei sitae and, on the other hand, movable property which either followed its possessor (England), was subject to the lex loci (Scotland), or had an emphasis on domicile (Germany).101 The importance of the issue gradually diminished only to reappear in the last quarter of the nineteenth century with

98 S. Jogarajan, ‘Prelude to the International Tax Treaty Network: 1815–1914 Early Tax Treaties and the Conditions for Action’, Oxford Journal of Legal Studies 31/4 (2011), 679–707. 99 Seligman, ‘Double et la coopération fiscale internationale’, 490ff. 100 T. Maullii, De Homagio, Reverentia, Obsequio, Operis, Auxilio et Aliis Juribus quae sunt inter dominos et Subditos ex Jure Diligens et Accurata Tractatio (Liège: Typis Lamberti Thonon, 1614), 51. 101 Seligman, ‘Double et la coopération fiscale internationale’, 490ff.

18 Marilyne Sadowsky the modern developments with which we are now familiar. The first modern tax treaty concluded in 1843 between France and Belgium concerned the exchange of information on deeds presented at registration.102 However, over time, the focus of cooperation gradually shifted from relief of double taxation to the prevention of tax evasion/fraud. The first need of a state is to obtain financial resources for its sovereign to satisfy their needs and those of their kingdom, empire, or state. But, as note earlier, the most archaic communities, from around 3000 bce, did not support themselves by levying taxes on land or income. One of the nine principles of ancient fiscal policy enunciated by the economic historian Hudson was elite tax evasion.103 Base erosion was a key concern and, in ancient Rome, tax evasion had already been criminalized and, for slaves, could even lead to death.104 The term that appears in the Digest is fraudis.105 The issue of tax evasion continued over the next few thousand years, which is unsurprising as people at different times and in different places have always had divergent ideas about the fairness of taxation—a matter that has been aggravated by the evolution of technology. It is a form of fiscal resistance that stems from a sense of injustice felt by the taxpayer.106 Economic models have shifted from a material to a digital economy and from national companies to international conglomerates, and people’s mobility has also increased. Even the vocabulary of tax treaties has changed to the now familiar term of tax ‘avoidance’. However, the Organisation for Economic Co- operation and Development (OECD) Model Tax Convention on Income and on Capital ignores the term.107 The increasing complexity of tax rules, in their design, in their application, and in their interpretation, has led to loopholes and mismatches between the systems of taxing countries, benefiting some taxpayers and creating a crisis of trust for others. This evolution shows the development of solidarity between states to protect their tax bases. At the time of the tax revolts, the general populace was rebelling against the state, but the focus now is the rebellion of states against the manipulation of tax bases by the people. States want to strengthen their relations, and international organizations remain at the heart of that change. In the recent past, the League of Nations, the United Nations (UN),108 and tax 102

Treaty signed on 12 August 1843 and confirmed by art. 14(2) of the inheritance tax treaty signed on 20 January 1959. 103 M. Hudson, ‘Mesopotamia and Classical Antiquity’, American Journal of Economics and Sociology 59/5 (2000), 3ff. 104 P. A. Brunt, Roman Imperial Themes (Oxford: Oxford University Press, 1990), 337–338. 105 Digest, XLVIII, Title XVIII, §20, 390: ‘In causa tributorum in quibus esse reipublicae nervos nemini dubium est, periculi quoque ratio, quod servo fraudis conscio capitalem pœnam denuntiat, ejusdem professionem extruat.’ 106 On the concept of tax resistance, see N. Delalande and R. Huret, ‘Tax Resistance: A Global History?’, Journal of Policy History 25/3 (2013). 301–307. 107 L. Friedlander and S. Wilkie, ‘Policy Forum: The History of Tax Treaty Provisions—And Why It Is Important to Know about It’, Canadian Tax Journal 54/4 (2006), 919. 108 M. Chrétien, ‘Contribution à l’étude du droit international fiscal actuel: le rôle des organisations internationales dans le règlement des questions d’impôts entre les divers États’, in Recueil des Cours 1954 II, vol. 86 (Leyde: A. W. Sijthoff, 1955), 5–116.

The History of International Tax Law 19 committees109 have been important in that regard. In the future, this will be the case for the OECD with the BEPS Project which aims to ensure that profits are taxed where the economic activities take place and where the value is created and also with the aim of reducing disputes. Multilateralism is based on the idea of socialization and the collaboration of states, which have the possibility of expressing reservations on certain issues. The 1922 multilateral tax treaty concluded between six states is often cited, but it is possible to go back to 1904 to find a multilateral tax treaty between twenty-nine countries which allowed tax exemption for ships in ports in wartime.110 Multilateralism has developed slowly, because negotiation between several states is a lengthy process if it is to have a successful outcome. The example of the MLI adopted in 2016 is a perfect example of this success. Here is a multilateral tax treaty covering ninety-nine jurisdictions and thousands of tax agreements.111 While model tax treaties have de facto authority over states,112 the MLI is progressing towards producing a ‘quasi-binding’ effect. Indeed, the BEPS actions include minimum standards to be implemented by members on treaty abuse (Action 6) and the resolution of international tax disputes (Action 14). Article 7(1) of the MLI (prevention of treaty abuse) relating to the principal purpose test has been adopted by all signatories.113 Therefore, all are bound by this minimum standard.114 We use the term ‘quasi-binding’ because while the MLI has a direct impact on the tax treaty network, its effect depends on the number of signatory jurisdictions and the matching choices made by those jurisdictions.115 This reinforces the idea that the OECD Model Convention is a quasi- source of law, already embodied in the practice of most states that rely heavily on model conventions. Here, the change concerns the state’s behaviour and reflect a ‘postmodern’ state.116 This postmodernity shows the capacity of international tax law to adapt to changing realities and to move towards the strengthening of interstate relations based on ever more binding rules. The BEPS reform and its ‘pillars’ propose an incremental change, adding new principles to the existing ones. Some authors even argue that there are no major new features in the agreement reached on 8 October 2021 by the 136 jurisdictions of the

109 C. T. Crobaugh, ‘International Comity in Taxation’, Journal of Political Economy 31/2 (1923), 262–275. 110 Convention sur les bâtiments hospitaliers, The Hague, 21 December 1904. 111 Information given on the OECD website (accessed 28 June 2022). 112 Chrétien, ‘Contribution à l’étude du droit international fiscal actuel’, 96ff. 113 D. G. Duff and D. Gutmann, ‘General Report: Restructuring the Treaty Network’, Cahiers de droit fiscal international vol. 105A (2020), 31. 114 Ibid., 47. 115 Ibid., 28. 116 J. Chevallier, ‘Vers un droit postmoderne’, in J. Clam and G. Martin, eds, Les transformations de la régulation juridique, Collection Droit et Société (Paris: Librairie Générale de Droit et de Jurisprudence, 1998), 21–46.

20 Marilyne Sadowsky Inclusive Framework on BEPS and that these developments are foreshadowed by history.117 In any event, while the evolution is now certain, the outcome is not. Even if changes in international tax law are not easily achieved by the OECD, it is time to make further progress either by repeating history or writing a new one. As Bob Dylan might have sung: international taxation, it is a-changin’.

117 R.

Avi-Yonah, ‘The International Tax Regime at 100: Reflections on the OECD’s BEPS Project’, Bulletin for International Taxation 75/11/12 (2021).

Chapter 2

From the ‘ 192 0s C om promise’ to t h e ‘ 2 02 0s C ompromi se ’ Roberto Bernales Soriano

2.1 Introduction This chapter presents an overview of the establishment of the so-called 1920s compromise, its evolution, consolidation into an international tax system, its subsequent crisis, and its revision and partial replacement by the new 2020s compromise. The scope and appraisal of the terms of the review are subject to different points of view. However, due to the number of topics concerned, the stakeholders involved, the results already achieved and expected to be achieved, and the instruments used for the implementation of the results, the review marks a turning point in the history of international taxation. In the chapter, first, we describe the circumstances and historical precedents of the 1920s compromise and its corollary of the 1930s, the arm’s-length principle (ALP). Then we describe how both were embedded and incorporated into an international tax system which entered into a long crisis from the 1990s. The 2008 financial crisis marked the need for a serious review of the whole system and the urgent need for a new compromise. We summarize the milestones of the process to achieve a new compromise, its content, the last issues to resolve in order to achieve the final agreement, and the subsequent developments to implement the agreement. Then, after a critical assessment, we compare the 1920s compromise with the new 2020s compromise, to reach the final conclusions.

22 Roberto Bernales Soriano

2.2 The Foundations of a New International Tax Order: The 1920s Compromise and Its Corollary of the 1930s The low tax rates existing during the nineteenth and beginning of the twentieth centuries were not a real concern among the economic operators and, until the decade of the 1920s, the demand for tax revenues by governments had remained relatively modest.1 After the First World War, the increase in taxation and therefore of double taxation became a concern for the international business community. As a consequence, the International Chamber of Commerce (ICC) proposed the organization of an International Financial Conference held in Brussels in 1920. The conference, among many other proposals, recommended that the League of Nations make progress on an international understanding which, while ensuring the due payment of everyone’s full share of taxation, would avoid the imposition of double taxation, which is at present an obstacle to placing investments abroad.2 The League of Nations initiated its work on international taxation and under its leadership four reports on international double taxation were produced in the 1920s: the 1923 Economists’ Report, and the reports on double taxation and tax evasion of the Committee of Technical Experts of 1925, 1927, and 1928. The foundations of the system as we know it today were laid in the 1923 Economists’ Report by the so-called four economists (Professors Bruins, from the Netherlands, Einaudi from Italy, Seligman from the USA, and Sir Josiah Stamp from the UK). The election of the economists was carefully made since two of them were from capital- importing countries (the Netherlands and Italy), one from a traditional capital- exporting country at the time (UK), and the last one from a former capital importer but the most important capital exporter at the time (USA).3 The Economists’ Report sought to define the jurisdiction to tax as a function of what it termed ‘economic allegiance’. Economic allegiance arose from the sense that the tax base, as a product of economic activity, must be understood not in terms of a taxpayer’s political or social connections to a country, but by their economic interaction with and within it.4 The report was something of a compromise. On the one hand, it put forward 1

R. S. Avi-Yonah, ‘Hanging Together: A Multilateral Approach to Taxing Multinationals’, in T. Pogge and K. Mehta, eds, Global Tax Fairness (Oxford: Oxford University Press, 2016), 113–128; V. Tanzi, ‘Lakes, Oceans and Taxes: Why the World Needs a World Tax Authority’, ibid., 251–264. 2 Avi-Yonah, ‘Hanging Together’. 3 R. S. Avi- Yonah, Advanced Introduction to International Tax Law, 2nd ed. (Northampton, MA: Edward Elgar, 2019), 4. 4 A. Christians and T. D. Magalhaes, ‘A New Global Tax Deal for the Digital Age’, Canadian Tax Journal 67/4 (2019), 1153–1178.

From the ‘1920s Compromise’ to the ‘2020s Compromise’ 23 the basic principle that the modern tax system is based on the ability to pay, which therefore favoured taxation by the country of residence. On the other hand, its analysis of national allegiance showed that international investments resulted in a division among different countries of the elements of allegiance which it identified—the most important factors being the source of the wealth created and the residence of its owner. The report accepted that agreement on the allocation of jurisdiction to tax could not be reached on the basis of any simple general principle.5 By international agreement, different types of tax could be assigned according to the primary economic allegiance of the income in question: hence, the source would have first claim on the profits of the business itself, while returns on investment, such as interest and dividends, could be treated as personal and taxed in the country of residence of the recipient.6 The works of the League culminated with the publication of the 1928 report and the 1928 conference held in October 1928 with representatives from twenty-seven countries. The result was the approval of model treaties on direct taxes, succession duties, and administrative assistance in assessment and on assistance in the collection of taxes.7 The three variants of the Convention on Direct Taxes were to provide models for bilateral agreements between states, to be adapted by negotiation around the tax systems and economic relationships of each set of treaty partners. Basic principles, following the Economists’ Report, were embodied in the three versions of the direct taxation model: profits made by a foreign company could only be taxed at source if made through a permanent establishment (PE) situated within the territory; the country of residence of an enterprise should be defined as that where the real centre of management was located; and affiliated companies or subsidiaries were not to be treated as PEs but as separate entities. The concept of a PE became established as a key factor in the compromise of tax jurisdiction. It established a separation between the taxation of business profits which could be attributed to a PE and taxed at source, and the taxation of investment profits, which could be treated as personal income and taxed in the country of residence of the investor.8

5

The 1923 Report identified four possible methods of reconciling the different national approaches. First, was a foreign tax credit such as that adopted by the USA, which in principle conceded priority to the source country, while retaining a residual right for the country of residence. Secondly, the converse approach would be for the source country to exempt non-residents. Thirdly, an international agreement could establish the basis for a division of taxes between source and residence countries, as had been done by Britain with its dominions. Finally, also by international agreement, different types of tax could be assigned according to the primary economic allegiance of the income in question: the source would have first claim to the profits of the business itself, while returns on investment (interest and dividends) could be treated as personal and taxed in the country of residence of the recipient. (S. Picciotto, International Business Taxation: A Study in the Internationalization of Business Regulation, (Cambridge: Cambridge University Press, 1992), available at https://www.taxjustice.net/cms/upload/pdf/Picciotto%201992%20In ternational%20Business%20Taxation.pdf). 6 Ibid. 7 S. Jogarajan, Double Taxation and the League of Nations, Cambridge Tax Law Series (Cambridge: Cambridge University Press, 2018), 325–331, DOI: 10.1017/9781108368865.031. 8 Picciotto, International Business Taxation.

24 Roberto Bernales Soriano These three 1928 drafts of model bilateral income tax treaties for the reciprocal relief of double taxation of international income have served as the common basis for the bilateral double taxation treaty (DTT) network in force throughout the world,9 and they constitute the foundational moment of global tax governance; that is, the set of institutions governing issues of taxation that involve cross-border transactions.10 The model treaties produced at that time were an innovative way of coordinating the different tax jurisdictions, and embodied the basic principles and the key terms which are still used today. The process to achieve this compromise had to get around the different interests of the countries involved and implied a successful limitation at that time of national sovereignty over tax matters.11 The system produced in the 1920s for taxing international income is referred to in the literature as the 1920s compromise, and is fundamentally based on the 1923 Economists’ Report. It is routinely characterized as allocating the taxation of business income to the country of its source and the taxation of portfolio income to the country of the capital supplier’s residence.12 Therefore, the 1920s compromise established the foundations of the international tax system built on the source and the residence principles: the former commensurate with the benefit principle, and the latter commensurate with the ability to pay principle. The source principle benefiting capital-importer countries and the resident principle benefiting capital-exporter countries.13

2.2.1 The Corollary of the 1920s Compromise: The ALP The 1928 conference left open two important questions: first, whether the source country could impose any tax on payments embodying a return on investments, notably interest and dividends. Secondly, where an enterprise had a PE in more than one state, the draft models merely provided that the competent administrations of the states concerned should come to an arrangement for the apportionment of profits between them. Thus, the 1928 draft conventions did not tackle the treatment of transactions between subsidiaries of a single enterprise. In 1933, the Fiscal Affairs Committee of the League of Nations approved a Draft Convention on the Allocation of Business Profits between States for the Purposes of Taxation. It was based on Carroll’s comparative report on the allocation of 9 M.

J. Graetz, ‘The David R. Tillinghast Lecture, Taxing International Income: Inadequate Principles, Outdated Concepts, and Unsatisfactory Policies’, delivered at New York University School of Law, 26 October 2000, published in Tax Law Review 54 (2001), 261–336. 10 P. Dietsch and T. Rixen T., eds, ‘Global Tax Governance: What is it and Why it Matters’, in Global Tax Governance: What is Wrong with it and How to Fix It (Colchester: ECPR Press, 2016), 1–23, 3. 11 Picciotto, International Business Taxation. 12 Graetz, ‘David R. Tillinghast Lecture’. 13 P. Genschel and T. Rixen, ‘Settling and Unsettling the Transnational Legal Order of International Taxation’, in T. Halliday and G. Shaffer, eds, Transnational Legal Orders (New York: Cambridge University Press, 2015), 154–184.

From the ‘1920s Compromise’ to the ‘2020s Compromise’ 25 taxable income to multinational enterprises and their subsidiaries and PEs in thirty- five countries,14 and was supported by the study by Jones on the allocation accounting for the taxable income of industrial enterprises.15 The draft was published again in 1935 as a model for bilateral conventions and established the principle that a PE should be considered and treated as an independent enterprise. Accordingly, a PE’s taxable income should be assessed based on separate accounts. The OECD revived this principle in its Authorized OECD Approach (AOA), introduced in the 2010 update of the OECD Model Convention (MC) on DTTs.16 The genesis of the ALP goes back to the Draft Allocation Convention, and was first recognized in international tax law within the limits of intra-company dealings but was later expanded to include transactions between associated companies. The convention allowed fractional apportionment as an alternative method but only between separate PEs of the same enterprises, not between separate subsidiaries.17

2.3 The Development and Consolidation of the 1920s Compromise The 1928 models were very successful. More than a hundred DTTs based largely on the 1928 models were concluded between 1929 and 1939.18 The later Mexico Model (1943) and the London Model (1946) were the result of the review of the 1928 models. The former being less source-country yielding of taxation rights than the later (postwar) London Model.19 The OECD MC is more in line with the London Model, and the UN Model Tax Convention is more the successor of the source country-predominant Mexico Model.20 The failure to adopt a new approach to international tax after the Second World War meant that, in fact, the solution adopted after the First World War continued by default.21 The UN failed to follow the work and the leadership of the League of Nations due to political divergence between the Western and the Eastern blocs, and also the different positions of the developing countries. Thus, the time for a radical change (if

14 M.

B. Carroll, ‘Methods of Allocating Taxable Income’, in Taxation of Foreign and National Enterprises, vol. IV (Geneva: League of Nations, 1933). 15 R. C. Jones, ‘Allocation Accounting for the Taxable Income of Industrial Enterprises’, ibid. 16 R. Bernales Soriano, ‘The Authorized OECD Approach: An Overview’, in Taxation of Business Profits in the 21st Century (Amsterdam: IBFD, 2013), 135–181. 17 H. Hamaekers, ‘Arm’s Length—How Long?’, ITPJ (Mar./Apr. 2001), 30–40. 18 Jogarajan, Double Taxation and the League of Nations. 19 M. B. Carroll, ‘International Tax Law: Benefits for American Investors and Enterprises Abroad: Part I’, International Lawyer 2/ 4 (1968), 692– 728, http://www.jstor.org/stable/40704519 (accessed 17 March 2021). 20 M. Lennard, ‘The Purpose and Current Status of the United Nations Tax Work’, Asia-Pacific Tax Bulletin (Jan./Feb. 2008), 23–30, 23. 21 Graetz, ‘David R. Tillinghast Lecture’.

26 Roberto Bernales Soriano it ever existed) vanished. The OECD (the successor of the Organisation for European Economic Co-operation (OEEC)) took the lead and published the 1963 OECD Draft MC. The 1963 OECD Model was largely based on the work of the League and especially the London Model and was the culmination of fifty years of development, rather than being a new departure.22 The consolidation of the 1920s compromise implemented through the extensive DTTs network gave rise to what is considered to be the existence of an international tax regime (ITR), meaning that the freedom of most countries to adopt international tax rules is constrained by the rules embedded both in the tax treaties and in customary law reflecting the regime.23 The regime’s norms determine which country has the right to tax (the source and/or the residence country), what the tax base should be, and what should be done about double taxation when both residence and source countries share the right to tax.24 The fundamental basis for the system consists of the international elements contained in domestic tax laws and an extensive network of bilateral tax treaties based on the OECD MC, plus some elements of soft law such as the Commentaries on the MC and the OECD Transfer Pricing Guidelines (TPGL) that are far less extensive than the common elements of the OECD.25 The ITR is based on two principles: the single-tax principle and the benefits principle,26 which are derived from the main rules and principles established in the 1923 Economists’ Report.27 The single-tax principle means that income should be taxed only once, and should never be untaxed. The principle is justified on the grounds of efficiency of cross-border investment. If income derived from cross-border transactions were taxed more heavily than domestic income, it would create an (inefficient) incentive to invest domestically.28 The benefits principle implies that active business income should be taxed at source and passive investment income should be taxed at residence because those are the respective locations where the taxpayer receives the benefits resulting in the taxable income.29 The benefits principle is more generally accepted than the single-tax principle and it is considered as one of the clear compromises achieved under the League of

22

R. J. Vann, ‘A Model Tax Treaty for the Asian-Pacific Region? (Pt. 1)’, Bulletin for International Fiscal Documentation 45 (1991), 99, 103. 23 R. S. Avi- Yonah, ‘Tax Competition, Tax Arbitrage, and the International Tax Regime’, Law & Economic Working Papers, University of Michigan Law School (2007). 24 L. Eden, ‘The Arm´s Length Standard: Making it Work in a 21st Century World of Multinationals and Nation States’, in T. Pogge and K. Mehta, eds, Global Tax Fairness (Oxford: Oxford University Press, 2016), 153–172. 25 R. Mason, ‘The Transformation of International Tax’, Public Law and Legal Theory Paper Series, University of Virginia School of Law, 2020-36 (2020), 4–5. 26 R. S. Avi-Yonah, ‘International Taxation of Electronic Commerce’, Tax Law Review 52 (1997), 507. 27 Avi-Yonah, ‘Tax Competition, Tax Arbitrage’. 28 Ibid. 29 Ibid.

From the ‘1920s Compromise’ to the ‘2020s Compromise’ 27 Nations in the 1920s.30 Nowadays, even its defenders consider that the motives behind it have become obsolete. Further, the principle does not indicate precisely how multiple states providing benefits to the same taxpayer should record their entitlement to tax income from multinationals (i.e. how to divide the corporate income tax base among the various jurisdictions providing benefits).31 Thus, it does not give clear guideline as to what proportion of income should be allocated to states in whose territory the factors of production are located. In the end, neither economic allegiance nor benefit theory resolves the base distribution issue.32 The existence of an ITR as such is controversial. It is rejected by some authors on both the grounds that there is no overriding and compulsory single-tax principle and on the legitimate use and benefits of tax arbitrage.33 Other authors consider that the single-tax principle has changed over time in different stages;34 or they criticize or even deny its inclusion in the ITR.35 However, in the aftermath of the coordinated actions against base erosion and profit shifting (BEPS), countries again support a ‘full taxation’ norm that implies that all of a company’s income should be taxed at the location of its real business activities. Thus ‘full taxation’ can be understood as an element of one of the compulsory principles of the regime, the single-tax principle.36

2.4 The Crisis of the 1920s Compromise Embedded in the International Tax System: The Precedents of the 2020s Compromise Up until the 1970s, capital movements had remained largely controlled but since that time capital flows have started to move freely around the world producing a new global financial system. The confluence of the abolition of controls on capital movements and

30 H. J. Ault, ‘Corporate Integration, Tax Treaties and the Division of the International Tax Base: Principles and Practices’, Tax Law Review 47 (1992), 565; M. J. Graetz and M. M. O’Hear, ‘The “Original Intent” of U.S. International Taxation’, Duke Law Journal 46 (1997), 1021. 31 Avi-Yonah, ‘Tax Competition, Tax Arbitrage’. 32 Ibid.; Mason, ‘The Transformation of International Tax’; Christians and Magalhaes, ‘New Global Tax Deal’. 33 H. D. Rosenbloom, ‘International Tax Arbitrage and the International Tax System’, Tax Law Review 53 (2000), 137; H. D. Rosenbloom, ‘Cross-Border Arbitrage: The Good, the Bad and the Ugly. Taxes’, Tax Magazine 85 (Mar. 2007), 155. 34 J. M. De Melo Rigoni, ‘The International Tax Regime in the Twenty-First Century: The Emergence of a Third Stage’, Intertax 45/3 (2017), 205–218. 35 L. Parada, Double Non-Taxation and the Use of Hybrid Entities: An Alternative Approach in the New Era of BEPS (Alphen aan den Rijn: Kluwer Law International, 2018). 36 Mason, ‘The Transformation of International Tax’.

28 Roberto Bernales Soriano the removal or mitigation of double taxation, together with the permanence of national tax sovereignty, brought with them tax evasion and tax competition. During the 1980s and 1990s, states fought one another for tax revenues in a process known as a race to the bottom. The international tax system became an international tax market, and the international financial markets turned into something akin to a large casino, with forces of supply and demand. Tax havens and offshore centres started offering their legal frameworks for tax planning, and the tax advisory industry offered the tax planning tools to implement the legislation and practices supplied by those jurisdictions. On the demand side of the system, multinational enterprises (MNEs) also began employing efficient and sophisticated tax planning techniques to reduce their aggregate tax liabilities to the minimum or even to nil.37 In the 1990s, tax planning (especially via treaty shopping) and the use of tax havens had already undermined both the regulatory effectiveness and the legitimacy of the international tax system. Both elements are still questioned to this day,38 and the general conclusion was and is that the ITR was unjust and in urgent need of reform.39 From an economic point of view, long-standing and fundamental international tax norms have made the system unsustainable. Thus, the problem is of a system whose core premises are inherently non-commercial and non-economic, making aggressive tax planning endemic.40 Further, from the domestic point of view, the deregulation and globalization of capital flows that started in the 1980s led to a strong trend towards the ‘refeudalization’ of tax systems where the relative share of taxes paid by economically strong actors is continuously shrinking, while the tax burden for the middle class and the poor has increased. This became crystal clear when the tax burden of the top business elite (especially the big digital companies) became public knowledge. Competition between states in order to obtain revenues forced larger countries to abandon some of the basic principles that had guided tax policies in previous decades, especially the principle that all sources of income should be taxed at the same rate. The result was that income from capital became taxed at significantly lower rates than wages.41 Thus, regulatory effectiveness and 37 Dietsch

and Rixen ‘Global Tax Governance: What is it and Why it Matters’, 6; N. Shaxson and J. Christensen, ‘Tax Competitiveness—A Dangerous Obsession’, in Pogge and Mehta, Global Tax Fairness, 265–297; V. Tanzi, Government versus Markets: The Changing Economic Role of the State (New York: Cambridge University Press, 2011). 38 Picciotto, International Business Taxation; T. D. Magalhäes, ‘What Is Really Wrong with Global Tax Governance and How to Properly Fix It’, World Tax Journal 10//4 (2018), 499–536; C. Peters, On the Legitimacy of International Tax Law (Amsterdam: IBFD, 2014); I. Mosquera Valderrama, ‘Legitimacy and the Making of International Tax Law: The Challenges of Multilateralism’, World Tax Journal 7/3 (2015), 344–366. S. Kingma, Inclusive Global Tax Governance in the Post-BEPS Era (Amsterdam: IBFD, 2020). 39 P. Hongler, Justice in International Tax Law: A Normative Review of the International Tax Regime (Amsterdam: IBFD, 2019). 40 E. D. Kleinbard, ‘Stateless Income and Its Remedies’, in Pogge and Mehta, Global Tax Fairness, 129–152. 41 Tanzi, Government versus Markets; Tanzi, ‘Lakes, Oceans and Taxes’.

From the ‘1920s Compromise’ to the ‘2020s Compromise’ 29 legitimacy have also been undermined in domestic tax systems since the mid-1980s due to the recurring difficulty of establishing fairness and efficiency in the taxation of income derived from different revenue flows.42 The social consequences of this have been especially visible in the growth of inequality and the diminishing of the welfare state.43 The 2008 global financial crisis marked a turning point in the international tax system. The devastating economic and social consequences of the 2008 crisis, together with the budgetary needs of governments, were the elements necessary to change the agenda of the international tax actors and to propel the multilateral BEPS Project.44 The remarkable lack of serf-criticism both from the tax advising industry and from MNEs on their social responsibility increased public outrage, sparked tax shaming of the big digital companies, and alarmed governments and international institutions. When the 2008 crisis unfolded, the international tax system was already seriously damaged and its main concepts had been questioned (residence, source, PE, ALP) but from that the first multilateral initiative to fix the system and put an end to harmful tax competition emerged.45 These precedents are connected to the foundations of the 2020s compromise.

2.4.1 The Reconsideration of Residence and Source and PE Concepts The residence and source concepts have been subject to clear erosion over the years because they had become easily manipulated and devalued, and reduced international equity to nothing.46 The whole concept of ‘source’ became less and less meaningful as a starting point for international tax allocation and, also, the notion of ‘residence’ no longer had the fully identifiable content it had in the past.47 However, both are still cornerstones of the international tax system for allocating taxing rights. The notion has been progressively accepted that international tax allocation should not be based only on the rigid PE concept. The PE threshold has also suffered erosion in its rigidity embodied in article 5 OECD MC. The OECD had already been gradually

42 Picciotto, International Business Taxation. 43

Shaxson and Christensen, ‘Tax Competitiveness’. Mason, ‘Transformation of International Tax’. 45 OECD, Harmful Tax Competition: An Emerging Global Issue (Paris: OECD Publishing, 1998). 46 W. Schön, ‘International Tax Coordination for a Second-Best World, (Part I)’, World Tax Journal 1/1 (2009), 67–114; Picciotto, International Business Taxation. 47 Graetz, ‘David R. Tillinghast Lecture’; A. J. Auerbach, M. P Devereux, and H Simpson, eds, ‘Taxing Corporate Income’ in Dimensions of Tax Design, vol. I of the Mirrlees Review: Reforming the Tax System for the 21st Century (Oxford: Oxford University Press, 2010), 837–913. Published online by the Institute for Fiscal Studies, https://ifs.org.uk/publications/mirrleesreview; Schön, ‘International Tax Coordination for a Second-Best World, (Part I)’. 44

30 Roberto Bernales Soriano moving away from that position and eroding the substantive requirements for the existence of a PE under that article.48 On the other hand, sales and services links as an alternative to the PE threshold have supporters who regard market penetration by sales and services to be a sufficient link for allocating taxing rights to the source country and replacing the PE requirement with a de minimis threshold, such as the minimum amount of revenue for the taxpayer from customers in a given jurisdiction.49 Therefore, the new consensus on the PE concept based on ‘significant economic presence’ included in the Final Report of Action 1 of the BEPS Project is a continuation of these precedents.50

2.4.2 The Reconsideration of the ALP The ALP is a building block of the ITR (the core of the transfer pricing regime—a ‘regime within the regime’ according to some authors).51 As early as the end of the 1970s, the controversy over worldwide unitary taxation in relation to international equity emerged and the principle was called into question.52 In the following decades, MNEs became highly integrated and centrally directed. The role of multinational corporations, which assemble final products from parts manufactured in several countries, increased sharply. Cross-country use of intellectual property also increased and the value of the use of intellectual property was difficult to determine, leading to further abuse of the system. The new century brought change to the products regarded as valuable: some MNEs were set up simply to produce and sell intangible products and, further, to produce and sell information given for free by their customers. All these factors exacerbated the problem of ‘transfer prices’ and attempts to rely mainly on the ALP became increasingly useless and anachronistic.53 According to this view, the principle is obsolete, not practical and unfair, leading to what it attempted to avoid (i.e. the profit shifting of benefits from high-to low-tax jurisdictions).54 A new paradigm is necessary and also possible based on abandoning 48

W. Schön, ‘International Tax Coordination for a Second-Best World, Part III’, World Tax Journal 2/ 3 (2010), 227–290. 49 B. J. Arnold, ‘Threshold Requirements for Taxing Business Profits under Tax Treaties’, Bulletin for International Taxation 57/10 (2003), 476–492; R. J. Vann, ‘Reflections on Business Profits and the Arm’s Length Principle’, in Arnold, Sasseville, and Zolt, The Taxation of Business Profits under Tax Treaties, 133– 170; Schön, ‘International Tax Coordination for a Second-Best World, (Part I); Schön, ‘International Tax Coordination for a Second-Best World, (Part III)’. 50 OECD, Addressing the Tax Challenges of the Digital Economy, Action 1—Final Report, OECD/G20 Base Erosion and Profit Shifting Project (Paris: OECD Publishing, 2015). 51 L. Eden, ‘Taxes, Transfer Pricing and the Multinational Enterprise’, in A. Rugman, ed., The Oxford Handbook of International Business, 2nd ed. (Oxford: Oxford University Press, 2009), 591–621; L. Eden, ‘The Arm’s Length Standard: Making It Work in a 21st Century World of Multinationals and Nation States’, in Pogge and Mehta, Global Tax Fairness, 153–172. 52 Picciotto, International Business Taxation. 53 Tanzi, ‘Lakes, Oceans and Taxes’. 54 Avi-Yonah, ‘Hanging Together’.

From the ‘1920s Compromise’ to the ‘2020s Compromise’ 31 both the separate entity fiction that MNEs are a loose collection of separate and independent entities in each country and the ALP fiction, and embracing the economic reality of MNEs being unitary entities that operate as single integrated firms under central direction.55 In other words, what is needed, according to some authors, is unitary taxation via formulary apportionment, which had already been considered by the League of Nations, based on the Carrol report, but finally disregarded because political agreement would have been impossible to achieve.56 From a very different point of view, there is not a transfer pricing problem but a problem caused by ‘perverse’ incentives—set in place by national tax authorities— that encourage MNEs to manipulate transfer prices to take advantage of differences in tax rates across jurisdictions. This is an ITR ‘design’ problem that is best handled by re-establishing the ITR, amending the source and residence rules, adopting common residency definitions, having stronger anti-avoidance rules (e.g. controlled foreign corporations (CFC) rules), and having greater transparency.57 Further, the assumption that corporate groups are largely integrated so that it is barely possible to identify individual contributions and transactions, or the resulting individual profits of group members, is not in line with the business models of many jurisdictions. Thus, for some, the ALP is still the most suitable tool for dealing with the issue although it is recognized that serious refinement is needed.58 The different views seem irreconcilable but, despite the formal respect and strong commitment in favour of the ALP, its defenders have been getting closer to the position of the defenders of formulary apportionment. The most recent changes to the OECD TPGL and its methods are a good example of this movement.

2.4.3 The Initiative to Curb Harmful Tax Competition The transformation of the international tax system into a tax market led not only to a significant amount of loss in tax revenues but also a great distortion in competition and international capital allocation.59 Moreover, tax competition resulted in a classic global market failure characterized by high transaction costs, free-riding, information asymmetries, and profitable anti-competitive collusion.60

55

S. Picciotto, ‘Towards Unitary Taxation: Combined Reporting and Formulary Apportionment’, in Pogge and Mehta, Global Tax Fairness, 221–237. 56 Carroll, ‘Methods of Allocating Taxable Income’. 57 Eden, ‘Taxes, Transfer Pricing and the Multinational Enterprise’. 58 Schön, ‘International Tax Coordination for a Second-Best World, (Part III)’; A. Turina, ‘Back to Grass Roots: The Arm’s Length Standard, Comparability and Transparency—Some Perspectives from the Emerging World’, World Tax Journal (May 2018). 59 Picciotto, ‘Towards Unitary Taxation: Combined Reporting and Formulary Apportionment’. 60 T. Dagan, International Tax Policy: Between Competition and Cooperation (Cambridge: Cambridge University Press, 2018).

32 Roberto Bernales Soriano By the late 1990s, it was clear that unilateralism was not enough to deal with the issue. Therefore, multilateral efforts to avoid double taxation were combined with efforts to mitigate tax competition. The OECD began its project on harmful tax competition61 by coming back to multilateralism.62 The project was successful on the issue of preferential tax regimes but failed with regard to the fight against tax havens due to a lack of interest from the USA and other OECD members.63 This was an attempt to resolve the so-called trilemma of international taxation. Governments might resolve both tax competition and double taxation issues but then national tax sovereignty (the third element of the trilemma) should be reduced at the international level. Global tax competition is thus the inevitable result of the decentralized nature of international taxation.64 However, sound tax competition, including respect for tax sovereignty, is widely accepted as a reality and as an efficient policy option by many authors.65 Nowadays, the implicit consensus about the benefits of international tax competition have been not substituted66 but simply modulated or diminished by a new appetite for cooperation and multilateralism between states.

2.5 The 2020s Compromise: Milestones, Content, and Issues to Be Resolved Within the BEPS Project, the 2020s compromise aims to achieve a new consensus on how to modify and adjust the traditional tax regime born in the 1920s to the new challenges arising from the new digital era. The need to change the foundations of the international tax system arises because more and more MNEs access local markets through digital platforms which do not qualify as PEs, therefore they avoid taxation in terms of the traditional threshold which allows source countries to tax their activities. By source countries, the BEPS Project means the so-called market jurisdiction—states where a significant number of customers are located (the destination country of the goods or service).67

61 OECD, Harmful Tax Competition. 62

Dietsch and Rixen, ‘Global Tax Governance: What is it and Why it Matters’, 7. Shaxson and Christensen, ‘Tax Competitiveness’. 64 Genschel and Rixen, ‘Settling and Unsettling the Transnational Legal Order of International Taxation’. 65 Schön, ‘International Tax Coordination for a Second-Best World, (Part I); W. Schön, ‘International Tax Coordination for a Second-Best World, (Part II)’, World Tax Journal 2/1 (2010), 65–94; Schön, ‘International Tax Coordination for a Second-Best World, Part III’. 66 Contrary to Peters, On the Legitimacy of International Tax Law. 67 W. Haslehner et al., Tax and the Digital Economy: Challenges and Proposal for Reform (Alphen aan den Rijn: Kluwer Law International, 2019). 63

From the ‘1920s Compromise’ to the ‘2020s Compromise’ 33 The Action 1 Final Report on Addressing the Tax Challenges of the Digital Economy of the BEPS Project68 sets out an analysis of the challenges for international taxation derived from the spread of the digital economy. The report considers that the digital economy is becoming the economy itself so that it would not be feasible to ring-fence the digital economy from the rest of the economy for tax purposes. This, in fact, includes the US position that defended the idea that there was no such thing as a ‘digital economy’ but, rather, a broad ‘digitalization of the economy’. On the other hand, the report recognizes that there are certain business models and key features of the digital economy that may exacerbate BEPS risks, for example the importance of intangibles combined with their mobility for tax purposes under existing tax rules, generates substantial BEPS opportunities. The Action 1 Report also identified a number of tax challenges relating to digitalization that go beyond BEPS—namely nexus, data, and characterization—and considered options that could address some of these broader challenges. The report considers that the two main policy challenges that need to be addressed are: first, the question of ‘nexus’ (i.e. where to tax profits) and, secondly, the question of value creation and profit allocation (i.e. how to attribute profit in the new digitalized business models driven by intangible assets, data, and knowledge). It is also recognized that several adjustments to the existing principles should be made, which would require substantial departures from existing standards for allocating profits within an MNE operating in multiple jurisdictions. Therefore, either substantial rewriting of the rules for the attribution of profits should be undertaken or alternative methods need to be considered, including methods based on fractional apportionment. After the report was produced, the process to reach a final agreement on the different options came to a halt because the differences between parties (and especially between the USA and other members of the Inclusive Framework) were too great. Recent US tax reforms69 have contributed to a new scenario which could reopen discussions.

2.5.1 The Main Milestones in the Process to Achieve a New Compromise Subsequent to the Action 1 Report, in March 2018 the Interim Report was delivered.70 It recognized discrepancies on whether and to what extent data and user participation 68 OECD, Addressing the Tax Challenges of the Digital Economy, Action 1. 69

In December 2017, the global intangible low-taxed income (GILTI), was introduced as an outbound anti-base erosion provision. It is a newly defined category of foreign income added to corporate taxable income each year. It is a tax on earnings that exceed a 10% return on a company’s invested foreign assets. GILTI is subject to a worldwide minimum tax of 10.5–13.125% on an annual basis. GILTI is supposed to reduce the incentive to shift corporate profits out of the USA into low-or zero-tax jurisdictions by using intellectual property. 70 OECD, Tax Challenges Arising from Digitalisation—Interim Report 2018: Inclusive Framework on BEPS (Paris: OECD Publishing, 2018).

34 Roberto Bernales Soriano represent contributions to value creation by enterprises. It also included guidance on the design of interim measures (e.g. digital services taxes (DSTs)) and set the 2020 deadline to achieve a consensus-based solution within the Inclusive Framework. On 23 January 2019, the Inclusive Framework issued a Policy Note (‘Addressing the tax challenges of the digitalisation of the economy’)71 which expressly introduced the current two pillars that were subsequently developed in the Programme of Work (PoW) delivered in May 2019. The BEPS 2.0 project took its current shape from the PoW which is also the base of BEPS in its current form. In October 2019, the OECD Secretariat published its proposal for a ‘Unified Approach’ of the three alternatives (‘user participation’, ‘marketing intangibles’, and ‘significant economic presence’) set out in the PoW under Pillar One72 with the purpose of making progress towards reaching a consensus solution for Pillar One. The approach covered highly digital business models but broadly focusing on consumer-facing businesses (CFBs). The scope of the proposal would be limited to large businesses that interact remotely with users in market jurisdictions above a certain threshold although some business sectors could be carved out for practical reasons. This implies the acceptance of the US position in the sense that the scope has been extended beyond the scope of highly digitalized businesses. Further, the new profit allocation rules would go beyond the ALP and the limitations on taxing rights determined by reference to physical presence. This was, according to the proposal, ‘a conscious deep departure from the established international tax rules’. In December 2019, the USA proposed the implementation of Pillar One on a safe harbour basis, and in January 2020 the OECD/G20 Inclusive Framework on BEPS issued a statement on the Two-Pillar Approach73 expressing rejection of the US proposal. The meeting of the Inclusive Framework planned for July 2020 was postponed due to the Covid-19 pandemic and, after the 8–9 October 2020 Cover Statement by the Inclusive Framework74 and the publication of Reports on the Blueprints of Pillar One75 and Pillar Two,76 the deadline for an agreement moved to mid-2021. In the meantime, a

71

OECD, ‘Public Consultation Document, Addressing the Tax Challenges of the Digitalisation of the Economy’ (13 February–6 March 2019). 72 OECD, Programme of Work to Develop a Consensus Solution to the Tax Challenges Arising from the Digitalisation of the Economy (Paris: OECD Publishing, 2019). 73 OECD, Statement by the OECD/G20 Inclusive Framework on BEPS on the Two-Pillar Approach to Address the Tax Challenges Arising from the Digitalisation of the Economy—January 2020, OECD/G20 Inclusive Framework on BEPS (Paris: OECD Publishing, 2020), http://www.oecd.org/tax/beps/statem ent-by-the-oecd-g20-inclusive-framework-on-beps-january-2020.pdf. 74 OECD, Cover Statement by the Inclusive Framework on the Reports on the Blueprints of Pillar One and Pillar Two as Approved by the Inclusive Framework at Its Meeting on 8–9 October (Paris: OECD Publishing, 2020). 75 OECD, Programme of Work to Develop a Consensus Solution to Tax Challenges. 76 OECD, Tax Challenges Arising from Digitalisation— Report on Pillar Two Blueprint: Inclusive Framework on BEPS, OECD/G20 Base Erosion and Profit Shifting Project (Paris: OECD Publishing, 2020), https://doi.org/10.1787/abb4c3d1-en.

From the ‘1920s Compromise’ to the ‘2020s Compromise’ 35 Public Consultation on the Reports on the Pillar One and Pillar Two Blueprints was held between October and December 2020 and January 2021.

2.5.2 The Last Issues to Resolve and the Final Agreement The deadline for reaching a compromise was July 2021. However, there were important issues, both political and technical, to be resolved that endangered that achievement. The most important political issue referred to the scope of Pillar One; that is, whether the new rules should go beyond the digital economy and apply to all those activities which have a more remote engagement with the market than they previously had, but recognizing that there are activities which have not changed very much with the advent of the digital economy (e.g. the way certain products are distributed—online or in a physical shop). Another important political issue was whether there might be a phased implementation (Pillar Two first and then Pillar One) and how to implement the agreement or agreements on the two pillars, since in many countries the adoption of a new multilateral instrument is a political issue owing to constitutional constraints. There were also technical difficulties, but the common feeling was that once the political problems were resolved, it would not be difficult to resolve the technical questions. It is worth noting the oscillating and erratic (and sometimes even irritating) position of the USA’s red lines to the negotiation that were changing (and contradictory) during the process.77 However, the change in the US position in announcing that it no longer required Pillar One rules to be considered as a safe harbour, and its commitment to the process to reach agreement on Pillar One,78 gave cause for optimism about reaching agreement within the deadline. Finally, on 1 July 2021, the OECD/G20 Inclusive Framework on BEP announced that 130 countries and jurisdictions (out of the 139 members of the Inclusive Framework) had joined and agreed the statement of the same date on a new two-pillar solution to address the tax challenges arising from the digitalization of the economy in order to reform international taxation rules and ensure that multinational enterprises paid their fair share of tax wherever they operated.79 The statement also announced that a

77

The different positions can be traced back through different declarations and letters, e.g. Treasury Deputy Assistant Secretary for International Tax Policy, L.G. ‘Chip’ Harter declarations at the 21st Annual Institute on Current Issues in International, Taxation, 13–14 December, Washington, DC; and Letters from the US Treasury Secretary Mr Mnuchin to the OECD Secretary-General Mr Gurría of 2019 and 2020. 78 Letter from the US Treasury Secretary Janet Yellen to the G20 officials of February 2021. 79 Statement on a Two-Pillar Solution to Address the Tax Challenges Arising From the Digitalisation of the Economy 1 July 2021, OECD/G20 Inclusive Framework on BEPS, https://www.oecd.org/tax/beps/ statement-on-a-two-pillar-solution-to-address-the-tax-challenges-arising-from-the-digitalisation-of- the-economy-july-2021.pdf.

36 Roberto Bernales Soriano detailed implementation plan together with the remaining issues would be finalized by October 2021. On 8 October 2021, a new Statement80 on a Two-Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy was published. It updated and finalized the political agreement reached in July by members of the Inclusive Framework to reform substantially the international tax rules. It was supported by all OECD and G20 countries. Four countries—Kenya, Nigeria, Pakistan, and Sri Lanka— did not join the agreement at that time. The October statement specifies some of the issues let open in the July statement. Probably the most important one was the agreement on the minimum 15% tax rate applied to some MNEs from 2023 (the July statement noted that it would be ‘at least’ 15%).81 Other significant issues that were specified were: (1) regarding Pillar One, 25% of residual profit (defined as a profit in excess of 10% of revenue) of in-scope MNEs should be allocated to market jurisdictions (in July the scope was 25–30%); and (2) regarding Pillar Two, the minimum rate for the subject to tax rule (STTR) was set at 9% (in July, the scope was 7.5–9%).82

2.5.3 The Content of the 2020s Compromise After these steps towards a final agreement, the compromise on tax reform for the twenty-first century will have two components:83 (1) Pillar One, which seeks to undertake a coherent and concurrent review of the nexus and profit allocation rules by expanding the taxing rights of market/user jurisdictions where there is an active and sustained participation of a business in the economy of that jurisdiction through activities in, or remotely directed at, that jurisdiction. It also seeks to improve tax certainty through dispute prevention and resolution measures. Pillar One includes two key concepts: (a) a new nexus rule in order to tax profits of MNEs even though they do not have a PE or any physical presence in a certain jurisdiction, because physical 80 Statement on a Two-Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy, 8 October 2021, OECD/G20 Inclusive Framework on BEPS, https://www.oecd.org/tax/ beps/statement-on-a-two-pillar-solution-to-address-the-tax-challenges-arising-from-the-digitalisat ion-of-the-economy-october-2021.pdf. 81 Statement of 8 October 2021, 4. 82 Ibid., 5. 83 For a summary of the two blueprints containing the two pillars, see A. Russo et al., ‘International Tax: Pillar One—Overview of “the Blueprint” ’, Baker McKenzie (2020), https://insightplus.bakermcken zie.com/bm/tax/international-tax-overview-of-blueprint; A. Russo et al., International Tax: Pillar Two— The New Normal for Effective Tax Rates’, Baker McKenzie (2020), https://insightplus.bakermckenzie. com/bm/tax/international-oecd-pillar-two-blueprint-released-full-summary-attorney-advertising.

From the ‘1920s Compromise’ to the ‘2020s Compromise’ 37 presence may not be enough to determine whether or not a company is actively present; and (b) new profit allocation rules, as a new layer added to the existing rules applicable to so-called routine profits, which would apply to the so-called residual profit (coming mostly from intangible assets) and which would be allocated to the market jurisdiction. Thus, the basic elements of Pillar One are a new taxing right for market/user jurisdictions based on a share of a business’s residual profits (‘Amount A’) and a fixed return for certain distribution and marketing activities physically in the market/user jurisdiction (‘Amount B’). This means that there is consensus on the fact that the ALP does not work properly and the need to change the transfer pricing rules mostly based on physical presence. (2) Pillar Two focuses on the remaining BEPS issues, and seeks to develop rules that will establish a minimum level of tax in global terms, by providing jurisdictions with a right to ‘tax back’ where other jurisdictions have not exercised their primary taxing rights or the payment is otherwise subject to low levels of effective taxation. This would be achieved through: (a) a new global minimum tax regime which aims to ensure a minimum effective tax rate of 15% across all jurisdictions, and is based on the Global Anti-Base Erosion (GloBE) Rules; that is, an income inclusion rule (IIR), which resembles a CFC rule, and the ‘undertaxed payments rule’ (UTPR), which would not apply where the income inclusion rule applies; and (b) imposing a minimum level of taxation (9%) on certain payments (interest, royalties, and a defined set of other payments) between connected persons (the STTR). Members of the Inclusive Framework (IF) that apply nominal corporate income tax rates below the STTR minimum rate would implement the STTR into their DTTs with developing IF members. The work on Pillar Two goes back to Action 3 (CFC) and Action 5 (harmful tax practices) of BEPS Project. According to the statements agreed in July and October, The GloBE rules will have the status of a common approach. This means that IF members are not required to adopt the GloBE rules but, if they choose to do so, they will implement and administer the rules in a way that is consistent with the outcomes provided for under Pillar Two, including in the light of model rules and guidance agreed to by the IF. Further, they accept the application of the GloBE rules applied by other IF members including agreement as to rule order and the application of any agreed safe harbours. The original deadlines for the implementation of both Pillar One and Pillar Two were set at mid-2022 with a high-level signing ceremony for the Multilateral Convention to implement Amount A of Pillar One, which would come into effect in 2023; and the Multilateral Instrument (MLI) for implementation of the STTR in relevant DTTs ready

38 Roberto Bernales Soriano for signing on the same date. Further, Pillar Two should be brought into law in 2022 to be effective in 2023 with the UTPR coming into effect in 2024.84 However, the deadlines were delayed due to the consultation process carried out with businesses, NGOs, and other stakeholders in order to refine and improve the rules.85

2.6 The Alternatives to the Failure to Achieve a Compromise The compromise that is at stake covers Pillar One and Pillar Two. There are voices that claim that agreement will have to be achieved on both pillars, while others consider that a separate agreement is both technically and politically possible. Thus, the alternative is either to reach an agreement on these progressive and fundamental changes or not to reach an agreement at all. The problem with not reaching an agreement is that there does not seem to be an alternative. The most immediate outcome would be a myriad of DSTs. But there is also the risk of different approaches to the concepts included in Pillar One with unilateral and uncoordinated measures taken by different countries. Regarding the nexus, the creation of different concepts of ‘digital’ PEs would also be on the agenda of unilateral measures. In addition, with regard to the amount of profits and their allocation, it would be possible to implement a whole range of deviations from the ALP. Regarding Pillar Two, the failure to complete the technical details of a coordinated framework for global minimum tax rules would result in a Pillar Two agreement that could also lead to tax uncertainty, tax disputes, and double taxation. The proposal for article 12B UN MC on income from automated digital services could be considered as a back-up plan in the event of there being no agreement. However, it lacks the representative legitimacy supported by the proposals within the OECD/G20 and the IF. Further, the way it is drafted does not seem to address all the issues and neither does it seem appropriate for some business. It is also a complicated article. Thus, countries would rather go for implementation of unilateral DSTs instead of negotiating treaty by treaty to implement article 12B. Moreover, in the absence of an agreed position, there would be an EU solution that would bring more uncertainty and chaos in the international tax system.

84

Brochure on the Two-Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy, October 2021, OECD/G20 Base Erosion and Profit Shifting Project, 5. 85 Public consultations on the implementation aspects of Pillar One and Pillar Two were opened in 2022. At the time of writing, it is still pending the public consultation on the Progress Report on Amount A of Pillar One opened in July, where interested parties were invited to send comments on the discussion draft no later than 19 August 2022. A public consultation meeting on the issue was due to be held on 12 September 2022.

From the ‘1920s Compromise’ to the ‘2020s Compromise’ 39

2.7 Critical Assessment of the 2020 Compromise within the New International Tax Governance The 2020s compromise deals with the tax challenges of digitalization but de facto the project goes beyond that objective. There has been an evolution from the starting point of dealing with profit shifting and tax competition and trying to close the loopholes of the international tax system. Pillars One and Two have undertaken a more fundamental structural change and, in fact, Pillar Two does not have anything specifically relating to digitalization and, hence, the reason why some call BEPS 2.0 a misnomer, because it goes beyond the original BEPS Project’s purpose. However, as we have seen, it is connected to Action 3 (CFC) and Action 5 (harmful tax practices) of the BEPS Project. Certain critics of the new 2020s BEPS 2.0 compromise (and of the BEPS Project in general) are sceptical of its narrow scope and of what they see as its remote chances of success.86 For them, the new global compromise looks very much like the old global tax deal, with a relatively modest redistribution of taxing rights among a few key states, thus missing an opportunity for meaningful reform. According to this view, the proposal in Pillar One is too conservative, since the compromise does not seek a review of the balance between residence and source taxation but instead seeks to rebalance taxing rights between a few states (OECD members in addition to a few other important states). Market jurisdictions would cover the location of users which seems universal since all jurisdictions appear to be locations of markets and consumers. Yet the threshold requirement of significance—together with the measures that will be introduced to simplify matters and provide certainty to taxpayers—may ultimately exclude most developing countries from the picture. Thus, the new deal will look very much like the old one, with just a slight redistribution of taxing rights between a few states. Critics claim, at least in the long run, that there should be a reconsideration of the overall structure of the rules that allocate taxing rights among residence and source states.87 Once it is clear that some changes have to be made to the system, the differences arise between those who think that the time has come to make fundamental changes to the whole system and those who support fundamental but incremental change. For the former, both the 2008 crisis and the Covid pandemic provided the momentum that was needed for radical changes. For the latter, the expected changes are qualitatively huge, resembling the Pareto principle (the 80–20 rule), because even though they will affect only 20% of transactions, they will impact 80% of the value at stake. The ALP will still apply to the remaining 80% of transactions.

86 87

Dietsch and Rixen, ‘Global Tax Governance: What is it and Why it Matters’, 13. Christians and Magalhaes, ‘New Global Tax Deal for the Digital Age’.

40 Roberto Bernales Soriano In any event, the reforms suppose a radical change in the international tax system. The target and principles of Pillar One, apart from the nexus rule, clearly depart from the international tax framework of determining nexus and profit allocation under the ALP through an evaluation of where value is created and substantial activities are undertaken, and departures also from the separate entity approach to a group approach. And both Pillar One and Pillar Two bring elements of a formulary apportionment moving away from the OECD-revered ALP. For the conservative critics, it is said that a departure from the long-standing global tax framework based on the ALP could give rise to significant tax uncertainty and increase tax disputes and widespread double taxation. However, this sounds a little ironic, given that the ALP is not considered to be an ‘exact science’, as it is subject to countless discrepancies between taxpayers and tax administrations. Further, the old nexus rules, the PE concept, and the ALP still exist and will overlap with the new concepts. Connecting both the old and the new is probably going to be more complicated than simply leaving out the old principles. Regarding Pillar Two, it is also important to note that the adoption of minimum tax is not linked with the existence of low or even no taxes in many countries which were previously connected with harmful tax competition. From the perspective of multilateralism, those who support a negative view find the attempts of the OECD/G20 and the Inclusive Framework to fix the system inadequate due to the absence of ‘fully global and inclusive’ multilateralism, the lack of an institutionalized enforcement, and states’ unwillingness to pool their fiscal sovereignty.88 However, this view can be considered too pessimistic, since, by 2021, the true extent of change had become clear and a new compromise seemed to be in the offing. The facts speak for themselves. The post-crisis developments have demonstrated a trend towards rapid change in the architecture of the international tax system and towards more radical, inclusive, and enforceable governance.89 Those who believe that international tax policy is simply about a fair or equitable allocation of taxation rights between two countries—with both seeking to get as big a slice of the cake as possible—and based on the defence of sound tax competition as an efficient policy option90 (whatever ‘sound’ may mean), should take note of the new paradigm. Most of the governments involved in the new compromise act in the conviction that there is a need for multilateral instruments to implement the 2020s compromise (clearly Pillar One and also most probably Pillar Two). Thus, the 2020s compromise expressly respects national sovereignty but the concept of a (restricted) coordinated and shared sovereignty underlies the whole compromise.

88

Dietsch and Rixen, ‘Global Tax Governance: What is it and Why it Matters’, 7. R. Corlin Christensen and M. Hearson, ‘The New Politics of Global Tax Governance: Taking Stock a Decade After the Financial Crisis’, Review of International Political Economy 26/5 (2019), 1068–1088, DOI: 10.1080/09692290.2019.1625802, https://doi.org/10.1080/09692290.2019.1625802. 90 Schön, ‘International Tax Coordination for a Second-Best World, (Part I and conclusions in Part III)’, 259. 89

From the ‘1920s Compromise’ to the ‘2020s Compromise’ 41 We argue that the way the process is conducted and the instrument in which the changes are reflected are part of the change itself. The Multilateral Convention to Implement Tax Treaty Related Measures to Prevent BEPS (MLI), signed on 7 June 2017 (entered into force 1 July 2018) can be considered as the turning point of the international tax system of the twenty-first century. The conventional perception existing not so long ago of a re-emergence of national tax sovereignty as the leading paradigm for international tax policy, and that international coordination would be achieved via DTTs and not by worldwide multinational agreements,91 should be subject to reconsideration. The new system, in which the 2020s compromise is included, is based on new legal forms starting from the MLI, which can be seen as having the potential to alter the architecture of international tax relations permanently.92 After the MLI, the second building block of the new ITR should be the multilateral instruments subsequent to the agreement of 2021 (i.e. the 2020s compromise).

2.8 The 1920s Compromise vs The 2020s Compromise Now that an agreement has been reached, we can summarize the similarities and differences between the 1920s compromise (and its corollary of the 1930s regarding the ALP) and the 2020s compromise. The two compromises arose in similar economic and political circumstances. In the case of the 1920s compromise, it included the consequences of the first wave of globalization, the aftermath of the First World War, the (many times forgotten) huge public health crisis of the so-called Spanish influenza which caused more casualties than the First World War itself, and the 1929 economic crash and its aftermath. In the case of the 2020s compromise, it included the second wave of globalization during the 1980s and 1990s, with the resultant globalization of the financial markets and a booming international tax planning market and also the 2008 financial crisis and the Covid pandemic, both of which exacerbated the need for more public revenue. In political terms, both compromises developed within an atmosphere of extreme political movements that threatened the democratic institutions and the international institutional framework. From a more technical point of view, both are formally based on two pillars. The 1920s compromise was based on the allocation of taxing rights consisting of two principles (residence and source), and a further subsequent pillar left open in 1928 on the attribution of profits to PEs and affiliates based on the ALP. The 2020s compromise groups the new taxing right and the allocation of profits based on new rules beyond the ALP in Pillar One and, as a further step of coordination, the minimum taxation in Pillar Two. 91 Ibid. 92

Mason, ‘The Transformation of International Tax’.

42 Roberto Bernales Soriano Both are also an attempt to limit national sovereignty on tax matters, very limited in the case of the 1920s compromise but more advanced in the case of the 2020s compromise. There are additional significant differences between the two compromises. In our view, the 2020s compromise can claim to be more innovative in respect of the 1920s compromise in how it creates new methods of economic allegiance and gives the source principle a new perspective for market jurisdictions. Further, it goes beyond the ALP, which was a logical consequence of the 1920s compromise, although the formulary apportionment method was already in the minds of its founders. The 1920s compromise took the PE as a key concept based on physical presence, and the ALP as the key principle for the allocation of profits, rejecting the formulary apportionment. The 2020s compromise goes beyond the physical presence to establish a nexus and admits formulary apportionment. The 2020s compromise is part of the compromises achieved under the implementation of the BEPS Project, especially on the four minimum standards and the MLI. Further, it has to be understood as a polyhedral compromise both in terms of the content of the two pillars that reflect the compromise, and in the framework of the development of the BEPS Project in which it is included (i.e. the new compromise is a cornerstone of the BEPS Project and a continuation of Actions 3 (CFC) and 5 (Harmful tax competition) of the project). The 2020s compromise has to be understood in its whole context. Unlike the 1920s compromise, the 2020s compromise takes into account new stakeholders. The number of governments involved is dramatically greater than the number behind the 1920s compromise. The business community, academia, and NGOs have been actors involved in drafting documents to be negotiated within the Inclusive Framework in the face of hostile public opinion. In this, NGOs have been quite remarkable, playing an important and novel role in raising awareness and shaping public opinion with campaigns on injustice within the ITR.93 The negotiation to achieve the 1920s compromise included only governments (with the UK as the great power and the United State as a late guest) and the lobbying of the ICC. The 1920s compromise was implemented via DTTs and consciously left its implementation to this uncoordinated international legal instrument, which proved to be a successful way of timidly limiting tax sovereignty. The 2020s compromise marks a step towards real multilateralism because it will be achieved through the Inclusive Framework and implemented—if finally achieved—via another multilateral instrument (or instruments). The targets of the 1920s compromise and that of 2020s are similar, but the 1920s compromise’s primary target was to avoid double taxation. The 2020s compromise is aimed at achieving fairness in the international tax field without causing double taxation. The order in stating the targets make a difference. The concept of ‘fairness’ is now

93 Hongler, Justice in International Tax Law; Eden, ‘Arm’s Length Standard’.

From the ‘1920s Compromise’ to the ‘2020s Compromise’ 43 first on the agenda and that is why BEPS is considered to represent a serious risk not only to tax revenues but also to tax sovereignty and tax fairness.94

2.9 Conclusion The 2020s compromise is the logical consequence of the development, consolidation via DTTs in an international tax system, and subsequent crisis of the 1920s compromise and its corollary of the 1930s concerning the ALP. Its foundations are rooted in the alternatives already taken into account by the founders of the 1920s compromise, as well as the initiatives for reform of the system which gained new momentum with the 2008 crisis and the BEPS Project. The 2020s compromise is part of a paradigm shift in international taxation; that is, the multifaceted BEPS Project standards and an innovative instrument (the MLI) to implement the agreed changes. Further, the compromise is a polyhedral compromise that goes beyond its precedent of the 1920s and is rooted in the fight against tax avoidance and in international equity, with a fair allocation of taxing rights and fair competition between jurisdictions. The concept of polyhedral compromise and a (restricted or limited) coordinated shared sovereignty are specific and essential parts of the 2020s compromise. The new compromise is part of the new world international tax order of the post- BEPS era, which may be compared to the new trade and financial architecture built after the Second World War. Even the least optimistic would acknowledge that formal changes may mean deep changes in the long run. The fate of our democracies lies in the ability to establish a new global tax order and a new global tax governance. The 2020s compromise can be a step forward to a new ITR and be key to international tax governance.

94 OECD, Addressing Base Erosion and Profit Shifting (Paris: OECD Publishing, 2013).

Chapter 3

Sou rces of L aw a nd Legal Meth od s i n Internationa l Tax L aw Rainer Prokisch

3.1 Introduction International tax law is a very broad concept that encompasses national law and international law in a very specific way of interaction. The term ‘international tax law’ has been used to refer to all international as well as domestic tax provisions relating to cross-border situations.1 It not only addresses material law issues, such as avoidance of double taxation and the allocation of taxing rights to tax jurisdictions, but also formal law questions, such as exchange of information or cooperation between tax administrations. It covers differing taxes like those on income of individuals and corporations, on inheritances and gifts, or on the turnover derived by enterprises, for example value added tax. International tax law consists of hard law, soft law, and case law. Hard law is the provisions of domestic tax law in combination with tax treaties and other international law provisions which have impacts on taxation. However, soft law also plays an important role. Tax treaties are frequently based on non-binding tax treaty models which find further explanation in commentaries (e.g. OECD Commentary on the OECD Model Tax Convention)2 that have major importance in regard to interpretation and 1 K.

Vogel, Klaus Vogel on Double Taxation Conventions, 3rd ed. (Boston, MA: Kluwer Law International, 1997), , Introduction, m.no. 5. 2 OECD, Model Tax Convention on Income and on Capital 2017 (Paris: OECD Publishing, 2017). The full version includes the articles, Commentaries, non- member economies’ positions, the Recommendation of the OECD Council, historical notes, and background reports. See also United Nations, Department of Economic & Social Affairs, Model Double Taxation Convention between Developed and Developing Countries, 2021 update (New York: UN, 2021). It includes an introduction,

46 Rainer Prokisch application of such tax treaties as far as the model has been used as a basis. The importance of the commentaries on the models can be explained by the fact that tax treaties frequently come without any further explanation. Typically, bilateral treaties are negotiated without recording any minutes and the accompanying unilateral documents cannot be used for interpretation and application unless they were signed by both parties.3 The commentaries are therefore the only source which can assist in finding the proper meaning of a treaty term. The OECD guidelines on transfer pricing,4 which are often used by countries to supplement and concretize the national transfer pricing rules although they are not directly legally binding, are also very significant. The recent project called Base Erosion and Profit Shifting (BEPS) has resulted in various OECD reports whose results were incorporated in the OECD Model Commentary in 2017 and the Multilateral Instrument signed in 2017. The BEPS reports are equally a valuable source for a deeper understanding of the problems connected with cross-border transactions, in particular in respect of tax avoidance, base erosion, and tax evasion.5 In the course of the ongoing BEPS Project, the OECD, entrusted by the G20, has presented two blueprints for a new allocation of taxing rights in respect of the digital economy and similar businesses.6 In October 2021, the OECD presented the results of the work on the tax challenges arising from the digitalization in a statement (‘landmark agreement’)7 that describes the components which were agreed on by the OECD/G20 Inclusive Framework on BEPS. The work on the technical rules of the new taxing right is ongoing and the OECD uses public consultations to gain insight from all stakeholders.8 Pillar Two introduces a 15% global minimum corporate tax rate. The text of the envisaged multilateral treaty is expected to be published in 2023. The elements of Pillar Two are planned to be brought into law on the basis of model rules9 in the near future.

the articles, and Commentaries on the articles. See also the 1987 Intra-ASEAN Model Double Taxation Convention signed by the members of the Association of South East Asian Nations (ASEAN). 3

Art. 31(2)(b) VCLT; as far as is known, the only explanatory notes that are signed by both parties are those made in connection with the treaty between Belgium and the Netherlands. 4 OECD, Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations (Paris: OECD Publishing, 2017); cf. J. Calderón, ‘The OECD Transfer Pricing Guidelines as a Source of Tax Law: Is Globalization Reaching the Tax Law?’, Intertax 35 (2007), 4–29. 5 The BEPS Project consists of fifteen actions, see http://www.oecd.org/tax/beps/beps-actions/. 6 OECD/G20, Inclusive Framework on BEPS, Tax Challenges Arising from Digitalisation—Report on Pillar One Blueprint (Paris: OECD Publishing, 2020); OECD/G20, Inclusive Framework on BEPS, Challenges Arising from Digitalisation—Report on Pillar Two Blueprint (Paris: OECD Publishing, 2020). 7 OECD/G20, Base Erosion and Profit Shifting Project, Statement on a Two-Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy (Paris: OECD Publishing, 2021) with an annex containing a detailed implementation plan. 8 Most recently, the OECD presented the Progress Report on Amount A of Pillar One, Two Pillar Solution to the Tax Challenges of the Digitalization of the Economy, OECD/G20 Base Erosion and Profit Shifting Project (Paris: OECD Publishing, 2022). 9 OECD, Tax Challenges Arising from the Digitalisation of the Economy—Global Anti-Base Erosion Model Rules (Pillar Two) (Paris: OECD Publishing, 2021); the Model Rules come with a Commentary and a sample of Illustrative Examples (both published 2022).

Sources of Law and Legal Methods in International Tax Law 47 The complexity caused by unsystematic methods of regulation, by the interaction of different layers of rules, and the non-coordination of national tax law systems may lead to double taxation, double non-taxation, and other adverse effects for taxpayers and tax authorities. Conflict resolution is therefore an extremely important feature of international tax law. First of all, national courts are appointed to protect legal positions, since an international tax court does not exist.10 When national judges have to decide on the application of tax treaties, they slip into a two-faced role: on the one hand, they have to turn to domestic law as usual, but on the other hand they have to apply tax treaties which deny, limit, or confirm the domestic tax charge. The latter function puts the judge in the position of a quasi-international judge who has to observe the law in the other contracting states and to consider differing language versions of a treaty. In parallel, taxpayers have the possibility of calling for mutual agreements between tax authorities which may be the pre-phase of an arbitration procedure. The European Court of Justice (ECJ) plays an important role in securing a common approach of applying European law and the abolishment of discriminations in transnational instances. The object of Section 3.2 is to provide an overview of legal sources relevant for international taxation. Section 3.3 will take a closer look at the interpretation and application of national, international, and European law. The chapter as a whole will focus on the taxation of income of individuals and companies.

3.2 Sources of International Tax Law 3.2.1 Domestic Tax Law Domestic tax law is the basis of taxation of cross-border activities. The charge of tax is found in the domestic tax law of each state. Typically, a domestic tax system will consider worldwide income to be subject to domestic income tax (i.e. the vast majority of states not only tax income from domestic sources but also from sources outside their own territory (principle of residence)). At the same time, states want to tax the income of non-resident persons if such income is derived from sources in the territory of a state (non-resident taxation or principle of source). Frequently, such taxes on non-resident persons will be levied in the form of withholding taxes on gross income. As a rule, the provisions which refer to cross-border transactions are scattered across the various tax laws of a given country. Some states have enacted specific laws that address international issues, such as Canada or Germany.11 10 Some authors recommend the foundation of an international tax court or a similar body. For a current proposal, see B. J. Arnold, ‘The Interpretation of Tax Treaties: Looking to the Future’, Bulletin for International Taxation 75/11/12 (2021) (published online 7 October 2021). 11 For Canada, see the Income Tax Conventions Interpretation Act of 20 December 1984; a German statute on Foreign Tax Relations (Außensteuergesetz) contains the legal basis for adjustments in transfer pricing cases, exit taxation, and CFC regulation.

48 Rainer Prokisch The consequence of taxing the worldwide income of residents and the source income of non-residents is that the tax claims of different states necessarily overlap and juridical double taxation will occur. Those states, therefore, need rules that avoid double taxation. In addition to this basic system, most states will have numerous legal provisions to address cross-border issues. First of all, most states nowadays have a so-called participation exemption in order to avoid economic double or multiple taxation and, by doing so, facilitate international trade and business. Legal rules that aim to avoid the erosion of the domestic tax base are also common. Typically, we find limitations on the deduction of interest and/or royalties if the payments are made in favour of a person abroad. Further, general anti-abuse rules may apply, supplemented by specific rules such as controlled foreign corporation (CFC) rules which apply in cases where profits are derived by subsidiaries in low-tax countries and, as a rule, are not repatriated to the parent company. There are an increasing number of provisions aimed at securing single taxation relative to hybrid entities or hybrid financing instruments. Administrative rules, which are needed to secure effective taxation relative to cross- border transactions, are also very important. Frequently, enterprises conducting cross-border business are required to provide more detailed information than in pure domestic situations. They also face numerous reporting obligations that do not exist in regard to domestic transactions. Finally, national constitutions may be relevant in regard to international taxation issues. Sometimes constitutions have a direct impact on taxation by requiring a strong rule of law or by setting the ability-to-pay principle as a benchmark for direct taxes, but they may also play a role in protecting the fundamental rights of citizens in more general terms (equity, fairness).

3.2.2 Tax Treaties National law does not always avoid double taxation effectively and its application without complementary rules will always lead to taxation of income in source states while the state of residence would be responsible for the avoidance of double taxation. One of the purposes of tax treaties is to fill the gaps—to allocate taxation rights to the states involved and to coordinate and integrate the unilateral approaches under domestic laws.12 Although the first bilateral treaties had been concluded before the First World War, tax treaties gained significance after the war as the income tax rates increased significantly in order to finance the war costs. Double taxation became painful and was recognized as a real threat to international trade. Nowadays, numerous treaties exist 12 Sometimes tax treaties are concluded in the form of multinational treaties (e.g. the Nordic Convention between Denmark, Finland, Iceland, Norway, and Sweden of 1983, later also extended to the Faroes) but this does not really change the method of legal application.

Sources of Law and Legal Methods in International Tax Law 49 and developed states have concluded treaties, at least with their most important trading partners. The need for a broader net of treaties was also the reason why the League of Nations began to work on a common basis for the conclusion of tax treaties in order to facilitate their negotiation and conclusion. In the late 1920s, the League of Nations published a number of model treaties which were mainly based on treaty practice. Before and after the publication of model conventions, sub-commissions produced numerous documents on various issues, until today a valuable source of many questions with which we are still concerned.13 During and after the Second World War, two further models were produced, the so-called Mexico and London Models. The UN, as successor organization of the League of Nations, did not continue this work despite the matter still being on the agenda. During the 1950s, the Organisation for European Economic Co-operation (OEEC) and its successor the OECD worked on a model for the most developed nations which were members of the organization. The work led to the publication of a model in 1963 that had great importance for the further development of tax treaties worldwide. Like the League of Nations models, the OECD Model incorporated commentaries on its various articles which have become a major source for interpreting and applying treaties. The OECD continued its work and, after revisions of the Model in 1977 and 1992, the Model and its Commentary became a continuous loose-leaf edition, subject to recurrent updates. The OECD Model is addressed to developed countries and does not consider the needs of developing countries. In the late 1970s, it was felt that there was a need for a model that considered the interests of developing countries in contrast to the standards used by developed countries. The UN took over the task of working on such a model, with the first version issued in 1980 and updates in 2001, 2017, and 2021.14 Some states, for example the USA, the Netherlands, Belgium, and Germany, also published their own national model conventions. These have a different purpose, aiming to provide information to members of parliament and the public in general about the treaty policy of their country. Nevertheless, they can be an important source of interpretation since they may show the object of specific rules in tax treaties. In particular, the US Model is accompanied by comprehensive explanatory notes15 that are not binding, being unilateral material, but may nevertheless be an extraordinary source for a better understanding of treaty provisions. Tax treaties benefit taxpayers by providing relief from double taxation or lower withholding tax rates. By doing so, the provisions of a treaty are transposed into

13

Many documents can be found at https://www.taxtreatieshistory.org. https://www.un.org/development/desa/financing/sites/www.un.org.development.desa.financing/ files/2022-03/UN%20Model_2021.pdf. 15 https://www.irs.gov/businesses/international-businesses/united-states-model-tax-treaty-docume nts. The US Model was updated in 2016 and the Internal Revenue Services planned to also update the explanatory notes. This, however, has not yet happened, so the relevant explanation still stems from 2006. 14

50 Rainer Prokisch taxpayer rights and the taxpayer may rely directly on treaty benefits. There is, however, no international court responsible for deciding conflicts between taxpayers and tax administrations therefore the domestic courts are the competent bodies in this respect. In this capacity, courts act as ‘international courts’. Therefore, general rules which govern the functioning of international courts—such as the Statute of the International Court16—are also relevant for domestic courts when they are called upon to interpret international conventions. Article 38 of the Statute takes into account ‘formal sources’, for example custom, treaties, general principles of law, jurisprudence, doctrine, and equity. The list is not exhaustive but is illustrative. Tax treaties are international conventions. Accordingly, the Vienna Convention on the Law of Treaties (VCLT)17 applies. The Convention is, in particular, important for the interpretation of tax treaties, but it also contains general rules such as those on the conclusion of treaties, their scope, on reservations, amendments, and modifications of treaties. Beginning in 2013, the OECD in cooperation with a number of non-member states began a comprehensive project to combat aggressive tax planning schemes which frequently used mismatches in national tax laws or the absence of specific anti-avoidance rules in domestic laws as well as in tax treaty law (the Base Erosion and Profit Shifting Project, BEPS). In regard to treaties, a new method to update tax treaties has been developed without the necessity to renegotiate individual treaties. The corresponding efforts of the international community led by the OECD resulted in a multilateral instrument (MLI) that has been signed and ratified by numerous states. After the publication of the final reports on fifteen BEPS actions, the G20 and the OECD decided to continue the work in cooperation with numerous non-member countries (Inclusive Framework on BEPS) in order to address the challenges arising from the digitalization of the economy (Pillar One) and to introduce a worldwide minimum tax regime (Pillar Two: Global Anti-Base Erosion (GloBE) Rules).18

3.2.3 European Law (Material Law) EU law is based on the EU Treaty (EUT)19 and on the Treaty on the Functioning of the European Union (TFEU)20 (so-called primary EU law). In regard to direct taxation, most important are the fundamental freedoms that became applicable on direct tax at the time the ECJ declared them as directly effective, when it broadened the scope of

16

https://www.icj-cij.org/en/statute. https://legal.un.org/ilc/texts/instruments/english/conventions/1_1_1969.pdf. 18 See OECD/G20, Base Erosion and Profit Shifting Project, Statement on a Two- Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy (Paris: OECD Publishing, 2021). 19 Consolidated version [2016] OJ C202/1, 15–45. 20 Ibid., 47–388. 17

Sources of Law and Legal Methods in International Tax Law 51 those freedoms to discriminations on the ground of residency (and not only of nationality), and when the Court had clarified that the freedoms applied irrespective of whether the EU Treaties provided legislative powers to the EU in the area of direct taxation. The case law of the Court since then has had a major influence on domestic tax systems with a certain degree of (negative) integration and the internationalization of taxation not only in Europe but also worldwide. In particular, the free movement of capital (arts 63–65 TFEU) also applies in relation to third countries. In respect of legislative powers, the EU is rather limited in regard to direct taxation, although there are exclusive legislative powers on customs and shared powers on value added taxation and excises. All powers that refer to taxation require unanimity in the Council of the European Union. The state aid rules of the TFEU (arts 107–109) also cover tax benefits provided by states to domestic enterprises. Those rules have become extremely relevant and also ensure fair competition in the field of taxation. Nevertheless, there is secondary EU law governing direct taxes. The EU has enacted several directives on the basis of article 115 TFEU for the purpose of the approximation of laws of member states that ‘directly affect the establishment or functioning of the internal market’. These are the Parent–Subsidiary Directive (1990)21 and the Interest– Royalties Directive (2003)22 which aim at the abolishment of withholding taxes on dividends, interest, and royalties within the EU. Such withholding of taxes is felt to hinder the functioning of the internal market as they may lead to double taxation. In addition, the Merger Directive (1990)23 was enacted to facilitate the reorganization of companies within the EU. As a consequence of the G20/OECD BEPS Project, the EU felt the need to harmonize the BEPS recommendations in order to avoid differing rules in the member states. The result of those efforts were two directives, the so-called ATAD I24 and ATAD II25 (anti- tax avoidance directives), which introduced anti-avoidance measures, for example interest limitation rules, exit taxation rules, a general anti-abuse rule, and CFC rules. ATAD I only addressed hybrid mismatches in the EU. However, it was felt that more

21

Council Directive 90/435/EEC of 23 July 1990 on the common system of taxation applicable in the case of parent companies and subsidiaries of different member states was subject to recast in 2011, see Council Directive 2011/96/EU of 30 November 2011 on the common system of taxation applicable in the case of parent companies and subsidiaries of different member states, amended through Directive 2014/ 86/EU of the Council of 8 July 2014, consolidated text [2015] OJ L96, 1. 22 Council Directive 2003/49/EC of 3 June 2003 on a common system of taxation applicable to interest and royalty payments made between associated companies of different member states as amended, consolidated text [2013] OJ L141, 30. 23 Council Directive 2009/133/EC of 19 October 2009 on the common system of taxation applicable to mergers, divisions, partial divisions, transfers of assets and exchanges of shares concerning companies of different member states and to the transfer of the registered office of an SE or SCE between member states as amended, consolidated text [2013] OJ L141, 30. 24 Council Directive (EU) 2016/1164 of 12 July 2016 laying down rules against tax avoidance practices that directly affect the functioning of the internal market [2016] OJ L193, 1–14. 25 Council Directive (EU) 2017/952 of 29 May 2017 amending Directive (EU) 2016/1164 as regards hybrid mismatches with third countries [2017] OJ L144, 1–11.

52 Rainer Prokisch extensive rules were needed and, consequently, ATAD II extended the scope to third countries. Finally, the EU aims to provide an effective system for resolving tax disputes over cross-border transactions. Already in 1990, the convention on the elimination of double taxation had been concluded but it was limited to transfer pricing disputes and related mutual agreements and arbitration procedures between member states (Union Arbitration Convention (UAC)).26 That convention was replaced by an EU directive on tax dispute-resolution mechanisms that has a broader scope of application and implies an improvement of the arbitration mechanism.27 It is also worth noting that the EU regulation of social security28 may be relevant for international taxation issues. Both fields of law are closely related and have impact, respectively. All EU law is subject to the EU Charter of Fundamental Rights.29 In respect of the directive on dispute resolution, article 47 of the EU Charter on fair trial protection can be relevant.

3.3 Exchange of Information and Cooperation of Tax Administrations As cross-border activities become more and more the norm, tax administrations have to cooperate in order to create a playing field on an equal footing with the private sector. In 1988, the OECD and the Council of Europe developed a multilateral instrument (Convention on Mutual Administrative Assistance in Tax Matters (CMAA) which was subject to a major amendment by a protocol published in 2010)30 that covers all forms of mutual tax cooperation in respect of the assessment and collection of taxes. Nowadays, there are very few countries worldwide that have not signed the Convention. Already by 1977, the EU had enacted a directive on mutual assistance in tax matters.31 That directive 26

Convention 90/436/EEC on the elimination of double taxation in connection with the adjustment of profits of associated enterprises—Final Act—Joint Declarations—Unilateral Declarations [1990] OJ L225, 10–24, as amended. 27 Council Directive (EU) 2017/1852 of 10 October 2017 on tax dispute resolution mechanisms in the European Union [2017] OJ L265, 1–14. 28 Consolidated text: Regulation (EC) No. 883/2004 of the European Parliament and of the Council of 29 April 2004 on the coordination of social security systems [2019] OJ L186, 21. 29 Charter of Fundamental Rights of the European Union [2012] OJ C326/02, 391–407). 30 OECD/Council of Europe, The Multilateral Convention on Mutual Administrative Assistance in Tax Matters: Amended by the 2010 Protocol (Paris: OECD Publishing, 2011), https://doi.org/10.1787/978926 4115606-en. 31 Council Directive 77/799/EEC of 19 December 1977 concerning mutual assistance by the competent authorities of the member states in the field of direct taxation [1977] OJ L336, 15–20). The directive was subject to several amendments, the latest consolidated version was published in 2007. Date of end of validity: 1 January 2013.

Sources of Law and Legal Methods in International Tax Law 53 was replaced by a new directive adopted in 2011 (DAC).32 The directive was subject to several amendments which were incorporated into the original document. Of utmost importance is the exchange of information between tax authorities. In the past, bank secrecy often hindered such exchange since tax authorities were not able to gain information from financial institutions. The exchange of information clauses contained in tax treaties had therefore been ineffective in that respect. The EU Savings Directive33 was a first attempt to regulate information exchange between EU member states without bank secrecy, on the one hand, and introduce of a type of minimum taxation of interest income in those states where bank secrecy existed, on the other hand. To the same end, the USA adopted the Foreign Account Tax Compliance Act (FATCA)34 which requires foreign banks and other financial institutions to provide certain information. In order to make the FATCA rules effective, the USA has concluded FATCA agreements with its most important trading partners. Concurrently, the OECD, at the request of the G20, developed the so-called common reporting standard (CRS). In 2014, the OECD member countries, as well as several other jurisdictions, committed themselves to implement the CRS (i.e. to automatically exchange relevant information). Other forms of providing information, mainly on request and spontaneously, make the exchange of information regime relatively comprehensive. As a consequence, in October 2014 the CMAA was supplemented by a CRS Multilateral Competent Authority Agreement (MCAA), which was signed by over sixty jurisdictions. In 2014, the EU also adopted the text of the CRS by amending the directive on administrative cooperation (DAC2).35 Another amendment contained rules on providing information about cross-border rulings and advance pricing arrangements.36 Further amendments referred to the exchange of country-by-country reports,37 beneficial ownership information that is collected on the basis of Directive (EU) 2015/849 (the Money Laundering Directive),38 and information on aggressive tax planning schemes.39 The latter EU directive also allows for carrying out simultaneous control of taxpayers

32 Council

Directive 2011/16/EU of 15 February 2011 on administrative cooperation in the field of taxation and repealing Directive 77/799/EEC [2011] OJ L64, 1–12. 33 Council Directive 2003/48/EC of 3 June 2003 on taxation of savings income in the form of interest payments [2003] OJ L157, 38, repealed by the Council on 10 November 2015. 34 The FATCA was passed as part of the HIRE Act of 2010. 35 Council Directive 2014/ 107/EU of 9 December 2014 amending Directive 2011/16/EU as regards mandatory automatic exchange of information in the field of taxation [2014] OJ L359, 1–29. 36 Council Directive (EU) 2015/2376 of 8 December 2015 amending Directive 2011/16/EU as regards mandatory automatic exchange of information in the field of taxation [2015] OJ L332, 1–10. 37 Council Directive (EU) 2016/ 881 of 25 May 2016 amending Directive 2011/16/EU as regards mandatory automatic exchange of information in the field of taxation [2016] OJ L146, 8–21. 38 Council Directive (EU) 2016/2258 of 6 December 2016 amending Directive 2011/16/EU as regards access to anti-money-laundering information by tax authorities [2016] OJ L342, 1–3. 39 Council Directive (EU) 2018/ 822 of 25 May 2018 amending Directive 2011/16/EU as regards mandatory automatic exchange of information in the field of taxation in relation to reportable cross- border arrangements [2018] OJ L139, 1–13.

54 Rainer Prokisch (DAC6). A new EU proposal intends to grant more room for ‘group requests’ and new disclosure rules are introduced relative to owners of digital platforms (DAC7).40

3.4 Other Sources Other sources that are not directly connected with taxation may nevertheless sometimes be relevant for international taxation. Although the imposition of tariffs on trading transactions in goods and services refers to indirect taxation and direct taxation is not clearly targeted by the General Agreement on Tariffs and Trade (GATT) and World Trade Organization (WTO) rules, those will, under specific circumstances, exclude direct tax concessions if they constitute a ‘subsidy’. Investment treaties or (older) treaties of friendship, commerce, and navigation41 may contain discrimination clauses which also cover tax issues. The European Convention on Human Rights42 and the jurisdiction of the European Court of Human Rights43 may also affect direct taxation. This is of specific relevance in countries whose judges are not allowed to challenge laws against the constitution, as is the case in the Netherlands. In this event, fundamental rights could be applied in many tax cases, for example protection of property or prohibition of discrimination. However, it seems that the general understanding of those rights in regard to their application on taxation is rather limited.44 The right to a fair trial, however, has resulted in several rulings against states in tax cases.

3.5 Methods in International Tax Law 3.5.1 General Aspects As with any law, tax law provisions have to be interpreted in order to discover their legal meaning. However, tax law in general has a long-standing special tradition in respect of interpretation and application. National constitutions also contain specific 40 European

Commission, Proposal for a Council Directive amending Directive 2011/16/EU on administrative Co-operation in the Field of Taxation, COM(2020)314final (15 July 2020); the Economic and Financial Affairs Council (ECOFIN) approved an amended version of the proposal on 25 November 2020; the Council adopted the directive on 22 March 2021. 41 e.g. Treaty of Friendship, Commerce and Navigation between the United States of America and the Federal Republic of Germany of 1955 (German BGBl. II 1956, 488). 42 https://www.echr.coe.int/documents/convention_eng.pdf. 43 https://www.echr.coe.int/Pages/home.aspx?p=home. 44 P. Baker, ‘Taxation and the European Convention on Human Rights’, British Tax Review (2000), 211–377.

Sources of Law and Legal Methods in International Tax Law 55 requirements in regard to tax rules. Tax claims require a legal basis, and the specific legal provision has to be precise and determined. Apart from those particularities, tax law, as with any other branch of law, is subject to the general methods of interpretation. There may be certain national traditions that lead to a stronger emphasis on the wording of the text or to a stronger influence of economic considerations. The fundamental principles of interpretation for international agreements, however, are not so different from those governing interpretation under domestic law.45 It is important to recognize that even if tax treaties are implemented into domestic law, international and domestic laws constitute separate law systems. Each system has different methods of how to apply certain concepts and how to interpret the terms concerned.

3.5.2 Interpretation of Tax Treaties 3.5.2.1 General Principles Tax treaties are international law conventions and the issue of interpretation of such conventions is regulated by articles 31–33 VCLT. Those provisions are regarded as an expression of customary international law.46 When the rules of the VCLT were created, treaty makers were confronted with a great divide between the objective and the subjective schools of thought. The VCLT made a clear choice in favour of the objective method.47 The objective method focuses on the text of the treaty, the wording is the main source of understanding and it also determines the limit of any interpretation. Of course, the reasonable sense of a term depends on the context and the object and purpose of the treaty as expressed in article 31(1) VCLT. In contrast, the intention of the parties is of minor importance, and may only be used if the intention can be derived indirectly from the object and purpose of a treaty provision. Intention has, however, found a place in article 31(4) VCLT which clarifies that the special meaning of a concept may be used if the parties so intended. This is especially of major importance in tax treaties since they use an ‘international tax language’ which frequently overrides the ordinary meaning of a word. The common intentions of the parties are expressed in the OECD Model and the UN Model as far as it can be

45 For more details, see K. Vogel and R. Prokisch, ‘The Interpretation of Double Taxation Conventions, General Report’, CDFI 78a (1993), 55–85, 57ff; R. F. van Brederode and R. Krever, eds, Legal Interpretation of Tax Law (Alphen aan den Rijn: Wolters Kluwer, 2014). 46 See, e.g., Pulau Ligitan and Pulau Sipadan (Indonesia v. Malaysia), ICJ Reports 2002, 645. 47 M. K. Yasseen, ‘L’interprétation des traités d’après la Convention de Vienne sur le droit des traités’, RCADI 151 (1976), 1ff; R. Kolb, The Law of Treaties (Northampton, MA: Edward Elgar, 2016), 131ff; F. G. Jacobs, ‘Varieties of Approach to Treaty Interpretation: With Special Reference to the Draft Convention on the Law of Treaties before the Vienna Diplomatic Conference’, International and Comparative Law Quarterly 18 (1969), 318–346.

56 Rainer Prokisch established that the contracting partners intended to follow those models. Accordingly, intention can play a greater role in bilateral tax treaties than in other international conventions. It should, however, be clear that substance-over-form approaches which have great merit when interpreting domestic law should only be used for the interpretation of tax treaty law with prudence. Further, it is obvious that it is not be possible to develop only one correct method of interpretation even in view of the rules of the VCLT. Persons interpreting a provision of a tax treaty are always part of a more or less distinct cultural environment.48 Context is a central concept of the interpretative process. First, it is addressed in the general rule of interpretation of article 31(1) VCLT and by doing this it refers to systematic interpretation (context in a narrow sense). Article 31(2) extends the context to agreements relating to the treaty and article 31(3) contains a further extension to documents created after the conclusion of a convention, such as subsequent agreements or subsequent practice (context in a broader sense). Other means of interpretation, for example the historical context or the preparatory work, are subsidiary (art. 32 VCLT) but will normally play a role in the course of the interpretative process. The rules of the VCLT do not hinder flexibility. In regard to tax treaties, article 3(2) OECD Model also uses the concept of context. This should be understood in a still broader sense also covering the Model and the Commentary on which the treaty is based, and even a mutual agreement concluded between the states’ tax authorities in regard to a certain treaty concept. Further, ‘the performance of treaties is subject to an overriding obligation of mutual good faith’.49 This means that a person who has to interpret a treaty should not take a unilateral perspective on tax treaty terms but, rather, a view that considers the interests and positions of the treaty partners.50 In particular, the interpreter should always ask themselves whether the other treaty partner would be expected to accept the interpretation in question. One of the consequences of the principle of good faith is the goal of achieving a ‘common interpretation’ as far as possible. Tax treaties aim to allocate tax claims equally between states and this can only be achieved if both parties apply the treaty consistently.51

3.5.2.2 The Reference to Domestic Law Tax treaties, as a rule, contain a specific interpretation provision. Article 3(2) OECD Model is a special rule of interpretation in relation to the general rules governing 48 S. A. Rocha, Interpretation of Double Taxation Conventions, General Theory and Brazilian Perspective (Alphen aan den Rijn: Wolters Kluwer, 2009). 49 A. D. McNair, The Law of Treaties (Oxford: Clarendon Press, 1961), 465. 50 E. Van der Bruggen, ‘Good Faith in the Application and Interpretation of Double Taxation Conventions’, British Tax Review 1 (2003), 25–68. 51 Vogel and Prokisch, ‘Interpretation of Double Taxation Conventions’, 62ff; D. Gosch, ‘Über die Auslegung von Doppelbesteuerungsabkommen’, ISR (2013), 87–95, 91finds that there is no legal basis for the common interpretation principle (‘Entscheidungsharmonie’); H. Hahn, ‘Gedanken zum Grundsatz der sog. Entscheidungsharmonie’, IStR (2012), 941 is also critical.

Sources of Law and Legal Methods in International Tax Law 57 interpretation of tax treaties. As such, it takes precedence over the general rules.52 Article 3(2) allows the application of domestic law if a treaty term is not defined in the treaty itself unless the context requires otherwise. In 1995, a clarification was added to the provision concerning the appropriate point in time and a rule on the ranking of different meanings of a term in the law of the state applying the treaty. The scope of the rule, however, is rather limited. It only concerns the interpretation of terms that are defined in domestic law pursuing the same purpose as the treaty term and it does not refer to general principles or legal fictions. In 2017, the OECD changed article 3(2) OECD Model by providing that a domestic law meaning should not be given to an undefined treaty term whose meaning has been agreed through the mutual agreement procedure pursuant to the provisions of article 25 OECD Model. This addition does not really imply a modification of the substance of the provision. Mutual agreements are only binding on the tax authorities that have concluded the agreement. Such agreements do not become law as they are not ratified to conform to constitutional requirements. They are neither ‘subsequent agreements’ in terms of article 31(3)(a) VCLT nor do they indicate a ‘special meaning’ as used in article 31(4). Accordingly, courts may come to a different interpretation and ignore a differing mutual agreement. Courts should be expected, however, to take the content of a mutual agreement seriously as being part of the context pursuant to article 3(2) OECD Model. Further, the provision is subject to other different understandings which may lead to different results. The root of the debate is the fact that the provision stems from Anglo-American law but also has to be applied by continental law jurisdictions.53 Being included in treaty models, the rule may acquire an ‘international meaning’ not dependent on the meaning it had in the law of the original countries. The debate refers to the question of which of the contracting partners ‘applies’ the treaty54 and whether the sentence ‘unless the context requires otherwise’ has an influence on the priority of domestic law interpretation over autonomous treaty interpretation or vice versa. Traditionally, contracting partners assumed that both contracting states apply the treaty.55 This has one major negative consequence since in many instances both states will attach different meanings to the concept and this may lead to double taxation or double non-taxation. Such understanding, therefore, fails to reach the object of tax treaties to avoid double and double non-taxation and is contrary to the purpose of the treaty to express the common understanding of the treaty partners. Nevertheless, the tax authorities and courts of many countries still follow this approach. 52 Vogel, Klaus Vogel on Double Taxation Conventions, Introduction, m.nos 58ff; art. 3, m.no. 61. 53

J. F. Avery Jones, ‘Treaty Interpretation—Global Tax Treaty Commentaries’ (accessed 1 June 2019), Global Topics IBFD (accessed 3 April 2021); J. F. Avery Jones, ‘A Fresh Look at Article 3(2) of the OECD Model’, Bulletin for International Taxation 74 (2020), 11. 54 See A. Rust, ‘Article 3’, in E. Reimer and A. Rust, eds, Klaus Vogel on Double Taxation Conventions, 4th ed. (Alphen aan den Rijn: Kluwer Law International, 2015), paras 116–117. 55 Cf. Vogel, Klaus Vogel on Double Taxation Conventions, art. 3, m.no. 65: any ‘tax question for which the treaty is considered or should be considered’.

58 Rainer Prokisch The OECD, on the basis of thoughts brought forward by Déry and Ward56 and Avery Jones,57 assumes that in the case of differences in domestic law the source state still taxes the income in accordance with the provisions of the treaty by interpreting and applying the treaty concept in question.58 The state of residence is consequently obliged to accept the understanding of the source state and to avoid double taxation pursuant to articles 23A and 23B OECD Model. Under the OECD approach, however, the reference to domestic law cannot claim priority over an alternative interpretation that follows from the context of the convention.59 Consequently, if the context—including the established intention of the parties under consideration of the meaning given to the term in the other contracting state (reciprocity)60—supports an alternative or autonomous interpretation, such meaning should be used for the application of the treaty. Other authors give priority to the second clause, ‘unless the context requires otherwise’.61 As a rule, an autonomous interpretation is possible and should be used either because the second clause has priority over the reference to domestic law or because the balancing of both interpretation results speaks in favour of an alternative interpretation. The OECD Model Commentary also uses this approach. Quite often, treaty terms are interpreted autonomously although the term as such is not defined in the treaty. It seems, therefore, that article 3(2) OECD Model has an international meaning which deviates from the original understanding in common law countries. An interpretation from the context has the advantage that a treaty term is possibly understood in both contracting states (or even worldwide) in the same way by exchanging the relevant arguments. Tax treaty interpretation calls for cooperation among all interpreters involved with the goal of arriving at a common understanding.

3.5.2.2.1 The role of model conventions and their commentaries The OECD Model and its Commentaries as source of interpretation It is generally acknowledged that the OECD models and especially the extensive explanatory Commentaries thereto are a most important source of interpretation of tax treaties. A substantial reason for this development is the need for a common understanding of

56 J.-M. Déry and D. A. Ward, ‘Interpretation of Tax Conventions: Canada National Report’, CDFI 78a (1993), 259–293. 57 J. F. Avery Jones, ‘Interpretation of Tax Conventions: United Kingdom National Report’, CDFI 78a (1993), 597–614. 58 See OECD Commentary on arts 23A and 23B, para. 32.3; cf. A. Nikolakakis, ‘Interpretation vs Qualification’, in G. Maisto, ed., Current Tax Treaty Issues (Amsterdam: IBFD, 2020), 319–357. 59 OECD Commentary on art. 3, para. 12. In contrast, Avery Jones endorses priority of the reference to domestic law. He argues that the concepts ‘requires’ and ‘unless’ describe a narrow exception (‘A Fresh Look at Article 3(2) of the OECD Model’, Bulletin for International Taxation, 74/11 (2020)). 60 OECD Commentary on art. 3, Para. 12. 61 M. Lang, ‘Qualification Conflicts— Global Tax Treaty Commentaries’ (accessed 6 August 2020), Global Topics IBFD; M. Lang, ‘Tax Treaty Interpretation— A Response to John F. Avery Jones’, Bulletin for International Taxation 74/11 (2020), 7; D. Gosch, ‘Über die Auslegung von Doppelbesteuerungsabkommen’, ISR (2013), 87–95, 88ff.

Sources of Law and Legal Methods in International Tax Law 59 treaty concepts.62 It is a matter of fact that at least the OECD member countries take the models as the basis of their own individual tax treaties as far as possible and that the courts, as well as the tax authorities, use the Commentaries frequently when they have to apply a treaty provision. The exact relationship, however, between the Commentaries and the actual treaties which are negotiated on the basis of the Model is still a much- debated question, and on which we are far from a consensus. In particular, it is quite unclear whether the approach to consider the Commentaries in the interpretative process of a certain treaty is a mere matter of state practice or whether there exists a legal foundation implying the necessity of the Commentaries’ consideration.63 It is obvious that the wording of articles 31–32 VCLT does not fit the Commentaries well. They are not part of the context as used by article 31 VCLT since it covers only the internal context of a treaty provision and not any external documents.64 They are certainly not an instrument that was made in connection with the conclusion of the treaty (art. 31(2)(b) VCLT), they are used solely as a basis for negotiations.65 Neither are they a supplementary means of interpretation in terms of article 32 VCLT since they are not made in connection with an individual treaty. A different approach refers to the doctrines of acquiescence and estoppel as general principles of international law. Insofar as OECD member countries have not made an observation on the interpretation of the provisions as made by the Commentaries and they have not expressed a different view when negotiating a new or revised treaty, they are supposed to ‘have acquiesced’ in the meaning indicated by the Commentaries so as to become legally bound by it. As a consequence, the Commentaries would become part of the context in terms of article 31(2)(a) VCLT.66 This opinion has not found support

62 Cf. para. 15 Introduction to the OECD Model 2017: ‘This [the Commentaries] has facilitated the interpretation and the enforcement of these bilateral conventions along common lines’; M. Lang, ‘Die Bedeutung des Musterabkommens und des Kommentars des OECD-Steuerausschusses für die Auslegung von Doppelbesteuerungsabkommen’, in W. Gassner, M. Lang, and E. Lechner, eds, Aktuelle Entwicklungen im Internationalen Steuerrecht (Vienna: Linde Verlag, 2000), 11–41. 63 See, e.g., A. J. van den Tempel, Relief from Double Taxation (Amsterdam: IBFD, 1967), 25; E. B. Switzer, ‘Exchange of Information Articles’, Canadian Tax Journal 26 (1978), 313; K. van Raad, ‘Interpretatie van belastingverdragen’, MBB 47 (1978), 55; J. F. Avery Jones et al., ‘The Interpretation of Tax Treaties with Particular Reference to Article 3(2) of the OECD Model’, British Tax Review (1984), 14ff. 64 Avery Jones et al., ‘The Interpretation of Tax Treaties’, 91ff, however, think that it would be possible to argue that the Commentaries were part of the Vienna Convention context; similarly, E. A. Driedger, The Construction of Statutes (Toronto: Butterworths, 1974), 83ff, who uses the concept of ‘intellectual context’, i.e. matters of common knowledge including commission reports, international agreements, and parliamentary debates not referred to in the statute itself. 65 The Commentary to the Vienna Convention (para. 13 ILC Final Draft, Commentary to art. 27) also speaks against such a broad understanding. The purpose of art. 31(2) VCLT has been to make it clear that a unilateral document cannot be regarded as forming part of the context unless the document is signed by both parties. 66 F. Engelen, Interpretation of Tax Treaties under International Law: A Study of Articles 31, 32 and 33 of the Vienna Convention on the Law of Treaties and Their Application to Tax Treaties (Amsterdam: IBFD, 2004); F. Engelen, ‘Some Observations on the Legal Status of the Commentaries on the OECD Model’, Bulletin for International Taxation 70 (2006), 105–109.

60 Rainer Prokisch in the literature or in the courts. There is no hard legal obligation that OECD member states have to react on certain interpretations in the Commentaries although it may be wise to do so. The recommendation of the OECD to the member states is limited to a rather weak suggestion to follow the Commentary.67 However, the ‘ordinary meaning’ of tax treaty terms is not necessarily that of everyday usage. As far as an ‘international tax language’68 has developed—widely expressed by the models and their Commentaries—the tax terms determine the ‘ordinary’ meaning of article 31(1) VCLT69 or, alternatively, the ‘special’ meaning of article 31(4).70 Therefore, when the negotiators followed the Model without any material change, it could be presumed that they intended to follow the Commentaries as well. The person who interprets a treaty concept can assume that the concept was understood by the treaty partners in line with the explanation by the Commentaries as long as the contracting states did not indicate that they found that assumption to be incorrect. However, since the OECD publishes the Model and the Commentaries in the form of a loose-leaf edition, and since it has become extremely difficult to identify which version of the OECD Model was used by treaty negotiators, the general assumption has been ‘destroyed’.71 Nevertheless, the Commentaries are certainly still an important source of interpretation similar to the opinions of high-ranking experts. Changes of Commentaries and their effect on previously concluded treaties The OECD Commentary has been amended from time to time, until 1992 occasionally and since then flexibly when it has been felt that amendments have become necessary. Such amendments raise the question whether they also have an impact on older treaties that were concluded before the relative amendment. The OECD is of the opinion that ‘existing conventions should, as far as possible, be interpreted in the spirit of the revised Commentaries’.72 Most authors and courts, however, contest this approach. The main argument is that tax treaties become domestic law by implementation or order as defined by the constitution. The understanding of a law provision cannot be determined by a statement that is made by an organization that is not part of the constitutional system of a nation state.73 The explanation of the role of the Commentaries described earlier 67 Cf.

D. A. Ward, ‘The Role of the Commentaries on the OECD Model in the Tax Treaty Interpretation Process’, Bulletin for International Taxation 70 (2006), 97–102. 68 This phrase was used by the High Court of Australia, 22 August 1990 in Thiel v. FCT, 21 ATR 531, 537 (1990); see also R. Prokisch, ‘Does It Make Sense If We Speak of an “International Tax Language”?’, in K. Vogel, ed., Interpretation of Tax Law and Treaties and Transfer Pricing in Japan and Germany (The Hague: Kluwer Law International, 1998), 103). 69 R. Prokisch, ‘Fragen der Auslegung von Doppelbesteuerungsabkommen’, SWI 4 (1994), 52. 70 H. G. Ault, ‘The Role of the OECD Commentaries in the Interpretation of Tax Treaties’, in H. Alpert and K. van Raad, eds, Essays on International Taxation: To Sidney I. Roberts (Deventer: Kluwer, 1993), 61, 65. 71 K. Vogel, ‘The Influence of the OECD Commentaries on Treaty Interpretation’, Bulletin for International Taxation 64 (2000), 612–616. 72 OECD Commentary 2017, Introduction, paras 33ff. 73 See M. Lang, ‘Haben die Änderungen der OECD—Kommentare für die Auslegung älterer DBA Bedeutung?’, SWI 5 (1995), 412–416.

Sources of Law and Legal Methods in International Tax Law 61 also hinders a ‘retroactive’ application of the Commentaries. If they are relevant for the determination of the ordinary or special meaning of terms as used in tax treaties, such meaning can only be the meaning at the time the treaty was concluded. Nevertheless, later Commentaries should at least be considered when a person interprets a treaty. The weight that will be given to the Commentary will depend on the circumstances,74 in particular on the reason for the change and whether the expectations of taxpayers will be frustrated. If later changes serve as clarification or if they harmonize the approaches used in a state’s practice, the Commentaries should be decisive in the case of doubt or uncertainty.

3.5.2.2.2 The language issue International treaties are generally concluded in the languages of both contracting parties if those states do not share the same language. As a rule, both versions are authoritative so that a judge has to consider both when interpreting the treaty (art. 33 VCLT). Sometimes states agree that a third language—usually English or French— shall be binding where the original language versions show differences and create uncertainties. The language of the treaty negotiations is not relevant. Discrepancies in meaning may arise. Article 33(4) VCLT stipulates in such cases that interpretation should follow the text which best reconciles both or all texts.75 If such an interpretation cannot be found, the contradiction leads to the non-application of the treaty.76

3.5.2.2.3 Administrative agreements and arbitration The mutual agreement procedure under article 25 OECD Model provides a useful mechanism for settling tax issues arising from the application or interpretation of a tax treaty. Such agreements are ‘international conventions’ but are normally not integrated into domestic law as they are not ratified in accordance with the requirements of constitutional law. However, article 25, in particular paragraph 3, can also be understood as a delegation of legislative power to the tax authorities of the contracting states.77 Most countries, however, do not accept this qualification since constitutional requirements on the determination and precision of delegation provisions are not met. In many countries, the binding effect on courts would therefore presume a sufficient statutory basis. Consequently, it would be desirable for the contracting partners to include such 74 J. F. Avery Jones, ‘The Effect of Changes in the OECD Commentaries after a Treaty is Concluded’, Bulletin for International Taxation 66 (2002), 102–109. 75 P. Arginelli, Multilingual Tax Treaties: Interpretation, Semantic Analysis and Legal Theory (Amsterdam: IBFD, 2015); R. X. Resch, The Interpretation of Plurilingual Tax Treaties (Hamburg: tredition, 2018). 76 For examples, see E. Reimer and A. Rust, Klaus Vogel on Double Taxation Conventions, Introduction, m.no. 88. 77 J. Sasseville, ‘The 2017 Change to Article 3(2) of the OECD Model: Comments on Professor Alexander Rust’s Presentation’, Bulletin for International Taxation 74/4/5 (2020) assumes that mutual agreements have binding force since art. 25 OECD Model delegates to the competent authorities the power to resolve difficulties or doubts by concluding agreements with the tax authorities of the other contracting state.

62 Rainer Prokisch agreements to the treaty itself (as an amendment or additional protocol). In any event, mutual agreements (and arbitration decisions) will be a valuable source for the person who has to interpret a treaty term. It is crucial in this respect that all types of rulings are publicly accessible but, unfortunately, article 25 OECD Model does not urge contracting states to publish such agreements.78

3.5.2.2.4 The multilateral instrument (MLI): interpretation and application The OECD Multilateral Convention (MLI)79 aims to implement the BEPS measures into tax treaties as far as the BEPS recommendations refer to tax treaties. In 2017, sixty-seven countries signed the convention. The MLI does not change the basic principles of the interpretation of tax treaties. However, if a treaty concept has its origin in the MLI, a multi-stage interpretation process makes the interpretation more complex and requires special attention from the interpreter.80 Still, the starting point of interpretation will be the tax treaty as amended by the MLI. As far as a concept originates in the MLI, article 2(2) MLI refers to the specific interpretation provisions of tax treaties (art. 3(2) OECD Model). This implies integration of both legal instruments. The MLI and the underlying BEPS documents become part of the context of article 3(2) OECD Model.

3.6 Conclusions Sources of international tax law are diverse and spread over numerous laws and official documents. A categorization covers three pillars: domestic law, international law, and European law. A speciality of international tax law are model conventions that serve as a basis for treaty negotiations and, together with their Commentaries, lead to a certain harmonization of tax treaties and have a major influence on the application and interpretation of treaties. As with any branch of law, international tax law is subject to a hermeneutic approach. Being international law, the international rules on treaty interpretation apply. However, tax treaties contain a specific treaty interpretation provision which

78

See R. Ismer, in Reimer and Rust, Klaus Vogel on Double Taxation Conventions, art. 25, para. 74. Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting, signed on 7 June 2017. The MLI is accompanied by an Explanatory Statement. 80 R. Prokisch, ‘Die Auslegung von DBA im Lichte des Multilateralen Abkommens’, in R. Ismer et al., eds, Territorialität und Personalität, Festschrift für Moris Lehner zum 70. Geburtstag (Cologne: Otto Schmidt Verlag, 2019), 195–208; R. Prokisch and F. Souza de Man, ‘Multilateralism and International Tax Law: The Interpretation of Tax Treaties in Light of the Multilateral Instrument’, in A. P. Dourado, ed., International and EU Tax Multilateralism: Challenges Raised by the MLI, (Amsterdam: IBFD, 2020), 199– 225; S. Wakounig, ‘Interpretation of Terms Used in the Multilateral Instrument’, in M. Lang et al., eds, The OECD Multilateral Instrument for Tax Treaties (Alphen aan den Rijn: Wolters Kluwer, 2018), 21–41. 79

Sources of Law and Legal Methods in International Tax Law 63 bridges the gap between international and domestic law. The application of this rule is still subject to debate and a common approach still needs substantial efforts by the international (tax) community. Model conventions and their commentaries play an important role in the interpretation process although their legal status still needs clarification.

Chapter 4

Ju risdict i ona l Underpinn i ng s of International Tax at i on H. David Rosenbloom and Fadi Shaheen

The modest purpose of this chapter is to summarize certain principles of public international law and raise related questions that may be useful to consider when efforts are made to reach a broad consensus on changes to the rules of international taxation.1 We approach this task as tax scholars trying to understand principles of public international law and their implications for jurisdiction to tax. We are mainly interested in the question whether, from a public international law standpoint, existing domestic tax laws and bilateral income tax treaties can rise to the level of customary international law and, as such, impose limitations on jurisdiction to tax. To explain why we highly doubt such a conclusion, it is necessary first to consider basic principles of jurisdiction.

4.1 Basic Principles The Restatement (Fourth) of the Foreign Relations Law of the United States defines jurisdiction as ‘the authority of a state to make, apply, and enforce law’.2 This definition comprises three separate aspects of national jurisdiction: jurisdiction to prescribe, jurisdiction to adjudicate, and jurisdiction to enforce. The scopes of these aspects of jurisdiction may, but do not necessarily, match. A state may have jurisdiction to prescribe—that is, to make law applicable to subject matters (persons, things, or conducts)—without having jurisdiction to apply that law through its court system (jurisdiction to adjudicate)

1

2

This chapter was written in the beginning of 2021. Restatement (Fourth) of the Foreign Relations Law of the United States, Part IV, Intro. Note (2018).

66 H. David Rosenbloom and Fadi Shaheen or to enforce that law using its general tools of enforcement (jurisdiction to enforce).3 Another state may have jurisdiction to apply or enforce the law. A state may exercise prescriptive jurisdiction if it has a ‘genuine connection’ to the subject matter. Although the Permanent Court of International Justice appeared to suggest in the Lotus case that any exercise of prescriptive jurisdiction is permitted in the absence of a prohibition,4 states generally seek an accepted basis for a genuine connection. The most common, but non-exhaustive, bases for genuine connection are territoriality, effects, active and passive personality, protective concerns, and universal concerns. Naturally, a state has jurisdiction to prescribe law with respect to subject matter within its territory. Prescriptive jurisdiction may extend, however, to extraterritorial conduct that has substantial effects within the prescribing state’s territory; and jurisdiction may extend as well, on the basis of active personality, to extraterritorial conduct, interests, status, and relations of nationals, domiciliaries, or residents, which include natural persons and corporations. The passive personality, protective, and universal bases of prescriptive jurisdiction relate mainly to crimes and acts of terrorism and are not directly relevant here. States may have concurrent prescriptive jurisdiction with respect to the same subject matter and, from a public international law perspective, there is no hierarchy of bases of prescriptive jurisdiction. States address undesirable consequences of concurrent prescriptive jurisdiction either unilaterally, as a matter of international comity, through domestic law, or through bilateral or multilateral treaties. It appears, therefore, that from a public international law perspective, jurisdiction to prescribe is broad enough to cover any reasonable tax law. It is hard to imagine a tax that would be seriously considered by any nation and that would fall outside the scope of standard prescriptive jurisdiction. A more interesting question is whether there are other international law limitations on a state’s power to prescribe when it has a genuine connection to the subject matter. An obvious practical, though not substantive, limitation is that of jurisdiction to enforce. As a matter of public international law, there is no limit on a state’s exercise of jurisdiction to enforce within its territory, although a state may not exercise jurisdiction to enforce its law if it had no jurisdiction to prescribe that law. Even if a state has jurisdiction to prescribe a law with extraterritorial effect, the state has no jurisdiction to enforce that law in the territory of another state without the consent of that other state. This rule was another aspect of the Lotus decision, where the Permanent Court of International Justice said: [T]‌he first and foremost restriction imposed by international law upon a State is that—failing the existence of a permissive rule to the contrary—it may not exercise its power in any form in the territory of another State. In this sense jurisdiction is

3

4

Reference to a ‘state’ here is to a nation state. SS ‘Lotus’ (France v. Turkey), 1927 PCIJ, Ser. A, No. 10, at 19 (7 September).

Jurisdictional Underpinnings of International Taxation 67 certainly territorial; it cannot be exercised by a State outside its territory except by virtue of a permissive rule derived from international custom or from a convention.5

This restriction helps to explain why, rather than directly taxing controlled foreign corporations (CFCs), taxing jurisdictions typically include covered income of a CFC in the taxable base of shareholders they consider to be residents.6 There appears to be a clear and legitimate connection between the home country of a shareholder and the shareholder’s CFCs that justifies prescribing a law directly taxing income of the CFCs.7 The effects basis appears to be on point here, but the list of bases for genuine connections is long and those bases seem to derive from a broader principle, as Jennings and Watts and others have argued: Although it is usual to consider the exercise of jurisdiction under one or other of more or less widely accepted categories, this is more a matter of convenience than of substance. There is, however, some tendency now to regard these various categories as parts of a single broad principle according to which the right to exercise jurisdiction depends on there being between the subject matter and the state exercising jurisdiction a sufficiently close connection to justify that state in regulating the matter and perhaps also to override any competing rights of other states.8

It would be hard to imagine a valid argument to the effect that the connection between a taxing jurisdiction and the CFCs of its resident shareholders is not ‘sufficiently close to justify’ the imposition of a tax directly on the CFCs. After all, a corporation is merely a sheet of paper, and a CFC is merely a foreign corporation controlled by local residents.9 The problem with such a tax would be that the taxing jurisdiction would have no jurisdiction to enforce it because a CFC is foreign with respect to the taxing jurisdiction and CFC rules are concerned with income derived outside the territory of the taxing jurisdiction. Because CFCs have limited liability, collecting tax from a CFC’s shareholders in the taxing jurisdiction requires that the CFC’s tax liability be imputed to shareholders. Such a design would be practically and conceptually less appealing than the current typical structure of CFC rules. In the absence of a general multilateral treaty, the rules governing jurisdiction to prescribe were developed as customary international law. Other sources of public international law include international agreements and general principles common to the 5

SS ‘Lotus’. S. Dueñas, ‘CFC Rules Around the World’, Tax Foundation Fiscal Fact No. 659 (June 2019); P. Baker, ‘Some Thoughts on Jurisdiction and Nexus, in G. Maisto, ed., Current Tax Treaty Issues, 50th Anniversary of the International Tax Group (Amsterdam: IBFD, 2020), 441, 443. 7 See R. S. Avi-Yonah, International Tax as International Law: An Analysis of the International Tax Regime (Cambridge: Cambridge University Press, 2007), 24–27. 8 R. Jennings and A. Watts, Oppenheim’s International Law, 9th ed. (London: Longman, 1992), 457– 458; Restatement (Fourth) of the Foreign Relations Law of the United States, § 407, Reporters’ Note 2 (2018) and references there. 9 See S. I. Roberts, ‘From the Thoughtful Tax Man’, Taxes 40/5 (1962), 355. 6 See

68 H. David Rosenbloom and Fadi Shaheen major legal systems.10 Although customary law and treaties are primary sources of international law, general principles are secondary. Therefore, it appears that general principles common to the major legal systems that have not risen to the level of customary or treaty law may not derogate from customary rules of prescriptive jurisdiction. This would explain why, in the absence of customary law or treaties, it would be hard to imagine a valid jurisdictional argument against novel taxes, such as digital services taxes.11 A treaty is binding on a state under public international law regardless of the treaty’s status from a domestic law standpoint,12 which may vary from state to state depending on each state’s constitutional and legal system. Income tax treaties allocate taxing rights between treaty partners and, as such, limit the partners’ jurisdiction to tax from a public international law perspective. ‘The maxim lex specialis derogat generali (the specific prevails over the general) is an accepted principle of international law.’13 A bilateral income tax treaty is clearly more specific than customary international law in regard to jurisdiction to prescribe, for two independent and separately sufficient reasons. Rules of prescriptive jurisdiction apply generally, as default rules, with respect to all fields of law, and income tax treaties apply only with respect to tax law. Moreover, the rules of prescriptive jurisdiction apply generally to all states, and bilateral income tax treaties apply only with respect to the treaty partners. Therefore, as a matter of public international law, allocations of taxing rights under income tax treaties must prevail over the typically more permissive rules relating to jurisdiction to prescribe. Article 38(1) of the Statute of the International Court of Justice, which is ‘commonly treated as an authoritative statement of the “sources” of international law’,14 lists treaties as the first of four sources of international law, and customary law second. Even though article 38(1) ‘does not provide for a hierarchy among sources, the priority of position given to treaties reflects the understanding [that it is] “natural” to apply a treaty in the first instance’.15 Still, the rule that the specific prevails over the general (which works in favour of tax treaties vis-à-vis customary international law governing jurisdiction to prescribe) may give priority to a custom that is more specific than a treaty. Also, a new rule of customary law might override a conflicting existing treaty if the parties’ intention

10

Statue of the International Court of Justice, art. 38(1). arguments why digital services taxes are not covered by treaties, see G. Kofler and J. Sinnig, ‘Equalization Taxes and the EU’s “Digital Services Tax” ’, Intertax, 47/2 (2019), 176, 195, and references there; F. Shaheen, ‘Income Tax Treaty Aspects of Nonincome Taxes: The Importance of Residence’, Tax Law Review 71/3 (2018), 583, 620ff, explaining why the residence-relevance element weighs heavily against treaty coverage of digital services taxes. 12 Vienna Convention on the Law of Treaties, art. 46, 23 May 1969, 1155 UNTS 331 (VCLT). 13 L. Fisler Damrosch and S. D. Murphy, International Law, Cases and Materials, 7th ed. (St Paul, MN: West Academic, 2019), 11. 14 Restatement (Third) of the Foreign Relations Law of the United States, §102, Reporters’ Note 1 (1987). 15 Damrosch and Murphy, International Law, 111. 11 For

Jurisdictional Underpinnings of International Taxation 69 to that effect is clearly manifested.16 Either way, such a custom might expand jurisdiction to tax under a treaty while cutting back on the general prescriptive jurisdiction rules, cut back on both, or—at least theoretically—expand both.

4.2 Can Tax Rules Rise to the Level of Customary International Law? It is accepted that two elements are required for a rule of customary international law to form: (1) a general and consistent practice of states; (2) opinio juris (i.e. that the practice is followed by states from a sense of legal obligation (opinion juris sive necessitates)).17 Clearly, the ‘practice of states’ includes unilateral tax legislation and tax treaties; it may also include tax initiatives taken in cooperation within an institution such as the UN or the OECD.18 However, such practices must be ‘general and consistent’. A state practice need not be universally or absolutely followed; it should be ‘extensive and virtually uniform’ among the states whose interests are affected.19 A practice that is followed by a limited number of states can satisfy the first element in the development of ‘particular customary law’, which, unlike general customary law, is binding only on the participating states. In any event, a state is not bound by a rule of customary international law if it declares its dissent during development of the rule. It is true that domestic tax laws around the world, especially those concerning cross- border taxation, are based on similar principles. But even rules that are based on similar principles still vary significantly, and the question is whether such laws constitute general and consistent practices of states. Tax treaties vary too, though they might be easier

16 As a matter of international law, a treaty might also be overridden by a conflicting peremptory norm of general international law (‘jus cogens’) (arts 53 and 64 VCLT). Under art. 53 VCLT, ‘a peremptory norm of general international law is a norm accepted and recognized by the international community of States as a whole as a norm from which no derogation is permitted and which can be modified only by a subsequent norm of general international law having the same character’. This does not appear to be relevant in the context of taxation. 17 See, e.g., North Sea Continental Shelf Cases (Federal Republic of Germany v. Denmark; Federal Republic of Germany v. Netherlands), ICJ Reports 1969 3, 44. Article 38(1)(b) of the Statute of the International Court of Justice refers to ‘international custom, as evidence of a general practice accepted as law’ as a source of international law. The Restatement (Third) of the Foreign Relations Law of the United States, §102(2) describes customary international law as the law that ‘results from a general and consistent practice of states followed by them from a sense of legal obligation’. 18 The ‘practice of states’ includes ‘diplomatic acts and instructions as well as public measures and other governmental acts and official statements of policy, whether they are unilateral or undertaken in cooperation with other states, for example in organizations such as the Organization for Economic Cooperation and Development (OECD)’. Restatement (Third) of the Foreign Relations Law of the United States, §102, Comment b. 19 North Sea Continental Shelf Cases, 43; Case Concerning the Military and Paramilitary Activities in and against Nicaragua (Nicaragua v. United States of America), ICJ Reports 1986 14, 88.

70 H. David Rosenbloom and Fadi Shaheen to qualify as a general and consistent practice of states. Treaties are an independent source of international law with respect to the treaty partners, and treaty law may develop into customary international law if the requirements are met.20 If that occurs, states may no longer withdraw from such obligations (even if they withdraw from a treaty), and obligations become binding on states that are not parties to the treaty.21 Given the large network of similar bilateral income tax treaties, it is plausible that certain tax treaty provisions would satisfy the first element of customary international law. A similar argument could be made with respect to guidelines promulgated pursuant to certain OECD tax initiatives, such as the transfer pricing guidelines and certain action plans of the BEPS Project.22 Those are interesting questions whose value may remain only theoretical given the difficulty in satisfying the opinio juris element with respect to any of the noted tax instruments. OECD guidelines are considered ‘soft law’ rather than binding obligations, and typically do not reflect opinio juris. As for treaties, the International Court of Justice explained that it is possible to treat a treaty provision as: a norm-creating provision which has constituted the foundation of, or has generated a rule which, while only conventional or contractual in its origin, has since passed into the general corpus of international law, and is now accepted as such by the opinio juris, so as to have become binding even for countries which have never, and do not, become parties to the [c]‌onvention. There is no doubt that this process is a perfectly possible one and does from time to time occur: it constitutes indeed one of the recognized methods by which new rules of customary international law may be formed. At the same time this result is not lightly to be regarded as having been attained.23

It is true that treaties are binding on the parties to them and, as such, are followed from a sense of legal obligation; but that is a matter of pacta sunt servanda—that is, that a ‘treaty in force is binding upon the parties to it and must be performed by them in good faith’24—not a matter of opinio juris.25 Administrative cooperation aside, bilateral tax treaties allocate taxing rights between treaty partners. Different treaties may allocate taxing rights differently—both quantitatively and, to a lesser extent, qualitatively. In so

20 North Sea Continental Shelf Cases, 41. See also Restatement (Third) of the Foreign Relations Law of the United States, §102(3) (‘International agreements . . . may lead to the creation of customary international law when such agreements are intended for adherence by states generally and are in fact widely accepted’). A treaty could also codify, develop, or crystallize customary international law. North Sea Continental Shelf Cases, 38. 21 North Sea Continental Shelf Cases, 41. 22 Initiatives that have been finalized by the time of writing do not appear relevant to the question of jurisdiction to tax. 23 North Sea Continental Shelf Cases, 41. 24 Art. 26 VCLT. 25 But cf. R. S. Avi- Yonah, ‘Does Customary International Tax Law Exist?’, Law & Economics Research Paper Series, Paper No. 19-005 (2019).

Jurisdictional Underpinnings of International Taxation 71 doing, tax treaties may operate as specific exceptions to the more generally applicable domestic tax laws of each treaty partner. A treaty rule may be consistent with, or similar to, a domestic law rule, but the rules may also differ significantly. An item of income might be fully taxable under the domestic laws of a state but partially or fully exempt under a treaty rule if the income is earned by a resident of a treaty partner. A treaty method of relief from double taxation (say, exemption) might differ from the domestic law method (say, foreign tax credit) that would apply if the treaty did not apply. It would be hard to maintain that bilateral treaties are entered into and followed from a sense of a legal obligation while the more generally applicable domestic rules are in place and the treaty rules operate as specific exceptions. One can further wonder whether ‘the need for such bilateral treaties arise[s]‌from the absence of a multilateral treaty on the subject, which in turn actually suggests a lack of acceptance by states of global rules on the matter’.26 The Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting (MLI) does not appear to undermine the validity of this question. It appears that the MLI does not codify pre-existing customs, nor does it create new norms of customary international law.27 It also does not appear to reflect a crystallization of a forming custom. The MLI is a multilateral instrument for the efficient modification of existing bilateral tax treaties and does not appear to have any effect on the treaties of states not parties to the MLI. Those modifications are also not necessarily uniform. Furthermore, article 30 of the MLI provides that its provisions are without prejudice to subsequent modifications to individual bilateral treaties. Finally, ‘the nearly incomprehensible volume of reservations and diverging options—as a consequence of the political compromise [the MLI] establishes—suggests that its content is not capable of the necessary general application’.28 If it would be hard to maintain that tax treaties satisfy the opinio juris requirement, this would be all the more true with respect to domestic tax legislation.

26

Damrosch and Murphy, International Law, 87–88 and references there. D. W. Blum, Normativity in International Tax (Amsterdam: IBFD, 2021). 28 Ibid. 27

Chapter 5

Internationa l Tax L aw and Cu stoma ry Internationa l L aw Elizabeth Gil García

5.1 Introduction Customary international law (CIL) refers both to the process through which certain rules of international law are formed and the rules themselves formed through such a process. No consensus has been reached, however, on its concept.1 The fact that custom is an intangible, unwritten source makes the topic debatable. Some authors consider that CIL arises spontaneously from the decentralized practices of states and that there is no clear identification criteria.2 Nevertheless, article 38 of the Statute of the International Court of Justice (ICJ Statute) specifies the two constituent elements of a customary rule when establishing the sources of international law. Moreover, the International Law Commission (ILC) has shed light on how CIL rules are identified. Traditionally, the literature has paid relatively little attention to custom in international tax matters, probably because international tax law (if it exists as such) is treaty- based. The use of double taxation conventions (DTCs) as (bilateral) mechanisms to prevent double taxation has given rise to a broad network of tax treaties. This latter network serves as a basis for an international tax regime, which, according to Avi-Yonah, can be considered to reach the level of CIL.3 1 See T. Treves, ‘Customary International Law’, in Max Planck Encyclopedia of Public International Law (Oxford: Oxford University Press, 2006), para. 1. 2 J. L. Goldsmith and E. A. Posner, ‘A Theory of Customary International Law’, University of Chicago Law Review 66/4 (1999), 1116. 3 R. S. Avi-Yonah, International Tax as International Law: An Analysis of the International Tax Regime, Cambridge Tax Law Series (Cambridge: Cambridge University Press, 2007), 5–8.

74 Elizabeth Gil García International double taxation can be understood as the beginning of the regime as well as the start of the ‘story’ of international tax law.4 Double taxation was actually the League of Nations’ main preoccupation in the 1920s. Since then, there has been a tendency to develop an ‘international tax language’ that is mainly reflected in tax treaties, allowing for their common and symmetrical interpretation.5 Even if international tax law is eminently based on treaty rules, the question is whether these treaty rules actually reflect CIL. In other words, does (or could) the network of bilateral tax treaties provide evidence of the existence of CIL rules? Or can evidence of CIL in the tax domain be found outside the treaty network. In this regard, international organizations play an important role, as countries usually discuss international tax issues during international forums. Thus, are the resolutions adopted by international organizations customary in nature? And are they having a legally binding effect on states? One may ask why it is important to determine the existence or not of CIL rules in tax matters. Even when a dense network of treaties does exist, it does not cover all the cases and situations that can arise in international tax relations. Indeed, the network of DTCs, extensive as it is, does not comprise all bilateral relationships. Therefore, the question is whether any customary rules exist that bind states in the regulation of international tax relations. In other words, different rules (whether national or international) may provide an order to regulate international tax relations. Is this also the case of CIL?

5.2 The Identification of CIL: Two Constituent Elements Legally binding international rules have traditionally been drawn from treaties and custom. In this regard, article 38 of the ICJ Statute was the starting point of international law and the lawmaking process. Among the sources of law that the Court can draw upon, article 38(1)(b) of the ICJ Statute lists: ‘international custom, as evidence of a general practice accepted as law’. For Crawford, this is prima facie defective wording as ‘the existence of a custom is not to be confused with evidence adduced in its favour’.6 That is to say,

4

D. Ring, ‘International Tax Relations: Theory and Implications’, Tax Law Review 60 (2007), 85. See ‘International Tax Law and its Influence on National Tax Systems’ in this volume. 6 J. Crawford, Brownlie’s Principles of Public International Law (Oxford: Oxford University Press, 2012), 23. Similarly, Wolfke had noted in 1964 the confusing wording of art. 38(1)(b) of the ICJ Statute, and clarified that the function of this provision is not to show what constitutes evidence of general practice, but what constitutes evidence of a customary rule (K. Wolfke, Custom in Present International Law (Wroclaw: Zaklad Narodowy im. Ossolińskich, 1964), 28). 5

International Tax Law and Customary International Law 75 it is general practice that constitutes evidence of international custom. Even if article 38 does not establish a hierarchy of international law sources, it is commonly held that treaties are the primary source of international law. Yet one can also consider (assuming a ‘custom primacy thesis’) that there cannot be treaty law without a pre-existing framework of customary law; that is, custom came first, treaties second.7 Examples include the law of diplomatic immunities, which rested on custom until the Vienna Convention on Diplomatic Relations came into force in the 1960s.8 In this vein, Sender and Wood consider the necessary and important role of CIL in the international legal system, as it is useful, for example, in the case of treaty gaps and to interpret treaties.9 To define international custom as such, it is necessary to evaluate whether a general practice is accepted as law. In this regard, custom is made up of two elements: on the one hand, general practice (usus or diuturnitas); and on the other, the conviction that such practice reflects or amounts to law (opinio juris), or is required by social, economic, or political exigencies (opinio necessitatis).10 This second element, often referred to as the subjective or psychological element, is the central CIL concept. Indeed, this conviction is the element that actually distinguishes a state act that is performed voluntarily from an act that a state follows because it is required to do so by law.11 In order to identify the existence of a CIL rule, the ILC has frequently engaged in a survey of all available evidence of the general practice of states as well as their attitudes or positions.12 In 2016, the Commission adopted certain (draft) conclusions on the identification of CIL.13 Accordingly, the two constituent elements should be separately ascertained, taking into consideration the context, the nature of the rule, and the particular circumstances of any evidence found.

7 M. Prost, ‘Hierarchy and the Sources of International Law: A Critique’, Houston Journal of International Law 39 (2017), 299. 8 D. M. Bodansky, ‘The Concept of Customary International Law’, Michigan Journal of International Law 16 (1995), 667 9 See O. Sender and M. Wood, ‘Custom’s Bright Future: The Continuing Importance of Customary International Law’, in C. A. Bradley, ed., Custom’s Future: International Law in a Changing World (Cambridge: Cambridge University Press, 2016), 360–370. 10 A. Cassese, International Law (Oxford: Oxford University Press, 2005), 156. 11 Goldsmith and Posner, ‘Theory of Customary International Law’, 1116. 12 The ILC was established by the General Assembly of the United Nations in 1947. The identification of customary international law is one of the topics included in the ILC’s work programme. In fact, in accordance with art. 24 of its Statute, the ILC ‘shall consider ways and means for making the evidence of customary international law more readily available, such as the collection and publication of documents concerning State practice and of the decisions of national and international courts on questions of international law’. 13 United Nations, Report of the International Law Commission, 68th session, A/7 1/10 (2016). See also United Nations, Text of the Draft Conclusions as Adopted by the Drafting Committee on Second Reading, 70th session, A/CN.4/L.908 (2018), https://documents-dds-ny.un.org/doc/UNDOC/LTD/G18/136/30/ PDF/G1813630.pdf?OpenElement (accessed 28 February 2021).

76 Elizabeth Gil García

5.2.1 State Practice as Evidence of CIL In regard to the characterization of state practice (the so-called objective or material element), uniformity and generality are key aspects to be considered when seeking to identify a CIL rule.14 In Colombian–Peruvian Asylum (1950), the Court refers to ‘a constant and uniform usage practised by the States in question’.15 The working paper prepared for the ILC by Manley Hudson in 1950 established that a ‘concordant practice by a number of states’ was necessary for the emergence of a CIL rule.16 The next step is then to define the sufficient number of states that need to be practising a custom rule to consider such a rule a general one; that is, one that has the force of CIL. There is no agreement as to how uniform state practice should be; and, even if one assumes that all (or almost all) world states should follow it, CIL is ‘usually based on a highly selected survey of state practice that includes only major powers and interested nations’.17 It can be argued that the practice by a single state is not enough. Indeed, when Wolfke affirmed in 1964 ‘a number of states’, he meant more than one, so the practice of a few states would be sufficient, provided that such practice is recognized by other states.18 The ILC concluded that a practice ‘must be sufficiently widespread and representative’.19 Moreover, Hudson’s 1950 report also required that the practice be continued over a considerable period of time. Yet the ILC has concluded that no particular duration is required if the practice is general.20 Regarding the forms of practice, the Commission has reached the conclusion that practice may take different forms, including conduct in connection with resolutions adopted by an international organization or conduct in connection with treaties.21 With regard to the latter, the requirement of practice means the practice of states, but in certain cases the practice of international organizations may also contribute to the formation (or expression) of CIL rules.22 The ILC considers that resolutions, decisions, or other acts adopted by international organizations do not, per se, constitute a CIL rule (as the documents are not legally binding), but they may have a ‘value in providing evidence of existing or emerging law’.23

14 ILC, Formation and Evidence of Customary International Law, 65th session, A/ CN.4/ 659 (2013), 9–10. 15 Asylum (Colombia v. Peru), Judgment of 20 November 1950, ICJ Reports 1950, 276. 16 UN, Yearbook of the International Law Commission 1950, vol. II (1950), 26 17 Goldsmith and Posner, ‘Theory of Customary International Law’, 1117. 18 Wolfke, Custom in Present International Law, 46. 19 United Nations, Report of the International Law Commission, 77. 20 Ibid. In the same vein, see Crawford, Brownlie’s Principles of Public International Law, 24. 21 United Nations, Report of the International Law Commission, 76. 22 Ibid. 23 Ibid., 106–107.

International Tax Law and Customary International Law 77

5.2.2 The Opinio Juris Evidence The second element that allows for distinguishing mere practice or usage from custom raises more issues as it is difficult to identify the reason why a state is following a certain practice. The ILC refers to the sense among states of the existence (or absence) of an obligatory rule (i.e. a rule’s obligatory nature). In some cases, however, it could also be the state’s recognition of the necessity of a rule. According to the ICJ, in North Sea Continental Shelf (1969), state practice should occur ‘in such a way as to show a general recognition that a rule of law or legal obligation is involved’.24 The ILC states that ‘acceptance as law’ should be distinguished from a state’s other possible extralegal considerations with regard to the practice in question,25 for example ‘courtesy, good- neighbourliness and political expediency’, as established by the Court in Colombian– Peruvian Asylum.26 Therefore, the ILC concluded that the relevant practice should be accompanied by a conviction that it is ‘permitted, required or prohibited’ by CIL; that is, that states believe themselves legally compelled or entitled to do so by reason of a CIL rule.27 Goldsmith and Posner do not consider that states comply with CIL norms out of a legal obligation (opinio juris), but out of self-interested policies on the international stage. According to their CIL theory, the authors identify coincidence of interest, coercion, cooperation, and coordination as four different strategic positions from which international behavioural regularities are likely to emerge, as they arise out of national self-interest.28 For Villiger, the question of self-interest is immaterial: what matters is what practice states actually engage in.29 In other words, acting out of self-interest does not imply acting out of a sense of legal duty.30 The Court stated in North Sea Continental Shelf that ‘acting, or agreeing to act in a certain way, does not of itself demonstrate anything of a judicial nature’. Hence, evidence of opinio juris is required to determine the existence of a CIL rule. Otherwise it will be state practice guided by other motivations, such as ‘comity, political expediency or convenience’.31 In this regard, the ILC refers to states’ statements, official publications, government legal opinions, or treaty provisions as indicative examples of acceptance of a law.32

24

North Sea Continental Shelf Cases, Judgment of 20 February 1969, ICJ Reports 1969, 43. ILC, Second Report on Identification of Customary International Law, 66th session, A/CN.4/672 (2014), 43. 26 Asylum (Colombia v. Peru), 285. 27 United Nations, Report of the International Law Commission, 97. 28 See Goldsmith and Posner, ‘Theory of Customary International Law’, 1122–1128, who indicate that other explanations for complying with CIL norms could be the state’s reputation (ibid., 1135). 29 M. Villiger, Customary International Law and Treaties: A Manual on the Theory and Practice of the Interrelation of Sources (The Hague: Kluwer Law International, 1997), 48. 30 North Sea Continental Shelf Cases, 45. 31 United Nations, Report of the International Law Commission, 97. 32 Ibid, 77. 25

78 Elizabeth Gil García Villiger also refers to states’ statements (as clear evidence) or to the fact that a non- contracting state engages in practice invoking a conventional rule. However, these cases where the subjective element depends on a rule of international law may undermine CIL’s flexible and general nature.33 The statements made by states truly require an express public act, for instance during international forum debates. In this vein, debates aimed at solving certain international matters have traditionally taken place (and still take place) within the framework of international organizations. As we saw, the practice of international organizations may contribute to the formation of CIL. In addition, even if the adopted resolutions do not constitute legally binding documents, they can provide evidence of existing customary rules. In particular, some authors highlight the value of resolutions adopted by the UN General Assembly (UNGA) because reaching a consensus (by a significant number of states) on the resolution contents may imply an expression of opinio juris.34 Even if they amount to general practice (e.g. almost all states are members of the UNGA), this does not immediately imply it is practice ‘accepted as law’. We must remember that states’ attitudes towards a given resolution are usually motivated by political or economic factors. In the field of international taxation, the work and activity of international organizations (the League of Nations in the 1920s and later, the OECD and the UN) have influenced the current design of international tax policy and tax rules in different ways.35 Indeed, the model tax conventions drafted by international organizations have served as a basis for the negotiation and signing of bilateral tax treaties. Two questions arise. Has such an influence contributed to the formation of CIL rules? And does the large network of tax treaties (normally based on such models) provide evidence of CIL?

5.3 Customary International Tax Law: In Search of Evidence International economic law considers international economic relations as part of public international law. In essence, according to Herdegen, international economic law refers to ‘the regulation of cross-border transactions in goods, services, and capital, monetary relations and the international protection of intellectual property’.36 For Zamora, international economic law comprises ‘a broad collection of laws and customary practices 33 Villiger, Customary International Law and Treaties, 50–51. 34

See N. Petersen, ‘Customary Law without Custom? Rules, Principles, and the Role of State Practice in International Norm Creation’, American University International Law Review 23/2 (2007), 292; M. P. Scharf, ‘Accelerated Formation of Customary International Law’, ILSA Journal of International & Comparative Law 2/2 (2014), 324. 35 D. Ring, ‘Who Is Making International Tax Policy: International Organizations as Power Players in a High Stakes World’, Fordham International Law Journal 33 (2010), 652. 36 M. Herdegen, Principles of International Economic Law (Oxford: Oxford University Press, 2016), 3.

International Tax Law and Customary International Law 79 that govern economic relations between actors in different nations’.37 Core areas of international economic law, in addition to, for, international trade law, include double taxation agreements.38 Thus, the taxation of income from cross-border activities and investments seems to fall into the domain of international economic law. As mentioned in Section 5.1, the prevention of double taxation through DTCs has led to a network of thousands of bilateral tax treaties, reflecting the success of international law in the field of taxation.39 This would thus imply that the basic rules of public international law, including the doctrine of sources, are applicable. Indeed, international tax law forms part of international economic law (which itself is part of public international law).40 As in the case of the definition of international economic law,41 in the author’s view, two approaches can be considered when defining international tax law: (1) the sources of law; (2) the object of regulation. That is to say, it could be defined by reference to the sources of international law, or also by other rules (either international or domestic) that regulate the international activity in question.42 The latter, according to the author, may be connected to the theory of international regimes. Indeed, this theory precisely explains the order that exists in the field of an international activity;43 that is, the existence of rules, principles, and decision-making procedures that create a sort of ordered system within an international activity.44 The sources of international law may serve not only to define international tax law but also as a basis for the international tax regime (i.e. for cross-border taxation regulation). In other words, the sources of international tax law can serve to regulate cross-border taxation issues. That said, it should be noted that the international tax regime (as well as international tax law) is mainly composed of treaty rules, and this dense treaty network is precisely the clearest manifestation of the international tax regime.45 37 S. Zamora, ‘International Economic Law’, University of Pennsylvania Journal of Economic Law 17/1 (1996), 63. 38 Herdegen, Principles of International Economic Law, 8. 39 H. D. Rosenbloom, ‘The David R. Tillinghast Lecture International Tax Arbitrage and the International Tax System’, Tax Law Review 53 (2000), 164. Similarly, C. Braumann, ‘Taxes and Custom: Tax Treaties as Evidence for Customary International Law’, Journal of International Economic Law 23 (2020), 750. 40 Braumann, ‘Taxes and Custom’, 750. 41 See B. Ziemblicki, ‘Some Reflections on Custom in International Economic Law’, Wroclaw Review of Law, Administration & Economics 8/2 (2018), 204–205. 42 In this regard, see ‘International Tax Law and its Influence on National Tax Systems’, Section 6.4 in this volume. 43 See E. Barbe, ‘Cooperación y conflicto en las relaciones internacionales (La teoría del régimen internacional)’, Afers Internacionals 17 (1998), 56. Krasner defines international regimes as ‘principles, norms, rules and decision-making procedures around which actor expectations converge in a given issue-area’ (S. D. Krasner, ‘Structural Causes and Regime Consequences: Regimes as Intervening Variables’, International Organization 36/2 (1982), 185). 44 E. Gil García, ‘The Single Tax Principle: Fiction or Reality in a Non-Comprehensive International Tax Regime?’, World Tax Journal 11/3 (2019), 311. 45 Ibid., 334–335.

80 Elizabeth Gil García In this regard, international organizations (in particular the League of Nations in the 1920s and the OECD since the 1960s) have played a key role in shaping international tax law through the drafting of model tax conventions. So much so, in fact, that international tax law (and therefore the international tax regime) is mainly constituted by treaty-based rules. Whether the treaty network may constitute evidence for CIL in the tax domain is a different matter. In this vein, the ILC has concluded that a treaty rule can reflect a CIL rule if it is established that the treaty rule: (1) has codified a CIL rule existing before the treaty was signed; (2) has led to crystallizing a CIL rule which had begun to emerge before the treaty was concluded; or (3) has given rise to a general practice that is accepted as law, thus generating a new CIL rule.46

5.3.1 Evidence of CIL in the Tax Treaty Network Some scholars have considered the customary status of certain rules of international tax law (typically included in tax treaties), such as the jurisdiction to tax, the permanent establishment (PE) threshold, the arm’s-length standard (ALS), or non-discrimination.47 That is to say, conventional rules which may reflect a CIL rule. The latter actually correspond to examples in which the so-called objective element (‘general practice’) can be found. However, the question is whether states follow these practices under an obligation of general international law,48 and whether a CIL rule thus emerges beyond mere usage.49 In other words, have some tax treaty rules given rise to a general practice accepted as law? Has an existing CIL rule been codified by a tax treaty rule or been crystallized in a DTC? As we saw in Section 5.2, assuming the ‘custom primacy thesis’, a pre-existing CIL framework would have resulted in treaty law. In the field of international tax law, the international community has traditionally been concerned with preventing double taxation—see, in particular, the work on double taxation initiated by the League of Nations in the 1920s and pursued by the OECD, with the further development of model tax conventions to prevent double taxation. Double taxation has precisely been a concern for states since the advent of the international tax regime, which explains why states have implemented either unilateral or bilateral means to avoid or reduce international double taxation. In spite of this, no international law rules oblige states to prevent double

46

United Nations, Report of the International Law Commission, 78. See Avi-Yonah, International Tax as International Law; Y. Brauner, ‘An International Tax Regime in Crystallization’, Tax Law Review 56 (2003), 281; Braumann, ‘Taxes and Custom’; C. Thomas, ‘Customary International Law and State Taxation of Corporate Income: The Case for the Separate Accounting Method’, Cornell Law Faculty Publications Paper 1108 (1996). 48 Avi-Yonah, International Tax as International Law, 5; C. M. López Espadafor, ‘La costumbre internacional en el derecho tributario’, Nueva Fiscalidad 1 (2006), 10. 49 As Crawford points out, ‘usage’ refers to a general practice which does not reflect a legal obligation (Brownlie’s Principles of Public International Law, 23). 47

International Tax Law and Customary International Law 81 taxation (beyond DTCs) or prohibit double taxation.50 In a nutshell, states tend to eliminate double taxation, but not because of a rule of general international law.51 The fact that states sign tax treaties to prevent double taxation but also to provide order in international tax relations (e.g. rules allocating taxing rights to residence and source states) is the best demonstration that no customary international (tax) law rule is regulating cross-border income taxation. However, some authors consider that even if no general international law rule forbids double taxation, CIL prevents states from taxing when there is no connection between the income and the taxing state.52 In particular, Herdegen considers that CIL requires ‘a personal or territorial link for the exercising of jurisdiction to tax’.53 The present author does not share such an assumption. States may indeed impose taxes based on the relationship between the taxable person and the state (i.e. personal or objective connections), but these connections are the expression of states’ tax sovereignty.54 In other words, domestic legislation is often based on the state’s policy, state sovereignty being the basis for the state’s power to tax.55 Since general international law places few limits on the states’ tax sovereignty, the same event may be taxed in two or more states, thus generating a double taxation scenario.56 The purpose of the treaty is therefore to prevent such scenarios from arising. Norr considers that tax treaties have two functions. On the one hand, they relieve taxpayers from double taxation. On the other, they ensure that while the same income is not taxed twice, it should be taxed at least once.57 Beyond the debate on the existence of a single- tax principle (which is, indeed, a different story), conventional rules clearly provide order to cross-border taxation (not only to prevent double taxation). But, in addition, the taxable income is divided between source and residence countries; that is, the DTC acts as a coordination mechanism between tax systems by allocating taxing rights and avoiding tax conflicts. For instance, the inclusion of the PE provision in a treaty implies 50 E. C. C. M. Kemmeren, ‘Principle of Origin in Tax Conventions: A Rethinking of Models’, Thesis dissertation (2001), 17; M. Norr, ‘Jurisdiction on Tax and International Income’, Tax Law Review 17 (1962), 431, 438. 51 While general international law normally refers to rules of a customary nature and other unwritten forms of law (e.g. principles of international law) that are binding erga omnes, specific international laws (e.g. international tax law) are based on conventional rules that are binding inter partes (see A. Gourgourinis, ‘General/ Particular International Law and Primary/ Secondary Rules: Unitary Terminology of a Fragmented System’, European Journal of International Law 22/4 (2011), 1012–1013). 52 A. Rust, ‘Double Taxation’, in A. Rust, ed., Double Taxation within the European Union (Alphen aan den Rijn: Wolters Kluwer, 2011), 3; Avi-Yonah, International Tax as International Law, 28. 53 Herdegen, Principles of International Economic Law, 346. 54 Martínez Laguna, among others, affirms that a state’s sovereignty enables it to tax through the tax system (F. D. Martínez Laguna, ‘Abuse and Aggressive Tax Planning: Between OECD and EU Initiatives—The Dividing Line between Intended and Unintended Double Non-Taxation’, World Tax Journal 9/2 (2017), 194). 55 S. Gadžo, ‘The Principle of “Nexus” or “Genuine Link” as a Keystone of International Income Tax Law: A Reappraisal’, Intertax 46/3 (2018), 202. 56 M. Lang, Introduction to the Law of Double Taxation Conventions (Vienna: Linde, 2013), 23–24. 57 Norr, ‘Jurisdiction on Tax and International Income’, 445.

82 Elizabeth Gil García that there is a ‘nexus’ that allows business profits to be taxed by the source state. It is therefore difficult to consider that without a PE the state will tax business profits earned by a foreign entity. This is not only because the PE establishes a connection between the taxable person and the state, but also because the PE imposes a threshold on source taxation. The fact of incorporating certain rules into tax treaties may suggest that such rules do not reflect the CIL (i.e. the fact that a treaty rule codified a previous CIL rule is not established), since states, as Christians mentions, would not be expected to enter into tax treaties if they believed that an existing law already applied.58 It is therefore difficult to assume the idea of a pre-existing CIL tax framework. Notwithstanding this, can the tax treaty network provide evidence of a general practice accepted as law, resulting in a new CIL rule? In international tax matters, we have seen that treaty-based rules take precedence over others, and the fact that states include provisions written in similar terms in their DTCs attests to general practice. For instance, returning to the case of the ALS, its extended application in the international arena may be evidence of a CIL rule. The League of Nations first developed the separate accounting approach, and it was later incorporated into model treaties and, therefore, into bilateral treaties. Thus, a general practice (the ‘objective element’) exists. The sheer abundance of bilateral conventions that faithfully implement the separate accounting principles of the OECD and UN model treaties provide, in Thomas’s view, strong evidence of a sense of obligation.59 However, the fact that a rule is included in several treaties (e.g. the ALS often appears in treaties) does not automatically imply that such a rule reflects a CIL rule.60 Avi-Yonah recognizes the difficulty in proving that, in the absence of a treaty, states consider themselves bound by the ALS. Despite this, the author of this chapter considers that the ALS is part of the CIL referring to a case where the US Treasury believed itself bound by such a standard, despite the absence of any treaty- based or statutory requirement.61 In the view of the author, it is indeed difficult to determine why a state follows a specific practice, but we must remember that to identify a CIL rule both constituent elements should be present. In 1964, Wolfke stressed that when identifying CIL, a practice must be sufficiently ripe to justify the presumption that it has been accepted by states as an expression of law, being formed as CIL from that moment onwards and then

58 A. Christians, ‘Hard Law, Soft Law, and International Taxation’, Wisconsin International Law Journal 25/2 (2007), 330. 59 Thomas, ‘Customary International Law and State Taxation of Corporate Income’, 130. 60 United Nations, Report of the International Law Commission, 78. 61 R. S. Avi-Yonah, ‘Does Customary International Tax Law Exist?’, Law & Economics Working Papers 161 (2019), 7–8, https://repository.law.umich.edu/law_econ_current/161 (accessed 27 February 2021). Similarly, in the domain of the US Air Force, the requirements stated in the Geneva Convention Protocol I were considered in the preparatory process of an intensive aerial bombardment campaign in Iraq in 1991. The USA had not ratified the convention and Vagts thus regards this as a set of CIL rules reflecting treaty requirements (D. F. Vagts, ‘International Relations Looks at Customary International Law: A Traditionalist’s Defence’, European Journal of International Law 15/5 (2004), 1038).

International Tax Law and Customary International Law 83 having binding effects.62 Thus, the formation of a CIL rule (e.g. the ALS) would require, first, widespread and consistent state practice in the absence of a tax treaty and, secondly, that opinio juris evidence can be ascertained.63 The ILC commented in 2016 that a state act that conforms to a treaty by which it is not bound (or the application of conventional obligations in relations with non- parties to a treaty), in the absence of any other explanation, may point to the existence of ‘acceptance as law’.64 Even if a state is following a treaty rule (which it is not bound by), there may be several ‘reasons for any widespread and consistent state practice emanating from treaties’, such as the adoption of international tax rules that are compatible with those used by most other states.65 Indeed, other extralegal motivations, such as an international comity or states acting out of self-interest, do not make them erga omnes binding rules. Thus, the present author agrees with Braumann’s conclusion that practice falling within the tax treaty network would presumably result from the state’s adherence to its treaty obligations according to pacta sunt servanda. Therefore, CIL in tax matters should be explored in state practice beyond the scope of any DTC obligation by looking at other types of evidence such as statements by states, states’ activities within international organizations, and states’ domestic practice.66 This chapter will now focus on states’ statements and the influence of international organizations in state practice.

5.3.2 The Role of International Organizations in the Formation of CIL The observance of states’ actions is what has traditionally served to identify general practice accepted as law. However, other elements, such as states’ statements contribute to determining whether a CIL rule exists (i.e. concrete evidence of CIL). To this end, one may consult the written material of the state act or behaviour; that is, the documentation of states actions, allowing one to ‘explore’ their practices. Is it possible, in the international tax arena, to find such types of materials that indeed attest to CIL? Some examples of statements where states show their position in regard to specific tax issues can indeed be found. For instance, even if model tax conventions (which serve as a basis for the signing of tax treaties) are not legally binding, states may introduce reservations to provisions in them (i.e. a sort of state’s statement). Reservations may be made for

62 Wolfke, Custom in Present International Law, 60. 63

Even if Avi-Yonah admits CIL exists in some cases of international law, they constitute relatively limited instances (‘Does Customary International Tax Law Exist?’, 10). 64 United Nations, Report of the International Law Commission, 98. 65 Braumann, ‘Taxes and Custom’, 765–766. 66 Ibid., 769.

84 Elizabeth Gil García different reasons,67 so the fact that a state makes a reservation is completely independent of the opinio juris element. A reservation regarding the model convention may, of course, be regarded as a deviation from widespread practice, but the opposite does not mean that a state follows such practice because it amounts to law. Indeed, model tax conventions are soft law instruments that turn into hard law when their provisions are included in a tax treaty, thereby becoming an inter partes (and not erga omnes) binding rule.68 Regarding the latter, it could be argued that the sense of obligation does not refer to the conventional rule itself but to the fact that states enter into treaties and include specific rules in them. The elaboration of model tax conventions by different international organizations is closely linked to the development of the international tax regime.69 We could thus say that there is some kind of commitment to enter into tax treaties, to address specific tax issues, and to follow model tax conventions for such a purpose. It is not clear whether this general practice is based on the opinio juris element. However, state practice could be considered as CIL when it is required by social, economic, or political exigencies (opinio necessitatis). Indeed, this is a more flexible subjective element. The reasons that states enter into tax treaties in a specific way may be economic and political. This accords with Goldsmith and Posner’s theory that, in this, states pursue their national self-interests,70 for example states’ interests in preventing double taxation, considered as a threat to national economic interests, and that may only be solved by global cooperation.71 However, even if national self-interests may explain why states act in a specific manner in a cross-border context, the application of an erga omnes binding rule does not seem sufficient to explain such actions. That is to say, as evidence of opinio juris. In the case of member states of international organizations (i.e. state- based organizations), this ‘sense of obligation’ is clear. For instance, there are currently thirty- six OECD member countries, so it may seem logical that these countries follow the OECD Model when entering into tax treaties. But, in addition, states feel they must comply with certain recommendations, guidelines, or reports prepared by international organizations (e.g. the 1998 OECD report on harmful tax practices). Christians points out that the fact that member and non-member states comply with such guidelines 67 On

why reservations to the OECD Model Convention are made, see A. Vega and I. Rudyl, ‘Explaining Reservations to the OECD Model Tax Convention: An Empirical Approach’, InDret Revista para el Análisis del Derecho 4 (2011), 13. 68 T. Dagan, ‘BRICS: Theoretical Framework and the Potential of Cooperation in BRICS and the Emergence of International Tax Coordination’, in Y. Brauner and P. Pistone, eds, BRICS and the Emergence of International Tax Coordination (Amsterdam: IBFD, 2015), s. 2.2.1. 69 In particular, OECD models have significantly influenced international tax matters as the basis for tax treaty negotiations, as well as for other model tax conventions such as the UN Model, which follows the OECD Model in several domains (Lang, Introduction to the Law of Double Taxation Conventions, 32–33). 70 Goldsmith and Posner, ‘Theory of Customary International Law’, 1122–1128. 71 A. Christians, ‘BEPS and the Power to Tax’, in S. Rocha and A. Christians, eds, Tax Sovereignty in the BEPS Era (Alphen aan den Rijn: Wolters Kluwer, 2017), 6.

International Tax Law and Customary International Law 85 suggests a further analysis to determine whether they have or may become law through some other process.72 That is to say, there is a general practice and a certain degree of willingness to comply with the OECD report. Of course, the question of what motivates such compliance remains. It is thus clear that resolutions or other documents adopted by international organizations are not, per se, evidence of CIL, but they may serve as an element to identify the existence or emergence of a custom rule. As seen earlier, some authors refer to the process by which a non-legally binding rule does become binding.73 For instance, the Final Report of Action 5 of the Base Erosion and Profit Shifting (BEPS) Project included (as a minimum standard) the alignment of intellectual property (IP) box regimes to the so-called (modified) nexus approach. The implementation of such a standard was actually regarded as a condition for the survival of these preferential tax regimes, and a deadline for the alignment was established. In consequence, several states have followed the standard and adjusted their IP box regimes to that effect. A nexus-compliant IP regime is therefore now accordingly mandatory in those jurisdictions. The question thus arises whether states believe themselves legally compelled to implement the nexus approach. The BEPS Action Plan and their Final Reports are naturally of a soft law nature. They require the states’ implementation to turn into hard law. Hence, the standards and recommendations included in the BEPS Final Reports are a way of introducing, or even to some extent enforcing, certain aspects in the existing international tax regime.74 Even when one expects the states to comply with these recommendations, there are other reasons that states may have to adhere to such practices. Moreover, developed and developing countries may take different views on international organizations’ adopted reports, resolutions, or decisions.

5.4 Conclusions This chapter addressed CIL in tax matters in order to determine whether CIL rules, that then become legally binding in the international tax arena, exist. It was assumed that international tax law exists as such and is made up not only of principles and rules of public international law but also domestic rules; that is, the different rules that regulate cross-border income taxation regardless of their origin. Therefore, the sources of international law enshrined in article 38 of the ICJ Statute may be considered in the regulation of international tax relations. Despite this, treaty-based rules (i.e. a particular international law) clearly takes precedence in the tax domain. The chapter then analysed the constituent elements of CIL and which requirements should be met to identify a customary rule. Even if both the material and the subjective 72

Christians, ‘Hard Law, Soft Law, and International Taxation’, 328. Ibid.; Dagan, ‘BRICS’. 74 Gil García, ‘Single Tax Principle’, 335. 73

86 Elizabeth Gil García elements should be present and ascertained in a separate way, the central aspect is actually the identification of evidence of opinio juris; that is to say, states’ acts are based on the belief that a practice is prohibited, required, or permitted. Identifying this aspect in the international tax arena is a difficult task to achieve. Indeed, the tax treaty network attests not only to the success of international law in the tax domain but also to the extended practice of entering into treaties in a specific way, revealing similar state behaviours regarding the regulation of cross-border tax matters. However, this is not enough to demonstrate that states believe themselves legally compelled or entitled to do so due to an erga omnes binding rule. There could be other reasons that explain states’ behaviours. One of the reasons that makes states act out of self-interest is that the international tax regime lacks coordination. Beyond this, the role of international organizations in the design of international tax rules (also domestic rules) should not be overlooked. In recent times, such a role seems more relevant (e.g. the BEPS initiatives carried out by the Inclusive Framework where more than 125 countries are present). It is within international organizations that states debate and discuss common interests in an international activity. Thus, the resolutions and decisions they adopt play a role in the lawmaking process of international tax law, or more specifically in the formation of the international tax regime. The international tax regime is mainly composed of treaty rules and, based on the analysis in this chapter, we cannot affirm that the network of thousands of bilateral tax treaties amounts to CIL. We cannot claim either that a custom-based international tax regime pre-existed, because states enter into tax treaties based on economic or political interests (i.e. to prevent certain scenarios from occurring in the international tax domain). Thus, in international tax law, it is only possible (at least for the time being) to talk about inter partes binding rules. In other words, states are legally compelled under the pacta sunt servanda once they ratify a treaty.

Chapter 6

Internationa l Tax L aw and its Influ e nc e on National Tax Syst e ms Craig Elliffe

6.1 Introduction International tax law has components drawn from different areas of law, some parts of which are peculiarly domestic or sovereign in the sense that they are the exclusive decision of a state to impose taxation.1 In contrast, other parts reside in public international law because they deal with the rules, norms, and standards generally accepted in relations between nations. This dual dynamic is central as to why international tax law is influential on national tax systems. States impose taxes and deal with their citizens (or for most countries, their residents) on their worldwide income, and non-residents in respect of income sourced in their jurisdiction. It is precisely because of this residence and source-taxing matrix, and the resultant double taxation, that states need to deal with other states to make the international tax regime more coherent and fair.2 In respect of public international law, most focus has been on the part of international tax that relates to bilateral and multilateral treaties.3 In contrast, the area of law focusing

1

e.g. A. Qureshi, The Public International Law of Taxation: Text, Cases and Materials, 2nd ed. (Alphen aan den Rijn: Wolters Kluwer, 2019). 2 The principal vehicle used to reduce or eliminate double taxation is, of course, that of the bilateral tax treaty, most of which are based on the OECD Model Convention in its various iterations over the years. 3 A conclusion reached by Y. Brauner, ‘The True Nature of Tax Treaties’, Bulletin for International Taxation (Jan. 2020), 28.

88 Craig Elliffe on international law from other sources, such as customary international law (CIL), receives less consideration.4 Why is this discussion important? Is it essential to understand how international laws and norms shape national tax systems and whether they constrain domestic taxation choices? In other words, are there limitations on states introducing domestic rules which cut across the obligations designed to make the international tax regime coherent? Also, to achieve fundamental, consensus-driven change, such as that in the area of the taxation of the digital economy, requires significant cooperation by states. Understanding these parameters will inform whether, how, and why states should get involved in the design and structure of the international tax architecture. This chapter initially explores what is meant by international tax law. The focus then shifts to that part of tax law which is public international law. Clearly, bilateral and multilateral treaties have a direct impact on the legal systems of participating states. Treaties are not the only source of international law even though in the tax arena they clearly are by far the most important. So important that some suggest they are ‘perhaps the source, of international law’.5 Other highly respected scholars suggest that CIL has a part to play in influencing national tax systems.6 This suggestion is controversial, and the weight of opinion does not support the view that CIL presently creates any binding tax obligations on states.7 This chapter proposes that there is, however, an emerging system based on common understandings, with the legal duties of cooperation and underpinning expectations of implementation which should be recognized as creating obligations upon states even though the system does not constitute CIL. In this chapter, this is described as ‘consensus international tax law’ (CITL). CITL is highly influential on national tax systems and is a phenomenon that recognizes both the sovereignty and politics of national tax laws, as well as the need for harmonization and reasonableness in international dealings with other states and their tax systems.

4 C.

Braumann, ‘Taxes and Customs: Tax Treaties as Evidence for Customary International Law’, Journal of International Economic Law 23 (2020), 748. 5 Brauner, ‘True Nature of Tax Treaties’, 28. 6 R. Avi-Yonah, International Tax as International Law (Cambridge: Cambridge University Press, 2007); Avi-Yonah, ‘Does Customary International Tax Law Exist?’, University of Michigan, Law and Economics Working Paper No. 19-005 (2019). 7 Braumann, ‘Taxes and Customs’, 767, where the author concludes that the behaviour of states in complying with their double-tax treaty network means that it is almost impossible to prove CIL. E. Gil Garcia, ‘The Single Tax Principle: Fiction or Reality in a Non-Comprehensive International Tax Regime?’, World Tax Journal 11/3 (2019), 305; Brauner, ‘True Nature of Tax Treaties’, 31, where he argues: ‘Recently, scholars have attempted to argue that some common norms and tax treaties have come close to the threshold of opinio juris, yet this is still a small minority opinion that is unlikely to hold at the present.’

International Tax Law and its Influence on National Tax Systems 89

6.2 What is International Tax Law? Putting together the words international and tax seems to create some controversy because international lawyers focus on the former and tax lawyers focus on the latter. As alluded to in the Introduction, there are scholars who believe that a state has jurisdiction to impose tax without limitation.8 According to this view, the jurisdiction to tax is only limited by a state’s enforcement powers.9 Such a view might focus too much on the concept of tax and too little on international law. According to Gadzo, the theory of unlimited tax jurisdiction lacks valid support from the perspective of public international law.10 The approach of unlimited jurisdiction confuses the theoretical basis for taxation with the practical enforcement aspects of tax jurisdiction. As noted by Jeffrey: ‘Just because a law cannot in practice be enforced does not in any way relate to its legality or otherwise.’11 The prevailing view in contemporary tax scholarship is that ‘general international law poses limits to the substantive facet of tax jurisdiction’.12 The right of an international jurisdiction to taxation is said13 to be founded either on the relationship to a person (described in the OECD Commentary as the taxpayer’s personal attachment to the state)14 or on the relationship to a territory.15 This connection of the taxpayer’s residence and the state’s right to tax is commonly known as residence taxation. Where there is a connection between the territory in which the income is earned and the taxpayer, this is known as source taxation. International tax law is primarily concerned with the taxation of this type of cross-border income: (1) where the resident of a jurisdiction is doing business overseas and the foreign jurisdiction concerned decides to impose source taxation (outbound investment). 8 Qureshi, Public International Law of Taxation. 9

Ibid., 31. S. Gadzo, ‘The Principle of “Nexus” or “Genuine Link” as a Keystone of International Income Tax Law: A Reappraisal’, Intertax 43/3 (2018), 199. 11 R. Jeffrey, ‘The Impact of State Sovereignty on Global Trade and International Taxation’, Indiana Journal of Global Legal Studies 7/2 (1999), 43. 12 S. Gadzo, ‘Principle of “Nexus” or “Genuine Link” as a Keystone of International Income Tax Law’, 199. At 2.2.1 and 2.2.2, he discusses the limitations of Qureshi’s views and details the opposing views held by other international tax scholars to conclude that the theory of unlimited tax jurisdiction is inferior to the view that there needs to be a tax nexus as a prerequisite for the assertion of tax jurisdiction. Generally speaking, where more than one state wants to exercise jurisdiction, discussions to resolve the issue will be based around the idea of reasonableness, an underlying principle in the exercise of jurisdiction across a range of different international law areas. 13 Avi-Yonah, International Tax as International Law, 28; C. Elliffe, Taxing the Digital Economy: Theory, Policy and Practice (Cambridge: Cambridge University Press, 2021), 5. 14 OECD, Model Tax Convention on Income and on Capital: Condensed Version 2017 (Paris: OECD Publishing, 2017), 105. 15 W. Schon, ‘Persons and Territories: On the International Allocation of Taxing Rights’, British Tax Review 6 (2010), 554. In this article, Schon successfully sets out to show the fragility of the concepts of personal and territorial attachment which purportedly underpin the power to tax. 10

90 Craig Elliffe The residence taxation of this overseas income is sometimes complicated because the source country’s tax laws may apply to that income and this will have consequential implications in the resident’s jurisdiction; (2) where the government of one country taxes the business being carried out in its jurisdiction by a non-resident (inbound investment). The imposition of source taxation is based on the nexus between the income and the taxing jurisdictions. When countries decide that they will continue to operate on worldwide residence-based taxation in respect of outbound investment and to tax non-residents on income sourced in their jurisdiction in respect of inbound investment, then double taxation must arise. A resident of one country earning income in another jurisdiction will be legitimately subject to tax twice: first, by the source jurisdiction where the income is earned and, secondly, by the country in which they are resident.16 Double taxation is, however, regarded as undesirable. Indeed, it was famously described as ‘evil’ by the League of Nations in the 1920s.17 To overcome, or at least to ameliorate, the consequences of double taxation involves states sorting out matters between themselves. They do so by entering into treaties, historically bilateral treaties, but more recently multilateral ones.18 Such treaties normally address double taxation by reducing domestic source taxation in respect of inbound investment.19 Treaties are very obviously part of international law, but what other areas need to be considered? Article 38 of the Statute of the International Court of Justice lists four areas of law within the competence of the Court, although they are generally regarded20 as the sources of international law:21

(1) international conventions (treaties); (2) international custom, as evidence of a general practice accepted as law; (3) the general principles of law recognized by civilized nations; and (4) judicial decisions and the teachings of the most highly qualified publicists of the various nations, as subsidiary means for the determination of rules of law. 16 Elliffe, Taxing the Digital Economy, 7–9.

17 When the Financial Committee of the League of Nations asked four economists to consider the economic consequences of double taxation (from the perspective of the equitable distribution of burdens and interfering with the free flow of capital), they were asked to propose any general principles to remove the ‘evil consequences of double taxation’; G. Bruins, L. Einaudi, R. Seligman, and J. Stamp, Report on Double Taxation, League of Nations Economic and Financial Commission, Doc. E.F.S.73.F.19 (Geneva: League of Nations, 1923). 18 An example is the MLI designed to implement the tax treaty- related changes proposed by the OECD’s Base Erosion and Profit Shifting Project adopted in late 2016 and signed by the majority of countries in 2017. https://www.oecd.org/tax/treaties/multilateral-convention-to-implement-tax-treaty- related-measures-to-prevent-beps.htm. 19 They also provide credits for source-based taxation to the entity in the other contracting state. 20 I. Brownlie, Principles of Public International Law, 9th ed. (Oxford: Oxford University Press, 2019), 3. 21 Statute of the International Court of Justice, https://www.icj-cij.org/en/statute.

International Tax Law and its Influence on National Tax Systems 91 International lawyers separate their field into ‘general international law’, the norms of which are applied erga omnes (towards all) sometimes in circumstances where all states have standing, and sometimes where standing to enforce may depend on the question of who is affected by the disputed action,22 and ‘particular’ areas of international law (a smaller group of states that are consenting to or participating in a custom or treaty).23 It is suggested that the norms of general international law are found in CIL,24 ‘while the norms of particular international law normally stem from the treaties’.25 In addition to the domestic taxing rules that deal with residents deriving foreign income, and non-residents deriving locally sourced income,26 there are rules in tax treaties designed to change the result of the domestic legislation.27 International tax law is mostly, but perhaps not exclusively, made up of these three areas which are therefore a mixture of domestic and international law. The big question is whether there is a fourth area of international law which can be logically included in the definition of international tax law. This chapter posits that there is, as discussed in Sections 6.3 and 6.4, and that this is CITL and not CIL.

6.3 Is CIL Part of International Tax Law? The preceding chapter in this volume discusses in detail the relationship between international tax law and CIL.28 Rather than traversing the rather contentious legal arguments29 about the nature and scope of CIL, this chapter adopts the conclusions of the International Law Commission (ILC, the UN’s expert body for international law).30 These conclusions summarize the constituent parts of CIL defined in the draft

22 R.

Wolfrum, General International Law (Principles, Rules, and Standards) (Oxford: Oxford University Press, 2010); C. Tomuschat, ‘General International Law: A New Source of International Law?’, in R. Mazzeschi and P. De Sena, eds, Global Justice, Human Rights and the Modernization of International Law (Cham: Springer, 2018), 185–204. 23 O. Elias, ‘The Relationship between General and Particular Customary International Law’, African Journal of International and Comparative Law 8/1 (1997), 68. 24 G. Tunkin, ‘Is General International Law Customary Law Only?’, European Journal of International Law 4/4 (1993), 534–541. 25 Gadzo, ‘Principle of “Nexus” or “Genuine Link” as a Keystone of International Income Tax Law’, 195. 26 Outbound and inbound investment are described in Sections 6.1 and 6.2. 27 Y. Brauner, ‘An International Tax Regime in Crystallisation’, Tax Law Review 56 (2003), 265. 28 See ‘International Tax Law and Customary International Law’ in this volume. 29 Braumann, ‘Taxes and Customs’, 752. 30 United Nations, Report of the International Law Commission, 73rd session, A/ 73/10 (2018). The ILC was established by the General Assembly in 1947, to undertake the mandate of the Assembly, under art. 13(1)(a) of the Charter of the United Nations to ‘initiate studies and make recommendations for the purposes of . . . encouraging the progressive development of international law and its codification’.

92 Craig Elliffe conclusions of the ILC on the topic of the ‘Identification of customary international law’ as ‘unwritten law deriving from practice accepted as law’.31 The ILC make it clear that according to the Statute of the International Court of Justice,32 and consistent with the judgments of the Permanent Court of International Justice,33 to determine a rule of CIL requires evidence of both of the following elements:34 (1) a general practice; and (2) acceptance of that practice as law (opinio juris). A rule will only be established if there is evidence of acceptance among states that ‘it may be considered to be the expression of a legal right or obligation (namely, that it is required, permitted or prohibited as a matter of law)’.35 Practice or compliance with the rule does not meet the test without acceptance that it is the law (opinio juris). Likewise, a belief that something is, or ought to be, the law is insufficient unless there is evidence of practice. The International Court of Justice in the North Sea Continental Shelf Cases36 helpfully stated that to demonstrate CIL required actions by states: not only must amount to a settled practice, but they must also be such, or be carried out in such a way, as to be evidence of a belief that this practice is rendered obligatory by the existence of the rule of law requiring it. The need for such a belief, i.e., the existence of a subjective element, is implicit in the very notion of the opinion iuris sive necessitatis. The states concerned must feel they are conforming to what amounts to a legal obligation.

Scholars have sometimes pointed to the enormous similarities in international tax rules in many different states created by compliance with bilateral treaties negotiated in accordance with the OECD Model Convention.37 It is a valid observation that ‘most rules comprising international taxation are very close to being de facto harmonised’.38 Is the reason for this harmonization the existence of a huge number of similar tax treaties 31 Ibid.,

General Commentary, para. 66(3), p. 122. Consistent with the American Law Institute’s Restatement of the Law (Third) (Foreign Relations Law of the United States) which states that customary international law is law that ‘results from a general and consistent practice of states followed by them from a sense of legal obligation’, s. 102(2), p. 24. 32 Statute of the International Court of Justice, https://www.icj-cij.org/en/statute, specifically art. 38(1)(b), which includes among the sources of public international law that derived from ‘international custom, as evidence of a general practice accepted as law’. 33 Discussed by T. Treves, ‘Customary International Law’, in Max Planck Encyclopedia of International Law (Oxford: Oxford University Press, 2006). 34 United Nations, Report of the International Law Commission, 125. 35 Ibid. 36 North Sea Continental Shelf Cases (Germany v. Denmark; Germany v. Netherlands), ICJ Reports 1969, 3, para. 77. 37 Avi-Yonah, ‘Does Customary International Tax Law Exist?’. 38 Brauner, ‘International Tax Regime in Crystallisation’, 263.

International Tax Law and its Influence on National Tax Systems 93 or a belief that states have acceptance of legal rights and obligations sufficient to constitute CIL?39 Some suggest the latter. For example, Avi- Yonah suggests that when the UK introduced it diverted profits tax (DPT), it did so because the permanent establishment threshold is CIL. The DPT was designed, Avi-Yonah asserts, so that it was not a corporate income tax and therefore not subject to the treaties, or CIL.40 Such an argument needs to be approached with care. If states truly believed that an entity resident in their jurisdiction would not be subject to tax on business profits earned in another contracting state unless the entity had a permanent establishment in that source state, they would not need to enter into a treaty at all. In negotiating a treaty, a state surrenders its sovereign right to tax in respect of business profits earned in its jurisdiction. There is no evidence that it believed that an absence of a permanent establishment meant that it could not impose taxation on a foreign entity. It is actually the reverse; the treaty rule establishes that it normally believes that it can impose tax if the foreign enterprise earns business income in its jurisdiction without a permanent establishment. Scholars doubt that the commonality among international tax systems is due to states feeling as though they are obliged to conform to CIL.41 Braumann makes the point that the more states ratify and are bound by a treaty rule, the more difficult it becomes to show that the customary rule exists independent of the treaty. This is because it must be shown that the customary rule applies when the treaty rule does not and, as discussed earlier, to do so it is necessary to establish evidence of opinio juris.42 Gadzo suggests that evidence of opinio juris is more likely to be found in domestic legislation and the documentation of international organizations involved in international tax coordination (such as the League of Nations, OECD, or UN). Domestic legislation aligned to the basic norms of tax treaties could be ‘evidence that a State believes it is obliged to abide by the norm even outside of the treaty context because it is part of Customary International Law’.43 In Gadzo’s view, domestic law provides evidence for a personal or

39 S.

Schill, The Multilateralization of International Investment Law (Cambridge: Cambridge University Press, 2009). 40 Avi-Yonah, ‘Does Customary International Tax Law Exist?’, at ‘3. The PE Threshold’. 41 Braumann, ‘Taxes and Customs’, 769, where she suggests that there is currently insufficient evidence to suggest that the opinio juris element is present on the features of international tax that she considered, or see Brauner, ‘True Nature of Tax Treaties’, 31, where he notes: ‘Recently, scholars have attempted to argue that some common norms and tax treaties have come close to the threshold of opinio juris, yet this is still a small minority opinion that is unlikely to hold at the present.’ Earlier sceptics include D. Rosenbloom, ‘International Tax Arbitrageurs and the “International Tax System”’, Tax Law Review 53/2 (2000), 137; M. Graetz, ‘Taxing International Income: Inadequate Principles, Outdated Concepts, and Unsatisfactory Policies’, Tax Law Review 54/3 (2001), 261. 42 Braumann, ‘Taxes and Customs’, 775. E. Gil, ‘The Single Tax Principle: Fiction or Reality in a Non-Comprehensive International Tax Regime?’, World Tax Journal 305 (2019), 333, agrees, saying, ‘Consequently, the international tax regime does not rise to the level of customary international law.’ 43 Gadzo, ‘Principle of “Nexus” or “Genuine Link” as a Keystone of International Income Tax Law’, 202.

94 Craig Elliffe territorial nexus which denotes ‘one rare occasion where international customary law is relevant’ in the income tax arena.44 The question is whether there is sufficient evidence to suggest that states conform to international tax norms because of their subjective belief that these rules exist in CIL. As indicated previously, most scholars who have looked for this evidence have failed to find it. This appears to be the case when examined through the lens of either a general or particular CIL.45 There are some other very plausible reasons for the similarity or convergence of the international tax rules found in national tax systems. These are reasons such as the wholesale adoption of the OECD Model as a basis for bilateral treaty negotiation, a pragmatic approach to the common-sense collection of tax, or the efficiency of copying a successful rule from another close jurisdiction (which in the case of a rule relating to the source of income has the benefit of reducing double taxation). In the absence of further evidence, one feels that there are other forces of law at work which may not yet have matured into CIL but which still significantly influence national tax systems. Let us now turn to discuss these issues.

6.4 CITL This chapter advances the theory that there is something additional to areas of domestic and treaty law discussed earlier, but that it has not yet developed into CIL. In other words, international tax law consists of domestic law regulating cross-border transactions, public international law constituted by treaties, and CITL. CITL can create mandatory binding obligations but it mostly creates norms of cooperation and consensus which can facilitate new international tax compromises between states. CITL is made up of, at least, three components that exist in our international tax infrastructure. These are the institutions, instruments, and interpretation parts of the system described later in more detail. These components are very strongly interrelated, meaning that they overlap. There is a common theme running through these three areas. Globalization and the impact of technology facilitate the possibility of consensus. Bentley discussed this nearly two decades ago, when he noted:46

44

Ibid., 208. United Nations, Report of the International Law Commission, where in Conclusion 16(2), it states that with a rule of particular customary international law (which applies to a limited number of states) it is still necessary to ascertain whether there is a general practice among the states concerned that is accepted by them as binding (opinio juris) between themselves. 46 D. Bentley, ‘Influence from the Shadows: The OECD, the Shape of Domestic Tax Policy and Lessons for Federal Systems’, Revenue Law Journal 13/1 (2003), 129. 45

International Tax Law and its Influence on National Tax Systems 95 It is in the exploration of that technical detail by policymakers, legislators, the executive and the judiciary that the ready access to a wide range of instant information, communication and influence has the potential to change dramatically the framework and the process of decision-making.

Globalization has led to an increase in cross-border transactions by both multinational companies and individuals and it has been suggested that it is responsible for a change in the balance between domestic tax systems and international tax rules.47 Communication has had the impact of making international tax, and more particularly the payment of tax by multinationals, a matter of domestic politics in most countries.48 This was the background to the Base Erosion and Profit Shifting (BEPS) Project of the OECD and arose from the global financial crisis of 2008.49 The introduction to the original BEPS discussion document in 2013 refers to how the media, the public, civil society, and NGOs had been pointing to the inadequacies of the international tax regime and suggesting that there was a perception that ‘domestic and international rules on the taxation of cross-border profits are now broken and the taxes are only being paid by the naïve’.50 This groundswell of public opinion transmitted into domestic politics with the expectation that the political system in a country could deal with a tax problem. As Sadiq explains, unfortunately the generally held view that a democratically elected state can deal with a tax problem does not hold true in the international tax arena:51 Most importantly, these domestic rules do not generally have an international influence. However, where a tax problem is international, there may be overriding international considerations as a government’s own tools to remedy the problem may prove insufficient. Ultimately, a government may have to turn to the ‘international tax regime’ to find a solution.

The international tax regime that governments ‘turn to’ involve the institutions, instruments, and interpretation of international tax which collectively make up CITL. The first component of CITL—the rise of the institutions of international tax—has been

47 I. Valderrama, ‘The Interaction of Tax Systems and Tax Cultures in an International Legal Order for Taxation’, Diritto e Pratica Tributaria Internazionale 5/2 (2008), 841–867, https://ssrn.com/abstract= 1550815. 48 See the historical description of the events that occurred with the financial crisis of 2008, multinational aggression on tax planning, public outrage, and political hearings in the USA and UK in R. Mason, ‘The Transformation of International Tax’, American Journal of International Law 114/3 (2020), 365. 49 OECD, Addressing Base Erosion and Profit Shifting (Paris: OECD Publishing, 2013), https://doi.org/ 10.1787/9789264192744-en. 50 Ibid., 13. 51 K. Sadiq, ‘The Inherent International Tax Regime and Its Constraints on Australia’s Sovereignty’, University of Queensland Law Journal 31/1 (2012), 139.

96 Craig Elliffe a key part of the broader international tax regime since the 1920s,52 but in more recent times it has assumed even greater significance.

6.4.1 The Institutions of International Tax For a variety of reasons, the current international organizations are possibly at their most powerful and influential since the coming together and creation of the 1920s compromise by the League of Nations.53 Mason writes of the ‘transformation of international tax’ describing the BEPS Project as having ‘profound implications’.54 The point she is making is that rather than viewing the BEPS Project as simply an opportunity to devise whole new sets of anti-avoidance measures to prevent multinationals shifting profits,55 the BEPS Project should really be regarded as a fundamental change to international tax involving new ‘participants, agenda, institutions, norms, and legal instruments’.56 Mason is right: there has been an acceleration of change to the point where it can be described as transformational. The central institution in this transformation and this theory of CITL is the OECD. The last decade or so has shifted the OECD from being what some have described as a shadowy influence57 to being an actor in the spotlight on the centre stage. Hearson points to OECD staff, civil servants of national governments, and professional stakeholders being consistently described as members of an ‘epistemic community’.58 An influential group whose individuals ‘diagnose and prescribe policy reforms that are informed by, and that play out within, national legal regimes’,59 knowledgeably negotiating and ‘providing a forum for discussion and providing a base of expertise to structure the debate’.60 However, keeping the debate highly technical between tax bureaucrats and business association representatives had the effect of excluding civil society from international tax matters.61 52 For a brief description of the 1920s compromise, see Elliffe, Taxing the Digital Economy, 5; for a more detailed discussion, see M. Graetz and M. O’Hear, ‘The “Original Intent” of US International Taxation’, Duke Law Journal 46 (1997), 1033. 53 Mason, ‘Transformation of International Tax’, 367– 369, where she discusses OECD officials receiving a ‘mandate’ from the G20 and the rise of policymakers from the EU member states acting as a counterweight to the USA in international tax relations. 54 Ibid., 354. 55 Ibid., 353, where she asserts that many academics have characterized the BEPS Project as ‘a mere technical project to close tax loopholes’, but that this characterization underestimates the enormous transformational changes that have occurred in international tax in recent times. 56 Ibid., 354. 57 Bentley, ‘Influence from the Shadows’, 129. 58 M. Hearsen, ‘Transnational Expertise and the Expansion of the International Tax Regime: Imposing “Acceptable” Standards’, Review of International Political Economy 25/5 (2018), 651. 59 A. Christians, ‘Networks, Norms and National Tax Policy’, Washington University Global Studies Law Review 9/1 (2010), 22. 60 D. Ring, ‘Who Is Making International Tax Policy?’, Boston College Law School Faculty Papers Number 264 (2010), 681. 61 M. Rixon, ‘Politicisation and Institutional (Non-) Change in International Taxation’, WZB Berlin Social Science Center (2008), 13.

International Tax Law and its Influence on National Tax Systems 97 The OECD has no formal power but it has enormous influence.62 It is unclear whether it is the OECD, or its member states, which pressure countries to conform to such things as OECD transfer-pricing standards, or which nominate countries for inclusion on lists where they are designated as tax havens or part of a harmful tax competition agenda.63 As Brauner notes, when discussing the legal source for the OECD’s influence, ‘These issues are simply ignored in most cases due to the power that the OECD and its dominating members impose over the international tax regime.’64 The point is that the OECD and its powerful members, in practice, wield influence and expect compliance, particularly from those countries that have been involved in the debate. During BEPS and then subsequently, the OECD’s influence has grown, first by receiving the mandate of the G20,65 and then subsequently through its leadership of the Inclusive Framework. Made up of 139 states, the Inclusive Framework is open to those countries that commit themselves to implementing BEPS minimum standards.66 Although the Inclusive Framework has been a key body in the implementation of BEPS through the multilateral instrument, it has also been heavily involved in the further work in BEPS 2.0 (the proposals to reform the taxation of the digital economy and the global anti-avoidance plans for minimum corporate taxation). The BEPS Project has been an enormous catalyst for multilateralism, with the OECD and its officials at the heart of international tax agenda-setting, but there are other institutions involved in CITL. For example, the UN operates its Tax Committee, made up of twenty-five experts nominated by governments,67 which is the body responsible for maintaining the UN Model Double Taxation Convention between Developed and Developing Countries as well as its associated manual. Many countries will have some reference to the UN Model in their treaty network (to a greater or lesser extent).68 The contribution made by the UN and it Tax Committee is significant in its key constituency—the developing countries—but it has also made a contribution in

62 Brauner, ‘True Nature of Tax Treaties’, 32. See also the general proposition that the Australian government adopts voluntarily international norms promulgated by the OECD as well as specific examples in the area of transfer pricing given by Sadiq, ‘Inherent International Tax Regime’, 139. 63 Brauner, ‘True Nature of Tax Treaties’, 32. 64 Ibid. 65 As noted by Mason, ‘Transformation of International Tax’, 365, the thirty-seven members of the OECD added a further eight countries through the imprimatur of the G20, but importantly in addition to having a collective population of 3.5 billion, those countries include particularly important developing countries. 66 As at February 2021, 139 countries were members. For further information, see https://www.oecd. org/tax/beps/beps-about.htm/. 67 See the document entitled ‘Selection of Membership of UN Tax Committee- Frequently Asked Questions’ (Feb. 2021), https://www.un.org/development/desa/financing/sites/www.un.org.developm ent.desa.financing/files/2021-02/FAQ%20Nominations%20Tax%20Committee_Feb2021.pdf. 68 Sadiq, ‘Inherent International Tax Regime’, 143, where she says, ‘Having said this, there are occasions when Australia also relies on the work of the UN’. This observation is also true for New Zealand.

98 Craig Elliffe areas such as technical services and taxation of the digital economy.69 Increasingly, organizations such as the African Tax Administration Forum (ATAF) have become a significant force in these institutions, particularly in representing developing countries.70 Completing the picture are NGOs, tax watchdog groups (e.g. the Tax Justice Network), and even journalists who have been very active in reporting on global corporate tax avoidance and evasion.71

6.4.2 The Instruments of International Tax As indicated in the introduction to Section 6.4, there is a great deal of overlap between the institutions and the instruments of international tax. There are numerous instruments of international taxation, including the OECD, UN, and G20 reports and recommendations;72 administrative agreements between governments;73 as well as both the OECD Model and its associated Commentary74 and the UN Model and its associated Commentary,75 but undeniably the most important instrument of international tax is the tax treaty. Brauner, when evaluating the sources of international law, concludes that, ‘treaties are, therefore, the key component of the international tax regime and the primary reason for its success’.76 Most countries use the OECD Model as a basis for their international double-tax agreements,77 and the importance of the Model can be seen by the fact that not only are the substantive and administrative provisions in tax treaties based upon this Model, but they also follow its pattern and incorporate the major principles on which it is based.78 Major norms of international tax (such as separate entity accounting, arm’s-length principles of profit attribution, exemption from taxation of business profits if there is no permanent establishment in the source state, limited source-state taxation for passive income, relief from double taxation, provisions for non-discrimination, and mutual 69 One

could point to art. 12A and the newly proposed art. 12B of the UN Model as evidence of foreshadowing the allocation of taxing rights to the market or destination jurisdiction without the non- resident entity having a permanent establishment in that source jurisdiction. 70 See further information at https://www.ataft ax.org/. 71 Mason, ‘Transformation of International Tax’, 369–370. 72 e.g. there are the many, the numerous interim and final reports of the BEPS Project. 73 e.g. the Australian and New Zealand administrative agreement on dual resident companies pursuant to art. 4(1) of the MLI, see https://www.ato.gov.au/Business/International-tax-for-business/In- detail/MLI-Article-4(1)-administrative-approach/. 74 OECD, Model Tax Convention on Income and on Capital (Full Version) (Paris: OECD Publishing, 2017), available at http://www.oecd.org/ctp/model-tax-convention-on-income-and-on-capital-full-vers ion-9a5b369e-en.htm. 75 United Nations, Model Double Taxation Convention between Developed and Developing Countries (New York: UN, 2017), available at https://www.un.org/esa/ffd/wp-content/uploads/2018/05/MDT_2 017.pdf 76 Brauner, ‘True Nature of Tax Treaties’, 31. 77 Bentley, ‘Influence from the Shadows’, 131. 78 Ibid.

International Tax Law and its Influence on National Tax Systems 99 agreement and assistance procedures) are therefore embedded in the many (greater than 3,000)79 treaties which are based upon this important Model.80 The use of the OECD Model can be seen as a sophisticated way of informally achieving a ‘coordinated and networked action by governments’ which can lead to new international law-and policymaking ‘without imposing undue restrictions on national sovereignty’.81 The use of the MLI82 to achieve modifications to countries’ tax treaty networks is an excellent example of ‘hard’83 international law adopting consensus positions while at the same time preserving some national sovereignty discretions. The MLI is a novel solution to update some of the ninety-five signatories’ bilateral tax treaties with a collective stroke of the pen. The MLI is targeted at making changes designed to prevent avoidance identified in the BEPS Project, but there is considerable scope for countries to exercise their own sovereignty through the ability to select various options or even completely opt out of certain changes. That said, there is a very obvious example of direct CITL in the MLI. As a combination of an institution (the Inclusive Framework) and an instrument (the MLI) working together, countries become bound to implement certain minimum standards. Implementation of these minimum standards (primarily aimed at tax treaty abuse and dispute resolution) is mandatory for members of the Inclusive Framework and, although they may be given some choice about how to effect the change, the requirement to make the changes is binding.84 There are some important specific examples here of the operation of these mandatory requirements. Most countries that are part of the Inclusive Framework elected to adopt the principal purpose test (a treaty- based general anti-avoidance rule (GAAR)), rather than the limitation-of-benefits clause, but either or both were permissible.85 The OECD Model Treaty, the resultant bilateral treaties, and the variations to these treaties created by the MLI are only part of the story. Certainly the treaties, in their various forms, bilateral or multilateral, become actual law once incorporated into a nation’s legal system. Other than these direct sources of law, some instruments of CITL play an important but indirect role in influencing and shaping domestic tax rules. These instruments, such as the OECD Model Commentary, the OECD Reports (such as the BEPS Action Reports), the Explanatory Statement to the MLI, and the transfer 79

This is the number referred to by Braumann, ‘Taxes and Customs’, 748. Lang, P. Pistone, J. Schuch, and C. Staringer, The Impact of the OECD and UN Model Conventions on Bilateral Tax Treaties (Cambridge: Cambridge University Press, 2012), 1–36. 81 A. Cockfield, ‘The Rise of the OECD as Informal “World Tax Organisation” through National Responses to E-Commerce Tax Challenges’, Yale Law Journal 8/1 (Spring 2006), 167. 82 Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting, https://www.oecd.org/tax/treaties/multilateral-convention-to-implement-tax-treaty-rela ted-measures-to-prevent-beps.htm (referred to as ‘Multilateral Instrument’ or ‘MLI’). 83 As opposed to ‘soft’ law. They can also be viewed as a direct as opposed to indirect source of law. 84 This is discussed in the FAQ section of the OECD MLI website where it is noted at question 12 that there is no flexibility in the MLI for certain amendments, see https://www.oecd.org/tax/treaties/MLI-fre quently-asked-questions.pdf. 85 C. Elliffe, ‘The Meaning of the Principal Purpose Test: One Ring to Bind Them All?’, World Tax Journal 11/1 (2019), 50. 80 M.

100 Craig Elliffe pricing guidelines, are often known as ‘soft law’. There is significant crossover between the role that these instruments play and their effect in influencing interpretation. Commentators, such as Brauner, point to the use of soft law as a ‘coordination solution’. By this, he means that countries were prepared to agree on certain principles for international tax (e.g. the exchange of taxpayer information and the elimination of double taxation) enshrined in the 1920s compromise because those principles would lead to greater standardization and coordination, but the countries would not agree to a formal harmonization of international tax laws.86 Soft law is a very important component of CITL because it leads to the development:87 of a sort of international common law that is a notch below customary law since not all countries agree to be bound by it, but which countries that believe in the benefit of more cooperation could support in the hope that the sovereignty cost of increased cooperation would be more than paid for by its benefits even if they could not reduce it to formal agreements.

This can lead to soft law being influential, for example when a state decides to formally adopt law. This phenomenon was discussed by Sadiq when she pointed to the wording of a press release when Australia adopted the OECD’s Transfer Pricing Guidelines in order to be ‘more in line with international best practice’.88 A country adopting an international regime developed by the OECD as a ‘best practice’ norm makes sense from a coordinated and integrated international tax perspective but, as Sadiq observes, it is an example of the institutions and instruments of international tax influencing domestic tax policy:89 In essence, as the November Press Release suggests, Australia is turning to what it considers to be international best practice through the work of the OECD on transfer pricing thereby proactively electing to adopt and be constrained by what could be seen as part of the international tax regime.

The second reason why soft law in its various forms is widely utilized in CITL is because of its flexibility. Hardwired tax treaties are generally thought of as being difficult, or time-consuming, and costly to amend. In comparison, changes to the OECD Model Commentary might be comparatively efficient since they potentially apply to all treaties that use terms which are described in the Commentary in an ambulatory way. It seems highly likely that the OECD Commentary is not binding, but that does not remove its importance as an interpretive tool, a point which is discussed in the next section. The 86

Brauner, ‘True Nature of Tax Treaties’,33. Ibid., 34, where Brauner acknowledges A. Guzman and T. Meyer, ‘International Soft Law’, Journal of Legal Analysis 2/1 (2010), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1353444#. 88 W. Shorten, Press Release no. 145, 11 November 2011 announcing the Australian adoption of new transfer pricing rules. 89 Sadiq, ‘Inherent International Tax Regime’, 139 (emphasis added). 87

International Tax Law and its Influence on National Tax Systems 101 issue is that revisions to the instruments of international tax law, such as the OECD Model Commentary, allow the epistemic group of the OECD Secretariat and the revenue officials of Working Party 1 to make changes without amending the Model or bilateral treaties. Whatever else this is, it is most certainly influential and a key part of CITL.

6.4.3 The Interpretation of International Tax The principal instrument of international tax law, a treaty, shall be interpreted in good faith in accordance with the ordinary meanings to be given to the terms of the treaty in their context and in the light of the treaty’s object and purpose.90 In most legal systems, this has led to a focus, first, on the clear meaning of the words used in the treaty. Subsequently, the focus on the text is combined with due consideration of the purpose of the particular article in the treaty as a whole.91 As the UK Supreme Court notes, treaty interpretation requires establishing, in an objective and rational way ‘the common intention of the parties’.92 In common law jurisdictions, at least, courts have generally taken the view that it is not a narrow domestic view of the words that are required, but rather a broad, purposive interpretation of the goals and objects of the convention.93 Such a purposive interpretation is likely to involve tax administrations, taxpayers, and most importantly courts primarily examining the OECD Commentaries, but also other instruments of international tax referred to earlier (see Section 6.4.2). As observed by the Committee on Fiscal Affairs (in the Commentary), ‘in many decisions, the Commentaries have been extensively quoted and analysed, and have frequently played a key role in the judge’s deliberations’.94 It is hard for tax lawyers to discern how much weight to give the Commentaries, but they can never be read literally in the same way as tax legislation and they certainly are not binding.95 The key question is whether the new Commentary amends, changes, or ‘clarifies’. It is clear that generally courts96 accept that later Commentaries 90

Vienna Convention on the Law of Treaties, art. 31(1). P. Baker, Double Taxation Conventions (London: Sweet & Maxwell, 2020), ch. E: ‘The Interpretation of Double Taxation Conventions’. 92 Anson v. Commissioners for Her Majesty’s Revenue and Customs [2015] UKSC 44, per Lord Reed. 93 See R v. Crown Forest Industries Ltd (1995) DTC 5,389, at 5393 per Iacobucci J, for the Canadian Supreme Court; United Dominions Trust Ltd v. CIR (1973) 1 NZTC 61,028, at 61,031 per McCarthy P, for the New Zealand Court of Appeal; and Thiel v. FCT (1990) 90 ATC 4717 at 4720 per Mason CJ, Brennan and Gaudron JJ, for the High Court of Australia. See also K. Keith, ‘Interpreting Treaties, Statutes and Contracts’, New Zealand Centre for Public Law (2009). 94 OECD, Model Tax Convention on Income and on Capital (Paris: OECD Publishing, 2014), at http:// dx.doi.org/10.1787/9789264239081-en, Introduction at 29.3, I-9. 95 J.Avery Jones, Treaty Interpretation—Global Tax Treaty Commentaries (Amsterdam: IBFD, 2014), at 1.1.2.1 where he says: ‘It would clearly not be possible for the OECD Commentaries to be binding, both because of the variety of material that they contain and because this would inhibit their development, as states would have to be very careful about what they contained.’ 96 Most recent cases in common law jurisdictions take the view that the ambulatory use of later Commentaries are the widely accepted guide to the interpretation and application of bilateral 91

102 Craig Elliffe (Commentaries amended and promulgated after the conclusion of the treaty), can be of considerable assistance in the interpretation of the treaty provided that they clarify or amplify the previous Commentary.97 Overall, the interpretation process which utilizes the Commentary and other instruments of CITL is useful, but it must be recognized that there are risks associated with the ‘casual, reflective reliance on them and other, less formal soft law instruments in the interpretation of tax treaties’.98 Occasionally, the process of ‘clarifying’ represents the ultimate in dramatic finessing. An unacceptable court decision (in the eyes of the OECD) might be effectively reversed on the basis that an interpretation by the court was not what was intended.99 In addition to this reliance on the instruments of CITL (such as the Commentary) in judicial decisions, there is an increasing tendency for courts in one country to use the judgments of another. Baistrocchi’s analysis of twenty-seven jurisdictions suggested that tax disputes were consistently patterned, no doubt reflecting the enormous similarity of the international tax regime created through the consistent use of the OECD Model.100 Klaus Vogel notes that common interpretation is necessary in order to consistently apply and allocate tax claims between contracting states.101 Vogel points to numerous instances where the courts of one jurisdiction have referred to the decision of another, while noting that ‘common interpretation’ does not mean that the case law of the other state must be accepted without review.102 Convergence in the interpretation conventions, provided they represent a fair interpretation of the Model Convention and do not conflict with the Commentary in existence at the time the convention was entered into. Of course, neither treaty partner should have registered an objection to the new Commentary; see, e.g., the approach of the Canadian Federal Court of Appeal in R v. Prevost Car Inc. [2009] FCA 57, [2010] 2 FCR 65. It is also appropriate to observe that many academic writers do not accept this approach, preferring the view that static interpretation is required so that only the Commentary available at the time the contracting states negotiated the treaty is relevant, and that subsequent Commentaries should be ‘seen in a different light’. See M. Lang and F. Brugger, ‘The Role of the OECD Commentary in Tax Treaty Interpretation’, Australian Tax Forum 23 (2008), 107. 97

For a discussion on the various academic views and the approach of high-level courts, see C. Elliffe, ‘Cross Border Tax Avoidance: Applying the 2003 OECD Commentary to Pre-2003 Treaties’, British Tax Review 3 (2012), 322–324. 98 Brauner, ‘True Nature of Tax Treaties’, 36, where he discusses the significant problems in treaty interpretation posed by the use of soft law. 99 Although it is ancient history, the New Zealand Court of Appeal decision in CIR v. JFP Energy Inc. [1990] 3 NZLR 536 (CA), is such a case, with the court deciding that the ‘borne by’ found in art. 15(2) necessitated a requirement that the expense be made in the accounts of the non-resident entity operating in the source jurisdiction. Subsequently, the OECD Commentary made it clear (currently contained in para. 7.1 of the OECD, Model Tax Convention on Income and on Capital 2014, on art. 15, at C-15, 11) that the term would also apply even in circumstances where the expense was not part of the accounts of the non-resident. 100 E. Baistrocchi, A Global Analysis of Tax Treaty Disputes (Cambridge: Cambridge University Press, 2017), 1455. 101 K. Vogel, Klaus Vogel on Double Taxation Conventions: A Commentary to the OECD-, UN-and US Model Conventions for the Avoidance of Double Taxation on Income and Capital, with Particular Reference to German Treaty Practice, 3rd ed. (Boston, MA: Kluwer Law International, 1997), 73–76. 102 Ibid., 75–76.

International Tax Law and its Influence on National Tax Systems 103 of terms found in international treaties is a desirable objective, so the Commentaries, and reference to relevant foreign judgments, can play an important part in international conformity, ‘given that the underlying objective of such international treaties and model laws is to achieve uniformity across jurisdictions’.103

6.5 Conclusion This chapter suggests that there is an international tax regime which has an enormous influence on national tax systems. The arguments do not suggest that states have the necessary ‘belief ’ to constitute CIL, but that may happen in the future. Meanwhile, there is an existing legal ecosystem involving multilateral discussions by states utilizing the institutions of international law such as the OECD, G20, and most recently the Inclusive Framework. These institutions are responsible for indirectly and, on occasions via the MLI, directly, producing the instruments of international law. These instruments include both hard law, such as the MLI and bilateral treaties through conformity with the OECD Model, and also soft law, through the OECD Commentary and numerous other reports. The subsequent process of interpreting these bilateral and multilateral instruments, using the OECD Commentary and other reports, together with reference to foreign judgments by courts from around the world, leads to the theory of CITL. How real is this theoretical CITL concept? Let us examine it through the lens of the BEPS Project. The consensus cooperation in the BEPS Project 1.0 (the development and implementation of the fifteen Action Plans to overcome base erosion) was ultimately undertaken by 139 countries, although it had originally been initiated by the OECD (with a mandate from the G20). The components of CITL involved a significant number of legal changes to both bilateral treaties and domestic law. Ultimately, this was implemented with some clear mandatory BEPS obligations, and some less certain expectations, in the countries which had voluntarily signed up to the Inclusive Framework. This is a good example of CITL working in the area of treaty abuse and tax avoidance. BEPS 2.0 (the taxation of highly digitalized businesses and consumer- facing businesses operated by large multinationals, together with proposals adopting a minimum tax on undertaxed income earned by entities in the multinational group) is the ultimate test of whether CITL will work on what might be regarded as the cornerstone principles of the international tax framework of the 1920s compromise; namely, separate entity accounting, the arm’s-length principle for determining profits, and the taxation of business profits in the market jurisdiction only where the enterprise carries on business through a permanent establishment in the source country. The fundamental changes

103

Dame Susan Glazebrook (New Zealand SC Judge) writing extrajudicially in a paper, ‘Judging in an International World’ (2010).

104 Craig Elliffe proposed to the existing international tax regime by BEPS 2.0 require the operation of CITL in design, implementation, and operation. Seen in the light of recent discussions on the taxation of highly digitalized businesses, it is still unclear whether a consensus will emerge (the 2020s compromise utilizing a new multilateral instrument to change the allocation of profits and taxing rights away from countries of origin to market or destination countries). If this consensus emerges, it will be further strong evidence of the existence of CITL.

6.6 Postscript The influence of international tax law on national tax systems views the interaction between states and international tax law from one direction, but it is also valid for reflecting on the influence of national tax systems in international tax law. The domestic tax system of the USA has had a huge impact on the design of the international tax regime. According to Avi-Yonah, for much of the past one hundred years, the USA has led international tax reforms in what he terms ‘constructive unilateralism’.104 This is an example of where domestic tax policy has influenced international tax policy. It is also notable that domestic legal systems reach into and dramatically influence the outcomes of international tax law. Two examples suffice to support this proposition. The first is an example of territorial limitation found in common law jurisdictions. Under this rule, a non-resident employer is required to have a trading presence in the UK in order to be subject to an obligation to deduct employment tax (PAYE) at source. The principle was considered in the 1983 House of Lords decision Clark,105 relating to whether a non-UK resident company was required to deduct income tax under the PAYE system from employment income paid to its employees working outside the UK but in the UK sector of the continental shelf. These employees were liable to income tax. Oceanic argued that it was not subject to the PAYE obligation because it was neither a UK resident nor did it have a taxable UK branch. The House of Lords held, by a narrow majority, that there was an implied territorial limitation on the statute in question. The problem faced by the court was that the clear words of the statute did not have any such limitation, suggesting only that a non-resident employer making a payment of a taxable emolument is subject to the PAYE obligation notwithstanding the place of payment, the currency in which it is made, or the residence of the payer. The majority decided that the statutory obligation was impractical and could not be enforced, leading to the conclusion that the territorial limitation was to be implied into the clear words of the relevant

104 R. Avi- Yonah, ‘The Interaction between Unilateralism and Multilateralism in International Tax’, in C. Elliffe, ed., International Tax at the Crossroads: Institutional and Policy Reform in the Error of Digitalisation (Cheltenham: Edward Elgar, forthcoming 2023). 105 Clark (Inspector of Taxes) v. Oceanic Contractors Inc. [1983] 2 AC 130 (HL).

International Tax Law and its Influence on National Tax Systems 105 part of the Act.106 The practical considerations of enforceability and collection led to an interpretation of the statute which restricted the cross-border application of the law. The second example is where domestic law reaches into and overrides international tax law. Whether overriding is acceptable is a contentious issue, although in circumstances where the taxpayer relies on international law for tax avoidance purposes, the general view is that the improper use of a tax treaty can be overridden by domestic anti-avoidance provisions.107 There are other situations where national tax systems influence international tax rules but perhaps such a discussion needs its own chapter?

106

Ibid., 148. van Weeghel, ‘Tax Treaties and Tax Avoidance: Application of Anti-Avoidance Provisions’, Cahier de Droit Fiscal International (2011), 22. The 2010 Congress of the IFA in Rome provided a forum for analysing the relationship between treaties and the GAARs in a number of different tax jurisdictions around the world. As a result of the Congress, forty-four country reporters considered their own country’s response to the issue. The General Reporter, Stef van Weeghel, concluded in his summary of these reports that the vast majority of the branch countries determined that their GAARs could be reconciled with their treaty obligations. By this he meant that while most countries have statutory or judge-made anti-avoidance rules (although there are a considerable number of differences in their application), these GAARs can, and do, apply to cross-border transactions. Van Weeghel, concluded that, ‘Without exception the GAARs can have international effect and there is no distinction in their application depending on the national or international effect’ (ibid.). 107 S.

Chapter 7

Internationa l Tax L aw and Persona l Ne x u s Michael Dirkis

7.1 Introduction As far back as 1776, Adam Smith in An Inquiry into the Nature and Causes of the Wealth of Nations identified the importance of individuals (persons) as a source of taxation revenue.1 He states: The revenue which must defray, not only the expence of defending the society and of supporting the dignity of the chief magistrate, but all the other necessary expences of government for which the constitution of the state has not provided any particular revenue, may be drawn either, first, from . . . the sovereign or commonwealth, and which is independent of the revenue of the people; or, secondly, from the revenue of the people.2

Smith concluded that ‘[e]‌very tax must finally be paid from . . . one or other of . . . three different sorts of revenue, or from all of them indifferently’, being rent, profit, and wages, all of which were the then existing key sources of revenue of individuals.3 In 1923, the Expert Committee Report to League of Nations Finance Committee also noted that the liability arising from the imposition of impersonal taxes is ‘ultimately defrayed by persons and, through the process of economic adjustment, ultimately affect the economic situation of the individual’.4 In the twenty-first century, the situation has not changed. Taxation which is levied upon the consumption of individuals or 1 A. Smith, An Inquiry into the Nature and Cause of the Wealth of Nations (1776), ed. E. Cannan (London: University Paperbacks, 1961). 2 Ibid., II, 34. 3 Ibid., 349–350. 4 G. Bruins et al., Report on Double Taxation, League of Nations Economic and Financial Commission, Doc. E.F.S.73.F.19 (Geneva: League of Nations, 1923), 18, https://adc.library.usyd.edu.au/view?docId=

108 Michael Dirkis remuneration for their personal service, or upon profits and gains from their direct and indirect ‘ownership’ (holding, interest (legal, contractual, equitable or beneficial entitlement, etc.))5 of real, personal, and intellectual property, and know-how is ultimately met by individuals.6 When jurisdictions seek to tax foreign and cross-border income, profits, gains, or consumption (international taxation), the tax consequences are also linked back to persons. It is this area of jurisdictional claims which gives rise to ‘international tax law’. What is commonly referred to as ‘international taxation law’ is the international consensus on acceptable tax policy and administration (the international tax ‘norms’ or ‘consensus-based framework’) reflected in OECD, UN, regional, and jurisdictional model conventions, agreements, and interpretative material, which are to varying degrees given legal effect in the domestic law of adopting jurisdictions.7 Jurisdictional tax claims over cross-border transactions can trigger conflicts between individual jurisdictions because they have differing tax policy approaches in the international components of their domestic taxation laws and administrative rules. As no two tax systems are exactly the same, non-alignment of domestic tax systems designed in isolation will result in mismatches.8 The OECD’s Addressing Base Erosion and Profit Shifting discussion paper (BEPS Report) and Action Plan on Base Erosion and Profit Shifting (BEPS Action Plan) noted that, since the 1920s, the continuation of jurisdictions asserting their sovereign right to establish their own tax rules has led to further gaps and frictions between the tax systems of different jurisdictions.9 Much of the original academic discourse on the interplay between international tax law and personal nexus, which was commenced by economists and lawyers in the late nineteenth century, has been in the context of seeking solutions to resolve the conflicts (double taxation) that can occur from jurisdictions adopting differing tax policy approaches.10 The reduction of these conflicts had also been a key focus of the League of Nations in the 1920s.

split/law/xml-main-texts/brulegi-source-bibl-1.xml;col lection=law;database=ozlaw;query=;brand= ozlaw. 5 These ‘interests’ in ‘legal fictions’ range from interests in corporations (body corporates) formed by statute (incorporated associations, government instrumentalities, labour unions, agricultural and housing cooperatives, property owners’ corporations, etc.), common law (corporation sole and body politic), royal charter, partnerships (common law, tax law, limited (Roman ‘societas’ and medieval ‘commenda’), incorporated), equitable arrangements (trusts), foundations (Stichtings, Privatstiftung, Anstalt, etc.). 6 See Bruins et al., Report on Double Taxation, 19. 7 The words ‘international’ taxation law seems to suggest that there is a global tax law created by a universally accepted body with universal application to all jurisdictions. That is not the case. 8 See OECD, Addressing Base Erosion and Profit Shifting (Paris: OECD Publishing, 2013), 34–36 and 39, http://dx.doi.org/10.1787/9789264192744-en 9 Ibid., 9–10. 10 See K. Vogel, ‘Which Method Should the European Community Adopt for the Avoidance of Double Taxation?’, Bulletin for International Fiscal Documentation 56/1 (2002), 4–10.

International Tax Law and Personal Nexus 109 In recent times, it has continued to be the focus of a number of international bodies/ forums, in particular the OECD, the G7, and the G20. In October 2015, the G20/OECD Base Erosion and Profit Shifting (BEPS) Project delivered fifteen final reports, which focused upon altering the international tax ‘consensus-based framework’ to deal with ongoing issues associated with double non-taxation or less than single taxation arising from gaps in the framework. This work has been continued since 2016 by the OECD/ G20 Inclusive Framework on BEPS. In the main, this work is focused upon reforming the international tax consensus-based framework in respect of cross-border corporate taxation and not issues associated with individuals. Thus, the domestic personal nexus tests (which also constitute the residency basis of bilateral tax treaties) have remained unchanged, leaving open the opportunities for future gaps and frictions. This chapter first briefly explores the limitation of sufficient connection on the sovereignty of jurisdictions to tax individuals. It then reviews the nexus criteria used to justify the taxation of individuals on their worldwide income, gains, or profits and used to justify location-based (territorial) taxation of individuals located in other jurisdictions. In doing so, it seeks to highlight the origins of the key personal nexus criteria, the commonality and diversity of the nexus criteria adopted by jurisdictions, and poses the question whether the development of more objective tests could lead to more consistent nexus tests for individuals in the context of international taxation.

7.2 Sufficient Connection with the Individual While, in theory, public international law does not impose any limitations on a government’s power to tax,11 it is widely accepted that a government’s power is effectively limited to those persons who have a ‘sufficient connection’ to the jurisdiction. Those limitations in the cross-border context are related to a jurisdiction’s ability to collect the information from the individual or the authorities in another jurisdiction to support any assessment under the tax laws. Then, more importantly, its ability to enforce the collection of the taxes assessed. Even with the development of comprehensive bilateral tax treaties and tax information exchange agreements (TIEAs) and multilateral agreements (e.g. Multilateral Convention on Mutual Administrative Assistance in Tax Matters (MCMAA), the OECD Automatic Exchange of Information Agreement (MCAA) (incorporating the OECD Common Reporting Standard (CRS)), and the OECD’s Multilateral Competent Authority Agreement on the Exchange of Country- by-Country (CbC) reports) these limitations remain due to a consensus on what is acceptable tax policy. This consensus is illustrated by many jurisdictions refusing to grant tax credits for foreign taxes paid where those taxes (e.g. ‘credit absorption’ and ‘unitary’ 11

e.g. see the US Supreme Court decision in McCullock v. Maryland, 4 Wheaton 316, 429 (1819).

110 Michael Dirkis taxes) contravene the international tax norms of taxing business income upon the basis of net income.12 Although sufficient connection can be established through the existence of many different linkages between the individual and a jurisdiction, it is possible to classify these various linkages into two broad categories. The first category of criteria is based upon an individual having strong personal links to the jurisdiction due to their nationality/citizenship and/or their residency/domicile in the jurisdiction. It is these personal links to a jurisdiction that form the justification for its worldwide taxation of the individual (i.e. for both the income generated within the jurisdiction and that arising in other jurisdictions). The second category of sufficient connection criteria is based upon the individual having either some physical presence in the jurisdiction or deriving gains from their direct and indirect ‘ownership’ of real, personal, and intellectual property, and know- how located in the jurisdiction (i.e. territorial taxation). The two traditional sufficient connection links used by jurisdictions to tax non-residents are the place where the income is deemed to arise (source) or where payments to a non-resident originate (origin). Another nexus criterion is the location of property (situs: the place where something exists or originates). However, as situs is often treated as part of the source nexus criteria, it will be discussed in that context. The scope of each of these connective criteria is explored in the next two sections. The scope of both connective criteria of the first category used to justify the taxation of individuals on their worldwide income, gains, or profits will be examined first. Following that, the scope of the second category of sufficient connection criteria used to justify location-based (territorial) taxation of individuals located in other jurisdictions will be explored.

7.3 Worldwide Taxation of the Individual As a pure form of either a worldwide or territorial taxation system has not been adopted by any jurisdiction, jurisdictions often will use a number of different nexus tests to define the scope of their jurisdictional claim. Further, as each of these sufficient connection links have inherent limitations, variations are made by jurisdictions aimed at addressing these shortcomings. So, in working through each of the connective criteria

12 A

‘credit absorption’ tax seeks to increase the tax rate to the extent the home jurisdiction grants benefits to the taxpayer on that foreign income. A ‘unitary’ tax is where a jurisdiction imposes tax on the global income of a company (and its associates) which is calculated on the relative levels of business income measured by a proportion of global payroll, property, and sales that occur in that country rather than the usual basis of net income derived by the company in that country.

International Tax Law and Personal Nexus 111 for worldwide taxation, the focus will be upon the scope of each criteria. In doing so, it is important to remember that, depending on the jurisdiction, the connective criteria being explored may be interlinked with other nexus criteria and/or tiered according to a descending scale of connectivity to the jurisdiction. In fact, they may even be unrelated to any of the other prescribed connective criteria to ensure a particular policy outcome (e.g. to deem certain categories of persons to be connected to a jurisdiction regardless of whether they are captured by the broader connective criteria). The first and oldest tax policy for taxing persons on their worldwide income arises from the exchange theory of taxation. It seeks to tax persons based upon their personal political allegiance or nationality (i.e. citizenship). Inherent in this connective criterion is the existence of the sovereign right of states to tax their citizens and nationals on income, profits, or gains derived by them from property and assets held in other jurisdictions.13 This exchange theory of taxation, as reflected in the extract from Adam Smith’s Wealth of Nations at the start of the chapter, arises from the philosophical social contract that arguably exists between government and the individual. The exchange theory has two underlying competing elements, the cost (sacrifice) theory and the benefit theory. Under the cost theory, taxes represent payment by the individual (the sacrifice) for the benefits (services) provided to the whole of society (the community) by government.14 The benefit theory prescribes that the level of taxation should depend upon the level of benefits conferred by the government upon the individual. However, the theory has more commonly been reinterpreted as the level of taxation should depend upon the totality of state services provided to all taxpayers in society, rather than being a fee for each individual service provided by governments.15 The main advantage of using citizenship as a method of linkage is ease of tax administration when compared to the other forms of personal nexus. Citizenship is an ascertainable fact where other nexus policies involve analysing the specific facts and circumstances of an individual to determine whether the individual satisfies that link. However, it creates problems of double taxation where citizens are located in other jurisdictions which adopt alternative nexus policies such as the residency or domicile of the taxpayer, the economic interest or economic allegiance of the taxpayer, or the source of the income, profit, or gain.16 The exchange theory fell out of favour due to theoretical and equity concerns. As early as the late 1800s, it had been suggested that the concept:

13

E. R. A. Seligman, Essays in Taxation, 3rd ed. (New York: Macmillan, 1900), 108. Although the cost theory can underpin worldwide taxation based upon citizenship, its acceptance in the twentieth and twenty-first centuries has been questioned—see K. Vogel, ‘Worldwide vs Source of Income—A Review and Revaluation of Arguments (Part III)’, Intertax 11 (1988), 394–395. 15 See generally, Smith, The Wealth of Nations, II, 34; Bruins et al., Report on Double Taxation, 18; Vogel, ‘Worldwide vs Source of Income (Part III)’, 394–395. 16 P. B. Gann, ‘The Concept of an Independent Treaty Foreign Tax Credit’, Tax Law Review 38/11 (1982), 63–64 and 68. 14

112 Michael Dirkis in the modern age of the international migration of persons as well as of capital, political allegiance no longer forms an adequate test of individual fiscal obligation. It is fast breaking down in practice, and it is clearly insufficient in theory.17

The concerns with the exchange theory included its inability to provide a methodology for apportionment of the cost or the benefit between jurisdictions where there was property of an individual in one jurisdiction, but the individual was located in another jurisdiction. Although property located in jurisdictions where the individual was located was protected by that jurisdiction, the property located in the second jurisdiction was protected by the second jurisdiction not the resident jurisdiction. Under both the cost and benefit theories, both jurisdictions have the right to tax the income, gains, or profits arising from that property, but the theories contained no mechanism to resolve the resultant international double taxation.18 As a result of this loss of support, the citizen–nationality nexus also lost favour as the acceptable theoretical basis for worldwide taxation. Currently only the USA uses citizenship as its primary nexus for worldwide taxation of citizens regardless of where the individual resides.19 It also seeks to preserve that linkage for non-resident US citizens seeking to renounce their citizenship, where it appears it has been done for tax avoidance purposes. A derivative of the use of nationality is to use an individual’s immigration/visa status. The USA deems individuals who have lawfully entered the country as permanent residents (resident aliens) to be US residents and liable for taxes on a worldwide basis (the so-called green card test). The justification for the continued use of citizenship by the USA has been the subject of much academic examination. Gann in her 1982 seminal work notes that the tax policy debate in respect of the continuation of the policy in the USA has been less about the acceptability of citizenship as the nexus for the taxation of individuals than arguments about its impact upon the competitiveness of US businesses located in lower taxing jurisdictions caused by the higher wage cost of employing US citizens due to the imposition of higher effective tax rates (the US tax rate) upon the income derived by those employees in those jurisdictions.20 Gann concludes that: the case for retaining citizenship jurisdiction is not strong. It likely does not produce significant revenues. Politically, Congress has been unwilling to apply it in a neutral and equitable fashion to all U.S. citizens living abroad . . . [this] raises a substantial question about the soundness of retaining citizenship jurisdiction. If Congress continues to fail to apply it neutrally and equitably, then it should be eliminated

17 Seligman, Essays, 109. His views are adopted word for word in Bruins et al., Report on Double Taxation, 19. 18 See Bruins et al., Report on Double Taxation, 19. 19 Internal Revenue Code, 26 USC § 1. A citizen is a person born or naturalised in the USA—CFR § 1.1–1(c). 20 Gann, ‘The Concept of an Independent Treaty Foreign Tax Credit’, 69.

International Tax Law and Personal Nexus 113 altogether, thereby simplifying in several ways U.S. taxation and conforming it to the clear international preference for residence jurisdiction.21

Despite the doubts about citizenship as a principal nexus test for individuals, it is used in other jurisdictions to impose tax on citizens living abroad22 in respect of certain income types or in limited situations where the citizen resides in a non-tax treaty country.23 Citizenship is used as a linkage in respect of exit taxes in some jurisdictions.24 Nationality is also the fourth tier criteria of the individual resident tiebreaker tests in subparagraphs 2(c) and (d) of article 4 of the OECD Model Tax Convention on Income and Capital (2017 OECD Model) and the United Nations Model Double Taxation Convention between Developed and Developing Countries (2017) (UN Model). Other jurisdictions also use immigration status, or are contemplating the adoption of such criterion, as the basis for determining whether individuals have retained their residency.25 With the demise of the exchange theory and the adoption of the faculty theory (i.e. every individual should contribute to the support of the public burdens according to their ability to pay), the resident concept has emerged as the most widely adopted personal nexus criteria. Unlike citizenship, it is based strongly upon the geographical attachment of an individual. It is generally not connected with immigration status, but immigration status may be a factor in certain jurisdictions. There is no single legal concept of ‘residence’; what constitutes residence varies in the domestic law of most jurisdictions.26 In some jurisdictions, even though the term may be used more than 400 times in legislation, it may not be formally defined, resulting in the adoption of the common law facts-and-circumstances resident concept.27 In other jurisdictions, it may

21

Ibid., 68–69. imposes a 2% tax on net income from employment, rent on property (movable and immovable), and professional services derived by Eritreans working abroad: Eritrean Proclamation Gazettes Proclamation No. 67/195; ‘Tax Payment Proclamation for Eritreans in Diaspora Who Have Income’, https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_d ata/file/559482/Eritrea-Appendix-I-II-III-V-Dec-2015.pdf. 23 Non- resident Myanmar citizens are taxed at a reduced flat rate of 10% in respect of self- employment and business income from overseas sources and a non-resident. Hungarian citizens will not be taxed on foreign source income if they are resident in a tax treaty country—see EYGM Ltd, ‘Worldwide Personal Tax and Immigration Guide 2020–21’ (Jan. 2021), 644 and 1078 (respectively), https://www.ey.com/en_gl/tax-guides/worldwide-personal-tax-and-immigration-guide. 24 See generally, L. De Broe, ‘General Report: The Treatment of Transfer of Residence by Individuals’, Cahiers de Droit Fiscal International, vol. LXXXVIIb (2002), 23–78. 25 Board of Taxation (Australia), ‘Reforming Individual Tax Residency Rules— A Model for Modernisation’ (Mar. 2019, released 12 Dec. 2019), Recommendation 5, https://taxboard.gov.au/sites/ taxboard.gov.au/files/migrated/2019/12/Tax-Residency-Report.pdf. 26 P. E. Nygh, Conflicts of Laws in Australia, 3rd ed. (Woburn, MA: Butterworths, 1976), 145. 27 In Canada, the term resident is not defined in the Income Tax Act RSC, c.1 1985 (5th Supp.). Canadian tax law relies principally on the common law (the source jurisdiction for interpreting these words being pre-2013 UK jurisprudence) to determine residency of individuals—see Thomson v. Minister of National Revenue [1946] DTC 812. 22 Eritrea

114 Michael Dirkis be a series of individual nexus tests both objectively and subjectively expressed,28 or be a range of objective factors that cumulatively deem an individual to be a resident.29 Underlying all of these various linkage methods is a desire to determine the point at which an individual has sufficient connection to a jurisdiction such that it can be said that they ‘have one’s home’ there.30 Generally, the structure of residence nexus tests follows a pattern. They start with a primary criterion that establishes residence (e.g. resident/residence or domicile). Often this criterion will include an integrated or stand-alone minimum duration presence rule, usually measured in terms of days, and adhesive criteria that ensure residence is retained when the individual is absent from the jurisdiction. Finally, as mentioned previously, such nexus tests will often contain a specialized criterion to deem residence for specific categories of persons who would not otherwise qualify as residents. In common law jurisdictions, the meaning of resident (resides) in their respective taxation legislation has evolved through the influence of 200 years of litigation before UK courts and litigation in their own jurisdictions. From this extensive litigation, it is possible to distil a number of factors, indicators, and ‘objective facts which are frequently relevant to the determination of the nature and quality of a person’s presence in or association with a particular location’ that can indicate where a person resides.31 They include:

• • • • •

physical presence; term of any employment or appointment; nature of the person’s family, business, and social ties; frequency and regularity of a person’s movements; and intention (purpose) of a visit or trip.

Not all these factors are present in all cases and in different situations some factors will be more important than others. As a result, in more complex fact situations the existence of residence can be difficult to ascertain. This reliance on facts and circumstances is not unique to common law jurisdictions (e.g. Australia, Canada, Ireland, and New Zealand). It is also found in the taxation law of civil law jurisdictions (e.g. Japan, Italy, Netherlands, Spain, and Switzerland). The level

28 In Australia, the statutory definition ‘resident or resident of Australia’ in s. 6(1) of the Income Tax Assessment Act 1936 (Cth) consists of a ‘resides’ test and three statutory deeming tests. 29 In the UK, the Finance Act 2013, s. 218 and Sched. 45 in 159 sections prescribe a basic test, the automatic residence test, ‘automatic UK tests’, ‘automatic overseas tests’, and ‘sufficient tie test’. 30 F. G. Fowler and H. W. Fowler, eds, The Pocket Oxford Dictionary of Current English, 4th ed. (Oxford: Clarendon Press, 1946), 691. The Macquarie Concise Dictionary, 3rd ed. (Sydney: Pan Macmillan, 1998), 987 defines resides in similar terms—‘to dwell permanently or for a considerable time; have one’s abode for a time’. 31 Harding v. Commissioner of Taxation [2018] FCA 837, [46].

International Tax Law and Personal Nexus 115 of reliance in each jurisdiction on facts and circumstances varies.32 Given that tax treaty definitions of resident are determined by the domestic laws of the contracting states, it is not surprising that similar factors (e.g. permanent home, personal and economic relations (centre of vital interests), habitual abode, and nationality) are used in the individual resident tiebreaker tests in subparagraphs 2(c) and (d) of article 4 of the 2017 OECD Model and in the 2017 UN Model. An often-used primary criterion included in a resident test for individuals is the concept of domicile. Although it is a commonly used term, its meaning in civil law jurisdictions is different to that in common law jurisdictions. Under civil law, the term domicile has different definitions located in the various civil codes, codes on private international law, fiscal codes, or codes on court organization and jurisdiction. For example, in Italy the term domicile means the individual’s principal place of business or interests, in Germany it means a home or dwelling maintained by the individual in the long term, and in France tax domicile is where an individual, who exercise a professional activity in France (whether salaried or not), has their home or the place of their main location in France, or who have the centre of their economic interests in France.33 Similarly, the existence of a ‘permanent place of abode’ or ‘a usual place of abode’ for an individual is used in some common law jurisdictions as a stand-alone nexus criteria. Domicile under the common law is a highly technical legal concept, developed through litigation in the UK and adopted in most common law jurisdictions. Domicile is a legal relationship between a person and a country by which the person is able to invoke the country’s laws as their own.34 In other words, a person’s domicile is their ‘permanent’ home rather than where they reside. Under the common law, there are three types of domicile: domicile of origin, domicile of dependence, and domicile of choice. At birth, a person acquires a domicile of origin. The actual domicile adopted at birth is the domicile of the person upon whom the infant is legally dependent (i.e. a domicile of dependency).35 A domicile of choice is acquired in a country where a person voluntarily fixes their sole or chief residence with the intention to continue to reside there indefinitely. There must be a combination of both of these elements for a domicile of choice to exist.

32 Board of Taxation (Australia), ‘Review of the Income Tax Residency Rules for Individuals’ (2017, released 9 July 2018), Annexure C, https://cdn.tspace.gov.au/uploads/sites/70/2018/07/T307956-income- tax-res-rules.pdf. 33 See Codice Civile (Italy), art. 43(1); German Fiscal Code (Abgabenordnung), § 8; Code général des impôts (France), art. 4B (respectively). In Belgium, domicile refers to a factual situation characterized by the actual residence or living quarters in Belgium: Code judiciaire, s. 36. 34 Henderson v. Henderson [1965] All ER 179. See J. J. Fawcett, ‘Residence and Domicile of Individuals in Private International Law’, in M. Guglielmo, ed., Residence of Individuals under Tax Treaties and EC Law (The Hague: IBFD, 2010), 3 and 7–17. 35 At common law, if the person is legitimate, they acquire the domicile of their father, if the person is illegitimate, they acquire the domicile of their mother, and if the person is a foundling, they acquire the domicile of the country in which they were found—see Udny v. Udny (1869) LR I Sc & Div 441.

116 Michael Dirkis These common law rules are modified in a number of common law jurisdictions by legislation (e.g. Australia and New Zealand). This legislation supplements and/or overrides the common law. It does this by replacing the concept of domicile based on legitimacy with domicile using the place of residence, removing domicile of dependency for married women, and abolishing the rule whereby the domicile of origin revives upon the abandonment of a domicile of choice without the acquisition of a new domicile of choice. The domicile continues until the person acquires a different domicile.36 This last change clarifies that a person may only have one domicile at any time. The third primary criterion included in a resident test for individuals is a period of presence in, or absence from, the jurisdiction. As with the other nexus criteria already examined, there are variations between jurisdictions in how this criterion has been adopted. At a high level, there is some commonality in approach. The periods of presence are usually expressed in terms of a time period in the jurisdiction during a specified measurement period, as are the periods of absence.37 However, the detail in terms of both the qualifying period and the specific measurement period vary. For example, in the case of presence they are variously expressed in terms of an individual being present in the jurisdiction for more than one-half of the year of income, for the greater part of the fiscal year, or where their presence exceeds six months over two years.38 This can be further complicated, for example, the US substantial presence test is satisfied where an individual was present in the USA on at least thirty-one days during the calendar year, and the sum of the number of days on which such individual was present in the USA during the current year and the two preceding calendar years (when adjusted by a statutory formula that discounts the past year days) equals or exceeds 183 days.39 Other variations include New Zealand’s use of a days- away method to determine abandonment of residence (a period of absences of more than 325 days in total in a twelve-month period),40 and the Canadian use of the concept of an individual ‘sojourning’ under their 183 days presence test, rather than merely being ‘present’.41 The UK makes extensive use of a number of time-related criteria, based upon specific numbers of days or hours either present in, or absent from, the UK, in four statutory automatic residence tests and five automatic overseas tests.42 36

e.g. see the Australian Domicile Act 1982 (Cth). Income Tax Act 2007 (NZ), s. YD1(6). 38 Australia (Income Tax Assessment Act 1936 (Cth), s. 6(1)(a)(ii); Belgium (Code des impôts sur les revenus 1992, art. 2); Italy (Codice Civile, art. 43(1)); Germany (German Fiscal Code (Abgabenordnung) § 8), respectively. Strictly speaking, there is not an exceeds six-months nexus criteria under German law. Rather, § 9 provides for a ‘habitual residence’ test, but a period of residence of more than six months deemed to give rise to habitual residence in Germany. 39 Internal Revenue Code, 26 USC § 7701(b)(3). 40 Income Tax Act 2007 (NZ), s. YD1(6). 41 Income Tax Act RSC, c. 1 1985, (5th Supp.), s. 250(1)(a). ‘A sojourner is a person who is physically present in Canada, but on a more transient basis than a resident. A sojourner lacks the settled home in Canada which would make him or her a resident’—see J. S. Wilkie, J. Li, and J. E. Magee, Principles of Canadian Income Tax Law, 9th ed. (Toronto: Thomson Reuters, 2017), para 3.2(b). 42 See UK Finance Act 2013, s. 218 and Sched. 45, paras 6–16. 37

International Tax Law and Personal Nexus 117 There are also a number of miscellaneous objective and subjective nexus criteria tests that deem individuals to be residents. These criteria include deeming individuals to be resident if they are enrolled on a national register of population,43 are state officials (diplomats, military personnel, etc.) who exercise their functions or are responsible for a mission in a foreign country and their families,44 are members of two prescribed government pension funds which no longer admit new members,45 or have their ‘seat of wealth’ located in the jurisdiction (i.e. the place from where the individual’s assets are managed).46 Finally, where nexus criteria were enacted prior to the creation of a tax treaty network, one jurisdiction placed qualifications on their domicile and presence tests aimed at avoiding possible double taxation.47 What is clear from this examination of personal nexus criteria is that, on the surface, there is commonality in labels (resident, domicile, days present in the jurisdiction) but what is enacted by jurisdictions varies. In some cases, similar terms have different meanings, and in some cases, they are used for completely different purposes. Regardless of these differences, the nexus criterion in most jurisdictions contains a balance of both ‘facts and circumstances’ tests and objective/bright-line tests. The use of facts-and-circumstances tests/requirements to determine an individual’s residence has become more contentious with the advent of revenue authorities adopting self-assessing regimes. Although it is claimed that the facts-and-circumstances tests are more ‘compatible with the political theory that government power comes from the consent of the governed’,48 they do not satisfy a number of tax policy objectives. Under some facts-and-circumstance tests minor variations in a taxpayer’s circumstances may result in taxpayers in similar circumstances being taxed differently and the horizontal equity criterion not being met. Secondly, facts-and-circumstance tests will generally fail the simplicity criterion. The case-by-case determinations of all key concepts mean that the rules can give rise to arbitrary outcomes, impose high compliance costs, create uncertainty, and be hard to administer.49 Finally, as the tests rely upon individual facts and circumstances, they are easy to manipulate therefore they generally fail the efficiency (anti-tax avoidance) criterion.50

43

Belgium (Code des impôts sur les revenus 1992, art. 2); Italy (Codice Civile, art. 43(1)). Income Tax Act RSC, c. 1 1985, (5th Supp.), s. 250(1)(c), (d.1), (f), and (g), Code général des impôts (France), art. 4B; Income Tax Act 2007 (NZ), s. YD1(7). 45 Australia: Income Tax Assessment Act 1936 (Cth), s. 6(1)(a)(iii). The spouse or a child under 16 years of age of such a person is also deemed by that relationship to be a resident under the test. 46 Belgium: Code des impôts sur les revenus 1992, art. 2. 47 See Australia: Income Tax Assessment Act 1936 (Cth), s. 6(1)(a)(i) and (ii). 48 M. McIntyre, The International Income Tax Rules of the United States, 2nd ed. (Stoneham, MA: Butterworths, 1992) 1–21. 49 P. Whiteman et al., eds, Whiteman on Income Tax, 3rd ed. (London: Sweet & Maxwell, 1988), 137. 50 M. Dirkis, Is It Australia’s? Residency and Source Analysed, Research Study No. 44 (Sydney: Australian Tax Research Foundation, 2005), specifically ch. 3 examining individual residency. 44 See

118 Michael Dirkis In an attempt to overcome these weaknesses, the UK has, and Australia is intending to, move to a residence test based upon specific factual criteria. These rules seek to remove any reliance on individual facts and circumstances in determining residence.51 In 2013, the UK inserted a statutory definition into UK tax law which replaced almost 200 years of reliance on the jurisprudence that constitutes the common law resident test. The 159 sections of complex legislation give effect to a web of seventeen interconnected tests, numerous detailed definitions (covering elements such as what constitutes a ‘day spent’, a ‘home’, ‘work’, ‘location of work’), and extensive specific anti-avoidance rules (many focused on countering actions to avoid the scope of the defined terms). The matrix starts with the ‘basic test’ which classifies a person to be a resident if they satisfy either the ‘automatic residence test’ or the ‘sufficient ties test’. If they satisfy neither, they are a non-resident. The ‘automatic residence test’ is met if the individual meets one of the four ‘automatic UK tests’ (which are a ‘183 day’ rule, ‘has a “home” in the UK where sufficient time is spent’ test, a ‘full time work in the UK’ test, and a ‘year of death’ test) and none of the five ‘automatic overseas tests’ are satisfied. Under the ‘sufficient ties test’, individuals will be resident if they do not meet any of the ‘automatic UK tests’ and the ‘automatic overseas tests’, and they have sufficient ‘UK ties’ (a family tie, an accommodation tie, a work tie, a ninety-day tie, and/or a country tie). The number of these five specified ‘UK ties’ needed for an individual to be classified as ‘sufficiently’ tied to the UK depends upon the number of days spent in the UK over a three-year period. The outcome of this new definition has been that the complexity of subjectiveness has been replaced with legislative complexity.52 One commentator has concluded that although the new rules bring ‘a greater certainty when dealing with the complex situations of internationally mobile individuals’, subjective areas still exist, and complexity remains due to the ‘maze of the tests combined with their subsidiary tests and the definitions which are either confusing or illogical at times’.53 The proposed Australian reforms, which were endorsed by the government in May 2021, share a number of similarities with the UK changes. Under the principal ‘183’-day test (the bright-line test), an individual will be a resident if they are present in Australia for 183 days or more in the income year. If the bright-line test is not satisfied, the next step in the matrix is dependent upon the individual’s prior income year residence status. If they were resident in the prior income year, then the ‘ceasing residency’ rule may apply. If not, the ‘commencing residency’ rule may apply. Under the ‘commencing residency’ rule, the individual will be a non-resident if they spent fewer than forty-five days in Australia in the prior year. If they spent forty-five days or more, they will be a resident 51 See UK Finance Act 2013, s. 218 and Sched. 45 and, for Australia, Board of Taxation, ‘Reforming Individual Tax Residency Rules—A Model for Modernisation’ (n.d.). 52 For a detailed analysis of the rules, see M. Dirkis, ‘Moving to a More “Certain” Test for Tax Residence in Australia: Lessons for Canada?’, Canadian Tax Journal 68/1 (2020), 143. 53 A. Florczak, ‘The New Statutory Definition of Residence for Individuals in the UK in the Light of the Tax Treaty Dual Residence Rules’, MA diss., Institute of Advanced Legal Studies, University of London (2013), 50, http://sas-space.sas.ac.uk/5887/1/Anna%20Florczak%20MA%20Dissertation.pdf.

International Tax Law and Personal Nexus 119 if they satisfy two or more of the four criteria (i.e. the individual has citizenship/permanent residency, has accommodation in Australia, has family in Australia, and economic connections in Australia) under the ‘factor’ test. The operation of the ‘ceasing residency’ rule varies depending upon whether the individual is classified as a ‘long-term’ resident (resident for the three prior years) or a ‘short-term’ resident (under three years). If they are a long-term resident, an adhesive ‘ceasing long-term resident’ test applies which will only deem non-residency if the individual has spent fewer than forty-five days in Australia in the current and two prior income years. Under the ‘ceasing short-term residency’ test, the individual will only be deemed to be non-resident if they have spent fewer than forty-five days in Australia in the current year and satisfy less than two of the factors under the ‘factor’ test. There is also a new ‘government officials’ test and an ‘overseas employment’ rule. So, as with the UK model, the proposed Australian reform highlights the legal complexity caused by moving away from a facts-and-circumstance model, particularly where compliance objectives override the touted objectives of simplicity and certainty. These changes are significant for international taxation as it is rare for jurisdictions to significantly alter their resident tests. The models adopted do offer an alternative, which could be adopted more broadly by jurisdictions in an attempt to create more commonality in the nexus test. This also has the potential to remove the facts-and-circumstance elements in the nexus criteria adopted in many jurisdictions and in the tiebreaker rules in tax treaties. Such changes could reduce the arbitrary outcomes, high compliance costs, uncertainty, and administrative difficulties that arise from case-by-case determinations. Overall, what is illustrated by the UK and proposed Australian rules is that, with the quest for certainty, the conversion of a complex facts-and-circumstances policy into a ‘certain’ outcome can come at a cost of complexity and a potential loss of equity.

7.4 Territorial Taxation of the Individual As mentioned earlier, the second category of sufficient connection criteria is based upon the individual having either some physical presence in the jurisdiction or deriving gains from their direct and indirect ‘ownership’ of real, personal, and intellectual property, and know-how located in the jurisdiction (i.e. territorial taxation). Territorial taxation jurisdictions use a number of different nexus tests, based upon source (including situs) or based upon origin, to define the scope of their jurisdictional claim over non-resident individuals. The first and traditional connective criteria tax policy for taxing non-resident persons used by jurisdictions is the place where the income is deemed to arise (source). Territorial source is based strongly upon geographical attachment (i.e. a transaction

120 Michael Dirkis being conducted within the jurisdiction or the location of property or where the income is deemed to arise). Source, according to Vogel, is the preferred nexus rule for taxation of income from labour ‘performed in a more than transitory way’ on both efficiency and equity grounds.54 However, he does not believe that is the case for all other forms of income and argues that in some circumstances taxing rights should be split between the state of residence and the source of the payment (e.g. income from rentals and royalties). Vogel’s approach of determining the most appropriate nexus test in relation to each specific type of income illustrates that the source nexus is dependent on the characterization of the income. Unlike residency, source is not purely a matter of geography as the source of income depends upon how it is produced, where it is produced, and who pays it.55 This fact is reflected in the varied way jurisdictions deal with source-taxing rights of non-resident individuals in respect of particular types of income and in the role of tax treaties which, in order to prevent double taxation and double non-taxation, allocate those source-taxing rights between contracting jurisdictions.56 In most common law jurisdictions, the determination of source for non-resident individuals relies heavily on guiding principles emerging from judicial decisions, physical or territorial location, and the legal characterizations of income. In some cases, this is supplemented by statutory source rules (usually origin based). Under generally accepted guiding principles, the source of income is determined by weighing up the individual facts and circumstances from the perspective of the non-resident individual. As noted previously, the determination of source is reliant upon establishing the nature of the income derived by the non-resident individual. As a result, it often focuses on the legal form of a transaction rather than its substance. Thus, a finding of source could turn on a minor variation in circumstances or, hypothetically, even on a difference in a single fact. So, the concerns expressed in relation to the use of facts-and-circumstances criteria in the context of the residence nexus tests apply to the determination of source. Again, seemingly arbitrary variations in outcomes in determinations impact upon horizontal equity and the need for the case-by-case determinations of source means the guiding principles impose high compliance costs, create uncertainty, and are hard to administer. In those jurisdictions where statutory source rules have been enacted to supplement the guiding principles, they tend to be narrowly focused, not necessarily governed by any general principles, and have their genesis in earlier tax legislation.57 Often withholding tax rules for dividends, interest, and royalties will be origin-based and do

54

See Vogel, ‘Worldwide vs Source of Income (Part III)’, 401–402. R. Couzin, Corporate Residence and International Taxation (Amsterdam: IBFD, 2002), 6. 56 Tax treaties also allocate resident taxing rights. The major difference between the OECD and UN Model Treaties is principally that the UN Model Treaty allocates greater taxing rights for the source country as opposed to those of the residence country. 57 e.g. Australia’s statutory source rules are limited to dividends, certain royalty payments, certain natural resource payments, overseas shipping payments, and certain insurance premiums. The ship charterer source rules were adopted from the Land and Income Tax Assessment Act 1891 (NZ), Sched. C (1). 55

International Tax Law and Personal Nexus 121 not determine source. They merely compel the withholding of tax due to the location of the payer (i.e. the payment to a non-resident individual originates in the jurisdiction).58 Most civil law jurisdictions prescribe in their domestic law the types of income produced or collected in the jurisdiction which are taxable. In doing so, they effectively incorporate the connecting factors (i.e. the circumstances and conditions that create a sufficient link to the jurisdiction) for each specified type of income received by the non- resident individual (e.g. income from real estate located in X).59 In that way, source allocation is achieved without specific source of income allocation rules. The embedding of the nexus criteria in the specific provisions for taxing non-resident individuals provides much more certainty than the guiding-principles approach adopted in common law jurisdictions. It also delivers lower compliance costs and assists in administration. Given the perceived benefits of codified source rules, some common law jurisdictions have proposed the adoption of a statutory code.60 There are already two common law jurisdictions that have extensive statutory source rules—New Zealand and the USA.61 Both were enacted in the early twentieth century and they present two very different models. The New Zealand model is solely focused on source rules for non-residents, while the US regime focuses on domestic and non-resident source rules.62 Under New Zealand’s rules, the classes of income covered are extensive, but not exhaustive. There can be resort to the common law guiding principles if the income in question is not expressly caught by the rules.63 The US source rules are more exhaustive. However, despite the width of the source regimes neither of the source regimes cover all classes of income. Both statutory schemes do improve clarity by setting out the criteria needed to make a determination of source in respect of a non-resident individual. However, the experience of the US code is that codification can lead to complexity. Further, the New Zealand model still relies on some facts-and-circumstance determinations under the common law, which can give rise to complexity. Thus, despite setting out the source rules in a more coherent manner, both models do not fully satisfy the simplicity criterion. That said, the two statutory schemes demonstrate that it is possible for common law

58 ‘The

justification for imposing income tax on non-residents on the basis of origin . . . rests on the “benefit” principle, [that is,] the non-resident’s income has been generated by economic activity conducted under the protection of the country of origin and relying on facilities provided, at least in part, at public expense’—see Taxation Review Committee, Commonwealth (Australia), Full Report (Canberra: AGPS, 1975) (Asprey Report), 276. 59 Belgium (Code des impôts sur les revenus 1992, 228 §§ 1– 2); E. Schoonvliet, ‘Belgium: Source and Residence: New Configuration of Their Principles’, Cahiers de Droit Fiscal International vol. 90a (2005), 190. 60 e.g. Review of Business Taxation (Australia), A Tax System Redesigned (Canberra: AGPS, 1999), Recommendation 23.2(c), 684–685. 61 New Zealand’s rules are principally contained in s. YD 4 of the Income Tax Act 2007 (NZ) and the USA in 26 CFR § 1.861–8(f). 62 The US rules enable residents to obtain foreign tax credits where taxes are paid in respect of income with a foreign source. 63 See G. A. Harris, New Zealand’s International Taxation (Auckland: Oxford University Press, 1990), 3.

122 Michael Dirkis jurisdictions to create simpler and more certain non-resident individual nexus criteria simply by codifying the guiding principles.

7.5 Observations and Conclusions This brief examination of the nexus between the person and international tax has confirmed that, as with domestic taxation, ultimately the tax liability is met by the individual. This occurs regardless of the scope of liability imposed (worldwide or territorial taxation). What is also evident is that, despite certain structural commonality, there is great variation between jurisdictions in respect of both the sufficient connective criteria used to justify the taxation of individuals on their worldwide income and to justify location-based (territorial) taxation of individuals located in other jurisdictions. These variations leave open the opportunities for gaps and frictions within the international tax consensus-based framework and they were not addressed by the G20/OECD BEPS Project. For both common law and civil law judications, replacing the facts-and-circumstances and the guiding-principles-based personal nexus criteria with more objective criteria could reduce arbitrary outcomes, high compliance costs, uncertainty, and administrative difficulties. The risk is that the jurisdictions will not be merely satisfied with simplification but will use it as an excuse to extend jurisdictional claims.

Chapter 8

Internationa l Tax L aw an d L ow- and Mi ddl e - Inc om e C ou nt ri e s Akhilesh Ranjan

8.1 Introduction No matter what any country may want to do with its tax system, or what anyone might think it should do from one perspective or another (ethical, political or developmental), what it does do is always constrained by what it can do.1

The tax policy of any country is influenced by a combination of factors, such as the strategy for growth, prevailing economic environment, and political imperatives. Developing countries, including the low-and middle-income countries (LMICs) are also guided by the same factors but, additionally, tax policy formulation in these nations must take into account the capabilities of the tax administration, technological capacity, social welfare needs, and, above all, the level of trust between the government and the people. The relative importance of each of these factors will vary significantly across the wide spectrum of developing countries, which range from small economies struggling with political strife to much larger ones like Brazil, India, and Indonesia. However, while these countries may face very different tax challenges, they will all be driven by the common objective of achieving sustained growth in revenue mobilization in a politically feasible manner, in order to meet the ever-increasing demands of public spending.

1 R. M. Bird, ‘Tax Challenges facing Developing Countries’, Inaugural lecture of the Annual Public Lecture Series of the National Institute of Public Finance and Policy, New Delhi (2008), http://ssrn.com/ abstract=1114084.

124 Akhilesh Ranjan Historically, the tax revenue mobilized by developing countries has averaged around 20% of gross domestic product (GDP), with some LMICs achieving a much lower 10– 12% of GDP, against the average of well over 30% across member states of the OECD. The second half of the twentieth century saw widespread efforts by developing countries towards reforming their tax systems and putting in place efficient tax policies aimed at maximizing revenues. The period also witnessed elaborate expositions from a number of well-meaning economists and tax experts on how developing countries could (and should) raise tax revenues. Global institutions such as the International Monetary Fund (IMF) and the World Bank (WB) examined the fiscal policy parameters of developing countries and made their own suggestions on fiscal measures that these countries ought to take, sometimes even linking these suggestions to the ability of developing countries to draw funds from them. At the turn of the millennium, the UN established a set of Millennium Development Goals for developing countries to be achieved by 2015 and commissioned the UN Millennium Project. The Project report submitted in 2005 inter alia identified the resource requirements of developing countries and indicated the need for a significant scaling up of the tax–GDP ratio. In this environment and development context, the focus of governments of LMICs in the area of tax policy has understandably been on domestic policy measures, with little attention paid to the international tax framework. The number of home-grown multinational enterprises (MNEs) has been very limited in most LMICs and consequently there has not been much concern over double taxation faced by domestic businesses, or about the possible erosion of tax bases or profit shifting by domestic enterprises. The taxation of foreign enterprises and non-residents has essentially been considered as dealing with a special category of cases, the scope of such taxation being in line with basic principles incorporated in the law but often without the benefit of comprehensive or well-defined source rules. Thus, the approach to international taxation over the years in developing countries has been largely to follow and try to apply the principles set out in the model tax conventions framed by the OECD and by the UN, utilizing the guidance contained in the commentaries accompanying these conventions. The exposure of tax administrations to international taxation has been confined to attending workshops and seminars conducted by the OECD and then attempting to import some of the best practices that could potentially be implemented in the respective countries. While many of the developing countries have entered into double taxation avoidance agreements (DTAAs) with economically advanced countries, the primary purpose of doing so has been to facilitate investment from the treaty partner country by agreeing to limited source- taxation rights along with mechanisms for eliminating double taxation. Indeed, several organizations and bodies have been critical of the wisdom of these countries signing certain tax treaties at all, including notably the IMF, which declared in 2014 that ‘developing countries would be well-advised to sign treaties only with considerable caution’.2

2

IMF, ‘Spillovers in International Corporate Taxation’, IMF Policy Paper (2014).

International Tax Law and Low- and Middle-Income Countries 125

8.2 The International Tax Architecture The architecture of international taxation had already been established globally by the time most of the LMICs came out of their colonial or monarchical existence to become sovereign republics. The two fundamental issues of allocation of taxing rights and elimination of double taxation had been addressed as long ago as 1923 in the Economic Experts’ Report to the League of Nations, recommending that the primary right of taxation be vested in the country to which the taxpayer owed ‘economic allegiance’. This place of economic allegiance was, in turn, to be determined by four factors; namely, the place of origin of wealth, the place where the rights to the wealth could be enforced, the location of the wealth, and the place where it was consumed or appropriated (or the place of residence or domicile). In a world that essentially comprised the imperial powers of North America and Europe and the poor colonies of Asia, Africa, and Latin America, the economic allegiance almost invariably rested with the imperial powers, the colonies being merely the location for sourcing raw materials and finding markets. On the allocation of income and profits, the Expert Group did not give credence to the common-sense approach of formulary apportionment (which, many believe, would also have been fairer and more reasonable). It proceeded to classify income into different categories and then assigned an allocation rule to each category on the basis of the factors identified as contributing to economic allegiance. The primary result of this exercise was that source countries—which provided the raw materials, labour, and markets—were left with taxing jurisdiction only to the extent of the operations carried out in those countries. This was justified on the ground that it would facilitate cross-border investment and trade by effectively mitigating the risk of double taxation, with source countries having the right to impose withholding taxes on specific items of income and the country of residence providing the corresponding tax credit. The development of the concept of ‘permanent establishment’ (PE) and the model tax conventions brought out in 1928 restricted the taxation of business income in source countries to income that could be said to be directly attributable to a PE. Over a period of time, the concept of PE itself was modified and became narrower and more restricted in scope. For instance, an agency PE not only required a dependent agent, but also an agent who had the authority to habitually conclude contracts on behalf of the enterprise. Subsequently, in 1960, eighteen European countries came together with Canada and the USA to form the OECD with the objective of mutual cooperation to achieve the highest sustainable economic growth for the member countries. The OECD Model Convention they agreed upon duly incorporated the strict conditions and restricted scope of the concept of PE. The convention, together with the Commentary on its provisions, has since formed the template for almost all tax treaties entered into by countries, including in the developing world.

126 Akhilesh Ranjan The principles set out by the OECD Model Convention were adopted by the UN Model Convention, with a few modifications to reflect some of the interests of developing countries. The UN had recognized the desirability of developing countries entering into tax treaties with developed countries as a means of enhancing international cooperation for achieving developmental goals through improved flow of trade and investment as well as transfer of technology. In 1967, the UN set up an Ad-hoc Group of Tax Experts, comprising members from developed as well as developing countries participating in their individual capacities. The recommendations of the Expert Group led to the publication of the UN Model Convention in 1980, which was quite similar to the OECD Model. The similarity was based on the need to maintain consistency with international tax principles that had been established by that time. Divergences were focused on reflecting differences in approach as exemplified in country practices. The UN Model essentially sought to achieve a balance between the taxing rights of the ‘residence’ and the ‘source’ countries by ensuring that taxing rights given under the domestic law of source countries were surrendered only under conditions that were politically acceptable and economically beneficial. Meanwhile, the expanding reach of MNEs gave rise to transfer pricing concerns for the tax administrations in LMICs that now had to try to ensure that these MNEs paid the right amount of tax in their jurisdictions, based on a fair and equitable allocation of profits made from the local operations. Again, the OECD had already considered these issues, at least from the standpoint of the countries in which the MNEs were headquartered. The OECD members realized that the residence-based or source- based tax systems established by different countries, when applied to MNEs, would not only lead to an inequitable allocation of profits but would also heighten the risk of double taxation. According to them, these risks could be mitigated by the separate entity approach where each entity of the group would operate on an arm’s-length basis and would be taxed on its income arising either on the residence or the source principle. The OECD Transfer Pricing Guidelines published in 1995 laid out detailed guidance on how the arm’s-length principle could be used with the separate entity approach to achieve a basis for allocation of income that best approximated open market conditions. However, while several LMICs did introduce transfer pricing rules in one form or another, they faced significant challenges in the application of the arm’s-length principle in accordance with the 1995 Guidelines. The near-complete absence of relevant information about the MNE group coupled with the difficulty in finding reliable comparators in the form of domestic enterprises made it virtually impossible to apply the arm’s-length methodology. Further, the economically relevant circumstances or comparability factors, which formed the basis of the application of the principle, underscored the primacy of the contract and placed a disproportionately high emphasis on supply-side factors or production variables. They did not place much importance on the availability of raw materials or to demand-side factors such as purchasing power, depth of the market, and marketing intangibles created by the market economies. The contribution of the low-and middle-income host countries in generating profits was therefore assessed as marginal, resulting in

International Tax Law and Low- and Middle-Income Countries 127 those countries getting little more than a small return on costs under the transfer pricing rules. The UN Tax Committee published its Practical Manual on Transfer Pricing, but this largely followed OECD guidance. The UN Manual in its present form includes an annex in which some larger developing countries explain how they take different approaches, including by adapting rules to reflect the strengths of their markets in order to allocate a greater share of the tax base to themselves. For instance, China and India talk about location savings and marketing intangibles, while Brazil has introduced a greatly simplified transfer pricing regime involving types of safe harbours. However, the smaller economies have generally been left to fend for themselves and implement the OECD guidelines to the best of their ability. Thus, the international tax architecture has developed over the years without much input from or consideration of developing countries, particularly low-and middle- income ones. Consequently, it represents a skewed framework that leans towards centres of production and exporters of capital and technology. There have been sporadic attempts to carry out much-needed reforms, but the solutions arrived at by developed countries are not always relevant to developing nations because of the different nature of economic activity, tax policy, and administrative capacity.

8.3 The Age of Cooperation More recently, however, and particularly after the global financial meltdown of 2008, countries both developed and developing have realized the importance and need for international tax cooperation in achieving sustainable growth and optimizing tax revenue mobilization. Some of the major multilateral initiatives and movements in this regard have been: (1) the G20/OECD (Base Erosion and Profit Shifting (BEPS) Project); (2) the exchange of information and mutual administrative assistance engendered and monitored by the Global Forum; and (3) the Addis Ababa Action Agenda set out in 2015 under the auspices of the UN for the achievement of specified Sustainable Development Goals (SDGs) by 2030. The BEPS Project, for the first time ever, saw G20 and OECD member states working together on an equal footing to develop international tax standards, giving non-OECD developing countries a seat at the table in the decision-making Committee on Fiscal Affairs of the OECD. This was in response to the call of the G20 leaders who recognized that ‘developing countries should be able to reap the benefits of a more transparent international tax system, and to enhance their revenue capacity, as mobilizing domestic resources is critical to financing development’.3 Later, to make the process more inclusive, fourteen more developing countries were invited to attend meetings of the

3

G20 Leaders’ Declaration, St Petersburg Summit, 2013.

128 Akhilesh Ranjan Committee on Fiscal Affairs as observers. Regional consultation events were held for countries in Latin America and Africa through organizations such as the African Tax Administration Forum (ATAF) and the Inter-American Centre of Tax Administrations (CIAT). Clearly, tax reforms of the nature and scale envisaged under the BEPS agenda could not be possible without the consent and cooperation of countries across the globe, including developing countries. The action items formulated for the project involved making changes in a large number of bilateral tax treaties as well as tax measures required to be taken in domestic law by the maximum number of countries in order to ensure coherence and a level playing field. While reporting to the G20 Development Working Group on how the BEPS Project would affect low-income developing countries, the OECD not only enumerated the key BEPS issues and corresponding action items that were more relevant to them but also discussed three issues outside the BEPS remit that were of concern to them and needed to be addressed separately:4 • the pressure faced to provide tax incentives in order to attract foreign investment; • the lack of ‘comparability’ data needed to administer transfer pricing, without which it was difficult for tax authorities to challenge the income declarations made by MNEs; and • the avoidance of capital gains tax by MNEs on the sale of valuable assets in their countries through the use of ‘indirect transfers’, especially in the extractive and telecommunications industries. Subsequently, toolkits giving guidance on each of these concerns were released by the Platform for Collaboration on Tax (PCT), a joint effort of the IMF, OECD, UN, and the World Bank Group (WBG) launched in April 2016 to provide capacity-building support and technical advice to developing countries in their pursuit of the SDGs. When the BEPS process moved in early 2016 from laying down standards to their actual implementation, the Inclusive Framework (IF) was created, allowing LMICs to participate in the implementation on an equal footing if they were ‘interested and committed’, which meant that they must agree to implement at least the four ‘minimum standards’ that the G20 and OECD countries had agreed upon. The Framework now has 139 members. Separately, a Multilateral Instrument (MLI) was negotiated by an ad hoc group of more than a hundred countries, to swiftly implement the treaty-related changes recommended by the various BEPS reports to the existing network of hundreds of bilateral tax treaties. During the same period, OECD and G20 members also decided to put in place systems that would allow exchange of information about taxpayers who have assets, income, or operations in multiple jurisdictions, particularly low- tax or no- tax jurisdictions. The Global Forum on Transparency and Exchange of Information, which 4 OECD,

Report to G20 Development Working Group on the Impact of BEPS in Low Income Countries (Part 1) (Paris: OECD Publishing, 2014).

International Tax Law and Low- and Middle-Income Countries 129 now has more than 140 jurisdictions as members, set the standards for such exchange through comprehensively revised information-exchange articles in DTAAs, stand-alone bilateral Tax Information Exchange Agreements, and the Multilateral Convention on Mutual Administrative Assistance and Exchange of Information (commonly referred to as MAC). The Forum continues to conduct peer reviews of countries’ compliance with the standards and is backed by the threat of ‘defensive measures’ from the G20. The MAC together with a set of other bilateral and multilateral agreements also provided the legal framework for automatic exchange of financial account information. About one hundred countries have signed on to these agreements, committing to the Common Reporting Standard (CRS) for Automatic Exchange of Information (AEOI) devised by the OECD in consultation with Global Forum members. At the UN, the additional financial resources required for achieving the specific development goals formulated were traditionally sought mainly in the form of development assistance from international organizations such as the IMF and WBG. A turning point came with the agreement reached in 2015 in the UN General Assembly on the adoption of the Addis Ababa Action Agenda for achieving a set of sustainable development goals by 2030. This brought a commitment from all countries to a set of universal, integrated, and transformational goals and targets codified in the Agenda. More importantly, the Agenda strongly emphasized the role of international tax cooperation in augmenting domestic resource mobilization. This, in turn, provided a fresh and strong impetus to developing countries, including the least developed countries, for joining and actively participating in the discussions at multilateral forums such as the IF, individually as well as through regional groupings like the ATAF. It also gave new direction and energy to the work being done by the UN Committee of Experts, although a proposal supported by almost all developing countries to upgrade the status of the committee from a grouping of individual tax experts to an intergovernmental body was not accepted at the Addis Ababa session. The committee started its own independent examination of current issues in international taxation instead of merely adapting the reports and guidance issued by the OECD to suit the needs of developing countries. A prominent example of such independent work is the inclusion of article 12A in the UN Model Convention as a separate article to allocate taxing rights in respect of fees for technical services. Further, the 2021 update to the UN Model Convention has incorporated another new article 12B, with a view to addressing concerns expressed by developing countries and providing a simplified manner of taxing income from automated digital services.

8.4 Tax Challenges of Low-and Middle-Income Countries Despite the stepping up of international tax cooperation, the economic imperatives peculiar to the LMICs continue to raise several critical tax issues that are difficult to

130 Akhilesh Ranjan resolve and result in these countries being unable to adequately exploit and utilize their tax bases. While substantial efforts were made by the OECD to take into account the perspectives of developing countries in analysing BEPS issues, it was always clear that the problem of tax avoidance and the resultant base erosion in developing countries, particularly in the LMICs, needed to be looked at in a different context. The OECD in its report to the G20 Development Working Group duly acknowledged this by stating that, ‘It is important to recognize that the risks faced by developing countries from BEPS, and the challenges faced in addressing them, may be different both in nature and scale to those faced by developed countries.’5 The LMICs are more concerned with less- than-fair source taxation of income rather than with the shifting of profits by domestic enterprises. Some of the major tax challenges faced by these countries are discussed in the following subsections.

8.4.1 Continuing Imbalance in Allocation of Taxing Rights The regional consultations that OECD launched in the initial months of the BEPS Project clearly indicated that the apparent unfairness of the international tax architecture in the way it allocates taxing rights between ‘residence’ and ‘source’ jurisdictions continues to be a major concern across developing countries. Several organizations and academic circles have been calling for a fundamental re-examination of the way in which tax treaties still deprive developing countries of significant taxing rights.6 The issue was not included in the BEPS agenda on the ground that it was not an issue of tax planning or tax avoidance. However, the OECD acknowledged in its report to the G20 Development Working Group that it called for a ‘legitimate debate’. The restrictive definition of PE, which revolves around the notion of a fixed place of business of the enterprise, has allowed MNEs to engage in several types of business ventures and projects in developing countries without triggering any taxable nexus in those jurisdictions. A prime example of such an arrangement is that of a large turnkey project, where the major part of the composite contract is stripped away into an offshore supply contract that cannot be taxed in the source jurisdiction in the absence of a PE, while the onshore installation and services contract carried out by a local company is rewarded with a nominal return on costs. Another example is that of telecasting companies that are able to distribute their channels across a jurisdiction and earn substantial advertising revenue without having to pay any significant tax in that jurisdiction in the absence of a fixed place of business. In either of these examples, it would also not

5 Ibid.

6 See, e.g., P. Pistone, ‘Tax Treaties with Developing Countries: A Plea for New Allocation Rules and a Combined Legal and Economic Approach’, in, M. Lang, P. Pistone, J. Schuch et al., eds, Tax Treaties: Building Bridges Between Law and Economics (Amsterdam: IBFD, 2010), 413–440.

International Tax Law and Low- and Middle-Income Countries 131 be possible to hold the local enterprise to be a dependent agent PE, as it would not have the authority to ‘habitually conclude contracts on behalf of ’ the foreign enterprise. Such business models could be prevalent in dealings between developed economies also, but for developing countries they lead to a substantial drain on resources without getting any share in the tax on the business profits made. As noted by Professor Klaus Vogel: If the flows of goods between the two countries involved—or rather, more accurately, the profits resulting from those flows—are balanced, the question of what principle should be applied when distributing taxation is of relatively little significance, and in such a case adoption of the permanent establishment principle is recommendable because it is practicable. But if the flows are in imbalance, the recipient State is justified in requiring to be allowed to participate in the taxation of the proceeds of the sales of the goods—in the same way as it participates where interest and royalties are involved.7

The BEPS Project did address the issue of artificial avoidance of PE in Action 7 of the BEPS agenda and made recommendations relating to commissionaire arrangements, anti-fragmentation rules, and the prevention of splitting of installation contracts. However, these issues all arose from artificial structures set up to avoid PE status and had nothing to do with the definitional issues that prevented legitimate source taxation of business profits. In any event, while the Action 7 recommendations were all incorporated in the MLI, these were not given the status of ‘minimum standards’ with the result that at present not many developed countries have chosen to adopt the relevant articles of the MLI in their treaties. The issue of unreasonably restricted source-taxation rights in respect of business income has been exacerbated by the digitalization of the economy that has enabled MNEs to penetrate large markets and carry out a sustained and intensive interaction with the economies of different jurisdictions, without having to maintain any kind of physical presence. The problem of taxation in a digitalized economy—which is currently under discussion in the IF constituted by the OECD in accordance with the mandate of the G20 group of countries (the Pillar One discussions)—has highlighted the inadequacies of the prevailing international tax system that prohibits the taxation of business profits in a source jurisdiction in the absence of a physical presence in that jurisdiction. Many LMICs that are members of the IF consider the ongoing Pillar One discussions to be an opportunity for the global community to set right the imbalance in source and residence taxing rights. When the IF began analysing the ‘marketing intangibles approach’ put forward by the USA and the ‘user contribution’ approach advocated by the UK, the G24 group of developing nations submitted its own model, which was taken on board for discussions as the third alternate approach. The G24 model specifically suggested 7 K.

Vogel et al., Klaus Vogel on Double Taxation Conventions, 3rd ed. (New Delhi: Wolters Kluwer India, 2010), 400.

132 Akhilesh Ranjan the establishment of an alternate taxing nexus in the form of a ‘significant economic presence’ that would be based on a measure of the level of interaction that a business had with the local economy in terms of volume of sales, number of users of digital services, and other factors. Consolidated profits of the business would then be allocated to every jurisdiction having a significant economic presence, using a fractional apportionment approach that would largely be based on the quantum of sales in a jurisdiction. Profits already being declared in the hands of existing group entities would get subsumed in the amount allocated by fractional apportionment, thereby minimizing the extent of possible double taxation. Most importantly, the model was simple to administer and could be easily implemented by developing countries. However, the ‘unified approach’ that was formulated by the OECD by combining elements of the three approaches did not incorporate the alternate nexus in the form of significant economic presence. The blueprint published in October 20208 after extensive discussion proposes a change in treaties that will recognize an alternate nexus based on the quantum of sales in a source jurisdiction. But this nexus has been designed to serve the limited purpose of allowing taxation, ‘unconstrained by physical presence’, of a portion of allocable profits (Amount A) that are determined under the Pillar One approach. The blueprint makes it very clear that the new nexus will not enable the taxation of any other income in any other circumstances. Further, while Amount A is to be allocated under a formulary approach, Amount A itself will be a fraction of the residual profit of the business, arrived at in an ad-hoc manner, and not the consolidated profit. The G20 group of nations had asked the OECD and the IF to endeavour to reach a consensus solution on taxation in a digitalized economy by July 2021. Recent developments, including the publicly announced change in position of the USA, raised hopes of at least a high-level political agreement. However, even if such a solution is in fact found, it is highly unlikely that it will address, in any meaningful way, the issue of allocation of taxing rights to source jurisdictions. LMICs will continue their struggle to realize from MNEs what they consider to be legitimate tax revenues, in the face of the prevailing imbalanced international tax architecture.

8.4.2 Differing Approaches to Attribution of Profits to PEs Prior to 2010, the provisions of the OECD and UN Model Conventions in regard to the attribution of profits to a PE were largely similar, the only significant difference being the inclusion of the ‘force of attraction’ principle in the relevant article 7 of the UN Model. Both models required the profits attributable to a PE to be the profits that the PE would make if it were a separate and independent entity, but allowed the use of a 8 OECD,

Tax Challenges Arising from Digitalization— Report on Pillar One Blueprint: Inclusive Framework on BEPS (Paris: OECD Publishing, 2020).

International Tax Law and Low- and Middle-Income Countries 133 method based on fractional apportionment if it was customary for a contracting state to use such a method. However, the 2010 update to the OECD Model Convention substantially altered article 7 by deleting the reference to fractional apportionment and specifically incorporating the arm’s-length standard in the provisions, so as now to require the functions performed, the assets used, and the risks assumed by the enterprise through the PE to be the basis for computing the profits attributable to the PE. Notably, the UN did not follow the OECD on this modification and no corresponding change was made in article 7 of the UN Model Convention. This approach based on functions, assets, and risks ignores the contribution of the market jurisdiction which provides the market infrastructure, the demand, and the market-generated intangibles. The proportion of profits attributed by this approach is consequently much lower than would normally be computed under a fractional apportionment approach that is based largely on the sales made and expenses incurred in the source jurisdiction. While most LMICs would be expected to negotiate for the attribution method incorporated in the UN Model, their limited bargaining power is not always sufficient to secure a fair result. Even otherwise, the difference in approaches would lead to disputes on the amount of profits attributed, thereby increasing the burden on the available dispute-resolution systems.

8.4.3 Royalties and Fees for Technical Services Payments of royalty and of fees for technical services (FTS) are two of the major base- eroding payments in the context of LMICs. The payments normally constitute deductible expenses in the hands of resident companies or PEs of non-resident enterprises and consequently lead to a reduction in the tax base. Where the payments are made to low-tax jurisdictions, they also constitute profit shifting that has not been specifically addressed by the BEPS recommendations. While these issues are equally valid in the case of developed countries, the problem is especially serious from the perspective of LMICs as they rely heavily on technology and know-how from more developed economies. The OECD Model Convention does not grant any source- taxation rights to jurisdictions where the royalty arises, except where it is effectively connected to a PE. This treatment is based primarily on the view that royalties should be taxed on a net basis in order to account for the expenses that are normally incurred for the development, maintenance, and enhancement of the right or property for which it is paid. On the other hand, the LMICs feel that the technology or processes once developed and provided to them do not require any significant marginal costs to maintain them and, therefore, royalties paid for the use of such technology or processes should be taxed on a gross basis in the source jurisdiction. The UN Model has effected a compromise between the two views by granting taxation rights to both residence and source jurisdictions but limiting the rate of tax to a figure to be mutually decided in cases where the recipient is the beneficial owner of the royalty. While most treaties between developed and

134 Akhilesh Ranjan developing countries do allocate taxing rights to the source state in respect of royalties, the widely differing perception of the nature of royalty income and the restrictive interpretation of the term in OECD member countries is a cause of frequent double taxation and continuing litigation, thus blocking tax revenue for the source countries and burdening their dispute-resolution systems. The problem of differing interpretations of what constitute royalties is best exemplified by the recurring issue faced by LMICs of whether payments for acquiring software under an end-user licence agreement can be taxed as royalties. The Commentary on the OECD Model Convention discusses the issue at length and concludes that the use of such software amounts to use of a copyrighted article that does not involve the exploitation of any right in the copyright and hence the consideration paid is not in the nature of a royalty. However, the UN Tax Committee has recently approved the inclusion of a (minority) view in the UN Model Commentary that does not agree with such a distinction and holds that copying software on to the hard disk of a computer, even if only to be able to use it, amounts to use of a copyright right. Similarly, fees paid for managerial, technical, or consultancy services are not considered by OECD member countries as constituting income that should be taxed on a gross basis. The OECD Model Convention does not contain a separate article for such income and treats it as part of normal business income taxable on a net basis under article 7 of the Model Convention. Developing countries, on the other hand, are of the view that FTS is normally paid as consideration for the sharing of technical know-how, experience, skills, and processes and is therefore of the same nature as royalty, except that FTS does not involve the transfer of any right or property. In addition to the loss of tax base, the non-taxability of FTS in the source country (in the absence of a PE) is also considered to be discriminatory against domestic businesses rendering the same or similar services and which are taxed in the normal course on such income. Several tax treaties between developed and developing countries do include technical services in the article dealing with royalties and provide for taxation in the source country on a gross basis. However, these inclusions are normally in respect of services that are only ancillary or subsidiary to the enjoyment of the right or property for which royalties are being paid (referred to as ‘included services’). In some other cases, the relevant article covers payment of FTS only where technical knowledge or experience or know-how, skill, or process is ‘made available’ to the payer, or the relevant designs or documentation are actually transferred to the payer. Since it is often difficult to establish that the technical services are in the nature of included services or that they make the technical know-how etc. available to the payer, such articles in treaties have generally resulted in a high level of disputes and consequent litigation. Taking note of these inconsistencies and uncertainties in the tax treatment of FTS, and considering the new business models emerging that could use technology to engage in substantial business activities in a source jurisdiction without having any physical presence there, the UN Model Convention specifically incorporated a new article 12A in 2017 to provide for source taxation of FTS on the same lines as for royalties. However, while the inclusion of the article in the UN Model Convention was supported by a

International Tax Law and Low- and Middle-Income Countries 135 majority of committee members, there was a significant number that strongly opposed the same, mainly on the ground that income from rendering such services should be taxed where it is performed and not where it is consumed. Further, the article defines FTS broadly to mean payments for services of a managerial, technical, or consultancy nature (with some exceptions). This definition is considered by many to be capable of differing interpretations that can lead to disputes as well as the possibility of double taxation not being fully relieved under the prevailing treaty mechanisms. There is also the possibility of the fees being grossed up by service providers thereby increasing the costs to the consumer, and of trade distortions that would be created by the differential treatment of goods and services. For all these reasons, it is difficult to reasonably expect many LMICs to be able to incorporate article 12A in its present form in their treaties, especially those with developed countries.

8.4.4 Issues in Transfer Pricing The effective use of transfer pricing has become a major focus for tax administrations of LMICs in trying to protect their tax bases. The OECD in its report of 2014 to the G20 group of nations enumerated the following major areas of concern in this regard: (1) Excessive or unwarranted inter-group payments typically related to financing (interest payments), the use of intangibles and technology (royalties), and various types of services (technical, managerial, or shared group services). While such payments could all represent normal business outgoings, an assessment of the extent to which the payments are reasonable is required to be made for ensuring that profits arising in the source country are not suppressed. (2) The extractive industry is a very important source of revenue for LMICs that are rich in natural resources, the revenue coming from a combination of royalties and taxes. It has been found that MNEs contribute a much larger share of tax revenue in these countries, especially where the extractive industry is large.9 In this context, the pricing of mineral exports by the mining entity to its MNE affiliates overseas needs to be adequately tested to ensure that the country of resource gets a fair price for parting with its wealth. (3) The increasing use of information and communications technology and the easing of financial flows globally have resulted in several MNEs restructuring their business models and operations. There has been a mushrooming of regional and global hubs carrying out centralized functions such as planning and strategy, procurement, and market analysis and the maintenance of intellectual property for the MNE group as a whole or for regional arms of these groups. Typically,

9 IMF,

Fiscal Regimes for Extractive Industries: Design and Implementation (Washington, DC: International Monetary Fund, 2012).

136 Akhilesh Ranjan these hubs are located in low-tax jurisdictions and the operational and entrepreneurial risks are contractually transferred to them in a process of ‘supply chain restructuring’. The group entities located in developing countries that actually carry out the functions of manufacture or distribution are characterized as low- risk or captive entities that are entitled to only nominal returns on their operational costs or other similar parameters, with the major part of the group income flowing into the hubs located in low-tax jurisdictions. Developing countries have taken note of several such arrangements but are unable to marshal the resources required to analyse the substance underlying these restructurings or the extent to which they reduce the tax base and shift profits overseas. The final reports of the BEPS Project addressed these issues, though with varying degrees of comprehensiveness and specificity. The recommendation in BEPS Action 4 on limiting the deductibility of interest does not deal with the determination of arm’s- length rates of return in respect of the various types of financial transactions that are typically entered into within MNEs and also does not address issues such as corporate guarantees that are of substantial importance for developing countries. The concern that developing countries have regarding profit shifting through payment of management fees and service charges is largely about the quantum of such payments and not merely about the mark-up charged on them. Further, the guidance set out in the Action 8, 9, and 10 reports relating to intangibles, transfer of risks and capital, recharacterization of transactions, and the use of profit-splits in the context of global supply chain restructuring is not of much help to LMICs. Tax administrations in these countries are yet to attain the required level of experience and skill necessary for them to be able to put the guidance into practice. They also have to deal with considerable information asymmetry as comprehensive information about the MNE group and its dealings in different jurisdictions is hard to come by. These countries also do not have well- developed corporate databases that could be used to benchmark transactions in the course of comparability analyses. While the country-by-country reports (CbCRs) and master files mandated by BEPS Action 13 can now provide substantial relevant information, countries are still in the process of devising strategies for effective utilization of the information. Over the last five years or so, and especially in the wake of the Addis Ababa Action Agenda that highlighted the importance of international cooperation in tax and transfer pricing matters, there have been significant efforts made by different organizations towards building capacity in LMICs and imparting the necessary skills to them in the area of transfer pricing: (1) The OECD in collaboration with organizations such as the WBG, ATAF, and other regional organizations has been conducting various capacity-building and skill- development programmes in these countries, including programmes such as ‘Tax Inspectors Without Borders’ (TIWB). A recently published case study has shown how Zambia, a middle-income country rich in copper resources, was able to carry

International Tax Law and Low- and Middle-Income Countries 137 out a detailed transfer pricing analysis of the price of copper sold by a mining company to its European associated enterprise and successfully defended in its Supreme Court a US$13 million tax demand raised against the mining company.10 (2) The Platform for Collaboration on Tax has brought out comprehensive toolkits on comparability and mineral-pricing issues11 as well as on transfer pricing documentation.12 (3) The UN Tax Committee in April, 2021 approved an updated Practical Manual on Transfer Pricing13 that contains a new section on financial transactions and comprehensively revises the guidance on various issues such as the use of the profit split method. These cooperative programmes and initiatives need to be continued in a sustained manner to enable an increasing number of LMICs to develop transfer pricing skills as well as to gainfully utilize the information now available in the form of CbCRs and master files.

8.4.5 Treaty Shopping The tax legislation of several LMICs provides for the levy of withholding taxes on payments of interest, royalties, dividends, and technical or management fees to non- residents. Indeed, these withholding taxes are major instruments of domestic resource mobilization and are viewed as blunt but effective and administrable anti-BEPS measures.14 However, these countries have also entered into a large number of tax treaties in the pursuit of foreign investment and the rates of tax withholding in several treaties are much lower than the domestic law rates, sometimes even nil. This opens the way for treaty shopping and consequent loss of legitimate tax revenues for LMICs. The problem can be particularly severe in countries that depend heavily on mining and mineral resources. The extractive industry typically involves a number of significant cross-border income streams and treaty shopping deprives the source jurisdiction of collecting its fair share of taxes from the income paid out to foreign enterprises.

10 OECD, Case Study: Building Capacity to Prevent Profit Shifting by Large Companies in Zambia (Paris: OECD Publishing, 2020). 11 Platform for Collaboration on Tax, A Toolkit for Addressing Difficulties in Accessing Comparables Data for Transfer Pricing Analyses (Washington, DC: IMF, OECD, UN, WBG, 2017). 12 Platform for Collaboration on Tax, Practical Toolkit to Support the Successful Implementation by Developing Countries of Effective Transfer Pricing Documentation Requirements (Washington, DC: IMF, OECD, UN, WBG, 2020). 13 United Nations, Practical Manual on Transfer Pricing for Developing Countries, 3rd ed. (New York, UN, 2021). 14 E. Mensah, ‘Mobilizing Domestic Resources for Development & International Cooperation: Ghana’s Perspective’, G24 Technical Group Meeting, Addis Ababa (2015).

138 Akhilesh Ranjan Treaty shopping has also been used to avoid source taxation of capital gains on alienation of shares in a company or comparable interests in a partnership or trust. Many treaties entered into by LMICs follow the OECD Model in this regard, which does not allow source taxation of capital gains on transfer of movable property except where it forms part of the assets of a PE. In some cases, these capital gains are not taxed even in the state of residence and the resultant double non-taxation makes treaty shopping even more lucrative. The prevention of treaty abuse including treaty shopping is one area where the BEPS Project can be said to have achieved significant success. The MLI enables countries to swiftly revise their treaties by incorporating suitable limitation on benefits (LOB) provisions that can potentially put an end to treaty shopping. Several LMICs which are members of the IF have signed and ratified the MLI and many more are in the process of doing so. However, implementing the new LOB provisions is a challenge by itself, especially in low-income countries that have neither the administrative resources nor the skills required to do so in an effective manner.

8.4.6 Indirect Transfer of Assets The tax treatment of ‘offshore indirect transfers’ of assets—that is, the sale of shares in the company owning the assets rather than the sale of the asset itself, or the sale of shares in another company that owns the shares of the first mentioned company—has been a long-standing issue of significance for most developed countries. It is of special importance to LMICs that depend heavily on their mineral resources for raising government revenues. The appreciation in value of mining rights or mineral licences that form the core of the business of extractive industries is often appropriated by the MNE group by selling the shares of the holding company at a profit, even while the business is running, with no capital gains tax payable in the local jurisdiction because the holding company is resident in another jurisdiction or the relevant treaty provides for taxing such profits and gains in the jurisdiction of residence of the alienator. The OECD as well as the UN Model Conventions contain provisions (art. 13(4) in both models) that allow source taxation of gains from alienation of shares of a company that derive their value substantially from immovable property situated in the source jurisdiction. However, as reported by the PCT, only about 35% of the 3,000-odd tax treaties in existence actually contain a provision that is akin to article 13(4) of either models. The PCT further found and noted with concern that the likelihood of such a provision being included in a treaty is significantly lower if one of the parties is a low-income, resource- rich country.15 The absence of such a provision in a treaty entered into by a developing country can now be fulfilled through article 9(4) of the MLI, which seeks to incorporate a reinforced 15 Platform

for Collaboration on Tax, The Taxation of Offshore Indirect Transfers— A Toolkit (Washington, DC: IMF, OECD, UN, WBG, 2020).

International Tax Law and Low- and Middle-Income Countries 139 version of the provision into treaties. However, article 13(4) of the Model Conventions deals with indirect transfers of only immovable property, the term being defined in the article itself as having the meaning assigned to it under the domestic law of the contracting state where the property is located. There are several instances where the shares or other comparable interests that are transferred derive their value substantially from valuable movable assets, such as oil and gas installations and telecommunication networks, which do not fall within the definition of immovable property. In this context, the economic rationale for limiting the article 13(4) treatment to immovable property is less than clear. Further, article 13(5) of the UN Model Convention does extend the scope of source-taxation rights over gains made by the transfer of shares or comparable interests of a company or other entity in certain circumstances. However, this applies only to cases of direct ownership of the shares and only where the holding company itself is resident in the source jurisdiction. In 2020, the UN Tax Committee decided to incorporate a new paragraph 6 into article 13 of the UN Convention, which directly addresses gains from offshore indirect transfers. This new paragraph grants source-taxation rights over gains derived from the alienation of shares or comparable interests in any company or entity, if the alienator holds a substantial proportion of the shares or interests in such company or entity and such shares or interests in turn derive at least 50% of their value from property situated in the source state, subject to the condition that the gains that would have arisen had such property itself been transferred would have been taxable in the source state under the treaty. The challenge now for the LMICs is to get these provisions included in their treaties.

8.4.7 Balanced Use of Tax Incentives The pursuit of development goals in an environment of scarce financial resources has led to the proliferation of tax incentives for investment in LMICs. For instance, Keen and Mansour16 found that in 1980, less than 40% of sub-Saharan Africa offered tax holidays. By 2005, this number had increased to more than 80%; further, 50% of the countries now also offered tax exemption through free zones. Apart from the resulting loss of tax revenue, the effectiveness of incentives in garnering additional investment remains largely unknown due to the lack of relevant data and the absence of analytical tools and skills. From a policy perspective, such incentives also detract from the efficiency of a tax system, undermining its horizontal equity and opening avenues for tax avoidance and litigation. Yet, LMICs are reluctant to discontinue these incentives for a variety of reasons including political convenience, vested interests, and, above all, tax competition with peer-group countries.

16 M.

Keen and M. Mansour, ‘Revenue Mobilization in sub- Saharan Africa: Challenges from Globalization’, IMF Working Paper (2009).

140 Akhilesh Ranjan The challenge for LMICs is to introduce tax incentives in a carefully calibrated and measurable manner in order to complement the other important factors impacting investment, such as adequate infrastructure, stable macroeconomic policies, and robust labour laws. As emphasized by the IMF, World Bank, OECD, and UN in their 2015 report to the Development Working Group of the G20,17 ‘striking the right balance between an attractive tax regime for domestic and foreign investment, by using tax incentives for example, and securing the necessary revenues for public spending, is a key policy dilemma’. An additional aspect that is likely to further confound this dilemma is the prospect of a global agreement on MNEs having to pay a certain minimum tax on their incomes in each jurisdiction in which they operate. The ongoing multilateral discussion happening in the IF around the Pillar Two proposals are likely shortly to result in an implementation framework for bringing into effect the model rules already formulated, and several developing countries may then have to reappraise the effective tax rates prevailing under their incentive schemes along with the substance-based hallmarks underlying such schemes.

8.4.8 Acquisition and Effective Processing of Relevant Information The revised standards of information exchange put in place by the Global Forum inter alia prohibited jurisdictions from denying taxpayer information on the grounds of bank secrecy, something that is of utmost relevance to LMICs facing the acute problem of illicit financial flows and hidden offshore assets of their residents. The information now available through AEOI is invaluable for LMICs in the context of tracking undisclosed wealth stashed overseas and, when processed along with the data from the CbCRs mandated by BEPS Action 13, has made it possible for developing countries to make a better assessment of tax risks and consequently more efficient utilization of resources for monitoring tax compliance. However, LMICs face a number of obstacles preventing them from benefiting fully from these initiatives. First, the basic legal framework required in the form of the multilateral convention on mutual administrative assistance or bilateral tax information exchange agreements with countries from which they want information may not be in existence or such countries may not be willing to sign these agreements. Secondly, developing countries may struggle to meet the prerequisites for information exchange, which include having the statutory authority to be able to collect information from their citizens and institutions and supply the same to treaty partners and to guarantee the confidentiality of information received. Thirdly, especially in the case of AEOI, developing countries need to invest significant resources to put in place systems allowing them to 17 IMF,

OECD, UN, and WBG, ‘Options for Low Income Countries’ Effective and Efficient Use’, https://www.imf.org/external/np/g20.

International Tax Law and Low- and Middle-Income Countries 141 collect, process, and transmit information in bulk and also to make use of the information they receive. The Global Forum has played a stellar role in guiding and assisting developing countries in fulfilling all these requirements. The peer-review process and the globally significant ratings of the extent of compliance by all member jurisdictions (numbering more than 140 countries) has been very effective in persuading countries to cooperate fully. However, several LMICs are still struggling to allocate the necessary resources and to put the required systems in place for efficient and useful exchange of tax-related information.

8.5 Conclusion Despite the systemic constraints and the deficiencies in skills and resources, it must be noted that the issues and challenges in international taxation being faced by LMICs are now becoming increasingly delineated and better appreciated by the global community. The sustainable development goals sought to be achieved by LMICs and the role of tax cooperation in achieving these goals are now getting increasingly assimilated in the wider context of the environmental, social, and governance standards that countries, developed and developing, are adopting in pursuance of their respective net-zero carbon commitments. LMICs should now be able to look forward to greater international tax cooperation from the larger economies as well as enhanced cooperation amongst themselves in building capacity and augmenting domestic resource mobilization. It is critically important, however, for developed economies pursuing their laudable goals of inclusive development to realize that developing countries, particularly the LMICs, view the fairness of the international tax system from a different perspective. For them, a fair system would not merely be one that ensures a minimum tax being paid by every MNE on its global profits, but a system that ensures a fair and equitable allocation of income and profits across all jurisdictions in which the MNE operates. It would mean a system that recognizes the inadequacies of the arm’s-length principle in achieving a fair allocation of income in the digitalized economy and that is willing to complement the same with alternate and simpler methodologies such as formulary or fractional apportionment. It would imply an acceptance of the need to move forward from the limited scope of the new nexus being contemplated under the Pillar One proposal and explicitly codify an alternate and independent taxing nexus that is unconstrained by physical presence and is based on a sustained and measurable interaction with the economy. It would require MNEs to carry out enhanced reporting and make greater disclosures that can better identify tax risks, on the one hand, and enable new processes of cooperative compliance to be set in motion, on the other hand. Above all, a fair system of international taxation would mean a system that allows LMICs to pursue stable tax policies and easy-to-administer rules that can lead to the generation of sustainable revenues in the long term.

Chapter 9

Internationa l Tax L aw Status Quo, Trends, and Perspectives Reuven S. Avi-Y onah

9.1 Introduction: Why Tax Cross- Border Income? In 1992, the Stanford economist and former government official Charles McLure published a visionary paper on the future of taxation in the twenty-first century.1 McLure envisaged a future in which all governments’ revenue needs would be fulfilled by Value Added Taxes (VATs) and by other consumption-based taxes. In such a world, McLure argued, the international tax regime (ITR) could simply disappear: all taxation would be on a destination basis, and there would be no need for tax treaties or for the imposition of taxes on extraterritorial income. Moreover, since the consumer base is relatively immobile, there would be no tax competition, and each country could set its consumption tax rate independently of other countries. If the only goal of taxation was revenue raising for the provision of public goods, McLure’s vision would be appealing. But taxation has two other goals, which cannot be satisfied by taxing only consumption. Beyond revenue raising, the second goal of taxation is (re)distribution of resources from the rich to the poor, in order to achieve a more equal allocation of resources than a capitalist system typically produces if left unchecked. Consumption taxes are regressive and cannot easily achieve distributive goals. Personal income tax (PIT) is widely acknowledged as the best tax for achieving progressivity.2

1 C. E. McLure, Jr., ‘Substituting Consumption- Based Direct Taxation for Income Taxes as the International Norm’, National Tax Journal 45 (1992), 145–154. 2 R. S. Avi-Yonah, ‘The Three Goals of Taxation’, Tax Law Review 60 (2007), 1.

144 Reuven S. Avi-Yonah The third goal of taxation is regulation. Taxes are frequently used to regulate market actors, and in many cases such regulation through taxation is more effective than direct command- and-control regulation or other types of market-based incentives (e.g. carbon taxes vs. cap and trade vs. direct regulation of emissions).3 The most important regulatory tax is corporate income tax (CIT), since corporations are the most important actors in the market. In addition, corporate tax is a backstop for the progressivity of PIT.4 If one assumes that countries in the twenty-first century will continue to rely on all three types of tax (consumption tax, PIT, and CIT), then it becomes clear why cross- border income flows must be subject to tax. If foreign-source income of individuals is untaxed, then the rich could defeat the progressivity goals of PIT by moving their income offshore (since they typically have more mobile types of income than the non- rich). And if foreign-source income of corporations is untaxed, multinationals could defeat the regulatory and progressivity goals of CIT by moving their income offshore. For these reasons, McLure’s vision of a simple world with no income taxation must be rejected, and we must continue to work to improve the ITR.

9.2 The Single-Tax Principle as the Goal of the ITR The ITR is based on two principles: the benefits principle (BP) and the single-tax principle (STP). The BP was the compromise reached by four economists in 1923 between the claims of residence and source jurisdictions. It gives the primary right to tax passive (investment) income to residence jurisdictions, and the primary right to tax active (business) income to source jurisdictions. The logic of the BP is that most investment income is earned by individuals, while most business income is earned by corporations, and the residence of individuals is determinable and meaningful in ways that corporate residence is not.5 The STP can be traced back to the adoption of the foreign tax credit by the USA in 1918 and to the 1927 model tax treaty developed by the League of Nations Committee of Technical Experts.6 The STP is a work in progress: unlike the BP, it is controversial and

3 R.

S. Avi-Yonah and D. M. Uhlmann, ‘Combating Global Climate Change: Why a Carbon Tax Is a Better Response to Global Warming than Cap and Trade’, Stanford Environmental Law Journal 28/3 (2009). 4 R. S. Avi-Yonah, ‘Corporations, Society and the State: A Defense of the Corporate Tax’, Virginia Law Review 90/5 (2004), 1193. 5 R. S. Avi-Yonah, ‘The Structure of International Taxation: A Proposal for Simplification’, Texas Law Review 74 (1996), 1301. 6 R. S. Avi-Yonah, ‘International Taxation of Electronic Commerce’, Tax Law Review 52 (1997), 507; R. S. Avi-Yonah, ‘Who Invented the Single Tax Principle? An Essay on the History of US Treaty Policy’, NYLS Law Review 59 (2015), 305.

International Tax Law: Status Quo, Trends, and Perspectives 145 has not yet been fully realized, although as I will argue later that it underlies the OECD Base Erosion and Profit Shifting (BEPS) Project. The STP states that all cross-border income should be subject to the rate of tax determined by the BP, which is the residence country rate for passive income (i.e. the PIT rate) and the average large source country rate for active income (i.e. the average G20 CIT rate, currently about 25%). The STP means that if the jurisdiction allocated the primary taxing right under the BP does not impose an adequate level of tax on cross-border income, the other jurisdiction (source for passive income and residence for active income) should collect the tax up to the rate dictated by the BP. One could argue that, until the 1980s, the ITR functioned reasonably well because there was not a great deal of cross-border individual investment and therefore residence countries could collect PIT, and there was not much tax competition, so source countries could collect CIT. Both of those prongs of the ITR were undermined by globalization, the relaxation of exchange controls, and the increasing intangibility and mobility of active income, culminating in the digitalization of the global economy since 1995.7 As a result, by the time the financial crisis of 2008–2009 hit the global economy, PIT and CIT revenues were significantly undermined by tax evasion (as revealed, e.g., by the Panama and Paradise Papers) and avoidance (e.g. the trillions in low-taxed income accumulated offshore by US-based multinationals). This chapter will argue that developments in the past decade have significantly bolstered the ITR, so that it does a much better job of protecting PIT and CIT from erosion due to cross-border tax evasion and avoidance than it did prior to 2010. Specifically, the adoption by the USA of the Foreign Account Tax Compliance Act (FATCA) and the consequent development of Automatic Exchange of Information (AEOI) and the Common Reporting Standard (CRS) have significantly protected PIT, while the OECD BEPS Project has significantly improved CIT, especially if the current BEPS 2.0 effort is successfully concluded.

9.3 FATCA and the AEOI Regime Before 2010, cross-border flows of passive income were generally not subject to tax either at source or at residence. From a source tax perspective, globalization and the consequent decline in withholding taxes meant that it was possible to avoid withholding taxes not just on interest (because of unilateral abolition), royalties (because of the treaties), and capital gains (because of the source rules), but also on portfolio dividends because of the rise of derivatives, which enabled portfolio investors to receive the economic equivalent of the dividend without being subject to withholding taxes. In addition,

7

R. S. Avi-Yonah, ‘Globalization, Tax Competition, and the Fiscal Crisis of the Welfare State’, Harvard Law Review 113 (2000), 1573.

146 Reuven S. Avi-Yonah it became clear that limits on the exchange of information such as bank secrecy, dual criminality, and the requirement that information only be exchanged on request meant that in most cases residence jurisdictions could not effectively tax foreign-source portfolio income (earned primarily by the rich). In 2005, Joe Guttentag and I estimated that the USA was losing $50 billion a year to such tax evasion, and that most other countries were in worse shape because the shadow economy was larger.8 The financial crisis of 2008–2009 and the Great Recession that followed led to millions losing their jobs and their homes, and frequently their families as well. Moreover, in Europe governments reacted to the pressure on the eurozone by imposing austerity and sharply cutting the social safety net. While the Obama administration made no such cuts, and the Affordable Care Act was a meaningful move towards bolstering the safety net, the size of the US fiscal stimulus was too limited and, while the banks were saved, millions of Americans suffered a decade of low growth and unemployment.9 The political reaction on both sides of the Atlantic was dramatic. It led directly to Brexit in the UK, the election of Donald Trump in the USA and of other right-wing populists in the EU, and the prospect of serious limits to globalization in the form of immigration restrictions, tariffs, and the re-enactment of exchange controls.10 The nation state was reasserting itself, and one of the instruments it used was taxation.11 In the USA, the focus on taxation was limited to the first couple of years after the crisis, since the Republican takeover of the House in 2010 meant that no tax measures could be enacted before 2017. But, in Europe, austerity meant a continued political focus on taxing both the rich and multinational enterprises (MNEs). In the USA, the ‘Double Irish Dutch Sandwich’ was once described in detail in 2010 on the NBC Evening News, but the topic faded thereafter. In Europe, taxes became front-page matter after 2008, and this political attention is ongoing. The result has been a series of developments that led to a significant enhancement in the ability of the ITR to capture cross-border income. On the passive income front, a key development was the UBS scandal of 2006–2008, which led directly to the enactment of FATCA in 2010. The UBS hearing before the US Senate Permanent Subcommittee on Investigations revealed that UBS sent bankers directly to the USA to solicit rich individuals to set up shell companies in the Caymans and

8 R. S. Avi-Yonah and J. Guttentag, ‘Closing the International Tax Gap’, in Max B. Sawicky, ed., Bridging the Tax Gap: Addressing the Crisis in Federal Tax Administration (Washington DC: Economic Policy Institute, 2005), 99. For more recent estimates, see https://gabriel-zucman.eu/files/Zucman2013 QJE.pdf and https://www.amazon.com/Hidden-Wealth-Nations-Scourge-Havens/dp/022624542X. 9 R. S. Avi-Yonah and O. Avi-Yonah, ‘Be Careful What You Wish For: Reducing Inequality in the 21st Century’, Michigan Law Review 116 (2018), 1001. 10 See K. Clausing, Open: The Progressive Case for Free Trade, Immigration, and Global Capital (Cambridge, MA: Harvard University Asia Centre, 2019). 11 ‘The current political priorities in international taxation highlight the need for ensuring that tax is paid where profits and value are generated. It is thus imperative to restore trust in the fairness of tax systems and allow governments to effectively exercise their tax sovereignty’ (Council Directive (EU) 2016/1164 of 12 July 2016 (ATAD)).

International Tax Law: Status Quo, Trends, and Perspectives 147 then reinvest the money through UBS into the USA. UBS claimed that even though it was a ‘qualified intermediary’ (QI) and knew who the real owners of the shells were, it was justified under the QI regulations in relying on a form that stated that the owner of the income was the Caymans shell and that it was foreign.12 The result was the enactment of the FATCA, which imposes a 30% withholding tax on the US income of any foreign financial institution (FFI) that knows or has reason to know that it holds accounts of US residents or citizens and does not reveal such information to the Internal Revenue Service (IRS). Because FFIs are frequently prohibited from directly revealing financial information to the IRS, the Obama administration negotiated over one hundred intergovernmental agreements (IGAs) that enable an FFI to turn over the information to its own government, which then exchanged it with the IRS under tax treaties and tax information exchange agreements (TIEAs). Many of the IGAs are reciprocal, so that the USA is also obligated (at least on paper) to exchange information about foreign residents. The IGAs, in turn, made countries develop a CRS for the automatic exchange of financial information, and the OECD then negotiated a Multilateral Agreement on Administrative Cooperation in Tax Matters (MAATM), which relies on the CRS to provide for AEOI without the ability to rely on bank secrecy or dual-criminality provisions. Most countries in the world, and all OECD members except the USA, have ratified the MAATM.13 The result has been that it is much more difficult to evade income taxation now than it was ten years ago. A potential evader has to worry that in almost every country information about their income may be collected and transmitted to their residence jurisdiction. In addition, they have to worry that the information may either be leaked by a whistle-blower (as in the Panama Papers) or hacked (as in the Paradise Papers). I would estimate that FATCA alone has led to a significant decrease in the international tax gap in the USA, well below my $50 billion estimate from 2005. Moreover, Thomas Rixen and his colleagues have shown that the average tax rate on dividends in OECD countries was 4.5 percentage points higher in 2017 than it would have been absent CRS and MAATM.14 This suggests that cooperation in this area has achieved its desired results.

12 R. S. Avi-Yonah, ‘Testimony on Banking Secrecy Practices and Wealthy American Taxpayers’, US House Committee on Ways and Means, Subcommittee on Select Revenue Measures (31 March 2009); R. S. Avi-Yonah, ‘Testimony for Hearing on Offshore Tax Evasion’, US Senate Finance Committee (3 May 2007); R. S. Avi-Yonah, ‘Testimony for Hearing on Offshore Transactions’, US Senate Permanent Subcommittee on Investigations (1 August 2006). 13 R. S. Avi- Yonah, ‘And Yet It Moves: Taxation and Labor Mobility in the Twenty-First Century’, Tax Law Review 67 (2014), 169; R. S. Avi-Yonah, ‘IGAs vs. MAATM: Has Tax Bilateralism Outlived Its Usefulness?’, CCH Global Tax Weekly 66 (13 Feb. 2014), 11 (with G. Savir). 14 L. Ahrens et al., ‘New Room to Maneuver: National Tax Policy under the Automatic Exchange of Information’, Socio-Economic Review (2020); L. Hakelberg and T. Rixen, ‘Is Neoliberalism Still Spreading? The Impact of International Cooperation on Capital Taxation’, Review of International Political Economy (2020), https://doi.org/10.1080/09692290.2020.1752769.

148 Reuven S. Avi-Yonah

9.4 BEPS 1.0 Following FATCA, the next major development in international taxation was BEPS 1.0, from 2013 to 2015. While the USA participated in BEPS 1.0, it was not the leader, ceding this role to the EU. The main reason was that the Great Recession was more severe in the EU than in the USA and the austerity policies adopted by EU governments led to public pressure on politicians to ensure than MNEs paid adequate tax. No such public pressure developed in the USA, despite similar congressional hearings (compare, e.g., the Starbucks case in the UK to the Apple hearing in the US Senate—Starbucks was condemned for legally reducing its UK tax while Apple was celebrated for doing the same in the USA). Nevertheless, it is clear that the US position influenced the outcomes of BEPS 1.0, for example, in which actions were determined to be minimum standards (all acceptable to the USA), and in the choices permitted (e.g. between the primary purpose test and the limitation-on-benefits provision in Action 7). BEPS 1.0 was an implementation of the STP, as can be seen from the new preamble to the OECD Model Tax Convention: [State A] and [State B] . . . Intending to conclude a Convention for the elimination of double taxation with respect to taxes on income and on capital without creating opportunities for non-taxation or reduced taxation through tax evasion or avoidance.15

This language embodies the OECD and G20’s official commitment to preventing both double taxation and double non-taxation (i.e. to the STP). While there were also statements that the OECD was not against low taxation per se—just the artificial separation of profit from underlying activities—the ultimate goal was to protect the CIT of the G20. In introducing the final BEPS package on 5 October 2015, OECD Secretary General Angel Gurria stated that: Base erosion and profit shifting affects all countries, not only economically, but also as a matter of trust. BEPS is depriving countries of precious resources to jump-start growth, tackle the effects of the global economic crisis and create more and better opportunities for all. But beyond this, BEPS has been also eroding the trust of citizens in the fairness of tax systems worldwide. The measures we are presenting today represent the most fundamental changes to international tax rules in almost a century: they will put an end to double non-taxation, facilitate a better alignment of taxation with economic activity and value creation, and when fully implemented, these measures will render BEPS-inspired tax planning structures ineffective.16

15

Emphasis added. presents outputs of OECD/G20 BEPS Project for discussion at G20 Finance Ministers meeting. 16 OECD

International Tax Law: Status Quo, Trends, and Perspectives 149 While this is no doubt over-optimistic, it is clear that BEPS 1.0 was conceptually intended to implement the STP. This goal can be seen in all of the BEPS action steps:

Action 1: Addressing the Tax Challenges of the Digital Economy

This step is designed to address the ability of multinationals to avoid taxation of active income at source by selling goods and services into an economy without having a PE. In a world in which most residence jurisdictions exempt or defer taxation of active income, changing the PE physical presence standard is essential to prevent double non-taxation.

Action 2: Neutralizing the Effects of Hybrid Mismatch Arrangements

This step is obviously designed to address double non-taxation by limiting tax arbitrage transactions designed to utilize hybrid mismatches to create double non-taxation. Check the box is a target.

Action 3: Designing Effective Controlled Foreign Company Rules

This step is intended to enforce effective residence-based taxation of income that is not taxed at source by limiting the scope of exemption and deferral to income that is subject to source-based taxation. Action 4: Limiting Base Erosion Involving Interest Deductions and Other Financial Payments This step is designed to enforce source-based taxation of active income by limiting interest and related deductions that erode the corporate tax base without corresponding inclusions at residence. Action 5: Countering Harmful Tax Practices More Effectively, Taking into Account Transparency and Substance This step is intended to reinforce source-based taxation of active income by putting limits on harmful tax competition involving special regimes, such as patent boxes and cashboxes, and by requiring real investment that raises the transaction costs. Action 6: Preventing the Granting of Treaty Benefits in Inappropriate Circumstances This action adopts the US limitation-on-benefits position that treaty benefits should not result in reduction of tax at source unless there is effective taxation at residence, including a ‘primary purpose test’ that states that the purpose of treaties is to prevent both double taxation and double non-taxation.

150 Reuven S. Avi-Yonah

Action 7: Preventing the Artificial Avoidance of Permanent Establishment Status

This action reinforces source-based taxation of active income and prevents the shifting of such income into low- tax jurisdictions through commissionaire and similar arrangements.

Actions 8–10: Aligning Transfer Pricing Outcomes with Value Creation

These actions build on earlier OECD work by limiting the ability to shift income to low- tax jurisdictions by transfer pricing.

Action 11: Measuring and Monitoring BEPS

This action attempts to incentivize governments to act on BEPS by measuring its magnitude (between $100 and $240 billion each year in tax avoided).

Action 12: Mandatory Disclosure Rules

This action seeks to prevent secret rulings that enable multinationals to pay very low effective tax rate in countries that appear to have high corporate tax rates. Action 13: Guidance on Transfer Pricing Documentation and Country- by- Country Reporting This action seeks to bolster transfer pricing by requiring country-by-country reporting by multinationals, so that tax avoidance can be measured and source taxation of active income upheld.17

Action 14: Making Dispute Resolution Mechanisms More Effective

This action builds on previous OECD work on mandatory arbitration in tax treaties to prevent double taxation. It is a necessary corollary to the steps that limit double non-taxation.

Action 15: Developing a Multilateral Instrument to Modify Bilateral Tax Treaties

This action is intended to improve coordination of the previous steps and implement them through the treaty network.

17 For an excellent example of how these data can be used, see K. Clausing, ‘5 Lessons on Profit Shifting from U.S. Country-by-Country Data’, Tax Notes 169 (9 Nov. 2020), 925; https://www.taxjustice. net/reports/the-state-of-tax-justice-2020.

International Tax Law: Status Quo, Trends, and Perspectives 151 Overall BEPS 1.0, despite its limitations (e.g. the failure to advance on Action 1 and the limited nature of Actions 8–10), was a very impressive achievement in a very short period of time.18 Most importantly, while BEPS will not eliminate double non-taxation in the foreseeable future, it demonstrated significant political commitment by the G20 and OECD to the STP. In the EU, BEPS was introduced as the Anti-Tax Avoidance Directive (ATAD) as part of the Anti-Tax Avoidance Package (ATAP), which generally came into effect in January 2019 and which among other measures requires all EU members to adopt strict controlled foreign corporation (CFC) rules (e.g. generally requiring residence-based taxation if the effective tax rate of the source jurisdiction is below 50% of the tax rate in the residence jurisdiction). This measure, in addition to the enactment of BEPS Action 2,19 means that it is much harder now to shift profits artificially out of EU member states.20 Another important measure in BEPS and ATAD is the primary purpose test (PPT), which requires that all tax treaties incorporate language that the treaty will not apply to transactions if a primary purpose of the transaction was tax avoidance.21

9.5 TCJA and BEPS 1.0 Until 2017, it could be argued that the USA was a laggard in terms of combating tax avoidance, because it took the position that it was already compliant with BEPS, rejected the PPT, and did not sign the MAATM.22 But the 2017 tax reform by way of the Tax Cuts and Jobs Act (TCJA) dramatically changed that. TCJA includes three measures that significantly increase taxation of US-based as well as foreign-based MNEs. First, TCJA imposed a one-time, hefty transition tax on the $3 trillion of past, accumulated earnings of US-based MNEs (although this tax was at 8–15.5% significantly lower than the full 35% pre-TCJA tax rate). Secondly, while TCJA provided for an exemption for certain future dividends from CFCs to their US parents, this exemption is strictly limited to a deemed 10% return on tangible property, which for most US-based MNEs is close to zero (because they rely heavily on intangibles). For any amount that exceeds this deemed return, TCJA imposes a current minimum tax of

18 On the limits of BEPS 1.0, see R. S. Avi-Yonah and H. Xu, ‘Evaluating BEPS: A Reconsideration of the Benefits Principle and Proposal for UN Oversight’, Harvard Business Law Review 6 (2016), 185. 19 See Council Directive (EU) 2017/952 of 29 May 2017 (ATAD II), applying the anti-hybrid rules to third countries. 20 Ibid. 21 See R. S. Avi- Yonah, ‘BEPS, ATAP and the New Tax Dialogue: A Transatlantic Competition?’, Intertax 46 (2018), 885(with G. Mazzoni). 22 Note, however, that the new US model tax treaty (2016) includes several provisions that directly implement BEPS 1.0, such as a rule that limits the reduction of withholding taxes at source if the income qualifies for a reduced tax rate at residence. See R. S. Avi-Yonah, ‘Full Circle: The Single Tax Principle, BEPS, and the New US Model’, Global Taxation 1 (2016), 12.

152 Reuven S. Avi-Yonah 10.5% (13.125% if foreign tax credits are included) on worldwide earnings of the MNE (Global Intangible Low-Taxed Income, GILTI). Third, TCJA imposes an alternative minimum tax (Base Erosion and Anti-Abuse Tax, BEAT) of 10% on both US-based and foreign-based MNEs by disregarding interest, royalties, and some other payments from the USA to the related foreign entity.23 In addition, TCJA limits the deductibility of payments on hybrid instruments (treated as deductible in the USA and exempt in the residence jurisdiction) or by hybrid entities (treated as corporations by the USA and transparent in the residence jurisdiction, or vice versa). TCJA also disallows the new participation exemption for hybrid dividends that are treated as deductible payments at source. These provisions implement OECD BEPS Action 2 in accordance with the STP. The result of these developments (BEPS 1.0, ATAP, and TCJA) is that both US and foreign MNEs are likely to be subject to significantly higher levels of tax on cross-border active income than they were before 2010.24 To give an example. The structure used by most US-based MNEs before 2017 for their foreign operations was to have a top-level CFC in a low-tax jurisdiction, with lower-tier CFCs in high-tax jurisdiction. The parent would transfer intellectual property (IP) to the top CFC via a cost-sharing agreement, and the top CFC would in turn license the IP to the lower-tier CFCs. The key to this structure was that under the US ‘check the box’ regulation, only the top CFC would be treated as a corporation, while all the lower CFCs would be disregarded (i.e. treated as branches of the top CFC).25 As a result, while for foreign tax purposes deductible royalties from the lower CFCs to the top CFC would be effective in shifting profits to the low-tax jurisdiction of the top CFC (and not subject to withholding under treaties), for US tax purposes these royalties did not exist and so did not trigger a deemed dividend to the US parent. In addition, deductible cost-sharing payments could be made from the US parent to the top CFC. This structure no longer works, for three reasons. First, under BEPS Action 2, as implemented by the EU ATAP, the royalties from the bottom CFCs to the top CFC would not be deductible because they are to a hybrid entity. Secondly, the cost-sharing payments from the US parent to the top CFC would be subject to the BEAT minimum tax. And, finally, the top CFC as well as all the disregarded entities below it would be subject to the GILTI minimum tax (10.5% or 13.125% with foreign tax credits) on a current basis. The result is that US-based MNEs need to restructure their foreign operations and are likely to be subject to a significantly higher worldwide effective tax rate than before 2018, despite the fact that both check the box and the Inland Revenue Code, section 23

Avi-Yonah, ‘BEPS, ATAP and the New Tax Dialogue’. See K. Clausing, ‘Profit Shifting Before and After the Tax Cuts and Jobs Act’, National Tax Journal 73/4 (Dec. 2020) (‘Estimates suggest that, once adjustment to the legislation is complete, it should reduce the U.S. affiliate corporate tax base in haven countries by about 20 percent, increasing the tax base in both the United States and in higher-tax foreign countries’). But the adoption of the participation exemption tends to cut in the opposite direction, and should be abolished. 25 Under the Internal Revenue Code, s. 954(c)(6), the payments would not trigger a deemed dividend even if they were not disregarded. 24

International Tax Law: Status Quo, Trends, and Perspectives 153 954(c)(6) have not been affected by the TCJA. There are, in fact, indications that the US tech companies as well as pharmaceuticals are paying significantly higher effective tax rates than before TCJA.26 While it is difficult to prove that the TCJA was influenced by BEPS 1.0 because the Republicans who wrote the law would not publicly acknowledge such an influence, it is hard to see TCJA as anything other than a move towards implementing the STP, which also underlies BEPS 1.0. Moreover, Lilian Faulhaber, who was at the OECD during the BEPS 1.0 negotiations, has written that GILTI was explicitly influenced by Action 3 (strengthening CFC rules) while Action 2 inspired the limits on hybrid payments in the TCJA.27

9.6 BEPS 2.0 BEPS 1.0 has some acknowledged limits. Specifically, consensus was not reached about taxing the digital economy (i.e. primarily the USA and potentially the Chinese tech giants). In addition, transfer pricing was not meaningfully reformed, and the permanent establishment (PE) threshold and arm’s-length standard (ALS) remained in place despite both being obsolete in a digital economy context. In addition, relatively few of the BEPS 1.0 actions were minimum standards that had to be adopted by all participants. The political pressure to do something about BEPS in the EU and in the developing world has not lessened, as manifested by the fast rise and adoption of digital services taxes and equalizations levies designed to bypass the treaty limits on taxing the digital economy. This, in turn, has led the OECD and G20, working with an inclusive framework of over one hundred countries, to propose BEPS 2.0, which was finalized in October 2021.28 BEPS 2.0 consists of two pillars, Pillar One and Pillar Two. Pillar One is designed to address the problem of taxing corporate income at source in accordance with the BP. Pillar One allows source jurisdictions to tax a limited amount of income without regard to the PE and ALS limits, and to tax an additional amount in the market jurisdiction

26 See M. A. Sullivan, ‘TCJA Not Enough to Shift Big Pharma Profits to U.S.’, Tax Notes International 100 (23 Nov. 2020), 1034; M. A. Sullivan, ‘The Effect of the TCJA on Big Tech’, Tax Notes International 100 (2 Nov. 2020), 605; T. Horst, ‘Financial Results and Effective Tax Rates for Ten Largest Pharmaceutical MNEs and Their Implications for U.S. International Tax Reform’, Tax Notes International 100 (21 Dec. 2020). 27 L. Faulhaber, ‘Diverse Interests and International Legitimation: Public Choice Theory and the Politics of International Tax’, American Journal of International Law (2020), https://www.cambridge.org/ core/journals/american-journal-of-international-law/article/diverse-interests-and-international-legit imation-public-choice-theory-and-the-politics-of-international-tax/DD28634CB3DB07094ACDA2566 D7B3B87. 28 OECD, ‘Statement on a Two- Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy’ (8 Oct. 2021) (the ‘Statement’).

154 Reuven S. Avi-Yonah subject to the PE and ALS limits. Pillar Two then directly implements the STP by ensuring a minimal level of tax in the residence jurisdiction if the source-country tax is insufficient, and if that is not enough, by ensuring a minimal level of tax in the source jurisdiction if the residence-country tax is insufficient. These provisions are based on but represent an improvement over GILTI and the BEAT. The main conceptual innovation in Pillar One is Amount A, which is 25% of residual profit (defined as profit in excess of 10% of revenue) of in-scope MNEs (MNEs with revenues over €20 billion and a pre-tax profit margin of 10%). Amount A will be allocated to market jurisdictions with nexus (at least €1 million in revenue) using a revenue-based allocation key (i.e. a single factor sales formula).29 Amount A eliminates the two features of the ITR that have long been identified as obsolete: the requirement that an MNE have a PE in a source jurisdiction and the ALS for calculating the amount of income attributable to the PE. The PE requirement is obsolete in a world in which MNEs can earn billions in a market jurisdiction with no physical presence. The ALS is unworkable for the residual profits of MNEs (defined here as profits above 10%) because there are typically no comparables, so that a formula is the best way to allocate them. The PE requirement and the ALS were both introduced into the ITR at an early stage, primarily through the work of Mitchell Carroll in the 1930s. It is high time for both to go in a way that ensures that large MNEs like Amazon, Apple, Facebook, and Google pay tax in the source country from which they derive billions in profits. Amount A is fully consistent with the benefits principle. While it can be argued that the residence country should also get a share since typically the algorithms that underlie the business model were developed there, this is reflected by the fact that 75% of the residual profit is not taxed in the market jurisdiction and therefore should be taxed in other jurisdictions based on Pillar Two. Pillar Two fully implements the STP. The STP was already envisaged in the original League of Nations Model from 1927 but was first implemented in the 1960s and then gradually accepted (with some steps back, such as the US portfolio interest exemption in 1984 and check the box in 1997) and implemented in BEPS 1.0 (2013–2015) and TCJA (2017).30 The income inclusion rule (IIR) reflects the ability of residence countries to implement the STP by taxing their MNEs on a residence basis. Since 95% of large MNEs are resident in G20 countries, this is expected to be highly effective. The under tax payments 29 These features are similar to the proposals in Avi- Yonah, ‘International Taxation of Electronic Commerce’; R. S. Avi-Yonah, ‘Between Formulary Apportionment and the OECD Guidelines: A Proposal for Reconciliation’, World Tax Journal 2 (2010), 3; R. S. Avi-Yonah, ‘Allocating Business Profits for Tax Purposes: A Proposal to Adopt a Formulary Profit Split’, Florida Tax Review 9 (2009), 497 (with K. Clausing and M. Durst). 30 For antecedents to Pillar Two, see Avi-Yonah, ‘International Taxation of Electronic Commerce’; R. S. Avi-Yonah, ‘Who Invented the Single Tax Principle? An Essay on the History of US Treaty Policy’, NYLS Law Review 59 (2015), 305; R. S. Avi-Yonah, ‘Stanley Surrey, the 1981 US Model, and the Single Tax Principle’, Intertax 49 (2021), 729 (with G. Mazzoni).

International Tax Law: Status Quo, Trends, and Perspectives 155 rule and subject to tax rule are designed to enable residual source taxation when residence taxation is ineffective. The minimum tax rate of 15% is low but was the best that could be expected from including so many countries. The G20 can be expected to use a higher rate for the IIR, especially if the USA takes the lead in raising the GILTI rate. The substance carve-out is unfortunate (since it allows for some double non-taxation in violation of the STP) but is quite limited. Together with the CRS regime that implements the STP for individuals, Pillar Two will ensure that the STP will apply to large MNEs as well.

9.7 Conclusion: Whither the ITR? The last decade has seen significant limits to tax evasion and avoidance and an advance towards achieving the STP. These steps are crucial to achieve the goals of the ITR and to protect both PIT and CIT. But in order to prevent further political damage, more needs to be done. First, additional changes to bolster the ITR are required. Secondly, the added revenues should be used to bolster the social safety net and prevent another Great Recession. There are three additional measures that I believe would strengthen the ITR. (1) In regard to passive income, despite CRS and MAATM, I do not think the solution can depend entirely on exchange of information and residence-based taxation. There are too many residence countries to cooperate effectively, and there will always be some non-cooperative tax havens to attract evaders. But the key point is that portfolio investments are limited to a small number of large jurisdictions. If the USA, EU, and Japan could cooperate to reinstitute withholding taxes on interest, a large part of the problem could be resolved.31 (2) In regard to active income, there are a limited number of residence countries of MNEs (over ninety of the Fortune 100 are resident in the G20). If all the G20 could agree to further strengthen CFC rules to eliminate exemption or deferral, most MNE income would be taxed currently.32 In the USA, this would mean that the GILTI provision should be revised to eliminate the 10% deemed return exemption and increase the rate to 21%.33 Strict anti-inversion rules (e.g. a managed and controlled residency test) would eliminate the ability of MNEs to artificially move out of the USA. 31 R.

S. Avi-Yonah, ‘What Goes Around Comes Around: Why the USA is Responsible for Capital Flight (and What It Can Do About it)’, Haifa Law Review 13 (2019), 321. 32 R. S. Avi-Yonah, ‘Hanging Together: A Multilateral Approach to Taxing Multinationals’, in T. Pogge and K. Mehta, eds, Global Tax Fairness (Oxford: Oxford University Press, 2016), 113 33 Clausing, Profit Shifting.

156 Reuven S. Avi-Yonah (3) Since active income should be taxed at source, and since tax competition does not affect the market jurisdiction, the EU proposals for eliminating the PE standard and substituting a virtual PE threshold for ‘significant digital presence’ should be adopted.34 In addition, a formula should be used to allocate residual profits under the ALS between source jurisdictions.35 These ideas build on BEPS 2.0 but advance it further. The key issue is that the USA and other G20 countries should grant foreign tax credits to such taxes. The fact that most G20 countries have similar tax rates should make such foreign tax credits acceptable. What should be done with the added revenues? I believe the first and necessary step would be to enhance the social safety net that was deeply hurt by the Great Recession and by the Covid-19 pandemic. In the USA, this requires universal health insurance, additional investment in education, and a massive infrastructure programme. The world faces a crucial choice in the 2020s. We can either continue retreating from globalization in favour of xenophobic nationalism, tariffs, immigration restrictions, and exchange controls. That road leads ultimately to war, as it did in the 1930s. Or we can revive globalization by investing in a robust social safety net, infrastructure, education, and job creation. While more needs to be done, we have made significant progress in curbing tax competition in the last decade. The key move now is to take the added revenue and spend it wisely.

34 R.

S. Avi-Yonah, ‘The International Implications of Wayfair’, Tax Notes International 91 (9 July 2018), 161; W. Hellerstein et al., ‘Digital Taxation Lessons from Wayfair and the U.S. States’ Responses’, Tax Notes (15 Apr. 2019). 35 R. S. Avi-Yonah, ‘Formulary Apportionment—Myths and Prospects’ World Tax Journal 3 (2011), 371 (with I. Benshalom); Avi-Yonah, ‘Between Formulary Apportionment and the OECD Guidelines’; M. P. Devereux et al., ‘Residual Profit Allocation by Income’, Oxford International Tax Group, Working Paper 19/01 (Mar. 2019).

Pa rt I I

R E L AT ION SH I P BET WEEN I N T E R NAT IONA L TAX L AW A N D OT H E R L E G A L A N D S O C IA L SP H E R E S

Chapter 10

Internationa l Tax L aw and Pri vat e Internationa l L aw Polina Kouraleva-C azals

10.1 Introduction International tax law (ITL) is a multidimensional branch of law. It governs relations between states and must, in this sense, respect public international law. It aims to ensure that taxes are levied, even in situations involving a foreign element, and, in this respect, it is unquestionably a matter of tax law. But it also aims to articulate the exercise of jurisdiction by the states concerned and, in this respect, it is very akin to private international law (PIL). In Europe, at the beginning of the twentieth century, even national tax law was not yet established as a specific and discrete branch of law. It was taught either as part of ‘financial sciences’1 or as part of other branches of law.2 It was even more the case for ITL, which was only beginning to emerge as a distinct set of legal rules, and it was not obvious how to define its relationship with other branches of law. Several authors tried to do so by comparison with other branches of international law, in particular with PIL. This comparison seemed particularly justified by the existence at the time of several international treaties which included clauses governing aspects of both PIL and ITL.

1

C. de la Mardière, ‘Histoire et actualité de l’enseignement du droit fiscal en France’, Revue européenne et internationale de droit fiscal 3 (2021), 476–482. 2 Thus, at the University of Louvain, it was initially taught as part of notarial law. See E. Traversa, ‘Des idées et des hommes: un regard historique sur l’enseignement du droit fiscal à l’Université catholique de Louvain’, https://dial.uclouvain.be/pr/boreal/object/boreal%3A142763/datastream/PDF_01/view.

160 Polina Kouraleva-Cazals Thus, trade or establishment agreements provided for equal treatment of nationals and non-nationals in both civil and tax matters.3 The relationship between the two fields depended, above all, on their respective definitions and scopes. For Niboyet, PIL was neither private (because it concerned relations between states) nor international (because at the beginning of the twentieth century it was governed essentially by domestic legal provisions). Its objective was, according to the author, to resolve conflicts between legislative and executive jurisdictions of different states.4 Its subject matter thus necessarily ‘encompassed’ ITL. In contrast, authors with a more restrictive view of PIL’s objective and scope found that a distinction should be made between ITL and PIL. Indeed, they recognized the similarity of subject matter between the two areas (conflicts of jurisdiction), but they also noted the differences in the way these conflicts were resolved. The main difference was based on the public, unilateral nature of tax law, which leads, in particular, to the non- application of foreign law.5 Thus, whereas PIL distinguishes between questions of the applicable law and those of the competent court, ITL deals only with the question of the jurisdiction to tax, which implies both the relevant law and the competent court.6 The public nature of ITL seems to have had another, unexpected, impact: the primacy of substance over method. In PIL, methods are the cornerstones while connecting factors are a secondary consideration, strongly depending on the method chosen. In ITL, nexuses are the cornerstones and methods are a purely formal matter, to be resolved according to the needs of the chosen nexus criteria.7 This framework sometimes leads ITL to very innovative solutions, but these are rarely considered in an overall analysis of the method.8 The aim of this chapter will be to take a methodological look at ITL, inspired by my (admittedly limited) readings in PIL. Indeed, the main methodological issues in the respective fields are rather similar. Thus, any circumstances with a 3 M. Chrétien, ‘Le fisc en face du droit international privé’, in Travaux du Comité français de droit international privé, 14–15e année, 1951–54 (Paris: Dalloz, 1955), 85, 105; J.-C. Gautron, ‘Les conventions d’établissement conclues par le Sénégal avec des entreprises’, Annuaire français de droit international 14 (1968), 654, esp. 665; P. Hay, P. J. Borchers, and S. C. Symeonides, Conflict of Laws, 5th ed. (Boulder, CO: West, 2010), 254, § 3.58. 4 J. P. Niboyet, ‘Les doubles impositions au point de vue juridique’, RCADI 31 (1930), 44. 5 M. Chrétien, ‘Le fisc en face du droit international privé’, 91–92. For a similar conclusion, based on partly different arguments, see O. Bühler, Les accords internationaux concernant la double imposition et l’évasion fiscale, vol. I (The Hague: Académie de droit international, Recueil des cours, 1936), 55, esp. 453– 455. For an excellent summary of comments on the relationship between the two branches, see A. Kallergis, La compétence fiscale (Paris: Dalloz, 2018), 299. 6 As is the case for crimes presenting foreign elements: D. Rebut, Droit penal international, 3rd ed. (Paris: Dalloz, 2019), 23, § 28; R. A. Leflar, ‘Conflict of Laws: Choice of Law in Criminal Cases’, Case Western Reserve Law Review 25/1 (1974), 47. 7 This difference has also been underlined by Kruger: T. Kruger, ‘The Quest for Legal Certainty in International Civil Cases’, Collected Courses of The Hague Academy of International Law 380 (2016), 302–303. 8 The difference in priority setting between the two subjects probably explains why authors working in PIL are very interested in the methodological innovations used in ITL, while the reverse is not true. ITL authors are much more interested in public than in private international law.

International Tax Law and Private International Law 161 foreign element immediately raise the question of the applicable civil and tax law. The methods used in PIL have evolved over time, whereas ITL has shown great methodological stability. However, beyond the issue of methods, in both branches there is an underlying, crucial issue of the influence of the will of the parties (Section 10.2). Once applicable laws are identified, they should be articulated in a consistent way, which is not as straightforward in ITL as it is in PIL (Section 10.3). These classic methods raise, both in ITL and in PIL, overwhelming difficulties when faced with a global phenomenon, such as that of multinational enterprises. These difficulties call for new approaches and methods in both fields (Section 10.4).

10.2 The Parties’ Autonomy and the Choice of Law The purpose of PIL is to settle questions of international relations between private persons. This leads to two important considerations: equality, insofar as possible, between the parties on the one hand, and the neutrality of the regulation, on the other. In some areas (e.g. contractual obligations) parties can even choose the applicable law,9 even though this autonomy is subject to some limits. Thus, parties may not abuse their freedom of choice by circumventing a regulation that would in principle apply to them.10 The chosen law can also be set aside on public policy grounds.11 More generally, the margin of choice left to individuals remains a subject of debate in PIL. Thus, Horatia Muir Watt pleads for ‘publicisation’, and increased control of the parties’ will.12 ITL can be defined as ‘the body of legal provisions . . . that covers the tax aspects of cross border transactions’,13 or as French authors put it, ‘la projection du droit fiscal sur le plan international’.14 This definition of ITL results in the preponderance of state interests as parties involved in the situations concerned. This difference undoubtedly has an impact on the determination of connecting factors and, in particular, taxpayers are not supposed to choose the applicable tax law. However, the extent of this impact remains to be defined both in terms of the influence of states (Section 10.2.1) and that of taxpayers (Section 10.2.2). 9 e.g. according to art. 3 of Regulation (EC) No. 593/2008 of the European Parliament and of the Council of 17 June 2008 on the law applicable to contractual obligations (Rome I): ‘1. A contract shall be governed by the law chosen by the parties.’ 10 See, e.g., s. 187(2)(b) of the Restatement and more generally on the limitations on party autonomy: Hay et al., Conflict of Laws, 1088ff. 11 Rome I, art. 9. 12 H. Muir Watt, ‘Droit public et droit privé dans les rapports internationaux (vers la publicisation des conflits de loi?), le privé et le public’, Archives de Philosophie de droit 41 (1997), 207, esp. §§ 8 and 16. 13 K. Holmes, International Tax Policy and Double Tax Treaties (Amsterdaqm: IBFD, 2014), 2. 14 ‘[T]‌he projection of tax law at the international level’ (author translation): G. Gest and G. Tixier, Droit fiscal international (Paris: PUF, 1985), 14.

162 Polina Kouraleva-Cazals

10.2.1 The States Autonomy and the Choice of Connecting Factors The specific feature of ITL is that states are interested parties (i.e. not third parties) with conflicting financial interests. Indeed, several states may wish to tax the same income or property, even though the taxpayer’s overall tax burden is usually capped by tax treaties. Consequently, the right to tax of one state reduces the amount that can be collected by another state. Therefore, unlike in PIL, the definition of connecting factors is not neutral, and inevitably leads to winners and losers.15 Thus, while in theory each state is free to choose the connecting factors, in practice this choice can be subject to the pressure exerted by other states (either directly or through an intergovernmental organization). The actual extent of the state’s autonomy in this area may then depend on its political and economic power. This clash of interests may lead to tensions between states and can limit progress in the search for effective solutions to tax jurisdiction conflicts. Several avenues have been pursued to try to overcome it. The first idea is to identify and promote the common interests of all states. This has led to significant progress being made in the Base Erosion and Profit Shifting (BEPS) Project, but its scope is limited mainly to procedural aspects (including anti-abuse measures). On the other hand, the idea of ‘tax fairness’ as a common ideal, regularly put forward by the EU over the last few years,16 seems to be more of a talking point than a real avenue of common interest.17 The second idea is that of mutual respect by states for each other’s interests. It calls for a form of regulation of relations between states. Paradoxically, states become less free than in civil matters. Thus, the phenomenon of regulations constraining companies in PIL18 is seen to constrain states in ITL, such as soft law provisions aimed at tackling harmful tax competition.19 This phenomenon is also apparent in recent cases of the European Court of Justice (ECJ) that construe the selection criteria for state aid purposes based on the idea of the consistency of national legislation,20 which is seen as 15

See, e.g., S. Wilkie, ‘The “Source” of the International Tax Conundrum’, Kluwer International Tax blog, 10 February 2020, http://kluwertaxblog.com/2020/02/10/the-source-of-the-international-tax- conundrum/. 16 See, e.g., the EU blacklist for violations of ‘fair taxation’ criteria; Commission communication of 17 June 2015, ‘A Fair and Efficient Corporate Tax System in the European Union: 5 Key Areas for Action’, COM(2015)0302; Package for fair taxation of the digital economy (21 Mar. 2018); Package for fair and simple taxation (15 July 2020); European Parliament resolution of 21 January 2021 on reforming the EU list of tax havens, 2020/2863(RSP). 17 See, on the vagueness of the tax fairness concept, in particular when applied outside the scope of the anti-abuse and anti-harmful tax competition provisions, A. Pirlot, ‘The Vagueness of Tax Fairness: A Discursive Analysis of the Commission’s “Fair Tax Agenda”’, Intertax 48/4 (2020), 402–415. 18 A.-M. Slaughter, A New World Order (Princeton, NJ: Princeton University Press, 2004), 187, cited in L. d’Avout, ‘L’entreprise et les conflits internationaux des lois’, in Collected Courses of the Hague Academy of International Law, vol. 397 (Leiden: Martinus Nijhoff, 2019), 843. 19 e.g. the EU list of non-cooperative jurisdictions: Council conclusions adopted on 5 December 2017. 20 ECJ, Case C-562/19 Commission v. Poland, 16 March 2021,, para. 43.

International Tax Law and Private International Law 163 evidence of a state’s goodwill in its handling of international situations, as opposed to what Advocate General (AG) Kokott called an ‘abuse in the exercise of national fiscal sovereignty’ of states; abuse which EU law could sanction,21 just as it would sanction abuse by taxpayers.

10.2.2 The Taxpayer’s Autonomy and the Implementation of Connecting Factors The question of the role of the will of taxpayers in ITL is very ambiguous. From a formal perspective, tax treaties do not seem to leave any room for the will of taxpayers. However, various (legal) tools and, above all, the concept of legal personhood, have allowed taxpayers to organize their affairs in such a way as to be able to choose the applicable tax law. This situation, largely accepted under PIL, has been identified as ‘profit shifting’, and the OECD has sought to address it. However, even under the post- BEPS framework, the situation remains ambiguous. Indeed, even though states consider that the taxpayer should not be able to choose the applicable law (or tax treaty), connecting factors defined by states refer to the taxpayers’ will. Of course, they do not refer to the taxpayers’ will in terms of the applicable law, but rather in terms of involvement in the local economy and the intention to make profits in the state concerned. Thus, both forms of the permanent establishment (the fixed place of business as well as the non-independent agent)—the state of source nexus—correspond to situations in which the taxpayer clearly intends to conduct business in that state, even if they do not necessarily intend to pay taxes there. Similarly, the nexus for passive income implies an intent to invest. Although this reference to the intention of the taxpayer is implicit and is probably just a consequence of the nexus criterion chosen to justify the levying of taxes, nevertheless, the choices thus made result—in a manner consistent, in my view, with the logic of tax law—in giving a decisive role to the will of the taxpayer, albeit indirectly. The situation is different if a tax is levied, as for example in Italy, on any derivative financial instrument, as long as it is based on a security issued by a company resident in the member state of taxation22. The investor may therefore have to pay taxes in a state where they do not specifically intend to settle or invest (they may intend to invest in the state of issue of the derivative financial instrument and not in the state of location of the underlying assets). However, these types of nexus are very rare in international tax law; they have been sharply criticized and have

21

Ibid., Opinion of Kokott AG, para. 42. 1(491), (492), and (494) of legge no 228— Disposizioni per la formazione del bilancio annuale e pluriennale dello Stato (Legge di stabilità 2013) (Law No. 228 laying down provisions for the establishment of the annual and multiannual State budget (Stability Law 2013)) of 24 December 2012 (Ordinary Supplement to GURI No. 302 of 29 December 2012, No. 212, p. 1) (‘Law No. 228/2012’). 22 Art.

164 Polina Kouraleva-Cazals even been taken to the ECJ,23 which recognized that there was a ‘link’ with the Italian state, but refrained from ruling on the relevance of the link regarding international law principles (probably considering, as AG Hogan did, that the issue of the relevance of the nexus was outside its jurisdiction24). Thus, while PIL sometimes determines the applicable law regarding only the consequences of one’s behaviour (the impact of an anti-trust violation on a market25), ITL refers only exceptionally to the accidental, unintended presence of the taxpayer in the territory of a state.26 In this respect, ITL gives more weight to the will of the individual than does private law. This emphasis on the taxpayer’s intention could prove particularly useful to the European Union. For some years now, the EU has been asserting that the nexus is one of the components of tax fairness pursued by the Commission.27 However, the Commission does not propose any standard of relevance of a nexus. The same problem was raised in the previously mentioned cases in which the Court of Justice refrained from ruling on the conformity with international law of the nexus chosen by the national legislating body referring to the origin of the underlying financial securities’ assets.28 In an opinion concerning a different case, AG Kokott has, however, considered the issue and decided that language was a sufficient nexus to justify the state’s competence to tax, without giving a clear criterion of a ‘sufficient territorial link’.29 The reference to the taxpayer’s intention to make profits in the state concerned would thus make it possible to avoid a nexus that is too random and unpredictable for the taxpayer. The determination of the applicable law, however, does not prevent interference with foreign laws.

23

ECJ, Case C-565/18 Société Générale SA, 30 April 2020,. Case C-565/18 Société Générale SA, Opinion of Hogan AG, para. 43. 25 ‘The effect doctrine’ first developed in USA: US Court of Appeal for the Second Circuit, 12 March 1945, United States v. Aluminum Co. of America, 148 F. 2d 416 (2d Cir. 1945) and was then accepted by the ECJ: Case C-413/14 P Intel v. Commission, 6 September 2017. See also B. Zelger, ‘EU Competition Law and Extraterritorial Jurisdiction—A Critical Analysis of the ECJ’s Judgment in Intel’, European Competition Journal (2020), 613–627, available at https://www.tandfonline.com/doi/pdf/10.1080/17441 056.2020.1840844. 26 Amount A of Pillar One may be seen as an example of an allocation of profits that takes less account of the real involvement of the company in the local economy. However, the revenue thresholds combined with the multinational nature of the enterprises concerned will probably result in the allocation of tax jurisdiction to the states where the taxpayer intended to conduct sales (or services). 27 Communication from the Commission to the European Parliament and the Council, ‘A fair and efficient tax system in the EU for the digital single market’ (21 Sept. 2017), COM(2017) 547 final, p. 7. 28 ECJ, Case C-565/18 Société Générale, 30 April 2020. The judicial position is clearly explained in the Opinion of Hogan AG, 28 November 2019, paras 40–46. 29 ‘The connection to the use of the country’s own official language provides a sufficient reasonable territorial link in principle. No one can deny that language forms an important part of national identity and is thus strongly connected with a State and its territory’ (Case C-482/18 Google Ireland Ltd, 12 September 2019, Opinion of Kokott AG, para. 50). 24

International Tax Law and Private International Law 165

10.3 Articulating Domestic Law and Foreign Law Most commonly, income or assets received in a cross-border context are subject not just to the civil and tax law of one state, but to the civil law of one state and the tax law of another state. Sometimes, they can even be subject to the civil law of several states (the state of the nationality and the state of residence of the deceased) and the tax law of several states (the state of tax residence of the deceased and that of location of inherited assets). It raises the problem of the articulation of national law and foreign law. The first possible solution is to avoid as far as possible the simultaneous application of domestic tax law and foreign civil law. Such a result can be achieved by aligning the scope of application of domestic tax law with the scope of application of domestic civil law. In France, during the first half of the twentieth century, the territorial scope of inheritance tax on stocks was delimited by reference to the territorial scope of French inheritance law, so that only stocks subject to French inheritance law were also subject to French inheritance taxes.30 Indeed, because the basis of assessment for inheritance tax is the estate, it was easier to calculate French inheritance tax on the estate as determined according to French civil law.31 This method had several weaknesses. On the one hand, the effect of the method was limited, as there could always be incidental issues that could still be subject to foreign civil law (e.g. the marital status of the surviving spouse). On the other hand, its restrictive effect on the territorial scope of the tax was indisputable. Therefore, the approach was abandoned, and the determination of the applicable tax law in France has become fully independent of the territorial scope of civil law.32 Other countries, such as the UK, have retained in their tax legislation some references to concepts from PIL, such as domicile,33 which are used to determine the applicable civil and tax laws. However, such examples are rare and their impact has been decreasing over time.34 Cross-border income can also be subject to several tax laws. The distributive provisions of tax treaties appear to be perfect examples of bilateral conflict-of-laws rules. They determine the applicable tax laws depending on legal categories (‘dividends’) 30 Art.

752 of the French General Tax Code, according to which shares were subject to French inheritance tax when they were part of an inheritance subject to French civil law. 31 C. Freyria, ‘L’imposition successorale des valeurs mobilières étrangères’, La Semaine Juridique, edition générale (1953), 1111. The author considered that ‘taxation must, in principle, respect the conflict of laws rule of private international law’. 32 Since then, French tax law can apply to an inheritance estate subject to foreign inheritance law. Such situations can bring up issues of articulation of French tax law and foreign civil law. L. Galliez, ‘Successions internationales: entre unité civile et morcellement fiscal’, La Semaine Juridique, edition notariale et immobilière 47 (22 Nov. 2013), 1271. 33 J. Schwarz, Booth and Schwarz: Residence, Domicile and UK Taxation, 19th ed. (London: Bloomsbury, 2017), 10. 34 The Finance Act 2008 has limited the impact of the civil domicile on tax residence, but has not completely abolished it (ibid.).

166 Polina Kouraleva-Cazals following a bilateral logic (‘is taxable in the state of residence’). However, they rarely rule out the simultaneous application of multiple tax laws and therefore do not, in the strict sense, lead to the choice of an applicable law. The conflict of jurisdiction is not resolved, several tax regimes continue to apply, but the result of this simultaneous application—double taxation—is neutralized by a relief granted in the state of residence of the taxpayer. In PIL, nowadays, the simultaneous application of more than one domestic legislation is also beginning to be accepted. Indeed, many authors have stressed that applying a single national law to an international situation would be tantamount to denying the situation’s specific features stemming from its foreign elements.35 Once the simultaneous application of several laws is established, the issue of their articulation arises, which has traditionally been presented as one of the main differences between PIL and ITL. Indeed, the distinction made under PIL between the court’s jurisdiction and the applicable law introduces the possibility of a court in one state applying the law of another state. Such a situation is unthinkable under ITL. Moreover, tax matters are cited, alongside criminal matters, as the main exception to the application of foreign law by the courts.36 This ‘revenue rule’ is closely linked to the territorial nature of state sovereignty. Indeed, as the Permanent Court of International Justice stated in the Lotus judgment: ‘the first and foremost restriction imposed by international law upon a State is that . . . it may not exercise its power in any form in the territory of another State’.37 Thus, interventions by foreign authorities on a given state’s territory are prohibited.38 However, the scope of the ‘revenue rule’ mainly refers to the refusal to enforce foreign tax legislation and help collect foreign taxes.39 Still, even outside the tax collection context, tax authorities remained very cautious about foreign tax law. Thus, in the 1993 Exxon case, the Internal Revenue Service (IRS) argued that allowing foreign law to be taken into account while applying US transfer pricing rules would be tantamount to allowing foreign law to dictate US tax law and policy. This argument was dismissed by the Tax Court, stating that referring to foreign law can be a ‘means of implementing 35 M.- P.

Weller, ‘Vom Staat zum Menschen: die Moethodentrias des internationalen Privatrechts unserer Zeit’, RabelsZ 81 (2017), 772 cited in L. d’Avout, ‘L’entreprise et les conflits internationaux des lois’, in Collected Courses of the Hague Academy of International Law, vol. 397 (Leiden: Martinus Nijhoff, 2019), 174–175. 36 I. Brown, Conflict of Laws (London: Old Bailey Press, 2001), 18; Hay et al., Conflict of Laws, 171. 37 Permanent Court of International Justice, SS ‘Lotus’ (France v. Turkey), 1927 PCIJ, Ser. A, No. 10, at 18. 38 The French Constitutional Council has found, e.g., that the power conferred on the Prosecutor of the International Criminal Court to take certain investigative steps in the territory of a state, without the presence of the authorities of that state, violated ‘the essential conditions for the exercise of national sovereignty’, even though the measures were to be carried out without coercive force. See the decision of 22 January 1999, no. 98-408 DC on the Treaty laying down the Statute of the International Criminal Court, para. 38, https://www.conseil-constitutionnel.fr/en/decision/1999/98408DC.htm. 39 See, e.g., US case law: Attorney General of Canada v. R. J. Reynolds Tobacco Holdings Inc., 268 F. 3d 103 (2d Cir. 2001), cert. denied 537 US 1000, 123 S. Ct. 513, 154 L. Ed 2d. 392 (2002) and UK case law: Government of India v. Taylor [1955] AC 491. See also B. Mallinak, ‘The Revenue Rule: A Common Law Doctrine for the Twenty First Century’, Duke Journal of Comparative & International Law 16 (2006), 79–124, https://scholarship.law.duke.edu/djcil/vol16/iss1/3.

International Tax Law and Private International Law 167 U.S. law’.40 Indeed, over time, governments have accepted their mutual dependency for international tax purposes. The question of articulation of domestic and foreign law has become one of the cornerstones of ITL. It is better dealt with by the legislature (Section 10.3.1), but can also be regulated by the judiciary (Section 10.3.2).

10.3.1 Articulation of Domestic Law and Foreign Law by the Legislature In PIL, as in ITL, the legal ground for the application or consideration of foreign law is of crucial importance. First, PIL distinguishes between that foreign law which directly claims enforcement in another territory (which is refused) and the foreign law whose application is designated by the national law of the jurisdiction concerned (which is then accepted).41 Similarly, even the ‘revenue rule’ can be set aside by tax treaties or EU directives requiring mutual assistance for the purposes of taxation. Secondly, if there is a domestic legal ground, then foreign law is a mere factual element, which makes its application much easier both from a theoretical and from a practical perspective. Thus, many domestic tax provisions (e.g. ‘controlled foreign company’ rules or provisions against hybrid instruments) expressly refer to the foreign law’s effects on the taxpayer. The articulation of domestic and foreign tax law is then made directly by the legislative body in the domestic legal provisions. If not, foreign law can still be taken into account while domestic provisions are implemented by tax authorities or the judiciary.

10.3.2 Articulation of Domestic Law and Foreign Law by the Judiciary In situations which fall under the civil and fiscal legislation of several states, the articulation of domestic law and foreign law by judges is both crucial and delicate. Judges seem to take different approaches, some of which are inspired by PIL and its distinction between questions of law and questions of fact (Section 10.3.2.1), whose transposition into ITL, however, raises difficulties. Therefore, judges in tax matters prefer to distinguish between foreign civil law and foreign fiscal law (Section 10.3.2.2).

10.3.2.1 Foreign law status: a matter of law or a matter of fact? For PIL purposes, when judges apply a legal provision, they distinguish between legal and factual standards required by the provision. The former constitutes an incidental 40 Exxon Corp., 66 TC Memo (CCH) at 1733, TC Memo (RIA) at 93-3261 cited in P. R. West, ‘Foreign Law in U.S. International Taxation: The Search for Standards’, Florida Tax Review 3/4 (1996), 169. 41 Brown, Conflict of Laws, 2.

168 Polina Kouraleva-Cazals question (i.e. a legal question) that can be subject to foreign law; in such a situation, judges usually have to ascertain the content of the applicable law.42 The factual standard, on the other hand, requires an assessment of a factual situation, which can be affected by foreign law.43 When a foreign law is a matter of fact, it is up to the party invoking it to prove it.44 ITL uses the same method and the same distinction (at least implicitly) but with some difficulties. Indeed, sometimes foreign law is clearly taken into account as a matter of fact. Thus, under French law, companies are set up to make profits. When they renounce them without reason, this constitutes an ‘abnormal act of management’ and justifies including such theoretical, expected profits in the company’s taxable income. When applying this French legal concept, judges agree to take into account the foreign law which imposes such a renunciation. Indeed, in these circumstances, the renunciation of profits cannot be considered ‘abnormal’.45 Similarly, when applying the US transfer pricing rule, judges consider that it is subject to an implicit condition that the price difference results from opportunities offered by the multinational establishment of an enterprise. Thus, faced with very low intra-group prices imposed by foreign law, US judges have considered that this condition has not been met and the taxable base should not be adjusted.46 In this case, the constraint resulting from the foreign law has been taken into account as a factual element leading to the conclusion that the legal standard of the US legislation has not been met. The status of foreign law is less clear in other cases. When tax treaties define a permanent establishment as a person who ‘has, and habitually exercises, . . . an authority to conclude contracts in the name of the enterprise’,47 it is not clear whether this ‘authority to conclude contracts’ is a legal concept or a matter of fact. French judges first ruled that one can commit in law and in fact, so that it can be a matter both of law and fact.48 42 In the application of the iura novit curia principle. See C. Esplugues and G. Palao, ‘Foreign Law, Application and Ascertainment’, in J. Basedow et al., eds, Encyclopedia of Private International Law, vol. 1 (Northampton, MA: Edward Elgar, 2017), 769 and subs. 43 E. Fohrer-Dedeurwaerder, ‘Le juge fiscal et les questions préalables entachées d’extranéité’, Revue de droit fiscal 28 (Dec. 2016), 604. 44 Esplugues and Palao, ‘Foreign Law, Application and Ascertainment’, 769 and subs. 45 CE (French Supreme Administrative Court), 7 September 2009, no. 303560, SNC Immobilière GSE. 46 Procter & Gamble Co. v. Commissioner, 95 TC 323, 325 (1990), aff ’d, 961 F. 2d 1255 (6th Cir. 1992) and Exxon Corp. v. Commissioner, 66 TC Memo (CCH) 1707, TC Memo (RIA) 93,616 (1993) cited in West, ‘Foreign Law in U.S. International Taxation’, 167–169. 47 art. 5(5) of the OECD Model Tax Convention 2010. The old version of the Model Tax Convention is quoted here as tax treaties considered by French judges in the following examples used the same wording. 48 CE, 20 June 2003, no. 224407, Sté Interhome in which AG Government Commissioner, Mr Austry, referred to the Commentaries on the 1994 OECD Model Convention, art. 5 at paras 32 and 33. Therefore, even if judges did not make clear which facts could amount to ‘an authority to conclude contracts’, it seems that, following the OECD Commentaries, it covers decisive participation in the conclusion of a contract that is only formally concluded by the main taxpayer (e.g. by means of an electronic signature). It is not clear whether the change of wording in art. 5(5) of the OECD Model Tax Convention in 2017 (now expressly referring to a person who ‘habitually plays the principal role leading to the conclusion

International Tax Law and Private International Law 169 However, some years later, judges seem to have renounced the possibility of considering the ‘authority to conclude contracts’ as a factual standard,49 contrary to the requirement according to which the authority should be ‘habitually’ exercised.50 Recently, French judges have returned to their original interpretation. While it does not expressly state that the authority to conclude contracts is a matter of fact, it takes into account the importance of the agent’s role in the conclusion of the contract and not only their legal capacity to do so.51 Even the marital status of spouses can be considered as a matter of law52 or as a matter of fact.53 The distinction between incidental questions and the factual elements seems too uncertain to ensure consistent implementation. It also seems rather pointless, since in most cases, even when dealing with incidental questions, foreign law (and in particular foreign tax law) is treated as a matter of fact.54 However, if judges consider a legal standard as an incidental question, what type of foreign law should it be subject to?

10.3.2.2 To what type of foreign law can an incidental question be subject? In ITL, there are frequently incidental questions subject to foreign civil law. The implementation of French tax law can require judges to verify how the factual situation is characterized under foreign civil law or whether a legal situation is validly formed under foreign civil law. For example, French judges have ruled on the tax consequences in France of a polygamous marriage between two Algerian nationals, as long as the marriage had been validly contracted under Algerian law.55 Similarly, Swiss judges also refer to the laws of the state of incorporation in order to determine whether a foreign entity has legal personality.56 However, when the incidental question concerns

of contracts that are routinely concluded without material modification by the enterprise’) will have an impact on the judicial approach of this issue as a matter of law or a matter of fact. 49

CE, 31 March 2010, nos 304715 and 308525, Sté Zimmer, see AG Government Commissioner Julie Burguburu on this issue: Revue Droit Fiscal no. 16, 2010, comm. 289, para 7. 50 CE, subss. 8e and 3e, 1 June 2005, no. 259617, Nouvelle-Calédonie v. Sté Allianz-Vie and no. 259618, Nouvelle-Calédonie v. SA Eagle Star Vie: Dr. fisc. 2006, no. 7, comm. 177, concl. P. Collin. 51 CE, 11 December 2020, no. 420174, Valueclick International Ltd. 52 Fohrer-Dedeurwaerder, ‘Le juge fiscal et les questions préalables entachées d’extranéité’, § 20. 53 J. Robbe, ‘Comment déterminer si des époux dont le mariage relève du droit italien sont séparés de biens au regard de la loi fiscale française?’, Note on the decision of the Administrative Court of Paris, 3 May 2016, no. 1504705, Canti, DF 2016, no. 23, comm. 362, no. 5. 54 Indeed, only a few countries, e.g. Italy, France, and Germany, have accepted the possibility of considering foreign law as an applicable law. In Italy, as in Germany, such equal treatment of foreign law is required by domestic law (Italy: arts 14 and 15 of Law 218/1995; Germany: art. 293 of the Code of Civil Procedure). In France, there is no such rule in the written law. Judicial opinion has changed over time, as initially foreign law was necessarily treated as a fact (Cass., 25 May 1948, Lautour). However, since 2005, foreign law has been treated as the applicable law (Cass., 28 June 2005). 55 Administrative Court of Appeal of Paris, 15 January 2015, no. 12PA03956, Mr and Mrs Zabour. 56 G. Lideikyte Huber, ‘Conceptual Problems of the Corporate Tax’, in Swiss–US Comparative Analysis (Amsterdam: IBFD, 2019), 48.

170 Polina Kouraleva-Cazals characterization under the foreign tax law, judges refuse to refer to foreign law, as that would amount to ‘accepting the application of foreign tax law’.57 Sometimes, however, ignoring foreign tax law and referring only to foreign civil law is detrimental to the coherence and effectiveness of domestic tax law. This is the case for the tax treatment of foreign entities. Some countries (e.g. the USA) introduce specific rules for foreign entities but most countries apply domestic general tax provisions to them. To do so, they have to classify foreign entities according to domestic corporate or tax categories. Under the first approach, the classification is made by reference to the foreign entity’s characteristics under foreign corporate law. Under the second approach, the classification is made by reference to the foreign entity’s tax regime under foreign tax law. Most states opt for the former, assimilating the foreign entity to a local entity (depending on its characteristics under foreign corporate law) and then applying the local corresponding tax regime.58 However, in such cases, there might be inconsistency between the tax treatment of an entity in the two states. This leads to the well-known problem of hybrid entities.59 Thus, the Netherlands has recently taken steps towards the classification of foreign entities based on their foreign tax regime.60 The impact of this reform on the legal security of taxpayers and the overall consistency of their tax treatment will be interesting to monitor. These difficulties of articulation raised by using traditional methods become particularly important when faced with global economic phenomena.

10.4 Legal Regulation of Global Economic Phenomena ITL, like PIL, is faced with the same global economic phenomena. The most striking example is that of multinational companies. ‘The private company confronted with the legislation of States; the global company confronted with local laws: this is the

57 E. Bokdam- Tognetti (‘rapporteur public’), opinion under Conseil d’État (French Supreme Administrative Court), 13 April 2018, no. 398271, Min. v. LVMH Moët Hennessy Louis Vuitton. 58 CE, 24 November 2014, no. 363556, Artémis; Lideikyte Huber, ‘Conceptual Problems of the Corporate Tax’, 51 (‘principle of resemblance’). On the comparison between the ‘resemblance test’ also chosen by Luxembourg and a general tax liability applied to all foreign entities in Italy, see J. Tschhurtschenthaler, ‘Théorie fiscale des entités dépourvues de la personnalité morale’, PhD thesis, Paris I Panthéon-Sorbonne University (2020), § 253. 59 See ‘Triangular Cases and Tax Treaties’ in this volume. 60 According to a recent legislative proposal, foreign entities for which there is no clear Dutch equivalent and which are not Dutch tax residents, will be treated in the Netherlands following their classification in their home state. This reform, which was initially supposed to take effect on 1 January 2022, should finally be enacted in 2023, with an effective date of 1 January 2024 (Dutch State Secretary of Finance, Dutch Tax Policy and Priorities, 3 June 2022, see https://www.stibbe.com/en/news/2022/june/ update-on-dutch-rules-on-the-classifi cation-of-dutch-and-foreign-legal-entities).

International Tax Law and Private International Law 171 tension that gives substance to the study; this tension is so strong that it perhaps calls for an overcoming in terms of cooperation between States and the emergence of authentically global institutions.’ This quotation is not from an introduction to the latest OECD tax project, but an introductory chapter to a book on PIL.61 In PIL, as in ITL, the difficulties encountered in determining and articulating laws covering these global phenomena have led to the ability to overcome conventional methods, which were aimed at determining applicable laws. Also in PIL, as in ITL, the trend is to move away from a state logic to an interstate logic, first by the development of worldwide unified substantive rules (Section 10.4.1) and, secondly, by the development of interstate dispute- resolution solutions (Section 10.4.2).

10.4.1 The Development of Multilateral Unified Substantive Rules The harmonization of substantive rules has many advantages. In particular, the process makes it possible to reduce or even eliminate the risks of ‘forum shopping’. Thus, in PIL, international conventions have been negotiated that provide for uniform substantive rules in certain areas, designed to avoid any conflict of laws. This is the case with the Vienna Convention of 11 April 1980 on contracts for the international sale of goods, as well as other conventions that replace domestic substantive rules on certain specific issues.62 In tax law, there is currently no multilateral treaty harmonizing substantive tax rules. But the OECD has recognized that any project aiming to eliminate the remaining profit-shifting problems should involve some form of multilateral substantive rules. Thus, the implementation of Pillar One will require a multilateral convention with a set of unified substantive norms.63 Pillar Two also includes uniform substantive rules even though, in formal terms, those rules will be incorporated into national legislation.64 However, in both PIL and ITL, it must be recognized that as long as international rules, even if they are substantive and unified, require the implementation of

61 L.

d’Avout, ‘L’entreprise et les conflits internationaux des lois’, in Collected Courses of the Hague Academy of International Law, vol. 397 (Leiden: Martinus Nijhoff, 2019), 28. 62 For more recent examples, see Convention for the Unification of Certain Rules for International Carriage by Air (the Montreal Convention), 28 May 1999; UNIDROIT Convention on international financial leasing, Ottawa, 28 May 1988. 63 OECD, Progress Report on Amount A of Pillar One, Two- Pillar Solution to the Tax Challenges of the Digitalisation of the Economy, OECD/G20 Base Erosion and Profit Shifting Project (Paris: OECD Publishing, 2022), https://www.oecd.org/tax/beps/progress-report-on-amount-a-of-pillar-one-july- 2022.pdf. 64 OECD, Tax challenges Arising from the Digitalisation of the Economy—Global Anti-Base Erosion Model Rules (Pillar Two), Inclusive Framework on BEPS (Paris: OECD Publishing, 2021), https://www. oecd.org/tax/beps/tax-challenges-arising-from-the-digitalisation-of-the-economy-global-anti-base- erosion-model-rules-pillar-two.htm.

172 Polina Kouraleva-Cazals qualification processes there is a risk of qualification conflicts.65 These may jeopardize the effectiveness of the system. Some authors argue for the use of the notion of autonomous concepts, both in PIL and ITL.66 However, considering the limited subject matter of tax treaties, autonomous concepts cannot completely replace all references to domestic law.67 Furthermore, as long as these autonomous concepts are used by domestic judges, they will give rise to diverging domestic interpretation unless the judges agree to follow the interpretation of the term used in previous case law by a foreign colleague. Indeed, sometimes judges do decide to consider the position of their foreign colleagues.68 They may even align themselves with it, as long as the foreign legal reasoning is convincing69 and if no domestic tax policy reasons contradict that reasoning.70 However, in many countries, as for instance in France, judges refuse to take foreign case law into consideration.71 In theory, judges’ resistance can be overcome by treaty provisions. Thus, under article 6(2) of the OECD Model Tax Convention, real estate income is qualified according to the domestic law of the source state, which implies that the judge of the state of residence refers to the qualification adopted by the judge of the source state. The state of residence should therefore accept this foreign qualification by virtue of the treaty and not by virtue of the recognition of foreign law.72 In practice, however, such a method implies a form of hierarchy between domestic qualifications which sovereign states have difficulty in accepting. France has thus expressed a reservation 65 For more detail on this issue for PIL treaties, see V. Espinassous, L’uniformisation du droit substantiel et le conflit de lois (Paris: LGDJ, 2010). According to the author, treaties with unified substantive rules do not supplement conflict-of-laws methods, which should still be used to determine the domestic law under which the treaty should be interpreted. 66 Recently, these questions have given rise to a very rich doctrinal debate on the application of art. 3(2) of the OECD Model Convention: see J. Avery Jones, ‘A Fresh Look at Article 3(2) of the OECD Model’, Bulletin for International Taxation 74/11 (2020); M. Lang, ‘Tax Treaty Interpretation: A Response to John Avery Jones’, Bulletin for International Taxation 74/11 (2020); J. Schwarz, ‘Article 3(2) of the OECD and UN Models: An International View’, Bulletin for International Taxation 75/1 (2021). 67 See, e.g., on the difficulty of determining an internationally acceptable concept of a fiscal attribution of income: J. Wheeler, ‘The Attribution of Income in the Netherlands and the United Kingdom’, World Tax Journal 3/1 (2011), 39–175. 68 J. Schwarz, Schwarz on Tax Treaties, 5th ed. (Alphen aan den Rijn: Wolters Kluwer, 2018), 141 69 Vogel and Prokisch, Interpretation of Double Taxation Conventions, 62. the authors give several examples of cases from various countries where judges have adopted the foreign qualification, although in each case the judges reserved the possibility of not following their foreign colleague’s qualification, if they found their reasoning unconvincing. 70 M. Lang, ‘Commentaries under Article 3 § 2 of the OECD Model Convention’, in R. Danon et al., eds, Modèle de convention fiscale OCDE concernant le revenue et la fortune (Basle: Helbing Lichtenhahn, Lefebvre, 2014), § 77. Lang gives several reasons in favour of the mutual independence of judges for the purpose of qualification. These reasons mainly pertain to tax policy and should indeed be taken into account, but do not necessarily justify a judicial position which would systematically ignore the qualification designated by foreign judges or foreign tax authorities. 71 See, e.g., CE, 13 April 2018, LVMH Moët Hennessy Louis Vuitton. 72 Similar observations can be made about the application of tax treaties to partnerships and other atypical entities: OCDE, The Application of the OECD Model Tax Convention to Partnerships, Issues in International Taxation, vol. 6 (Paris: OCDE Publishing, 1999); art. 1(2) of the 2017 OECD Model Convention.

International Tax Law and Private International Law 173 under the Commentaries on article 6 aimed at retaining the possibility of applying the provisions of French domestic law pertaining to the taxation of income from shares or in rem rights, which are treated therein as income from immovable property. Furthermore, it does not resolve all the problems related to conflict of qualification, as qualification is sometimes a preliminary step in the choice of the applicable distributive provision. Thus, when a real property is held by a company, the qualification of the company’s shares is a prerequisite for the choice of the applicable article. Traditional judicial remedies do not seem to resolve these problems and the search for effective results leads to the development of alternative solutions to resolve interstate disputes.

10.4.2 The Development of Solutions Aimed to Resolve Interstate Disputes Interstate disputes are widespread in ITL. The difficulty lies in the fact that taxation is the very essence of state sovereignty, while the resolution of interstate conflicts necessarily calls for interstate solutions, which may limit the freedom and independence of states in that area. Thus, tax treaties traditionally provided for a mutual agreement procedure, conducted entirely by states and with no obligation to actually resolve the conflict. In relation to PIL, disputes in this area are mostly between private parties and not between states, therefore most disputes are effectively resolved by the domestic courts. However, faced with global phenomena, such as multinational enterprises, PIL highlights the need for interstate cooperation and for taking public interest considerations into account in private disputes. PIL is thus also opening up to alternative methods of settling conflicts of jurisdiction. These alternative methods can be divided into two groups: those that keep the procedure in the states’ hands but make the achievement of a consensus mandatory, and those that entrust the resolution of the conflict to a third party. Among the solutions for the first group, it is worth mentioning the OECD proposal made under Pillar One which provides for the establishment of a panel of tax experts from states that have MNEs, with the aim of agreeing on a harmonized tax treatment for each MNE concerned.73 As far as is known by this author, there is no similar solution in PIL, although d’Avout argues for the application of a ‘refrain or associate’ principle, according to which a state should either refrain from interfering or do its best to involve other the states concerned.74

73 OECD,

Tax Challenges Arising from Digitalisation—Report on Pillar One Blueprint: Inclusive Framework on BEPS, OECD/G20 Base Erosion and Profit Shifting Project (Paris: OECD Publishing, (2020), 712ff, https://doi.org/10.1787/beba0634-en. 74 L. d’Avout, ‘L’extraterritorialité du droit dans les relations d’affaires’, La Semaine Juridique (12 Oct. 2015), 1112.

174 Polina Kouraleva-Cazals As for the second group, in ITL this mainly covers arbitration which has recently been promoted by the OECD. However, in most cases, the arbitration solution is not mandatory for states as such (with the notable exception of the tax treaty between Germany and Austria which provides for arbitration before the ECJ75). States still have the possibility to find a different solution to their dispute. Only if they are not able to do so, are they required to implement a decision by arbitration. Yet the implementation of such a solution raises other complications76 which, interestingly, seems to be a source of inspiration for PIL.77 Similarly, Louis d’Avout suggests the establishment of a ‘global legislative extraordinary proceeding’78 introducing an internationally representative body of a permanent nature and entrusting it with the power to resolve conflicts of state competence and, additionally, to formulate new rules of international law. Such a body is reminiscent of the standing committee for the settlement of disputes whose constitution is authorized by the directive of 10 October 2017 (with the difference that the standing committee is not supposed to formulate any new rules),79 and whose creation was put back on the agenda by the Commission’s latest tax package.80 The composition of these bodies and the legal status of the decisions thereby rendered approach the limits of law as a means of resolving conflicts of jurisdiction. At the beginning of the twenty-first century, as was the case at the beginning of the twentieth century, PIL and ITL operate in different contexts and may pursue different objectives. Nevertheless, the various processes involved have gradually led to a certain convergence in the means of approaching international situations in both areas. Thus, even if the direct transposition of the solutions adopted by the one into the fields of the other does not seem desirable, the two branches can be mutually inspiring, offering new perspectives on old difficulties. In particular, thinking about the current challenges of ITL from a methodological perspective—so dear to PIL—leads to a reflection on consistency in ITL. Having steadily come to the fore of judicial review in recent years (in a purely domestic context, due to the importance of the goals and objectives in judicial interpretation, as well as

75

Art. 25 § 5 of the Convention between the Republic of Austria and the Federal Republic of Germany for the avoidance of double taxation with respect to taxes on income and capital) of 24 August 2000 (BGBl. III, 182/2002). The article led to the judgment of the ECJ in Case C-648/15 Republic of Austria v. Federal Republic of Germany, 12 September 2017. However, this is an isolated example. 76 On the extent of legal constraints on states in the context of arbitrage procedure, see P. Kouraleva- Cazals and A. de Nanteuil, ‘Quel(s) arbitrage(s) en fiscalité internationale?’, Fiscalité Internationale 4- 2022, n°02.5. 77 An expert in PIL recommending interstate arbitration on private initiative, as recommended by the OECD in its most recent Model Convention and MLI: d’Avout, ‘L’entreprise et les conflits internationaux des lois’, 832. 78 Author translation of the term ‘référé législatif mondial’ in ibid., 839. 79 Art. 10 of Council Directive (EU) 2017/ 1852 of 10 October 2017 on tax dispute resolution mechanisms in the EU. 80 Communication from the Commission to the European Parliament and the Council, ‘An action plan for fair and simple taxation supporting the recovery strategy’, COM (2020) 312 final (15 July 2020), 13–14.

International Tax Law and Private International Law 175 the central role played in ECJ tax case law),81 such consistency may prove to be a useful standard in the construction of ITL. Consistency can be sought both in the determination of the nexus and in the articulation of the different applicable tax laws. It may also prove to be the key to finally ensuring legal certainty, a major (if not desperate) concern in recent years.

81

ECJ, Case C-562/19 Commission v. Poland, 16 March 2021, para. 43.

Chapter 11

Internationa l Tax L aw and P u bl i c Internationa l L aw Christiana HJI Panayi and Katerina Perrou

11.1 Introduction The power to levy taxes is inextricably linked with the sovereign state. Indeed, there is no such thing as an ‘international tax’; only ‘national taxes’. As a consequence, there is no international ‘tax system’ as such; each country has its own domestic tax system.1 At the same time, the existence of ‘international tax law’ and its classification as part of international law is not disputed.2 In particular, international tax law is considered to be part of international economic law, along with international investment law.3 But if there is no ‘international tax’ and no ‘international tax system’, then what does ‘international tax law’ stand for? The term ‘international tax law’ is used to describe a country’s legal rules concerning its tax jurisdiction over transactions that cross its borders; this can refer to the economic

1 L. Oats and E. Mulligan, Principles of International Taxation, 7th ed. (Oxford:, Hart Publishing, 2019), s. 2.1; J. Schwarz, On Tax Treaties, 3rd ed. (Alphen aan den Rijn: Wolters Kluwer, 2013), paras 9–150. On the existence of an international tax system, see D. Rosenbloom, ‘The David R. Tillinghast Lecture: International Tax Arbitrage and the “International Tax System”, Tax Law Review 53 (1998), 137; and the contemporary analysis on the same issue by W. Schön, ‘Is There Finally an International Tax System?’, World Tax Journal 13/3 (2021), 357–384. 2 A. Qureshi and A. Kumar, The Public International Law of Taxation, 2nd ed. (Alphen aan den Rijn: Kluwer Law International, 2020), s. 1.02. 3 A. Nollkaemper, ‘Unilateralism/ Multilateralism’ (last updated Mar. 2011), paras 27– 28, in A. Peters and R. Wolfrum, eds, The Max Planck Encyclopedia of Public International Law (Oxford: Oxford University Press, 2008), https://www.mpepil.com (accessed 8 August 2022); Qureshi and Kumar, Public International Law of Taxation.

178 Christiana HJI Panayi and Katerina Perrou activities of foreigners within the territory of a state and the economic activities of the state’s own residents outside its territory.4 In other words, international tax law provides for the national taxation of transnational fiscal facts.5 The subject of international taxation is the issue of whether, and to what extent, a country has the right to tax a person— in other words, what is its jurisdiction to tax. This is a matter of public international law.6 Public international law, on the other hand, is the law that mainly governs the relations between subjects of international law.7 It is the law that determines the rights and obligations that subjects of international law have vis-à-vis each other. The principal subject of international law is the state, possessing full international personality. International organizations also possess limited international law personality. Individuals and companies have only limited rights and obligations in certain areas of international law. Both the rights and obligations of international organizations as well as of private parties under international law are derived from the state. Thus, international law is primarily concerned with the state, whereas national law is concerned chiefly with the individual within the state. Since international tax law is part of public international law and at the same time is rooted in domestic systems and premised on national fiscal autonomy, it is imperative to define what components of public international law are relevant to international taxation. This exercise will define the framework within which the latter is required to operate and elucidate the consequences attached to the quality of a norm being characterized as a ‘public international law norm’. The combination of these two prima facie opposite forces, forms the area of the ‘public international law of taxation’.8 According to Qureshi, the public international law of taxation is comprised of those international law norms, whether tax or non-tax specific, that touch on domestic taxation. These norms are mainly to be found in treaties and are essentially concerned with the management of fiscal sovereignty as well as the jurisdiction of states.9 In essence, these norms function as a coordinator of the exercise by individual states of their tax jurisdiction. As such, the characteristic feature of the public international law of taxation is that it is more facilitative than normative. It is facilitative in the sense that it enables coordination in the application of domestic tax laws, with a view to avoiding excessive taxation but also under-taxation; it enables cooperation between states in order to allow the effective exercise of their tax jurisdiction; and it provides for the avoidance and resolution of conflicts between states in the field of taxation. Certain fundamental areas that belong to this group of rules (i.e. public international law rules that are not specific to tax but apply to tax law and provide a facilitative function for international 4

Y. Margalioth, ‘Taxation, International’ (last updated Apr. 2011), in ibid., para. 1. Qureshi and Kumar, Public International Law of Taxation. 6 Oats and Mulligan, Principles of International Taxation. 7 Qureshi and Kumar, Public International Law of Taxation, s. 1.01. 8 A. Qureshi, ‘Coherence in the Public International Law of Taxation: Developments in International Taxation and Trade and Investment Related Taxation’, Asian Journal of WTO & International Law Health Law and Policy 10 (2015), 193, https://ssrn.com/abstract=2600132. 9 Ibid. 5

International Tax Law and Public International Law 179 tax law) are, especially, the rules on the sources of international law, the rules on tax treaty interpretation, and the rules on dispute resolution. These issues will be discussed in Section 11.2. On the other hand, from a normative perspective, there is little room for rules of public international tax law to develop, as national fiscal sovereignty can only be restricted on a voluntary basis.10 However, there are certain areas of public international law that interact with international tax law, and, because they are norms of public international law, have an impact on both the design and substance of international tax rules. Such rules are the rules on the jurisdiction privileges and immunities, the rules of international human rights law, and the rules of international investment law (especially the rules on protection from expropriation). These issues will be discussed in Section 11.3.

11.2 The Facilities of Public International Law to International Tax Law The aim of this section is to discuss some rules of public international law, which although not specific to taxation, nevertheless provide the framework within which international tax law is shaped and organized. These are, in particular, the rules on the sources of international law, the rules on tax treaty interpretation, and the rules on dispute resolution.

11.2.1 Sources of International Tax Law: Treaties, Custom, and General Principles of Law The international law system, as opposed to national legal systems, has no single legislature and no executive. A disparate network of sui generis courts and tribunals apply international law specific to their differing jurisdictions.11 In this framework, article 38(1) Statute of the International Court of Justice (ICJ Statute) that lists the sources of international law has almost constitutional significance, since it is universally accepted and applied.12 10

For a discussion on the development of what can be described as an international tax regime, see R. S. Avi-Yonah, International Tax and International Law, An Analysis of the International Tax Regime (Cambridge: Cambridge University Press, 2007); R. S. Avi-Yonah, ‘Does Customary International Law Exist?’, in Y. Brauner, ed., Research Handbook on International Taxation (Cheltenham: Edward Elgar, 2020), 2ff. 11 G. Boas, Public International Law (Cheltenham: Edward Elgar, 2012), 45. 12 Ibid.

180 Christiana HJI Panayi and Katerina Perrou According to article 38(1) ICJ Statute, the primary sources of international law are international conventions, international custom, and the general principles of law recognized by civilized nations. In addition to the primary sources, the ICJ Statute also recognizes certain subsidiary means for the determination of rules of law; that is, judicial decisions, subject to the binding result of res judicata, and ‘the teachings of the most highly qualified publicists of the various nations’.

11.2.1.1 International tax treaties Treaties are agreements between sovereign nations. According to article 2 Vienna Convention on the Law of Treaties (VCLT),13 a treaty is an international agreement concluded between states and governed by international law. International tax law has been traditionally developed through (mostly bilateral) tax treaties. The first double tax treaty that applied to income taxes was entered into between Prussia and Austro-Hungary in 1899.14 A handful of further bilateral tax treaties of a similar nature were concluded in Central Europe before the First World War. The bilateral tax treaty network continued to grow, and it now exceeds 3,000 worldwide. The conclusion of the Base Erosion and Profit Shifting Project (BEPS) Project, that included changes to the existing tax treaty rules, led to a multilateral tax treaty (the MLI)15 as a tool to amend existing treaties in an efficient and effective way. Although multilateral conventions do exist in other areas of international law and multilateral treaties— especially in areas of administrative cooperation—are not uncommon in international taxation, the adoption of the MLI was a major step for international tax law.16

13

The convention was adopted on 22 May 1969 and entered into force on 27 January 1980. as Harris observes, it was apparently based on the German Imperial Double Taxation Law of 1870, which sought to relieve double taxation between German states; see P. Harris, International Commercial Tax, 2nd ed. (Cambridge: Cambridge University Press, 2020), s. 1.2.2. 15 See Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting (Multilateral Instrument or MLI), https://www.oecd.org/tax/treaties/multilateral-con vention-to-implement-tax-treaty-related-measures-to-prevent-BEPS.pdf. 16 See Y. Brauner, ‘McBEPS: The MLI— The First Multilateral Tax Treaty That Has Never Been’, Intertax 46 (2018), 6. On the functioning of the MLI under public international law, see the Note by the OECD Directorate for Legal Affairs entitled ‘Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting: Functioning under Public International Law’, https://www.oecd.org/tax/treaties/legal-note-on-the-functioning-of-the-MLI-under-public- international-law.pdf; R. Garcia Anton, ‘The Concept of Substantive Multilateralism’, in P. Dourado, ed., International and EU Tax Multilateralism: Challenges Raised by the MLI (Amsterdam: IBFD, 2020), 15ff; R. Szudoczky and D. Blum, ‘Unveiling the MLI: An Analysis of Its Nature, Relationship to Covered Tax Agreements and Interpretation in Light of the Obligations of Its Parties’, ibid., 125ff; R. Prokisch and F. Souza de Man, ‘Multilateralism and International Tax Law: The Interpretation of Tax Treaties in the Light of the Multilateral Instrument’, ibid., 199ff; M. Lang et al., eds, The OECD Multilateral Instrument for Tax Treaties, Analysis and Effects (Alphen aan den Rijn: Wolters Kluwer, 2018); N. Bravo, A Multilateral Instrument for Updating the Tax Treaty Network (Amsterdam: IBFD, 2020); J. Hattingh, ‘The Multilateral Instrument from a Legal Perspective: What May Be the Challenges?’, Bulletin for International Taxation 71/3/4 (2017), s. 2 (online); Nollkaemper, ‘Unilateralism/Multilateralism’, para. 28. 14 Although,

International Tax Law and Public International Law 181 Tax treaties, whether bilateral or multilateral, are therefore instruments of public international law and at the same time sources of international tax law. They serve as a bridge between the tax systems of the contracting states.17 Their role is to regulate the way in which the taxing powers of each contracting state are exercised. Based on reciprocity, they provide a basis on which the two contracting states divide between them their taxing powers to ensure that double taxation is alleviated or mitigated. As a general rule, double tax treaties do not increase the tax liability of taxpayers; they limit the taxing powers of the contracting states in a mutually agreed way. The MLI itself is a relative newcomer in the international tax field and, arguably, does not go very much beyond the bilateral tax treaty paradigm.18 At best, the MLI can be viewed as being of a hybrid nature, on the border between multilateral and bilateral frameworks, with its predominant aim being to modify bilateral tax treaties. Although its efficacy and viability are yet to be assessed, as an attempt to coordinate the fight against tax evasion of individual states on a global scale, it shows that multilateralism as a means for agreeing upon (substantive) international tax rules is not only possible but also desirable.

11.2.1.2 International Customary Tax Law The issue whether international customary law exists is a controversial one. Avi-Yonah holds the view that the existence of over 3,000 bilateral tax treaties (which are about 80% identical) might, in limited instances, arguably create customary international tax law in four areas: jurisdiction to tax; the permanent establishment threshold; the arm’s-length standard; and non-discrimination.19 Such customary international tax law, if it is found to exist, would then be binding on states, even in the absence of a double tax treaty. The argument that customary international law of taxation exists could be supported by empirical evidence showing that there is a clear trend towards convergence in the international legal language adopted in the bilateral tax treaties. This fact may suggest that countries are guided by transnational legal considerations.20 However, at this stage of development, it appears that despite the dense network of double tax treaties and the convergence in the legal language adopted in them (at least in some areas), the uniformity observed in tax treaty practice is mainly due to the fact that this is seen as being in the best interests of states and is not due to any legal conviction (i.e. an obligation to follow a certain practice).21 17 B. Arnold, International Tax Primer, 4th ed. (Alphen aan den Rijn: Kluwer Law International, 2019), s. 8.1. 18 See the analyses by various contributors in Dourado, International and EU Tax Multilateralism. 19 Avi-Yonah, ‘Does Customary International Law Exist?’. 20 E. Ash and O. Marian, ‘The Making of International Tax Law: Empirical Evidence from Natural Language Processing’, University of California Irvine School of Law, Legal Studies Research Paper Series No. 2019–02 (2019), available at https:///ssrn.com/abstract=3314310. The authors stop short of concluding that customary international law of taxation exists. 21 C. Braumann, ‘Taxes and Custom: Tax Treaties as Evidence for Customary International Law’, Journal of International Economic Law 23/3 (2020), 747–769, 766; R. Codorniz Leite Pereira, ‘The Emergence of Transparency and Exchange of Information for Tax Purposes on Request as an

182 Christiana HJI Panayi and Katerina Perrou Recently, international tax policymaking appears to be shifting from the states acting individually to a new public international institution, the so-called Inclusive Framework.22 The impact of the creation of the OECD’s Inclusive Framework23 on international tax law remains to be seen.24 It must be pointed out, though, that already jurisdictions seeking to participate in the Inclusive Framework must commit to implementing the four BEPS minimum standards and be subject to peer-review monitoring. Currently, 140 countries and jurisdictions are members of the Inclusive Framework and work together with a view to developing new, consensus-based rules for international taxation. The Inclusive Framework has made major progress by developing a two-pillar solution to address the tax challenges arising from the digitalization of the economy. As of 16 December 2022, 138 OECD/G20 Inclusive Framework member jurisdictions have joined the new two-pillar solution to reform international taxation rules.25 This practice, when assessed in the light of public international law, suggests that perhaps the mechanism for the creation of international customary tax law has already been set in motion in two areas: on the reallocation of some taxing rights over multinational enterprises (MNEs) from their home countries to the markets where they have business activities and earn profits; and on the introduction of a global minimum corporate tax rate as a global anti-base erosion mechanism.

11.2.1.3 General principles of international tax law As regards the third category of primary sources of international (tax) law, the general principles of law recognized by nations, it should be noted that the recognition of such general principles of law lies with the judiciary. It has been argued that taxation based on fiscal sovereignty and taxation based on reasonable nexus to a state belong to this category.26 At this stage of development of international tax law, it may not be easily proven

International Tax Custom’, Intertax 48 (2020), 624; E. Gil García, ‘The Single Tax Principle: Fiction or Reality in a Non-Comprehensive International Tax Regime?’, World Tax Journal 11/3 (2019), 497–556. 22 For a discussion of various aspects of the Inclusive Framework, see I. Ozai, ‘Institutional and Structural Legitimacy Deficits in the International Tax Regime’, World Tax Journal 12/1 (2020), 53–78; A. Christians and L. van Apeldoorn, ‘The OECD Inclusive Framework’, Bulletin for International Taxation 72/4/5 (2018), 226–233; I. J. Mosquera Valderrama, ‘Output Legitimacy Deficits and the Inclusive Framework of the OECD/G20 Base Erosion and Profit Shifting Initiative’, Bulletin for International Taxation 72/3 (2018), 160–170. 23 OECD, ‘All Interested Countries and Jurisdictions to Be Invited to Join Global Efforts Led by the OECD and G20 to Close International Tax Loopholes’ (23 Feb. 2016), https://www.oecd.org/ctp/all-int erested-countries-and-jurisdictions-to-be-invited-to-join-global-efforts-led-by-the-oecd-and-g20-to- close-international-tax-loopholes.htm 24 OECD, ‘Inclusive Framework on BEPS, Background Brief ’ (Jan. 2017), https://www.oecd.org/tax/ beps/background-brief-inclusive-framework-for-beps-implementation.pdf. 25 See statement available at https://www.oecd.org/tax/beps/statement-on-a-two-pillar-solution-to- address-the-tax-challenges-arising-from-the-digitalisation-of-the-economy-october-2021.pdf. 26 See the analysis and discussion of previous literature on the matter in S. Gadzo, Nexus Requirements for Taxation of Non-Residents’ Business Income—A Normative Evaluation in the Context of the Global Economy (Amsterdam: IBFD, 2018), s. 2.2.

International Tax Law and Public International Law 183 that the arm’s-length principle on a stand-alone and abstract basis constitutes a general principle of international tax law. However, the arm’s-length principle in the framework of the OECD Transfer Pricing Guidelines may have been elevated into such a principle. Two examples might be used to support this thesis. In Unilever Kenya Ltd,27 the Kenyan court examined the value of the OECD Transfer Pricing Guidelines. Since the domestic law provision was considered ambiguous, the taxpayer argued that the OECD Transfer Pricing Guidelines should be used as an interpretative aid. The court held that in the absence of relevant guidelines in Kenya, it was necessary to look elsewhere, as one ‘cannot overlook or sideline what has come out of the wisdom of taxpayers and tax collectors in other countries’. Although the court was not legally bound to apply the OECD Transfer Pricing Guidelines, it was not able to ignore them either. The court concluded that it would apply the OECD Transfer Pricing Guidelines, as ‘[t]‌hese have been evolved in other jurisdictions after considerable debates and taking into account appropriate factors to arrive at results that are equitable to all parties . . . and it would be fool-hardy for any court to disregard internationally accepted principles of business as long as these do not conflict with our own laws’.28 A similar issue arose in the UK in the case of DSG Retail Ltd & others v. HMRC.29 In this case, Special Commissioners J. F. Avery Jones and C. Hellier ruled that although in a particular domestic provision the OECD Model Tax Convention was not expressly incorporated, it seemed to them that ‘in determining the arm’s length price, the approach of the OECD model is a useful aid which we should apply in the absence of any other guidance as they are the best evidence of international thinking on the topic’.30 In both cases, the judges examined what was happening at international level. Although the OECD Transfer Pricing Guidelines were not recognized as legally binding rules in determining the content and application of the arm’s-length principle, they were nevertheless considered as a normative framework that the judges were able (perhaps, even required) to follow, on the sole premise that it was ‘internationally recognised’. Although we do not go as far as to submit that the arm’s-length principle, as supported in the OECD Model and in the OECD Transfer Pricing Guidelines, constitutes a general principle of law in the sense of article 38(1) ICJ Statute, what is obvious is that the courts in these and other cases are often swayed by the international consensus over the OECD Transfer Pricing Guidelines to fill the gaps in domestic law.

27 Judgment

of the High Court of Kenya at Nairobi of 5 October 2005 on the Income Tax Appeal no. 753 of 2003, Unilever Kenya Limited v. The Commissioner of Income Tax [2005] eKLR, http://kenyalaw. org/caselaw/cases/view/12804/. 28 Ibid., 8. 29 DSG Retail Ltd & others v. HMRC TC00001 [2009] UKFTT 31 (TC), https://www.bailii.org/uk/ cases/UKFTT/TC/2009/31.html. 30 Ibid., [77].

184 Christiana HJI Panayi and Katerina Perrou

11.2.2 The Impact of the VCLT on the Interpretation of Double Tax Treaties; On ‘Context’ and ‘Object and Purpose’ Double tax treaties are closely linked with the domestic tax legislation of the contracting states. Although tax treaties do not, in principle, impose any taxes, nor do they allocate any free-standing taxing rights to the contracting states (these taxing rights are derived from domestic law and as such are a manifestation of state sovereignty), tax treaties interact with domestic tax laws as they limit the taxes otherwise imposed by a state.31 Therefore, in many aspects the interpretation of tax treaties is similar to the interpretation of domestic tax legislation. Indeed, the fundamental principles on the interpretation of international agreements, seen as a whole, are not so different from those which would govern interpretation under domestic law.32 At the same time, however, tax treaties are instruments of international law and, as a consequence, their interpretation is governed by the Vienna Convention on the Law of Treaties (VCLT).33 As the rules on interpretation contained in the VCLT constitute codification of existing international customary law, they are binding on all states, even if a state has not acceded to the VCLT.34 Moreover, the fact that the VCLT was drafted in 1969 and entered into force in 1980, makes no difference as to the binding effect of the rules on interpretation. As the VCLT codifies pre-existing customary law, it applies to all tax treaties irrespective of when they were concluded between states (i.e. before or after the contracting states ratified the VCLT).35 The rules for the interpretation of international treaties are found in articles 31–33 VCLT. The general rule of interpretation is established in article 31 VCLT. Article 32 provides for supplementary means of interpretation which are subordinate to the means under article 31, whereas article 33 establishes the rule for the interpretation of treaties authenticated in two or more languages. According to the general rule of interpretation established in article 31(1) VCLT, a treaty shall be interpreted in good faith in accordance with the ordinary meaning

31 Arnold, International Tax Primer, s. 8.2.2.

32 K. Vogel and A. Rust, ‘Introduction’, in E. Reimer and A. Rust, eds, Klaus Vogel on Double Taxation Conventions, 4th ed. (Alphen aan den Rijn: Kluwer Law International, 2015), 35ff. 33 See Arnold, International Tax Primer, s. 8.6.2; Vogel and Rust, ‘Introduction’, 34ff; Schwarz, On Tax Treaties, paras 12-050ff; for a thorough analysis of tax treaty interpretation under international law, see F. Engelen, Interpretation of Tax Treaties under International Law (Amsterdam: IBFD, 2004); for comparative aspects, see M. Lang, ed., Tax Treaty Interpretation (The Hague: Kluwer Law International, 2001); IFA, ‘Interpretation of Double Taxation Conventions’, Cahiers de Droit Fiscal International, vol. 78a (1993). 34 Vogel and Rust, ‘Introduction’, 24, with references to the case law of international and national courts. 35 Compare art. 4 VCLT; see also Commissioner of Taxation v. SNF (Australia) Pty [2011] FCAFC 74, para. 113.

International Tax Law and Public International Law 185 to be given to the terms of the treaty in their context and in the light of its object and purpose.36 Relief from double taxation, however, is not the only objective pursued by double tax treaties.37 Since 1977, the Commentary on article 1 OECD Model Tax Convention has been modified to provide expressly that tax treaties were not intended to encourage tax avoidance or evasion.38 The 1992 OECD Model Tax Convention marked a significant shift, as a shorter title of the convention was used without the (restrictive) reference to the elimination of double taxation. At the same time, however, it was also understood that the practice of many states was still to include in the title a reference to either the elimination of double taxation or to both the elimination of double taxation and the prevention of fiscal evasion.39 In 2003, paragraph 7 of the Commentary on article 1 was amended to clarify that the prevention of tax avoidance was also a purpose of tax treaties. As a result of work undertaken as part of the OECD/G20 BEPS Action 6 minimum standard, the title of the convention was amended in 2017.40 A preamble was included for the first time clarifying that the objectives of double tax treaties are not limited to the elimination of double taxation and that the contracting states do not intend the provisions of their tax treaties to create opportunities for non-taxation or reduced taxation through tax evasion and avoidance.41 As mentioned in the Final Report on BEPS Action 6, the changes to the title and the preamble of the model treaty were made having specifically in mind that these would be relevant for the interpretation of double tax treaties.42 Indeed, the clear statement of the intention of the signatories to a tax treaty that appears in the preamble is relevant to the interpretation and application of the provisions of that treaty according to the basic rules of interpretation of treaties under the VCLT (art. 31(1)). The context of the treaty includes its preamble.43

36 For a discussion of the terms ‘context’ and ‘object and purpose’ of a double tax treaty, see Vogel and Rust, ‘Introduction’, 39; Schwarz, On Tax Treaties, paras 12-150ff. 37 The 1963 and 1977 OECD Model Tax Conventions only referred to the avoidance of double taxation; see the text of the 1963 Model and the text of the 1977 Model. 38 1977 OECD Model Commentary on art. 1, para. 7. 39 See Model Tax Convention on Income and on Capital September 1992, para. 16 of the Introduction. 40 See 2017 OECD Model Tax Convention, 27, ‘Title of the Convention’, which now reads as follows ‘Convention between (State A) and (State B) for the elimination of double taxation with respect to taxes on income and on capital and the prevention of tax evasion and avoidance’. 41 See 2017 OECD Model Tax Convention, 27, ‘Preamble’, which reads as follows:

(State A) and (State B), Desiring to further develop their economic relationship and to enhance their cooperation in tax matters, Intending to conclude a Convention for the elimination of double taxation with respect to taxes on income and on capital without creating opportunities for non-taxation or reduced taxation through tax evasion or avoidance (including through treaty-shopping arrangements aimed at obtaining reliefs provided in this Convention for the indirect benefit of residents of third States), Have agreed as follows. 42 43

See BEPS Action 6 Final Report (2015), para. 77. See also the discussion in paras 16.1 and 16.2 of the 2017 OECD Model Tax Convention.

186 Christiana HJI Panayi and Katerina Perrou

11.2.3 International Tax Dispute Resolution International tax disputes arise when taxpayers disagree with tax authorities on the interpretation or application of a certain international tax rule. This rule may be a unilateral rule established by domestic legislation, or a double tax treaty rule if a treaty has been agreed between the states. For the resolution of this type of dispute, remedies provided in the domestic law of the state concerned are often available to the taxpayers. International tax disputes also arise when states disagree between them as to the application and interpretation of a provision contained in a double tax treaty. This kind of dispute, being a dispute between states, is a pure international law dispute, primarily governed by public international law. Under public international law, there is no obligation on the states to settle disputes, and the procedures for settlement by formal and legal procedures rest on the consent of the parties involved. Accordingly, judicial settlement in international relations is rather exceptional.44 Public international law broadly distinguishes between two categories of mechanisms for the settlement of international disputes: diplomatic means and arbitral/judicial means. The main difference between the two lies in the fact that when diplomatic means are used, the parties retain control over both the procedure itself and, more importantly, the outcome of the procedure, since any solution proposed by a third party will not be automatically binding upon them. By contrast, in the case of arbitration or adjudication, the parties accept as binding the final solution adopted by the international arbitrator or judge. Article 2(3) Charter of the United Nations45 provides that the states shall settle their international disputes by peaceful means. According to article 33(1) UN Charter, the parties to any dispute shall, first of all, seek a solution by negotiation, enquiry, mediation, conciliation, arbitration, judicial settlement, resort to regional agencies or arrangements, or other peaceful means of their own choice. The principle of peaceful resolution of international disputes is further elaborated in the 1982 Manila Declaration on the Peaceful Settlement of International Disputes.46 The Manila Declaration enunciates the duty of states to ‘act in good faith’, with a view to avoiding disputes among themselves likely to affect friendly relations among states.47 44

According to J. Crawford, Brownlie’s Principles of Public International Law, 9th ed. (Oxford: Oxford University Press, 2019), 706, from 1946 to 31 July 2018, the International Court has dealt with some sixty- nine judgments on the merits, thirty-one preliminary objections, eight judgments on jurisdiction and admissibility, and thirty-seven requests for provisional measures, as well as twenty-six requests for advisory opinions. As of August 2018, there were seventeen contentious cases and one request for an advisory opinion pending. State reluctance to resort to the Court may be explained by several factors, such as the political fact that hauling another state before the Court is often regarded as unfriendly; the general conditions of international relations; and a preference for the flexibility of arbitration versus judicial process. 45 Charter of the United Nations and Statute of the International Court of Justice 1945, https://www. un.org/en/charter-united-nations/index.html. 46 Text available at https://legal.un.org/avl/ha/mdpsid/mdpsid.html. 47 Manila Declaration, s. I, para. 1.

International Tax Law and Public International Law 187 States acting in good faith and in a spirit of cooperation shall seek to reach an early and equitable settlement of their international disputes by any of the peaceful means that may be appropriate to the circumstances and nature of their dispute.48 In addition, it is provided that states shall in accordance with international law implement in good faith all the provisions of agreements concluded by them for the settlement of their disputes.49 Double tax treaties contain a special mechanism for the resolution of international tax disputes: the mutual agreement procedure, which is built on the principles of public international law on the peaceful resolution of international disputes and operates within the framework set by public international law. As a result, the mutual agreement procedure is traditionally viewed as a state-to-state procedure in which the taxpayer has no standing.50 The preferred means used in the mutual agreement procedure are the non-judicial means provided for in article 33 UN Charter and, in particular, direct negotiations between the parties.51 Under article 25(5) OECD Model, the competent authorities may, under conditions and following the parties’ agreement, refer any unresolved issues of an individual mutual agreement procedure to mandatory binding arbitration, which is another of the means provided for in article 33 UN Charter.52 For disputes between EU member states, the Court of Justice of the European Union can serve as arbitrator, as is the case of the 2000 Austria–Germany double taxation convention.53 Moreover, within the EU, under the 2017 Dispute Resolution Directive,54 member states are also encouraged to use other means of dispute resolution during the final stages of the mutual agreement procedure, such as mediation or conciliation.55 48

Ibid., para. 5. Ibid., para. 11. 50 H. Ault and J. Sasseville, ‘2008 OECD Model: The New Arbitration Provision’, Bulletin for International Taxation 63/208 (2009), 210. While the mutual agreement procedure is often seen as a state-to-state procedure for the so-called general mutual agreement procedure of art. 25(3) of the OECD Model, the same cannot be convincingly supported for the ‘specific case’ or ‘individual mutual agreement procedure’ of art. 25(1). See, in general, K. Perrou, Taxpayer Participation in Tax Treaty Dispute Resolution (Amsterdam: IBFD, 2014). 51 In the 1963 OECD Model, the Commentary on art. 25 at para. 9 suggested that as the provisions of the Convention and the Commentary were the result of close international joint work with the Fiscal Committee, the Fiscal Committee could be asked to give an opinion on the correct understanding of the provisions, where difficulties of interpretation arose. 52 On the use of arbitration for the resolution of tax treaty disputes, see A. Salehifar, ‘Rethinking the Role of Arbitration in International Tax Treaties’, Journal of International Arbitration 37 (2020), 87; M. Lang et al., eds., Tax Treaty Arbitration (Amsterdam: IBFD, 2020); A. Majdanska and L. Turcan, eds, OECD Arbitration in Tax Treaty Law (Vienna: Linde, 2018). 53 See Austro-German Convention, art. 25(5). This mechanism has only been used once: see Case C- 648/15 Republic of Austria v. Federal Republic of Germany, ECLI:EU:C:2017:664. 54 Council Directive (EU) 2017/ 1852 of 10 October 2017 on tax dispute resolution mechanisms in the European Union [2017] OJ L265/1. See C. HJI Panayi, European Union Corporate Tax Law (Cambridge: Cambridge University Press, 2021), ch. 2, 2.5. 55 See recital 6 of the Dispute Resolution Directive. The directive seems to provide for an enhanced taxpayer role, see K. Perrou, ‘Dispute Resolution and Taxpayer Participation’, in C. HJI Panayi, W. Haslehner, and E. Traversa, eds, Research Handbook on European Union Taxation Law 49

188 Christiana HJI Panayi and Katerina Perrou Last but not least, the International Court of Justice (ICJ) also has competence to adjudicate on international tax disputes.56 However, the jurisdiction of the ICJ, as with any form of adjudication under public international law, depends entirely on whether the parties have chosen to submit their disputes to it; in practice this has rarely happened.57 In conclusion, it is argued that public international law provides a framework for the resolution of cross-border tax disputes. However, certain limitations within this framework, especially as far as the participation of taxpayers is concerned, need to be re- examined in the light of developments on dispute-resolution mechanisms in other areas of public international law, especially in the area of international investment law.58

11.3 The Impact of Public International Law on the Substantive Rules of International Tax Law The aim of this section is to discuss in brief the rules of public international law that, while not being specific to taxation, nevertheless have an impact on the taxing rights and

(Cheltenham: Edward Elgar, 2020), 541ff. The directive also provides for the establishment of a standing committee to deal with the resolution of tax treaty disputes within the EU; for a discussion of various issues of the implementation of this provision, see S. Piotrowski et al., ‘Towards a Standing Committee Pursuant to Article 10 of the EU Tax Dispute Resolution Directive: A Proposal for Implementation’, Intertax 47 (2019), 678; Fiscalis Project Group (FPG) 093, ‘Working Paper on the Implementation of Article 10 of Directive (EU) 2017/1852 on Tax Dispute Resolution Mechanisms in the European Union’ (Aug. 2019), https://ec.europa.eu/taxation_customs/system/files/2019-10/2019-tax-dispute-resolution- fiscalis-project-group-report.pdf. 56 On the jurisdiction of the ICJ, see L. Nobrega e Silva Loureiro, ‘Mutual Agreement Procedure: Preventing the Compulsory Jurisdiction of the International Court of Justice?’, Intertax 37 (2009), 529; E. van der Bruggen, ‘Compulsory Jurisdiction of the International Court of Justice in Tax Cases: Do We Already Have and “International Tax Court”?’, International Tax Review 29 (2001), 250; E. van der Bruggen, ‘About the Jurisdiction of International Courts to Settle Tax Treaty Disputes’, in Lang and Züger, Settlement of Disputes in Tax Treaty Law, 500ff. 57 See, e.g., art. 41(5) of the 1992 Convention between the Kingdom of Sweden and the Federal Republic of Germany for the Avoidance of Double Taxation with respect to the Taxes on Income and Capital as well as on Inheritances and Gifts and Concerning Mutual Administrative Assistance in Tax Matters (unofficial translation), IBFD Treaties and Models; the first sentence of art. 41(5) provides that ‘The provisions of Chapters I, II and IV of the European Convention on the Peaceful Settlement of Disputes of 29 April 1957 shall be applied to settle international conflicts arising under this Convention’; according to art. 1 of the latter convention, the parties shall submit to the judgment of the ICJ all international legal disputes which may arise between them. 58 On investment treaty arbitration, see J. A. Mortera Cavazos, ‘Arbitration of Tax Matters Based on Bilateral Investment Agreements’, in Majdanska and Turcan, OECD Arbitration in Tax Treaty Law, 613ff; see also the contributions in M. Lang et al., eds, The Impact of Bilateral Investment Treaties on Taxation (Amsterdam: IBFD, 2017); A. Gildemeister, L’ arbitrage des différends fiscaux en droit international des investissements (Paris: LGDJ, 2013).

International Tax Law and Public International Law 189 obligations of the state (i.e. they have the power to dictate the way the taxing jurisdiction of a state is exercised). These are, in particular, the rules on jurisdiction privileges and immunities, the obligations of the states under international human rights law, and the obligations arising from international investment law, especially as regards protection from expropriatory taxation.

11.3.1 Fiscal Privileges and Immunities under International Law In accordance with the principle of sovereign equality provided in article 2(1) UN Charter, public international law prohibits the taxation of a sovereign state by another sovereign state.59 Under international customary law, states enjoy immunity for acts performed in the exercise of sovereign authority. Income from activities of a commercial nature is not exempt from tax under the rules of general international law; in these cases, taxation is governed exclusively by the provisions of double tax treaties.60 This fiscal immunity is also extended to members of diplomatic missions and consular posts. The fiscal privileges for members of diplomatic missions and consular posts form part of international customary law and are primarily recorded by two international instruments, the Vienna Convention on Diplomatic Relations of 18 April 1961 and the Vienna Convention on Consular Relations of 24 April 1963.61 These instruments grant different degrees of tax privileges, depending on the rank and function of the respective person; they range from full to limited privileges. Under article 28 OECD Model, it is provided that the provisions of a double taxation convention shall not affect the fiscal privileges of members of diplomatic missions or consular posts under the general rules of international law or under the provisions of special agreements. This provision aims to ensure that members of diplomatic missions and consular posts shall, under the provisions of a double taxation convention, receive no less favourable treatment than that to which they are entitled under international law or under special international agreements.62 Although not explicitly mentioned, the protection under article 28 also extends to staff of international organizations who enjoy similar privileges. A second aspect of the application of the doctrine of sovereign immunity to the exercise of government functions is found in the tax treatment of other remuneration paid to government officials, in respect of government services, under article 19 OECD

59 See also the discussion in Gadzo, Nexus Requirements for Taxation of Non- Residents Business Income. 60 R. Ismer and D. Endres, ‘Article 28’, in E. Reimer and A. Rust, eds, Klaus Vogel on Double Taxation Conventions, 4th ed. (Alphen aan den Rijn: Kluwer Law International, 2015), 1979ff. 61 Ibid., 1983 62 OECD Commentary on art. 28, para. 1.

190 Christiana HJI Panayi and Katerina Perrou Model.63 The distributive rule adopted in article 19 confers to the paying state the exclusive right of taxation of remuneration from government services as well as government pensions. Fiscal immunities, especially the diplomatic immunities pursuant to international customary law, have always been part of international law. These public international law principles affect international taxation rules, and—in some shape or form—have found their way into the OECD Model.

11.3.2 International Human Rights Law and International Tax Law Double tax treaties do not contain any provisions regarding the protection of taxpayer rights. This can be, at least partially, explained by the fact that the models, on the basis of which the vast majority of double taxation conventions have been negotiated, predate the main developments in human rights law.64 Moreover, arguably, it is not really a matter of international tax law to provide guarantees for the protection of the taxpayer. This is a matter for the domestic law of each state and also a matter regulated universally by international human rights law, which is another branch of public international law.65 However, public international law recognizes rights for non-state actors. In tax matters, these include taxpayers and other private persons involved in the levying of taxes. Their fundamental rights are human rights, which must be effectively protected even when there is a general community interest in the collection of tax.66 The human rights system, based on the Universal Declaration of Human Rights and elaborated in the various regional human rights conventions,67 provides for worldwide coverage. As an example, in the EU there exist multiple levels of protection of taxpayer rights. Domestically, taxpayers enjoy the protection granted by their member

63 Schwarz, On Tax Treaties, para. 19-600. 64 P.

Baker, ‘Double Taxation Conventions and Human Rights’, in P. Baker and C. Bobett, eds, Tax Polymath: A Life in International Taxation (Amsterdam: IBFD, 2011), 63ff, 63. 65 C. HJI Panayi and K. Perrou, ‘Tax Justice in the Post-BEPS Era: Enhanced Cooperation among Tax Authorities and the Protection of Taxpayer Rights in the EU’, in P. Harris and D. de Cogan, eds, Tax Justice and Tax Law (Oxford: Hart Publishing, 2020), 203ff, 205. 66 For a comprehensive worldwide analysis of human rights and a discussion of the issues that arise in national and cross-border situations in connection with tax procedures, substantive law, and sanctions, see J. Kokott and P. Pistone, Taxpayers in International Law (Munich/Oxford/Baden-Baden: Beck/Hart/ NOMOS, 2022). 67 Following the adoption of the Universal Declaration on Human Rights (1948) and of the International Covenant on Civil and Political Rights (ICPPR) (in force since 1976), three major regional human rights systems were developed: the European system in the 1950s (the European Convention for the Protection of Human Rights and Fundamental Freedoms (ECHR)); the Inter-American system in the 1970s (see the American Convention on Human Rights ‘Pact of San Jose, Costa Rica’); and the African system in the 1980s (see the ‘African Charter on Human and People’s Rights’, also known as the Banjul Charter).

International Tax Law and Public International Law 191 state constitutions. At the EU level, after the entry into force of the EU Charter of Fundamental Rights68 on 1 December 2009, taxpayers enjoy the protection of the rights and guarantees provided in the Charter. In parallel, the European Convention for the Protection of Human Rights and Fundamental Freedoms (ECHR)69 is also applicable in the EU, as all member states are signatories and subject to the jurisdiction of the European Court of Human Rights (ECtHR). Even though the scope of application of the two instruments is not the same as regards the rights and guarantees included therein, their territorial scope covers the whole of the EU and, subject to the conditions, these instruments are also available to taxpayers in the EU. Taxpayers have relied on the ECHR to challenge the compatibility of various domestic tax measures.70 There is no reason why international tax provisions, as provided for in double tax treaties, cannot be scrutinized for compatibility with human rights norms, just like any part of a tax system. Indeed, as Philip Baker argues, there is no reason why tax treaties should be immune from scrutiny to ensure consistency with international human rights instruments. Nor is there any reason in principle why taxpayers cannot challenge the application of tax treaty provisions on the grounds that those provisions infringe human rights norms.71 Both substantive and procedural aspects of international tax law have been the subject of interesting case law by the ECtHR, showing that rules of international tax law should also be compatible with human rights law, respecting, however, the margin of appreciation that states enjoy under human rights law.72 The most prominent areas where human rights law has had an impact on international tax law are in the context of exchange of information and dispute resolution. As far as exchange of information on the basis of article 26 OECD Model is concerned, there have been clashes with certain procedural rights and guarantees derived from article 6 ECHR, which provides for the right to a fair trial, and from article 8 ECHR, which provides for the right to respect for private and family life, home, and correspondence.73 As far as the mutual agreement procedure under article 25 OECD Model is concerned, again, and to the extent that the nature of the procedure lies closer to a dispute-resolution procedure than

68

Charter of Fundamental Rights of the European Union [2012] OJ C326/391. Council of Europe, Convention for the Protection of Human Rights and Fundamental Freedoms (1950). 70 See e.g. the factsheet published by the ECtHR, ‘Taxation and the European Convention on Human Rights’ (Apr. 2019), https://www.echr.coe.int/Documents/FS_Taxation_ENG.pdf. 71 Baker, ‘Double Taxation Conventions and Human Rights’, 64. 72 For an overview of general tax law cases decided under the ECHR, see the 2021, 2020, 2019, 2018, and 2015–2017 general and special reports on the protection of taxpayers’ rights by the IBFD Observatory for the Protection of Taxpayers’ Rights, https://www.ibfd.org/Academic/National-Reports-Observatory- Protection-Taxpayers-Rights; C. Endresen, ‘Taxation and the European Convention for the Protection of Human Rights: Substantive Issues’, Intertax 45 (2017), 508; G. Kofler, M. Poiares Maduro, and P. Pistone, eds, Human Rights and Taxation in Europe and the World (Amsterdam: IBFD, 2011). 73 For an analysis of the relevant issues and discussion of pertinent case law, see HJI Panayi and Perrou, ‘Tax Justice in the Post-BEPS Era’, 206ff. 69

192 Christiana HJI Panayi and Katerina Perrou a treaty negotiation, the procedural rights and guarantees derived from article 6 ECHR dealing with the right to a fair trial are applicable.74 It appears that the enhanced powers of cooperation between tax authorities following recent developments in international tax law will increase the need for effective scrutiny of the rules as to their compatibility with international human rights norms.

11.3.3 International Investment Law and Protection from Expropriatory Taxation International investment law is a specific branch of public international law. Whereas international tax treaty law limits the contracting states’ right to tax on the basis of reciprocity, bilateral investment treaties also provide, under certain conditions, for a limitation of the right of the host state to tax the foreign investor. The rules of bilateral investment treaties and bilateral tax treaties often overlap, especially in the area of expropriatory taxation measures. As each instrument serves different purposes, they apply in parallel and neither of them can substitute the other. The relationship between international investment agreements and double tax treaties is governed by specific clauses that are usually found in international investment agreements and can range from complete carve-outs (maximizing the regulatory autonomy of the host state to tax foreign investors), selective carve-outs, and conflict clauses (safeguarding the application of double tax treaties in all cases to the exclusion of the international investment agreement).75 The term expropriation in international law refers to the taking of foreign-owned property by a state whether for public purposes or other reasons. The state has to satisfy certain conditions in order for its actions to be considered legitimate.76 The interaction between taxation measures and expropriation was considered in detail by the ICSID tribunal in the case of Burlington v. Ecuador.77 The tribunal noted that in the absence of guidance in the applicable treaty as to the relationship between taxation and

74 For an analysis of the relevant issues and case law, see K. Perrou, ‘Dispute Resolution and Taxpayer Participation’, in HJI Panayi, Haslehner, and Traversa, Research Handbook on European Union Taxation Law, 541ff. 75 See the discussion and classification in V. Vasudev, ‘Interactions between Taxation Measures and International Investment Agreements’, in J. Chaisse et al., eds, Handbook of International Investment Law and Policy (Singapore: Springer, 2021), 2035–2053; P. Pistone, ‘General Report’, in Lang, Impact of Bilateral Investment Treaties on Taxation, 17ff. 76 On international law on expropriation, see R. R. Babu, ‘Standard of Compensation for Expropriation of Foreign Investment’, in Chaisse et al., Handbook of International Investment Law and Policy, 419–435. 77 Burlington Resources Inc. v. Republic of Ecuador (2012), ICSID Case No. ARB/08/5, https://www.ita law.com/sites/default/files/case-documents/italaw1094_0.pdf. For an analysis of the issues posed in the Burlington case, see A. Lazem and I. Bantekas, ‘The Treatment of Tax as Expropriation in International Investor–State Arbitration’, Arbitration International (2015), 1–46.

International Tax Law and Public International Law 193 expropriation, the tribunal would consider the limits on the host state’s ability to impose taxation measures under customary international law. The tribunal noted that customary international law imposed two limitations on the power to tax: taxes may not be discriminatory and taxes may not be confiscatory, that is they may not ‘take too much from the taxpayer’.78 If the amount of tax required was so high that taxpayers were forced to abandon property or sell it at a distress price, the tax was confiscatory.79 The concept of confiscatory taxation was related to that of expropriatory taxation under investment treaties.80 In the Burlington v. Ecuador case, the standard for expropriatory taxation was set rather high as it required a ‘substantial deprivation’ of the investment;81 it required that the investment as a whole become unviable and unable to generate a commercial return. While loss of profits in one year could indicate that the investment had become unviable, an assessment of the capacity to earn a commercial return in future years was necessary in order to establish expropriatory taxation.82 In other words, a tax measure constituted expropriation if it resulted in a ‘radical deprivation’ of the investment;83 the financial suffering caused by a tax measure is not sufficient to establish expropriation if the investor could ‘continue to function profitably and to engage in normal range of activities’.84 Therefore, the standard for expropriatory taxation under international investment treaties is rather high. Yet, the right of a state to tax is still potentially limited by the protection granted to investors under international investment agreements.

11.4 Concluding Remarks: Towards Greater Internationalization of Tax Law In recent years, an unprecedented internationalization of tax law has taken place.85 This chapter has explored the interface between aspects of public international law and

78

Burlington, para. 393 (‘Taxation is, in a sense, a partial confiscation’). A. R. Albrecht, ‘The Taxation of Aliens under International Law’, British Yearbook of International Law 29/145 (1952), 173 79 Burlington, para. 393. 80 Ibid., para. 394; see also the discussion in P. Pistone, ‘General Report’, in Lang, Impact of Bilateral Investment Treaties on Taxation, 30ff. 81 Burlington, para. 396 82 Ibid., para. 399. 83 Cargill, Incorporated v. United Mexican States (2009), ICSID Case No. ARB(AF)/05/2, paras 366 and 368 84 EnCana Corporation v. Republic of Ecuador (2006), LCIA Case No. UN3481, UNCITRAL, para. 174. 85 See the discussion in J. Kokott, P. Pistone, and R. Miller, ‘Public International Law and tax Law: Taxpayers’ Rights; The International Law Association’s Project on International Tax Law-Phase 1’, Georgetown Journal of International Law 52/2 (2021), 381–426. .

194 Christiana HJI Panayi and Katerina Perrou international taxation. Apart from the international taxation rules described in the previous sections, the institutional framework of international taxation is also affected. It is not disputed that the states, and to a limited extent the European Union as a supranational legal order separate from its members, are the driving force of the developments in international taxation. However, at the same time, the role of intergovernmental organizations, which also have standing under public international law, especially the role of the OECD (in particular, the Committee on Fiscal Affairs) and the UN (in particular, the UN Tax Committee), in shaping international tax law cannot be ignored.86 The creation of the OECD’s Inclusive Framework is an important development with wide-ranging ramifications, as it broadens the scope of countries participating in the implementation of the BEPS proposals and (to an extent) levels the playing field when it comes to influencing outcomes. Any consensus-based recommendations involving the Inclusive Framework are likely to be seen as enjoying more legitimacy and are more likely to be adopted by states, potentially leading to the creation of international customary tax law. Arguably, the litmus test is likely to be whether the OECD/G20 manage to achieve international consensus over the detailed proposals to revise international tax rules under Pillars One and Two and whether these proposals will be implemented uniformly by the members of the Inclusive Framework. It is undeniable that recently the multilateral process for reaching consensus has been given a powerful and welcome boost: on 8 October 2021 it was announced that the Inclusive Framework had agreed the two-pillar solution to address the tax challenges arising from the digitalization of the economy. This was adopted by the G20 on 31 October 2021.87 Even though as of 16 December 2022 not all members of the Inclusive Framework had joined, 138 member jurisdictions had agreed on the joint statement laying out the components of each pillar.88 One of the characteristics of Pillar One, which will be implemented by a multilateral agreement once the detailed rules are finalized, is that states will be required to remove all digital services taxes and other similar unilateral measures. They will also be required to commit not to introduce such measures in the future. The impact of Pillar Two on the freedom of states to shape their tax policy and legislation is also likely to be considerable. Pillar Two consists of a set of domestic rules (the so-called Global Anti-Base Erosion (GloBE) rules) and a complementary treaty-based rule. The GloBE rules have the status of a common approach. This means that members of the Inclusive Framework are not obliged to adopt the GloBE rules; should they choose to do so, however, they

86

For a discussion on the role of the OECD in developing international tax law, see A. Tychmanska, ‘The OECD as the Future International Tax Organization: An Inevitable Course of Events?’, Intertax 49 (2021), 614. 87 See the G20 Rome Leaders’ Declaration, https://www.g20.org/wp-content/uploads/2021/10/G20- ROME-LEADERS-DECLARATION.pdf. 88 See the statement at https://www.oecd.org/tax/beps/statement-on-a-two-pillar-solution-to-addr ess-the-tax-challenges-arising-from-the-digitalisation-of-the-economy-october-2021.pdf.

International Tax Law and Public International Law 195 would be required to implement and administer those rules according to the standards set by the Inclusive Framework. They will also be required to accept the application of the GloBE rules applied by other members of the Inclusive Framework. Undoubtedly, the agreement on the two-pillar solution marks a clear and major shift towards greater internationalization of tax law. Of course, much will depend on the detailed scope of the final proposals and whether they will be implemented in a uniform manner by countries around the world.

Chapter 12

Internationa l Tax L aw and C orp orat e L aw Marcos André Vinhas Catão and Verônica Souza

12.1 Introduction Public international law plays an important role in international taxation as it recognizes states’ original jurisdiction and sovereignty to tax.1 Cross-border transactions are deeply affected by public international law, as it determines the delimitation of states’ territories, international waters, airspace and outer space, the international effects of nationality, and other matters.2 The jurisdictional power of a country to tax the income derived from international transactions is first established by customary principles of international law. Under these principles, the exercise of the jurisdictional power to tax may be based on different elements, which include nationally, domicile or residence, and presence, or doing business within the country.3 Such elements are recognized limitations on the power to tax; and yet the most relevant effect of public international law on international taxation is the application of tax conventions or treaties.4 International tax treaties are the most important source of international tax law. They restrict fiscal sovereignty but do not ignore that each contracting state is able to apply its own law even though such treaties do establish limits to the application of the

1 K.

Vogel, Klaus Vogel on Double Taxation Conventions, 3rd ed. (Boston, MA: Kluwer Law International, 2013), 16. 2 J. Schwarz, Schwarz on Tax Treaties, 3rd ed. (Kingston-upon-Thames: Wolters Kluwer, 2013), 1. 3 C. H. Gustafson et al., Taxation of International Transactions: Materials, Text and Problems, 4th ed. (Minnesota: West Academic, 2010), 1060. 4 The terms ‘tax treaty’ or ‘treaty’ are used in this chapter, except in reference to the OECD Model Convention.

198 Marcos André Vinhas Catão and Verônica Souza contracting states’ domestic laws.5 However, in private international law, tax treaties do not have to choose between domestic and foreign law. Corporate law refers to the body of laws that govern relations between companies, organizations, and businesses. While tax law and international tax law are matters of public law, corporate law or company law form part of private law. This is the first difference between these fields. Private international law is that which governs private individuals or enterprises of different countries where there is a conflict between the laws of two or more countries.6 That said, no treaty can prevail over domestic law provisions. There is, of course, interplay between international tax law and corporate law—and they affect one other in many practical aspects. The management of a corporation will take taxes into consideration, for example when it undertakes a corporate reorganization, whereas the tax administration in a country will tackle corporate structures that have been created for the sole purpose of reducing or avoiding taxation. This chapter reviews the practical aspects of the interaction between corporate and tax laws, looks at the tax outcomes of international corporate events, and analyses how international tax law and the domestic laws of particular jurisdictions treat cross-border transactions. Section 12.2 addresses the tax aspects of the two ways of financing a company—equity and debt—as well as the impact of thin capitalization and earnings-stripping rules on shareholders’ decisions. Section 12.3 focuses on the relevant tax issues of corporate distributions, the tax treatment given to cross-border dividends payments by different jurisdictions, and the solution of conflict by the application of a tax treaty. Section 12.4 analyses the tax implications of international corporate reorganizations in respect of capital gains and exit taxes. Section 12.5 deals with types of entities and their liability for tax with a focus on fiscally transparent entities. Section 12.6 provides some final remarks.

12.2 Corporate Financing: Tax Aspects of Equity and Debt When a corporation is formed and the capital stock is subscribed and paid in cash, the cash amount received from the shareholders will be accounted for as paid-in capital. In general, no taxation is due at that point. However, the event is relevant for tax purposes because it defines a shareholder’s tax acquisition cost, which will be used to calculate future capital gains if there is an alienation or any other transaction related to the shareholder’s investment. 5 Vogel, Klaus Vogel on Double Taxation Conventions, 16. 6 Schwarz, Schwarz on Tax Treaties, 6.

International Tax Law and Corporate Law 199 Any type of asset that is susceptible to monetary assessment can be the object of a capital contribution. Depending on whether the asset appreciates or depreciates, the transfer may generate a gain or loss to the shareholder. In some tax systems, taxation is required at that point, while in others deferment of tax is permitted.7 Capital contribution can also be made through debt in the form of loans, bonds, or other obligations. As a rule, interest on debts is taxed in the hands of the shareholder but is deductible against the invested corporation’s income tax basis. The decision to invest in a company through debt or equity depends on many economic aspects, including the expectation of earnings, the age of the company’s assets, the need for capitalization in the short or medium term, and the risk that the investor is willing to take. Tax implications are also highly relevant as, clearly, different implications derive from investments made through equity or debt. In respect of debt, a country’s tax rules often seek to limit the deductibility of interest and prevent the unrestricted distribution of interest. In addition, if the invested company and the investor are resident in different countries, the domestic law of both jurisdictions needs to be analysed, as a low tax burden or tax benefit granted in one country may be neutralized by higher taxation in another. Depending on the tax system in question, the equity–debt classification may or may not originate in corporate law. For example, in the USA the distinction between debt and equity comes from case law and the courts do not consider themselves bound by commercial law when it comes to the characterization of an investment.8 On the other hand, civil law countries traditionally base their considerations on commercial law and apply special rules to certain types of stocks.9 The UK, despite being a common law country, follows commercial law to distinguish between debt and equity and has complex statutory rules covering specific situations. French tax legislation characterizing debt and equity investments is also based on commercial law.10 Debt and equity classification is, then, crucial for the application of a tax treaty or for the domestic laws of the countries where the entities are established. For example, credit and exemption methods to avoid double taxation may require an investor to meet certain conditions regarding the percentage of their participation or a fixed debt–equity ratio. This is the case in most participation-exempt regimes and with thin capitalization rules. Therefore, investors should consider the existence of thresholds and requirements when financing a company in a specific jurisdiction, either through equity or with debt. Although the difference in treatment of debt and equity by jurisdictions may cause economic distortions,11 it is well known that the level of debt in a multinational 7 H.

J. Ault et al., Comparative Income Taxation: A Structural Analysis, 4th ed. (Alphen aan den Rijn: Wolters Kluwer, 2020), 742. 8 Ibid., 743. 9 Ibid., 744. 10 Ibid., 745. 11 R. Wood, ‘The Taxation of Debt, Equity, and Hybrid Arrangements’, Canadian Tax Journal 47/1 (1999), 49.

200 Marcos André Vinhas Catão and Verônica Souza enterprise can be used to erode its income tax base by the use of excessive deductions for interest payments. In this context, a corporation is considered thinly capitalized if the amount of debt is considerably more than the value of its capital. Since payments of interest can be used to reduce the income tax basis of an invested company, many countries impose restrictions on the deductibility of such payments. Thin capitalization rules traditionally use a debt– equity ratio to limit the amount of debt on which interest is deductible. However, the ratio approach then became the subject of Action 4 of the Action Plan on Base Erosion and Profit Shifting (BEPS Action Plan, OECD, 2013), ‘Limiting base erosion involving interest deductions and other financial payments’.12 In its report, the OECD recommended that the asset-based ratio should be replaced with a ratio which limited a company’s net deductions for interest, and payments economically equivalent to interest, to a percentage of its earnings before interest, taxes, depreciation, and amortization (EBITDA) in a range of 10–30%.13 According to Action 4, unlike the asset-based ratio, the EBITDA approach ‘ensures that an entity’s net interest deductions are directly linked to the taxable income generated by its economic activities’.14 EBITDA-based ratio rules are known as earnings- stripping rules. From a practical perspective, safe harbour rules are easy to implement and can be effective in tackling base erosion, depending on the percentage used to set the level of excessive interest payments. A common aspect of the asset-based and EBITDA-based ratios is that both consider a company’s debt level as a whole, and thus ignore the fact that a multinational enterprise may have different activities which justify different ratios. A negative aspect of earnings-stripping rules is that they can lead to economic double taxation when non-deductible interest is not requalified as a dividend because, unlike dividends, interest does not benefit from deductibility. Following BEPS, the EU 2016 Anti-Tax Avoidance Directive (ATAD) required member states to adopt interest-deduction restrictions based on safe harbours, regardless of whether the interest was paid to a resident or non-resident shareholder or corporation.15 Under the directive’s provisions, interest deduction is limited to a threshold of 30% of the taxable EBITDA or €3 million, if it is higher. In the light of this, some countries have supplemented their domestic laws to include the earnings-stripping rules in addition or in replacement of the traditional asset-based thin capitalization rules. In the USA, the 2017 tax reforms limited interest deductions to 30% of a taxpayer’s annual adjusted taxable income.16 The UK amended its domestic law and now has

12 OECD, Limiting Base Erosion Involving Interest Deductions and Other Financial Payments, Action 4—2016 Update (Paris: OECD Publishing, 2016), 15. 13 Ibid. 14 Ibid. 15 See Council Directive (EU) 2016/1164 of 12 July 2016. 16 Ault et al., Comparative Income Taxation, 1048.

International Tax Law and Corporate Law 201 earnings-stripping rules which follow BEPS Action 4 and ATAD.17 Germany also replaced its thin capitalization rules with earnings-stripping rules following ATAD which does not permit the deduction of interest expenses exceeding 30% of a company’s EBITDA.18 Hybrid financial instruments are also an alternative means for a company’s capital structure. In general, hybrid financial instruments combine features of equity and debt. Typical examples of such instruments are convertible bonds and preferred convertible securities, with the levels of debt and equity depending on the instrument, some having more debt-like characteristics and others being more equity-like. They can also differ in other dimensions, such as maturity, voting rights, and returns.19 The combination of debt and equity in a hybrid instrument gives rise to the possibility of tax avoidance by multinational groups which can benefit from deductive interest expenses in the source country, and dividend tax exemptions or reductions in the country of residence. The OECD chose to depart from the concept of hybrid instruments for both accounting rules and corporate law, and instead focused on arrangements which exploit the difference in tax treatments in two or more jurisdictions to artificially reduce a multinational’s global tax burden. The Action 2 Final Report contains recommendations to avoid such practices,20 a subject that was later addressed by ATAD, as amended by ATAD II,21 and then implemented by the EU member states. In the context of tax treaties, the effects hybrid instruments will be qualified as interest or dividends. Qualification will define the applicable withholding tax, in the country of source, and the tax credit or exemption, in the country of residence. Some treaties, especially those concluded after the BEPS Action 2 Final Report, define the nature of specific hybrid instruments, either in article 3 or in the Protocol,22 and the domestic law definition is not applied.

17

Ibid., 1051.

18 Ibid. 19 E.

Flores, ‘Are Hybrid Financial Instruments Debt or Equity? International Evidence’, PhD dissertation, São Paulo, University of São Paulo (2017), 8. 20 OECD, Neutralising the Effects of Hybrids Mismatch Arrangements, Action 2— 2015 Final Report (Paris: OECD Publishing, 2015). 21 Directive (EU) 2017/952. 22 This is the case with Brazil which has a unique legal hybrid instrument under which a corporation can remunerate its shareholders, which is called interest on equity. Interest on equity is a deductible expense for tax purposes and is regarded as a dividend for corporate purposes. Since interest on equity is deductible in the source country (Brazil), the resident countries now deny the application of a participation exemption to interest on equity, especially EU member states following ATAD II (Netherlands and Luxembourg are examples). The most recent tax treaties signed by the Brazilian government provide that interest on equity is treated as interest (and not as a dividend) to settle the qualification of the income.

202 Marcos André Vinhas Catão and Verônica Souza

12.3 Corporate Distributions: The Tax Treatment for Dividends As a rule, corporate law provides that payments to shareholders in respect of their investment in a corporation’s capital can be either by way of dividends or with a return on invested capital. The perception of dividends generally relates to the distribution of profits. Corporate law and accounting rules are significant in the characterization of such distributions as dividends, as they often limit the payment of dividends to the profits earned by a company. If the distribution exceeds earnings and profits, it is treated as a return on capital invested in the company and not as a dividend. From an international tax perspective, it is difficult to find define dividends as different jurisdictions have varying types of companies. The OECD recognized this in its Commentary on article 10 of the Model Convention,23 and highlighted that its definition aims to include examples found in the majority of the member states’ domestic laws. Since there are divergences between corporate law and tax law, it is not possible to arrive at a definition independent of the member states’ domestic laws. Having said that, dividends are defined by the OECD as: income from shares, ‘jouissance’ shares or ‘jouissance’ rights, mining shares, founders’ shares or other rights, not being debt-claims, participating in profits, as well as income from other corporate rights which is subjected to the same taxation treatment as income from shares by the laws of the State in which the company making the distribution is a resident.24

The word ‘company’ is defined in article 3(1)(b) of the Model Convention as ‘any body corporate or any entity that is treated as a body corporate for tax purposes’.25 However, the definition of a company will vary depending on the domestic law of a given country and on whether the company is domiciled in a civil or common law country. For instance, civil law countries distinguish between public and private companies, each with their own set of rules. Common law countries do not make that distinction, viewing the two types as basically the same, although different rules do apply.26 Also, whereas civil law public companies have shares, civil law private companies are similar to partnerships.27

23 OECD,

Model Tax Convention on Income and on Capital: Condensed Version (Paris: OECD Publishing, 2017), 231. 24 Ibid., 241. 25 P. Baker et al., ‘The Definition of Dividends and Interest in the OECD Model: Something Lost in Translation?’, World Tax Journal 1/1 (2009), 2. 26 Ibid., 3 27 Ibid.

International Tax Law and Corporate Law 203 This is relevant for the qualification of a company’s distribution. All tax systems have rules covering the tax applicable to distributions of profits, even if that means no taxation at all. In the UK, the concept of a distribution is crucial in determining its treatment for tax purposes. In general, if the distribution is classified as a dividend in corporate law, it will not be deductible by the corporation but will be subject to a lower tax rate for the receiving shareholder.28 In the Netherlands, a corporation can make a lawful distribution to its shareholders from surplus held by the corporation. Such distribution has the nature of an anticipation of profits in the future and the amount is then taxed as a dividend despite the lack of any current profits.29 China’s legislation adopts a broad concept of corporate distributions; dividends and other returns on equity are defined as ‘income derived by an enterprise from investees on account of equity investments’.30 Dividends distributed to a resident company are exempt from corporate income tax if the shares have been held for at least two months. The exemption does not apply to non-resident companies unless the company has a permanent establishment (PE) in China and it is limited to the dividends connected to the PE.31 In the USA, the tax consequences of distribution of profits also depend on whether the distributing corporation has had earnings and profits. Dividends paid to individuals have always been subject to normal income tax rates, but are currently taxed at the same preferential rates as capital gains.32 Dividends paid to corporate shareholders are taxed as regular income, but the dividends-received deduction tends to reduce the overall tax burden due to the relief granted to the corporation being based on the percentage of stock held in the distributing company.33 Brazil exempts from taxation the distribution of dividends to individuals or corporate shareholders.34 Distributions are limited to the earnings and profits of the company in the same tax year, and benefit from exemption; but the exemption does not apply to dividends from foreign investments that are received by companies domiciled in Brazil, in which event corporate income tax applies. Within the European Union, the Parent–Subsidiary Directive prevents double taxation of dividends distributed between member states’ companies.35 The directive

28

Ault et al., Comparative Income Taxation, 752.

29 Ibid. 30

See China, EIT Regulations, art. 17(4). Ault et al., Comparative Income Taxation, 754. 32 Ibid., 750. 33 Ibid. 34 Art. 10 of Law no. 9.249/1995. 35 See Council Directive 90/435/EEC, as amended by Council Directive 2003/123/EC and Council Directive (EU) 2015/121. 31

204 Marcos André Vinhas Catão and Verônica Souza requires member states to either exempt dividends from corporate income tax or permit a credit for the underlying tax due in another member state.36 Where the payment of dividends occurs at the international level, the effects for taxation will be governed the provisions of any tax treaty concluded between the relevant countries where the provisions cover the avoidance of any double taxation involved, whether by credit or exemption. Participation exemption, which relates to an exemption from taxation for a shareholder on dividends received, also plays a role in international corporate structures with respect to the distribution of dividends. Some jurisdictions apply the participation exemption mechanism to foreign income under their own domestic laws. A well-known example is the Dutch participation exemption system, under which dividends received from a foreign subsidiary by a shareholder with at least a 5% shareholding are exempt from corporate income tax. In a certain way, participation exemption regimes allow intra-group dividend relief for international structures to be extended. Many jurisdictions apply degrees of intercompany dividend relief based on the percentage of stock.37 The relief aims to avoid ‘cascading’ levels of taxation on corporate distributions.38 In other instances, this can be achieved by permitting a deduction of the dividend received or by the application of reduced rates.39 If the ultimate shareholder is domiciled in a jurisdiction where it can benefit from a participation exemption regime relating to foreign investments, then the domestic intra-group relief is extended to the international level. Distributions that exceed earnings and profits are often considered to be a return on capital and not a dividend. There is nothing contentious about returns on capital because, as a matter of principle, they should be tax-free as they represent a return on the shareholder’s tax cost in the company’s capital. Where the distribution exceeds the shareholder’s investment, income tax will be due on the capital gain. For cross-border transactions, even where a return on capital is equal to the shareholder’s tax cost, there may be capital gains on foreign exchange. A foreign exchange gain or loss commonly arises from the acquisition or disposal of an asset denominated in a foreign currency, which includes participations/shares/interests in an entity.

36 There are some requirements to be met by companies: to satisfy the definition of a parent company, it must be resident in the member state, subject to tax, and hold at least 10% of the shares in the subsidiary domiciled in the other member state. 37 This is the case of the USA, UK, and France, for example. 38 Ault et al., Comparative Income Taxation, 761. 39 Ibid., 429.

International Tax Law and Corporate Law 205

12.4 Tax Results from International Corporate Restructurings In this section, references to international corporate restructuring mean any cross- border transaction between companies that results in corporate changes or cross- border restructuring that results in changes in the economic structure of business.40 The term business restructuring generally refers to cross-border changes in a company’s operations, which may include realignment of ownership, assets, or capital structure to improve performance, synergy, and profit generation.41

12.4.1 Capital Gains Ideally, corporate restructuring should be tax-free. However, for many reasons, corporate restructuring may result in a taxable capital gain. Taxes on capital gains from corporate reorganizations vary from jurisdiction to jurisdiction. For example, in some OECD member countries, capital gains are taxed as ordinary income earned by an enterprise on the disposal of its assets, while other countries apply special rates in such cases.42 Many jurisdictions have tax-free rules for corporate restructuring or reorganization. A comparative analysis shows that, in some jurisdictions, there are detailed statutory rules which allow tax deferrals, roll-over of tax costs, and even exemptions, while in other jurisdictions, the tax-free result falls under broader general principles, such as continuity of business or ownership interests, neutrality of the tax law. Yet other jurisdictions use a combination of the two approaches. EU member states extend their ‘tax-free rules’ to corporate restructuring among the member states based on the freedom of establishment principle provided in article 49 of the Treaty on the Functioning of the EU. 40 The OECD Transfer Pricing Guidelines mention that there is no universally accepted definition of business restructuring:

There is no legal or universally definition of business restructuring. In the context of this chapter, business restructuring is defined as the cross-border redeployment a multinational enterprise by a multinational enterprise of functions, asserts and/or risks. A business restructuring may involve cross-border transfers of valuable intangibles, although this is not always the case. It may also or alternatively involve the termination or substantial renegotiation of existing arrangements. OECD, Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations (Paris: OECD Publishing, 2010), 235. 41 P. H. Blessing, Tax Planning for International Mergers, Acquisitions, Joint Ventures and Restructurings, 2nd ed. (The Hague: Kluwer Law International, 2014), 3. 42 OECD, Model Tax Convention, 293.

206 Marcos André Vinhas Catão and Verônica Souza In Germany, corporate reorganizations may be partially tax-free. Although the transfer of assets and shares should be carried out at fair market value, there are cases where the historic tax cost can be rolled over to the new shares received by a shareholder in an exchange of shares transaction, subject to certain limitations.43 In the UK, if the transaction qualifies as a reorganization for tax purposes, the historic tax cost can be rolled over to the new shares as a continuation of the original shares.44 On the other hand, if a target company is wound up as part of a merger due to corporate law provisions, then the target’s shareholders will be deemed to have received a capital distribution in respect of their shares, which will result in a capital gain for tax purposes.45 In the USA, formal statutory requirements must be met so that a corporate reorganization is qualified as tax-free. However, the rule covering capital gains on corporate restructuring is heavily guided by case law, and there are several judicial doctrines which limit the scope of such rules and subject the transaction to a business purpose test, even when the formal statutory requirements have been met.46 China’s legislation on the subject has specific reorganization rules that allow a partial or full deferral of the capital gains or a roll-over on the tax cost applied to certain tax regimes. For example, the main transaction’s purpose should not be the avoidance of tax, the business should be continued for a determined period, and at least 75% of the total assets or equity of the company should be transferred.47 In Latin America, Argentina’s legislation establishes, inter alia, that corporate restructuring may be tax-free if the target company has been operating for at least the previous eighteen months and the investment is to be held for at least two years after the transaction.48 In Brazil, the tax-free result of a corporate reorganization stems from a general rule that allows the transfer of assets or interests to be carried out at book value.49 Corporate reorganizations carried out at book value do not trigger the recognition of capital gains taxation at either corporate or shareholder level. Where the fair market value should be applied due to regulatory or any other specific provision, the capital gain is generally taxed. Brazilian legislation also has a specific rule for capital gains earned by non- residents which provides that, where an asset is located in Brazil, its sale, transfer, or any type of disposal is subject to income tax in the country.50 Such rule is applied even if the transaction involves only non-residents. In that event, the non-resident’s attorney collects the tax. 43

Ault et al., Comparative Income Taxation, 792. Ibid., 794. 45 Ibid. 46 Ibid., 788. 47 Ibid., 797. 48 See La Renta Financiera en el Impuesto a las Ganancias a partir de la Reforma Tributaria de la Ley no. 27.430, 2018, http://www.argentina.gob.ar. 49 Art. 22 of Law no. 9.249/1995. 50 Art. 26 of Law no. 10.833/2003. 44

International Tax Law and Corporate Law 207 The corporate reorganization of a multinational enterprise can generate a loss. This can occur if a company is dissolved, or other type of legal extinction, and the difference between the tax cost of the shares and the return on capital is negative. From an accounting perspective, the loss in investment will be reflected in the shareholder’s net equity and will incur a capital loss usually registered in the company profits and losses. If, in the same circumstances, a capital gain would have been taxed, then the capital loss should be deductible. However, it is usual for tax systems to restrict the deductibility of capital losses to a maximum percentage (or other type of threshold) to prevent taxpayers incurring losses while deferring capital gains.51 In general, capital losses are deductible only against capital gains.52 This brief glimpse of domestic legislations illustrates how conflicts between the tax rules of different countries may arise when a cross-border/international corporate restructuring takes place. The key question is how any resulting capital gains will be qualified in the context of cross-border transactions. Tax systems should elect whether corporate reorganizations will be tax-free, partially taxed, or fully taxed, and then provide rules to qualify that revenue as capital gains, ordinary income, or even as dividends. For example, if there is a transfer of interests in a company, it is common sense for the revenue from such a transaction to be considered a gain, or a loss. However, if an asset is removed from a company, some tax systems consider that to be similar to a dividend distribution to a shareholder. The tax implications of this can differ, depending on the nature of the shareholder or whether any relief is available.53 Capital gains are generally referred to as non-recurring gains and/or gains unrelated to the company’s core business. Nonetheless, any difference in treatment for certain classes of capital gains may lead to statutory complexity and jurisprudential difficulty in attempting to delimit the classification of such gains.54 The qualification of capital gains is extremely important in the application of a tax treaty that aims to resolve conflicts in multinational corporate reorganizations, especially if there is no relevant definition in the treaty. As determined by article 3(2) of the OECD Model Convention, in that instance the parties should rely on domestic legislation. Article 13 does not provide a detailed definition of capital gains and the OECD Commentary recognizes that there is a potential for problems where there is no agreed definition of capital gains between the contracting states.55 Under article 10 of the Model Convention, we have seen that dividends are generally referred to as a distribution of profits and income derived from shares. However, the broad definition contained in article 10 can create problems, for example where 51 Controlled

Foreign Corporation rules may not only restrict but prohibit the deductibility of the losses of a foreign subsidiary. This happens in Brazil where art. 25, para. 5 of Law no. 9.249/1995 provides that losses derived from foreign subsidiaries are not deductible for the Brazilian parent company. 52 This is the case in the USA, the UK, and Australia. See Ault et al., Comparative Income Taxation, 546–554. 53 Ibid., 786. 54 Ibid., 545. 55 OECD, Model Tax Convention, 295.

208 Marcos André Vinhas Catão and Verônica Souza dividends resulting from corporate restructuring are regarded as capital gains. In addition, the qualification of income in a cross-border corporate reorganization is relevant because dividends and capital gains may not receive the same relief under the participation exemption regime in the country of residence.

12.4.2 Exit Taxes Exit taxes can also result from corporate reorganizations. Where the tax residence of a company is transferred to another country under a transfer of assets, transfer of seat, merger, or any other corporate event, exit tax liability is generally triggered on the unrealized capital gains on the company’s assets, which are deemed to be disposed of at market value. In intra-EU reorganizations, the case law of the Court of Justice of the European Union (CJEU) allows exit taxes to be moderated by the freedom of establishment; and any domestic law considered incompatible with EU law must be amended. Exit tax rules are also influenced by the Merger Directive56 and the ATAD. The Merger Directive exit tax rule prohibits exit taxes on the transfer of seat of a Societas Europaea (SE) resulting from a merger, division/partial division, transfer of assets, or transfer of registered office. The ATAD, following the CJEU’s case law,57 requires member states to impose exit taxes on the value of transferred assets in specified circumstances where there is potential tax evasion. The ATAD represents a shift in regulation based on the case law on taxes on unrealized gains.58 The directive now allows exit tax to be paid in instalments and deferral for up to five years.59 In the UK, where a company ceases to be resident in a country, it is deemed to have disposed of all its assets for capital gains purposes and to have immediately acquired them at market value before ceasing to be resident. In the USA, such reorganization is treated as an outbound transfer and is not subject to tax-free treatment. The gains on the transfer of assets and on the share exchange by the shareholders are deemed to be realized and, hence, taxable events.60 China does not have explicit exit tax rules for companies that transfer residence out of a country, although deregistration requires tax clearance.61 Similarly, Brazil does not have exit rules for a company ceasing to be resident in a country. 56

Council Directive 2009/133/EC. See Case C-371/10 National Grid Indus, 29 November 2011 and Case C- 164/ 12 DMC, 23 January 2014. 58 As in Case C-9/02 Lasteyrie du Saillant, where the Commission decided that different treatments for cross-border transfers within the EU and domestic transfers are, in principle, an unjustifiable infringement on freedom of establishment. The infringement can be justified if the levy can be deferred until the time the capital gains are actually realized (and not deemed realized). 59 See Council Directive 2016/1164 of 12 July 2016, art. 5. 60 Ault et al., Comparative Income Taxation, 910. 61 Ibid. 57

International Tax Law and Corporate Law 209 Due to the high burden of exit taxes, in some cases it may be less onerous to liquidate a company in the country where it is intended to cease residency and reincorporate it in another. With regard to tax-efficiency, liquidation triggers corporate taxes on undistributed profits, and evaluation of stock investments, but they are subject to the relevant tax treaties and participation exemption regimes, whereas exit taxes are not.

12.5 Types of Entities and Liability for Tax Business organizations are created under and defined by corporate law. The tax system of a country stipulates which entities will be subject to corporate tax. It is common for tax systems to have a general rule outlining the relevant characteristics and requirements. Based on such rules, an entity may be liable for taxes. In that respect, it is common for the rules to contain a general description of the necessary characteristics and requirements rather than having an exclusive statutory list,62 since the latter would certainly need to be constantly updated. In the USA, all entities incorporated under domestic corporate law are regarded as corporations for federal tax purposes.63 The USA also has the well-known ‘check the box’ mechanism, under which other types of entities—such as partnerships and limited liability companies—can elect to receive the same tax treatment as corporations. The French tax code enumerates types of organizations subject to corporate tax regardless of their activity, but also establishes a general provision under which all entities that have legal personality according to civil law and carry out profitable activities are subject to tax.64 The UK has a broad definition of company and taxes all types of incorporated bodies and unincorporated associations except for partnerships.65 In China, until the 1980s, corporations were associated with capitalism. State- owned, collective-owned, and privately owned organizations do not have the form of corporations due to the legacy of China’s command economy.66 Following economic reform, domestic enterprises and enterprises with foreign investment became subject to enterprise income tax.67 The Brazilian income tax code does not contain a list of entities subject to tax. Some of its articles, though, use the word ‘corporation’ to describe tax events, for which the 62

Germany uses an exclusive statutory list of entities subject to tax. These entities are basically those which are treated as legal entities for civil law and corporate law purposes (ibid., 730). 63 United States Treasury Regulations, s. 301.7701-2(b)(1)–(7). 64 Ault et al., Comparative Income Taxation, 731. 65 In the UK, Income Tax Act 2010, s. 1121 and Income Tax Act 2007, s. 992. 66 Ault et al., Comparative Income Taxation, 733. 67 Ibid.

210 Marcos André Vinhas Catão and Verônica Souza concept of corporate law is used (Law no. 6.404/1976). Brazil’s tax system does not have ‘pass-through’ entities or transparent entities which are taxed only at shareholder level. Many jurisdictions have special rules for the tax treatment of business activities in organizations which allow their income and losses to be passed directly to the organization’s members. Corporate law classifies such organizations as partnerships. The basic structure of a partnership varies from one country to another, but tax systems generally use the pass-through mechanism, under which the partnership and partners are regarded as separate subjects for tax purposes.68 In international taxation, it is important to determine whether a tax treaty is applicable to partnerships and other non-corporate entities. As mentioned in Section 12.3, the OECD’s definition of a company in the Model Convention is provided in article 3(1)(b) as ‘any body corporate or any entity that is treated as a body corporate for tax purposes’. On that definition, Avery Jones et al.69 point out that the OECD Commentary on article 3(1) implies that all bodies corporate are taxable entities, ignoring that some entities are tax transparent.70 Therefore, problems may arise where partnerships or other non-corporate entities are transparent for tax purposes in the state where they are managed or organized. In that event, article 10 cannot avoid double taxation of profit distributions.71 Conflicts exist where the source state and the state of residence, relying on their respective domestic laws, apply different articles of a tax treaty to non-corporate entities. The OECD’s Committee on Fiscal Affairs has in the past addressed the issue of the application of the Model Tax Convention to partnerships.72 The OECD then confirmed that partnerships should be considered a body of persons for the purposes of article 3, which was subsequently amended accordingly,73 and, in addition, that partnerships should be resident in one of the contracting states. When a person is resident in a certain contracting state, it means that the person is fully liable to tax in that state both for income produced in that state and for worldwide income. According to the Commentary on article 1 of the Model Convention, the concept of ‘fiscally transparent’ refers to situations where, under the domestic law of a contracting state, the income of the entity is not taxed at the level of the entity and, instead, the entity’s shareholders or quota holders are the ones who are taxed.74

68

Ibid., 862. Baker et al., ‘Definition of Dividends and Interest in the OECD Model’, 2. 70 In 1956, the 14th Committee of the Organisation for European Economic Co-operation (OEEC), which was in charge of the definition of company, considered that all legal entities were taxable, which was probably the case at the time (Baker et al., ‘Definition of Dividends and Interest in the OECD Model’). 71 Ibid., 2. 72 OECD, The Application of the OECD Model Tax Convention to Partnerships (Paris: OECD Publishing, 1999). 73 P. Baker, ‘The Application of the Convention to Partnerships, Trusts and Other, Non-Corporate Entities’, GITC Review 2/1 (2002), 2. 74 OECD, Model Tax Convention. 69

International Tax Law and Corporate Law 211 The OECD considered that where a partnership is treated by a state as wholly or partly fiscally transparent, it is not a resident of that state for the purposes of the convention.75 Where a non-corporate entity is granted full tax transparency, it is not regarded as being liable to tax (or if it is granted with partial transparency, it is not liable to tax for the extent of the income subject to such a regime). Tax transparency is, then, inconsistent with the notion of full tax liability based on worldwide taxation.76 Following the OECD BEPS Action 2 Report, modifications were made to the Model Convention’s articles and Commentaries concerning the entitlement to tax treaty benefits of tax transparent entities. The main modification is the new article 1(2),77 which provides that when an entity is treated as wholly or partially transparent by a contracting state, the income received through or by such entity will be considered as the income of a resident of the contracting state that considers the entity to be tax transparent, to the extent that the recipient is liable to tax in that state. In a nutshell, the article clarifies that the tax treaty’s provisions are applicable to a fiscally transparent entity whose partner(s) is (are) resident(s) in the state where the entity is fiscally transparent. Conversely, the benefit granted by the tax treaty should be denied if the income is received by a fiscally transparent entity whose partner(s) is (are) not resident(s) in the state which considers the entity fiscally transparent. Although the introduction of article 1(2) seems to be an objective solution, it tends to favour the state of residence over the source state.78 There are other types of entities which are non-corporate, such as trusts, limited liability companies, funds, and joint ventures. These entities can benefit from tax treaty provisions if they are qualified as bodies of persons and fiscally transparent in the states where they are organized and managed. In addition, if the partners (or members in a broader concept) are resident in the state which considers the non-corporate entity as fiscally transparent, then the tax treaty will be applicable to that entity in relation to received income.

75

The report conclusions are reflected in the Commentaries on art. 1 of the Model Convention. Parada, ‘Tax Treaty Entitlement and Fiscally Transparent Entities: Improvements or Unnecessary Complications?’, in J. Wheeler, ed., The Aftermath of BEPS (Amsterdam: IBFD Publishing, 2020), 67. 77 The provision states: 76 L.

For the purposes of this Convention, income derived by or through an entity or arrangement that is treated as wholly or partly fiscally transparent under the taxation law of either Contracting State shall be considered to be income of a resident of a Contracting State but only to the extent that the income is treated, for purposes of the taxation by that State, as income of a resident of that State. 78

Parada, ‘Tax Treaty Entitlement and Fiscally Transparent Entities’, 24.

212 Marcos André Vinhas Catão and Verônica Souza

12.6 Conclusion Based on the foregoing discussion, it is not possible to identify all practical aspects of the relationship between international tax law and corporate law. There is considerable interplay between them, as a state’s taxes and the organization of businesses are closely connected. Businesses are created in accordance with corporate law, and their related entities, acts, and transactions are all subject to taxation. From the time of a corporation’s formation, tax law applies to its activities, profits, and shareholders’ remuneration. Corporate income tax is the relevant tax, as it is levied on earnings, profits, and other types of income which are usual for corporations and their shareholders. It is in the world of multinational corporations where international tax law and corporate law meet. Conflicts in international tax law are resolved by the application of a tax treaty, whose main role is to restrict the fiscal sovereignty of the contracting states and allocate the tax rights to prevent both double taxation and tax avoidance. This chapter first pointed out the tax consequences in financing a corporation by equity or debt where the entity and the shareholder are resident in different countries. We then looked at taxation of corporate distributions through cross-border payments of dividends. International corporate reorganizations were explored, including qualifications for capital gains that are relevant when applying tax treaties to such reorganizations. The chapter concluded by looking at different types of organization, and conflict that stems from using fiscal transparent entities in multinational corporations. The conclusion drawn is that international tax law and corporate law will keep company as long as businesses are created. The structuring of a multinational corporation involves not only the aspects addressed here, but also other aspects such as a group’s economic health which is intrinsically linked to taxation. It is, then, the role of the tax administrations and international organizations to find solutions to conflicts which arise from the way corporations in different jurisdictions transact with each other.

Chapter 13

Internationa l Tax L aw and Internat i ona l Trade L aw Servaas Van Thiel

13.1 Introduction This chapter is based on a greatly simplified understanding of ‘international trade law’ as the law of the 1994 World Trade Organization (WTO), and of international tax law as the national and international rules on the basis of which states impose direct or indirect taxes on cross-border economic transactions.1 It places both disciplines in the ‘integration law’ perspective of removing obstacles to cross-border economic activity, and explores how states have collectively agreed to limit their international taxing powers through the multilateral WTO agreements2 applying to goods (General Agreement on Tariffs and Trade (GATT)), services (General Agreement on Trade in Services (GATS)), and subsidies (Agreement on Subsidies and Countervailing Duties (SCM)), including their powers to impose tariffs and similar duties (Section 13.2), indirect taxes (Section 13.3), direct taxes in relation to trade in goods (Section 13.4), and trade in services through foreign-owned branches and subsidiaries (Section 13.5). WTO agreements will be referenced by their acronyms and WTO/GATT cases by year, DS number, and shortened name.3 Conclusions are in Section 13.6. 1 This

chapter uses the legal (but economically imperfect) distinction between indirect taxes (e.g. tariffs, other charges, sales taxes, excise duties) and direct taxes (e.g. income taxes, wealth taxes, withholding taxes, capital gains taxes). It focuses on cross-border trade and establishment (not labour). 2 In 2000: DS108 US Foreign Sales Corporation, the AB held that: ‘A Member of the WTO may choose any kind of tax system it wishes—so long as, in so choosing, that Member applies that system in a way that is consistent with its WTO obligations’ (point 179). 3 All WTO agreements, in particular the General Agreement on Tariffs and Trade (GATT), the General Agreement on Trade in Services (GATS), and the Agreement on Subsidies and Countervailing

214 Servaas Van Thiel As a preliminary remark, it is noted that while international trade law traditionally focused on customs duties and indirect tax discrimination and international tax law on double taxation and direct tax discrimination, European law has addressed all tax obstacles, and it may thus serve as a source of inspiration on possible ways to further strengthen international trade law and solidify a global economic integration process based on a multilateral rules-based system. The author is well aware that this is not necessarily high on today’s political agenda, but he is nevertheless convinced that a strengthening of the substance and enforcement of multilaterally agreed rules and procedures is the necessary legal underpinning for a sustainable future.

13.2 Tariffs and Similar Duties GATT (1947 and 1994)4 uses a ‘standstill’ and a ‘rollback’ to gradually reduce tariffs (customs duties) and other duties on imports and exports (charges with equivalent effect). It first provides for members to bind maximum tariff levels and to list other duties and charges in their schedules of concessions (stand-still of art. II).5 It subsequently obliges members to apply tariffs and other duties on a most-favoured-nation basis (MFN obligation of art. I), and to apply other duties, neither in excess of the approximate cost of services rendered nor in a way constituting ‘an indirect protection to domestic products’ (art. VIII). Finally, GATT encourages members to agree to further tariff reductions in successive rounds of multilateral trade negotiations (rollback of art. XXVIII) such as the 1987–1994 Uruguay Round (which established the WTO). One weakness of GATT is that the standstill is incomplete and that it pushed the rollback towards successive rounds of negotiations so that a zero tariff level was never reached. The world is still negotiating on the tariff phase-out6 and relies in the meantime on the MFN obligation to at least prevent tariff discrimination of imports from different countries of origin. A second weakness is the ‘horizontal’ dispute-settlement system under which only members are allowed to appeal to ad hoc panels and a Measures (SCM) are available at: http://www.wto.org/english/docs_e/legal_e/legal_e.htm. All WTO cases are available at http://www.wto.org/english/tratop_e/dispu_e/dispu_status_e.htm. All GATT cases are available at http://www.wto.org/english/tratop_e/dispu_e/gt47ds_e.htm. 4 GATT 1947 was the leftover when consensus broke down over the Havana Charter for an International Trade Organisation, which addressed a much broader range of policies (1948 Final Act of the UN Conference on Trade and Employment, http://www.wto.org/english/docs_e/legal_e/havana_ e.pdf). GATT 1994 consists of GATT 47 and a number of protocols, decisions, and understandings. 5 Understanding on the Interpretation of Article II:1(b) of the General Agreement on Tariffs and Trade 1994, http://www.wto.org/english/docs_e/legal_e/07-2-1-b_e.htm (accessed 2 February 2021). All duties and charges were bound as of 15 April 1994. 6 The ongoing Doha Round seeks ‘to reduce, or as appropriate eliminate tariffs, including the reduction or elimination of tariff peaks, high tariffs, and tariff escalation’. WTO, 2016, Doha Ministerial Declaration, WT/MIN(01)/DEC/1 of 20 November 2001, para. 16, http://www.wto.org/english/thewto_ e/minist_e/min01_e/mindecl_e.htm#marketaccess (accessed 2 February 2021).

International Tax Law and International Trade Law 215 standing appellate body (AB) and the winner’s final enforcement option is to retaliate against an offending member, which provides large traders with a certain degree of impunity.7 Every case thus necessarily becomes a weighing of economic, political, and cost factors and, unsurprisingly, case law is limited, even though there are some rather interesting past decisions on the standstill,8 the MFN obligation,9 and the article VIII obligation to keep the level of charges commensurate with the approximate cost of the service rendered.10 Since 2018, there has unfortunately been an increase in the number of tit-for-tat cases concerning the GATT compatibility (standstill and MFN) of the ‘Trump Tariffs’11 and the respective countermeasures,12 and of tariffs imposed for political reasons (e.g. complaints by Ukraine against Russia and by Qatar against its neighbours),13 or tariff preferences withdrawn for political reasons (e.g. EU complaints against Russia).14 Following the EU example, the way forward for the WTO would be to agree a complete phase-out of all customs duties in a time-limited transitional period. In the longer term, the WTO should ideally strengthen the dispute-settlement mechanism, for example by allowing private parties to challenge customs duties (vertical dispute settlement) and by ensuring stronger enforcement, for example by means of an obligation on states to refund any charges collected in violation of their WTO commitments and to repair any damages that may have been caused by their WTO-incompatible action. When considering such a complete phase-out, however, special rules are needed for low-income countries for whom tariffs are a significant source of revenue and who may

7 In the 1995–2015 period, for instance, the USA, the EU, and Canada often failed to implement WTO decisions leaving their adversaries no choice but to ask for authorization to suspend concessions. R. Arie, ‘The Effectiveness of the WTO Dispute Settlement System: A Statistical Analysis’, 17. EUI Working Papers 2017-11, https://cadmus.eui.eu/bitstream/handle/1814/47045/LAW_2017_11.pdf?sequence=1 (accessed 24 February 2021). 8 1998: DS56 Argentina Textiles and Apparel; 2005: DS269/286 EU customs classification of chicken cuts; 2005: DS302 Dominican Republic Import and sale of Cigarettes; 2008: DS360 India Duties on Imports from the US; 2010: DS375/377 EU classification of IT products. See the WTO cases at http://www.wto.org/ english/tratop_e/dispu_e/dispu_status_e.htm and the GATT cases at http://www.wto.org/english/trato p_e/dispu_e/gt47ds_e.htm. 9 2004: DS246 EC Tariff Preferences. See also the WTO and GATT cases ibid. 10 1988: US Customs fee; 1998: DS 56 Argentina Textiles and Apparel. See also the WTO and GATT cases ibid. 11 Ongoing cases: DS544, DS547, DS548, DS550, DS551, DS552, DS554, DS556, DS564, DS587. See also the WTO and GATT ibid. 12 Ongoing cases: DS557, DS558, DS559, DS560, DS561, DS566 and DS585. But see also ongoing cases against India (DS518; DS582; DS584; DS585 DS588) and Russia (DS566). See also the WTO and GATT cases ibid. 13 See, e.g., ongoing complaints between Ukraine and Russia (DS532 and DS525) and ongoing complaints by Qatar against its neighbours (DS 526, DS527, DS528). See also the WTO and GATT cases ibid. 14 See, e.g., the 20 January 2022 EU request for consultations with the Russian Federation with regard to the termination of tariff-rate quotas on exports of wood products (DS608: Russian Federation— Measures Concerning the Exportation of Wood Products, https://www.wto.org/english/tratop_e/dispu_e/ cases_e/ds608_e.htm).

216 Servaas Van Thiel still need to protect their infant industries. They could replace import tariffs with an internal tax which should, in principle, increase revenue, because it would be imposed not only on imports but also on domestic supplies.15 However, empirical research has shown that, in the process of replacing easy-to-collect border taxes (tariffs) with harder-to-collect value-added taxes,16 especially low-income countries have suffered significant revenue losses.17 This experience may help explain why certain developing countries with limited tax collection resources resist undertaking binding tariff reduction commitments in their GATT schedules of concessions.18 Any proposals to phase out customs duties should therefore foresee the necessary transitional periods and technical assistance for low-income countries, and exceptions for least developed countries, to ensure that they will have the required capacity to collect the indirect taxes necessary to replace trade tax revenue.

13.3 Indirect Tax Discrimination and Excessive BTA To ensure trade neutrality of taxes, GATT prohibits (direct and) indirect tax discrimination against imports as well as (direct and) indirect tax subsidies for exports. On the import side, GATT prohibits a higher tax on imported than on like domestic goods (national treatment, art. III, 1 and 2) as well as discrimination between imports from different origin countries (MFN art. I). On the export side, it prohibits (direct and) indirect tax subsidies (art. 3 SCM and Illustrative List).19 Indirect tax discrimination and indirect tax subsidies for exports may arise in many different ways, and GATT has specific rules for border tax adjustments (BTA) applied

15 M.

Keen and J. Ligthart, ‘Coordinating Tariff Reductions and Domestic Tax Reform’, Journal of International Economics 56/2 (2001), 489–507, http://www.sciencedirect.com/science/article/pii/S00221 99601001234. 16 L. Seelkopf, H. Lierse, and C. Schmitt, ‘Trade Liberalization and the Global Expansion of Modern Taxes’, Review of International Political Economy 23/2 (2016), 208–231, https://doi.org/10.1080/09692 290.2015.1125937 or http://www.tandfonline.com/doi/full/10.1080/09692290.2015.1125937?scroll= top&needAccess=true (accessed 2 February 2021). 17 T. Baunsgaard and M. Keen, ‘Tax Revenue and (or?) Trade Liberalization’, Journal of Public Economics 94/ 9– 10 (Oct. 2010), 563– 577, http://www.sciencedirect.com/science/article/pii/S00472 72709001479 (accessed 2 February 2021); International Monetary Fund, ‘Dealing with the Revenue Consequences of Trade Reform’, Background Paper for Review of Fund Work on Trade, Prepared by the Fiscal Affairs Department In consultation with Other Departments. Approved by Teresa Ter-Minassian (15 Feb. 2005), http://www.imf.org/external/np/pp/eng/2005/021505.pdf (accessed 2 February 2021). 18 Whereas developed countries bound 99% of their tariffs in the Uruguay Round, developing countries bound around 73%. WTO at: http://www.wto.org/english/thewto_e/whatis_e/tif_e/agrm2_ e.htm. 19 Art. 3 SCM prohibits export subsidies and the Illustrative mentions both direct tax export subsidies (under e and f) and indirect tax export subsidies (under g and h).

International Tax Law and International Trade Law 217 in destination-type indirect taxes.20 Since excessive BTA would amount to, respectively, a customs duty on imports and an export subsidy, GATT allows BTA but prohibits excessive indirect tax BTA (i.e. a higher tax on imports than on like domestic products and a refund higher than the amount of indirect tax imposed on domestic supplies).21 GATT also prohibits BTA for direct taxes (art. XVI GATT and SCM) as discussed in Section 13.4. GATT rules on indirect taxes have given rise to litigation and academic discussion on the meaning of indirect tax discrimination and, more recently, on the GATT compatibility of CO2 BTA. The very meaning of the concept of indirect tax discrimination22 has been disputed because of the rather complex wording of article III GATT. The question was whether it simply prohibited any taxes on imports ‘in excess’ of taxes on like domestic products (art. III.2 first sentence), or whether it only prohibited ‘excess’ internal taxes that were imposed with a protectionist motive (reference to ‘so as to afford protection’ in art. III.2 second sentence and chapeau), thus allowing an excess tax on imports for other policy reasons. Both views found support in the pre-WTO case law, with one of the last GATT panels in 1994, Gas Guzzler, holding, rather explicitly, that the words ‘so as to afford protection’ suggested both aim and effect and that a national measure could be said to have the aim of affording protection if it was the declared domestic policy goal to change the competitive opportunities in favour of domestic products.23 In 1996, however, the newly created Advisory Body (AB) settled the issue in DS 8, 10, 11 Japan Alcoholic Beverages. It held first that if imported and domestic products were ‘like’ (depending on criteria such as physical characteristics, tariff classification, and end use), article III.2 first sentence prohibited any excess tax without de minimis exception (point 23). Secondly, it held that if imported and domestic products were not ‘like’, but directly competitive and substitutable, article III.2 second sentence prohibited a more than de minimis tax difference ‘so as to afford protection’ (points 26–30), whereby neither the subjective nor declared intent of the legislator nor the actual effects on the 20 Because

destination-type indirect taxes are imposed on domestic supplies and imports, while exports are exempt, they require BTA in the form of a tax on import and a refund on export. 21 The interpretative notes to arts III and XVI GATT clarify that non- excessive BTA within destination-type taxes are not GATT-incompatible. See also the Report of the Working Party of 2 December 1970, L/ 3464, BISD 18S/ 97- 109 and the Secretariat’s consolidated overview of existing practices in L/3389. See also SCM for illustrative list of subsidies. 22 For some general background, see J. Tudor, ‘Discriminatory Internal Taxation in the European Union: The Power of the European Court of Justice to Limit the Tax Sovereignty of Member States under Article 110 of the TFEU’, Willamette Journal of International Law and Dispute Resolution 23/1 (2015), 141– 186, http://www.jstor.org/stable/26210443 (accessed 3 February 2021). G. M. Grossman, H. Horn, and P. C. Mavroidis, ‘The Legal and Economic Principles of World Trade Law: National Treatment’, Research Institute of Industrial Economics. IFN Working Paper No. 917 (2012), http://www.ifn.se/wfiles/wp/ wp917.pdf; T. Cottier and M. Oesch, Direct and Indirect Discrimination in WTO Law and EU Law, Working Paper No. 2011/16 (Apr. 2011), Swiss National Center of Competence and Research, https://pdfs. semanticscholar.org/cec5/0872506cf2971aa2163d378698669f70b6e2.pdf. 23 1994: US—Taxes on Automobiles. See the WTO cases at http://www.wto.org/english/tratop_e/disp u_e/dispu_status_e.htm and the GATT cases at http://www.wto.org/english/tratop_e/dispu_e/gt47ds_ e.htm.

218 Servaas Van Thiel market were decisive, but rather the regulatory non-neutrality.24 This approach was dutifully followed in subsequent WTO case law,25 which clarified that ‘so as to afford protection’ could follow from objective elements such as the sheer size of the tax differential or the design and architecture of the tax (DS75, 84 Korea Alcohol point 150; DS87, 110 Chile Alcohol point 6.2) and that the actual situation on the market was less relevant than the regulatory non-neutrality (DS87, 110 Chile Alcohol point 6.7).26 The approach of the AB was criticized in the literature, most notably by Horn and Mavroidis, who argued that a much more sophisticated econometric analysis was needed of the question whether imported and domestic products were ‘like’, and that in order not to unduly restrict the sovereignty of members, the stated intent of the policy should be decisive to determine whether a national measure is meant to protect domestic products.27 We believe that the requirement of a sophisticated econometric analysis would unnecessarily complicate dispute settlement. After all, in the bulk of WTO cases (many concerning alcoholic beverages) the regulatory non-neutrality has been rather evident and panels came to straightforward conclusions that some imported and domestic spirits were ‘like’ while others were directly competitive and substitutable, without the need for a complicated analysis of the market. Also, a return to Gas Guzzler that discrimination requires the stated protectionist intent of the lawmaker, will empty the GATT national treatment provision of all its substance, because states could easily avoid WTO scrutiny by claiming a non-protectionist intent. In this respect, the sovereignty argument is also unconvincing, because the whole point of GATT is that members did agree to limit their sovereignty to eliminate indirect tax discrimination in international trade. This does not prevent them from applying destination-type indirect taxes and from pursuing genuine policy objectives as long as they choose non-discriminatory means. Even if discriminatory tax policies are unavoidable, they are possibly justifiable under article XX GATT. We agree that states have formulated this article in a narrow way (requiring legitimacy on limited grounds, necessity, and proportionality) but, talking about sovereignty, this is what they agreed in 1947 and reconfirmed in 1994. Therefore, rather than going back to Gas Guzzler, a possible way forward could be either for members to renegotiate article XX (but consensus is unlikely) or for the AB to interpret the article XX exceptions

24 Pre-WTO case law had already clarified that an analysis of the actual trading effects on the market was not necessary because art. III protects expectations on the competitive relationship between imports and domestic products. See, e.g., 1987: US Taxes on petroleum (L/6175 GATT DOC BISD 34S/136 at point 5.1.9) in which the USA argued that a higher tax on imports (to finance a fund from which clean-up operations could be paid) had insignificant trade effects. See also the WTO and GATT cases ibid. 25 See, e.g., 1998: DS54 Indonesia car tax; 1999: DS75, 84 Korea tax on alcohol; 2000: DS87, 110 Chile tax on alcohol; 2005: DS308 Mexico tax on soft drinks; 2010: DS371 Thai tax on cigarettes. See also the WTO and GATT cases ibid. 26 See also Ongoing DS537: Canada Measures Governing the Sale of Wine. See also the WTO and GATT cases ibid. 27 H. Horn and P. C. Mavroidis, ‘Still Hazy after All These Years: The Interpretation of National Treatment in the GATT/WTO Case-law on Tax Discrimination’, EJIL 15/1 (2004), 39–69.

International Tax Law and International Trade Law 219 extensively and teleologically in the light of the broader objectives of ‘sustainable development’ to which the preamble of the WTO agreement explicitly refers. That would be a dynamic forward-looking solution that would be much better than handing the discrimination decision back to WTO members, as Horn and Mavroidis seem to suggest. A more recent discussion concerned the GATT compatibility of CO2 BTA.28 Assuming that carbon taxes are indirect taxes,29 that question very much turns on whether an imported product from a country that does not price the CO2 emitted by its production process, is ‘like’ a domestic product that carries a carbon tax in its production price. If only ‘physical characteristics’ or ‘consumer preferences’ are decisive, a CO2 tax on imported products risks contravening article III GATT if the CO2 tax on the production process is the only difference. In that case, we would argue that a CO2 tax on imports would be justifiable under article XX GATT because the different treatment serves the legitimate objectives of protecting human, animal, and plant life or health (art. XX b) and of conserving the exhaustible natural resource of ‘clean air’ (art. XX g, already recognized by the AB).30 It would also be necessary (and proportional or least trade-restrictive) because there is a broad consensus that carbon pricing is the most effective way to reduce emissions which is crucial for the world to survive. Finally, it would neither constitute disguised discrimination nor a restriction on trade because, in fact, the ‘like’ national products also have a CO2 tax incorporated in their production cost. Alternatively, the imported product could be considered ‘unlike’ the domestic product because it was produced or processed without the carbon content being taxed. In that case, a BTA on import (equivalent to the carbon tax paid by domestic producers) would not constitute discrimination. There has been a long discussion in the literature on the relevance of process and production methods (PPMs) for the ‘likeness’ of products and the case law supports both views,31 but we firmly support taking account 28 For a review of the legal and economic literature, see A. Cosbey et al., ‘Developing Guidance for Implementing Border Carbon Adjustments: Lessons, Cautions, and Research Needs from the Literature’, Review of Environmental Economics and Policy, Association of Environmental and Resource Economists 13/1 (2009), 3–22. http://www.journals.uchicago.edu/doi/10.1093/reep/rey020; A. Pirlot, Environmental Border Tax Adjustments and International Trade Law: Fostering Environmental Protection, New Horizons in Environmental and Energy Law series (Cheltenham: Edward Elgar, 2017); J. Hillman, ‘Changing Climate for Carbon Taxes: Who Is Afraid of the WTO’, Georgetown Law Faculty Publications and Other Works (2013), https://scholarship.law.georgetown.edu/facpub/2030 (accessed 7 February 2021). W. H. Maruyama, ‘Climate Change and the WTO: Cap and Trade versus Carbon Tax?’, Journal of World Trade 45 (2011), 679–726. 29 See fn. 58 SCM which defines indirect taxes as including ‘all taxes other than direct taxes and import charges’, and direct taxes as ‘taxes on wages, profits, interests, rents, royalties and other forms of income, and taxes on the ownership of real property’. 30 1996: DS2 US Gasoline point 6.37; 1998: DS58 US Shrimp Turtle AB points 50 and 141; 2001: DS135 EC Asbestos point 115. See also the WTO cases at http://www.wto.org/english/tratop_e/dispu_e/dispu_s tatus_e.htm and the GATT cases at http://www.wto.org/english/tratop_e/dispu_e/gt47ds_e.htm. 31 1992: US Tuna I and 1994 US Tuna II, concerning a US ban on imports of tuna caught with nets that also trapped dolphins, held that art. III calls for a comparison of the treatment of imported tuna as a product with that of domestic tuna as a product and that ‘regulations governing the taking of dolphins incidental to the taking of tuna could not possibly affect tuna as a product’. In 2001 DS58: US Shrimp

220 Servaas Van Thiel of environmentally relevant PPMs in the likeness analysis, because this would allow WTO members to link trade and environmental policy in a meaningful way, under the scrutiny of the AB.32 The issue resurfaced when the European Commission followed up on its proposals for an overhaul of the EU Energy Tax Directive and the Council agreed the creation of a Europe-wide carbon border adjustment mechanism in March 2022.33

13.4 Direct Tax Discrimination and Trade in Goods: Trade- Related Investment Measures and Export Subsidies The relation between GATT and direct taxes34 has long been considered to be remote because direct taxes are on persons and not on ‘products’ and ‘imports’,35 but GATT 47

Turtle the AB held that PPM-based differential treatment could in principle be justified under art. XX (see also art. 21 Panel points 5.87–5.88, confirmed in AB point 153). See also the WTO and GATT cases ibid. 32 D. Sifonios, Environmental Process and Production Methods (PPMs) in WTO Law, EYIEL Monographs—Studies in European and International Economic Law (Cham: Springer, 2018). 33 See European Commission, ‘European Green Deal: What Role Can Taxation Play?’ (2019), https:// ec.europa.eu/taxation_customs/commission-priorities-2019-24/european-green-deal-what-role-can- taxation-play_en. See ‘Council agrees on the Carbon Border Adjustment Mechanism (CBAM)’, Council Press Release of 15 March 2022, https://www.consilium.europa.eu/en/press/press-releases/2022/03/15/car bon-border-adjustment-mechanism-cbam-council-agrees-its-negotiating-mandate/. 34 For some general background, see M. Daly, ‘Is the WTO a World Tax Organization? A Primer on WTO Rules for Tax Policymakers’, IMF Technical Notes and Manuals (2016), http://www.imf.org/exter nal/pubs/ft/tnm/2016/tnm1602.pdf. J. E. Farrell, ‘The Interface of International Trade Law and Taxation’, IBFD Doctoral Series No. 26 (2013); J. Herdin-Winter and I. Hofbauer, The Relevance of WTO Law for Tax Matters (Vienna: Linde, 2006); M. Daly, ‘The WTO and Direct Taxation’, Discussion Paper 9, WTO (2005), http://www.wto.org/english/res_e/booksp_e/discussion_papers9_e.pdf; M. Lang, J. Herdin, and I. Hofbauer, WTO and Direct Taxation (Vienna: Linde, 2005); W. Schön, ‘World Trade Organization Law and Tax Law’, Bulletin for International Taxation (2004), 283ff; P. R. McDaniel, ‘The David R. Tillinghast Lecture, Trade Agreements and Income Taxation: Interactions, Conflicts, and Resolutions’, Tax Law Review 57/2 (2004), 275 (in-depth review concluding that FSC/ETI decision was correct from legal and economic perspective), https://heinonline.org/HOL/Page?handle=hein.journals/taxlr57&div=14&g_s ent=1&casa_token=&collection=journals. 35 J. H. Jackson, World Trade and the Law of GATT (Indianapolis: Bobbs-Merrill, 1967), 297; K. Dam, The GATT Law and International Economic Organization (Chicago: Chicago University Press, 1970), 124 and 125; J. Fisher-Zernin, ‘GATT versus Tax Treaties? The Basic Conflicts between International Taxation Methods and the Rules and Concepts of GATT’, Journal of World Trade Law 1 (1987), 42. For the opposite view, see J. Slemrod and R. S. Avi-Yonah, ‘How Should Trade Agreements Deal with Income Tax Issues?’, Tax Law Review 4 (2002), 536. See also UN, Report on the discussions in Sub-Committee A of the Third committee of the Havana Conference, E?CONF.2/C.3?A?W.32 (1948), 1–2; 2001: DS 155 Argentina—exports of bovine hides and import of finished leather.

International Tax Law and International Trade Law 221 already prohibited the use of (direct and indirect) tax incentives to promote import substitution (art. III.4 and 1994 TRIMs) and exports (art. XVI and 1994 SCM).

13.4.1 Trade-Related Investment Measures As regards import substitution, article III:4 GATT prohibits regulatory discrimination against imports; that is, any less favourable treatment of imported goods compared to like domestic goods ‘in respect of all laws, regulations and requirements affecting their internal sale, offering for sale, purchase, transportation, distribution or use’. Early GATT practice recognized that income tax measures could be used to discriminate against imports. In 1952, for instance, Austria complained against an Italian remission of income tax (23 Lire per kg) to Italian shipyards that used domestically produced instead of imported plates.36 In 1987, the EC complained about discriminatory US tax credits and accelerated depreciation rules for small passenger aircraft assembled in the USA, but the panel was never established because the US measures were temporary.37 The issue resurfaced during the Uruguay Round of negotiations which resulted in the 1994 Agreement on Trade-Related Investment Measures (TRIMs Agreement) which phased out existing investment measures that stimulate the use of local rather than imported content or restrict the use of imported inputs in local production or exports. Though TRIMs did not explicitly refer to income tax incentives, the WTO Panel in the 2001 ETI case confirmed earlier GATT practice by holding in respect of the contested income tax exemption: ‘Thus nothing in the plain language of the provision specifically excludes requirements conditioning access to income tax measures from the scope of application of Article III, which deals with “national treatment on internal taxation and regulation” ’. It concluded that ‘by reason of the foreign articles/labour limitation the act accords less favourable treatment within the meaning of Article III:4 of the GATT 1994 to imported products than to like products of US origin’.38 There were a number of subsequent decisions in which tax advantages, conditioned on the use of local instead of imported content, were considered incompatible with

36 1952:

Italy ship plates, complaint by Austria against an Italian remission of income tax to firms that used domestically produced instead of imported ship plates (L/875 SR.13/12, SR.15/17, SR.16/9, SR.17/ 5, SR.18/4). A similar decision, which concerned a credit instead of a tax, is 1958: Italy agricultural machinery, in which the UK complained about Italian lower interest loans for purchasing domestically produced tractors (L/833 7S/60). See the WTO cases at http://www.wto.org/english/tratop_e/dispu_e/ dispu_status_e.htm and the GATT cases at http://www.wto.org/english/tratop_e/dispu_e/gt47ds_e.htm. 37 1987: United States—Tax Reform Legislation for Small Passenger Aircraft (L/6153). See also the WTO and GATT cases ibid. 38 2001: DS108 US Foreign Sales Corporations, art. 21.5 Panel Report, point 8.158. The Panel and AB thus decided that the USA treated imported products less favourably by limiting the property which could generate exempt income to ‘qualifying foreign trade property’ which again was defined as property not more than 50% of the fair market value of which was attributable to articles produced or direct labour performed outside the USA. See the WTO and GATT cases ibid.

222 Servaas Van Thiel article III.4 GATT,39 and in some of these cases the contested measures were also seen as export subsidies inconsistent with the SCM agreement.40 GATT thus prohibits any income tax incentives that promote the use of local content over imported goods and there is likely to be a steady future trickle of WTO dispute-settlement cases.41 It is interesting to note, finally, that local content requirements (i.e. the use of domestic instead of foreign services and services suppliers) may also be contrary to the article XVII GATS national treatment obligation.42

13.4.2 Export Subsidies Under article XVI GATT and articles 1 to 3 SCM, subsidies include any financial contribution by a government (including uncollected taxes)43 which confers a benefit44 and is specifically granted to certain as opposed to all enterprises or industries.45 The SCM prohibits export subsidies (‘contingent in fact or in law upon export performance’ or ‘upon the use of domestic over imported goods’)46 and has an illustrative list that

39 e.g.

2019: DS510 US Renewable Energy; 2017–18: DS472 Brazil Taxation (concerning indirect and direct tax incentives for the local production of cars and information technology products). See also the WTO and GATT cases ibid. 40 See, e.g., 2019: DS510 US Renewable Energy, and ongoing DS563: US Renewable Energy. See also the WTO and GATT cases ibid. Note that art. 3 SCM defines prohibited export subsidies as subsidies contingent upon export performance or subsidies contingent upon the use of domestic over imported goods, which overlaps with the definition of a TRIM. 41 Participants in the 2004 Rust Conference identified a range of possible TRIMs including deductions, depreciation, and tax credits available only in respect of domestic products, tax incentives for domestic cultural productions (e.g. films), and anti-avoidance limitations of deductions for costs made in low-tax jurisdictions. See S. van Thiel, ‘General Report’, in Lang, Herdin, and Hofbauer, WTO and Direct Taxation, 13–49. 42 In 1998–2000: DS 139, 142 Canada Automotive Industry, the Panel considered that an import duty exemption conditioned on a local content requirement (that could be filled by using locally produced goods or services) was inconsistent with the national treatment obligation of art. XVII GATS (but the AB reversed the Panel’s findings on the ground that it had not established that the contested measure ‘affected’ trade in services and was thus within the scope of GATS). See the WTO cases at http://www. wto.org/english/tratop_e/dispu_e/dispu_status_e.htm and the GATT cases at http://www.wto.org/engl ish/tratop_e/dispu_e/gt47ds_e.htm. 43 2002: DS108 US Foreign Sales Corporations, art. 21.5 AB Report at points 81–106. See also the WTO and GATT cases ibid. 44 Meaning that it places the beneficiary in a better position. 1999: DS70 Canada Export of civilian aircraft, AB Report points 149–161. See also the WTO and GATT cases ibid. 45 C. Micheau, State Aid, Subsidy and Tax Incentives under EU and WTO Law, Kluwer Series on International Taxation vol. 45 (Alphen aan den Rijn: Wolters Kluwer, 2014) ; R. Luja, Assessment and Recovery of Tax Incentives in the EC and the WTO: A View on State Aids, Trade Subsidies and Direct Taxation (Antwerp: Intersentia, 2003). 46 Other subsidies that have ‘adverse effects’ are ‘actionable’ both at national level (the affected state may impose countervailing duties) and at WTO level (dispute settlement). See the 2020 Commission proposal to protect the internal market against the completion distorting effects of foreign subsidies including those that facilitate the acquisition of EU undertakings, COM(2020) 253 final (17 June 2020),

International Tax Law and International Trade Law 223 explicitly mentions export BTA in the form of (indirect47 or) direct tax incentives, such as export-related exemptions, remissions, deferrals, and deductions of direct taxes or social welfare charges.48 BTA for direct taxes featured prominently in the decade-long transatlantic dispute over US corporate income tax exemptions for export earnings (cases on domestic international sales corporations or DISC, foreign sales corporations or FSC, and extraterritorial income or ETI).49 The USA argued that European countries enjoyed a competitive advantage because they relied more heavily on indirect taxes (BTA for exports) and on the exemption method of avoiding double taxation (foreign-source income exempt from direct tax). To restore the balance, it allowed US industries to channel their exports through a separate (US or foreign) company which would not pay income tax on part of their export earnings. European countries claimed, and panels and the AB repeatedly agreed, that the US direct tax exemption constituted a GATT and SCM-incompatible export subsidy.50 As part of this transatlantic dispute, the US filed counterclaims in 1976 against several EU member states, arguing that the exemption method of avoiding double taxation constituted an illegal export subsidy.51 Even though, surprisingly, the panels agreed with the USA, this was later overturned by a GATT Council Understanding,52 which was subsequently incorporated in a clarification in footnote 59 of the SCM, that the prohibition of export subsidies does not limit the use of the credit or the exemption method to avoid double taxation (it being understood that members applying the exemption

White Paper on levelling the playing field as regards foreign subsidies, https://ec.europa.eu/competition/ international/overview/foreign_subsidies_white_paper.pdf. 47 Though not further discussed in this chapter, the SCM also prohibits indirect tax incentives for exports. See the WTO Analytical Index on the SCM at: https://www.wto.org/english/res_e/publications_ e/ai17_e/subsidies_annii_oth.pdf. 48 Fn. 58 SCM defines direct taxes as taxes on wages, profits, interests, rents, royalties, and all other forms of income, and taxes on the ownership of real property. This differs from the definition used in the GATS. 49 For an overview, see D. L. Brumbaugh, ‘A History of the Extraterritorial Income (ETI) and Foreign Sales Corporation (FSC) Export Tax–Benefit Controversy’, Congressional Research Service Report for Congress (2006), http://www.everycrsreport.com/reports/RL31660.html (accessed 22 February 2021). 50 1976: US Tax Legislation (DISC) Panel (L/ 4422 BISD 23S/98). 2001–2: DS 108 US Foreign Sales Corporations, Panel and Appellate Body. See the WTO cases at http://www.wto.org/english/tratop_e/ dispu_e/dispu_status_e.htm and the GATT cases at http://www.wto.org/english/tratop_e/dispu_e/gt47d s_e.htm. 51 1976: Income Tax Practices Maintained by Belgium (L/442-23S/127), France (L/442-23S/114), and Netherlands (L/442-23S/137). In 1998, the USA also filed counterclaims against five EU member states but these disputes remained at the level of consultations without panels being established. 1998: DS131 France, DS130 Ireland, DS129 Greece, DS128 Netherlands, and DS127 Belgium. See also the WTO and GATT cases ibid. 52 In a 1981 Understanding, the GATT Council agreed that territorial systems of income taxation which apply the exemption method were not GATT-incompatible. 1982 BISD 28S/114.

224 Servaas Van Thiel method should ensure that related parties are taxed in accordance with the arm’s-length principle).53 An interesting sequel to this transatlantic discussion was the 2016 US Tax Blueprint proposal for a destination-based cash flow tax (DBCFT).54 It was considered by many a WTO-incompatible export subsidy,55 but not by all,56 with the nuanced view of Wolfgang Schön that a fundamental tax reform with a switch from a residence/source- based tax to a sales-based tax does not itself constitute a subsidy (because it is not a deviation from but a wholesale change of the ‘benchmark’ tax system), but that the BTA of a DBCFT would be contrary to the GATT prohibition of tax discrimination and export subsides.57 Other high-profile cases on direct tax subsidies include various claims of illegal subsidies to the aircraft industry (pitting the EU and the USA as well as Brazil and Canada against each other).58

53 A combination of the exemption method and a weak enforcement of arm’s-length transfer pricing rules may result in de facto export subsidization because companies could by means of non-arm’s-length transfer prices transfer income to low-tax jurisdictions and ensure that this ‘foreign source’ income remains untaxed at home because it is covered by the exemption method. 54 The proposal was to replace corporate income tax with a destination- based cash flow tax, i.e. imposed on net receipts (i.e. receipts from domestic sales and imports of goods and supplies of services in the jurisdiction (but not from exported products and services supplied outside the jurisdiction)) minus real costs, including labour costs. For the perceived advantages, see ‘A Better Way, Our Vision for a Confident America: Tax’ (24 June 2016), http://abetterway.speaker.gov/_assets/pdf/ABetterWay-Tax- PolicyPaper.pdf. 55 See C. McLure and W. Hellerstein, ‘Does Sales- Only Apportionment of Corporate Income Violate the GATT?’, NBER Working Paper 9060 (2002) (see also: 3 CES/ifo Forum, 2002, issue 4 at 3); N. Shaxson, ‘Ten Reasons Why the Destination Based Cash Flow Tax Is A Terrible Idea’, , Tax Justice Network Blog, 2019, http://www.taxjustice.net/2019/03/19/ten-reasons-why-the-destination-based- cash-flow-tax-is-a-terrible-idea/; R. S. Avi-Yonah and K. Clausing, ‘Problems with Destination-Based Corporate Taxes and the Ryan Blueprint’, Columbia Journal of Tax Law 8/2 (2017), 229–255; C. Wei, ‘Destination-Based Cash-Flow Taxation: A Critical Appraisal’, University of Toronto Law Journal 67/ 3 (2017), 301; J. Hillman, ‘Why the Ryan–Brady Tax Proposal Will Be Found to Be Inconsistent with WTO Law’, Institute of International Economic Law, Issue Brief no. 3 (2017); R. S. Avi-Yonah, ‘Back to 1913?: The Ryan Blueprint and Its Problems’, University of Michigan Law and Economics, Working Paper No. 16-026 (2016), https://ssrn.com/abstract=2871801; J. Becker and J. Englisch, ‘A European Perspective on the US Plans for a Destination Based Cash Flow Tax’, Oxford University Centre for Business Taxation, Working Paper 17/03 (2017), http://eureka.sbs.ox.ac.uk/7294//WP1703.pdf. 56 See A. Pirlot, ‘Don’t Blame It on WTO Law: An Analysis of the Alleged WTO Law Incompatibility of Destination-Based Taxes’, Florida Tax Review 23/1 (2019), 432–494, https://heinonline.org/HOL/ TextGenerator?handle=hein.journals/ftaxr23&collection=journals§ion=10&id=432&print=sect ion§ioncount=1&ext=.txt. Memorandum from Timothy Keeler and Warren Payne, Mayer Brown LLP (on behalf of Caterpillar, Inc.) to Kevin Brady, Chairman, House Committee on Ways & Means (22 May 2017), reprinted in Tax Notes Doc. 2017-46486. 57 W. Schön, ‘Destination-Based Income Taxation and WTO Law: A Note’. Max Planck Institute for Tax Law and Public Finance Working Paper 2016-03, http://www.journals.uchicago.edu/doi/10.1093/ reep/rey020. 58 See 2011–20: DS316, 317, 353 US Large Civil Aircraft, and 2010–19: DS316 EC Large Civil Aircraft. 2009–10: DS70, DS71 Canada Export Civil Aircraft and ongoing DS522 Canada Trade in Commercial Aircraft (Brazil complained about Canadian support to Bombardier including a municipal property tax exemption, a Federal R&D investment tax credit, and Quebec’s refundable tax credits for

International Tax Law and International Trade Law 225 In summary, (both indirect and) direct tax incentives may qualify as illegal import substitution subsidies (if conditional on using local over imported content) or export subsidies (if conditioned on export performance), and it is irrelevant whether the advantage is granted by means of reduced rates, exemptions, deductions, or credits (but foreign tax credits or foreign income exemptions to avoid double taxation are allowed).

13.5 Direct Tax Discrimination and Trade in Services: Taxation of Foreign- Owned Branches and Subsidiaries The GATS is a framework agreement with some general obligations (including MFN art. II) and the possibility for members to negotiate further liberalization commitments (art. XIX) on market access (art. XVI) and national treatment (art. XVII).59 GATS potentially has a very wide scope of application because it applies to ‘measures’ ‘affecting’ ‘trade’ in ‘services’ (art. I) whereby: • ‘services’ include any service in any service sector (except public services, public procurement, and air traffic services);60 • ‘international trade in services’ includes cross- border supply and consumption (modes 1 and 2) and cross-border establishment (mode 3), and the supply of services includes the production, distribution, marketing, sale, and delivery of services (art. XXVIIIb); • ‘measures’ include all laws, regulations, rules, procedures, decisions, administrative actions, etc. including direct tax measures as defined in article XXVIII;61 • the term ‘affect’ services has the broader sense of having ‘an effect on services’, and not the narrower sense of specifically seeking to regulate or govern trade in services.62 pre-competitive research). See also the WTO cases at http://www.wto.org/english/tratop_e/dispu_e/dis pu_status_e.htm and the GATT cases at http://www.wto.org/english/tratop_e/dispu_e/gt47ds_e.htm. 59

This part builds on previous work of the author. See S. van Thiel, ‘The General Agreement on Trade in Services and Income Taxation’, in K. Byttebier, ed., WTO Obligations and Opportunities: Challenges of Implementation (London: Cameron May, 2007), ch. 15, 385ff. 60 Art. I:3 (b) as widely interpreted by the AB in 1997, DS27 EC bananas point 220: ‘We also note that Article XXVIII (b) of the GATS provides that the supply of a services includes the production, distribution, marketing, sale and delivery of a service. There is nothing at all in these provisions to suggest a limited scope of application for GATS.’ 61 Art. XXVIII(o) GATS defines ‘direct taxes’ as comprising all taxes on total income, on total capital, or on elements of income or of capital, including taxes on gains from the alienation of property, taxes on estates, inheritances, and gifts, and taxes on the total amount of wages or salaries paid by enterprises, as well as taxes on capital appreciation (see different definition in the SCM). 62 2000: DS139 Canada Automotive, AB Report point 155. 1997: DS27 EC Bananas AB Report point 220. 2015: DS453 Argentina, Panel Report points 7.99ff. See also the WTO cases at http://www.wto.org/

226 Servaas Van Thiel GATS thus, in principle, applies to the tax treatment of a foreign-owned subsidiary, branch, or representative office through which a foreign service supplier delivers services in the host or destination member, but its impact on direct taxes is nevertheless limited by the GATS itself in two main ways. First, the MFN obligation (art. II) is subject to general exceptions, including for ‘bilaterally agreed provisions for the avoidance of double taxation’ (art. XIVe).63 This means that WTO members would in principle be allowed to continue to apply their different tax treaty-based measures to avoid double taxation to investors from different home countries, including different withholding taxes on outgoing investment income flows (because the source state withholding tax combined with the residence state credit allocates tax jurisdiction to avoid double taxation).64 It remains to be tested, however, to what extent the MFN obligation affects other tax rules such as domestic anti-avoidance measures (e.g. controlled foreign corporations or limits on the deductibility of expenses) directed only at certain low-tax or blacklisted jurisdictions,65 and tax treaty clauses that are not covered by the carve-out (e.g. because of providing substantive benefits).66 Interesting, in this respect, is the recent literature on the WTO compatibility of the OECD base erosion and profit shifting (BEPS)measures.67 One question, for instance, concerns the national treatment compatibility of the BEPS Action 5 rule that taxpayers should only benefit from intellectual property regimes if they

english/tratop_e/dispu_e/dispu_status_e.htm and the GATT cases at http://www.wto.org/english/trato p_e/dispu_e/gt47ds_e.htm. 63

Art. XIV introduction and sub (e) together read:

Subject to the requirement that such measures are not applied in a manner which would constitute a means of arbitrary or unjustifiable discrimination between countries where like conditions prevail, or a disguised restriction on the trade in services, nothing in this agreement shall be construed to prevent the adoption or enforcement by any Member of measures (e) inconsistent with Article II, provided that the different treatment is the result of an agreement on the avoidance of double taxation or provisions on the avoidance of double taxation in any other international agreement or arrangement by which the Member is bound. 64

In Case C-336/96 Gilly [1998] ECR I-2793, the ECJ explicitly held that in tax treaty rules the allocation of tax jurisdiction cannot be considered to constitute discrimination, not even if based on the criterion of nationality. 65 See, e.g., J. B. Gross, ‘OECD Defensive Measures against Harmful Tax— Competition Legality under WTO’, Intertax 31/11 (2003), 390–400, https://heinonline.org/HOL/LandingPage?handle=hein.klu wer/intrtax0031&div=76&id=&page=; B. Chilala and R. Grynberg, ‘WTO Compatibility of the OECD “Defensive Measures” against Harmful Tax Competition’ (2001), https://heinonline.org/HOL/LandingP age?handle=hein.journals/jworldit2&div=28&id=&page. 66 For an overview of the arguments in the discussion on an MFN obligation under EU law, see S. van Thiel, Free Movement of Persons and Income Tax Law: The European Court in Search of Principles (Amsterdam: International Bureau of Fiscal Documentation, 2002). 67 See, e.g., P. Schoueri, Conflicts of International Legal Frameworks in the Area of Harmful Tax Competition (Amsterdam: WU Institute for Austrian and International Tax Law—Tax Law and Policy Series, IBFD, 2019); W. Zhang, ‘When Global Tax Reform Meets International Trade Rules: An Inquiry into the Intersection of the GATS and the BEPS Package’, CTEI Working Paper CTEI-2016-5 (10 June 2016), https://ssrn.com/abstract=2946820 or http://dx.doi.org/10.2139/ssrn.2946820.

International Tax Law and International Trade Law 227 incurred the qualifying R&D expense themselves locally, which could be considered a WTO-incompatible local content requirement. Another question concerns the rule that (to avoid double non-taxation in case of mismatches) countries should deny a deductible expense if not taxed in the hands of the recipient, as well as the controlled foreign corporations rule that must be applied to companies established in countries with markedly lower effective tax rates. Both seem incompatible with the MFN obligation and, to the extent that they affect trade in goods and services, would need to fit the narrowly formulated exceptions of GATT article XXd and GATS article XIVd (‘necessary to secure compliance with laws or regulations which are not inconsistent with the provisions of this Agreement’68 without constituting arbitrary discrimination or a disguised restriction on trade). Relevant is the 2015 Panel Report in DS453, which concluded that Argentine anti- tax haven legislation, which had different tax basis rules for payments to and from ‘cooperative countries’ and ‘non cooperative countries’ (without exchange of information), was contrary to the GATS MFN obligation. That finding was reversed by the AB but on a different ground (i.e. because the Panel had not established whether services and service suppliers of ‘non cooperative countries’ and ‘cooperative countries’ were ‘like’).69 Secondly, the GATS national treatment obligation (art. XVII) applies only if members have undertaken specific commitments in their schedules, and provided the measure does not fall either within the scope of a tax treaty70 or within the scope of the horizontal carve-out (art. XIV sub d) which allows members to apply different treatment aimed at ensuring the equitable or effective imposition or collection of direct taxes in respect of services or service suppliers of other members. This carve-out includes tax measures, which distinguish between residents and non-residents, ensure the imposition or collection of taxes, prevent the avoidance or evasion of taxes, or apportion income between related persons,71 but these are not allowed if they would constitute arbitrary discrimination or a disguised restriction on trade. 68 GATS clarifies that these include measures relating to: (1) the prevention of deceptive and fraudulent practices or to deal with the effects of a default on services contracts; (2) the protection of the privacy of individuals in relation to the processing and dissemination of personal data and the protection of confidentiality of individual records and accounts; (3) safety. 69 2015: DS453 Argentina. See the WTO cases at http://www.wto.org/english/tratop_e/dispu_e/dispu_ status_e.htm and the GATT cases at http://www.wto.org/english/tratop_e/dispu_e/gt47ds_e.htm. 70 Art. XXII 3 GATS provides that:

A Member may not invoke Article XVII, either under this Article or Article XXIII, with respect to a measure of another Member that falls within the scope of an international agreement between them relating to the avoidance of double taxation. In case of disagreement between Members as to whether a measure falls within the scope of such an agreement between them, it shall be open to either Member to bring this matter before the Council for Trade in Services. The Council shall refer the matter to arbitration. The decision of the arbitrator shall be final and binding on the Members. 71 Fn. 6 to GATS (art. XIV sub d) lists these measures and further specifies that the tax terms used in art. XIV para. d and in the footnote are determined according to tax definitions and concepts, or equivalent or similar definitions and concepts, under the domestic law of the member taking the measure.

228 Servaas Van Thiel The GATS National Treatment clause (art. XVII) could, in principle, prohibit host country income tax discrimination against foreign services suppliers doing business through a local branch or a subsidiary, once members have made specific market access commitments allowing inbound establishment.72 A very interesting question in that case is whether future WTO case law will go in the direction of article 24 of the OECD Model Convention or in the direction of EU case law. If the latter, several elements of the OECD Model Commentary on article 24, for instance concerning non-discrimination of permanent establishments,73 could be WTO- incompatible, and foreign- owned branches could claim equal treatment as domestic enterprises in respect of participation exemptions and credits (irrespective of whether available under national law or tax treaties),74 reduced corporate income tax rates on certain types of income,75 host country measures to avoid economic double taxation of dividends (lower split rates of corporate income tax to avoid economic double taxation of dividends76 and host country imputation credits),77 and procedural rules (e.g. interest payments in the case of a late refund of overpaid tax).78

13.6 Conclusions The two disciplines of international tax law and international trade law have long developed in almost splendid isolation from each other, one in Paris focusing on direct taxes and seeking to develop an OECD Model Double Taxation Convention, and one in Geneva focusing on indirect taxes and seeking to eliminate tariffs and indirect tax discrimination. However, in 1995, international trade law made a giant step with the establishment of the WTO which was ratified by 164 states. This contribution illustrated how, by becoming WTO members, states agreed to multiple, and sometimes unexpected, limits on the way in which they exercise their tax jurisdiction, going well beyond the traditional tariff ceilings. For instance, WTO members reconfirmed a broad prohibition of indirect tax discrimination (GATT MFN and NT) which may well affect their right to introduce new indirect taxes, such as carbon taxes with BTA and destination-based

72

See van Thiel, ‘General Report’, 13–49. 33–72 of the Commentary (https://read.oecd-ilibrary.org/taxation/model-tax-convention- on-income-and-on-capital-2017-full-version_9e0b6e02-en#page1). 74 Case C-251/98 C. Baars v. Inspecteur der Belastingen Particulieren/ Ondernemingen Gorinchem, ECLI:EU:C:2000:205; Case C-307/97 Compagnie de Saint- Gobain, Zweigniederlassung Deutschland v. Finanzamt Aachen-Innenstadt, ECLI:EU:C:1999:110. 75 Case C-311/97 Royal Bank of Scotland v. Elliniko Dimosio, ECLI:EU:C:1999:216. 76 Case C-253/03 CLT-UFA SA v. Finanzamt Köln-West, ECLI:EU:C:2005:227. 77 Case 270/83 Commission v. French Republic (Avoir Fiscal), ECLI:EU:C:1986:37. 78 Case C-330/91 The Queen v. Inland Revenue Commissioners, ex parte Commerzbank AG, ECLI:EU:C:1993:303. 73 Points

International Tax Law and International Trade Law 229 cash flow taxes. Moreover, through GATT, TRIMs, and SCM, WTO members also accepted a prohibition of indirect and direct tax incentives contingent on the use of local over imported content, as well as of export subsidies, including BTA for direct taxes. Furthermore, they accepted a GATS MFN obligation, which may undermine their efforts to combat tax avoidance and evasion (including measures envisaged under the OECD BEPS framework). Finally, WTO members accepted a GATS national treatment obligation which, once they undertake specific market access commitments, may prohibit direct tax discrimination against foreign-owned branches and subsidiaries, in a much broader way than envisaged in article 24 of the OECD Model Convention. No doubt, finance ministries all over the world will keep invoking the concept of ‘tax sovereignty’ to resist any limits on the traditional ways in which they tax cross-border transactions and activities. However, as this Handbook illustrates, they should better realize that their freedom to design national tax systems is already limited by many other areas of international law, and they should accept that, because, in a globalized world, states pool their sovereignty to set up rules based global governance systems for the common good.

Chapter 14

Internationa l Tax L aw and Ec onomi c Analysis of L aw Werner Haslehner

14.1 Introduction The economic analysis of law is generally concerned with questions of an optimal allocation of resources through the legal system, also referred to as economic efficiency.1 To that end, it investigates the effects of legal rules on human behaviour, taking as a foundational assumption the reactions of a rational welfare-maximizing individual in a world of scarce resources. In this world, any legal rule has an incentive effect on each such individual, leading to collective outcomes that follow logically from interaction of their individual responses to such incentives. But economic analysis of law in general has often not concerned itself with international tax questions. For example, Richard Posner’s seminal book Economic Analysis of Law2 contains a chapter on taxation, but does not touch on cross-border questions. Similarly, foundational contributions on the ‘economics of tax law’ are liable to omit references to international tax questions.3 Why is that?4 As much of the economic analysis of law is focused on individual choices, the issue of the international allocation of taxes will strike many law and economics scholars as a question outside the scope of 1

See R. A. Posner, Economic Analysis of Law, 8th ed. (New York: Aspen Publishers, 2011), ch. 1. Ibid., ch. 17. 3 This is true even if the contribution is written by most prolific thinkers on questions of international tax law. See, e.g., D. N. Shaviro, ‘The Economics of Tax Law’, in F. Parisi, ed., The Oxford Handbook of Law and Economics, vol. 3 (Oxford: Oxford University Press, 2017), 106–126. 4 This phenomenon is not unique to international tax law, but for international law more generally. For an analysis of the dearth of scholarship on international law and economics, see J. L. Dunoff and J. P. Trachtman, ‘Economic Analysis of International Law’, Yale Journal of International Law 24/1 (1999), 1–55. 2

232 Werner Haslehner the study, as it may appear to be primarily concerned with states’ actions and the division of tax revenue. However, this would overlook the impact that both the existence of overlapping tax jurisdictions, the risk of double taxation, and the form of relief from it have on the behaviour of taxpayers. Furthermore, where the traditional law and economics framework is limited to analysis of the efficiency impacts within one jurisdiction, the problem of international taxation appears inherently inter-jurisdictional. Yet such a view would not only ignore the fact that much of what constitutes international tax norms is in fact firmly rooted in domestic legislation rather than treaties,5 it would also miss the potential of economic analysis of the choices of countries in the interest of the overarching search for globally efficient tax rules. This latter aspect has traditionally been at the core of the economic analysis of international tax law, leading to what has been called an ‘alphabet soup’6 of acronyms such as CEN, CIN, NN, CON. Even in this latter respect, the law and economics scholarship on the question of optimal international tax rules has not been able to find definitive answers. This is, in part, due to a lack of clarity as to what benchmark to apply for defining an ‘optimum’: in contrast to the literature on trade law and economics, where the optimum for one country conveniently achieves the optimum for all countries, national welfare and worldwide welfare are decidedly different when the question concerns the allocation of tax revenue.7 If the primary concern for international tax policy is to decide how to coordinate tax claims by different countries, economic analysis is ill-equipped to provide a helpful result, since the question is, in essence, distributional.8 Despite this lack of clear outcomes, the application of an economic perspective to the problems of international taxation remains undoubtedly valuable for a proper understanding of the issues at stake. This chapter aims to address four core questions of international tax law that lend themselves to the application of this perspective: the definition of double taxation and its impact; the (lack of) economic foundations for the allocation of (source) taxing rights among states; the use of relief mechanisms and the meaning of ‘neutralities’ in international tax law; and the problem of international tax competition.

14.2 Economic Analysis of Double Taxation The core problem of international taxation is regularly identified in double taxation. Although it is frequently framed in purely legal terms as a matter of interpersonal justice 5

See Y. Brauner, ‘An International Tax Regime in Crystallization’, Tax Law Review 56/2 (2003), 259. D. N. Shaviro, Fixing US International Taxation (Oxford: Oxford University Press, 2014), 14. 7 See A. Raskolnikov, ‘Accepting the Limits of Tax Law and Economics’, Cornell Law Review 98/3 (2013), 523 at 565. 8 See M. A. Kane, ‘A Defense of Source Rules in International Taxation’, Yale Journal on Regulation 32/ 2 (2015), 322–326. 6

International Tax Law and Economic Analysis of Law 233 that is based on the—frequently unstated—assumption that cross-border economic activity is normatively equal to domestic economic activity and thus should be treated the same, it can also be (possibly more coherently) analysed in economic terms, as a problem of efficient resource allocation. Clearly, double taxation is not the only concern of international tax law, and not the only element to be analysed from an economic perspective. The distribution of taxing rights and its effect on the final tax burden faced by a cross-border investment is similarly relevant. Indeed, the relevance of double taxation can easily be overstated: a rational taxpayer will prefer paying taxes twice at less than half the ordinary tax rate applicable to domestic income to paying tax once at the ordinary rate.9 Double taxation can impose a prohibitively high (transaction) cost on an otherwise profitable investment. Under the standard assumption that resource allocation would be efficient in the absence of the distortions introduced by taxation (i.e. assuming the absence of market failure), the result is that capital will not be located where it would be employed most productively. While that is an effect of taxation in general, it would appear inevitably stronger if double taxation were not avoided.10 As a consequence, not only would worldwide output be lower than in a scenario without double taxation, but also each taxpayer and each individual country would likely be worse off. After this first approximative analysis of the importance of double taxation from an economic perspective, it is necessary to take a more precise look: what exactly is meant by ‘double taxation’, and is the standard definition as it is applied in domestic and treaty law economically coherent? As commonly used in tax policy circles, international (juridical) double taxation is defined as the imposition of taxes by two states on the income earned by one taxpayer.11 It is thus distinguished, on the one hand, from the situation in which only one state by itself imposes two tax charges on the income of only one 9

See Shaviro, Fixing US International Taxation, 5. Note that as the extent of the distortion depends on the cumulative marginal tax rate faced by an investment rather than the number of instances in which a tax is applied on it, this is true under the assumption of uniform tax rates applied to domestic and foreign income, although it may not be correct where several instances of taxation are combined with lower effective tax rates to keep the marginal burden of alternative investments in different locations the same—either through lower nominal rates applied to foreign income or relief mechanisms that achieve the same result. 11 See OECD Model Convention (OECD MC), Commentary, Introduction para. 1, which adds the element of those taxes relating to the same time period. Whether this can really be considered to be a constitutive element of a normative understanding of double taxation is subject to doubt, however. On the one hand, double taxation conventions typically do not address the timing question in the relevant provision to relieve double taxation (i.e. art. 23 of the OECD MC). In practice, relief may not be granted in cases of mismatching recognition of income among contracting states. On the other hand, the lack of explicit attention to such scenarios does not rule out bringing them within the scope of application of such a tax treaty (see further J. Wheeler, ‘Double Tax Relief and Time’, in W. Haslehner et al., eds, Time and Tax (Alphen aan den Rijn: Kluwer Law International, 2018), 37–41, 27). Instead, object and purpose of tax treaties strongly suggest including cases where the same investment return is merely recognized in different periods in the contracting states. After all, tax treaties leave many key elements undefined, including the term ‘income’, which inevitably must be circumscribed in the domestic law of each contracting state. 10

234 Werner Haslehner taxpayer and, on the other, from the situation in which two tax charges are imposed on the income earned by two separate taxpayers (whether or not this involves two states). The former situation of multiple tax charges within one jurisdiction is rather obviously not considered an issue of international taxation, while the latter—so-called economic double taxation—is also recognized as a concern in the international sphere that is not, however, tackled consistently in bilateral agreements in a way akin to the concern of juridical double taxation. This curious circumstance is probably best explained by the fact that economic double taxation is not exclusively a problem that arises in cross-border situations. But is there any difference between the two from an economic perspective? The answer must be in the negative. Taking an economic point of view here means to look at the real-world effects of taxation on decision making by economic actors. So- called economic double taxation most commonly arises as a consequence of the separate entity principle applied to corporate taxation, which builds on the recognition of a legal entity as the taxable unit. It occurs either because income may be attributed to different entities by two states, or because states characterize a cross-border transaction in a way that leads to different inclusion or deduction of elements of income or capital, or simply because income is taxed when earned by one entity and again when distributed to that entity’s owners.12 Substantively, however, the legal entity is not the unit at which decision making takes place. More appropriately, the legal entity can be viewed as the vehicle at whose level a decision is being legally operationalized (although laws on corporate governance also contribute to the decision-making process by imposing a certain structure, at least theoretically). In the absence of taxes and the risk of bankruptcy, the legal form would be immaterial for business decisions. Furthermore, the effects of any tax on any agent’s behaviour ultimately depends on whether the tax burdens that agent; it is a question of tax incidence.13 Yet whether a tax economically falls on the shoulders of a particular person does not depend on whether the law intends it to be so, but reflects the preferences of the decision-making units involved in the transaction that gives rise to taxation. Given the difficulty of separating juridical and economic double taxation from an economic point of view, the continued importance of the distinction for applying the rules of international taxation is surprising and challenging. Double-taxation conventions primarily aim to address the overlap of tax jurisdiction relating to one and the same taxpayer: built into the fabric of the norms delineating taxing jurisdiction to avoid double taxation is a duality of residence and source that results in one particular type of such overlap (without effectively preventing all its instances).14 Economic double 12 K.

Vogel and W. Haslehner, ‘Introduction’, in E. Reimer and A. Rust, eds, Klaus Vogel on Double Taxation Conventions, 5th ed. (Alphen aan den Rijn: Kluwer Law International, 2021), 14. 13 For foundational explanations of the concept and its implications, see R. A. Musgrave and P. B. Musgrave, Public Finance in Theory and Practice, 5th ed. (New York: McGraw-Hill,1989), 236; A. B. Atkinson and J. E. Stiglitz, Lectures on Public Economics (Princeton, NJ: Princeton University Press, 2015), 132; J. A. Mirrlees, ed., Tax by Design (Oxford: Oxford University Press, 2011), 27. 14 See, e.g., OECD Commentary on arts 23A and 23B, n. 1 confining the application of the article to instances of juridical double taxation.

International Tax Law and Economic Analysis of Law 235 taxation, by contrast, is tackled only very sparingly with respect to transactional profit adjustments under article 915 and classification mismatches in article 1(2) of the OECD Model Convention. To the extent that economists consider the imposition of several tax charges on a given stream of income to be problematic (i.e. harming real-world outcomes in terms of measurable economic output from a certain perspective), the distinction between juridical and economic double taxation is more than a mere distraction. It is actively harmful in that it introduces an additional distortion, since it treats investments that are economically equivalent but structured differently with regard to their legal form in a different way, creating a bias towards the applicability of tax treaties that provide coveted relief from double taxation. An even more fundamental problem exists, however, with the standard definition of international double taxation. Despite its wide usage in both tax scholarship and tax policy circles, it is generally accepted that ‘double taxation’ is not a very clearly circumscribed idea with concrete boundaries as to when it is fully, partially, or not at all present.16 Various different taxation outcomes from both from the taxpayer’s and the state’s perspective can be described as avoiding double taxation—most notably both the exemption of foreign income and its deductibility from the domestically owed income tax are credited (as it were) with preventing double taxation completely, while granting a deduction from the domestic tax base is not.17 The different consideration given to the credit and the deduction method with regard to their prevention of double taxation neatly illustrates the lack of coherent economic rationale behind the term: if a resident earns foreign income that is subject to tax in the source jurisdiction at a lower rate than the residence jurisdiction, and the applicable rules lead to the application of the credit method, the taxpayer will still be taxed twice on the same income—literally falling within the standard definition of international juridical double taxation—since they will have to pay the difference in the foreign and the domestic tax rate on foreign income to the residence state.18 Despite this fact, it is not considered double taxation because the overall tax burden facing the taxpayer is no higher than would be in the case if they earned the same income in the residence state. The standard definition of double taxation is thus incoherent insofar as it appears to be overinclusive in capturing this rather typical situation. In reaction, several scholars have correctly identified the need to add a quantitative element to an operational legal definition of double taxation, which would link it to an increased tax burden as a consequence of the tax imposition by two states.19 From an economic perspective, it is not only reasonable to consider a type of dual taxation that does not increase the overall tax burden compared to the situation of taxation 15

See OECD Commentary on art. 9, n. 5. e.g., M. Lehner, ‘Grundlagen des Abkommensrechts’, in K. Vogel and M. Lehner, eds, Doppelbesteuerungsabkommen Kommentar, 7th ed. (Munich, C. H. Beck, 2021), 135; see also R. Ismer and J. Ruß, ‘What Is International Double Taxation?’, Intertax 48/6/7 (2020), 555. 17 Ismer and Ruß, ‘What Is International Double Taxation?’, 557; see further on the economic effects of different relief methods and their link to different ‘neutralities’, Section 14.4. 18 See Shaviro, Fixing US International Taxation, 5. 19 See most recently Ismer and Ruß, ‘What Is International Double Taxation?’, 557–558. 16 See,

236 Werner Haslehner in a single state to be not harmful to economic activity; it may, in fact, be rather desirable to have a form of residual taxation in the residence state in addition to the tax raised by the source state despite it resulting in two parallel charges.20 Thus, a combination of the deduction method and a lower tax rate applied by the residence state on foreign income in such a way that the overall tax burden does not exceed the tax imposed in the case of domestic income would be excluded from the scope of an economically more coherent notion of ‘double taxation’, as it would be equally unproblematic from the perspective of a disincentivization of cross-border investment.21 Otherwise, economically equivalent or even preferable outcomes may be subject to legal challenges on the basis of an overly narrow treaty definition of double taxation that imposes an obligation to provide relief beyond the economic optimum. The importance of a coherent definition of double taxation—a definition which, as is argued here, should be based on an economic rather than a semantic analysis—has further increased with the introduction of anti-abuse rules in tax treaties. Especially general anti-abuse rules characterized by a principal purpose test–general anti-avoidance rule (PPT–GAAR) as included by the Multilateral Instrument (MLI)22 reinforce an understanding according to which tax relief through a treaty is only to be granted where this is in line with the object and purpose of the treaty provisions.23 As that object is probably identified primarily in the prevention of double taxation as commonly understood, it has thereby further entrenched the need to identify an object and purpose that is ill- defined both legally and economically. It has done so by creating a teleological link between the goal of relieving double taxation and the notion of achieving single taxation, not least by including this objective in the new preamble to double tax treaties.24 And yet, mirroring the confusion over the meaning of double taxation, the concrete content of the single taxation or full taxation norms25 remains unclear along the margins that matter for the efficiency analysis (i.e. the location and rate of taxation). 20

See further Section 14.4. the concrete positive efficiency-based arguments in favour of such an approach, see Shaviro, ‘Fixing U.S. International Taxation’, 145–150. 22 Multilateral Convention to implement tax treaty related measures to prevent base erosion and profit shifting, https://www.oecd.org/tax/treaties/multilateral-convention-to-implement-tax-treaty-rela ted-measures-to-prevent-BEPS.pdf (accessed 2 May 2021). 23 Note that anti-abuse provisions lend themselves to an economic analysis, whereby the uncertainty introduced by a vague provision deliberately imposes a cost on taxpayers who have an incentive to minimize their tax burden. But this analysis is not unique to international taxation, nor even taxation in general. See Posner, Economic Analysis of Law, 665, who broadens the issue to the analysis of ‘standards vs rules’. See also D. A. Weisbach, ‘An Economic Analysis of Anti-Avoidance Doctrines’, American Law and Economics Review 4 /1 (2002), 88. 24 See D. G. Duff, ‘Tax Treaty Abuse and the Principal Purpose Test—Part 2’, Canadian Tax Journal 66/4 (2018), 947. 25 Mason argues that single taxation, as a principle, encompasses the norm against double taxation as well as a norm in favour of ‘full taxation’, which ‘dictates that all of a company’s income should be taxed in places where it has real business activities’ (R. Mason, ‘The Transformation of International Tax’, American Journal of International Law 114/3 (2020), 353, 370. As such, the concept is agnostic about both the concrete place and rate of tax that ought to be imposed, focusing primarily on the number of jurisdictions legitimately imposing a tax. For criticism of the notion as a coherent normative concept, 21 For

International Tax Law and Economic Analysis of Law 237

14.3 Economic Foundations of Source Taxation Rights Does the allocation of taxing rights through the various distributive rules found in bilateral tax treaties have an economic rationale? To be more precise, taking as a given the rationality of taxing residents on their domestic income, how should one think about taxing non-residents on domestic income and residents on their foreign income? The latter question is addressed more fully in the next section. Here the focus is on the question under what conditions it is efficient for a country to tax the income of a non-resident. Practical, political, and legal considerations are at the forefront of this question: taxing non-residents on their foreign-sourced income would be highly ‘efficient’ from the taxing country’s perspective insofar as it would not impose any cost on its own economy, but is practically infeasible. Public international law arguably also indicates a requirement for a ‘genuine link’ between a country and a stream of income for the latter to be subjected to a tax charge.26 This condition, together with the practical concerns surrounding the enforcement of tax jurisdiction over foreigners, has led to the development of the substantial basis to recognize taxing rights ‘at source’. The historical and theoretical legal foundations of this principle are discussed at length in Part I of this volume and shall not be repeated here. It shall suffice to note that taxation at source as it is sanctioned in the vast majority of currently existing tax treaties does not actually follow a coherent economic rationale. Could such coherent rationale be found, in particular, in economic theory? The foundational work on allocation of taxing rights completed by four economists on behalf of the League of Nations in their famous 1923 report attempted to find such a basis.27 The underlying idea taken as the starting point to allocate taxing rights has been that of a taxpayer’s economic allegiance, which in turn can be divided into four factors: ‘the acquisition of wealth, the location of wealth, the enforceability of the rights to wealth and the consumption of wealth’,28 which in turn can be translated into four locations where taxes may properly be imposed. These are place of origin of wealth, place of its situs, place of the enforcement of rights, and place of residence.29 Of these, the main biggest see L. Parada, ‘Full Taxation: The Single Tax Emperor’s New Clothes’, Florida Tax Review 24/2 (2021), 729–783. 26 See J. Kokott, ‘The “Genuine Link” Requirement for Source Taxation in Public International Law’, in W. Haslehner et al., eds, Tax and the Digital Economy (Alphen aan den Rijn: Kluwer Law International, 2018), 9–23. 27 See the 1923 League of Nations Report on Double Taxation, as described and discussed in E. R. A. Seligman, Double Taxation and International Fiscal Cooperation (New York: Macmillan, 1928). 28 1923 League of Nations Report, 23; for a detailed review of the report, see S. Jogarajan, ‘Stamp, Seligman and the Drafting of the 1923 Experts’ Report on Double Taxation’, World Tax Journal 5/3 (2013), 368–392. 29 Ibid.

238 Werner Haslehner concern—which coincides with the familiar dichotomy of source-versus residence- based taxation—is the identification of the ‘origin of wealth’, which in modern-day (post-base erosion and profit shifting, BEPS) parlance can be equated with the identification of the place of ‘value creation’.30 Unfortunately, despite its easy appeal as a foundational principle for a fair global allocation of taxing rights, the concept does not lend itself to precise content that is economically meaningful. The most fundamental reason for this lies in the fact that—as was already clear in 1923—the place of creation of value through cross-border activity is inevitably split between different locations where the various functions that combine for such activity are carried out,31 and the contribution of each element cannot be reliably quantified, since a dependable measure of value can only be assigned at the end of the production (value) chain. Yet even if the various inputs could be quantified32 and the resulting profit split in the proportion of those inputs, it would not—as the recent debate over allocating taxing rights to ‘market jurisdictions’ in the context of the digitalization of business operations has shown—satisfy all countries’ taxing claims based on the idea of ‘value creation’ for the simple reason that they would argue—correctly—that the emergence of any ‘value’ inevitably requires more than a producer: it also requires a buyer, and a market where the transaction can take place.33 Confining the analysis to a production-based view, the idea of value creation may have some use in that it can help fill the notion of the benefit principle—under which a state’s claim to tax derives from its contribution to the productive process leading to the generation of income34—with somewhat more substance.35 Yet, since there is neither a direct link between benefits from state action and income generation—and nor does it appear at all measurable—economic analysis cannot, ultimately, fill the idea of source with precise content—whether or not it is fortified with the idea of value creation.36 This may not be a huge surprise given the fact that these interrelated principles—source, benefit, value creation—all properly belong to a perspective that is geared towards 30

See W. Haslehner and M. Lamensch, eds, Taxation and Value Creation (Amsterdam: IBFD, 2021). 1923 League of Nations Report referred to the creation of wealth by ‘human agency’ as split into four separate elements: the management of the labour and organization; the agencies for transport; the controlling power deciding the enterprise’s policy; the selling activity by agents in the market jurisdiction. 32 E.g. on a mere cost basis, assuming that all inputs were themselves provided on the basis of open market transactions so that no adjustment would be required to arrive at a reliable valuation of those inputs. 33 J. Becker and J. Englisch, ‘Taxing Where Value is Created: What’s “User Involvement” Got to Do with It?’, Intertax 47/2 (2019), 161; D. N. Shaviro, ‘Digital Services Taxes and the Broader Shift from Determining the Source of Income to Taxing Location-Specific Rents’, NYU Law and Economics Research Paper No. 19-36 (2020). 34 For a recent monographic exploration of the idea, see E. Escribano, Jurisdiction to Tax Corporate Income Pursuant to the Presumptive Benefit Principle (Alphen aan den Rijn: Kluwer Law International, 2019), 31–45. 35 W. Haslehner, ‘Value Creation and Income Taxation: A Coherent Framework for Reform?’, in Haslehner and Lamensch, Taxation and Value Creation, 49. 36 Ibid., 42–48. 31 The

International Tax Law and Economic Analysis of Law 239 finding a fair and equitable allocation of taxing rights, to which economic analysis is ill- equipped to contribute.37 An alternative perspective, for which economic analysis is more adapted to be useful, considers the issue of allocation not one of overall fairness, but one of optimization. Recast in this way, the key question then is not what a country is entitled to tax, but what it can and should tax in order to maximize its welfare. The starting point of the analysis must then be that each state seeks to balance the cost and benefit of attracting foreign investors: whenever the marginal benefits of new investment by a foreign business exceed the marginal cost of that investment, a state would allow it in order to maximize its own welfare. In this balancing act, the negative side includes the cost of public goods provision that may have to increase due to additional activity in its territory as well as indirect costs to society, for instance from environmental harm; the positive side of the equation includes tax revenue that can be gained from taxing the foreign investor as well as the increased income of resident businesses and employees interacting with that business following the investment.38 This optimization analysis carried out by the state is matched by that made by the investor, who weighs the expected return of an investment in the country against the cost, among which a source tax may feature prominently. Because the investor can choose between different countries with different offers of potential returns and costs, the optimization exercise for both parties becomes a dynamic interaction between different jurisdictions vying for investment and investors looking for returns. The competition between countries would lead to an expectation that countries indeed set their source tax as the ‘price’ for location of investment at their marginal cost. If the net cost of foreign investment is zero, or even negative due to positive wealth effects of such investment, the optimal tax rate would thus be zero (or negative, e.g. in the form of direct subsidies limited to new investment). If, however, a country has some market power so that it need not offer access to foreign investors at marginal cost, a positive tax rate would be optimal (from a national welfare perspective) despite it reducing the amount of foreign investment.39 Essentially, to the extent that foreign investors can earn a higher return from investment in a particular jurisdiction than from investment elsewhere, that jurisdiction can impose a tax up to the excess return over that of any such alternative investment. This is the logic behind the efficiency-based argument for taxing location-specific rents40—as well as quasi- 37

Kane, ‘A Defense of Source Rules in International Taxation’, 353. W. Schön, ‘Value Creation, the Benefit Principle and Efficiency-Related Allocation of Taxing Rights’, in Haslehner and Lamensch, 163–164. 39 This is analogous to the familiar model result of the monopolist maximizing its own profit by selling a reduced quantity of its products at above marginal cost. Just as with the monopolist’s action, setting tax above marginal cost will create deadweight loss, but that loss would be imposed on foreigners (at least partially) so that the gain from tax revenue exceeds the loss to domestic residents from the lower level of investment. 40 Location-specific rents refer to returns on investment that exceed the long-term marginal costs, to the extent that they are tied to a specific location, i.e. that higher return is available only for investments 38 See

240 Werner Haslehner rents41—at source.42 To the extent that such capital return can be taxed without distorting the choice of investment, it would generally be efficient to impose a tax on it; the (efficiency) case for the country to whose territory the rent is geographically tied arises from the possibility that it is connected to sovereign investment.43 This analysis is somewhat incomplete, however, as it does not take into account the effect of coordination between residence and source states. The situation is significantly altered, for example, if the residence state of a foreign investor imposes a tax on (from that state’s perspective) foreign-source income coupled with a credit for foreign income tax. In this situation, the source country can tax not only rents, but all returns from the foreign investment up to the tax rate imposed by the other state without facing any reduction in the amount of such investment. The question then arises why a residence state would give up on its own tax revenue in favour of the source state in this manner. From an economic perspective focused on national welfare, it is not easy to make sense of it. In fact, the most commonly applied explanation is linked to an ideal of global welfare, as will be discussed in the next section concerning optimal taxation rules in the residence state and the application of relief mechanisms for double taxation.

14.4 Economic Analysis of Residence Taxation and Relief Looking at the international tax system from the perspective of the residence state, which has both the legitimacy and practical ability to tax residents on worldwide income, economic analysis can be usefully employed to investigate the optimal policy to be adopted towards foreign-source income. Economists analysing international taxation have done so, for the most part, from the perspective of a globally efficient allocation of resources, applied to the taxation of international capital income.44 The general framework to do so goes back to Peggy Musgrave (née Richman),45 who analysed the necessary conditions to achieve a neutralization of tax differentials in addition to preventing made in that geographic area. A standard example relates to super-normal returns available from the exploitation of natural resources. 41 Schön, ‘Value Creation’, 167 makes the point that a tax would also be efficient in this sense to the extent that the investor would face higher costs from changing the location of investment due to sunk cost in the initial investment. 42 Shaviro, Fixing US International Taxation, 160; Shaviro, ‘Digital Services Taxes’; Kane, ‘A Defense of Source Rules in International Taxation’, 354; Schön, ‘Value Creation’, 167. 43 See Kane, ‘A Defense of Source Rules in International Taxation’, 354. 44 For a review of the different positions from an economic perspective, see M. Devereux, ‘Taxation of Outbound Investment’, in J. G. Head and R. Krever, eds, Tax Policy in the 21st Century—A Volume in Memory of Richard Musgrave (Alpen aan den Rijn: Kluwer Law International, 2009), 499. 45 P. B. Richman, Taxation of Foreign Income: An Economic Analysis (Baltimore, MD: John Hopkins Press, 1963).

International Tax Law and Economic Analysis of Law 241 the harmful effects of double taxation. The economic analysis is built on the insight that the form and extent to which relief from double taxation is granted matters not only to the taxpayer and both the countries involved, but also from a global welfare perspective. Taking a global view, the purpose of designing intelligent international tax rules is to maximize worldwide production efficiency—rather than, say, the maximization of welfare in any particular country (although this may be—more realistically—what each government may aim to achieve).46 Given the constraints that follow from the necessity of taxation to fund state activities and the absence of a world government imposing globally harmonized taxes, the best one could hope to achieve is for international tax rules to distort investment choices as little as possible. This has led economists in the past generally to favour a tax system that ensures so-called capital export neutrality (CEN), which is fulfilled when an investor is subject to the same effective tax rate whether they earn a capital return in their residence state or abroad. Where this condition holds, the result should be locational neutrality of investment decisions and thereby the maximisation of global output (welfare).47 The (still) most common alternative standard introduced by Musgrave is capital import neutrality (CIN), which is fulfilled where investments in the same country face the same effective tax rate. Where this is achieved, a form of competitive neutrality for activities in that country would be reached, which would reduce distortions between domestic and foreign firms operating within the same economy and could thereby also contribute to the efficient functioning of such a market.48 For CEN to be achieved, the taxpayer’s residence state must grant a full credit for any foreign taxes (including reimbursements where the foreign tax is higher than the domestic tax on any portion of income). For CIN to be achieved, the taxpayer’s residence state would exempt foreign-earned income and the respective source state would have to ensure that foreign resident taxpayers are taxed at the same rate as domestic residents on their local-source income. Taking the perspective of aiming to maximize national rather than global welfare, a state may, however, opt for neither CEN nor CIN, but instead pursue ‘national neutrality’ (NN). Under this standard, a country aims to take account of the fact that taxes paid to another country are lost to the national economy, while taxes paid at home may be reintroduced into the economic cycle. Thus, from the perspective of national

46 See Section 14.5 on the issue of tax competition, which follows from self-interested policy decisions of countries. 47 More technically, this result is achieved by equalizing pre-tax returns from all available investment assets across the involved jurisdictions, which is the main condition for production efficiency: if one asset produced a higher pre-tax return, output could be increased by additional capital being invested into that asset. 48 Economists have pointed to another efficiency- enhancing aspect of CIN, the so-called savings neutrality: by equalizing post-tax returns (rather than pre-tax returns), CIN would lead to the same distortion at the margin of reinvestment vs consumption, ensuring that all investors’ choices are affected equally by the existence of a tax. See M. S. Knoll, ‘Reconsidering International Tax Neutrality’, Tax Law Review 64/99 (2011), 107–109.

242 Werner Haslehner production, capital outflows represent a cost insofar as their returns are diminished by foreign taxes. At the same time, a capital outflow would still be beneficial to the investor’s residence state if the pre-tax return earned abroad exceeded the available domestic pre-tax return by more than the foreign tax rate; in this case, the positive return (which accrues to the domestic economy) after deduction of the foreign tax would still be greater than in the case of domestic investment. Thus, to maximize national welfare, a country should allow a deduction of foreign taxes from the global tax base of all its residents.49 These classic neutrality notions were expanded by more elaborate economic analysis in the early 2000s, building on an analysis of the assumptions in the models that led to the previously described results. In particular, scholars pointed to the need to investigate some assumptions underpinning the standard descriptions to achieve CEN and CIN. Those ideas were built around the supposition of direct investment flows that amount to new production in the sources state as well as the hypothesis that such investment abroad would be a substitute for investment in the home jurisdiction. The notion of competitive neutrality achieved by CIN was furthermore premised on the idea that foreign-owned businesses compete with local businesses in the source state, rather than competing on a global market. Yet these assumptions are increasingly unrealistic as such investments represent an ever-smaller share of global capital flows. This has important implications for the practical application of the traditional neutrality ideas. First, if instead of the creation of new capital assets investment flows primarily reflect changes in ownership of existing assets, CEN no longer leads to global efficiency. This is so because of the assumption that it is asset ownership rather than asset location that determines its optimal use in production. Indeed, a large amount of cross-border capital flows concerns consolidation in industries, which leads to efficiency gains in the use of existing assets. Thus, if ownership is the relevant criterion, it is capital ownership neutrality (CON) that maximizes global output.50 But achieving CON is more difficult than the more traditional neutralities: although in a certain guise it can be reached by simultaneous exemption of foreign income in all countries, it would—if applied more generally with a view to reaching true market neutrality—ultimately require the considerably more challenging global harmonization of both source and residence corporate taxes.51 Secondly, the assumption of outbound investment being a complete substitute for domestic investment does not hold in practice. This is so because as outbound investment in this case counts any equity investment, meaning any investment involving a change in ownership—even partially—of productive assets. Such investment does not depend on the amount of domestic capital, since it can also be financed through capital inflows,

49 Shaviro, Fixing US International Taxation, 145. 50 For

the introduction of the term, see M. A. Desai and J. R. Hines, ‘Evaluating International Tax Reform’, National Tax Journal 56/3 (2003), 487–502, citing and building on M. P. Devereux, ‘Capital Export Neutrality, Capital Import Neutrality and Capital Ownership Neutrality and All That’, IFS Working Paper (1990). 51 Devereux, ‘Taxation of Outbound Investment’, 510.

International Tax Law and Economic Analysis of Law 243 and, for any ‘small’52 open economy, will be so financed at the margin. Consequently, the prevailing national welfare standard of NN might even be achieved through exemption of foreign-source income;53 or, for non-small economies, a tax rate below that imposed on domestic income.54 Thirdly, it is important to consider the fact of trade and competition between firms on a global market for goods and services. If the locus of competition is not the ‘source’ country, the idea that CIN actually achieves neutrality with respect to competition in a significant way is undermined. In addition, the presence of international trade also proves that the CON conditions of simultaneous exemption in all countries only achieve such neutrality when combined with equal taxation rules in the residence and source states.55 Shaviro has criticized the use of ‘neutralities’ altogether as unhelpful, since it conflates two relevant decisions (‘margins’): which tax rate to apply to foreign-source income, and how to treat foreign-source tax. While the marginal tax burden on returns from foreign direct investment has an impact on the location of profits that can be taxed, by influencing both the amount of such outbound foreign investment and the amount of inbound foreign investment,56 the marginal rate of reimbursement influences the incentive of resident taxpayers to reduce their exposure to foreign-source taxes. Realizing this, it becomes possible to find additional ways to find solutions beyond the triptych of exemption, credit, and deduction. Seeking to trade off distortions at the different margins and preferring a national welfare framework, Shaviro concludes that the tax rate imposed on foreign-source income should be lower than the tax rate imposed on domestic-source income (i.e. a tax rate lower than under a classic credit system) although what precise rate is to be applied depends on the level of substitution between domestic-and foreign-source income. At the same time, a tax system should not offer a full credit in order to retain taxpayers’ incentives to minimize their foreign tax burden.57 In sum, as one would expect from the ‘dismal science’, there is no clear answer as to the best approach to foreign-source income that is subject to source taxation. This is not a failure of economic analysis, however: rather, the different answers relate to different questions. As policymakers face different circumstances (small vs non-small economies) and aim to optimize for different outcomes (global vs national welfare), it is unsurprising that the responses differ. Yet it is clear that the analysis provides a much clearer picture than one would be able to have without it, relying purely on intuitions of equity and fairness.

52 i.e.

an economy that is small relative to worldwide demand for capital and thus has no power to influence the global return on investment. 53 Ibid., 513–515. 54 Shaviro, Fixing US International Taxation, 172. 55 Devereux, ‘Taxation of Outbound Investment’, 510. 56 The logic behind this, as explained earlier, is that outbound foreign investment is itself financed, at the margin, by inbound foreign investment. 57 Shaviro, Fixing US International Taxation, 168–172.

244 Werner Haslehner

14.5 International Tax Competition Finally, an economic approach also must be brought to bear on the problem of tax competition. Simply put, this concerns the interaction of tax law frameworks of states which are aiming to attract investment in different forms, with the ultimate goal of increasing their own (citizens’) welfare. This welfare increase can—but does not necessarily have to—come from an increase in tax revenue. In other words, tax competition refers to the use of tax law for a state’s strategic ends in general, rather than the use of tax law for the purpose of increasing tax revenue. Recent studies highlight the disagreement among legal and economic scholars concerning its meaning, its existence, its effects, and the possibilities and merits of its regulation.58 In legal literature, Roxan defines tax competition as the ‘response of sovereign governments to pressure to reduce taxation because of the behaviour of mobile taxpayers’.59 Steichen, by contrast, defines it as ‘a process whereby the fiscal decisions of one country affect the economic welfare of another’.60 Each definition highlights a different crucial element of the concept: the first emphasizes the precondition for the emergence of tax competition—mobility of taxpayers; the second focuses on the effect of governments’ reaction to this precondition; both include the additional element of governments taking self-interested decisions. Tax competition as discussed in legal literature can be said to result from governmental responses to taxpayer mobility, and to result in changes in the welfare of other jurisdictions. Among economists, Devereux and Loretz analyse the ‘uncooperative setting of source-based taxes on corporate income where the country is constrained by the tax setting behaviour of other countries’.61 More broadly, Keen and Konrad refer to tax competition as the ‘class of games’ in which national tax policymakers are involved with one another.62 This lack of agreement on a definition is probably partly to blame for widespread disagreement over the effects of tax competition; the latter disagreement almost certainly also colours the differences in defining what it is. At the heart of the debate surrounding tax competition lie two basic questions, which are, however, closely intertwined. First, whether tax competition is economically

58 L. V. Faulhaber, ‘The Trouble with Tax Competition: From Practice to Theory’, Tax Law Review 71/2 (2018), 311; M. P. Devereux and S. Loretz, ‘What Do We Know about Tax Competition?’, National Tax Journal 66/3 (2013), 745, 746; M. Keen and K. A. Konrad, ‘The Theory of International Tax Competition and Coordination’, in A. J. Auerbach et al., eds, Handbook of Public Economics, vol. 5 (Amsterdam: Elsevier, 2013), 257, 321; for an instructive overview of theories about tax competition, see J.D. Wilson, ‘Theories of Tax Competition’, National Tax Journal 52/2 (1999), 269–299. 59 I. Roxan, ‘United Kingdom’ in W. Schön, ed., Tax Competition—EATLP (Amsterdam: IBFD, 2003), 485. 60 Alain Steichen, ibid., 45. 61 Devereux and Loretz, ‘What Do We Know about Tax Competition?’, 746. 62 Keen and Konrad, ‘Theory of International Tax Competition and Coordination’, 258.

International Tax Law and Economic Analysis of Law 245 beneficial or harmful to states participating in it. Secondly, whether tax competition is also an ethical problem or merely an economic one. The ethical aspects of tax competition are at least as difficult to tackle as the economic one, but crucially depend on the conclusion as to the economic impact: if all countries were better off without tax competition, the ethical case to curb it would be beyond doubt. If all countries were better off as a consequence of tax competition, the ethical case to accept it would be established equally easily. If, as is most plausible, tax competition has both positive and negative economic effects—and those effects are unequally distributed among countries with different characteristics—the ethical question becomes very difficult to resolve. Other authors have approached that issue,63 which this chapter will not attempt to answer. It can only provide a brief overview of the different economic arguments that necessarily underlie any further analysis.64 The economic case for the harmful effects of tax competition builds on the assumption that, in the absence of tax competition—if we imagine only one country to exist, for instance—tax burdens65 would be set at an optimal level. This means, at a level that allows government to have enough revenue to provide public goods so that social benefits equal the social cost of taxation. Taking this starting point, any reduction in taxation would inevitably66 result in a lower provision of such public goods. Of course, it is neither obvious that governments would actually set taxes at the social optimum, nor that all taxes are significantly affected by tax competition: if it only concerns globally mobile capital, states could still collect any amount of revenue they require from other less mobile sources.67 Yet, even then, a marginal cost–benefit analysis suggests that tax burdens for mobile capital might be set below the global social optimum. The reason for this is that each state would set its tax rates in such a way that the loss of revenue from existing capital subject to its tax jurisdiction is matched by an increase in revenue from capital attracted from other countries. In this calculation, each country would ignore the social benefits lost to the country from the outflow of capital, creating an external cost that would lead to an inefficient under-provision of state resources.68

63 See,

e.g., A. Christians, ‘Sovereignty, Taxation and Social Contract’, Minnesota Journal of International Law 18/ 1 (2009), 99; P. Dietsch, Catching Capital: The Ethics of Tax Competition (Oxford: Oxford University Press, 2015). 64 On the interaction of economic and normative (constitutional) perspectives, see W. Haslehner, ‘Time and Tax: Constitutional versus Economic Perspectives’, in Haslehner et al., Time and Tax, 271. 65 Tax competition is often discussed merely in terms of tax rates, as this is the most salient aspect of tax laws both for multinational companies and the public. It is also the easiest to study. Rational actors are not supposed to consider the nominal rate, but base their decisions exclusively on the effective tax burden they face. 66 The state could, of course, fund public goods via public debt, but the assumption must be that such debt ultimately needs to be repaid with taxes as well. 67 This may actually have beneficial consequences since, as seen in Sections 14.3 and 14.4, the tax rate on both inbound and outbound capital may ideally be set to zero under certain assumptions, e.g. the absence of economic rent. 68 See Wilson, ‘Theories of Tax Competition’, 275; Keen and Konrad, ‘Theory of International Tax Competition and Coordination’, 267–269.

246 Werner Haslehner A case for beneficial effects of tax competition can also be built on various economic arguments: one builds on the Tiebout model for the efficient provision of public goods, which showed that allowing taxpayers to choose fixed bundles of public goods and tax burdens by ‘voting with their feet’ was an efficient solution to the problem of concealed preferences.69 The full application of this idea to the international tax arena fails on practical grounds, such as sufficient mobility and taxpayer’s ability to ‘unbundle “packages” of sovereignty’.70 Another powerful argument lies in the general point that a low or zero corporate tax rate may actually be close to optimal due to the benefits of leaving cross- border mobile capital undistorted71 and the (unrelated) result from optimal tax theory that capital income should not be taxed if alternatives are available in order to avoid the distortion of intertemporal investment/consumption decisions.72 Finally, public choice theory of economics stresses another benefit from tax competition: constraining the state from socially suboptimal over-taxation.73 It should be noted that, although often described as such,74 tax competition is not a simple case of a prisoner’s dilemma. If it were, cooperation through the coordination of tax rates would not only increase global revenue, but also the revenue for each participating economy, aligning global and national welfare considerations.75 However, in order to apply the logic of the prisoner’s dilemma to the tax competition framework, the participants in the game would have to face uniform costs and benefits. But economies differ in size, resources, and existing capital, among other things. Economic models capture this by modelling asymmetric competition, which suggests optimal tax rates for smaller jurisdictions to be lower than for larger countries.76 This is intuitive: under the assumption of a fixed global capital stock available for investment in all countries, the potential gain from attracting foreign capital is much greater for small countries relative to the potential loss from tax on their domestic capital.77 As a consequence, in the absence of direct transfers of tax revenue from the true ‘winners’ of cooperation to the ‘losers’, even assuming that a capital tax should be imposed, the case for

69 C.

M. Tiebout, ‘A Pure Theory of Public Expenditures’, Journal of Political Economy 64/5 (1956), 416–424. For an argument built partially on the insights by Tiebout, see D. Elkins, 'The Merits of Tax Competition in a Globalized Economy‘, Indiana Law Journal 91 (2016) 906–954. 70 T. Dagan, ‘International Tax and Global Justice’, Theoretical Inquiries in Law 18/1 (2017), 1 at 14; Faulhaber, ‘The Trouble with Tax Competition’, 318–319. 71 See Sections 14.3 and 14.4; A. J. Auerbach, M. P. Devereux, and H. Simpson, ‘Taxing Corporate Income’, in IFS, ed., Dimensions of Tax Design—The Mirrlees Review (Oxford: Oxford University Press, 2010), 837, 868. 72 See Wilson, ‘Theories of Tax Competition’, 282; for an extensive (and critical) discussion, see e.g. J. Banks and P. Diamond, ‘The Base for Direct Taxation’, in IFS, Dimensions of Tax Design, 548–648. 73 Wilson, ‘Theories of Tax Competition’, 298. 74 See Elkins, 'The Merits of Tax Competition in a Globalized Economy‘, 908. 75 For an instructive explanation of the prisoner’s dilemma in the international tax setting, see Shaviro, Fixing US International Taxation, 137–139. 76 Devereux and Loretz, ‘What Do We Know about Tax Competition?’, 749. 77 Wilson, ‘Theories of Tax Competition’, 279.

International Tax Law and Economic Analysis of Law 247 cooperation by harmonizing tax rates (unlike the case for agreeing to a global minimum tax that all countries abide by) is weak. Luckily, however, economic analysis provides the basis for a solution other than global coordination: tax competition can be effectively removed by replacing a classic source-based capital tax with an alternative form of taxing mobile capital that is by design not subject to the pressures that give rise to tax competition in the first place, since they are imposed in locations in respect of which there is less mobility than for the illusive ‘source’:78 individual (rather than corporate) residence, or destination (i.e. the place of final consumption).

14.6 Conclusion Economic analysis does not provide as clear an answer to the pressing questions of international tax law as one might wish. This is not the fault of the discipline, or of the scholars applying it, but rather follows from the fact that the questions asked in tax law are not mere questions of efficiency, but involve important distributional concerns. Yet, in order to appropriately address those concerns, applying the insights from economic analyses are invaluable as it is only through those that informed decisions on equity grounds can be based. This chapter has shown how such economic analysis can inform the necessary debate on four such decisions.

78

Which, as used here (as commonly in economic analyses), must be understood as a reference (also) to corporate residence.

Chapter 15

Internationa l Tax L aw and L ang uag e Florian Haase

15.1 Introduction All across the world, behaviourists, psychologists, and cognitive scientists alike have struggled with the conundrum where language comes from and how it evolves. Is language acquired by reinforcement and repetition,1 or learned through interaction?2 Is it simply an innate ability3 or an instinct?4 Is it part of the overall development5 when we grow up or ‘an instinctive tendency to acquire an art’,6 as Charles Darwin once put it? Leaving this aside, it is no exaggeration to say that human experience in general is scarcely imaginable without language. Language is unique to humanity, and we come across it in many different forms (e.g. spoken language, written language, sign language, language dancing). Language is arguably the cultural gift that differentiates human beings from any other species, and on some accounts, language is from a historical perspective the main symbolic behaviour7 that allowed human peculiarities like the arts, religion, or science to develop in the first place. Language is also a necessary form of any legal system,8 as we know from the philosophy of law.9 In the early days of legal positivism, linguistic acts were observed as 1

B. F. Skinner, Verbal Behavior (La Jolla, CA: Copley, 1957). S. Bruner, ‘From Communication to Language: A Psychological Perspective’ Cognition 3 (1975), 255. 3 N. Chomsky, ‘Review of Skinner’s Verbal Behavior’, Language 35/1 (1959), 26–58. 4 S. Pinker, The Language Instinct (New York: Harper, 1994). 5 J. Piaget, The Mechanism of Perception (New York: Basic Books, 1969). 6 C. Darwin, The Descent of Man (Cambridge: Cambridge University Press, 1871), 58. 7 J. Huttenlocher and E. T. Higgins, ‘Issues in the Study of Symbolic Development’, Minnesota Symposia on Child Psychology 11 (1978), 98. 8 H. L. A. Hart, The Concept of Law, 3rd ed. (Oxford: Oxford University Press, 2012), 14. 9 See for details, A. Marmor, The Language of Law (Oxford: Oxford University Press, 2014). 2 J.

250 Florian Haase empirical phenomena, and they were made an essential element of the theory of law.10 Law as a social system of order is decisively dependent on behavioural coordination and consequently on communication, language, and culture. However, the comparison of laws and the segregation into major legal families in the world, as we know from comparative law,11 has not yet been thoroughly investigated for tax law purposes.12 Regardless of the question whether tax law is part of the public law or the civil law of a country’s legal system,13 it will in most cases be based on codified provisions which form part of the respective national law. International double taxation treaties, other conventions with a link to taxation as well as tax-related papers published by the OECD, the IMF, the WTO, the World Bank, the UN, or other supranational organizations are of course also presented in written language, and the same is first and foremost true of the legislative acts that have, directly or indirectly, their roots in EU law. In short, language is key to tax law, in theory and practice. Moreover, ‘legislative criticism does not exclude any area of law, but tax law in general is in many countries mentioned as a reference area for particularly “terrible” laws, quasi as a synonym for bad legislation’.14 This is, also from an overall economic point of view, particularly regrettable because studies have shown that tax evasion is less prevalent in countries where legislators seek clear and comprehensible language in tax statutes and where taxpayers are therefore better able to understand the purpose of certain rules.15 Within this setting, ‘the’ language of international tax law is clearly English. English is not only spoken as a first or second language in most parts of the world, but is also the main working language of the OECD, the EU, and other important international organizations that work in tax-related fields. Most official documents, tax journals, materials, and provisions with respect to international tax law are (at least also) available in English. This brings us to the general problem of multilingualism in international tax law, which increases the intricacies in interpreting international tax law rules dramatically. The influence that language has on the application of law in general16 and tax law in particular cannot be underestimated. If law, by and large, equates to language, it is 10

J. Bentham, cited in J. H. Burns and H. L. A. Hart, eds, Bentham: A Fragment on Government—The New Authoritative Edition (Cambridge: Cambridge University Press, 1988), 14. 11 K. Zweigert and H. Kötz, An Introduction to Comparative Law, 3rd ed. (Oxford: Oxford University Press, 1998); see also ‘Comparative Tax Law’, Section 16.2 in this volume. 12 See for first steps, P. Chuenjit, ‘The Culture of Taxation: Definition and Conceptual Approaches for Tax Administration’, Journal of Population and Social Studies 22/1 (2014), 14. 13 But see C. Harlow, ‘ “Public” and “Private” Law: Definition without Distinction’, Modern Law Review 43 (1980), 241; J. H. Merryman, ‘The Public Law–Private Law Distinction in European and American Law’, Journal of Public Law 17 (1968), 3. 14 Translated from J. Hey, ‘Steuergesetzgebungslehre’, Steuer und Wirtschaft 96/1 (2019), 3. 15 P. Webley et al., Tax Evasion: An Experimental Approach (Cambridge: Cambridge University Press, 1991). 16 See F. Bowers, Linguistic Aspects of Legislative Expression (Vancouver: University of British Columbia Press, 1989); D. R. Klinck, The Word of the Law: Approaches to Legal Discourse (Ottowa: Carleton University Press, 1992); L. M. Friedman, ‘Law and Its Language’, George Washington Law Review 33 (1964), 563; B. S. Jackson, Making Sense in Law: Linguistic, Psychological and Semiotic

International Tax Law and Language 251 inherent that the well-known problems of usage and abusage of language (e.g. its vagueness, its manipulative effects) inevitably also influence the results of the application and interpretation of law. In this chapter, I first elaborate on language and its interpretation in double taxation treaties (Section 15.2), where additional problems come into play due to the fact that such tax treaties are usually concluded in at least two binding, authentic languages. I then investigate the languages that are used in EU tax law and by the institutions of the EU in tax matters (Section 15.3). The chapter closes with a discussion and suggestions on how problems arising from language in international tax law could be avoided or at least mitigated (Section 15.4).

15.2 Double Taxation Treaties 15.2.1 General Double taxation treaties are treaties that determine, that is divide or restrict taxation rights between two contracting states, mainly in bilateral situations, but also in relation to third states (e.g. art. 21 OECD Model Convention (MC)). They make use of a tax- specific, rather technical language which may in some scenarios also be used outside those treaties (but then often with a different meaning in national tax law). This is true regardless of whether the tax treaty is based on an internationally accepted model (such as the OECD MC or the UN MC), a merely regional MC,17 a national MC that countries preferably use when negotiating individual treaties18 or no MC at all. The following scenarios can be observed in practice: • The treaty uses terms that are defined in the ‘general definition article’ of the treaty, article 3(1) OECD MC. These terms may or may not be used in the national (tax) laws of the contracting states (Case 1). • The treaty uses terms that are not defined in the ‘general definition article’, but are defined either in a specific definition article (e.g. art. 5 OECD MC) or for the purposes of a specific distributive rule (e.g. art. 6(2) sentence 2 OECD MC or art.

Perspectives (Liverpool: Deborah Charles Publications, 1995); P. Goodrich, Languages of Law: From Logics of Memory to Nomadic Masks (London: Weidenfeld and Nicolson, 1990); R. Hiltunen, Chapters on Legal English: Aspects Past and Present of the Language of the Law (Helsinki: Suomalainen Tiedeakatemia, 1990). 17

See ‘Agents in International Tax Treaties’, Section 28.3 in this volume. the German Model Tax Convention (‘Basis for negotiation for agreements for the avoidance of double taxation and the prevention of fiscal evasion with respect to taxes on income and on capital’), available in English at https://www.bundesfinanzministerium.de/Content/DE/Standardartikel/Themen/ Steuern/ Internationales_Steuerrecht/Allgemeine_Informationen/2013-08-22-Verhandlungsgrundlage- DBA-englisch.html. 18 e.g.

252 Florian Haase 10(3) Alternatives 1 and 2 OECD MC). These terms may or may not also be used in the national (tax) laws of the contracting states (Case 2). • The treaty uses terms that are neither defined in the ‘general definition article’ nor in a specific definition article nor a specific distributive rule, but the applicable treaty provision refers to the national laws of one of the contracting states (e.g. art. 6(2) sentence 1 OECD MC or art. 10(3) Alternative 3 OECD MC). These terms may, again, be used or may not be used in the national (tax) laws of the respective contracting states (Case 3). • The treaty uses terms that are defined neither in the ‘general definition article’ nor in a specific definition article nor a specific distributive rule, and the applicable treaty provisions do not explicitly refer to the national laws of one of the contracting states, but the country that applies the treaty defines the term in its national (tax) law (Case 4). • The treaty uses terms that are defined neither in the ‘general definition article’ nor in a specific definition article nor a specific distributive rule, the applicable treaty provisions do not refer to the national laws of one of the contracting states, and the country that applies the treaty does not define the term in its national (tax) law (Case 5). The legal basis for solving these cases is article 3(2) OECD MC. This provision refers, for the interpretation of terms not defined in the OECD MC which are also used in the respective national laws of the contracting states, to the tax provisions of the state of application with precedence and, subsidiarily, to the (tax) provisions of the other contracting state. The legal status at the time of the application and not the legal status at the time of the signing of the agreement is decisive.19 A restriction, however, of the reference to the laws of the contracting states exists in cases where the context requires a different interpretation. Article 3(2) thus deals with the use of expressions used but not defined in the OECD MC. With reference to the cases described earlier, one can categorize the scenarios as follows: how should terms and definitions be interpreted which are defined in the treaty, and which may or may not be used in the national (tax) laws of the contracting states (Cases 1 and 2)? How should terms and definitions be interpreted in cases in which there is an explicit referral to the national laws of a contracting state (Case 3)? What is the role of article 3(2) OECD MC in cases in which, in the absence of an explicit referral to the national laws, terms and definitions are not defined in the treaty, but the country that applies the treaty does or does not define the terms and definitions used in the treaty in its national (tax) law (Cases 4 and 5)?

19

See also the OECD Model Commentary on art. 3(2), para. 11.

International Tax Law and Language 253

15.2.2 The Role of Article 3(2) OECD MC There are very different views on the effect of article 3(2) OECD MC. According to some authors, the purpose of this regulation is to safeguard the interests and sovereignty of the contracting states.20 The reference to the domestic (tax) law of the state that applies the treaty, however, entails the risk of different interpretations. For this reason, reference is sometimes made to the principle of decision-making harmony or regulatory homogeneity to advocate cautious application of this provision, also due to its subsidiarity. Other authors justify such an approach with the restriction expressly mentioned in article 3(2) OECD MC to cases in which the context does not require otherwise.21 In my opinion, the reference in article 3(2) OECD MC is the result of the fact that the agreements have an intermediate position in the light of the legal hierarchy. Double taxation treaties are by nature both international treaties and national tax law at the same time. This inherent link also leads to specific features in the interpretation of such treaties. They modify the national tax law and the tax obligations resulting from it as well as the taxation competences of the contracting states. In addition, the use of domestic terms leads to a considerable simplification in the application of the agreements. Of course, this only applies to the extent that no independent definition is contained in the agreement or no other primary understanding of the term results from the context of the agreement or accompanying documents. Individual areas of regulation are already reserved for national law in the absence of a corresponding provision in the agreement. This concerns, in particular, the determination of profits or income as well as the attribution of income to individual taxpayers.

15.2.3 Interpretation Rules Double taxation treaties are often referred to as a separate legal family.22 Nevertheless, due to the function of these agreements, there is a close interlocking with domestic law. This modification function obviously suggests that the contracting states refer to their national (tax) law for the purposes of interpretation under article 3(2) OECD MC or even implicitly. Domestic concepts are therefore regularly of major importance in the

20 K.

Vogel and R. Prokisch, ‘General Report: Interpretation of Double Taxation Conventions’, Cahiers de Droit Fiscal International, vol. 78a (1993), Note 1, 81: ‘Art. 3(2) expresses the idea that the tax sovereignty of contracting states should as far as possible be left untouched.’ 21 See the references in E. van der Bruggen, ‘Unless the Vienna Convention Otherwise Requires: Notes on the Relationship between Article 3(2) of the OECD Model Tax Convention and Articles 31 and 32 of the Vienna Convention on the Law of Treaties’, European Taxation (May 2003), 142, 143ff. 22 This is true of countries that follow a dualist approach under which international law and domestic law are separate legal systems—see F. Engelen, Interpretation of Tax Treaties under International Law (Amsterdam: IBFD, 2004), 518–519.

254 Florian Haase interpretation of treaties. But, conversely, the terms used in the agreements have no meaning for the application of domestic law. The OECD MC and the Model Commentary23 are the most important sources for interpreting treaty provisions. They provide an idea of how treaty terms are most commonly understood internationally. It is generally assumed that the contracting states are also familiar with the agreed term, unless a different understanding is expressed. This can be done, for example, in the Model Commentary by including reservations. Contracting states will only have recourse to MC comments that were available at the time the agreement was concluded.24 The resulting difficulties in determining which understanding of a term was known on the basis of the express wording of the OECD MC and the Model Commentary are rightly pointed out in literature.25 OECD sources only serve to research the possible content of the term. This must then be further substantiated in the context of interpretation according to the respective national principles. Double taxation treaties are, by their legal nature, treaties under public international law. They are therefore subject to the Vienna Convention on the Law of Treaties (VCLT) of 1969.26 According to article 31(1) VCLT, international treaties are to be interpreted in good faith. The interpretation of the wording shall be based on the usual meaning and context. The grammatical interpretation, however, will basically represent the boundary of interpretation. The aims and purposes of a double taxation treaty must also be obeyed strongly. In addition to the wording of a double taxation treaty, the preamble and annexes and other accompanying documents to the treaty must also be taken into account.27 Furthermore, agreements concluded subsequently by the contracting parties and corresponding tacit agreements are likewise relevant. Where treaties have been concluded in several languages, all languages are equally authoritative under article 33(1) VCLT, unless the parties have agreed otherwise. In case of remaining language conflicts, article 33(4) again points to the meaning and context in good faith, but even this provision does not provide for a final solution or binding interpretation rule between the contracting states. 23 M. Lang and F. Brugger, ‘The Role of the OECD Commentary in Tax Treaty Interpretation’, Australian Tax Forum 23 (2008), 95. 24 M. Lang, ‘Later Commentaries of the OECD Committee on Fiscal Affairs, Not to Affect the Interpretation of Previously Concluded Tax Treaties’, Intertax 25/1 (1997), 7. 25 U. Linderfalk and M. Hilling, ‘The Use of OECD Commentaries as Interpretative Aids— The Static/Ambulatory-Approaches Debate Considered from the Perspective of International Law’, Nordic Tax Journal 1 (2015), 34; C. Garbarino, Judicial Interpretation of Tax Treaties: The Use of the OECD Commentary (Cheltenham: Edward Elgar, 2016). 26 Vienna Convention on the Law of Treaties (with annex). Concluded at Vienna on 23 May 1969. Authentic texts: English, French, Chinese, Russian, Spanish. 1155 UNTS 331, 8 ILM 679. 27 See for interaction with art. 3(2), D. O. Corredor Velásquez, ‘Tax Treaty Interpretation: Interaction Between Article 3(2) Organization for Economic Co-Operation and Development Model Convention and Article 31 Vienna Convention’, Intertax 44/12 (2016), 960; van der Bruggen, ‘Unless the Vienna Convention Otherwise Requires’, 142.

International Tax Law and Language 255 If one of the contracting states is an EU member state, the question arises whether EU law is relevant to the interpretation of a double taxation treaty.28 Some are of the view that double taxation should be avoided solely on the basis of bilateral agreements (at least at the current stage of the harmonization of direct taxes within the EU), according to which European law and the fundamental freedoms granted therein do not play a role in the interpretation of double taxation agreements.29 Others are of the view that EU law must also be taken into account in order to avoid infringements of EU law.30 Although EU member states are free in how they regulate the division of tax sovereignty in relation to another contracting state in a bilateral agreement, they must nevertheless apply such agreements in compliance with EU law.31 This includes, in my opinion, an interpretation of the agreement in conformity with European law (primary, secondary, and tertiary law). In the context of interpretation, it seems natural for any state to base the text of the agreement on its own language. However, double taxation treaties regularly provide for the respective languages of the contracting states to be equally binding (i.e. authentic). These can be two or more languages if, for example, a state has more than one national language. A translation can never succeed in producing an identical version in the other language, since despite the fact that there are cognates between languages, the words available in the other language usually do not have the same meaning. There will always be nuances that make a difference. An interpretation which is only based on wording would therefore regularly lead to a different result in the other language version. Some authors therefore argue that the languages of the respective underlying model agreements should also be taken into account when interpreting in all binding languages.32 In this respect, however, a clear distinction must be made between the interpretation of the specific double taxation agreement to be applied and that of a MC. The interpretation result of the MC is not necessarily the basis for the application of 28 For the interdependencies between EU law and tax treaties in general, see the working document issued by the European Commission, Ref. TAXUD E1/FR DOC (05) 2306 (9 June 2005). 29 See, with further references, J. F. Avery Jones et al., ‘The Interpretation of Tax Treaties with Particular Reference to Article 3(2) of the OECD Model (Parts I and II)’, British Tax Review (1984), 18 and 90. The underlying question here is whether the referral in art. 3(2) OECD MC is limited to national law (as art. 3, para. 13.1 of the Model Commentary seems to suggest) or not (and, if the latter, whether EU law can be ‘law’ in this sense). 30 This is particularly argued by F. Avella, ‘Using EU Law To Interpret Undefined Tax Treaty Terms: Article 31(3)(c) of the Vienna Convention on the Law of Treaties and Article 3(2) of the OECD Model Convention’, World Tax Journal 4/2 (2012), 95 (with further references); Engelen, Interpretation of Tax Treaties. 31 See, for a discussion, G. Beck, ‘The Macro Level: The Structural Impact of General International Law on EU Law: The Court of Justice of the EU and the Vienna Convention on the Law of Treaties’, Yearbook of European Law 35/1 (2016), 484; P. Andrés Sáenz de Santa María, ‘The European Union and the Law of Treaties: A Fruitful Relationship’, European Journal of International Law 30/3 (2019), 721. 32 See for a comprehensive overview, G. Maisto, Multilingual Texts and Interpretation of Tax Treaties and EC Tax Law, vol. 1 in the EC and International Tax Law Series (Amsterdam: IBFD, 2005), with further references; F. Arginelli, Multilingual Tax Treaties: Interpretation, Semantic Analysis and Legal Theory (Amsterdam: IBFD, 2015).

256 Florian Haase a double taxation agreement in an individual case in practice. Rather, the content determined within the framework of the interpretation of the MC serves the interpretation of a specific agreement which is based on the corresponding MC. The result is that the interpretation of the wording always reaches its limits where an interpretation in another likewise binding language version leads to a different result. In these cases, a comparison of the different language versions shows that the interpretation objective of determining the content of a provision cannot be achieved on the basis of a (pure) wording interpretation. For the taxpayer, the greatest challenge will often be to rebut a particular result of interpretation achieved by a national tax authority or tax court relative to a contract in one national language with reference to another authentic version. In many cases, this can only be achieved, if at all, with the participation of the other contracting state (e.g. within the framework of a mutual agreement procedure). The OECD MC expressly defines individual terms in various places, in particular in articles 3(1) and 5 OECD MC. These definitions can be comprehensive or partial. Article 3(2) OECD MC also refers to the law on taxes of the contracting state which applies the treaty, unless the context requires otherwise. There should be agreement on the fact that terms defined in the agreement itself take precedence (i.e. recourse to the national law of one of the contracting states is not necessary). Terms that are used in an agreement or which can be interpreted in context within the agreement, but which are also used in the law of the state that applies the treaty or in the other contracting state, are to be interpreted by comparing the respective term contents. In this respect, it should be noted that treaties and domestic law must be regarded as different conceptual levels; that is, there may be no conceptual identity, but at most a conceptual parallelism. If an interpretation based on an agreement and the interpretation according to the respective national understanding of the terms lead to different results, it must be clarified whether the interpretation found from the context of the agreement is necessary and therefore has priority. In the opinion of many states, if there is a definition missing in the agreement, the first port of call is to see whether the term can be determined from the context.33 Failing this, where there is a conflict between the domestic tax law of one contracting state (applying a treaty) and that of the other, then the domestic tax law of the former prevails (article 3(2) OECD MC. If terms not defined in an agreement are not used or defined in the national tax law of the state that applies the agreement, the interpretation must be based on the agreement; that is, recourse to the other laws of the state that applies the agreement is not possible and is ruled out. In this respect, it must be taken into account that the terms used may only be the result of the respective translation or that the terms are purely

33 See

the references in D. A. Ward et al., The Interpretation of Income Tax Treaties with Particular Reference to the Commentaries on the OECD Model (2005); F. van Brunschot, ‘The Judiciary and the OECD Model Tax Convention and Its Commentaries’, Bulletin for International Taxation 59/1 (2005), 5, 6; M. Erasmus-Koen and S. Douma, ‘Legal Status of the OECD Commentaries: In Search of the Holy Grail of International Tax Law’, Bulletin for International Taxation 61/8 (2007), 339, 342.

International Tax Law and Language 257 coincidentally identical with those of national law and that no identity of terms with regard to content was intended. If an interpretation of an agreement and an interpretation according to the respective national understanding of the term lead to different results, it must be clarified whether the interpretation found from the context of the agreement is necessary and therefore of priority. In most cases, in the event of a missing definition in the agreement, it must first be examined whether the term can be determined from the context of the provision of the agreement. An interpretation based on an agreement always takes precedence, and this extends to official accompanying documents. As to which version of national law is to be used as the basis for interpretation, the wording of article 3(2) OECD MC and article 3 paragraph 11 of the Model Commentary to that provision state that the time of application of the law and not the time of conclusion of the treaty must be taken into account. It is therefore a so-called dynamic reference, since the law on which the interpretation is based may evolve or change. The dynamic reference and the possibility of amending agreements as a result may, however, be limited by the agreement itself. Such a limit may, for example, lie in a partial definition contained in the agreement, provided that this contradicts the interpretation on the basis of the domestic law of the state that applies the treaty. In addition, an agreement may contain limitations on a reference to the law of a contracting state. For example, article 6(2) OECD MC contains a restriction on the reference to the laws of the state where immovable property is located in that it expressly excludes ships and aircraft from the definition of immovable property and thus excludes their inclusion on the basis of the law of that state. However, the permissibility of such reference is problematic in those states in which a double taxation treaty does not per se become effective through the conclusion of the treaty, but must first be transposed into national law.

15.2.4 Restrictions The interpretation of terms and definitions which are used in a double taxation treaty is limited to the extent that any other interpretation required by the context takes precedence (‘unless the context otherwise requires’). According to article 3 paragraph 12 of the Model Commentary to article 3(2) OECD MC, both the ideas or intentions of contracting parties when signing an agreement and the content of the reasoning in the other contracting state (expression of the so-called international decision harmony) are of equal importance with respect to interpretation.34 34

There is an interesting case decided by the CJEU (as an arbitrator) on referral in art. 3(2) OECD MC to the national laws of a contracting state (see Case C-648/15 Republic of Austria v. Federal Republic of Germany, 12 September 2017). The Court argued that such a rule of interpretation by a single state was not a rule intended to arbitrate between divergences of interpretation between the two states and that this understanding would deprive the arbitration provisions of art. 25(5) of ‘all practical effect’.

258 Florian Haase A new restriction with the same legal consequence came into effect through the 2017 update to the OECD MC. If the competent authorities agree to a different meaning pursuant to article 25 OECD MC, this meaning will prevail over the understanding in the sense of that state for the purposes of the taxes to which the convention applies. The Report of Action 14 of the Base Erosion and Profit Shifting (BEPS) Project35 called for these changes—which are intended to remove any doubt that, in a case where the competent authorities have agreed on a common meaning of an undefined term, the domestic law meaning of that term would not be applicable—as part of the follow-up work on the BEPS minimum standard Action 14 to clarify the legal status of a competent authority mutual agreement.

15.3 European Union (Tax) Law The language of EU law is as diverse as the economic and social background of the member states. As regards EU primary law, sedes materiae for multilingualism is article 55(1) Treaty on European Union (TEU) and article 225 Treaty establishing the European Atomic Energy Community respectively, according to which each text is equally binding. Therefore, we have on principle an equal legal status of all versions of the EU Treaties in the twenty-four recognized languages of the EU today.36 As regards EU secondary law, a regulation establishes this—Regulation No. 1 determining the languages to be used by the European Economic Community37—which is still valid and binding law. The effect of this—cum grano salis—‘statutory multilingualism’ within EU law cannot be underestimated. The aim of the language regime is to make EU law available to all EU citizens in their own language, which is not only relevant if we take the EU provisions literally, but also when we interpret them particularly in the light of the doctrine of effet utile.38 Real multilingualism, however, can only arise if there is more than one authentic language, as is the case with article 55(1) TEU and Regulation No. 1. By contrast, the translations of the EU Treaties into other official EU languages, as permitted under

35 https://w ww.oecd.org/tax/making-dispute-resolution-mechanisms-more-effective-action-14- 2015-final-report-9789264241633-en.htm. 36 In practice, English and French prevail. EU law and other legislative texts are published in all official languages, except for Irish, for resource-related reasons (only regulations adopted by both the Council of the European Union and the European Parliament are currently translated into Irish). 37 Council Regulation of 15 April 1958 [1958] OJ 385/58, as amended by and currently valid in the sense of Council Regulation 517/2013 [2013] OJ L158/1; art. 1 of that regulation currently lists twenty-four official working languages, and under art. 4, regulations and other documents of general application are drawn up in the official languages of the EU. 38 See in this regard, U. Šadl, ‘The Role of Effet Utile in Preserving the Continuity and Authority of European Union Law: Evidence from the Citation Web of the Pre-Accession Case Law of the Court of Justice of the EU’, Journal of European Legal Studies 8 (2015), 18, 22.

International Tax Law and Language 259 article 55(2) TEU, are not legally binding. There are also certain legal acts that are legally binding in one language only, such as decisions addressed only to certain EU member states or companies of such member states. In these texts and documents, divergences between the authentic version and translations can have only very limited significance in terms of content—for example, when the authentic version only becomes comprehensible on reading a translation. In principle, however, their interpretation can only be based on the only authentic language version. Similar considerations are true of judgments of the EU Courts (i.e. the Court of Justice of the European Union and the General Court). They are formally binding only in the language of proceedings. How they are to be interpreted follows from a different version, namely the language of the hearings (i.e. French).39 That is the only language that has been examined and authorized by all the judges responsible. Only French can therefore provide reliable information on what an EU Court has decided. Divergences may, however, come to light due to translation errors. These cannot call into question the content of the judgment, but can at most give rise to a correction of the translation. Moreover, when referring to earlier judgments, it is unlikely that the EU Courts will consider versions of these other than the French. This would only occur where the parties or the Opinion rendered by the Advocate General explicitly pointed out divergences. The key challenge of multilingualism is to ensure coherence among different language versions. While the EU institutions must aim to ensure that all language versions of all texts are consistent with one another,40 in practice it is almost impossible to achieve this objective in total. Accordingly, even in cases before the EU Courts, there are more or less significant divergences between the language versions identified. No single version can be applied in isolation in order to avoid the application of different rules in different language areas. In order not to impair the uniform effect of EU law, it is necessary to develop a common interpretation of all the language versions of a provision. The principle is that all language versions should be given the same value.41 There is therefore no language that enjoys priority, nor does the value of a language version vary with the size of the population of member states using that language.42 It also does not matter whether a particular variant of a legal provision is used in the majority of language versions. All these methods would be incompatible with the equal status of languages.

39

M.-A. Gaudissart, ‘Le régime et la pratique linguistiques de la Cour de Justice’, in D. Hanf et al., eds, Langues et construction européenne (Brussels: Peter Lang, 2010), 137, 149ff. 40 M. Guggeis, ‘Multilingual Legislation and the Legal– Linguistic Revision in the Council of the European Union’, in B. Pozzo and V. Jacometti, eds, Multilingualism and the Harmonisation of European Law (Alphen aan den Rijn: Kluwer, 2006), 109. This is necessary for uniform application of EU law: see Case 29/69 Stauder, ECLI:EU:C:1969:57, para. 3 and Case C-188/03 Junk, ECLI:EU:C:2005:59, para. 33. 41 See, inter alia, Case 283/81 Cilfit et al., ECLI:EU:C:1982:335, para. 18; Case C-296/95 EMU Tabac et al., ECLI:EU:C:1998:152, para. 36; Case C-152/01 Kyocera, ECLI:EU:C:2003:623, para. 32. 42 See Case C-296/95 EMU Tabac et al., ibid, para. 36; Case C-257/00 Givane et al., ECLI:EU:C:2003:8, para. 36.

260 Florian Haase However, unanimous consent from the viewpoint of all authentic languages is difficult to reach, particularly for taxpayers.43 The Court of Justice always looks at the individual case. In cases in which only one language version differs from all the others, the Court has held that such a divergence could not reflect the real will or intention of the EU legislator.44 Conversely, under special circumstances, there may also be cases in which the Court may reach an interpretation which is expressed in only one language version.45 Therefore, the wording of an EU Treaty provision in the different languages only sets the framework within which the result of the interpretation must be developed with emphasis on the remaining methods of interpretation. These are, in particular, the origin of the provision, its systematic context, and its objectives, as it will be known in principle from the interpretation of tax laws in most countries. The latter is very important, because in practice even the Court of Justice does not try to distil the true meaning of a tax rule in any case and at any price, because despite the available resources in terms of highly qualified personnel46 this would be extremely costly and a rather lengthy, time-consuming process. Therefore, the Court limited the obligation to take into account other language versions to cases of doubt.47 The Court moved even further and only deals with the multilingualism of the applicable rules in cases it sees an opportunity to achieve a different result.48 Typically, the participants in a proceeding who pursue a certain subjective interest are particularly motivated to find arguments for their position in the language versions. However, the interest in objectively correct decisions will often also lead to a look into other language versions if the version primarily used is vague, otherwise unclear, or leads to results that contradict one’s own previous understanding.

43 Apart from seeking advice from tax lawyers regarding different language versions, taxpayers could, in principle, pursue their rights under art. 24(4) TFEU and art. 41(4) EU Charter of Fundamental Rights and enquire with EU institutions in an official language of their choice about linguistic divergences in EU law. The Commission undertook to provide such a reply within three weeks or at least to indicate within that period when it expected to be able to reply (Code of Good Administrative Behaviour, adopted by the Commission on 13 September 2000 [2000] OJ L267/63). 44 See Cases C-261/08 and C-348/08 Zurita García and Choque Cabrera, ECLI:EU:C:2009:648, para. 56; P. Pescatore et al., eds, ‘Interprétation des lois et conventions plurilingues dans la Communauté européenne’, Les Cahiers du Droit 25 (1984), 989, 997ff. 45 See Case C-567/10 Inter-Environnement Bruxelles et al., ECLI:EU:C:2012:159, paras 28ff. 46 The members of the Court and their staff cover all official languages with native speakers. It also has a scientific service and a translation service, which are staffed by lawyers from all EU member states. 47 See Case C-511/08 Handelsgesellschaft Heinrich Heine, ECLI:EU:C:2010:189, para. 51; Case C-412/10 Homawoo, ECLI:EU:C:2011:747, para. 28; M. Derlén, Multilingual Interpretation of European Union Law (Alphen aan den Rijn: Kluwer, 2009), 32–36. 48 This can be derived from the judgment in Case C- 298/ 12 Confédération paysanne, ECLI:EU:C:2013:630.

International Tax Law and Language 261

15.4 Discussion: Risks and Opportunities of Multilingualism in International Tax Law The traditional methods of interpreting law may in most cases in practice help to specifically resolve qualification conflicts or to avoid double taxation in general, but they cannot resolve language conflicts that result from the fact that two or more languages are authentic with respect to, for example, a double taxation treaty or an EU directive. Even in cases where it is theoretically clear that the law of just one state is decisive, as is, for instance, the case under article 6(2) sentence 1 OECD MC where the national law of the source state takes precedence, the other contracting state may still argue that from the viewpoint of the language used in that state the real property in question does not fall within the scope of that provision (apart from the fact that even the national law may not be precise in any case). The inevitable consequence from the perspective of that state would be that the source state proceeds with a taxation which is not in line with the tax treaty, and therefore the state of residence would probably not exempt the underlying rental income from taxation or would not grant a tax credit. This result is clearly to the detriment of the taxpayer, but is it also a good reason to abolish multilingualism from international tax law? The fact that there are several language versions available opens up opportunities for interpretation, as the wording boundary may be extended compared to a single language version, so it may in fact help to find out the ‘real’ meaning of a term. At the same time, of course, determining this wording boundary becomes the greatest practical challenge, since today it ideally requires the analysis of all authentic language versions that are applicable to a cross- border scenario. However, the chances of interpreting divergent versions are not entirely positive. For many users of law, as well as with respect to the rule-of-law principle of legal certainty and the balance of powers, they are primarily a challenge. This challenge lies in the difficulty of precisely determining the content of a regulation. Tax court rulings show that, in practice, in the case of linguistic divergences there is a large number of factors that may be decisive. Therefore, it is argued in academic literature that the vagueness of international law resulting from multilingualism is problematic with regard to human rights.49 Interventions in human rights generally require a sufficiently clear legal basis. If equally valid language versions of a regulation have different contents, one can certainly doubt this clarity.

49 T. Schilling, ‘Beyond Multilingualism: On Different Approaches to the Handling of Diverging Language Versions of a Community Law’, European Law Journal 16/1 (2010), 47, 66 (specifically for EU law).

262 Florian Haase At the very least, different language versions jeopardize the uniform application of international tax law. Indeed, if they do not suspect anything, legal practitioners normally trust and apply their own language version. Differences in language versions then inevitably lead to a different application of the law in question. The solution to these challenges lies formally in the interpretative competence of a tax court, as is the case with the Court of Justice with respect to the interpretation of EU law. Textual ambiguities in EU law open up the opportunity for the Court of Justice to provide legal clarity and to facilitate the uniform application of EU law. However, the exercise of this opportunity depends above all on the national courts fulfilling their responsibility to refer cases. The resulting strengthening of the judiciary, especially the Court of Justice, is also a challenge for the separation and balance of powers.50 Even if the Court of Justice makes intensive efforts to explore the ‘true will’ of the democratically legitimized legislature, especially by examining the history of the origins of a legislative act, the steering power of the legislature is diminished accordingly. The EU or contracting states to a double taxation treaty could meet these challenges by reducing the number of mandatory languages or even by giving a decisive weight to one language.51 The latter solution was, for instance, chosen in the Berne Convention for the Protection of Literary and Artistic Works, where the French text is decisive in disputes concerning the interpretation of the various texts.52 However, it is very difficult to imagine such an approach politically in the EU, nor would a contracting state of a tax treaty surrender to the binding effect of the language of the other country without need. In addition, there are also legal obstacles. Multilingualism is based on the consideration that, as a general rule, citizens of a country cannot be required to observe mandatory provisions laid down in a language that is foreign to them. This too is a manifestation of the principle of legal certainty. The prohibition of discrimination on the basis of nationality points in the same direction, since a reduction in the number of binding language versions would at least indirectly disadvantage those citizens for whom those languages are foreign. Accordingly, article 21(1) of the EU Charter of Fundamental Rights expressly prohibits any discrimination on the basis of language. Within the EU, to some extent, multilingualism is regarded as a constitutional principle,53 for which article 22 of the EU Charter must be consulted, which demands respect for the languages of the member states. In any event, the challenges in terms of legal certainty, uniform application, and the balance of powers oblige legislators to minimize linguistic divergences wherever possible. At the same time, however, there is an opportunity in the approach because

50

See E. Paunio, ‘The Tower of Babel and the Interpretation of EU Law: Implications for Equality of Languages and Legal Certainty’, in T. Wilhelmson et al., eds, Private Law and the Many Cultures of Europe (Alphen aan den Rijn: Kluwer, 2007), 401. 51 This has been, inter alia, suggested by T. Schilling, ‘Beyond Multilingualism’, 47, 65. 52 See art. 31(1)(c) of the Berne Convention for the Protection of Literary and Artistic Works of 9 September 1886, completed at Paris on 4 May 1896, revised at Paris on 24 July 1971. 53 Pescatore et al., ‘Interprétation des lois et conventions’, 989, 997ff.

International Tax Law and Language 263 efforts to ensure the coherence of language versions can help to improve the quality of regulations. These efforts are a further argument against an objection solution in combination with non-binding translations. Despite their best intentions, legislators as well as tax authorities would certainly be less committed to the quality of non-binding translations than to the coherence of multilingual law. One would therefore have to reckon with considerably more translation errors and with the associated difficulties of the uniform application of international tax law. The pros and cons of multilingualism demonstrate the difficulties in finding a straightforward solution. This is very unfortunate, because apart from the vagueness of language in general even serious translation errors do happen in practice quite often, particularly on an EU level where the quality of legislation should be rather high due to highly qualified ‘back-office’ staff. In the Proposal for a Council Directive laying down rules relating to the corporate taxation of a significant digital presence,54 for instance, as well as in the Proposal for a Council Directive on the common system of a digital services tax on revenues resulting from the provision of certain digital services,55 the term ‘ “entity” means any legal person or legal arrangement that carries on business through either a company or a structure that is transparent for tax purposes’. The English definition of the same term was therefore identical in both directives, but in the German version of the directives it was translated on the one hand as ‘ “Rechtsträger” eine juristische Person oder Rechtsgestaltung, die eine Geschäftstätigkeit über ein Unternehmen oder eine steuerlich transparente Struktur ausübt’ and as ‘ “Rechtsträger” bezeichnet eine juristische Person oder ein juristisches Konstrukt, die bzw. das entweder durch ein Unternehmen oder mittels einer für steuerliche Zwecke transparenten Struktur Geschäfte tätigt’ on the other hand. Another translation error occurred recently in the context of the Anti-Tax Avoidance Directive (ATAD).56 In article 7 paragraph (2) (a)(vi) of the English version, so-called invoicing companies are defined as companies that ‘earn sales and services income from goods and services purchased from and sold to associated enterprises’, whereas in the German version the connecting word ‘and’ is translated as ‘or’. If one looks at the obligations and duties that are imposed on taxpayers nowadays, particularly in the EU, and also the meticulousness that is demanded from them, this cannot be regarded as the mere negligence of translators and/or the legislator, since it causes real problems in practice. Tax authorities cannot be convinced easily that the version of a directive (or any other legal provision) in their own mother tongue actually contains a translation error, even if a comparison of the language versions proves them wrong (this was also the case with the German ‘or’ in the ATAD). So, what should be done in the light of this? Above all, legislators must pay attention in much greater detail to minimize translation errors. Secondly, states should agree 54

COM(2018) 147 final (21 Mar. 2018). COM(2018) 148 final (21 Mar. 2018). 56 Council Directive (EU) 2016/1164, 12 July 2016; see ‘Cross-Border VAT Aspects: The EU Approach and Evolving Trends’, Section 40.5 in this volume. 55

264 Florian Haase to just one authentic language wherever politically possible. This is probably easier to achieve with double taxation treaties than in EU law, but it is at least worth a try. Thirdly, another chance lies in broadening the definition section contained in double taxation treaties, as we know it from contracts that are concluded in common law countries and that usually contain extensive definitions of the terms relevant for the contract. This would also accord with the convention that the rules of a tax treaty should be interpreted autonomously. Another idea is the abolition of the general ‘renvoi rules’ in article 3 paragraph 2 OECD MC. Double taxation treaties should not allow the possibility of interpreting terms of an agreement under the national law of a contracting state, because this contradicts the character of the treaty as a contract under public international law. At least contracting states should agree on the material scope of the reference which is made to the laws of the contracting states. It should be clarified whether this reference only covers tax provisions or also other legal regimes and provisions of the legal system in question. It is also conceivable to establish an ‘interpretation panel’ in the sense of a standing committee with delegates from both contracting states or with neutral persons such as tax law professors, who come together once or twice a year to discuss and to reconcile general or even specific interpretation difficulties. Last but not least, the aforementioned should be accompanied by the long-awaited implementation of a ‘world tax court’ which would have final jurisdiction over the interpretation of tax statutes and agreements that are applicable in cross-border scenarios. At least countries should be encouraged to make more use of the possibility to install the CJEU as an arbitrator, just as the parties did successfully in Republic of Austria v Federal Republic of Germany.57

57

Case C-648/15, 12 September 2017.

Chapter 16

C omparative Tax L aw Marco Barassi

16.1 Introduction The birth of modern legal comparison goes back to the period between the end of the nineteenth century and the beginning of the twentieth.1 During the first international conference on comparative law, held in Paris in 1900, there were discussions on the object, nature, methods, and applications of comparative law, and its place among the social sciences.2 The topics arose because, unlike other legal fields, comparative law is not concerned with a specific area of law3, for example private law, constitutional law, or tax law: comparatists do not study comparative statutes that do not exist but instead compare rules, institutions, and legal orders.4 Comparative law has been defined as an intellectual process that has law as its object and comparison as its instrument;5 hence, why scholars have long debated whether comparative law is a science or a method. The goal then became that of overcoming a lack of tangible consequences.6 Comparative law is, in fact, both science (since it is taught in university courses) and method (because, for

1 R. David and C. Jauffret-Spinosi, I grandi sistemi giuridici contemporanei (Padua: Cedam, 2004), 3; A. Pizzorusso, Sistemi giuridici comparati (Milan: Giuffrè, 1998), 145. 2 David and Jauffret-Spinosi, I grandi sistemi giuridici contemporanei, 3. 3 A. Watson, Il trapianto di norme giuridiche. Un ‘approccio’ al diritto comparato (Naples: Edizioni Scientifiche Italiane, 1984), 1. 4 As O. Kahn-Freund states, the subject of comparative law ‘has by common consent the somewhat unusual characteristic that it does not exist’ (‘Comparative Law as an Academic Subject’, Law Quarterly Review 82 (1966), 41). 5 K. Zweigert and H. Kötz, Introduzione al diritto comparato (Milan: Giuffrè, 1998), 2. 6 On the futility of the issue, see P. Stanzione, ‘Introduzione’, in M. Ancel, ed., Utilità e metodi del diritto comparator (Naples: Jovene, 1974), XIIff, with a reference list; R. David and J. E. C. Brierely, Major Legal Systems in the World Today (London: Stevens & Sons, 1954), 4.

266 Marco Barassi comparison, one need to use specific methods that differ from those used by jurists who study the law of a single jurisdiction).7 Comparative law was mainly developed by private law scholars and so, when it spread to fields of law such as constitutional law, criminal law, or administrative law, the issue was whether the approach to comparison followed by private lawyers (e.g. the methods and classification of different legal orders into families) was also useful in those other areas of law. The theoretical questions included whether the purposes, methods, and classifications of legal families were the same or whether they differed according to the area of law. The same questions apply to comparative tax law.

16.2 Comparative Tax Law Comparisons among different tax jurisdictions have long been made. In the past, scholars would read and quote from foreign colleagues and foreign law,8 especially among countries with cultural ties. Examples of the use of comparisons in tax law include the imitation of a foreign model that may provide the solution to a particular tax system’s problem or their use to attain a broader knowledge of the rules and institutions of foreign tax systems. Although the use of comparisons in tax law had long been common, theoretical reflections on comparative tax law are a more recent phenomenon9 as scholars began to question whether simply describing foreign tax rules was, in fact, comparative tax law. 10 What are its purposes, what is it, and what is its object? How should comparisons be made?11 These topics will be discussed in more detail in the following subsections.

7 A. Gambaro, P. G. Monateri, and R. Sacco, ‘Comparazione giuridica’, in Digesto Discipline privatistiche sezione civile (Turin: Utet, 1988), 52. 8 For examples of Italian scholarship, mainly related to indirect taxes (registration duty), see A. Malgarini, Della coordinazione delle imposte secondo la legislazione comparata (Parma: Luigi Battei, 1885); G. Vignali, Le tasse di registro nella teoria e nel diritto positivo italiano (Milan: Società Editrice Libraria, 1907); A. Uckmar, La legge del registro (Padua: Cedam, 1958). 9 K. Brooks, ‘An Intellectual History of Comparative Tax Law’, Alberta Law Review 57/3 (2020), 649, traces the history of comparative tax law outlining the different phases through which it developed. 10 J. M. Moessner, ‘Why and How to Compare Tax Law’, in Liber amicorum Luc Hinnekens (Brussels: Bruylant, 2002), 306; also in C. Sacchetto and M. Barassi, eds, Introduction to Comparative Tax Law (Soveria Mannelli: Rubettino, 2008), 13. 11 Examples are O. Y. Marian, ‘Discursive Failure in Comparative Tax Law’, American Journal of Comparative Law 58 (2010), 415; M. Barassi, ‘Comparazione (dir. Trib.)’, in S. Cassese, ed., Dizionario di diritto pubblico (Milan: Giuffrè, 2006), 1070; P. Selicato, ‘La comparazione nel diritto tributario: riflessioni sul metodo’, in Dal diritto finanziario al diritto tributario. Studi in onore di Andrea Amatucci (Bogotà and Naples: Editorial Temis SA, Jovene Editore, 2011), 37; for further references, see Brooks, ‘An Intellectual History of Comparative Tax Law’.

Comparative Tax Law 267

16.2.1 The Aim of Comparative Tax Law: Why Compare? The basic purpose of comparative law—including, of course, comparative tax law—is to attain knowledge of foreign and national law.12 Comparative law studies different rules and institutions to establish to what extent they are similar or different.13 A curious and well-known result of comparative law is that it may lead to a better understanding of the law of one’s own country and its evolution in comparison with other legal systems, as the structures hidden in one system (if viewed from the perspective of national doctrine) may be more visible in another system14 and therefore be easier to identify. Comparative tax law may also have specific and practical aims. International business clearly requires a knowledge of foreign tax laws, even though the study of a foreign tax law is not actually ‘comparative’ tax law, as will be explained in more detail later. An example of this is where a tax lawyer in country A has a question for a tax lawyer in country B and would obviously find it helpful to have a basic understanding of the tax law of country B.15 Comparative tax law can clearly be seen in the EU context, where the harmonization and approximation of EU legislation (e.g. concerning dividends, restructuring operations such as mergers, and divisions) involved a comparative analysis of the domestic laws of the member states. Comparative tax law also finds extensive application in the imitation of foreign legal rules or institutions and in legal transplants,16 as changes in tax law are often made by transplanting foreign tax rules into a domestic system.

16.2.2 The Object of Comparative Tax Law: What to Compare? Since comparative tax law involves a comparison, it requires the comparison of homogeneous objects. Tax law as a field of law encompasses rules relating to the creation and application of taxes. Therefore, the focus here is the notion of tax.17 A tax is usually

12

P. Arminjon, B. B. Nolde, and M. Wolff, Traité de Droit Comparé, vol. I (Paris: Librairie Génerale de Droit et de Jurisprudence, 1950), 15; Zweigert and Kötz, Introduzione al diritto comparato, 17; R. Sacco, ‘Legal Formants: A Dynamic Approach to Comparative Law’, American Journal of Comparative Law 39 (1991), 1–6. 13 Sacco, ‘Legal Formants’, 5. 14 Similarly, Moessner, ‘Why and How to Compare Tax Law’, 306. 15 Ibid. 16 M. Barassi, ‘La circolazione dei modelli tributari e la comparazione’, in C. Sacchetto, ed., Principi di diritto tributario europeo e internazionale (Turin: Giappichelli, 2016), 47ff; I. J. Mosquera Valderrama, ‘The Study of the BEPS 4 Minimum Standard as a Legal Transplant: A Methodological Framework’, Intertax 48/8/9 (2020), 719. 17 In some civil law countries, compulsory contributions are defined by law, judges, or scholarship; in common law countries, it seems that there is a more cautious approach to providing a definition: G. Morse and D. Williams, Davies: Principles of Tax Law (London: Sweet & Maxwell, 2004), 3, underline three characteristics of tax: ‘compulsory levy imposed by an organ of government for public purposes’,

268 Marco Barassi levied to raise public revenue18 to allow a state to pursue its aims and, to levy a tax, a legal basis is required. That principle is written in the constitution of many countries.19 The question then arises of whether the taxes in question include the same type of levy. A typical problem of comparison concerns the translation of the relevant texts.20 In some countries, a general word is used to mean fiscal levies (or compulsory contribution):21 tributo in Italy and Spain, abgabe in Germany. In these countries, tributo and abgabe mean fiscal levies that consist mainly of taxes (imposta, impuesto, steuer) and fees (tassa, tasa, gebühren).22 Taxes, as opposed to fees,23 form a relatively homogeneous group: they are compulsory, because the taxpayer is unable to choose whether or not to pay once the taxable event occurs; their main aim is to raise revenue for government expenses and to allow the government to promote its general interests; and they are paid without any benefit received by an individual taxpayer in exchange for the payment. Nevertheless, differences do exist. For example, in some countries social contributions are classified as taxes, whereas in others they are not.24 Therefore, if the word ‘tax’ indicates pure tax—such as income tax or VAT where nothing is received by the individual taxpayer in return for the payment25—the meaning of tax may change from one country to another, revealing one of the basic risks of comparison; that is, the same words having a different meaning in different jurisdictions. For example, a comparison of the income taxes of two countries may reveal that the notions of income are different; one has a comprehensive notion of the income tax base, or an accretion concept (e.g. the USA),26 and the other a source concept (e.g. many European countries).27

quoting a ‘rare judicial discussion’; A. Shipwright and E. Keeling, Textbook on Revenue Law (London: Blackstone Press, 1998), 2ff, underline the difficulty in defining tax. 18 There

may be exceptions, e.g. when taxes are levied for different aims such as protection of the national economy with import duties. 19 F. Vanistendael, ‘Legal Framework for Taxation’, in V. Thuronyi, ed., Tax Law Design and Drafting (Washington, DC: International Monetary Fund, 1996), 16. 20 On translation problems, see Sacco, ‘Legal Formants’, 10ff; V. Thuronyi, K. Brooks, and B. Kolozs, Comparative Tax Law (Alphen aan den Rijn: Wolters Kluwer, 2016), 4, with examples of tax terms that may have different meaning: capital gains, income, partnerships, trusts, dividend, and tax avoidance; C. Sacchetto, ‘Translation and Tax Law’, in R. Sacco, ed., L’interpretation des textes juridiques rédigés dans plus d’une langue (Turin: L’Harmattan Italia, 2002), 203ff. 21 According to the expression suggested by Thuronyi, Brooks, and Kolozs, Comparative Tax Law, 41. 22 For a discussion on the meaning of ‘tax’, see ibid., 41ff. Other examples of expressions that may have different meanings are provided by M. Livingston, ‘Law, Culture and Anthropology: On the Hopes and Limits of Comparative Tax’, Canadian Journal of Law and Jurisprudence 18/1 (2005), 119, 121 who cites capital gains and non-taxable reorganizations. 23 M. Barassi, ‘The Notion of Tax and the Different Types of Taxes’, in B. Peeters, ed., The Concept of Tax (Amsterdam: IBFD, 2005), 64ff.; L. Del Federico, ‘The Notion of Tax and Para-Commutative Taxation’, ibid., 74ff. 24 V. Thuronyi, ed., Tax Law Design and Drafting (Alphen aan den Rijn: Wolters Kluwer, 2000), 332. 25 As it is well known, this is not the case for social contributions. 26 Thuronyi, Brooks, and Kolozs, Comparative Tax Law, 213. 27 The source concept refers to periodic income from a source.

Comparative Tax Law 269 The boundaries of comparative tax law, however, are not easily defined.28 Comparative tax law may be applied to income tax, VAT, inheritance tax, or other taxes or compulsory contributions whilst also taking account of the specific rules, institutions, and tax structures of the systems being compared. Unlike taxes, the category of fees seems to be less homogeneous. Their main characteristic is that the taxpayer receives something in return for their payment, for example the issue of a licence or passport or payment in return for a public administration’s duty to act (e.g. action by the relevant body following payment for said passport). It is not easy to distinguish between a compulsory contribution that falls inside the definition of a fee and what falls outside the category (e.g. a charge). The same service may be financed in one country by a fee and in another by a charge.29 The difference lies in the divergent legal rules that apply to a fee and to a charge. Comparison may concern detailed elements of a tax or follow a more general approach. Comparative analysis has recently been addressed to tax incentives for international donations30 and has previously covered personal income tax31 and various aspects of income tax such as fringe benefits, deductions, and business organization.32 Lastly, the object of comparative tax law may depend on knowledge of a foreign language, the availability of the sources of law, and the interests or work of a single scholar as opposed to a team of researchers.

16.2.3 How to Compare: The Methods of Comparative Tax Law With the development of studies on comparative tax law, scholars began to analyse the methods to apply in this field.33 Methods refer to the ‘how’ to compare (i.e. the procedure followed in the comparison) and encompass the identification of sources of law, the study of foreign law, and the real comparison. Thus, it is necessary to apply a 28

As Marian, ‘Discursive Failure’, 452ff, explains. Beltrame and L. Mehl, Techniques, politiques et istitutiones fiscales compares (Paris: Puf, 1984), 40, state that France municipalities may choose between a waste fee (that is established by law, and is levied even if the subject does not use the service provided) and a redevances d’enlèvement des déchetes calculated based on the actually services provided. 30 R. Buijze, ‘The Categorization of Tax Jurisdiction in Comparative Tax Law Research’, Erasmus Law Review 4 (2016), 189ff. 31 Livingston, ‘Law, Culture and Anthropology’, 119. 32 These are only a few examples; many articles and books address selected objects of comparative analysis. See H. Ault, B. Arnold, and G. Cooper, Comparative Income Taxation: A Structural Analysis (Alphen aan den Rijn: Wolters Kluwer, 2020),, Part II. 33 M. Barassi, ‘Comparazione giuridica e studio del diritto tributario straniero’, in Uckmar (a cura di), Corso di diritto tributario (CEDAM, Padova, 2005), 1499; Barassi, ‘Comparazione (dir. trib.)’, 1070ff; C. Garbarino, ‘An Evolutionary Approach to Comparative Taxation: Methods and Agenda for Research’, American Journal of Comparative Law 57/3 (2009), 677; Marian, ‘Discursive Failure’, 415; Moessner, ‘Why and How to Compare Tax Law’. 29 P.

270 Marco Barassi method, select the subject of comparison, identify the sources of law (statutes, case law, tax authorities’ official views), and decide how to study the sources and interpret the results (comparison of a foreign legal rule with a domestic one in order to reveal similarities and differences). However, the method or methods used are not always expressly mentioned; in other words, the author compares but does not always explain how the comparison was carried out. On the other hand, some authors specifically analyse comparative tax law theory or methods.34 Academics have traced the history of comparative tax law and identified the different phases of its development, outlining the authors who contributed to the development of comparative tax law, including those with an interest in the methods used.35

16.2.3.1 Method Method is crucial in comparative law as it has proper characteristics compared to the methods typically used by domestic jurists. While a domestic jurist studies their country’s law, a comparatist jurist studies and compares different laws or legal orders. In doing so, a comparatist is a ‘neutral observer’ of the foreign law: they take note of the legal rules actually applied, including whether they differ from the rules set down by statute. Study of the object—the foreign law—requires specific techniques: a domestic jurist in studying their own law is influenced by the concepts and implicit unwritten rules which form part of the culture of the jurist, who considers certain elements to be obvious and uses and applies them unconsciously. These elements are called cryptotypes36 and would be clear to a foreign jurist. Examples of cryptotypes in tax law would be the principle of taxpayer reliance37 or the highly legalistic approach to tax rules in some countries that can make it impossible to develop anti-avoidance doctrines while in other countries the principle of substance over form is applied.38 Another successful result of comparative law scholarship is the ‘legal formant’.39 As Rodolfo Sacco states, a local jurist seeks one rule: identifying which, among statutes, case law, and scholarship is the correct rule. Whereas a comparatist jurist looks for the different legal formants that have an influence on the rules applied. In tax law, legal formants are statutes, case law, and the opinions of scholars and of tax authorities.40 Legal formants are the sources of law; however, that expression in comparative law has an operational meaning rather than a formal one41—everything that affects the relevant law (the so-called law in action) is a source. This means, for example, that in civil law countries the decisions of the courts are not, formally speaking, sources of law—as they 34

See all references ibid. Brooks, ‘Intellectual History of Comparative Tax Law’. 36 The use of cryptotypes in comparative law is due to Sacco (‘Legal Formants’, 384ff). 37 Barassi, ‘Comparazione (dir. trib.)’, 1072. 38 Garbarino, ‘Evolutionary Approach to Comparative Taxation’, 697. 39 Sacco, ‘Legal Formants’, 22ff. 40 These are the most important formants in the Western legal tradition. 41 For sources of law in comparative law, see David and Brierley, Major Legal Systems in the World Today, 14. 35

Comparative Tax Law 271 are not binding—but they are nonetheless of great importance since lawyers, taxpayers, tax authorities, and judges take court decisions into account even if they are not legally binding. This similarly applies to tax authority opinions issued in the form of circulars, rulings, and academic writings. In addition, the sources of law must be interpreted according to the criteria applicable under the foreign law and not according to the criteria of the comparatist’s own law. A comparative analysis may show that legal formants in a jurisdiction converge or diverge. An example is provided by the agreements between taxpayers and tax administrations (tatsaechliche Verstaendingung) in German tax law.42 Although these agreements are not covered by statute, they are taken into account in the case law, scholarship, and tax administrations. Thus, tax law is based on the principle of legality and belongs to public law rather than private law. It is in the latter where agreements between parties are relevant. In a civil law country, where statute law is the most important source of law, legal formants that differ from statutes and provide solutions using efficiency and compliance with applicable principles are relevant. Thus, formants diverge, since case law (as well as scholarship and tax administrations) recognizes that which a statute omits (agreements). The result is that formants will have been influenced by elements other than those which influenced the statutes. If agreements between taxpayers and the tax administration were not taken into consideration at the time a statute was drafted, but formants later recognized the usefulness of those agreements for a more effective tax system, those agreement are applied despite thei lack in the statute., . The theory of legal formants has a practical application with the so-called law in action method rather than law in the books.43 While law in action is law concretely applied and may be influenced by elements other than the formal sources of law, including tax culture,44 law in the book is law formulated in statutes and codes.45 An interesting example of this method relating to tax law is provided by the extra-statutory concessions in the UK:46 in a legal order that proclaims parliamentary supremacy and requires the consent of Parliament to impose taxes,47 rules enacted by the tax

42

Tatsaechliche Verstaendingung is dealt with in K. Tipke and J. Lang, Steuerrecht (Cologne: Dr. Otto Schimdt KG, 2018), 1310ff. 43 The expression, as noted, is from Roscoe Pound’s article, ‘Law in Books and Law in Action’, American Law Review 44/1 (1910), 12. 44 For examples of tax culture, see Marian, ‘Discursive Failure’, 461, quoting M. A. Livingston, ‘From Milan to Mumbai, Changing in Tel-Aviv: Reflections of Progressive Taxation and “Progressive” Politics in a Globalized But Still Local World’, American Journal of Comparative Law 54/3 (2006), 555, and in particular, M. A. Livingston, Tax and Culture. Convergence, Divergence, and the Future of Tax Law (Cambridge: Cambridge University Press, 2020). 45 G. Gorla, ‘Diritto comparato’, in Enciclopedia del diritto (Milan: Giuffrè, 1963), 932ff. 46 Extra-statutory concessions are ‘departures from the strict letter of the law, but obviously always in favour of the taxpayer. Their purpose is to allow relief which is in the spirit of a particular piece of legislation but which, for some reason or other, is not actually given by it’: Shipwright and Keeling, Textbook on Revenue Law, 28. 47 A constitutional principle in the UK is that the Bill of Rights of 1688 provides that taxes cannot be imposed without the consent of Parliament: Morse and Williams, Davies, 35.

272 Marco Barassi authority can modify statute when the statute produces an overly onerous burden on the taxpayer.48

16.2.3.1.1 Micro and macro-comparisons Comparisons can be made at the micro or macro level. The former is targeted at rules and institutions, while the latter concerns the tax system as a whole.49 An example of a micro-comparison is a comparison of different group taxation solutions (or anti- avoidance rules or rules concerning the deduction of interest) where the analysis is detailed and in-depth. In a macro-comparison, the level of research is more general and the analysis aims to discover the main characteristics and permanent elements of a tax system. Macro-comparisons thus classify tax systems into families, grouping together those systems with similar features. The classification of legal systems into families is well known in comparative law scholarship50 and depends on the area of law being studied. As classifications were first undertaken by private law comparative scholars, the results may differ where classification occurs in other fields of law51 and may also differ at different points in time.52 For instance, Victor Thuronyi proposed a classification for tax law53 that was extremely wide, included almost all countries, and was carried out with reference to the tax system as a whole. Some elements used by private law comparatists to classify legal families are also relevant for classifications in tax law: while the outlook in civil law countries is formal and systematic, statutes are drafted in greater detail in common law countries.54 Classifying tax law as a legal family retains its main characteristics and permanent elements which are not affected by frequent changes in tax legislation (e.g. those contained in Finance Acts) which usually affect single rules and not the most fundamental elements of a legal order. In addition to classification based on the elements mentioned earlier, scholarship has highlighted both that taxation has its own taxonomy and that classification may follow

48 Ibid.,

37, who define extra-statutory concessions as ‘truly curious’ since they are not law but, in practice, prevail over law. 49 The notions of micro-and macro-comparisons are debated. While Zweigert and Kötz, Introduzione al diritto comparato, 5, refer to macro-comparison as the analysis of drafting legislation, style of case law, interpretation of law, etc., Gambaro and Sacco, Sistemi giuridici comparati, 12, define this as sistemology—i.e. the analysis of legal orders to clarify the main rules that characterize a legal order beyond the detailed differences. Comparative tax law examples are provided by Marian, ‘Discursive Failure’, 428 and 449ff. 50 Famous examples of classification are those provided by Arminjon, Nolde, and Wolff, Traité de Droit Comparé; R. David, Le grand systèmes de droit contemporains (Paris: Dalloz, 1964). 51 Zweigert and Kötz, Introduzione al diritto comparato, 80; Thuronyi, Brooks, and Kolozs, Comparative Tax Law, 20. 52 Zweigert and Kötz, Introduzione al diritto comparato, 82. 53 Thuronyi, Brooks, and Kolozs, Comparative Tax Law, 21ff. 54 Ibid., 23.

Comparative Tax Law 273 the normative ‘defining elements’ of equity and redistributive justice (with reference to democratic taxation and income tax).55 The two approaches (one stating that some elements relevant in private law comparisons are also relevant in tax law and the other indicating the ‘defining elements’ of taxation) are not incompatible and one does not preclude the other. According to first approach, classification assumes that tax law is part of a legal system with which it shares certain elements (e.g. drafting statutes, the training and reasoning of judges), whereas the second approach is based on the defining elements of a tax system which usually apply to a specific area of tax law (i.e. income tax, with reference to equity and distributive justice) rather than to tax law as a whole. Depending on the perspective, a tax system may be classified as belonging to one group (with which it shares similar elements concerning drafting of statutes, etc.) or as belonging to another (those systems with similar defining elements).

16.2.3.1.2 The steps of comparison Comparing tax law necessitates that the object to be studied must be identified, different tax laws must be studied, and they must then be compared to highlight similarities and differences; some scholars add to this an explanation of the similarities and differences56 and an evaluation of the laws compared.57 The word ‘method’ covers the procedures and instruments of comparison, and includes techniques for acquiring a knowledge of foreign law (e.g. by studying statutes or court decisions, issuing questionnaires to relevant jurists, considering problems of translation), the degree of detail that the specific research requires (micro or macro- comparisons), and the criteria to be used in comparative analysis. As mentioned, when comparative law spread from private law to other areas of law, comparatists faced the problem of method: are private law methods also useful in other areas of law or does each area require a specific method? Although there are different approaches to the comparative tax law method, those followed in private comparative law (as evidenced by the frequent reference to private comparative law scholarship) appear to prevail.58 This is due to the fact that tax law is a well-established area of law 55 W.

Baker, ‘Expanding the Study of Comparative Tax Law to Promote Democratic Policy: The Example of the Move to Capital Gains Taxation in Post-Apartheid South Africa’, Penn State Law Review 109/3 (2005), 713. 56 L. J. Constantinesco, Die rechtsvergleichende Methode, Band II, Rechtsvergleinchung (Cologne: Heymanns, 1972); A. Procida Mirabelli di Lauro, ed., Il metodo comparativo (Turin: Giappichelli, 2000), 10. 57 G. Gorla, ‘Diritto comparato e straniero’, in Enciclopedia Giuridica (Rome: Treccani, 1989), 3ff; Zweigert and Kötz, Introduzione al diritto comparato, 7. 58 Thuronyi, Brooks, and Kolozs, Comparative Tax Law, 3ff, outlining the comparative method with reference to well-known comparative scholars; V. Thuronyi, ‘Studying Comparative Tax Law’, in L. Gustaf et al., eds, International Studies In Taxation: Law and Economics. Liber amicorum Leif Mutén (Alphen aan den Rjin: Kluwer Law International, 1999), 340: ‘There is no reason why the [private comparative law] methodology cannot be applied to tax law’; Ault, Arnold, and Cooper, Comparative Income Taxation, in the introduction, after quoting Otto Kahn-Freund’s definition of comparative law as a method of looking at one’s own law, state that in general Kahn-Freund’s view is equally applicable

274 Marco Barassi and therefore the key elements of legal comparison methods apply (e.g. identifying the object of comparison, the sources of law, their translation and interpretation, etc.). However, since tax law is a separate area of law, it is governed by specific principles and rules (e.g. the principle of legality) that may have an effect on method. From a theoretical perspective, two approaches prevail. Some scholars believe that even in comparative tax law, the general comparative method can be used and, based on that, they underline the general characteristics of tax law, highlighting the peculiarities and difficulties of tax law analysis and its comparison;59 other scholars move from a comparative law approach to methodology (though not always in the same way), noting the peculiarities of tax law and constructing a bespoke comparative law methodology for tax law.60 Moving on from these different approaches, but bearing in mind the steps of comparison—particularly the study of foreign law and comparative analysis—in the author’s opinion, comparative tax law requires some ‘methodological indications’ (as opposed to ‘methods’ which differ from those applied in comparative law) as set out below.

16.2.3.1.3 The study of foreign law Studying a foreign tax law is difficult as tax law is complex and frequently changes.61 Once general information on the foreign legal system has been acquired, foreign tax law can be studied in two ways: by analysing the legal formants, if possible in the original language, and by reading publications by domestic jurists. While the second is easier,62 if the first is done with the awareness of comparison (unlike the work of a domestic jurist) it provides a wider understanding of the implicit and non-verbalized elements of the system (the so-called cryptotypes).63 As mentioned earlier, this is due to the fact that notions that may be implicit or non-verbalized in one legal order, may be explicit or verbalized in another legal order and thus recognized by the comparatist.

to tax law; Garbarino, ‘Evolutionary Approach to Comparative Taxation’, 679, refers to comparative law scholarship. 59 Ault,

Arnold, and Cooper, Comparative Income Taxation, Introduction; Thuronyi, Brooks, and Kolozs, Comparative Tax Law, 5ff, deal with the comparative methods applied to tax law and the peculiarities of tax law that make comparisons difficult. 60 It is not possible here to give an account of the varying opinions of scholars: however, see in general, Moessner, ‘Why and How to Compare Tax Law’, 311ff; F. Vanistendael, ‘The Merits and Demerits of the Comparative Study of Taxation’, presented at the conference ‘La comparazione nel diritto Tributario: metodi, applicazioni e orientamenti in ambito internazionale’, Rome, 25–26 March 2002; Marian, ‘Discursive Failure’, 456ff; Barassi, ‘Comparazione (dir. trib.)’, 1073ff; Baker, ‘Expanding the Study of Comparative Tax Law’, 706–707. 61 V. Thuronyi, ‘Comparative Tax Law’, in Encyclopaedia of Comparative Law (Cheltenham, UK: Edward Elgar Publishing, 2006). 62 In some cases, the support of a textbook is unavoidable since statute law is highly complex. 63 Gambaro, Monateri, and Sacco, ‘Comparazione giuridica’, 53.

Comparative Tax Law 275 Moreover, direct knowledge of a foreign law’s legal formants enables the style of the legal order to be understood, including the drafting of statutes and the reasoning of judges, scholars, and tax authorities.64 In studying foreign law, the rules of interpretation of the foreign law must be used, setting aside those of one’s own jurisdiction, including its concepts and framework. One particular characteristic of tax law should be noted here. Unlike private law, tax law does not cover relationships between private parties but takes as taxable events those facts or acts that have an economic basis and may be governed by another field of law.65 Since the rules of the other field of law may also affect the tax rules, comparative tax law requires a knowledge of the foreign law not limited only to the tax law. Knowledge of other disciplines may therefore be required.66 For example, a civil law jurist studying UK taxation of income from land will need to have knowledge of UK property law;67 or it may be necessary to study inheritance law to understand inheritance tax; or, when accounting and taxation is at issue, a knowledge of accounting rules is usually necessary.

16.2.3.1.4 Comparative tax law analysis In the view of this author, there are several methods of analysis that can be applied in comparative tax law which do not differ from the methods used in comparative law. It is first necessary to identify the most appropriate method for the object and then to apply the analysis to tax law principles. Starting from a macro-comparison, the comparative analysis may have as its purpose the classification of tax systems into families, in which event a relatively large number of systems will need to be examined. The outcome of the classification will depend on the similarities and differences detected; classification may be based on single taxes (e.g. income tax68 and wealth tax)69 or on tax systems as a whole.70

64 Zweigert and Kötz, Introduzione al diritto comparato, 76ff and 83ff, where they indicate the elements that characterize the legal order’s style. It is certainly true, as Barker states, that the style of legal families may be less important in tax law compared with private law since the former is a recently new field of law, but the differences in style of legal orders are quite clear when comparing statutes, case law, or textbooks: e.g. the approach in civil law countries tends to be more systematic and refers to abstract concepts (‘Expanding the Study of Comparative Tax Law’, 712). As Thuronyi, Brooks, and Kolozs underline, these characteristics are also found in tax law (Comparative Tax Law, 23). 65 Furthermore, tax law is part of public law which focuses on the duties and relationships between individuals, government, and society, as Barker explains in ‘Expanding the Study of Comparative Tax Law’, 707. 66 Thuronyi, Brooks, and Kolozs, Comparative Tax Law, 6 underline the interdisciplinarity of tax tradition and the need for knowledge of the general legal tradition. 67 It is not possible to give further detail here as UK property law has a long history dating back to the Middle Ages and includes issues such as estate and tenure that are not easily compared to analogous institutions in a civil law country. 68 Thuronyi, Tax Law Design and Drafting, 479. 69 M. Lehner, ‘The European Experience with a Wealth Tax: A Comparative Discussion’, Tax Law Review 53/4 (2000), 615ff. 70 Thuronyi, Brooks, and Kolozs, Comparative Tax Law, 19.

276 Marco Barassi Using a micro-comparison, the comparative analysis may have as its object a more limited number of countries, specific rules, or institutions. One type of method that is useful in different fields of comparative law, including comparative tax law, is a historical one. In tax law, a historical method may be used to understand the rules and institutions from their development over time which allows for more in-depth knowledge of the subject: for example, history can help to understand the different notion of income in the UK as compared to that in the USA.71 The functional method allows comparison purely of legal rules which perform the same functions.72 Examples are the rules governing the taxation of business income and their relationship with financial statements73 or the existence of agreements between a taxpayer and the tax authority to resolve a dispute without recourse to the courts (e.g. the conditions for concluding the agreement and the legal effect the agreement produces). This method may also be useful in comparative tax law as once a tax problem has been highlighted, the comparatist will need to identify the solutions provided by different tax systems. Solutions in different systems to the same problem may converge or diverge.74 If the object of comparison is a single tax, other methods may be more useful. For example, in comparing income tax it is useful to compare: (1) the concept of income; (2) the inclusion of income realized or accrued; and (3) the relevance of capital gains. This approach will be structural since it concerns the structure of the tax. The factual approach, developed by Rudolf Schlesinger during seminars at Cornell University,75 consists of questionnaires sent to jurists from different countries. The method is also known as a common core comparison because in comparative private law convergent solutions from different countries have been highlighted, notwithstanding that there may be formal differences. Apart from the common core approach,76 the method of acquiring knowledge of foreign law through questionnaires sent to domestic jurists is also useful in comparative tax law, but conditional on asking questions in 71 In fact, statutes reflect the cultural and historical differences between two countries. Since it was introduced in 1864, US income tax has applied to ‘all income from whatever source derived’ (Internal Revenue Code, § 61), whereas UK income tax only applies to those categories specified by statute (schedular system). Likewise, since its introduction, US income tax has applied to capital gains whereas, in the UK, capital gains tax was introduced only in 1965; the difference being due to the history of the circulation of assets, particularly land, which was common in the USA but limited in the UK at the time income tax was introduced. Since the Middle Ages, land in the UK has been a primary source of wealth and was traditionally income-producing and not sold; the idea of a gain arising from the sale of an asset such as real property was therefore not a tradition in the UK. In the USA, land was frequently sold therefore the idea of realizing a gain from the sale of land reflected economic reality; see K. Holmes, The Concept of Income: A Multidisciplinary Analysis (Amsterdam: IBFD, 2001), 221. 72 Zweigert and Kötz, Introduzione al diritto comparato, 37. 73 See M. Grandinetti, ed., Corporate Tax Base in the Light of the IAS/IFRS and EU Directive 2013/34: A Comparative Approach (Alphen aan den Rijn: Wolters Kluwer, 2016) . 74 E.g. group taxation. 75 R. B. Schlesinger, ed., Formation of Contracts: A Study of the Common Core of Legal Systems (Dobbs Ferry, NY: Oceana Publications, 1968). 76 Although there is a convergence of solutions, as noted later, tax systems provide different solutions to similar problems: group taxation, interest deduction, etc.

Comparative Tax Law 277 concrete terms. For example, asking: (1) how fringe benefits are taxed; (2) whether capital gains include only gains realized or also those accrued; (3) which legal instruments are used to resolve disputes between taxpayer and tax authority, etc.77 Questions must therefore refer to a tax law problem but set aside that law’s conceptual structure. In law and economics, legal events are analysed on a cost–benefit basis, with rules and institutions studied for the economic effects they produce. Law and economics may be applied to tax law,78 for example, in relation to enforcement of tax. As well as the legality of a possible solution to a tax problem, its efficiency (on a cost–benefit basis) should also be considered. This approach could affect the choice of jurisdictional dispute-resolution models or agreements between a taxpayer and the tax authority. There is one further point of note in relation to methodology which has been considered by many scholars, although in different ways and with different emphases: the relevance of tax law principles in comparative tax law. As mentioned in Section 16.2.3.1.1, some scholars believe that since it is a separate field of law, tax law is characterized by specific principles, such as the rule of law, equality, ability to pay, coherence of tax system, prevention of abuse of law.79 Whereas other scholars follow a more ideological approach80 and, with reference to income tax law, point out the relevance of an income tax system’s defining elements such as equity and redistributive justice as normative underpinnings of democratic taxation.81 These principles and defining elements are instruments that can be used to compare different tax laws with a reference model (so-called tertium comparationis): this affects both the activity of classifying the rules and institutions into homogeneous categories and the evaluation of them against a reference model. The first consists of classifying the rules and institutions of different laws into groups, each of which have similar characteristics: for example, countries that limit corporate interest deduction with a fixed ratio rule or that subject it to a thin capitalization rule. The second activity may follow the first and provide an evaluation of each group of rules and institutions (models) according to its compliance or otherwise with the above-mentioned principles. Is the model compliant with the equality principle? Does it prevent abuse of law? Is it manageable by the taxpayer and the tax administration?

16.2.3.1.5 A comparative tax analysis: family taxation Family taxation is an example of the methodological approach described above.82 Family is relevant from different perspectives. First, it is a natural fact that has a human 77 Examples of questions in concrete terms are the questionnaires in by the general reports prepared by the European Association of Tax Law Professors on annual topics. 78 Concurring Thuronyi, Brooks, and Kolozs, Comparative Tax Law, 5. 79 Moessner, ‘Why and How to Compare Tax Law’, 315; Thuronyi, Brooks, and Kolozs, Comparative Tax Law, 59ff, include some principles in the framework of legal contexts. 80 Concurring Marian, ‘Discursive Failure’, 458. 81 Baker, ‘Expanding the Study of Comparative Tax Law’, 713. 82 What follows is an example of comparative analysis, but omitting the steps relating to the study and description of foreign tax law. See Ault, Arnold, and Cooper, Comparative Income Taxation, 427ff.

278 Marco Barassi dimension. It also has a legal dimension since moral and economic interests relating to family justify domestic and international legal rules. Income tax law must first define what is understood by family (spouses, cohabiting couples, same-sex couples/ partnerships, de facto partners, children) and then look at taxation of the family as a whole or of its members, consider the effects of income tax progressivity on single- income families, and decide, where family members are not taxed separately, whether to promote, penalize, or tax the family in the same way as individuals. Political choices may take into account principles such as ability to pay, equality, family protection,83 and aims pursued by the legislator which may be influenced by economic and social factors (e.g. to promote spouses’ work, to provide childcare services, to increase birth rates). The model of a family has changed with the times—through economic and social influences—and also has an important historical and sociological perspective. In the 1960s, the prevailing family model had the husband as the breadwinner but, in the twenty-first century, the framework changed dramatically with both spouses working, de facto partners and same-sex couples/partnerships becoming more common, and separation and divorce no longer viewed as unusual. Therefore, the law, including tax law, must be able to adjust accordingly and the notion of family must, as a consequence, include legally constituted cohabiting couples such as pacte civil de solidarité in France or the civil partnership in the UK. Since these couples have reciprocal legal rights and duties comparable to married couples, tax law cannot discriminate against those in similar situations as that would infringe the principle of equality. However, de facto couples are not wholly comparable with those who are legally married:84 although their factual situations may be the same, there are two reasons for differential tax treatment. First, if the relationship is formalized in law, it has legal relevance and, secondly, from a tax law viewpoint, a formalized relationship is easier for a tax administration to assess. The first element is substantive: rights and duties with legal relevance may be appreciated under the ability-to-pay perspective of the person to whom they relate (e.g. tax allowances for dependants). The second relates to enforcement: to be applied, tax law needs rules that do not require complex factual findings. The choice of family tax treatment largely depends on identifying a family’s ability to pay (as exemplified later) and on the need to tax equally those with the same ability to pay, with those who are not equal being treated in a different manner. However, the constitutional relevance of the family may sometimes justify more favourable treatment of a family compared to another taxpayer with the same ability to pay. Different countries’ tax laws show a trend towards taxation of the individual members of a family and not of the family as a whole: the constitutional courts of Germany (BVerfG, 17 January1957), Spain (Tribunal Constitutional 45/1989), and Italy (Corte costituzionale 179/1976) considered as unconstitutional the joint taxation of one of the spouses on the incomes of both. Nevertheless, ‘splitting’ and family quotient systems are common. While Italy

83

That sometimes has a constitutional relevance, as in Germany or Italy.

84 ECtHR, Burden v. United Kingdom, App. No. 13378/05.

Comparative Tax Law 279 and the UK tax individual family members, other countries tax them individually or jointly with special brackets for married couples (USA), and yet others apply family quotients (France) or splitting (Germany). For each model, there is a corresponding evaluation of the ability to pay: if the individual is taxed, the ability to pay exists but expenditure incurred in relation to dependants may entitle the taxpayer to deductions (on the assumption that such expenditure reduces their ability to pay); if splitting or family quotients apply, a family unit is assumed in relation to income and expenditure; and if joint taxation is applied, income is assumed. The countries mentioned here apply both the equality principle, which usually has a constitutional relevance, and the ability-to-pay principle, which is only in some cases explicitly or implicitly included in the constitution. Comparative analysis therefore seems to show that similar principles are applied differently, and it does not seem that the difference stems from the interpretation of apparently similar principles but with different content. In fact, the Italian Constitution (art. 53) specifies an ability to pay whereas the German Constitution does not, although the Federal Constitutional Court (Bundesverfassungsgericht) derives it from the equality principle (Basic Law, art. 3): Italy applies individual taxation of family members, while Germany applies splitting. The idea that this difference stems from the explicit inclusion of the principle in the constitution rather than deducing it from another principle does not hold true because although Italy and the UK apply individual taxation of family members, the UK does not have a written constitution. It, therefore, follows that the ability-to-pay principle does not have any constitutional relevance. Two reasons could justify this outcome. First, the choice of using a legal model depends not only on the ability-to-pay principle but also on the equality principle, and on the protection of family enshrined in some constitutions. The joint application of the ability-to- pay and equality principles justifies the use of family quotients and splitting. Secondly, the ability-to-pay principle is wide and general sufficient to prevent it being breached by different tax systems. If personal taxation and the ability to pay expressed by income are emphasized, the model of individual taxation will prevail; but if a family unit, as an entity where family members’ incomes converge and decisions on expenditure are made, is stressed, then joint taxation may be applied. Both solutions comply with the ability-to- pay principle, but they stem from different approaches. The choice of model may also be influenced by two opposite theories: according to the first, family gives rise to economies of scale and a greater ability to pay compared to the individual, while according to the second, income produced by the family members must also meet the needs of the other family members. However, while the effects of the first theory are difficult to measure, the second emerges as a justification in some countries (France and Germany). In addition, the second theory seems to have a firmer legal footing: in fact, if family members owe a duty of maintenance and care to each other, any income produced is not entirely available but, rather, is available only for the part not intended to satisfy the family’s basic needs. In such cases, family quotients and splitting may be justified as well as deductions for dependants. In summary, comparative analysis shows that family taxation is subject to debate in different countries where tax rules have changed over time to accommodate different goals.

280 Marco Barassi

16.2.4 Circulation of Models Tax models (rules and institutions) circulate among countries for different reasons, one of which is as follows. A tax system covers tax rates, degree of complexity, a tax administration’s approach (which may affect its appeal to foreign investors), and the level of competitiveness of the country. Hence, there is tax competition among countries, especially when they share similar characteristics as is the case with many European countries, and this induces a convergence of similar models.85 On the other hand, countries seek to protect their tax base against erosion and profit shifting. The OECD/G20 Base Erosion and Profit Shifting (BEPS) Project that has been implemented at the EU level with the Anti-Tax Avoidance Directive86 is based on the inadequacy of single states to face international tax evasion and avoidance carried out by aggressive tax planning, base erosion, and profit shifting. An effective response requires coordination which has been realized by the OECD and the EU, on the one hand, with tax cooperation,87 and, on the other hand, with the actions included in the BEPS Project’s action plan and the ATAD.88 Although the legal effect of the BEPS Project and the ATAD outcomes are different—the former is soft law, the latter is binding—the trend is quite clear and shows an international coordination of tax rules to protect the tax base (against base erosion and profit shifting) and to develop tax cooperation. For these reasons, comparative tax law is affected by globalization which has increased attention on the convergence of different systems instead of on traditional research into similarities and differences.89

85 There are internal constraints, e.g. public finance, due to the expenditure that taxes cover and external ones, e.g. prohibition of state aid (art. 107 TFEU). Examples of such models are patent box regimes, participation exemption, group taxation, and tax rules to promote residence for non-residents. 86 Council Directives (EU) 2016/1164 and 2017/952 (ATAD 1 and ATAD 2). 87 e.g. with automatic exchange of financial information through common reporting standard (Standard for Automatic Exchange of Financial Account Information in Tax Matters, published by the OECD, July 2014). 88 To counter aggressive tax planning, the OECD, in 2008 and 2013, highlighted the importance of disclosure of information by taxpayers and recommended a model of cooperative compliance to move the relationship between taxpayers and tax administrations to a cooperative framework based on trust rather than an adversarial one. Another example of convergence is the two-pillar solution proposed by the OECD to address the tax challenges arising from the digital economy with the global deal to ensure that large multinational enterprises pay a minimum tax rate of 15%. See Tax Challenges Arising from the Digitalization: Report on Pillar One Blueprint: Inclusive Framework on BEPS (Paris: OECD Publishing, 2020) and Tax Challenges Arising from the Digitalization: Report on Pillar Two Blueprint: Inclusive Framework on BEPS (Paris: OECD Publishing, 2020). As concerns Pillar Two, the EU Council directive 2002/2523 was approved on 14 December 2022. 89 M. R. Ferrarese, ‘Il diritto comparato e le sfide della globalizzazione. Oltre la forbice differenze/ somiglianze’, Rivista critica di diritto privato 3 (2013), 370.

Comparative Tax Law 281

16.3 Conclusion During its history, comparative tax law has been analysed in depth in its different aspects, particularly the methodology. Although it is hoped that this analysis will continue, the task of scholarship, in the author’s view, should aim to provide extensive scientific comparative tax studies that can be used for different purposes, among which the circulation of tax models seems to be one of the most common applications. Difficulties exist but the different steps in comparisons—for example, the study of foreign laws— could be divided between jurists in order to broaden the scope of studies to more countries, since a comparative study by a single scholar is objectively limited.

Pa rt I I I

SE L E C T E D I S SU E S ON TAX T R E AT I E S A N D I N T E R NAT IONA L TAX L AW

Chapter 17

Qualification C onfl i c ts and Tax Tre at i e s Gianluigi Bizioli

17.1 Introduction Qualification conflicts in tax treaty law raise two different issues. The first involves defining the boundaries of the topic and, in particular, the differences and the overlapping of the contiguous terms and activities of interpretation, classification, and characterization. This issue is common to every legal order, although it assumes specific dimensions in international taxation since tax treaty law relies on, and is strictly interlocked with, domestic law, and the conflicts are always due to a varying interpretation by two jurisdictions. The second, differently, involves the consequences of these conflicts on international taxation, in particular double taxation or double non-taxation of the same income in the hands of the same taxpayer by two (or more) different jurisdictions. The outlined features of qualification conflicts in international taxation show the need for an analytical treatment of the issue, highlighting the stipulative definitions adopted and the solutions provided by international organizations and tax literature. Accordingly, Section 17.2 provides a literature review of the issue, with particular emphasis on the OECD and the UN documents, on the commentaries, and on monographs on the topic. Section 17.3 deals with the critical analysis of the causes as well as the solutions provided, their coherence with the principles of international taxation, and ability to pursue the objectives. Section 17.4 draws some conclusions.

286 Gianluigi Bizioli

17.2 Definition of the Issue 17.2.1 The Position of the OECD MC and the UN Model Although the initial formalization has been provided by the Partnership Report,1 the most comprehensive definition of the expression ‘conflicts of qualifications’, and contiguous matters, is addressed by the Commentary on the 2017 OECD Model Tax Convention (OECD MC).2 According to paragraph 32.3 of the latter, conflicts of qualification are caused by ‘differences in the domestic law between the State of source and the State of residence’, which give rise to the application of different provisions of the OECD MC ‘to the same item of income or capital’. From the Commentary’s perspective, therefore, these conflicts are triggered by the reference made by tax treaty law to the domestic (tax) laws of the contracting states and they are the consequence of the domestic (tax) law disparities between the source and residence jurisdictions which lead to the application of different tax treaty provisions ‘to the same item of income or capital’.3 Accordingly, these conflicts find their roots in the tax treaty law dependence on domestic law(s) and, logically, the dimension of the issue is strictly linked to the number of references made. The following paragraphs of the Commentary explain the distance between the conflicts of qualification and those conflicts arising from interpretative or applicative issues of tax treaty law,4 implicitly suggesting that they belong to the traditional realm of the theory of (tax) treaty interpretation and find positive guidelines in articles 31ff of the Vienna Convention on the Law of the Treaties (VCLT). Clearly, the taxonomy provided by the Commentary is closely connected to the goals of the OECD MC (and should be interpreted along this line). As highlighted by paragraph 32.1 (and, after the amendment introduced in 2017, para. 32.2), the previously mentioned definitions should be considered in the light of the scope of application of 1 OECD,

The Application of the OECD Model Tax Convention to Partnerships, Issues in International Taxation, no. 6 (Paris: OECD Publishing, 1999), para. 94 (hereinafter, ‘Partnership Report’). 2 The 2017 Commentary on the Articles of the United Nation Model Double Taxation Convention follows the same position of the OECD MC (Commentary on ch. V, Methods for elimination of double taxation, in particular Section E., Conflicts of qualification). Thus, any reference to the OECD MC is also understood as a reference to the UN Model, unless otherwise specified. 3 This is confirmed by para. 94 of the Partnership Report, which defines the conflicts of qualification as the situations ‘where the residence and source States apply different articles of the Convention on the basis of differences in their domestic law’. It is worth noting that the language adopted by the Commentary is not always consistent. Para. 32.2 uses the term ‘classification’ as a synonym of ‘qualification’ and of the emerging conflicts. 4 See para. 35.5 of the Commentary which reads as follows: ‘conflicts resulting from different interpretation of facts or different interpretation of the provisions of the Convention must be distinguished from the conflicts of qualification described in the above paragraph where the divergence is based not on different interpretations of the provisions of the Convention but on different provisions of domestic law’; see also Partnership Report, para. 108.

Qualification Conflicts and Tax Treaties 287 article 23A or B and, in particular, of the interpretation of the phrase ‘may be taxed in the other Contracting State’. In this sense, the taxonomy has the purpose of eliminating double taxation and of preventing double non- taxation through tax evasion or avoidance, and, prima facie, is applicable only to those cases and provisions where the elimination of double taxation (or the prevention of double non-taxation) require the application of article 23A or B. Finally, the Partnerships Report addresses the issue of allocation conflicts arising from the identical classification of an entity (e.g. partnerships, trusts, funds) by two jurisdictions, which, however, treat the same entity differently for tax purposes (para. 18).5 The entity shares the same legal classification (according to civil or common law) but is treated differently for fiscal purposes by the source and residence jurisdictions.

17.2.2 The Position of (the) Tax Literature The reaction of international tax literature can be represented—in a simplified framework, but without forfeiting the basic elements—along two different lines. The first recognizes the (legal) difference and the utility of the distinction between interpretation and qualification. Assuming that interpretation is the ‘cognitive process’ aimed at unfolding a text and making it clear, the qualification problem in international tax law arises ‘when a tax treaty uses terms derived from domestic law of the Contracting States (especially when the term has a different meaning in the domestic laws of both Contracting States)’.6 Consequently, the qualification issues belong to the realm of the interpretative activity and the conflicts are the result of the reference to (or the dependence on) domestic 5 The extension to entities other than partnerships was made expressly by the 2017 version of the MC: ‘the paragraph not only serves to confirm the conclusions of the Partnership Report but also extends the application of these conclusions to situations that were not directly covered by the report’ (Commentary on art. 1, para. 4). 6 A. Rust, ‘Introduction’, in E. Reimer and A. Rust, eds, Klaus Vogel on Double Taxation Conventions, vol. I (Alphen aan den Rijn: Wolters Kluwer, 2015), 35 and 5 (paras 73 and 120) (emphasis in original). The author recognizes that the expression ‘conflict of qualification’ finds its roots in the field of private international law and has been transplanted by the German tax literature into the realm of international taxation. However, in the latter it has acquired a different meaning since it does not define the (domestic or foreign) law applicable to the situation but the issues arising from the dependence of tax treaty law on domestic law (para. 120). The same theoretical approach has been adopted by W. Haslehner, ‘Introduction’, in E. Reimer and A. Rust, eds, Klaus Vogel on Double Taxation Conventions, vol. I (Alphen aan den Rijn: Wolters Kluwer, 2021), 44 and 63 (paras 91 and 137). In the same vein, C. Pleil and S. Schwibinger, ‘Confronting Conflicts of Qualification in Tax Treaty Law: The Principle of Common Interpretation and the New Approach Revised’, World Tax Journal 10 (2018), 427, 428 (‘[c]‌onflicts of qualification occur where treaty terms are interpreted differently by the contracting states’); F. Potgens (‘[c]onflicts of qualification arise due to differences in domestic law between states’), in G. Kofler and F. Potgens, ‘Article 23—Methods for Elimination of Double Taxation’, in Global Tax Treaty Commentaries, Global Topics (IBFD online), para. 2.16.2, https://research.ibfd.org/#/doc?url=/collections/gttc/html/ gttc_article23_chaphead.html (accessed 30 April 2021).

288 Gianluigi Bizioli law by tax treaty law, either due to the general reference made by article 3(2) of the OECD MC and by the UN Model or because of specific tax treaty provisions (such as art. 6(2) or art. 10(3) OECD MC and the UN Model). The conflicts arise, therefore, from the different meanings assigned to the same tax treaty terms or to the different characterization of the same facts and they are caused by the simultaneous use of the domestic (tax) laws of the two contracting states. In this sense, qualification conflicts occur at the end of the interpretation process and their normative value should be appreciated for the solutions adopted to avoid them.7 A marginal deviation from this position maintains the distinction between conflicts of qualification ‘in a narrow sense’—which arise ‘on the basis of differences in their domestic law’—and in a ‘broad sense’—when the different interpretation is the consequence of the application of different methods of interpretation by the two jurisdictions.8 The devaluation approach is supported by different arguments. On the one hand, ‘conflicts of qualification are a non-issue’ because even if the same term has ‘a different meaning in each state . . . the meaning applies in only one state at any given time’.9 In detail: [m]‌y preferred reading (assuming that credit or exemption are given under the treaty rather than by reference to domestic law) would be to say that the source state categorizes the income for the purpose of deciding whether the treaty prevents it from taxing, which logically comes first, and the residence state merely has to ask whether (if it is not prevented from taxing) the ‘income may be taxed [in the source state] in accordance with the provisions of this Convention’—one of those provisions being that the source state applies its domestic law to answer the question of whether it is prevented from taxing.10

In this theoretical framework, the normative uselessness of the category ‘conflicts of qualification’ stems from the interpretation of article 3(2) of the OECD MC and of the UN Model as a special interpretative rule which refers to domestic law for the meaning 7 The

same author recognizes that the allocative conflicts are separate legal issues from the qualification ones, and, following the German tradition, labels them as problems of ‘classification’ (Rust, ‘Introduction’, 55, para. 121), and quotes the decision of the German Bundesfinanzhof of 12 January 1973, BFH III R 30/72, BStBl. II 440, 442 (1973) which considers this activity as ‘Einordnung’. 8 A. Ramos Huerta, ‘Conflicts of Qualification and the Interpretation of Tax Treaty Law’, in E. Burgstaller and K. Haslinger, eds, Conflicts of Qualification in Tax Treaty Law (Vienna: Linde, 2017), 21, 25, follows the OECD MC and the UN Model position. 9 This is the well- known position backed by J. Avery Jones. The quotations are taken from the author’s latest article (J. Avery Jones, ‘A Fresh Look at Article 3(2) of the OECD Model’, Bulletin for International Taxation 74 (2020), 654, 656–657), but, similarly, ‘Current Issues on Treaty Interpretation’, in J. Monsenego and J. Bjuvberg, eds, International Taxation in a Changing Landscape. Liber Amicorum in Honour of Bertil Wiman (Alphen aan den Rijn: Wolters Kluwer, 2019), 15. 10 Avery Jones, ‘Fresh Look at Article 3(2)’, 659. This approach dates back to 1984 when the author, together with a group of international tax lawyers, published ‘The Interpretation of Tax Treaties with Particular Reference to Article 3(2) of the OECD Model-I’, British Tax Review 14 (1984), 14. Similarly, K. van Raad, ‘Interpretation and Application of Tax Treaties by Tax Courts’, European Taxation 1 (1996), 4.

Qualification Conflicts and Tax Treaties 289 of undefined terms and the pre-emption assigned to the source jurisdiction in the application of tax treaty law.11 On the other hand, Lang holds that ‘[t]‌he term “qualification conflicts” is mostly used to describe interpretation issues over allocation rules in tax treaties . . . it is hardly or not possible to differentiate qualification conflicts from other interpretation conflicts’ and, therefore, ‘it is not very productive to waste a great amount of effort on exactly defining this term and distinguishing it from other interpretation conflicts’.12 Conflicts of qualification are, therefore, the result of an interpretative activity which means, ‘on the one hand, to identify the facts of the case and, on the other hand, to interpret the legal provisions to establish which provision is applicable to the identified facts of the case’.13 In this sense, they are caused by the different interpretative results deriving from the classification of facts and/or of the provisions by two (or more) jurisdictions.14 However, Lang recognizes that within the interpretative issues, although ‘not referred to as qualification conflicts’, there are other types of conflicts ‘resulting from treaty provisions modelled on article 1 and article 4(1) of the OECD Model’. The latter are named as ‘allocation conflicts’ since they ‘involve issues of personal treaty entitlement’.15

17.2.3 Conclusions The described alternatives of either enhancing or devaluing the autonomy of the category of the qualification conflicts depend on a specific view of the tax treaty interpretation process and are not without their consequences. 11

This is also the position held by R. X. Resch, Tax Treaty Interpretation: The Meaning and Application of Article 3(2) (Hamburg: tradition, 2020), 181ff, 188: the ‘precept of common interpretation’ based on the fundamental principle of unity means for tax treaties that terms defined in the treaty (or to be given a contextual meaning) have an unalterable common meaning, whereas terms to be defined by recourse to domestic law must be defined at any point in time according to the domestic law of one contracting state and may never be defined differently by both contracting states having recourse to their domestic laws simultaneously. The latter constitutes a misapplication of the treaty by at least one contracting state. 12 M. Lang, Qualification Conflicts, Global Tax Treaty Commentaries, Global Topics (IBDF online) ), para. 1.2, https://research.ibfd.org/#/ (accessed 30 April 2021). A critical position is also taken by A. Nikolakakis, ‘Interpretation vs Qualification’, in G. Maisto, ed., Current Tax Treaty Issues (Amsterdam: IBFD, 2019), who, referring to the OECD position, states that ‘[i]‌t is submitted that this is very confusing’ (para. 9.3). 13 Lang, Qualification Conflicts, para. 2.1.1. 14 See also M. Lang, The Application of the OECD Model Tax Convention to Partnerships. A Critical Analysis of the Report Prepared by the OECD Committee on Fiscal Affairs (Vienna: Linde, 2000), 28. 15 Lang, Qualification Conflicts, respectively paras 1.3 and 4.1. It is worth mentioning, however, that the author is also sceptical about the distinction between qualification and allocative conflicts since, assuming the OECD position (summarized in Section 17.2.1), he warns that ‘it makes a big difference whether the conflict in question is a qualification or allocation conflict’, although ‘[i]‌t is often hardly possible . . . to distinguish between the two kinds of conflicts’ also because ‘allocation conflicts can be the consequence of a different qualification’ (para. 4.2).

290 Gianluigi Bizioli The autonomous nature of the category is backed by those who consider that article 3(2) of the OECD MC and the UN Model prescribe a precise duty to refer to domestic (tax) law when a term is ‘undefined’ by the treaty. Following this line of reasoning, article 3(2) of the OECD MC and the UN Model establish a sharp distinction between the interpretation of the ‘defined’ tax treaty terms and the interpretation of the ‘undefined’ terms, because only the former are subject to the rules provided by the 1969 VCLT, whereas the latter follow primarily the meaning arising within the domestic (tax) system. Since this position eventually leads to double taxation or double non-taxation, the Commentary has adopted a specific interpretation of the expression ‘income or . . . capital which may be taxed in the other Contracting State in accordance with the provisions of this Convention’ contained in article 23A and B(1) of the OECD MC and the UN Model, fostering a solution which privileges the qualification provided by the source jurisdiction. Nevertheless, this conclusion is not shared by those who consider that only the source jurisdiction applies tax treaty law according to the wording of article 3(2) of the OECD MC and to the UN Model. In this understanding, any conflict (of qualification) is prevented by the fact that the residence jurisdiction must follow the characterization of the income provided by the source jurisdiction. The opposite position is based on the substantial autonomy of tax treaty law, seen as an autonomous legal system, which leads to the conclusion that reference to domestic (tax) law should be understood as exceptional and, according to article 3(2), a common or autonomous interpretation of undefined terms of tax treaty law should be privileged. Having thus summarized the different positions of the OECD, the UN, and international tax literature, a few doubts arise as to whether the expression ‘conflicts of qualification’ describes an interpretative issue; that is, a heuristic process aimed at providing a meaning to (written) words or sentences in a legal document or at characterizing facts according to legal provisions.16 So said, the value of this expression is limited to the mere description of a phenomenon as long as a specific legal treatment is provided. In other words, the normative value of the expression depends on the existence of an autonomous legal regulation of the effects caused by the different interpretation of the undefined terms provided by the source and the residence jurisdictions.

16 Ramos Huerta, ‘Conflicts of Qualification’, 26, excludes from the qualification issues the ‘conflicts concerning the interpretation of facts’ without providing any arguments for this position. However, in broad terms, legal characterization of facts can be considered as an interpretation process because it requires the application of legal provision to facts or, in practical terms, it is not easy to distinguish between the interpretation of provisions and the application of provisions to facts (for further considerations, see Section 17.3). This is also the position held by Resch, Tax Treaty Interpretation, 11, who, supporting the OECD taxonomy, states that ‘[s]‌ome authors and courts appear to apply the terminology indiscriminately to all types of mismatches in interpretation. Such indiscriminate use should be abstained from in order to avoid misunderstandings’. However, if both situations belong to the process of interpretation and—as theory and practice suggest—it is difficult to distinguish between interpretation and legal characterization of the facts, it is hard to understand which misunderstandings may arise.

Qualification Conflicts and Tax Treaties 291 Nonetheless, if the qualification issues are understood from a theoretical perspective, they should be considered as a matter of interpretation. Shedding light on this topic will be the task of Section 17.3. Before that, however, two further remarks should be made. Reference to domestic law is contained in specific tax treaty provisions too. This is the case of article 6(2) and article 10(3) of the OECD MC and the UN Model. In the former, the term ‘immovable property’ has the meaning of the law where the property is situated; in the latter, ‘income from other corporate rights which is subjected to the same taxation treatment as income from share’ should be interpreted according to the ‘laws of the State of which the company making the distribution is a resident’. These situations may lead to double taxation (or double non-taxation) if the source and residence jurisdictions do not share the same interpretation of the domestic law. Indeed, in these cases, qualification issues do not depend on the interpretation of article 3(2), but on a precise choice made by tax treaty law. Secondly, both the OECD MC and international tax literature converge on the distinction between qualification issues and allocation issues, the latter concerning the different characterization (or classification) of the same entity by two jurisdictions. The result of the conflicts of allocation is the attribution of income to two different taxpayers by two jurisdictions.

17.3 A Matter of Interpretation International tax literature and domestic cases unanimously agree that tax treaty law should be interpreted according to international customary law as detailed, in particular, by articles 31–33 of the 1969 VCLT. This conclusion is clearly justified by the legal and logical assertion that tax treaty law belongs to international law and, as such, is wholly ruled by the customary laws on the treaties.17 However, as mentioned in Section 17.2, tax treaty law usually contains a special rule on interpretation, the content of which is based on the following main elements: (1) it is applicable only to ‘any term not defined’ by the treaty law ;and (2) ‘unless the context

17 There is an abundance of (tax) literature on this point. For a clear and outstanding position, see K. Vogel and R. G. Prokisch, ‘General Report: Interpretation of Double Taxation Conventions’, Cahiers de Droit Fiscal International vol. 98a (1993), 55, 67:

[a]‌s international treaties, double taxation conventions are governed by the Vienna Convention on the Law of Treaties (VCLT) from 23 May 1969, which came into effect when the 35th state joined on 27 January 1980 . . . It is widely accepted that Article 31–33 VCLT codifies already valid customary law and hence all international treaties would—as they are already based on this customary law—be governed by the rules of interpretation of the VCLT in an unrestricted manner. The latter is also the stance of the International Law Commission; see the Commentary on the VCLT (1969), para. 5.

292 Gianluigi Bizioli otherwise requires’ that term has the meaning of the domestic law of the state which applies the convention. Due to the prevalence of lex specialis over generalis, article 3(2) of the OECD MC and the UN Model take precedence over the general rules enshrined in the VCLT. The scope of application of this provision does not raise specific issues since article 3(2) of the OECD MC and the UN Model apply exclusively to undefined terms. First, the word ‘term’ should be interpreted in a broad manner, including not only definitions but also concepts and it should not be restricted to identical (domestic) words.18 Secondly, undefined terms are also those contained in the course of the definition of a defined term and/or where there is a partial definition.19 In both cases, the terms or the expressions are not defined by tax treaty law and, therefore, article 3(2) of the OECD MC and the UN Model apply. In this sense: (1) Terms and expressions defined by tax treaty law shall be interpreted according to the rules provided by the 1969 VCLT (e.g. art. 5 of the OECD MC and the UN Model). (2) Terms and expressions defined by article 3(1) of the OECD MC and the UN Model shall be interpreted according to the rules provided by the VCLT, ‘unless the context otherwise requires’. (3) Terms and expressions not defined by tax treaty law shall be interpreted according to article 3(2) of the OECD MC and the UN Model. Thus, the core issue arising from the interpretation of article 3(2) of the OECD MC and the UN Model concerns the effects of the provision, that is, the meaning of the ‘context’ and which state applies the convention. These are decisive elements when dealing with the definition of the existence and the boundaries of the conflicts of qualification. According to a first scenario, the context takes precedence over the reference to domestic (tax) law. The meaning of undefined tax treaty terms should, therefore, be inferred within the system of the treaty and only exceptionally by reference to domestic (tax) law.20 This interpretative solution of article 3(2) of the OECD MC and of the UN 18

J. Hattingh, ‘Treaty Interpretation’, in Global Tax Treaty Commentaries (IBDF online), para. 4.3.2.1, https://research.ibfd.org/#/ (accessed 7 June 2021). 19 Ibid., para. 4.3.2.2. 20 In this sense, see Lang, Qualification Conflicts, para. 3.1.1. Art. 3(2) of the OECD MC and the UN Model focuses, inter alia, on interpretation with regard to the ‘context’ of the treaty. This must be understood as an implicit reference to the international law interpretative principles included in the VCLT. In the absence of evidence to the contrary, it can be assumed that states that have concluded a tax treaty with full knowledge of the interpretative rules of the VCLT have also implicitly assumed that the provisions contained therein are also relevant for the interpretation of that treaty. This also seems to be the position of the OECD MC Commentary on art. 3(2). Para. 12 reads: ‘paragraph 2 specifies that the domestic law meaning of an undefined term applies only if the context does not require an alternative interpretation and the competent authorities do not agree to a different meaning pursuant to the provisions of Article 25’.

Qualification Conflicts and Tax Treaties 293 Model leads to the development of autonomous meanings of the tax treaty terms and expressions and of an autonomous international tax language. The immediate results of this position are the substantial irrelevance of the conflicts of qualification, which are limited to those exceptional cases where the interpretation of the undefined terms and expressions is based on the reference to domestic (tax) law, and, from a theoretical perspective, the substantial uselessness of such a category. Taken to extreme consequences, this position considers article 3(2) of the OECD MC and the UN Model essentially superfluous21 (as well as the amendments that the OECD Fiscal Committee has introduced in the Commentary on article 23). A second scenario considers that ‘[t]‌he history and development of the provision confirm that an interpretation contrary to the meaning a term has under domestic law must constitute an exception’, although ‘the mandate to interpret a tax treaty “in the light of its object and purpose” (Article 31(1) VCLT) leads to the requirement that States should seek the treaty interpretation that is most likely to be accepted in both Contracting States (the goal of “common interpretation”)’.22 The expression ‘common interpretation’ is, however, ambiguous. Considering the result of the interpretative process, the definition encompasses both those processes aiming at establishing autonomous tax treaty meanings (i.e. meanings independent from domestic tax concepts) and those processes that lead to meanings which are most likely to be accepted in both jurisdictions. In this framework, in both cases the outcome of the interpretative activity is a single meaning applicable by the two contracting states, although this meaning is established through different processes, the former based on the systematic interpretation of tax treaty law and the latter (also) on the analysis of the domestic (tax) law of the two jurisdictions. In other words, if the outcomes of the activity are considered, the expression ‘common interpretation’ also includes the ‘autonomous’ interpretation; differently, considering exclusively the approach adopted, the expression is usually understood as ‘the interpretation of tax treaties that avoids conflicts of qualification through a consistent application of the OECD Model Convention by the authorities and court of both contracting states’.23 21 M. Lang, ‘Die Bedeutung des originär innerstaatlichen Rechts für di Auslegung von Doppelbesteuerungsabkommen (Art. 3 Abs. 2 OECD-MA’, in G. Burmeister and D. Endres, eds, Aussensteuerrecht, Doppelbesteuerungsabkommen und EU-Recht im Spannungsverhältnis. Festschrift für Helmut Debatin (Munich: C. H. Beck, 1997), 283, 303, quoted by Resch, Tax Treaty Interpretation, 202. 22 See, respectively, A. Rust, ‘Article 3’, in E. Reimer and A. Rust, eds, Klaus Vogel on Double Taxation Conventions, vol. I (Alphen aan den Rijn: Wolters Kluwer, 2015), 212 para. 123; and Rust, ‘Introduction’, 41–42, para. 90 (emphases in original). 23 Pleil and Schwibinger, ‘Confronting Conflicts of Qualification in Tax Treaty Law’, 429–430, who assign this lack of uniform understanding to the different legal traditions:

[w]‌hile especially in Germany, common interpretation is understood to promote an autonomous interpretation of a tax treaty and its terms that is detached from domestic law, common law scholars have emphasized that a common interpretation, even though desirable, can only be reached for treaty issues not depending on the meaning of terms since, according to article 3(2) of the OECD Model Convention, each contracting state can interpret undefined treaty terms differently depending on its domestic law, at least unless the context otherwise requires.

294 Gianluigi Bizioli In the case of ‘common interpretation’ too, conflicts of qualification are substantially non-existent since both jurisdictions apply the same meaning of the term. Focusing on the result, this conclusion can be extended to the method of treaty interpretation backed by Avery Jones (see Section 17.2.2), where undefined terms should be understood according to the terms and definitions provided by the domestic law of the source jurisdiction. Finally, a third scenario involves the lack of any coordinated interpretation between the source and the residence jurisdictions and, therefore, the freedom, for both jurisdictions to interpret undefined terms according to their own domestic (tax) law. As mentioned (see Section 17.2.1), this is the paradigmatic situation described in the OECD and the UN Commentary under heading E, Conflicts of qualification, of article 23.24 In these cases, the taxation of income ‘in accordance with the provisions of this Convention’ is provided by the source jurisdiction or, in other words, as interpreted and applied by the state of source. Conflicts of qualification arising from the difference in domestic (tax) law should, consequently, be relieved from double taxation by the state of residence even if the latter does not agree on the interpretative results and, as such, on the treaty provision applicable. In the reverse situation, where the source jurisdiction considers that the provisions of the convention preclude it from taxing an item of income ‘which it would otherwise have had the right to tax . . . even though the State of residence would have applied the Convention differently so as to have the right to tax that income if it had been in the position of the State of source’, the state of residence should avoid exempting the item of income (para. 32.7 of the Commentary on art. 23). It is worth mentioning that in the perspective adopted by the Commentary, double taxation arising from ‘different interpretation of facts’ (para. 32.5 of the Commentary on art. 23) does not need to be eliminated by the residence jurisdiction, because in these cases the latter ‘can argue that [the source jurisdiction] has not imposed its tax in accordance with the provisions of the Convention’. Therefore, these situations should be tackled according to the mutual agreement procedure.

This is the position traditionally upheld by Vogel, ‘Doppelbesteuerungsabkommen und ihre Auslegung (I)’, Steuer und Wirtschaft 59 (1982), 122; Vogel and Prokisch, ‘General Report’, 62; Rust, ‘Introduction’, 41, para. 90: [t]‌ax treaties are meant to allocate tax claims equally between the Contracting States. This goal can only be achieved if the treaty is applied consistently by the authorities and courts in both Contracting States. Therefore, the mandate to interpret a tax treaty ‘in the light of its object and purpose’ (Article 31(1) VCLT) leads to the requirement that States should seek the treaty interpretation that is most likely to be accepted in both Contracting States (the goal of ‘common interpretation’). 24 According to para. 32.3, conflicts of qualification arise ‘[w]‌here, due to differences in the domestic law between the State of source and the State of residence, the former applies, with respect to a particular item of income or capital, provisions of the Convention that are different from those that the State of residence would have applied to the same item of income or capital’.

Qualification Conflicts and Tax Treaties 295 As already said, this position cannot be backed and would be subject to review. The justification for this position can be based on the main argument: that the interpretation of facts differs from the interpretation of (tax treaty) law and, in a radical perspective, that the interpretation of facts is not a legal activity (but belongs to the real world). A second argument that may be raised is that the interpretation of facts does not involve domestic (tax) law. Both arguments are ungrounded.25 Legal theory has largely demonstrated that the distinction between interpretation of legal provisions and application of legal provisions to facts exists only in the books26 and that a precise line of demarcation cannot be drawn in practice.27 In other words, the interpretative activity does not differ from the application of legal norms to facts because both activities require the attribution of a meaning to legal provisions and the characterization of facts according to the legal order (and not to the real world).

17.4 Conclusions According to the general understanding, the expression ‘conflicts of qualification’ in the context of tax treaty law relates to the different meaning that the source and residence jurisdictions attribute to a term caused by the reference to their domestic (tax) law. In this sense, these conflicts depend on (or are the effect of) the interpretative process and, therefore, can be considered as a matter of interpretation. This definition raises two preliminary thoughts. It is difficult to make clear distinctions within the interpretative process. Consequently, it is not possible to follow the taxonomy developed by the OECD Fiscal Committee (and fostered by the UN Model), which, under the umbrella of the phrase ‘may be taxed in the other Contracting State’, identifies ‘conflicts’ arising from: (1) differences in domestic (tax) law; (2) different interpretation of facts; and (3) different interpretation of the provisions of the convention. The boundaries between these situations are extremely fragile (or even non-existent) and this great uncertainty, due to the consequences arising from the wrong qualification, is unacceptable. The proper conflicts of qualification require the residence jurisdiction to adopt the interpretation

25 Lang, Qualification Conflicts, para. 3.3.1.2 states that the distinction between conflicts of qualification and those arising from the interpretation of facts or the interpretation of tax treaty law ‘is problematic in many ways’. 26 W. Flume, Allgemeiner Teil des Bürgerlichen Rechts II (Berlin: Springer-Verlag, 1992), 317 states that the ‘art’ of interpretation can by learned only by practice. 27 For a comprehensive analysis of this conclusion, see T. M. J. Möller, Legal Methods: How to Work with Legal Arguments (Munch/Oxford/Baden-Baden: Beck/Hart/Nomos, 2000), 112ff and 492ff (‘[t]‌he convincing element of this is that the legal norm must be regularly substantiated before it can be applied’).

296 Gianluigi Bizioli of the state of source; differently, in cases (2) and (3), the conflicts should be resolved through the mutual agreement procedure. Although similar considerations should be extended to the category of conflicts of allocation, the value of the latter concerns the difference between the tax treatment of an item of income and the attribution of an item of income to a person. Whereas the conflicts of qualification always involve the interpretation of tax treaty law, the conflicts of allocation require the allocation of the income according to national (tax) law. In other words, the distinction between conflicts of qualification and conflicts of allocation is extremely relevant since the attribution of income is always a matter of domestic (tax) law. In this sense, even if there are a few doubts as to whether the attribution of income belongs to the interpretative process,28 the usefulness of this category lies in the precedence it takes over the application of tax treaty law. In this theoretical framework, therefore, the heuristic value of the category ‘conflicts of qualification’ depends on which interpretative approach is adopted to tax treaty law (as described in Section 17.3). More in detail, if the special nature of article 3(2) of the OECD MC and of the UN Model (compared to arts 31–33 of the 1969 VCLT) is shared, the destiny of the category is strictly linked to the interpretation of the reference to the ‘context’ contained therein. This issue involves two elements, the primacy of the context over the reference to domestic (tax) law (or vice versa) and the meaning of ‘context’. In the author’s view, article 3(2) of the OECD MC and the UN Model should be interpreted in such a way as to privilege the development of an autonomous meaning of undefined tax treaty terms, so that the context has a broad enough meaning to include any material connected to the (specific) tax treaty. In this scenario, one of the most important instruments for developing this common meaning of undefined terms is the Commentary, as well as the observations and reservations made. This conclusion is essentially based on two arguments, one positive and the other negative. According to the first, in line with the author’s legal background, each tax treaty and, in particular, tax treaty law, represents a system of provisions fostering one (or more) common goal(s). These elements allow any interpretative issue (or conflict) in the system to be solved through the ordinary means of interpretation (literal, systematic, and teleological). In other words, tax treaty law—as a system of provisions—is essentially a complete legal text whose conflicts of interpretation should be solved according to (the) traditional legal arguments. As to the second argument, prioritising the meaning of the domestic law of the source jurisdiction results in a (fiscal) advantage for this jurisdiction, in the sense that the broader the interpretation it adopts, the greater the advantage for such state. Needless

28

This is also the position of Lang, Qualification Conflicts, para. 4.2.

Qualification Conflicts and Tax Treaties 297 to say, this situation leads to the breach of the taxing rights distribution agreed by the jurisdictions signing the tax treaty. We are fully aware that the interpretative solution drawn in these conclusions also derives from the author’s legal background and, in a certain sense, from the author’s hope that a common international tax language will finally be developed.

Chapter 18

Triangul ar Case s a nd Tax Treati e s Paolo Arginelli

18.1 Introduction: What are Triangular Cases? Although not defined by law, the term ‘triangular case’ is a well-known technical term in the field of international taxation, in particular within the domain of tax treaty law.1 1

Due to space constraints, notes are kept to the minimum. In particular, the most relevant literature on triangular cases is listed exclusively in this note: K. van Raad, Nondiscrimination in International Tax Law (Deventer: Kluwer Law International, 1986); K. van Raad, ‘Issues in the Application of Tax Treaty Non-Discrimination Clauses’, Bulletin for International Fiscal Documentation 42/8/9 (1988), 347–351; K. van Raad, ‘Dual Residence’, European Taxation 28/8 (1988), 241–246; R. Betten, ‘Denial of Certificate of Residence to a Dual Resident Company’, European Taxation 29/11 (1989), 371–373; K. van Raad, ‘Dual Residence and the 1977 OECD Model Treaty Article 4(1), Second Sentence’, European Taxation 30/ 1 (1990), 27–29; C. J. A. M. van Gennep, ‘Dual-Resident Companies: The Second Sentence of Article 4(1) of the OECD Model Convention of 1977’, European Taxation 31/5 (1991), 141–146; J. F. Avery Jones et al., ‘The Non-Discrimination Article in Tax Treaties’, European Taxation 31/10 (1991), 310–347; OECD Committee on Fiscal Affairs, ‘Triangular Cases’, in Model Tax Convention: Four Related Studies (Paris: OECD Publications, 1992), 28–41; K. van Raad, ‘The 1992 OECD Model Treaty: Triangular Cases’, European Taxation 33/9 (1993), 298–301; P. M. Smit, ‘Taxation of Dividends Distributed by a Dual Resident Company’, European Taxation 33/1 (1993), 36–40; F. A. García Prats, ‘Triangular Cases and Residence as a Basis for Alleviating International Double Taxation. Rethinking the Subjective Scope of Double Tax Treaties’, Intertax 11 (1994), 473–491; K. Vogel, ed., Klaus Vogel on Double Taxation Conventions: A Commentary to the OECD, UN and US Model Conventions for the Avoidance of Double Taxation on Income and Capital: With Particular Reference to German Treaty Practice (The Hague: Kluwer Law International, 1997); M. Lombardi, ‘Triangular Situations: A Case of Double Source Taxation of Interest and Royalties’, Bulletin for International Fiscal Documentation 51/4 (1997), 177–182; J. F. Avery Jones et al., ‘Tax Treaty Problems Relating to Source’, European Taxation 38/3 (1998), 78–93; J. F. Avery Jones, ‘The David R. Tillinghast Lecture: Are Tax Treaties Necessary?’, Tax Law Review 53/1 (1999), 1–38; J. F. Avery Jones and C. Bobbett, ‘Triangular Treaty Problems: A Summary of the Discussion in Seminar E at the IFA Congress in London’, Bulletin for International Fiscal Documentation 53/1

300 Paolo Arginelli With regard to income tax, that term is generally intended to denote any situation in which two or more (generally three) states are entitled to tax the same item of income based on different connecting factors, such as different rules on tax residence, permanent establishment, lex rei sitae, the place of residence, or establishment of the payor. The multiplicity of connecting factors may concern either the personal connections of the taxpayer with two or more states, which lead to cases of dual or multiple tax residence, or the plurality of sources of the income. In the latter case, in addition to the (1999), 16–20; H. E. Kostense, ‘The Saint-Gobain Case and the Application of Tax Treaties. Evolution or Revolution?’, EC Tax Review 9/4 (2000), 220–232; A. J. Martín Jiménez, F. J. García Prats, and J. M. Calderón Carrero, ‘Triangular Cases, Tax Treaties and EC Law: The Saint-Gobain Decision of the ECJ’, Bulletin for International Fiscal Documentation 55/6 (2001), 241–253; S. Damen, ‘Netherlands Supreme Court Rules on the Residence of Dual Resident Companies under Tax Treaties with Third Countries’, Bulletin for International Fiscal Documentation 55/7 (2001), 290–292; J. W. J. de Kort, ‘HR 28 February 2001, nr 35.557: The Supreme Court of the Netherlands Reaches a Questionable Decision in a Triangular Dividend Withholding Case’, Intertax 29/12 (2001), 402–406; A. Rust, ‘Situs Principle v. Permanent Establishment Principle in International Tax Law’, Bulletin for International Fiscal Documentation 56/ 1 (2002), 15–18; P. M. Smit, ‘Treaty Residence of a Company in a Triangular Situation: Decision of the Supreme Court of 28 February 2001’, European Taxation 42/4 (2002), 155–158; R. Papotti and N. Saccardo, ‘Interaction of Articles 6, 7 and 21 of the 2000 OECD Model Convention’, Bulletin for International Fiscal Documentation 56/10 (2002), 516–521; B. J. Arnold and J. Sasseville, ‘Source Rules for Taxing Business Profits under Tax Treaties’, in B. J. Arnold et al., eds, The Taxation of Business Profits under Tax Treaties (Toronto: Canadian Tax Foundation, 2003), 109–131; K. Vogel, ‘Tax Treaty Monitor: Tax Treaty News’, Bulletin for International Fiscal Documentation 57/6 (2003), 23523–6; P. J. Hattingh, ‘The Role and Function of Article 1 of the OECD Model’, Bulletin for International Fiscal Documentation 57/11 (2003), 546–553; F. P. Sutter and U. Zehetner, eds, Triangular Tax Cases (Vienna: Linde, 2004); R. Pereira Ribeiro, ‘Dual Source Situations and the Multiple Taxation of Dividends and Interest’, IV Diritto e Pratica Tributaria Internazionale 1 (2004), 151–190; M. Gusmeroli, ‘Triangular Cases and the Interest and Royalties Directive: Untying the Gordian Knot?’—Part 1’, European Taxation 45/1 (2005), 2–13; J. F. Avery Jones, ‘The Interaction between Tax Treaty Provisions and Domestic Law’, in G. Maisto, ed., Tax Treaties and Domestic Law (Amsterdam: IBFD, 2006), ch. 6; J. Sasseville, ‘A Tax Treaty Perspective: Special Issues’, in ibid., ch. 3; B. J. Arnold, ‘At Sixes and Sevens: The Relationship between the Taxation of Business Profits and Income from Immovable Property under Tax Treaties’, Bulletin for International Fiscal Documentation 60/1 (2006), 5–18; E. A. Madeira and T. C. Neves, ‘Exploring the Boundaries of the Application of Article 10(5) of the OECD Model’, Intertax 38/8/9 (2007), 473– 483; F. P. G. Pötgens, ‘The Netherlands Supreme Court Again Excludes Credit of Withholding Tax in a Triangular Case’, European Taxation 48/4 (2008), 210–215; R. Vann, ‘ “Liable to Tax” and Company Residence under Tax Treaties’, in G. Maisto, ed., Residence of Companies under Tax Treaties and EC Law (Amsterdam: IBFD, 2009), 197–271; B. Kosters, ‘Triangular Cases in Tax Treaties’, Asia-Pacific Tax Bulletin 15/6 (2009), 372–377; K. Van Raad, ‘2008 OECD Model: Operation and Effect of Article 4(1) in Dual Residence Issues under the Updated Commentary’, Bulletin for International Taxation 63/5 (2009), 187–190; J.-P. van den Berg and B. van der Gulik, ‘The Mutual Agreement Tiebreaker: OECD and Dutch Perspectives’, Tax Notes International 54/5 (2009), 417–428; G. Zhai, ‘Triangular Cases Involving Income Attributable to PEs’, Tax Notes International 53/12 (2009), 1105–1123; S. Yong, ‘Triangular Treaty Cases: Putting Permanent Establishments in Their Proper Place’, Bulletin for International Taxation 64/3 (2010), 152–164; E. Fett, Triangular Cases: The Application of Bilateral Income Tax Treaties in Multilateral Situations (Amsterdam: IBFD, 2014); G. Maisto et al., ‘Dual Residence of Companies under Tax Treaties’, International Tax Studies 1/1 (2018), 57–62,, M. Tenore and P. Arginelli, ‘Italian Supreme Court Clarifies Taxation of Capital Gains Effectively Connected to an Italian Permanent Establishment Arising from the Alienation of Immovable Properties in the State of Residence’, Bulletin for International Taxation 75/6 (2021), 294–301.

Triangular Cases and Tax Treaties 301 standard source represented by the lex rei sitae or the tax residence of the payor, the item of income is sometimes regarded as taxable by a different state based on the fact that it is attributable to a permanent establishment situated in its territory. Due to constraints of space and the overall goal of this volume, this chapter will only deal with ‘classic’ triangular cases, which involve three different states and which may be summarized as follows: (1) Permanent establishment cases, that is cases where the relevant item of income is: (a) sourced in one state, based either on the lex rei sitae or the place of residence of the payor; (b) owned by a taxpayer that is a tax resident of a second state; and (c) attributable to a permanent establishment of the taxpayer situated in a third state. (2) Dual resident cases, that is cases where the relevant item of income is: (a) sourced in one single state; and (b) owned by a taxpayer that is regarded as a tax resident of two states (other than the state of source); (3) Reverse permanent establishment cases, that is cases where the item of income: (a) owned by a taxpayer that is a tax resident of a state is regarded as; (b) sourced in two states (different from the state of residence of the taxpayer) due to the fact that the payor is a tax resident of one of them and has in the other a permanent establishment to which the payment is connected. (4) Reverse dual resident cases, that is cases where the item of income: (a) owned by a taxpayer that is a tax resident of a state is regarded as; (b) sourced in two states (different from the state of residence of the taxpayer) due to the fact that the payor is regarded as a tax resident of them. In the remainder of the chapter, it is assumed that, unless otherwise explicitly provided: (1) tax treaties in force between the various states involved are bilateral treaties on income taxes reproducing the 2017 OECD Model Tax Convention on Income and on Capital (OECD Model); (2) tax residence under domestic law triggers tax treaty residence under article 4(1) of the relevant treaty; and (3) permanent establishments qualify as such under both domestic and treaty law. Moreover, although reference is generally made to entities (in particular, companies), the rules and principles analysed therein are generally also applicable to individuals. Finally, the cases analysed will factor in neither issues of allocation of income to different taxpayers, nor issues of characterization of the income or of the taxpayer (as fiscally opaque or transparent). In this respect, all persons involved are deemed to be tax opaque under the law of each state. The main issues stemming from the application of tax treaties to triangular cases derive, first, from the fact that tax treaties are generally bilateral, that is they bind in their reciprocal relations two states, while triangular cases are typically trilateral, that is they involve the exercise of taxing rights by three (or more) states. The effect is that while more than one tax treaty usually applies to prototypical (trilateral) triangular cases, the effects stemming from the application of one of those treaties are often not fully factored in by the other relevant treaties, leading to awkward results. Examples

302 Paolo Arginelli thereof are the double sourcing of the income under the treaties concluded by the residence state of the taxpayer with two source states; or the fact that, for the purpose of applying the treaty between the residence state of the taxpayer and the residence state of the person paying interest to the taxpayer, the obligation of the former state to exempt the interest under the treaty concluded with the state where the taxpayer has a permanent establishment (to which the interest is attributable) is generally not taken into account. Such misalignments, coupled with the specific rules applicable under the domestic laws of the states involved, may cause instances of double taxation and non-taxation. This effect is compounded by the fact that most tax treaties provide for multiple competing sourcing rules. While in a pure bilateral case, governed by the rules of one single treaty, the treaty rules aimed at resolving those conflicts are effective, in triangular cases, where more than one treaty applies, they often do not eliminate those conflicts since they are not drafted to apply in such circumstances. A third factor that contributes to cause awkward results is the role played by permanent establishments in the tax treaty context. Indeed, although the concept of permanent establishment is mainly purported to perform as a connecting factor to establish in which cases a state other than the residence state of the taxpayer may tax business profits, it is unquestionable that the profits that the state of the permanent establishment may tax under tax treaties theoretically include items of income that—under the sourcing rules provided for in other articles of the treaty—could be regarded as sourced in other states. In this respect, the state of a permanent establishment is entitled to tax the worldwide income that is attributable to that permanent establishment, somehow similarly to what would happen if the latter were a resident taxpayer. This chapter will focus on such issues and on how they should be tackled through a reasonable interpretation and application of the relevant tax treaties. In particular, Section 18.2 will discuss permanent establishment cases, Section 18.3 will deal with dual resident cases, Section 18.4 will analyse reverse permanent establishment cases, and Section 18.5 will examine reverse dual resident cases. Finally, Section 18.6 will conclude.

18.2 Permanent Establishment Cases 18.2.1 In General In permanent establishment cases, the relevant item of income, which is sourced in one state (hereinafter ‘state S’, or simply ‘S’) generally on the basis of the lex rei sitae or the place of residence of the payor, is owned by a taxpayer that is a tax resident of a second state (hereinafter ‘state R’, or simply ‘R’) and attributable to a permanent establishment of the taxpayer situated in a third state (hereinafter ‘state PE’, or simply ‘PE’). In most instances, triangular cases involve dividends, interest, royalties, and capital gains. To a

Triangular Cases and Tax Treaties 303

C income

R

PE S

Figure 18.1 The Permanent Establishment case.

lesser extent, the relevant item of income could be real estate income, or business profits sourced in state S under the domestic law of that state. See Figure 18.1 below. As a general rule, all states involved would apply their tax on the relevant income, here supposedly dividends. In particular, S would tax based on the tax residence of the payor, R would tax based on the tax residence of the recipient, while PE would tax because the income is attributable to the permanent establishment that the taxpayer has in PE. For the sake of the analysis, we assume that neither R nor PE provide for foreign tax credit or exemption under their respective domestic laws. Thus, before applying the relevant tax treaties, a triple taxation is at stake. Of the three tax treaties at stake (i.e. R–S, R–PE, and PE–S), the latter does not apply to limit the taxation of the relevant item of income in PE and in S. Indeed, the requisite for the application of the PE–S tax treaty, under article 1 thereof, is not satisfied since the relevant income is not owned by a tax resident of either contracting state, but by a person (C) who is a resident of a third state (R). This result is built on an interpretation of article 1 of the OECD Model according to which the ‘persons’ referred to therein are only those owning the relevant items of income. Such an interpretation stems from a systematic construction of article 1 in the context of the treaty as a whole and, in particular, of articles 6–21, all of which2 explicitly require—in order to apply—the owner of the relevant income to be a tax resident of either contracting states. The effects stemming from the application of the R–S and R–PE treaties are analysed in the following sections.

18.2.2 The R–S Treaty The first effect of the R–S treaty is that S may be bound to relinquish or reduce the tax to be levied under its domestic law. For instance, with regard to dividend distributions, article 10 applies since the company paying the dividends is a resident of S and the dividends are paid to a resident of R (C). Provided that C is the beneficial owner of the income, article 10(2) limits the taxation in S to 5% or 15% of the dividends, depending on the circumstances.

2

Except for art. 10(5), which is of limited application and that will be discussed in Section 18.4.

304 Paolo Arginelli The fact that the income is attributable to a permanent establishment in PE has generally no effect on the taxation in S, unless a provision similar to article 29(8) of the 2017 OECD Model is included in the R–S treaty, or the arrangement is regarded as abusive. Indeed, the treaty provisions—other than article 29(8)—dealing with the attribution of income to a permanent establishment refer only to permanent establishments situated in a contracting state. This is the case, for example, under article 10(4), which makes article 7 applicable if the ‘holding in respect of which the dividends are paid is effectively connected with [a]‌permanent establishment [in the other contracting state]’. As previously mentioned, certain R–S treaties include a minimum tax requirement for the application of the relevant treaty benefits, a prototype of which is article 29(8) of the 2017 OECD Model. Where R exempts the profits attributable to the permanent establishments situated in PE, those treaties make the application of the reduced taxation (or exemption) in S subject to the condition that the tax on the relevant income in PE is at least equal to a certain effective tax rate, either computed in absolute value or as a percentage of the standard tax rate applicable in R. Under article 29(8) of the 2017 OECD Model, treaty benefits are nonetheless granted if: (1) the relevant income emanates from, or is incidental to, the active conduct of a business carried on through the permanent establishment in PE;3 or (2) the competent authority of S determines that granting such benefits is nonetheless justified.4 The second effect of the R–S tax treaty is that R is bound to eliminate the R–S double taxation either by crediting the tax levied by S in accordance with the treaty, or by exempting the income. If source taxation is allowed under article 10, 11, or 12, R is generally obliged to grant a tax credit.

18.2.3 The R–PE Treaty The R–PE tax treaty is relevant in three respects. First, it usually confirms the right of PE to tax—under article 7 (or 13)—the relevant item of income, if the latter is attributable to a permanent establishment that the taxpayer has in PE. Secondly, if PE is granted a right to tax under article 7 (or 13), R is bound to relieve the taxpayer from international double taxation by way of exemption, or credit, under article 23(A) or (B). Thirdly, PE may be also bound under article 24(3) to eliminate the double taxation stemming from the overlapping of the taxes levied by S and PE. With regard to the first aspect, under article 7 the profits attributable to the permanent establishment in PE are taxable therein without any limitation, except for those stemming from the prohibition of discrimination set out in article 24(3). Thus, if those profits include the income sourced in S, PE is entitled to tax it. Article 7, however, does 3 Other than the business of making, managing, or simply holding investments for the enterprise’s own account, unless those activities are banking, insurance, or securities activities carried on by a bank, insurance enterprise, or registered securities dealer (see art. 29(8)(b) of the 2017 OECD Model). 4 See art. 29(8)(c) of the 2017 OECD Model.

Triangular Cases and Tax Treaties 305 not affect the provisions of other articles of the treaty, if the profits of the taxpayer include income dealt with separately in those articles. Where the income is sourced in a third jurisdiction (PE), the application of other articles is generally limited to articles 8, 13, and 21. The interaction among articles 6, 7, 13, and 21 is discussed separately in Section 18.2.4. With regard to the second aspect, it appears worth discussing how the obligation for R to relieve double taxation under article 23 of the R–PE treaty interacts with its duty to grant credit (or, less frequently, exemption) under the R–S treaty. In this respect, where R is bound to exempt the income under one of the two treaties (R–S or R–PE), no credit for the taxes levied by PE or S should be granted under article 23 of the other treaty. Indeed, under the OECD standard wording, the tax credit to be granted by R ‘shall not . . . exceed that part of the income tax . . . as computed before the deduction is given, which is attributable . . . to the income . . . which may be taxed’ in the other contracting state. In this respect, if the relevant income is exempt by R under the provision of one treaty (e.g. R–PE), no income tax is levied by R on that income and, therefore, no R tax is attributable to that income for the purpose of article 23 of the other treaty (e.g. R–S). As a consequence, R will not be bound to grant any credit under the latter treaty provision (e.g. for the taxes levied by S). Similarly, if both treaties require R to exempt the income, the exemption granted by R under either treaty removes any other duty for R to exempt an equivalent amount of income under the other tax treaty, since under the OECD Model that obligation is specifically linked to the relevant item of income.5 Finally, the question arises whether PE is bound to exempt the income sourced in S, or to grant a credit for the taxes levied in S, under article 24(3) of the R–PE treaty. On the one hand, it is generally recognized that, based on the clear-cut wording of that article, if exemption or foreign tax credit is granted to resident enterprises under the domestic law of PE, the same exemption (or credit) must be granted by PE to the local permanent establishments of R’s resident persons. The Commentary on article 24 of the OECD Model plainly confirms that reading.6 On the other hand, the issue is subject to debate in cases where the credit or exemption is granted by PE to its resident enterprises only by virtue of the PE–S tax treaty. In that case, according to the OECD Commentary, the ‘majority of [OECD] member countries are able to grant credit . . . on the basis of their domestic law or under paragraph 3 [of article 24]’.7 It is the author’s view that, under public international law, the wording, object, and purpose of article 24(3) require PE to extend the effects of the tax treaty exemption and credit to domestic permanent establishments of enterprises that are residents of R. Indeed, article 24(3) of the R–PE treaty focuses on the PE’s effective taxation of the permanent establishment, requiring that it is no less favourable than the taxation levied 5

Art. 23A of the 2017 OECD Model provides that ‘[w]‌here a resident of a Contracting State derives income . . . which may be taxed in the other Contracting State in accordance with the provisions of this Convention . . . the first-mentioned State shall . . . exempt such income . . . from tax’ (emphasis added). 6 See 2017 OECD Commentary on art. 24, para. 67. 7 Ibid., para. 70.

306 Paolo Arginelli on PE’s enterprises carrying on the same activities, while it does not distinguish between the various sources of law that regulate the taxation of those resident enterprises. Thus, the effects of the PE–S treaty on PE resident enterprises must be taken into account in order to establish the maximum taxation that PE may levy, under article 24(3) of the R– PE treaty, on the local permanent establishments of enterprises that are residents of R.8

18.2.4 Income from Immovable Property Triangular cases sometimes concern income derived from the sale or the use of immovable property. The specific issues of interpretation concerning those cases are dealt with in this section.

18.2.4.1 Gains from the alienation of immovable property Assuming that the immovable property sold is situated in S and is attributable to a permanent establishment in PE of an enterprise resident of R, the R–S and R–PE treaties come into play. The application of the R–S treaty does not pose significant interpretative issues. If the property qualifies as immovable property under article 6(2), article 13(1) takes precedence over articles 7 and 21 and allows S to tax the gain arising from the sale. R is then bound to grant a credit or exemption under article 23. With regard to the R–PE treaty, assuming that the sale is characterized as business profits under the tax law of PE, article 7 theoretically applies. However, under article 7(4), the provisions of other articles remain unaffected by article 7, where the relevant income is dealt with separately in such articles. This rule applies to the case at hand, since gains from alienations are covered by article 13. However, article 13(1) of the R–PE treaty certainly does not apply, since the property sold is not situated in a contracting state (but in S). As a consequence, the question arises whether the gain is covered by paragraph 2 or 5 of article 13. While article 13(5), which attributes exclusive taxing rights to R, applies to all gains derived from the alienation of property other than those referred to in paragraphs 1–4, article 13(2) only applies to gains ‘from the alienation of movable property forming part of the business property of a permanent establishment’. Since article 13(5) applies only if article 13(2) does not, the key question is whether, for the purpose of the R–PE tax treaty, the real estate situated in S may be regarded as 8 Where PE is bound to grant a credit for S taxes under art. 24(3) of the R–PE treaty, one should consider the possible existence of different source tax limitations under the R–S and PE–S treaties. Where the R–S treaty establishes a source tax limitation for S lower than that imposed by the PE–S treaty, PE is required only to grant a credit for the (lower) tax effectively levied by S (under the R–S treaty). Where the tax levied by S under the R–S treaty is higher than the tax that S could have levied under the PE–S treaty, art. 24(3) of the R–PE treaty only requires PE not to tax the permanent establishment of an enterprise of R less favourably than it would tax an enterprise of PE carrying on the same activities, i.e. to grant a credit only for the lower tax that S might levy under the PE–S treaty (if the beneficial owner of the income were a resident of PE).

Triangular Cases and Tax Treaties 307 ‘movable property’ forming part of the business property of the permanent establishment that the taxpayer, resident of R, has in PE. By construing article 13(2) in isolation and based on the apparent ordinary meaning of its terms, the answer to that question would seem straightforward: real estate cannot be categorized as ‘movable property’ and, therefore, article 13(2) may not apply. This interpretation, however, is too simplistic. First, the term ‘movable property’ is juxtaposed to the term ‘immovable property’, as used in paragraph 1. Therefore, contextual interpretation requires the former term to be construed by taking into account the meaning of the latter. In this respect, one might take the view that the term ‘movable property’ denotes any property other than the ‘immovable property’ covered in article 13(1), which refers to article 6 for definition purposes. Since article 6(2) defines ‘immovable property’ primarily through a reference to the meaning which it has under the law of the contracting state in which the property is situated and by adding to that core meaning a list of other items that the term is deemed to denote, one should conclude that either the real estate situated in S falls within one of the items expressly included in the definition of article 6(2)—for example, rights to which the provisions of general law respecting landed property apply—or it does not qualify as ‘immovable property’ for the purpose of article 13(1), since it is not situated in a contracting state (but in S). In the latter case, if the term ‘movable property’ used in article 13(2) were construed as encompassing any property that does not qualify as ‘immovable property’ under article 13(1), as the Commentary on article 13 of the OECD Model seems to suggest,9 the real estate situated in S would fall within the scope of application of article 13(2), with the effect that PE would be entitled to tax the gain arising from the sale thereof. This interpretation could be further supported by taking into account that article 21(2), when dealing with the taxation of income other than income from immovable property, specifies that the latter is ‘defined in paragraph 2 of Article 6’. An alternative solution is to regard the term ‘movable property’ as an undefined term that should be interpreted in accordance with article 3(2); that is, primarily through a renvoi to the law of the state applying the treaty (i.e. PE), unless the context requires a different interpretation. From this perspective, one should start by looking at the meaning of the term ‘movable property’ under the law of PE and—as would generally be the case—if the real estate situated in S were not to be regarded as ‘movable property’ under the law of PE, or under the law of S, which would be taken into account as part of the context,10 the conclusion would be drawn that article 13(2) of the R–PE tax treaty does not apply while article 13(5) does, with the effect of denying the right of PE to tax the gain. The context could be said to further support this conclusion, as the Commentary on article 13 of the OECD Model establishes that article 13(5) does ‘apply to gains derived from the alienation of immovable property . . . situated in a third State’.11 9

See of 2017 OECD Commentary on art. 13, para. 24. This conclusion would hold true where the real estate qualified as one of the items expressly listed in the definition provided for by art. 6(2) of the R–PE treaty. 11 2017 OECD Commentary on art. 13, para. 22. 10

308 Paolo Arginelli While, from a policy perspective, the author is inclined to prefer a solution where PE is given the taxing right over the gain stemming from the alienation of the immovable property in S, the fact that article 13(2) employs the term ‘movable property’—which is an undefined term that should be interpreted primarily according to the meaning it has under the law of PE—makes it preferable to construe that article as precluding PE to tax the gain from the sale of the real estate situated in S.

18.2.4.2 Income from the use of immovable property Where, under circumstances similar to those discussed in the previous section, the income is derived from the use of the immovable property, the application of the R–S treaty is straightforward. If the property qualifies as immovable property under article 6(2), article 6 takes precedence over articles 7 and 21 and allows S to tax the income. R is then bound to grant credit or exemption under article 23. With regard to the R–PE treaty, assuming that the relevant income is characterized as business profits under the tax law of PE, article 7 theoretically applies. Under article 7(4), however, the provisions of other relevant articles remain unaffected. Article 6 clearly does not apply since the real estate is not situated in a contracting state (but in S). Conversely, scholars have long debated whether article 21 applies. A first solution is to consider that article 21(1) does not come into play, since it applies only to income ‘not dealt with in the foregoing Articles’ of the treaty and the relevant item is already dealt with in article 7. Under this approach, article 7 of the R–PE treaty applies and PE is entitled to tax the income and bound to grant relief under article 24(3). This solution, however, seems to make article 21(2) partially meaningless, where it refers back to article 7 with regard to income attributable to a permanent establishment situated in PE, except for ‘income from immovable property as defined in paragraph 2 of Article 6’. If one takes the view that article 21 might apply, it then becomes necessary to interpret the term ‘immovable property’ as employed in article 21(2). The first possibility, in this respect, is to take the reference to the definition of ‘immovable property’ provided by article 6(2) literally. From this perspective, either the real estate situated in S falls within one of the items expressly listed in article 6(2), or it does not qualify as ‘immovable property’ since it is not situated in a contracting state as required by article 6(2). In the latter case, article 21(2) refers back to article 7, which allows PE to tax the income. This construction of article 21(2) is coherent with the first interpretation of article 13(2) highlighted in the previous section. The second possibility is to look at the meaning of the term ‘immovable property’ under the domestic laws of PE (the state applying the treaty) and S (the state where the real estate is situated) and to disregard the requirement under article 6(2) for the real estate to be situated in a contracting state. Where, from this perspective, the real estate in S qualifies as ‘immovable property’, article 21(1) applies and PE is forbidden to tax the income. This construction of article 21(2) is coherent with the second interpretation of article 13(2) highlighted in the previous section. Also in this case, from a policy perspective, the author favours a solution where PE is given the right to tax the income. However, the last interpretation provided above

Triangular Cases and Tax Treaties 309

C R1 R2 income

S

Figure 18.2 The Dual Resident case.

appears the more reasonable in the light of the current wording of the OECD Model and Commentary.12

18.3 Dual Resident Cases 18.3.1 In General In typical (trilateral) dual resident cases, the income is sourced in one state (hereinafter ‘state S’, or simply ‘S’) and owned by a taxpayer who is a resident of two other states (hereinafter ‘state R1’ and ‘state R2’, or simply ‘R1’ and ‘R2’) for domestic and tax treaty purposes. See Figure 18.2. As a general rule, all states involved apply their tax to the relevant income, which for present purposes we assume to be a dividend distribution. In particular, S taxes based on the tax residence of the payor, while R1 and R2 tax based on the tax residence of the recipient. Assuming that R1 and R2 do not provide tax credit or exemption in respect of foreign income, a triple taxation is at stake. The effects stemming from the application of those treaties are analysed in the following sections.

18.3.2 The R1–R2 Tax Treaty 18.3.2.1 The treaty includes the place-of-effective-management test as a tiebreaker rule under article 4(3) The application of the R1–R2 treaty first requires that the dual residence status of C is resolved under article 4(3),13 which here is assumed to provide a tiebreaker rule based on the place-of-effective-management criterion, such as in article 4(3) of the 2014 OECD Model. Under that rule, once the dual residence status is resolved in favour of one of the contracting states, the taxpayer is regarded as being exclusively resident of that state for the purposes of the treaty and the other provisions thereof are applied accordingly. 12 13

See also the 2017 OECD Commentary on art. 6, para. 1. The relevant tiebreaker rule would be art. 4(2) if the taxpayer were an individual.

310 Paolo Arginelli In particular, if it is assumed that C is a tax resident solely of R1 under article 4(3) and that it has no permanent establishment in R2, or that the relevant item of income is not attributable to the permanent establishment that C has in R2, article 7 precludes R2 from taxing that item of income. Conversely, where the relevant item of income is attributable to a permanent establishment that the taxpayer has in R2, the situation is akin to that analysed in Section 18.2.

18.3.2.2 The treaty includes a provision analogous to article 4(3) of the 2017 OECD Model The 2017 update to the OECD Model replaced the tiebreaker rule provided by the previous article 4(3) with a provision requiring dual residence situations to be resolved through a mutual agreement procedure between the competent authorities of the contracting states. Most importantly, for the purpose of this chapter, new article 4(3) rules that, absent an agreement between the competent authorities, the reliefs and exemptions laid down by the treaty shall apply only to the extent and in such a manner as may be agreed by those authorities. Thus, the new provision no longer incorporates a substantive tiebreaker rule that may be applied unilaterally by the tax authorities of a contracting state or by national judges, but requires the competent authorities to find an agreement through a procedure that entails a certain degree of discretion. If the competent authorities strike an agreement on the tax treaty residence of the taxpayer, the situation becomes similar to that described in the previous section. However, where (and as long as) such an agreement is not reached, the reliefs and exemptions laid down by the treaty shall not apply. This means that both R1 and R2 will continue taxing the income of C, without granting any relief under article 23 of the R1– R2 treaty.

18.3.3 The Treaties between the States of Residence (R1 and R2) and S 18.3.3.1 The R1–R2 treaty includes the place-of-effective-management test as a tiebreaker rule under article 4(3) Assuming that article 4(3) of the R1–R2 treaty breaks the residence tie in favour of R1, which becomes the sole state of residence for the purpose of that treaty, the R1–S and R2–S treaties apply as follows. The R1–S treaty applies as described in Section 18.2.2 with regard to the R–S treaty. With reference to the R2–S treaty, the last sentence of article 4(1) comes into play, which provides that the term ‘residence of a Contracting State’ does ‘not include any person who is liable to tax in that State in respect only of income from sources in that State or capital situated therein’. The Commentary on article 4 of the OECD Model, taking a substantive view on the meaning of that sentence, provides that it ‘excludes

Triangular Cases and Tax Treaties 311 companies and other persons who are not subject to comprehensive liability to tax in a Contracting State because these persons, whilst being residents of that State under that State’s tax law, are considered to be residents of another State pursuant to a treaty between these two States’.14 This entails that, due to the effect of article 4(3) and the distributive rules15 of the R1–R2 treaty, C shall not be regarded as a resident of R2 for the purpose of the R2–S tax treaty, which, as a consequence, shall not apply to the relevant income.

18.3.3.2 The treaty includes a provision analogous to article 4(3) of the 2017 OECD Model If the competent authorities of R1 and R2 find an agreement under article 4(3) of the R1–R2 tax treaty in respect of the residence of C, the application of the R1–S and R2–S treaties follows the same pattern discussed in the previous section. Conversely, if (and as long as) such an agreement is not reached, the following pattern emerges. Since article 4(3) of the R1–R2 treaty does not prevent the application of the tax treaties in force between either of the residence states and S, both the R1–S treaty and the R2–S treaty apply. In particular, since S is bound by both the R1–S treaty and the R2–S treaty, it shall apply the treaty provision that most restricts its taxing rights. For instance, if article 10(2) of the R1–S treaty provides for a maximum source tax rate of 10%, while article 10(2) of the R2–S treaty provides for a 15% maximum rate, S is bound to tax no more than 10%. In addition, new article 4(3) may trigger certain issues of temporal application of the treaty relief by S. More specifically, while, pending the mutual agreement procedure concerning the treaty residence of C, S is bound to apply the most restrictive treaty, once the competent authorities of R1 and R2 find an agreement with retrospective effects, so that the loser residence state (e.g. R2) is bound to refund the taxes that it has levied on C pending the procedure, S might have to adjust its previous taxation to reflect the right to tax laid down by the tax treaty concluded with the winning state. Domestic statute of limitations, however, may prevent (wholly or partially) the application of these retroactive adjustments. Symmetrically, absent or pending the mutual agreement procedure, both R1 and R2 are bound to grant relief under their respective treaties with S, since article 4(3) of the R1–R2 treaty only denies tax relief under that treaty, but does not extend that denial to the tax reliefs provided under the treaties concluded by the residence states with S. The tax levied by S, thus, shall be credited against both R1 and R2 tax. This double relief is not prevented by any anti-abuse rule or principle, as the credit is used to offset the taxes imposed on the same income by both R1 and R2.16

14

2017 OECD Commentary on art. 4, para. 8.2. i.e. arts 6–21 of the R1–R2 treaty. 16 See by analogy the dual deduction matched by a dual inclusion allowed under the BEPS Final Report on Action 2.e 15

312 Paolo Arginelli C R

income

S SPE C1

Figure 18.3 The Reverse Permanent Establishment case.

18.4 Reverse Permanent Establishment Cases 18.4.1 In General In typical (triangular) reverse permanent establishment cases, an item of income owned by a resident of state R (or simply ‘R’) is sourced in two other states, due to the fact that the payor is a tax resident of one of them (hereinafter ‘state S’, or simply ‘S’) and has in the other a permanent establishment to which the payment is connected (hereinafter ‘state SPE’, or simply ‘SPE’). In most instances, reverse permanent establishment cases concern dividends, interest, and royalties. See Figure 18.3. It is assumed that the relevant item of income is a dividend distribution, or an interest/ royalty payment. It is also assumed that S taxes based on the tax residence of the payor, R taxes based on the tax residence of the recipient, and SPE taxes because the payment is connected to the permanent establishment that the payor has in SPE. Finally, for the sake of the analysis, it is supposed that R does not provide for credit or exemption in respect of foreign income under its domestic law, so that a triple taxation is at stake. The S–SPE treaty may apply only with regard to dividends, due to the exceptional scope of article 10(5) of the OECD Model which operates subject to the sole condition that the company distributing the dividends is a resident of a contracting state deriving profits or income from the other contracting state. Conversely, the application of article 10(5) is not subject to the further condition that the beneficial owner of the dividends is a resident of a contracting state. With regard to interest and royalty payments, no similar provision is included in the S–SPE treaty. Therefore, since the owner of the income (C) is not a tax resident of either contracting state (but of R), the requirement for the application of the S–SPE treaty under article 1 thereof is not satisfied.17 The effects stemming from the application of the three tax treaties are separately analysed in the following sections.

17

See the analysis carried out in Section 18.2.

Triangular Cases and Tax Treaties 313

18.4.2 The R–S Tax Treaty Under the R–S tax treaty, S may be bound to relinquish or reduce the tax to be levied under its domestic law. With regard to dividends, provided that C is the beneficial owner of the income, the limitation set forth in article 10(2) applies and, therefore, the maximum tax to be levied by S shall not exceed 5% or 15% of the dividends, depending on the circumstances. A similar result is achieved in respect of interest/royalty payments. Under the first sentence of article 11(5), and where the relevant tax treaty provides for source taxation on royalties, the analogous provision included in article 12, ‘[i]‌nterest [and royalties] shall be deemed to arise in [S] when the payer is a resident of [S]’. The consequence of this rule is that the interest (or royalty) is deemed to arise in S, which triggers the application of article 11(2) and, thus, the limitation of the taxing right of S.18 The second sentence of article 11(5) of the OECD Model—and the corresponding provision of article 12, where the relevant tax treaty provides for source taxation on royalties—states that, where ‘the person paying the interest . . . has in a Contracting State a permanent establishment in connection with which the indebtedness on which the interest is paid was incurred, and such interest is borne by such permanent establishment, then such interest shall be deemed to arise in the State in which the permanent establishment is situated’. This provision does not trigger any effect on the application of the R–S treaty, since it clearly applies only insofar as the permanent establishment is situated in a contracting state, and not in a third state (e.g. SPE). This conclusion would change where the second part of article 11(5)—and the corresponding provision of article 12, where existing—provided that the income should be deemed to arise in the state of the permanent establishment (SPE), and not in S, also where the former is not a contracting state.19 As it is shown in Section 18.4.3, the standard wording of article 11(5) leads to possible instances of double taxation at source, since both S and SPE retain their taxing rights under the relevant treaties (i.e. R–S and R–SPE respectively). The second effect deriving from the application of the R–S treaty is that R is bound to eliminate the double taxation on the income sourced in S, either by way of credit or by exempting the income. Under the OECD Model, if the relevant income falls within the scope of articles 10 and 11 (and 12, if this is the case), R is obliged to grant a credit for the taxes that S levies in accordance with those articles. A relevant question, in this respect, is whether that tax credit should apply in parallel with the credit or exemption that R has to allow under article 23 of the R–SPE treaty. This issue is dealt with in the following section.

18

19

e.g. down to 10% under art. 11 of OECD Model-based tax treaties. See the 2017 OECD Commentary on art. 11, paras 28–30.

314 Paolo Arginelli

18.4.3 The R–SPE Tax Treaty With regard to dividends, SPE does not have any taxing right under article 10. Indeed, dividends distributed by C1 to C fall outside the scope of application of article 10, which applies only to dividends paid by companies that are a resident of a contracting state. Since C1 is not a resident of either contracting state, article 10 does not apply, while article 7 does, with the effect of attributing an exclusive taxing right on the dividends to R. A side effect of the application of article 7 is that R is bound to apply neither credit nor exemption under article 23. The situation is different where interest (and royalties, if the treaty provides for source taxation) payments are at stake. As previously mentioned, the second part of article 11(5) and, where the relevant tax treaty provides for source taxation on royalties, the analogous provision included in article 12, provides that, where: the person paying the interest . . . has in a Contracting State a permanent establishment in connection with which the indebtedness on which the interest is paid was incurred, and such interest is borne by such permanent establishment, then such interest shall be deemed to arise in the State in which the permanent establishment is situated.

This provision deems the interest (and royalties) to be sourced in SPE, with the effect of making article 11 (and article 12) applicable and, therefore, limiting the taxing right of SPE. As previously mentioned, this poses possible instances of double taxation at source, since both S and SPE—under the OECD standard wording of article 11(5)—retain their taxing rights under the relevant tax treaties. With regard to interest and royalty payments, under the R–SPE treaty R is bound to grant a credit for the taxes levied by SPE. This obligation to grant a foreign tax credit applies in parallel with the one imposed on R by the R–S treaty. Nothing in either treaty (R–S and R–SPE) relieves R from this double-tax-credit requirement. Therefore, R is bound to grant an aggregate credit to C, in respect of the total taxes levied at source by S and SPE, up to the amount of tax that R imposes on the relevant income (interest or royalty payment) under its domestic law.

18.4.4 The S–SPE Tax Treaty Under article 10(5) of the S–SPE tax treaty, SPE cannot tax the dividends distributed by C1, although such dividends are paid out from profits derived through the permanent established in SPE. Indeed, article 10(5) provides that SPE: may not impose any tax on the dividends paid by [a]‌company [resident of S], except insofar as such dividends are paid to a resident of [SPE] or insofar as the holding in respect of which the dividends are paid is effectively connected with a permanent

Triangular Cases and Tax Treaties 315 establishment situated in [SPE] . . . even if the dividends paid . . . consist wholly or partly of profits or income arising in [SPE].

As the Commentary on article 10 clarifies, article 10(5) rules out the extraterritorial taxation of dividends; that is, the practice by which states tax dividends distributed by a non-resident company solely because the corporate profits from which the distributions are made originated in their territory, such as in the case of profits realized by non- resident companies through a permanent establishment situated therein. Article 10(5) has no analogous provision in the OECD Model with respect to other items of income, in particular interest and royalties.

18.5 Reverse Dual Resident Cases 18.5.1 In General In typical (trilateral) reversal dual resident cases, the relevant item of income is sourced in two states (hereinafter ‘state S1’ and ‘state S2’, or simply ‘S1’ and ‘S2’) and owned by a taxpayer that is a tax resident of a third state (hereinafter ‘state R’, or simply ‘R’). In most instances, reversal dual resident cases concern passive income (i.e. dividends, interest, and royalties). See Figure 18.4. It is assumed that the relevant item of income is a dividend distribution, or an interest/ royalty payment. It is also assumed that both S1 and S2 tax because the payor is regarded as a tax resident of either state, and R taxes on the basis of the tax residence of the recipient. Finally, for the sake of the analysis, it is supposed that R does not provide for credit or exemption in respect of foreign income under its domestic law, so that a triple taxation is at stake. For the reasons highlighted in Section 18.2.1, the R–S1 treaty and the R–S2 treaty always apply to the relevant income. Conversely, the S1–S2 treaty applies only with regard to dividend distributions and insofar as the dual residence status of C1 is resolved under article 4(3) of that treaty, due to the exceptional scope of article 10(5) of the OECD Model.

C R

income

S1

S2 C1

Figure 18.4 The Reverse Dual Resident case.

316 Paolo Arginelli The effects stemming from the application of those treaties are separately analysed in the following sections.

18.5.2 The S1–S2 Tax Treaty The proper application of the S1–S2 treaty, as well as of the R–S1 and R–S2 treaties, requires that the dual residence status of C1 is resolved under article 4(3) of the S1–S2 treaty.20 Where article 4(3) reproduces the corresponding provision of the pre-2017 OECD Model, the dual residence status is resolved in favour of the contracting state where the place of effective management of C1 is situated. Assuming that S1 is the sole state of residence under that rule, article 10(5) precludes S2 from taxing the dividends distributed by C1.21 As previously mentioned, the S1–S2 treaty does not apply in cases of income other than dividends. The same result occurs where article 4(3) of the S1–S2 treaty follows the corresponding provision of the 2017 OECD Model and the competent authorities of the two contracting states strike an agreement on the treaty residence of C1. Conversely, where article 4(3) of the S1–S2 treaty follows the corresponding provision of the 2017 OECD Model and the competent authorities do not reach any agreement (or as long as such agreement is not found), article 10(5) cannot apply since it is not possible to establish which of the two contracting states should be regarded as the residence state for the purpose of that rule. In that case, the S1–S2 treaty does not affect the taxation of the dividends by the two contracting states.

18.5.3 The R–S1 and R–S2 Tax Treaties Where the dual residence status of C1 is resolved under the S1–S2 treaty, the following consequences are triggered in the application of the R–S1 and R–S2 treaties. With regard to dividend distributions, the loser state (e.g. S2) is not entitled to tax the dividends paid by C1 under article 10 of the R–S2 treaty,22 since under the last sentence of article 4(1) of that treaty, C1 is not regarded as a company resident of S2 because of the effects of the S1–S2 treaty.23 Article 7 applies, with the effect of attributing the exclusive taxing right to R. Conversely, with regard to interest and royalty payments, it becomes essential to establish whether C1 is deemed to have a permanent establishment in the loser state (e.g. S2) for the purpose of applying the R–S2 treaty. If such permanent establishment does 20

The relevant tiebreaker rule would be art. 4(2) if the taxpayer were an individual. See by analogy Section 18.4.4. 22 See by analogy Section 18.4.3. 23 See Section 18.3.3. 21

Triangular Cases and Tax Treaties 317 not exist, or the payment of the interest or royalties is not borne by the permanent establishment, the relevant item of income is not regarded as arising in S2.24 As a consequence, articles 11 and 12 do not apply, while article 7 does, with the effect of attributing the exclusive taxing right to R. If, on the other hand, a permanent establishment exists in S2 and the relevant item of income is borne by that permanent establishment, the situation is akin to the one described in Section 18.4.3. In both cases, under the R–S1 treaty, S1 will be entitled to tax the income under article 10, 11, or 12, and R will be bound to grant a credit for the taxes levied at source.25 If the dual residence status of C1 is not resolved under the S1–S2 treaty, which could be the case where article 4(3) of that treaty reproduces the corresponding provision of the 2017 OECD Model and the competent authorities do not find an agreement, both source states are entitled to tax the relevant income under articles 10, 11, and (where this is the case) 12 of the R–S1 and R–S2 treaties, as countries of residence of the payor.26 This entails that R is bound to grant credit for the taxes levied by both S1 and S2 under their respective treaties, insofar as the sum of the taxes levied at source in accordance with the provisions of those treaties does not exceed the tax to be levied by R on the relevant income.

18.6 Conclusion The analysis carried out in the previous sections show that triangular cases pose a wide range of issues of interpretation (and, therefore, application) of bilateral tax treaties. A good number of those issues can be resolved through an appropriate construction of the relevant treaty provisions, on the basis of the general principles of interpretation codified by the Vienna Convention on the Law of Treaties and the guidance provided by the OECD (and UN) Commentary. From a policy perspective, however, it would be advisable in the long run to bring certain changes into the Model with a view to streamlining the application of tax treaties in triangular cases. In this respect, it is the author’s perspective that the best solution going forward would be to treat permanent establishments—for treaty purposes—as if they were resident persons of their contracting state of establishment. This solution, which would build upon the separate entity approach that already characterizes the concept of permanent establishment under tax treaties would significantly shift the nature and function of the permanent establishment concept in the

24

See by analogy Section 18.4.3. See by analogy Section 18.4.2. 26 The 2017 OECD Commentary on art. 4, para. 24.4 clarifies, in this respect, that C1 would continue to be regarded as resident of both S1 and S2, even for the purpose of the S1–S2 treaty, with reference to the taxation of income that it pays to another person. 25

318 Paolo Arginelli treaty context from a sourcing rule to a territorial connection akin to residence. As a result, most of the issues discussed in this chapter would be eliminated. Finally, although this chapter is not the appropriate place to analyse them in detail, it is nonetheless worth highlighting that treating permanent establishments like resident persons under tax treaties triggers five main policy questions, which should be answered in order to make the proposed solution effective: (1) how to ensure and verify that the specific income is actually attributed (and attributable) to the permanent establishment that is entitled to the relevant treaty benefits; (2) whether the entity legally owning the income should continue to be entitled to apply the tax treaties concluded by its state of residence in respect of the relevant income or, conversely, whether only the permanent establishment to which that income is attributed should be entitled to the tax treaty benefits; (3) whether also the paying entity should be regarded as the payor of the income for tax treaty purposes, in addition to the permanent establishment that bears the relevant payment; (4) whether the new treaty characterization of the permanent establishment should also apply to the ‘internal dealings’ between the head office and the permanent establishment; (5) how to ensure the effectiveness and efficiency of this change of paradigm in a network of far more than 3,000 tax treaties.

Chapter 19

The Fu tu re of Avoi di ng D ouble Tax at i on Martin Berglund

19.1 Introduction The two main methods of mitigating international double taxation—exemption and foreign tax credit (hereinafter ‘the methods’)—have been an essential part of international tax law since this area of law was established. In the international discussion over the past decade, the design and application of the methods may at first glance seem to have faded into the background. In the OECD Base Erosion and Profit Shifting (BEPS) Project and parallel developments in national legal systems, the focus has instead been on countering tax avoidance in the form of double non-taxation and other forms of undesired low taxation that arise in international business and transactions. However, as will be described, the design of the methods is also an important aspect of this development. The methods are regulated in the OECD Model Tax Convention (OECD Model) in the two alternative articles 23A and 23B, and in national legislation of different countries. The rules in the OECD Model aim at mitigating international juridical double taxation, and thereby at achieving a situation where the income in question is instead subject to single taxation. However, these objectives are not always fulfilled in the concrete application. The rules in articles 23A and 23B should be distinguished from mitigation of various forms of international economic double taxation. The same methods are often applied in national law, and sometimes also in tax treaties, to provide relief for the economic double taxation that can arise in international company groups. Further, a special form of international economic double taxation arises as a result of transfer pricing adjustments, which is dealt with in article 9(2) of the OECD Model. The purpose of this chapter is to assess what impact the OECD BEPS Project has had so far, and can be expected to have in the future, on the methods of mitigating international double taxation. This refers to both the OECD BEPS reports published

320 Martin Berglund in 2015 and the continued work within the Inclusive Framework, where in 2020 two published drafts—the Pillar One and Two Blueprints—marked the starting point for further changes in international tax law as a result of digitalization of the economy. In the chapter, I also seek to discuss whether the current changes appear to be consistent with how the methods have traditionally been designed. I focus on the application of the methods of mitigating international juridical double taxation, but, as will be apparent, it is to some extent necessary to compare with the handling of the previously mentioned two forms of international economic double taxation. The chapter has the following outline. In Section 19.2, I first describe the basic effects of applying exemption and credit to mitigate international double taxation. Section 19.3 deals with the design and applicability of the method articles in the OECD Model. The policy question of choosing between exemption and credit is described in Section 19.4. There I also address the use of switchover rules which in different situations require a method change from exemption to credit. In Section 19.5, I describe more specifically the impact, or lack of impact, of the 2015 BEPS reports on the design and application of the method articles. Sections 19.6 and 19.7 discuss how some of the proposals introduced in the Pillar One and Two Blueprints affect the methods for mitigating international double taxation. Finally, I make some concluding remarks in Section 19.8.

19.2 Basic Effects of the Methods Application of the exemption method generally means that the foreign income must be excluded from the tax base in the residence state. Some variations are possible. A full exemption can be applied, where the foreign income in no way affects the taxation in the residence state. The design used in article 23A of the OECD Model is exemption with progression, where the foreign income—even though it is not taxable in the residence state—is taken into account when calculating the tax rate on the income that is still taxable in that state. An alternative design of the exemption method, which gives basically the same result as exemption with progression, is that the residence state includes the foreign income in the tax base but reduces the tax according to a calculation of how much of that state’s tax is considered to amount to the foreign income.1 When applying the exemption method, a difficult question is whether a deduction should be allowed for foreign losses. Since foreign income is exempted, foreign losses should for symmetrical reasons not be deductible. There is sometimes the possibility of allowing deductions according to national law, but the deduction will often be recaptured if the foreign activities later yield a profit.2

1

See OECD Commentary on art. 23, para. 37. See ECJ Case C-157/07 Krankenheim, ECLI:EU:C:2008:588, where such rules were considered to be in line with the fundamental freedoms in the Treaty on the Functioning of the European Union. 2

The Future of Avoiding Double Taxation 321 When applying the credit method, the foreign income is included in the tax base in the residence state, but the foreign tax will be deducted from the tax imposed. Various alternative designs of the credit method are conceivable. An alternative that is usually referred to as a full credit provides that the foreign tax must always, in full, be deducted (or repaid) in the residence state. The variant used in article 23B of the OECD Model is an ordinary credit, where a credit limitation applies. According to this rule, the maximum credit is equal to the residence state’s tax, before credit, that amounts to the foreign income. Hence, the credit limitation is in principle identical to the alternative design of the exemption method described earlier. In detail, the credit limitation can be designed in different ways. It is possible to use an overall limitation, according to which a single maximum credit is calculated for the taxpayer’s total foreign income. Another possibility is to calculate a separate maximum credit for income from each foreign jurisdiction (per country limitation). Finally, it is conceivable to calculate a separate maximum credit for different categories of income, such as active and passive income, or for each specific item of income. In the OECD Model, a per country limitation applies.3 A special use of the methods takes place in the case of dividends (and other returns) on substantial shareholdings within the corporate sector. When the recipient state taxes the dividends, it causes both international juridical double taxation in relation to foreign withholding taxes on dividends, and international economic double taxation in relation to the foreign taxes on the profits in the foreign company. If the exemption method is applied by making the dividend tax-free, both of these forms of double taxation can be mitigated at once. However, in order to achieve the same objective with the credit method, not only the credit described earlier is required, but also an indirect credit that allows the crediting of foreign taxes on the company profits of the underlying companies. A comparison between exemption and credit shows that exemption gives a more favourable result for the taxpayer if the foreign tax is lower than the tax in the residence state. In such a situation, the credit method means that the residence state has a residual tax claim after the foreign tax has been set off. However, in the variants used in the OECD Model—exemption with progression and ordinary credit—the reverse situation has the same tax consequence regardless of the method used. If the foreign tax is higher than the tax in the residence state, it is the tax level of the source state that will be decisive for the overall tax burden. It has traditionally been considered that exemption promotes capital import neutrality—or competitive neutrality—because it is the tax level of the source state (or establishment state) that decides the tax outcome. Full credit has instead been considered to promote capital export neutrality, as an investor does not get different tax outcomes depending on where the investment is destined. However, the ordinary credit does not follow any uniform neutrality principle, apart from the fact that it counteracts distortions in general that international double taxation is mitigated.

3 See

K. Vogel, Klaus Vogel on Double Taxation Conventions, 3rd ed. (Boston, MA: Kluwer Law International, 1997), art. 23 m.no. 167.

322 Martin Berglund Instead, the ordinary credit may be justified by other considerations. An important consideration is that the residence state has a responsibility to eliminate international double taxation, but not to allow further tax relief than that. The double taxation is considered to be eliminated if the foreign income is taxed in total in accordance with the tax rate of either the residence state or the source state.

19.3 Design and Applicability of the Method Articles The early international work to create internationally accepted methods to avoid double taxation was carried out by the League of Nations. The first treaty model, the Geneva Model, provided for the credit of foreign taxes in the residence state.4 The design of this method has clear similarities with what is today called ordinary credit. Article 23 of the Mexico and London Models also provided for the application of ordinary credit in the residence state, although in particular the Mexico Model also contains significant elements of exclusive withholding tax.5 Both exemption with progression and ordinary credit were included in articles 23A and 23B, respectively, of the 1963 OECD Model. The design of these articles has since remained largely unchanged for many decades. In the 2017 OECD Model, the following applies. In article 23A, which contains the exemption method, paragraph 1 expresses the basic provision that foreign income should be exempted from tax in the residence state. Paragraph 2, however, provides for a change of method to the credit in respect of income which may be taxed in the residence state in accordance with articles 10 (dividends) and 11 (interest). Finally, the safeguard of progression is regulated in paragraph 3. The fourth paragraph of the article, which was introduced in the OECD Model in 2000, states that exemption shall not be allowed if ‘the other Contracting State applies the provisions of this Convention to exempt such income or capital from tax or applies the provisions of paragraph 2 of articles 10 or 11 to such income’.6 Article 23B, containing the credit method, begins with a first paragraph stating that, under certain conditions, foreign tax is to be deducted from the income tax in the residence state. The last part of the first paragraph prescribes a credit limitation. The second paragraph of the article provides for a similar progression safeguard as in article 23A(3). 4 League of Nations, Double Taxation and Tax Evasion. General Meeting of Government Experts on Double Taxation and Tax Evasion (Oct. 1928), 11–12 and 16–17. It should be mentioned that the foreign tax credit was introduced as early as 1918 in the US federal income tax system; see S. Surrey, ‘Current Issues in the Taxation of Corporate Foreign Investment’, Columbia Law Review 56 (1956), 819. 5 League of Nations Fiscal Committee, London and Mexico Model Tax Conventions: Commentary and Texts (Geneva: League of Nations, 1946), 66–69. 6 Such a requirement was proposed in the OECD report, The Application of the OECD Model Tax Convention to Partnerships (Paris: OECD Publishing, 1998). The issue is discussed in e xample 14 of the report; see esp. at 113.

The Future of Avoiding Double Taxation 323 If an item of income has been exempted from tax according to the tax treaty’s allocation rules, it can still be taken into account when applying the credit limitation. A general condition for the application of both articles 23A(1) and 23B(1) is that the income ‘may be taxed in the other Contracting State in accordance with the provisions of this Convention’. This means that the residence state should only mitigate double taxation in a situation where the income that is taxable in the residence state is covered by an allocation rule that gives taxation rights to ‘the other state’ (which can usually be characterized as a source state). There is thus a condition that it must be at least a potential international juridical double taxation in order for the methods to be applied. On the other hand, there is no general condition that such double taxation should actually have occurred. The applicability of the exemption method in the residence state is dependent only on the other state being allowed to tax the income in accordance with the tax treaty. In principle, there is no need for actual double taxation, but on the other hand, the applicable allocation rule of the treaty must allow the income to be taxed in both states. However, difficulties arise when states apply the agreement in different ways. This can lead to unresolved double taxation or double non-taxation. How these situations are to be handled depends on the interpretation of the earlier condition ‘may be taxed’ and article 23A(4). This is discussed in more detail in connection with the changes introduced in the method articles in connection with the BEPS Project. In order for the credit method to apply, it must as a main rule be a situation where the same tax subject has been taxed in two countries for the same income. It is not possible to allow a credit for anything other than foreign taxes that have been paid, and so an actual double taxation is required for the credit method to be applied. In the same way as with the exemption method, difficulties arise in assessing situations where the contracting states have considered the income asymmetrically and applied the tax treaty differently. The income may have been attributed to different taxpayers, or categorized under different allocation rules. Unlike the exemption method, however, such differences mainly entail that there is a risk of unresolved double taxation.

19.4 Exemption vs Credit: Switchover Rules The choice between exemption and credit is largely a choice between a global and territorial tax system. With credit, the residence state’s global tax claim is maintained at least as a starting point, while exemption means that this claim is circumvented. Modern income taxes do not constitute purely global or territorial systems, but contain elements of both to varying degrees.7 7 Concerning the use of exemption and credit in general, see the comparative overview in H. Ault, B. Arnold, and G. Cooper, Comparative Income Taxation: A Structural Analysis, 4th ed. (The Hague: Wolters Kluwer, 2020), 582–589.

324 Martin Berglund The exemption method appears in some cases—especially when it comes to various forms of business activities—as the fundamentally correct alternative, for example from a neutrality perspective. Another argument for the exemption is that it is often easier to apply than the credit. At the same time, the credit has, as previously mentioned, a built-in protection against (intended or unintended) tax avoidance in the form of double non-taxation and undesired low taxation. Relief is only allowed if there is a foreign tax, and the residence state keeps a residual tax claim. The exemption, on the other hand, can regularly lead to double non-taxation and undesired low taxation, since there is no general requirement that an income has been taxed in the source state. Various measures to prevent such outcomes have been introduced in many countries. This means that the exemption is not necessarily easier to apply than the credit. In individual income taxation, the safeguard of progression—or the alternative exemption—means that the exemption also requires similar calculations as the credit. When it comes to dividends (and other returns) on substantial cross-border holdings in the corporate sector, the exemption (‘participation exemption’) is nowadays the dominant approach. Large Western economies such as the USA, UK, and Germany have in recent years and decades changed to exemption systems instead of the indirect credit. Regarding the USA, the change was implemented in the 2017 tax reform (the Tax Cuts and Jobs Act). China should be mentioned as an example of a major economy that uses the indirect credit. As for the use of the methods to mitigate international juridical double taxation, the picture must be described as relatively fragmented. However, it appears that the exemption method is relatively common when it comes to permanent establishments and employment income. The credit has instead had a major impact on the taxation of passive income. Partly due to the increased use of exemption, but also as a result of the increased focus on tax avoidance, there has been a gradually more extensive use of different types of switchover rules where there, under certain circumstances, is a change of method from exemption to credit. Such a change typically takes place when the foreign income is low-taxed. In the case of major holdings in foreign companies—controlled foreign companies (CFCs)—these rules not only mean that there is a change of method, but also that you disregard the legal form and include the foreign company’s income in the domestic owner company’s tax base. CFC taxation regimes have been in place in many jurisdictions for a long time, but have been expanded in recent years. Special mention should be made here of the US Subpart F legislation, which was introduced in the 1960s.8 Since 2017, the US taxing claims on CFCs have been expanded by a provision on

8 Internal Revenue Code, ss. 951–964. It should be noted that prior to the reform in 2017, these rules rather led to a switchover from one form of credit (in the case of taxation of dividends on major holdings in foreign companies) to another form of credit (in the case of current taxation on such holdings).

The Future of Avoiding Double Taxation 325 Global Intangible Low-Taxed Income (GILTI).9 A CFC taxation regime has also been introduced at the EU level in the EU Anti-Tax Avoidance Directive (ATAD).10 Similar switchover rules have increasingly been applied in certain cases involving the methods to mitigate international juridical double taxation. This can be the case where a foreign permanent establishment is covered by the exemption method and subject to low taxation in the state of establishment.11 This is proposed in the OECD BEPS Action 3 Report.12 A regulation of this type is also found in article 7 ATAD, which will be briefly described here. According to this article, a member state must, under certain conditions, treat not only ‘an entity’ but also ‘a permanent establishment of which the profits are not subject to tax or are exempt from tax in that Member State’ as a CFC. The condition is that the permanent establishment in the state of establishment is taxed at a lower effective rate than the difference between the corporation tax that had been imposed on the permanent establishment in the taxpayer’s residence state and the tax actually imposed in the state of establishment. In practice, this means that an effective taxation in the source state that is less than half of what the residence state’s tax would have been leads to an obligation for the residence state to apply a switchover from exemption to credit. Switchover rules regarding permanent establishments entail a higher risk of tax treaty override than the CFC rules described previously for major holdings in foreign companies. Although the BEPS Project has not resulted in an implementation of switchover rules in the OECD Model, it has resulted in the adoption of article 5 of the Multilateral Instrument which contains three optional alternative provisions on a method change from exemption to credit. According to the first rule, exemption must be switched to credit in situations where the source state applies the provisions of a tax treaty to exempt the income from tax or to limit the maximum tax rate to which the income is subjected. The provision is similar to article 23A(4) in the OECD Model, but with the difference that a switchover to credit is expressly provided for. The second rule prescribes a change from participation exemption to credit, if the payer of a dividend has been granted a deduction in the source state. This rule is obviously only relevant if the tax treaty provides for participation exemption to be granted in certain situations, which, as mentioned, is not the case for tax treaties based on the OECD Model. The third rule does not prescribe a switchover rule in the sense that credit will be applied instead of exemption in certain

9 Internal Revenue Code, s. 951A. Regarding the GILTI regime, see P. Gautrin, ‘United States—US Tax Cuts and Jobs Act, Part 1, Global Intangible Low-Taxed Income (GILTI)’, Bulletin for International Taxation 73/1 (2019). 10 ATAD, arts 7–8. 11 An example is the German rules that were tried in ECJ Case C- 298/05 Columbus Container Services, ECLI:EU:C:2007:754. Under the German tax rules that were applicable in the national court case, a change from exemption to credit took place if the income in question would have been subject to CFC taxation if it had belonged to a foreign subsidiary. Even today, there seem to be different types of switchover rules in German national law; see Ault, Arnold, and Cooper, Comparative Income Taxation, 583–584. 12 OECD BEPS Action 3 Report, 22, para. 28.

326 Martin Berglund situations. Instead, a general method change from exemption to credit is prescribed for tax treaties that contain the exemption method.

19.5 Impact of the 2015 BEPS Reports on the Method Articles So far, no major changes have been introduced to the method articles in the OECD Model as a result of the BEPS Project. The reason for this is that the project has largely dealt with counteracting various forms of tax avoidance that can give rise to double non-taxation and other undesired forms of low taxation. As tax treaties aim to mitigate double taxation, and as a rule can only limit the application of national tax law, they are not always an appropriate tool to counteract double non-taxation and low taxation that arise as a result of the interaction between national tax systems. An example is hybrid mismatches, which are largely to be resolved through legislation at national level. The changes introduced in the 2017 OECD Model after the first BEPS reports are mainly addressed at counteracting risks of tax avoidance that may arise through the application of the tax treaty itself. An example is the revised article 4(3). In order to assess why only limited changes were introduced in the method articles, it is necessary to first discuss which parts of the articles may function as a protection against (intended or unintended) tax avoidance. As stated in Section 19.3, the exemption method is exposed to the risk of being applied to situations that do not constitute double taxation. Such a situation may arise if the applicable distribution article allocates the taxing rights to both states, while the method article consists of article 23A. If the source state does have any taxing claims according to its national law, double non-taxation arises. The application of the credit method cannot, in principle, give rise to double non-taxation or similar effects, since an imposed and paid foreign tax is required for credit to be allowed. In special cases, however, situations may arise where the credit is not applied to avoid double taxation.13 Article 23A counteracts certain forms of double non-taxation and undesired low taxation. First, it may be noted that according to article 23A(2), a switchover to credit is prescribed when the income is covered by one of the treaty’s tax rate limitations in articles 10 and 11. Secondly, as mentioned in Section 19.3, the article requires that the source state has a right according to the treaty to tax the income. The traditional design of article 23A thus entails counteracting such double non-taxation and undesired low taxation that arise due to the application of the tax treaty. If, on the other hand, an item of income is for some reason exempted according to national law in the source state, double non-taxation can arise. Tax treaties are traditionally not designed to counteract

13

See e.g. the example in OECD BEPS Action 2 Report, 281–283.

The Future of Avoiding Double Taxation 327 double non-taxation and undesired low taxation that arise due to the application of national law. Article 23A can also counteract cases where double non-taxation arises due to different applications of the treaty allocation rules in the contracting states. As mentioned, a general condition for the application of the method articles is that the income ‘may be taxed’ in the source state. According to the OECD approach, a distinction must be made depending on the reason why the treaty has been applied differently.14 If the income ends up in different allocation rules due to differences in the national law of the countries, the residence state must consider the application of the source state in assessing whether the income may be taxed in that state. If, instead, the asymmetrical application is due to the states’ assessing the circumstances in different ways or interpreting the articles differently, the residence state shall not consider the source state’s application of the agreement. However, the application of the source state must always be followed if it means that that state has no right to tax the income according to the agreement. This means that double non-taxation cannot arise because the articles of the tax treaty are applied differently in the different states. A similar condition is provided for in article 23A(4), which refers to a specific situation where the two states apply the allocation rules differently. The residence state may have concluded that there is a right to tax the income of the source state, so the conditions in article 23A(1) are met. At the same time, the source state may have made the assessment that another distribution article should be applied, which either prevents the source state from taxing the income or limits the source state’s tax collection. In this case, exemption will not be provided. In a situation where the source state has no tax claims due to national law, there is no possibility of denying exemption in the residence state.15 The difference from the rules dealt with in Section 19.4 is that no switchover is provided for credit if the residence state does not have to allow an exemption under article 23A(4). However, in situations where the source state applies a distribution article that limits the tax rate (article 10 or 11), it is considered possible to apply article 23A(2) by analogy.16 In connection with the BEPS Project, only a limited addition was introduced in the method articles, which aims to ensure that the methods are only applied to double- taxation situations. The condition that a certain income ‘may be taxed in the other Contracting State in accordance with the provisions of this Convention’ has been provided with an exception, which states that this condition is not fulfilled if ‘these provisions allow taxation by that other State solely because the income is also income derived by a resident of that State’. This reservation refers to hybrid entities.17 Assume

14

OECD Commentary on art. 23, paras 32.1–32.7. Ibid., paras 34 and 56.2. 16 A. Rust, ‘Article 23’, in E. Reimer and A. Rust, eds, Klaus Vogel on Double Taxation Conventions, 4th ed. (Alphen aan den Rijn: Kluwer Law International, 2015), art. 23 m.no. 103. See Section 19.4, regarding art. 5 of the Multilateral Instrument. 17 See OECD Commentary on art. 23, paras 11.1–11.2. 15

328 Martin Berglund that an entity, typically a partnership, is considered a resident of one state while its partners are considered residents of another state. The state in which the partnership is a resident treats it as a separate tax subject, while it is considered transparent in the other state. Both states thus tax the partnership income as if it were earned by a resident taxpayer. The new addition to the articles means that a state only has an obligation to mitigate double taxation if the tax treaty would confer taxing rights to the other state even if that state would not have been a residence state. It is thus a kind of hypothetical test to be applied in these situations. In my opinion, it followed implicitly from the previous wording that the method articles could only be applied if the other state was entitled to tax the income in its capacity as ‘the other state’ than the residence state. All in all, there are thus certain mechanisms in the method articles that ensure that only double-taxation situations are covered. However, there has traditionally been a reluctance to address situations where it is exclusively due to national law that no double taxation has arisen. In my view, this reluctance can be explained by the fact that it has not been considered part of the function of tax treaties to counteract tax avoidance in general. It is mainly if double non-taxation or undesired low taxation has arisen as a consequence of the application of the tax treaty itself that it has been considered justified to counteract it. In these situations, it would disrupt the balance in the treaty to allow double tax relief. Such undesired effects were already dealt with in tax treaties before the first BEPS reports were published. My impression is that these reports, which in parts were implemented in the 2017 OECD Model, followed the traditional view that double non-taxation and undesired low taxation stemming from national law should not be dealt with in tax treaties. Correspondingly, the changes introduced as a result of the first BEPS reports had a very limited impact on the design and application of the method articles.

19.6 Potential Impact of OECD Pillar One on the Methods The methods for mitigating international juridical double taxation are generally not applicable in the area of transfer pricing. If a state makes a transfer pricing adjustment, international economic double taxation arises, which is generally dealt with through the application of a tax treaty provision identical to article 9(2) of the OECD Model. If the other state, which has previously taxed the corresponding profit, considers that the adjustment was correct, that state shall eliminate the double taxation by reducing the taxable profit. To some extent, this is can be characterized as the application of a conditional variant of the exemption method. According to the OECD Pillar One Blueprint, however, the methods of mitigating international juridical double taxation seem to be given an important role in the allocation of taxing rights between

The Future of Avoiding Double Taxation 329 the different countries to which the activities of a multinational enterprise (MNE) are connected.18 The OECD Pillar One Blueprint seeks to adapt the taxation of MNEs to the digitalization of the economy. The proposal suggests that a certain proportion (referred to as ‘Amount A’) of the residual profits of certain MNEs should be allocated to market jurisdictions.19 Since the individual companies in the group are also subject to ordinary corporation tax in different jurisdictions, double taxation arises within the group. To some extent, the group’s profits are thus taxable within the framework of both Amount A and the ordinary corporation tax. To resolve this double taxation, the OECD seeks to apply a method to allocate Amount A to specific entities in the group that has been taxed.20 The idea seems to be that this allocation will make it a question of international juridical double taxation that can be mitigated with methods similar to articles 23A and 23B of the OECD Model. The residual profit that is taxable as Amount A must therefore be broken down from a group level to a company level. It is not considered possible to allocate Amount A to individual companies by using the same method as proposed for calculating Amount A at a group level (non-routine profits before taxes/revenue). The individual companies in the MNE may have intra-group transactions that affect earnings. Different accounting standards can also be applied in the different companies. Instead, a four-step test is advocated to assess to which companies in a particular group Amount A should be assigned. First, a qualitative determination must be made of which companies in the group contribute to creating residual profits within the group. This is strongly influenced by transfer pricing principles, and the OECD Transfer Pricing Guidelines are relevant. Secondly, Amount A should only be attributed to entities that have the capacity to bear the tax liability in relation to residual profits. Amount A should not be attributed to entities with low income or losses. It must be decided which entities in the group have residual profits, whereby financial accounts are used rather than tax accounts. Thirdly, entities must have a sufficient connection to a market jurisdiction. Finally, a pro rata distribution of Amount A may be carried out in situations where it is not possible to allocate the amount sufficiently according to the previous steps. The conclusion in the report is that the method described above for distributing Amount A to the individual companies within an MNE entails that international 18

After finalizing this chapter, on 11 July 2022 the OECD released a report entitled Progress Report on Amount A of Pillar One. The report contains detailed proposed rules on Amount A. The present section is based on the general recommendations in the OECD Pillar One Blueprint. 19 Concerning the personal scope of the rules, see OECD, Public Consultation Document, ‘Pillar One—Amount A: Draft Model Rules for Domestic Legislation on Scope’ (4 Apr. 2022). The proposed personal scope of the rules has been subject to several changes; see G. Cooper, ‘Building on the Rubble of Pillar One’, Bulletin for International Taxation 75/11/12 (2021), s. 4.2. 20 OECD Pillar One Blueprint, 135–148. See in this regard, S. van Weeghel, ‘Have the OECD Model and the UN Model Served Their Purpose? Are They Still Fit for Purpose’, Bulletin for International Taxation 75/11/12 (2021), s. 3.1.3, who states that ‘a bridge will have to be constructed between the Amount A world with formulary apportionment—a financial accounting tax base and a “consolidated” taxpayer—on the one hand, and the separate enterprise method—the single taxpayer of the existing treaty construct—on the other’.

330 Martin Berglund juridical double taxation arises at the level of the companies that have part of Amount A allocated to them.21 As far as I understand it, this conclusion is due to the fact that the allocation method aims to identify which entities in the group of companies actually have such residual profits to which Amount A refers. The result is that tax is levied in a market jurisdiction to which the company belongs, while it is also subject to corporate taxation on the residual profits in the residence state. The additional taxation that Amount A entails would therefore result in an imposition of taxes in two jurisdictions on the same person in respect of the same income. Since this constitutes international juridical double taxation, it is proposed that the two methods—exemption and credit— should be used to mitigate this. In the report, it is noted that it would have been more appropriate to use the method in article 9(2) of the OECD Model if it had been a question of international economic double taxation. It is not clear to what extent the design and application of exemption and credit in these situations are intended to be based on existing rules in the OECD Model and national law. In the discussion, the report initially refers to articles 23A and 23B of the OECD Model and the UN Model. The proposal is, however, that the taxing rights to Amount A should not be regulated by existing tax treaties, but by new national rules in conjunction with a new multilateral tax treaty.22 Such a treaty should also include mechanisms for mitigating double taxation. Although the report discusses whether existing national rules would be sufficient to ensure that double taxation is eliminated, it notes that this is not always the case. An example mentioned is where the residence state has source rules that must be applied when mitigating double taxation.23 The application of such rules may lead to the conclusion that the income taxed in the market jurisdiction in the form of Amount A does not have its source in that state. This could mean, for example, that the residence state denies a credit for the foreign tax. Thus, as I understand the proposal, the new multilateral tax treaty should introduce separate methods for countering double taxation arising from the taxation of Amount A. These methods would nevertheless be aimed at mitigating international juridical double taxation in the residence state for the company that has also been taxed in a market jurisdiction. The suitability of this design can be discussed. The four-step method proposed for allocating Amount A to specific companies in a group can hardly be a guarantee that the residual profits that constitute Amount A have, in fact, been subject to corporate taxation in those companies.24 It is sufficient to consider the pro rata distribution to be applied in the last resort. The method would, in many cases, mean that a certain entity is taxed in a market jurisdiction, even though the corresponding profits have not been

21

OECD Pillar One Blueprint, 148–152. Ibid., 201–202, paras 822–831. See J. Wheeler, ‘Tax Treaties: What Are We Going to Do with Them?’, Bulletin for International Taxation 75/11/12 (2021), s. 6, on the complexity that may arise due to such a multilateral treaty. 23 OECD Pillar One Blueprint, 202, para. 831. 24 For a similar criticism, see R. Collier, M. Devereux, and J. Vella, ‘Comparing Proposals to Tax Some Profit in the Market Country’, World Tax Journal 13/3 (2021), s. 4.1.2. 22

The Future of Avoiding Double Taxation 331 included for the same company within the framework of corporate taxation in the residence state. If credit is to be applied, situations could therefore arise where the tax in the residence state is not sufficient to set off the foreign tax. This is particularly true given that these rules are to be applied in parallel with the ordinary method articles in existing tax treaties. As I see it, it would have been better to accept the fact that double taxation that may be involved can often be economic in nature, since the profits on which Amount A may be attributed differently in ordinary company taxation than according to the four-step method. It would have been preferable to design a special rule to deal with this double taxation, instead of starting from the existing methods in the OECD Model that are not designed to deal with this type of double taxation.25

19.7 Potential Impact of OECD Pillar Two on the Methods The OECD Pillar Two Blueprint mainly proposes rules to ensure that MNE groups are taxed at an effective tax rate that at least amounts to a certain minimum level. This is ensured through various mechanisms.26 Important parts of this proposal were in 2021 implemented in the OECD Global Anti-Base Erosion Model Rules. A central component of the Pillar Two Blueprint proposal and the Model Rules is the so-called income inclusion rule, which means that the parent company of a group must pay a top-up tax on income in low-taxed companies within the group, in order to ensure that the income is taxed at a minimum rate.27 As noted in the Blueprint, this is similar to CFC taxation. In this current taxation of low-taxed subsidiary income, the parent may credit the taxes that have been levied abroad. The introduction of an income inclusion rule thus entails an extended application of a switchover from exemption to credit. When applying the income inclusion rule, the parent company’s permanent establishments that are exempted from tax in the residence state must be treated in the same way as directly owned subsidiaries.28 This means that income in such a permanent establishment that is low-taxed abroad must be included in the residence state, and a minimum tax is imposed. Technically, this is resolved in the Model Rules by generally including permanent establishments (PEs) as ‘constituent entities’, and allocating

25 The Inclusive Framework statement from 1 July 2021 does not clarify this issue. The following is stated: ‘Double taxation of profit allocated to market jurisdictions will be relieved using either the exemption or credit method. The entity (or entities) that will bear the tax liability will be drawn from those that earn residual profit’ (2). See https://www.oecd.org/tax/beps/statement-on-a-two-pillar-solut ion-to-address-the-tax-challenges-arising-from-the-digitalisation-of-the-economy-july-2021.pdf. 26 There are three important parts of the proposal: an undertaxed payments rule, an income inclusion rule, and a subject-to-tax rule. In this chapter, I focus on the two latter rules. 27 OECD Pillar Two Blueprint, ch. 6 and OECD Pillar Two Model Rules, art. 2.1–2.3. 28 Ibid., 121, paras 453–454.

332 Martin Berglund taxes on PE income to the PE.29 The effect of these rules has similarities with the rules discussed in Section 19.4, regarding switchover rules from exemption to credit, which today exist in the national law of some countries and in the ATAD. In the report, it is noted that this type of rule may be in conflict with tax treaties, if the basis for the permanent establishment being exempted in the residence state is that the exemption method is applied in the applicable tax treaty.30 Therefore, it is proposed that the Inclusive Framework should continue to work on the design of a switchover rule which, as far as I understand, could be included in current tax treaties.31 One way to introduce such a switchover rule would be to include it in the Multilateral Instrument. In a footnote in the Blueprint, it is also considered that such a switchover rule could be designed more broadly, and thereby involve a general change from exemption to credit if income in an exempted permanent establishment is low-taxed abroad.32 In the Pillar Two Blueprint, there is also a proposal for a subject-to-tax rule, which is to be introduced in tax treaties (possibly also in the Multilateral Instrument).33 This rule is a type of supplement to the income inclusion rule. The idea is that the source state in the case of certain cross-border payments (mainly interest and royalties) must ensure that the payment is taxed at a minimum level in the recipient’s residence state. When designing the subject-to-tax rule, the question arises as to what relevance is to be attached to the residence state applying the exemption method according to the tax treaty’s method article. In this connection, the OECD seems to mean that the source state should not take this circumstance into account when calculating the minimum tax to be imposed on the payment.34 This means that the source state’s taxation of the payment must be calculated before any exemption or credit that may be granted in the recipient’s residence state. The justification for this is that the subject-to-tax rule should not reallocate taxing rights under the treaty, but only ensure that income is taxed at a minimum level. Considering how the method articles are currently designed, however, the residence state of the recipient would need to take into account the tax imposed in the source state on the basis of the subject-to-tax rule when applying the methods.35 This means that any tax that have been imposed in the residence state, and which was taken into account when the source state applied the subject-to-tax rule, will be reduced by

29

OECD Pillar Two Model Rules, arts 1.3, 3.4, and 4.3. Ibid., 121–122, paras 455–456. 31 See also OECD Commentary on Pillar Two Model Rules, 24. 32 Ibid., 122, fn. 3. 33 Ibid., ch. 9. 34 Ibid., 165, paras 645–648. 35 Ibid., 166– 167, paras 652–660. Taxes imposed under the subject-to-tax rule also need to be considered and credited when applying the income inclusion rule; see ibid., 16, para. 20 and 171, para. 671. See also H. Loukota, ‘Have the OECD and UN Models Served Their Purpose?’, Bulletin for International Taxation 75/11/12 (2021), s. 2.4.2; L. E. Schoueri, ‘Some Considerations on the Limitation of Substance-Based Carve-Out in the Income Inclusion Rule of Pillar Two’, Bulletin for International Taxation 75/11/12 (2021), s. 3. 30

The Future of Avoiding Double Taxation 333 applying the method article. Either the income is exempted from tax under article 23A, or the tax that the source state imposed under the subject-to-tax rule may be credited in the residence state under article 23B. In certain situations, this would mean that the effect of the subject-to-tax rule is that the taxation in the residence state, that has been exercised in accordance with the tax treaty would only be transferred to the source state. Paradoxically, this also means that the total tax burden would be higher if the residence state has no taxing claims on the income at all, than if the residence state has a claim to tax the income at a low rate. The report is therefore considering the possibility of introducing a rule stating that the obligation to grant exemption or credit does not apply when the source state exercises a taxing right on the basis of the subject-to-tax rule. In situations where the source state exercises its usual taxing rights under a tax treaty, but where the total tax burden is still too low, it is also proposed that the residence state should not completely mitigate the double taxation. Taken together, the proposals in the OECD Pillar Two Blueprint suggest that the design of the tax treaty methods of mitigating international double taxation should be adjusted to some extent to prevent double non-taxation or undesired low taxation that stems from lacking taxing claims in national law. This constitutes a departure from how tax treaties have traditionally been designed to handle such risks in the application of the method articles. As noted in Section 19.5, rules in the tax treaties to limit the application of the method articles have traditionally been introduced if it would disrupt the balance in the treaty to allow double tax relief.

19.8 Concluding Remarks The design of the methods for eliminating international double taxation in international taxation is traditionally characterized by certain distinctions and demarcations around which the system has been structured. The work of reforming the field of international tax law within the framework of the BEPS Project has gradually meant that this traditional structure has been weakened, and in the long run it may be abandoned altogether. The following are some examples of this development, in the light of the discussion in previous sections. First, the credit method seems to be increasingly given an overarching function of ensuring that the methods of mitigating double taxation do not yield a significantly more favourable result than is required to achieve single taxation.36 This is reflected in the various switchover rules that have been introduced in different national laws,

36 Regarding post- BEPS developments on the ‘single tax principle’, see e.g. R. Avi-Yonah, ‘The International Tax Regime at 100: Reflections on the OECD’s BEPS Project’, Bulletin for International Taxation 75/11/12 (2021), s. 3.2; W. Schön, ‘Is There Finally an International Tax System?’, World Tax Journal 3/3 (2021), s. 3.2.

334 Martin Berglund EU law, and the Multilateral Instrument, and also proposed in the OECD Pillar Two Blueprint.37 Secondly, the distinction between international juridical and economic double taxation has traditionally played a significant role in designing international tax law, as there are largely different considerations behind the mitigation of the various forms of double taxation. Increasingly, however, international juridical double taxation that arises for foreign permanent establishments is handled in a coordinated manner in relation to corresponding rules for major cross-border company holdings. This coordination includes, in particular, the switchover rules which entail that the exemption method is changed to the credit method in certain situations, mainly when the activity in question is low-taxed abroad. Another example is the proposed approach to eliminating double taxation on Amount A that, according to the OECD Pillar One Blueprint, is to be allocated to market jurisdictions. This is really a question of profit distribution in multinational company groups, which can give rise to both legal and economic double taxation. Thirdly, the method articles in tax treaties have traditionally not generally addressed the risks of such double non-taxation or undesired low taxation that may arise solely from the application of national laws of the contracting states. An exception is that it is inherent in the credit method that only actual taxation in the source state can be credited. This demarcation seems to have been maintained even after the 2015 BEPS reports and the changes that these entailed in the 2017 OECD Model. On the other hand, the proposals included in the OECD Pillar Two Blueprint mean that the method articles would need to be designed to a greater extent to counteract the mentioned risks. Finally, the methods of mitigating double taxation are regulated by an increasing number of sources of law, which make the area complex. This complexity increases if, in accordance with the OECD Pillar One Blueprint, a new multilateral treaty is introduced regarding Amount A, with special rules for the mitigation of double taxation which are based on the OECD Model. The same can be said concerning the changes in the methods following the OECD Pillar Two Blueprint proposal of an income inclusion rule and a subject-to-tax rule, which could possibly be included in the Multilateral Instrument. If these various developments continue, in my opinion it is necessary to review the design of the method articles in the OECD Model, and perhaps even the function of tax treaties in general.

37 P. Kaka, ‘From the Avoidance of Double Taxation to the Avoidance of Double Non-Taxation: The Changing Objectives of Tax Treaties’, Bulletin for International Taxation 75/ 11/ 12 (2021), s.3, who considers whether a general shift from exemption to credit could be an alternative to the Pillar Two proposals.

Chapter 20

Charities i n Tax C onvent i ons Sigrid Hemels

20.1 Introduction The OECD observed in its 2020 tax policy study Taxation and Philanthropy that most countries provide tax incentives for philanthropic giving and philanthropic entities.1 Examples are exemptions from corporate income tax and gift and inheritance tax and personal income tax deductions for donations to charities. Definitions and tax legislation applicable to charities vary widely.2 The terminology to refer to these organizations also differs. This includes philanthropic organizations, non-profit organizations, public benefit organizations, and NGOs. This chapter uses the term ‘charities’ to refer to this type of non-governmental public benefit organization. Charities cross borders in various ways including through international initiatives, investments, and fundraising. This may result in double taxation, both by the resident state of the charity and by the source state of the income. For example, a charity that invests in foreign real estate, shares, or bonds may be confronted with foreign corporate income tax, dividend withholding tax, or interest withholding tax while at the same time the income is taxable in the resident country. If the resident state exempts the charity, the foreign taxation renders such exemption less effective. Where the resident state wants to 1 OECD, Taxation and Philanthropy, OECD Tax Policy Studies no. 27 (Paris: OECD Publishing, 2020), 22–23. 2 See, e.g., S. Heidenbauer, Charity Crossing Borders: The Fundamental Freedoms’ Influence on Charity and Donor Taxation in Europe (Alphen aan den Rijn: Kluwer Law International, 2011); A. Cédelle, ‘The Taxation of Non-Profit Organizations after Stauffer’, in W. Haslehner, G. Kofler, and A. Rust, eds, European Tax Law Classics (Alphen aan den Rijn: Kluwer Law International, 2015), 207–233; F. Vanistendael, ed., Taxation of Charities (Amsterdam: IBFD, 2015); OECD, Taxation and Philanthropy; Philanthropy Advocacy, Legal Environment for Philanthropy in Europe (Brussels: European Foundation Centre, 2020).

336 Sigrid Hemels exempt the charity from corporate income tax because of its beneficial activities, the charity is still confronted with tax because of its activities abroad. In that case, there is only single taxation, but in a situation where the objective was to have no taxation. Similarly, a national gift or inheritance tax exemption for charities is less effective if a cross-border gift or inheritance is taxed in the country where the donor or deceased was resident. Another complication for charities that want to attract cross-border donations is that under the national legislation of their donor’s country of residence, gift deduction may only be allowed on donations to resident charities. Many countries have concluded bilateral treaties for the avoidance of double taxation (double taxation conventions (DTCs)) to solve or mitigate situations of double income taxation. Model conventions (MCs) usually serve as the starting point for negotiations. The IBFD Tax Research Platform3 includes over 9,000 DTCs on income and capital. Many of those are based on the Income and Capital Model Convention of the OECD (OECD MC) or the UN Model Double Taxation Convention between Developed and Developing Countries (UN MC). In addition, the US Model Income Tax Convention (US MC) is influential. MCs provide various solutions for international taxation problems which charities might encounter, for example by only allowing the source state to tax proceeds from real estate,4 and by reducing the amount of tax a source state can levy on outgoing dividends,5 interests,6 and royalties.7 DTCs to prevent double gift or inheritance taxation are not as common as DTCs on income and capital. Fewer than 200 are included in the IBFD Tax Research Platform. Again, MCs can serve as a basis for treaty negotiations. These include the 1982 OECD Model Convention for the avoidance of double taxation with respect to taxes on estates and inheritances and on gifts (OECD EIGMC) and the 1980 US Estate and Gift Model Convention (US EGMC). The central question in this chapter is whether charities have access to these DTCs and, if so, whether they have a special position. This question is answered using the MCs as a proxy. Although MCs are not legally binding, their wording is often incorporated in the text of bilateral treaties. Therefore, commentaries and technical explanations that help in interpreting an MC may shed light on the meaning of a treaty based on that model. For that reason, the latest version of the OECD MC and Commentary, the 2017 version, the 2017 UN MC and Commentary, and the 2016 US MC are used as references. As two states that have agreed on a certain DTC are usually referred to as ‘contracting states’, that terminology is used in this chapter. Before discussing the access of charities to and their position in DTCs, the chapter analyses the reason why many countries restrict tax incentives to resident charities as

3

https://research-ibfd-org.eur.idm.oclc.org/#/. OECD MC, art. 6. 5 Ibid., art. 10. 6 Ibid., art. 11. 7 Ibid., art. 12. 4

Charities in Tax Conventions 337 this is often the cause of double or single taxation where the purpose of a tax incentive is no taxation.

20.2 Rationale for Only Applying Tax Incentives to Resident Charities In 2020, the OECD observed that available data suggest that cross-border philanthropy is growing, but that most countries do not permit tax relief for donations to foreign philanthropic entities.8 The same applies to other tax incentives for charities. Historically, most countries restricted the application of tax incentives to resident charities. Outside the EU, this is still the norm: Japan, the USA, and Australia only provide for incentives for domestic charities. The explanation that countries offer for this restriction is often based on the so-called public policy rationale for tax incentives for charities. This rationale argues that charities perform functions that governments would otherwise have to perform themselves. That explains why governments are willing to forgo tax revenue.9 The US House Committee on Ways and Means formulated this as follows in 1939: The exemption from taxation of money or property devoted to charitable and other purposes is based upon the theory that the government is compensated for the loss of revenue by its relief from financial burden which would otherwise have to be met by appropriations from public funds, and by the benefits resulting from the promotion of general welfare.10

Similarly, the German Federal Fiscal Court observed in 1966 that the purpose of the tax deductibility of gifts was in fulfilling domestic tasks.11 The court deemed it unclear on what reasonable grounds the German tax legislator should have favoured donations to foreign corporations. Based on the public policy rationale, it would not make sense to give incentives to foreign charities as they do not perform functions which the national government of the donor country would otherwise have to perform. Following the logic of this rationale, such activities should be funded by the government of the other country. This rationale was used by various governments when the restriction was challenged before the Court of Justice of the European Union (CJEU) in the first decade of this millennium. Germany and the UK argued in the Stauffer case that resident and non- resident charities are not in a comparable situation because a resident charity plays an 8 OECD, Taxation and Philanthropy, 33–34. 9

B. R. Hopkins, The Law of Tax-Exempt Organizations (Hoboken, NJ: Wiley, 2011), 13. H. Rep. No. 1860, 75th Cong., 3d Sess, 19 (1939). 11 Bundesfinanzhof, Urt. v. 11.11.1966, Az.: VI R 45/66. 10

338 Sigrid Hemels active role in its country of residence.12 The governments followed the logic of the public policy rationale stating that resident charities perform duties which would otherwise have to be carried out by government and burden the state budget and that the charitable activities of non-resident charities do not concern that country. In the Persche case, Germany, Spain, and France made a similar argument: if a member state abstains from tax revenue by granting tax incentives for gifts to resident charities, it is because doing so absolves that member state of certain charitable tasks which it would itself otherwise have to fulfil using tax revenues.13 However, most countries do not apply the public policy rationale in a consistent way. In 1969, Van Hoorn observed that in most countries the place where the philanthropic funds are used is not decisive.14 In the Stauffer and Persche cases, the CJEU came to the conclusion that the public policy rationale was not applied consistently.15 Von Hippel noted that no EU member state ever required a domestic connection so strict that it was necessary that the state was directly financially relieved of burdens or obligations by philanthropic activity.16 He pointed out that all EU member states permitted tax-privileged philanthropic activities that benefit foreigners living abroad. Similarly, the OECD observed that many countries allow domestic entities to operate abroad without losing their tax-favoured status.17 For that reason, the CJEU obliged member states to apply tax incentives to charities resident in other member states that otherwise met the requirements for the tax incentive. The CJEU does not require mutual recognition of charities. Countries may still impose their own requirements as long as these are not linked to residency or nationality. Notwithstanding the application of the fundamental freedoms, the tax problems of charities and their donors are not solved in the EU.18 Some countries such as Spain,19 Portugal,20 and Romania21 did not change their legislation and still apply a residency requirement. Germany has restricted the scope of tax incentives for philanthropic organizations by requiring a link with Germany.22 Some member states introduced strict 12

Case C-386/04 Stauffer, ECLI:EU:C:2006:568, 14 September 2006, paras 33–34. Case C-328/07 Persche, ECLI:EU:C:2009:33, 27 January 2009, para. 42. 14 J. van Hoorn, ‘General Report: The Possibilities and Disadvantages of Extending National Tax Reduction Measures, If Any, to Foreign Scientific, Educational or Charitable Institutions’, Cahiers de Droit Fiscal International vol. 54b (1969). 15 For detailed discussion of the CJEU jurisprudence on cross- border philanthropy, see S. J. C. Hemels, ‘Charitable Organisations’, in C. H. J. I. Panayi, W. Haslehner, and E. Traversa, eds, Research Handbook on European Union Taxation Law (Cheltenham: Edgar Elgar, 2020), 248–268. 16 T. von Hippel, Cross Border Philanthropy in Europe After Persche and Stauffer: From Landlock to Non-Discrimination? (Brussels: EFC and TGE, 2014). 17 OECD, Taxation and Philanthropy, 9. 18 R. Buijze, Tackling the International Tax Barriers to Cross- Border Charitable Giving (Amsterdam: IBFD, 2020). 19 M. Márquez de la Calleja, 2020 Legal Environment for Philanthropy in Europe: Spain Country Profile (Brussels: European Foundation Centre, 2020), 29. 20 R. Gonçalves, M. Potes, and O. Taylor, 2020 Legal Environment for Philanthropy in Europe: Portugal Country Profile (Brussels: European Foundation Centre, 2020), 27. 21 OECD, Taxation and Philanthropy, 110. 22 Jahressteuregezetz 2009, of 19 December 2008, Bundesgesetzblattes part I, no. 63, p. 2794. 13

Charities in Tax Conventions 339 requirements for providing proof that foreign charities meet the national requirements. Surmatz and Forrest qualified the processes to gain equal treatment as burdensome, lengthy, and costly, a claim they substantiated with anecdotal and statistic evidence.23 For charities and donors outside the EU, the fundamental freedoms do not provide a solution for problems of double taxation, single taxation where the home state aimed for an exemption, and non-deductibility of gifts to foreign charities.24 In paragraph 70 of the Persche case, the CJEU observed that as a rule it is legitimate to refuse to grant tax incentives to philanthropic organizations in a non-EU member state if it is impossible to obtain the necessary information from that country, in particular because it is not under any international obligation to provide such information. This is where DTCs may come into the picture.

20.3 OECD MC The OECD MC does not mention charities. However, the 2017 Commentary makes reference to charities in several sections.

20.3.1 Scope of DTCs: Residency Article 1(1) of the OECD MC restricts the application of the treaty to residents of one or both contracting states. To assess whether a charity has access to a DTC, it must be established whether it is a resident. Article 4 of the OECD MC defines ‘resident of a Contracting State’ as any person who, under the laws of that state, is liable to tax therein by reason of their domicile, residence, place of management, or any other criterion of a similar nature. It does not include persons only liable to tax on income from sources in that state or capital situated therein. For charities benefiting from a corporate income tax exemption in their resident state, this raises the question whether they can be considered residents for DTC purposes and thus benefit from the DTC, for example from reduced withholding taxes. Section 8.11 of the Commentary on article 4 of the OECD MC notes that many states consider a person liable to tax even if tax is not actually imposed. The Commentary gives as an example charities that may be exempt from tax, but only if they meet all requirements specified in the tax laws. According to the Commentary, they are subject to tax. The Commentary observes that if they do not meet the requirements, they must pay tax. It argues that most states would view such entities as residents for the purposes 23 H. Surmatz and L. Forrest, Boosting Cross- Border Philanthropy in Europe (Brussels: European Foundation Centre, 2017). 24 T. Ecker, ‘Taxation of Non- Profit Organizations with Multinational Activities: The Stauffer Aftermath and Tax Treaties’, Intertax 35/8/9 (2007), 450–459.

340 Sigrid Hemels of the DTC. By this reasoning, exempt charities can benefit from the DTC. However, section 8.12 of the Commentary observes that some states do not consider exempt charities liable to tax or as residents unless they are expressly covered by the convention. This issue may be addressed in bilateral negotiations. This can be illustrated by the experiences of the Netherlands. The Netherlands strives to include in DTCs that all exempt entities are considered residents for tax treaty purposes.25 The reason the Netherlands takes this position is a 2009 ruling by the Dutch Supreme Court.26 The Supreme Court ruled that a Dutch association that was exempt from Dutch corporate income tax was not a resident of the Netherlands under the 1992 US–Netherlands DTC. Article 4(1) of the DTC deviates from the OECD MC. It states that an exempt pension trust (art. 35 of the DTC) or an exempt organization (art. 36) that is considered a resident according to national law, is considered a resident for DTC purposes. The exempt association did not qualify under articles 35 and 36 and therefore the Supreme Court did not deem it to be a resident under the DTC. Questions were raised on the implications of this judgment for DTCs with an article 4(1) in line with the OECD MC. Therefore, the Dutch tax treaty policy since 2011 is to strive for clarity. The Netherlands is of the opinion that, except for situations of abuse, there is no reason why an exemption provided by the state of residence should lead to an extension of the taxing rights of the source state. This is regarded as probably not being in line with the residence state’s aim pursued by an exemption for charities.27 Gooijer concludes that of the twenty-one DTCs the Netherlands has entered into (or that were amended) from 2012 up until 2019, eleven included a resident definition that deviated from the OECD MC and met the Dutch tax treaty policy.28 More specifically, article IV of the Protocol to the Switzerland–Netherlands DTC provides that the term ‘resident’ includes an organization that is established and operated exclusively for religious, charitable, scientific, cultural, sporting, or educational purposes (or for more than one of those purposes) and that is a resident of that state according to its laws, notwithstanding that all or part of its income or gains may be exempt from tax under the domestic law of that state. This provision is specifically aimed at making clear that exempt charities are residents for that DTC. Ten of the twenty-one DTCs did not include a deviation from article 4(1) of the OECD MC. However, in six cases the negotiations were on specific changes in an existing DTC and did not regard the residency provision. One of the five other DTCs is the Ireland–Netherlands DTC. According to the Dutch government, the Netherlands and Ireland agree that exempt entities qualify as treaty

25 Notitie Fiscaal Verdragsbeleid 2020 [Note on Tax Treaty Policy], Kamerstukken II, 2019–20, 25087, no. 256, s. 4.2. 26 Hoge Raad, 4 December 2009, 07/10383, ECLI:NL:HR:2009:BF0938. 27 Notitie Fiscaal Verdragsbeleid 2020, s. 4.2. 28 J. Gooijer, ‘De stichting en vereniging als inwoners voor toepassing van belastingverdragen’, Maandblad Belasting Beschouwingen 91/11 (2020), MBB 2020/39.

Charities in Tax Conventions 341 residents in line with paragraph 8.6 of the Commentary on article 4 of the OECD MC. For that reason, it was not deemed necessary to include a specific provision.29 For the DTC with China, the Dutch government notes that even though China did not want to deviate from article 4(1) of the OECD MC, it acknowledged the issue and the different views on residency of exempt entities. China was willing to be sympathetic if interpretation problems arose in concrete cases. These problems can then be addressed in a mutual agreement procedure. China did not deem it likely that differences in interpretation would arise as both countries deem, inter alia, that resident non-profit organizations are residents under the DTC.30 Only Iraq explicitly refused to agree on regarding exempt charities as residents, because it was deemed to be an extension of the personal scope of the DTC.31 Based on this Dutch experience, there is reason to believe that in many cases exempt charities will be regarded as tax residents under a DTC, but that this will not be the case under all DTCs.

20.3.2 Special Tax Regimes The Commentary on article 1 of the OECD MC (persons covered) includes a section on provisions to deny the application of specific treaty provisions, such as those of article 11 (reduction of interest withholding tax) and article 12 (reduction of withholding tax on royalties). The application of these DTC provisions would be denied with respect to income benefiting from so-called ‘special tax regimes’.32 The definition of ‘special tax regime’ can be included in the list of general definitions in article 3(1) of the OECD MC.33 Such regimes would, for example, result in preferential tax rates or reductions in tax bases. This could raise questions of the applicability of these rules to charities that are exempt from corporate income tax. If the exemption is regarded as a special tax regime, the charity cannot apply for a reduction of interest withholding tax in the source country. However, the Commentary notes that a regime that applies principally to, among others, organizations that are established and maintained exclusively for religious, charitable, scientific, artistic, cultural, or educational purposes would not meet the definition of a ‘special tax regime’.34 Regimes resulting in a special regime for charities would, therefore, fall outside the scope of this restriction.

29

Staten Generaal I/II, 2019–20, 35318, no. A; 1, p. 12. Kamerstukken II, 2012–13, 33718, no. 3, p. 8. 31 Staten Generaal I/II, 2019–20, 35324, no. A; 1, p. 10. 32 Sections 85–100. 33 Section 86. 34 Subdivision (iv) of the definition in s. 86. 30

342 Sigrid Hemels

20.3.3 Limitations on Benefits Clauses As a result of the OECD/G20 Base Erosion and Profit Shifting (BEPS) Project,35 countries agreed to include a general anti-abuse rule, the principal purpose test (PPT), and/or a specific anti-abuse rule—the limitation on benefits (LOB) provision. The latter is a detailed rule addressing a large number of treaty-shopping situations in which non- residents of the contracting states might establish an entity that would be resident in one of the states to reduce or eliminate taxation in the other state through the benefits of the DTC. These options to reduce the entitlement to benefits of the DTC are included in article 29 of the OECD MC.36 The LOB provision restricts the general scope of the DTC’s benefits to so-called qualified persons. Article 29, paragraph 2 defines qualified persons. Any person to which that paragraph applies is entitled to all benefits of the DTC. These include ‘certain non-profit organizations’. The Commentary explains that the DTC should include an agreed description of the relevant non-profit organizations found in each contracting state. These organizations must qualify as resident in one of the states. Sections 37 and 40 of the Commentary note that such non-profit organizations would generally correspond to those that do not pay tax in their state of residence and that are constituted and operated exclusively to fulfil certain social functions (e.g. charitable, scientific, artistic, cultural, or educational). According to the Commentary, the description of such entities would typically refer to the provisions of the domestic law of each country or to the domestic law factors that allow the identification of those entities. Thus, charities can benefit from the DTC and are not restricted by the LOB clause.

20.3.4 No Mutual Recognition Notwithstanding the Non-Discrimination Article Article 24 of the OECD MC provides for the elimination of tax discrimination in certain circumstances. Article 24, paragraph 3 states that a permanent establishment may not be less favourably taxed than resident enterprises. This article has a limited scope for charities. Section 1 of the Commentary on article 24 notes that all tax systems incorporate legitimate distinctions. The non-discrimination provisions seek to balance the need to prevent unjustified discrimination with the need to take account of these legitimate distinctions. According to section 47 of the Commentary on article 24(3), it does not oblige a state that gives special taxation privileges to non-profit institutions whose activities are performed for purposes of public benefit that are specific to that state, to extend the same privileges to permanent establishments of similar institutions of the 35 OECD, Preventing the Granting of Treaty Benefits in Inappropriate Circumstances, Action 6 (Paris: OECD Publishing, 2015), para. 22. 36 OECD MC, art. 29, paras 9 and 1–7, respectively.

Charities in Tax Conventions 343 other state whose activities are not exclusively for the first-mentioned state’s public benefit. This means, for example, that if income from an investment in real estate would be tax-exempt if such an investment were made by a resident charity, income from such investment may be taxed if it is being made by a foreign charity. This is a reflection of the public benefit rationale discussed in Section 20.2. Some DTCs, however, extend tax benefits to charities in the other contracting state on a reciprocal basis as is discussed in Section 20.5.4. Furthermore, section 2 of the Commentary on article 24 explains that the provision cannot be interpreted to require most-favoured-nation treatment. This means that if a country has concluded a bilateral or multilateral agreement which gives tax benefits to donors of charities in contracting state(s) party to that agreement, donors resident in a third state may not claim those benefits by reason of a similar non-discrimination provision in the DTC between the third state and one of the contracting states. The Commentary explains that as DTCs are based on the principle of reciprocity, a tax treatment granted by one contracting state under a bilateral or multilateral agreement to a resident of another contracting state party to that agreement by reason of the specific economic relationship between those states, may not be extended to a resident of a third state under the non-discrimination provision of the DTC with the third state.

20.3.5 Donors to Charities The OECD MC does not include a provision allowing a tax deduction for donations to charities in the other contracting state. The OECD observed that in Barbados the inclusion of a provision on charitable donations seems to be part of the tax treaty policy.37 In the tax treaty between Barbados and the Netherlands, contributions to an organization constituting a charitable organization under the income tax laws of the other jurisdiction are deductible under the same terms and conditions as those applicable to contributions to resident charitable organizations where the competent authority of the state of the donor agrees that the organization qualifies as a charitable organization for the purposes of granting a deduction under its income tax laws. A similar provision is included in the tax treaties Barbados has concluded with Mexico, the Seychelles, Mauritius, and Ghana.

20.3.6 Conclusion on the Position of Charities under Treaties Based on the OECD MC DTCs that are based on the OECD MC and that follow the Commentary will usually regard charities as residents, even if they are exempt from corporate income tax. 37 OECD, Taxation and Philanthropy, 113.

344 Sigrid Hemels As such exemption is not regarded a ‘special tax regime’ and as qualifying non-profit organizations are not hit by the LOB provisions, this will usually mean that charities that receive income from investments in another contracting state, such as interest or dividends, will be given a reduction of withholding tax if they meet the other requirements for such reduction included in the DTC. Most DTCs do not extend tax benefits for resident charities or their donors to charities in the other contracting state. The DTCs that extend benefits on a reciprocal basis cannot be invoked by (donors to) charities resident in third countries, as the non-discrimination provision is not a most-favoured-nation clause. For charities raising funds abroad or investing in real estate abroad, DTCs that are based on the OECD MC will usually not mitigate the tax disadvantages of not being a resident charity.

20.4 UN MC It does not seem to make much difference for charities whether a DTC is based on the OECD MC or the UN MC. As with the OECD MC, the UN MC does not include a reference to charities. The Commentary on the UN MC refers to non-profit organizations and charitable activities in a similar way as OECD MC and refers to the Commentary on the latter MC. As paragraph 6 of the Commentary on article 4 of the UN MC refers to paragraph 8.6 of the Commentary on the same article in the OECD MC, exempt charities can be regarded as residents under the UN MC. The Commentary on article 1 of the UN MC states that among other things, section 86 of the Commentary on article 1 of the OECD MC is also relevant for the purposes of the UN MC.38 This means that under the UN MC, regimes that principally apply to organizations that are established and maintained exclusively for religious, charitable, scientific, artistic, cultural, or educational purposes are also excluded from the definition of ‘special tax regimes’. In the same way as the OECD MC, the UN MC includes non-profit organizations in the definition of a qualified person to whom the LOB clause does not apply.39 In addition, the wording of the non-discrimination article (art. 24 UN MC) is similar. As the UN MC Commentary refers to, among others, sections 2 and 47 of the OECD MC, this non-discrimination provision does not give a right to the same beneficial treatment as resident charities. If a specific DTC gives such right, that right does not extend to charities in third countries (no most-favoured-nation treatment).

38 39

Commentary on art. 1 of the UN MC, s. 118. Ibid., art. 29(2)(e)(i).

Charities in Tax Conventions 345

20.5 US MC The US has its own MC. The provisions relevant for charities are similar to those in the OECD MC and the UN MC, but there are some differences which are discussed in the following sections.

20.5.1 Scope: Residency Article 4(2) of the US MC states that the term resident includes an organization that is established and maintained exclusively for religious, charitable, scientific, artistic, cultural, or educational purposes, notwithstanding that all or part of its income or gains may be exempt from tax under the domestic law.40 Until now, no technical explanation accompanying the 2016 US MC has been published. The technical explanation on article 4(2) of the 2006 US MC (that has the same wording) clarifies that charitable organizations will be regarded as residents of a contracting state regardless of whether they are generally liable to income tax in the state where they are established. This means that a tax-exempt organization under section 501(c) of the Internal Revenue Code, that is organized in the USA, is a US resident for all purposes of the DTC. Thus, in the US MC charities are explicitly included in the definition of resident in the DTC itself, whereas this only follows from the Commentary in the OECD MC and UN MC. On the other hand, the US MC is more restrictive as only organizations having ‘exclusively’ charitable activities qualify, where the Commentaries on the OECD MC and UN MC seem to allow for broader definitions. For example, in the Netherlands an organization is charitable if its activities are at least 90% charitable. Nevertheless, the 1992 DTC between the Netherlands and the USA is aligned with the US MC in its definition of exempt organizations in that it requires that the organization is operated ‘exclusively’ for religious, charitable scientific, educational, or public purposes. The list of activities in this DTC deviates from the current US MC in that ‘artistic’ and ‘cultural’ activities are not mentioned in the US–Netherlands DTC and that ‘public purposes’ are added in that DTC.

20.5.2 Special Tax Regimes Article 3(1)(l)(iv) of the 2016 US MC explicitly excludes from the definition of ‘special tax regimes’ regimes that principally apply to organizations that are established and maintained exclusively for religious, charitable, scientific, artistic, cultural, or

40

US MC, art. 4(2).

346 Sigrid Hemels educational purposes. Again, the difference with the OECD MC and UN MC is that this provision is included in the US MC instead of in a Commentary.

20.5.3 Limitation on Benefits Article 22(2)(e) of the US MC includes residents defined in article 4(2) of the US MC in the definition of qualified persons. Therefore, the LOB clause does not apply to those charities under the US MC. A difference with the OECD MC and the UN MC is that the US MC gives a specific definition whereas the other MCs refer to ‘certain non-profit organizations’ and leave the definition of the term to the contracting states.

20.5.4 Non-Discrimination and Reciprocal Exemption of Charities The technical explanation accompanying the 2006 US MC does not include the statement of the Commentary on the OECD MC that a contracting state is not obliged to give special taxation privileges to non-profit institutions in the other state. Given the similar wording of the non-discrimination provision, it is likely that the US MC should not be interpreted as giving such privileges. However, the USA has concluded DTCs with several countries, such as Canada, Germany, and Mexico that allow for a reciprocal tax exemption for charities that are exempt in the other contracting state. Article 21(1) of the 1980 US–Canada Income and Capital DTC provides that income derived by a religious, scientific, literary, educational, or charitable organization is exempt from tax in a contracting state if it is resident in the other state, but only to the extent that such income is exempt from tax in that other state. Article 27 of the 1989 US–Germany DTC, article 36 of the 1992 US–Netherlands DTC, and article 22(1) of the 1992 US–Mexico DTC include a similar provision but add as a requirement that the income will be exempt from tax in the source state if received by an organization recognized in the source state as exempt from tax as an organization with religious, scientific, literary, educational, or other charitable purposes. Where the US– Canada DTC provides for a mutual recognition of charities (source state must follow recognition by residence state), these DTCs apply a double test as both countries must— theoretically in case of the source state—regard the entity as an exempt charity. Furthermore, article 21(5) of the US–Canada DTC and article 22(4) of the US–Mexico DTC provide that a religious, scientific, literary, educational, or other charitable organization which is resident in Canada, or respectively Mexico, and which has received substantially all of its support from persons other than US citizens or residents, is exempt from US excise taxes imposed with respect to private foundations.

Charities in Tax Conventions 347

20.5.5 Donors The US model does not include a provision allowing a tax deduction for donations to charities in the other contracting state. In this respect, the US MC has the same effect as the OECD MC and the UN MC. However, the DTCs that the USA concluded with Canada, Israel, and Mexico allow, if certain requirements are met, the deduction of gifts to charities in the other country. Article 15A of the 1975 US–Israel DTC provides that contributions to a charitable organization in the other state are deductible in the state of the donor insofar as the contribution would also have been treated as a charitable contribution had the charity been resident in the state of the donor. The entity must, therefore, meet the requirements of both countries for the donor to be able to deduct their gift. The US technical explanation of this DTC mentions that in notes exchanged at the time of the signing of the treaty, it was agreed that an effort would be made to develop procedures to simplify the determination that an organization in the state is eligible to receive charitable contributions from residents of the other state. The competent authorities would review the procedures and requirements of the other state for certifying an organization of that other state as eligible to receive charitable contributions. If the competent authorities found that the procedures did not substantially differ, it was contemplated that the authorities of each state would accept certification by the other and would not require recipient organizations to qualify separately in both states. This would mean a type of mutual recognition. However, the deduction does not apply to contributions to a charity in the other state in excess of 25% of taxable income from sources in the state where the charity is resident. This means that if the donor does not have income from the state where the charity is resident, they cannot claim any gift deduction for charitable contributions to a charity in that other state. Similarly, articles 21(6) and 21(7) of the US–Canada DTC provide that contributions by a citizen of one of the contracting states to an exempt entity in the other contracting state that could qualify to receive deductible donations (USA)/as a registered charity (Canada) had it been resident (USA)/established or created (Canada) in the other contracting state will be treated as charitable contributions (USA)/gifts to a registered charity (Canada). Again, the deduction is limited to a percentage of the income the donor has from sources in the resident state of the charity. However, this does not apply to charitable contributions/gifts to a college or university at which the donor or a member of their family is or was enrolled: in such event, the donations are deductible even if the donor does not receive income from the charity’s resident state. The US and Canadian competent authorities exchanged notes agreeing that if a contracting state determined that the other state maintains procedures to determine charitable status and rules for qualification that are compatible with such procedures and rules of the first state, such state will accept the charity certification of the other state.

348 Sigrid Hemels Article 22(2) and (3) of the US–Mexico DTC is formulated more conditionally: only if the two states agree that the law of the other state provides standards for organizations authorized to receive deductible contributions that are essentially equivalent to the standards for charities of the state in which the donor is resident, are contributions to a charity in the other state treated as charitable contributions in the state of the donor. Again, the limitation to a percentage of the income from the resident state of the charity applies. The exception for colleges and universities is not included in this DTC.

20.6 OECD Model Convention on Taxes on Estates, Inheritances, and Gifts The 1982 OECD EIGMC tries to solve the problem of double estate, inheritance, and gift taxation. Charities may be exempt from these taxes in their country of residence. If they receive donations or legacies from abroad, they might be confronted with foreign taxation as, based on the public policy rationale, the country of the donor or deceased does not apply the domestic exemption to foreign charities. Unfortunately, the OECD EIGMC does not provide a solution for these problems. In fact, it does not include any reference to charitable organizations.

20.6.1 Commentary on the OECD EIGMC The Commentary on the OECD EIGMC refers to charitable institutions, but as Deblauwe et al. rightly commented: Although the Commentary touches upon foundations and trusts in general as special features of the domestic law of certain member countries, charitable foundations and trusts are not mentioned at all in this respect. In the Commentary on Article 4 of the OECD Model Convention (concerning fiscal domicile) and article 10 (concerning non-discrimination) hardly any attention is given to charitable institutions and non-profit-making institutions whose activities are performed for purposes of public benefit.41

Paragraphs 15 and 32 of the Commentary on article 4 on fiscal domicile only observe that for charitable institutions the criterion of (place of effective) ‘management’ may often not apply. According to paragraph 32, that term refers more to business enterprises. One wonders whether this is the case, as charities, as with any other organization, need

41 R.

Deblauwe et al., ‘Gift and Inheritance Tax with Regards to Charities’, in F. Vanistendael, ed., Taxation of Charities, (Amsterdam: IBFD 2015), 92–3.

Charities in Tax Conventions 349 some type of management. This is probably not the most urgent problem that charities encounter. Nevertheless, for the purposes of article 4(1), establishing the domicile of a charity, the Commentary suggests in paragraph 15 using some similar criterion such as the ‘statutory seat’ or the ‘establishment under the law of a certain State’. For the purposes of article 4(3) on double domicile, paragraph 32 of the Commentary suggests that in ‘the very rare cases where a charitable institution is deemed to be domiciled in both states’ the two states should rely on the mutual agreement procedure provided for in article 11 of the OECD EIGMC. The article that could have been relevant for charities receiving donations and inheritances from abroad—the non-discrimination provision of article 10 OECD EIGMC—does not provide a solution. One might be inclined to interpret that article as obliging contracting states to treat charities in the other state in the same way as resident charities. However, paragraph 7 of the Commentary on that article notes that this provision does not oblige a state to extend benefits to charities in the other state. Paragraph 9 argues that if a state gives privileges to certain private non-profit-making institutions, this is clearly justified by the very nature of these institutions’ activities and by the benefit which that state and its nationals will derive from those activities. This observation indicates that the drafters of the Commentary relied on the public policy rationale for tax incentives for charities. Based on that rationale, the Commentary explicitly allows states to differentiate between domestic and foreign charities. Paragraph 10 of the Commentary mentions that states are free to adopt another approach, especially with regard to taxes on estates, inheritances, and gifts. The Commentary observes that circumstances may prompt states to treat the charitable institutions of both countries in the same way. It is suggested that similarity of treatment can be effected by including the following wording in article 10: Organisations of a Contracting State shall be entitled in the other Contracting State to any exemptions from, or reductions of tax accorded to, organisations of the same type of that other State. The type of the organisations shall be determined according to the domestic law of the State imposing the tax.

This solution seems to be very similar to the approach the CJEU has taken and that was briefly mentioned in Section 20.2: the state imposing the tax (and providing a tax incentive) may impose the requirements to qualify for an incentive and is not obliged to go along with the qualification of the charity’s resident state.

20.6.2 Examples of Gift and Inheritance Treaties Referring to Charities As discussed in Section 20.1, there are relatively few gift and inheritance treaties. The treaties that include a provision on charities are even fewer in number, but several such treaties can be found in the IBFD tax treaty database. I will discuss some examples.

350 Sigrid Hemels Article 17(2) of the 1990 France–Italy Inheritance and Gift Tax Treaty includes a provision that is similar to the alternative wording of article 10 of the OECD EIGMC. France and Switzerland concluded a specific agreement in 1979 on the tax treatment of gifts made for non-profit-making purposes. It provides that gifts and inheritances transferred or bequeathed to wholly non- profit- making organizations which are engaged in activities, inter alia, in the scientific, artistic, cultural, or charitable fields are exempt, provided that the exemption is granted to organizations of the same nature established or organized in the other state. Article 11(2) of the 1994 France–Sweden Inheritance and Gift Tax Treaty also includes a provision on charities: Public institutions and non-profit organisations, whatever their denomination, which are founded or organised in a Contracting State and pursue religious, scientific, artistic, cultural, educational, or charitable activities, shall benefit in the other Contracting State, subject to the conditions provided by the laws of that other State, from the exemptions, reductions or other tax advantages granted to similar entities founded or organised in that other State.

The 1994 France–Portugal Inheritance and Gift Tax Treaty uses similar wording with the difference that ‘religious’ is not included in the enumeration of charitable activities. Article 28(2) of the 1992 Income, Capital, Inheritance and Gift Tax Treaty between Germany and Sweden uses slightly different wording as it is divided into one part on Sweden and another part on Germany. Except from the country names, however, these parts are the same. The treaty requires that the organization must exclusively pursue the purposes mentioned and that the organization should also have been exempt had it been resident in the other state. In this respect, the provision resembles that of the US model (see Section 20.7). Article 10(2) of the 1993 France–Austria Inheritance and Gift Tax Treaty is stricter than the treaties mentioned previously, as the application of the tax advantages requires a case-by-case agreement with the competent authorities. Article 10(2) of the 2001 Austria–Netherlands Estate, Inheritance and Gift Tax Treaty again has different wording: Tax exemptions and reductions granted under the laws of a Contracting State in respect of taxable events to which this Convention applies for the benefit of organisations of that Contracting State, whose activities are exclusively carried on for religious, philosophical, charitable, cultural or scientific purposes, or for the common good, shall also apply to similar organisations of the other Contracting State.

This provision includes an ‘exclusivity’ requirement and explicitly includes ‘philosophical’ purposes in the enumeration. Article 10 of the 1964 Italy–Greece Inheritance Tax Treaty deviates:

Charities in Tax Conventions 351 Legal persons having a religious, social, or charitable purpose legally established or to be established in either of aforesaid States, shall be exempt from any tax on gifts and successions irrespective of the place where the asset inherited or donated is situated or of the elements of which it is composed. The same rule shall apply to the Contracting States and local authorities thereof, provided that the gifts or successions concerned are used for religious, social, charitable, or cultural purposes.

Article VI of the 1968 Italy–Israel Estate and Inheritance Tax Treaty is very specific. It requires approval by the resident state of the charitable status and includes mutual recognition. If the resident state deems the organization to be charitable, it is recognized as such by the other state. The other state cannot impose its own requirements. The enumeration specifically includes welfare purposes. It may be that in other treaties welfare purposes are included under ‘charitable’ purposes. Article IX of the 1962 Israel–Sweden Death Duties Treaty includes ‘social welfare’ and it is sufficient if the organization is established ‘mainly’ for such purposes. From these examples, it is clear that states do not seem to use standard wording for this provision, but negotiate it on a case-by-case basis. Even France, which has concluded a relatively large number of gift and inheritance tax treaties which include a provision on charities, does not seem to use one standard form. The enumeration of purposes differs slightly between treaties.

20.7 US Estate and Gift Model Convention Unlike the OECD EIGMC, the 1980 US EGMC includes a provision on charities. Article 8(3) provides that transfers of property for religious, charitable, scientific, literary, or educational purposes that are tax-exempt in one contracting state will be exempt from tax in the other contracting state, provided the transfer would be tax-exempt if made to a similar entity in that other state. The accompanying technical explanation observes that US domestic tax legislation would not allow a deduction for the transfer of property by a non-resident alien to a foreign charity. Under article 8(3), such deduction is allowed. The provision requires that the transfer would be exempt in both states, meaning that the requirements of both states must be met. This provision can be found in article 9 of the 1983 US–Denmark Inheritance and Gift Tax Treaty and article 8 of the 1983 US–Sweden Inheritance and Gift Tax Treaty. In the latter treaty, ‘literary’ is not included in the enumeration of purposes. Article 10 of the 1980 US–Germany Inheritance and Gift Tax Treaty also includes that provision but adds that the competent authorities of the contracting states shall settle the application of the provision by mutual agreement. Furthermore, instead of ‘literary’, ‘public’ purposes are

352 Sigrid Hemels included in the enumeration. Article 10 of the 1978 US–France Inheritance and Gift Tax Treaty also deviates from the US EGMC as it includes an additional requirement in article 10(2)(c): the charity must receive a substantial part of its support from contributions from the public or governmental funds. This means that charities that mainly derive funds from one individual or company would not qualify.

20.8 Conclusion In many cases, charities have access to DTCs, but this does not always solve the problem of single taxation where no taxation was envisaged by the resident state. Mutual recognition of charities is rare. The public policy rationale for tax incentives for charities is influential in DTCs. One might wonder whether the OECD and its member states should follow the example set by the CJEU and reconsider the validity of this rationale. In 1969, Van Hoorn had already criticized the restriction of benefits within the home country. He observed: in many circles there is a feeling that the restricted application of tax concessions to domestic institutions constitutes an obstacle to furthering of interests which in the present world are no longer nationally bound . . . By itself a nationally bound application of tax concessions is to be rejected in an era in which international integration is ever progressing.42

Van Hoorn was of the opinion that the criterion of national interest had lost every significance, pointing out that in certain tax treaties benefits are extended to institutions in the other country without there being any requirement other than that such institutions must further the general interest.43 With globalization, the feeling described by Van Hoorn has probably become even stronger. In 2007, Koele deemed it ‘completely archaic to deny that there is a significant number of non-profit institutions that pursue supranational objectives which are not limited to the nation in which they are organized’.44 However, at present the public policy rationale still seems to be deeply rooted, and it is not likely that in the near future DTCs will change in that respect.

42

Van Hoorn, ‘General Report’, paras 6, 57, 76. Ibid., para. 64. 44 I. A. Koele, International Taxation of Philanthropy (Amsterdam: IBFD 2007), 8. 43

Chapter 21

Excha nge of Informat i on and Tax Tre at i e s Xavier Oberson

21.1 Introduction The date 13 March 2009 has been described as the ‘big bang’ of exchange of information (EOI).1 Following that date, many double-taxation treaties (DTTs) around the world were adapted or concluded with the purpose of implementing the OECD standard. At the same time, tax information exchange agreements (TIEAs) were also widely signed. As of 2014, this development led to the adoption of Automatic Exchange of Information (AEOI) as the new OECD standard. In this chapter, after a general description of the international existing rules of EOI (Section 21.2), we will focus on the main issues raised in the context of EOI under DTTs (Section 21.3).

21.2 International Legal Instruments of EOI Historically, the main international legal instruments of EOI in tax matters are DTTs. Most DTTs around the world implement the rule of article 26 of the OECD or the UN Model. This provision reflects the so-called ‘standard on transparency and effective EOI or tax purposes’.2

1 X. Oberson, International Exchange of Information in Tax Treaties, 2nd ed. (Cheltenham: Edward Elgar, 2018), 7. 2 T. Falcão and A. Lara Yaffar, ‘General Report: Exchange of Information: Issues, Use and Collaboration’, Cahiers de Droit Fiscal International vol. 105B (2020), 206.

354 Xavier Oberson TIEAs represent another source of EOI. They were adopted around the 1990s, in order to obtain information from countries that did not usually cooperate in tax matters, namely tax havens. In 2002, the OECD published a bilateral and a multilateral Model TIEA. At the outset, only a small number of jurisdictions entered into these agreements. After the ‘big bang’ of 2009, the situation changed dramatically.3 Between the end of 2008 and 2018, more than 1,000 TIEAs were signed. In June 2015, a Protocol to the TIEA was adopted by the OECD in order to also include spontaneous and automatic EOI. In parallel, the number of states ratifying the multilateral OECD/Council of Europe Convention on Mutual Administrative Assistance in Tax Matters (CMAA) has continued to grow. The Multilateral Competent Authority Agreement (MCAA) was introduced in 1988 and amended by a Protocol of 2010 in order to open it to non-OECD members and implement the new standard of EOI. The CMAA is the most comprehensive legal instrument for EOI. It covers both EOI upon request, spontaneous and automatic. As of 2020, more than 136 jurisdictions have signed it. The possibilities offered by the MCAA do not limit or are not limited by those contained in existing or future international agreements between the parties. In order to facilitate the implementation of the US Foreign Account Tax Compliance Act (FATCA), as a unilateral tax enforcement measure with extraterritorial effects, the USA introduced intergovernmental agreements (IGAs). There are basically two model IGAs. Under Model 1, adopted by most countries, foreign financial institutions (FFI) have to report US account holders directly to their local tax authorities, which then pass on the information to the Internal Revenue Service (IRS), under a DTT or an IGA or another convention. By contrast, a Model 2 IGA requires a direct agreement between the FFI and the IRS. This system was adopted notably by Switzerland and Japan. With the global acceptance of AEOI as the new standard, it started to be implemented globally, as of 2014. The OECD ‘common reporting standard of financial accounts information in tax matters’, published in July 2014, sets the minimal standard of due diligence. The implementation of AEOI requires three elements. First, an international legal basis is necessary (typically, MCAA, or a multilateral or bilateral treaty). Secondly, an agreement between competent authorities (CA) is required in order to implement the AEOI. For that purpose, the OECD has introduced a multilateral and a bilateral Model CA agreement. Thirdly, a domestic legal basis has to be adopted in order to introduce the system of common reporting standards (CRS) in each state. The first AEOI took place in September 2017, within the ‘early adopters’ jurisdictions. Since then, the network of participating jurisdictions continues to grow globally, with the main exception of the USA, which did not adopt the CRS but relies on the FATCA Model, which is criticized because it does not offer the same level of reciprocity among states.4 There are other international instruments providing for exchange of information, notably as a consequence of the adoption of the G20/OECD Base Erosion and Profit

3 Oberson, International Exchange of Information, 63. 4

Ibid., 195.

Exchange of Information and Tax Treaties 355 Shifting (BEPS) Project. An automatic exchange of country-by-country reporting, as a result of BEPS Action 13 has been introduced, following the OECD bilateral and multilateral Model CA agreements. In addition, following BEPS Action 5, taxpayers have to report tax rulings, which are defined under six categories. Thirdly, BEPS Action 12 also recommends the implementation of reporting by taxpayers and advisors of aggressive tax-planning arrangements. In the EU, exchange of information exists under various directives. The more comprehensive system is provided for by the Directive on administrative cooperation in the field of taxation (DAC), adopted on 15 February 2011.5 The DAC has been modified many times, notably in order to introduce: AEOI on financial information (DAC 2), exchange of information on rulings (DAC 3), country-by-country reporting (DAC 4), exchange of anti-money laundering information (DAC 5), AEOI on reportable cross- border arrangements (DAC 6), and AEOI on digital platform operators (DAC 7). The chapter will then focus on the legal issues that EOI DTTs may raise, based notably on the existing and growing case law in this context, in particular following the case law of the Swiss courts.6 The status and progress of implementation of the EOI upon request among various jurisdiction may also be consulted in the Global Forum’s peer review reports.7

21.3 Scope of EOI under DTT 21.3.1 Material Scope of EOI 21.3.1.1 ‘Foreseeable relevance’ of the request 21.3.1.1.1 In general Article 26 paragraph 1 of the OECD Model provides that the competent authorities of the contracting states shall exchange such information as is foreseeably relevant for carrying out the provisions of this convention or for the administration or enforcement of the domestic laws concerning taxes of every kind and description imposed on behalf of the contracting states, or of their political subdivisions or local authorities, insofar as the taxation thereunder is not contrary to the convention. The standard of ‘foreseeable relevance’ was introduced in 2005 in article 26 paragraph 1 (first sentence) and replaced the wording ‘as is necessary’ in the old version. This new wording shows a willingness to implement a rather broad exchange of information. Indeed, according to the OECD, the standard of ‘foreseeable relevance’ is intended to

5

[2011] OJ L64/1. to Falcão and Lara Yaffar, ‘General Report: Exchange of Information: Issues, Use and Collaboration’, 207, Switzerland is one of the very few countries with such an extensive case law in EOI. 7 See ibid. 6 According

356 Xavier Oberson provide for exchange of information in tax matters to the widest possible extent and, at the same time, to clarify that contracting states are not at liberty to engage in ‘fishing expeditions’ or to request information that is unlikely to be relevant to the tax affairs of a given taxpayer.8 In a leading case, the Swiss Federal Supreme Court has ruled that this condition is considered to be met, to the extent that, at the time of the request of exchange of information, there is a ‘reasonable possibility’ that such information would appear to be relevant. It follows that the appreciation of the standard of foreseeably relevance is initially a competence of the requesting state.9

21.3.1.1.2 Tax status of the taxpayer or of the procedure Various requests pertain to the issue of defining the effective place of residence of individuals (potential dual residents). In particular, the Swiss Supreme Court confirmed that the French tax authorities could request information on two individuals, residents of Switzerland, to the extent that the requesting state had elements that could also trigger tax residence in France. Indeed, such information is foreseeably relevant because, notably, it may assist the application of the ‘tiebreaker’ rules in article 4 of the treaty.10 In a further case, the Supreme Court, with similar reasoning, also admitted a request for information from the French tax authorities about apparent Swiss residents, also potentially resident in France, for whom third states, namely Spain and then Luxembourg, also claimed potential resident status.11 In other cases, the Supreme Court has confirmed that requests about the status of Swiss residents taxed under the regime of taxation on expenses on a lump-sum basis (so-called forfait fiscal) are foreseeably relevant.12 The Supreme Court has ruled that requests about so-called status updates, that is information about the state of the appeal process of the taxpayer, are relevant.13 In particular, the requesting state does not have to check whether the domestic procedural rules of the requesting state have been correctly applied, such as the possibility of a revision process,14 or the impact of a statute of limitation.15

21.3.1.1.3 Transfer pricing In practice, requests as to the application of transfer pricing rules have also been accepted. For example, a request aiming to verify the economic substance of a Swiss resident company, controlled by a French resident (unique shareholder), asking for information about its activities, premises, and number of employees, was foreseeably 8

OECD Commentary on art. 26, para. 5. Federal Supreme Court (FSC), 24 September 2015, ATF 142 II 161, 165. 10 FSC, 24 September 2015, ATF 142 II 161. 11 FSC, 5 April 2016, ATF 142 II 218. 12 FSC, 1 February 2019, ATF 145 II 112; FSC, 7 June 2019, 2C_764/2018; FSC, 22 July 2019, 2C_1053/ 2018. 13 FSC, 3 November 2017, ATF 144 II 130. 14 FSC, 7 April 2017, 2C_241/2016, paras 5.4 and 5.6. 15 FSC, 4 October 2017, 2C_1162/2016, para. 6.3 s. 9

Exchange of Information and Tax Treaties 357 relevant.16 In the same vein, names and address of employees and clients may be foreseeably relevant to evaluate the applicable transfer pricing rules for transactions between closely related Swiss and French companies.17 The Supreme Court also admitted the transfer of various balance-sheet and tax allocation rules of Swiss companies and permanent establishments associated with a French group.18

21.3.1.2 Identification of the taxpayer 21.3.1.2.1 In general In practice, it will sometimes be difficult to identify properly the taxpayers subject to the request. The tax authorities of the requesting state do not necessarily know their names, but possess other means of identification, such as account numbers of banks or credit cards, other electronic identification data, or other elements that may identify or link them with assets in the requested state. Such request remains foreseeably relevant, to the extent that the requesting state provides sufficient information to identify the taxpayer.19 Eventually, this issue has evolved. This question has, indeed, been raised not only in the case of individual taxpayers but has also started to include groups or even list of taxpayers (so-called collective requests).

21.3.1.2.2 Group requests A first example of a group request can be found in the UBS case of 5 March 2009.20 At that time, the ‘old’ treaty of 1996 was applicable between the USA and Switzerland and the exchange of information required a ‘tax fraud and the like’ on the side of the concerned taxpayers. In this important case, which is the first of a series of cases pertaining to the UBS saga in the USA,21 the IRS was asking for the names of about 250 US taxpayers. In particular, the IRS could identify a repetitive and systematic ‘pattern of behaviour’ by the unknown taxpayers which, by the interposition of offshores entities (British Virgin Islands or Liechtenstein foundations), aimed at circumventing the qualified intermediary regime, with collaboration from the bank. The Swiss Federal Administrative Court, which was the only competent court at the time, admitted the request. Later, as of July 2012, the OECD Commentary on article 26 OECD Model DTT would be adapted and, in a way, confirm the admissibility of such group requests. Seven years later, in a leading case,22 the Swiss Supreme Court was confronted with a new group request stemming from the Dutch tax administration, against various unknown Dutch individuals who were holding accounts with a Swiss bank. and the Dutch

16

FSC, 1 March 2016, ATF 142 II 69, para. 3.2. FSC, 15 March 2016, 2C_690/2015, paras 3.3, 3.5, and 4.4. 18 FSC, 13 February 2017, ATF 143 II 185. 19 OECD Commentary on art. 26, para 5.1 . 20 FAC, 5 March 2009, A-7342/2008 and A-7426/2008. 21 For a description of the whole UBS saga in the USA, see Oberson, International Exchange of Information, 48 ff. 22 FSC, 12 September 2016, ATF 143 II 136. 17

358 Xavier Oberson tax administration did not respond to a request from the bank requesting confirmation of their tax-compliant status. Despite the absence of the names of the taxpayers in the request—as the wording of the treaty required—the court found that it was not a fishing expedition because the request was sufficiently limited. The court, in accordance with the OECD Commentary on article 26, in the version of 12 July 2012,23 ruled that, in order to be admissible, a group request had to comply with three cumulative conditions: (1) it must provide for a detailed description of the group, and the facts and circumstances motivating the request; (2) it should describe the relevant applicable tax law and the reason behind the suggestion that the taxpayers belonging to the group have not complied with their tax obligations; and (3) it should demonstrate that the information requested is adequate to achieve compliance with those obligations. In this case, the three conditions were met, and the request was therefore allowed.

21.3.1.2.3 Collectives (or bulk) requests The same year, the Supreme Court had to deal with a request from Norway, based on nine credit cards numbers of unknown individuals.24 Retrospectively, this interesting case may be seen as the first collective request. In this case, the court did not characterize the request as a group request—since the number of taxpayers was clearly determined and their identification occurred through a credit card number—but rather as a ‘collective request’. In any event, in order to allow the request, the Supreme Court developed an argumentation similar to the group request analysis, considering the difficulty of the distinction between group and collective requests.25 Another example of a bulk request involved a request from Germany in which taxpayers resident there had been identified by numbered accounts, discovered during a search in the Swiss subsidiary of a German bank.26 In a landmark case of 26 July 2019, which involved a French request targeting about 40,000 taxpayers, holding directly or indirectly bank accounts at UBS Switzerland,27 the Swiss Supreme Court was again confronted with the interpretation of the concept of a collective request. Such request was based on article 28 of the French–Swiss treaty and on an agreement of 25 June 2014, which entered into force on 30 March 2016 (hereafter the 2014 agreement), amending the Protocol to that treaty. The request relied on

23

See OECD Commentary on art. 26, para. 5.2. . FSC, 1 September 2017, ATF 143 II 628. 25 See L. Papadopoulos, ‘Echange de renseignements: vers une admission massive des demandes groupées et collectives?’, Archives de droit fiscal 88 (2019/20), 287ff. 26 FSC, 29 October 2018, 2C_695/2017. 27 FSC, 26 July 2019, ATF 146 II 150; for a critical analysis of this case, see A. Opel, ‘Neuausrichtung der schweizerischen Abkommens- politik in Steuersachen: Amtshilfe nach dem OECD- Standard’, in M. Zweifel, M. Beusch, and S. Oesterhelt, eds, Kommentar zum Schweizerischen Steuerrecht, Internationales Steuerrecht, Amtshilfe (Basle: Helbing Lichtenhahn, 2020), § 10 n. 125; L. Papadopoulos, ‘Switzerland: Article 26 of the OECD Model and the Broad Interpretation of “Foreseeable Relevance” ’, in E. Kemmeren et al., eds, Tax Treaty Case Law around the Globe 2020 (Munich: C. H. Beck, 2021), s. 4; X. Oberson, Droit fiscal Suisse, 5th ed. (Basle: Helbing Lichtenhahn, 2021), 379 ss. 24

Exchange of Information and Tax Treaties 359 two attached lists, namely list ‘B’ (year 2006) and list ‘C’ (year 2008), which the French authorities had obtained from Germany. At a public hearing, the Supreme Court allowed the transfer of information to France.28 In a nutshell, the Supreme Court concluded that despite the fact that it was targeting about 40,000 French taxpayers, the request was an admissible collective request and not a group request. To reach that conclusion, the court argued as follows. First, the Supreme Court recognized that there is a ‘fine line’ of distinction between a group and a collective request. In essence, as described previously, a group request is based on a ‘pattern of behaviour’ (‘modèle de comportement’; ‘Verhaltens-muster’).29 By contrast, a collective request is an addition of individual requests,30 in which the name of the taxpayers involved are unknown, but whose identification is possible by other means (bank account numbers, digital references, credit card numbers, etc.). Since it is a collective request, the temporal limits of article XI paragraph 3 lit. a of the 2014 agreement are not applicable. Indeed, this provision only applies to group requests, but does not cover collective requests based on a list. As a consequence, the French collective request may still apply to facts existing as at 2010. Secondly, the issue was whether this collective request was an inadmissible ‘fishing expedition’.31 Surprisingly, in order to allow such a request, the court referred to the criteria of group requests and concluded that the three cumulative conditions for the admissibility of such requests were met.32 In order to justify the request, and notably to determine whether the (unknown) taxpayers on the list were arguably non-compliant, the Supreme Court referred to a statistical analysis based on the so-called list A, in which the names of the taxpayers were known, but was otherwise similar to lists ‘B’ and ‘C’, in which the taxpayers were unknown. Finally, in accordance with the condition of specialty, the last issue was whether this information could also be used for criminal purposes against the taxpayers involved, respectively against the bank itself.33 In this instance, article 28 paragraph 2 of the French– Swiss treaty did not allow the use of information exchanged for other purposes, such as criminal procedures or illicit banking solicitations. Since the Swiss federal tax administration had obtained a guarantee from the French competent authorities that the information exchanged would not be used for criminal purposes, the Supreme Court had no reason to doubt that the agreement would not be respected.

21.3.1.3 Transfer of information about third parties Requests for information not only target directly involved taxpayers, but also third parties, whose link with the procedure requires a further analysis. In practice, the foreseeable relevance of requests concerning third parties necessitates a specific additional

28

FSC, 26 July 2019, 2C_653/2018, ATF 146 II 150. FSC, 26 July 2016, 2C_653/2018, ATF 146 II 150, para. 4.4. 30 In this sense, FSC, 26 July 2016, 2C_653/2018, para. 4.3. 31 See also art. IX para. 2 of the Protocol to the French–Swiss DTT. 32 For the conditions of admissibility of group requests, see Section 21.3.1.2. 33 See Section 21.4. 29

360 Xavier Oberson link with the relevant case. In a nutshell, a distinction should be drawn between third parties concerned and not concerned by the request. According to the principle of proportionality, the requested state should proceed to a balance of interests, and delete information relating to third parties (e.g. names of bank employees), who are not concerned and whose names are not relevant for the case.34 In some instances, in particular the communication of names of bank employees, has been regarded as irrelevant.35 This issue has been especially delicate in the framework of the so-called Department of Justice (DOJ) Program with the Swiss Banks, based on a 2013 Joint Statement of the US DOJ and the Swiss Department of Finance.36 The Swiss Supreme Court was confronted with this specific issue, among others, in one case in the context of a request for a non-prosecution agreement.37 In this particular case, the court admitted that the names of a lawyer/public notary and of bank employees mentioned in the documentation were not foreseeable relevant and should therefore be suppressed (‘caviardés’). The third parties whose names have been forwarded do not lose any protection. Indeed, the use of information on third parties obtained by the requesting state may also, under some circumstances, be restricted by the principle of specialty.38

21.3.2 Personal Scope of EOI Since 1977, EOI under article 26 is not restricted by article 1 of the OECD Model. It follows that an exchange of information is possible not only for residents in the contracting states, but also about residents of third states or individuals or entities that are not subject to taxation according to article 4 OECD Model.39

21.3.3 Taxes Covered Since 2000, the scope of article 26 is not limited by article 2 of the OECD Model. Exchange of information may thus cover taxes of ‘any kind and description’ and not only taxes under the scope of the Model DTT, such as income (profit) or capital (wealth) taxes.

34

FSC, 24 September 2015, ATF 142 II 161, para. 4.6.1. FSC, 24 September 2015, ATF 142 II 161, 180. 36 On this Program, see Oberson, International Exchange of Information, 52, 217. 37 FSC, 18 December 2017, ATF 144 II 29. 38 See Section 21.4. 39 Oberson, International Exchange of Information, 26; M. Engelschalk, ‘Commentary to Art. 26 OECD MC’, in K. Vogel and M. Lehner, eds, Doppelbesteuerungs-abkommen Kommentar, 5th ed. (Munich: C. H. Beck, 2008), n. 57 on art. 26. 35

Exchange of Information and Tax Treaties 361

21.4 Competent Authorities, Confidentiality, and Specialty Principle Exchange of information occurs between competent authorities. In most cases, it will be the Ministry of Finance, which in turn delegates some of its powers to the tax administration. The competent authorities must keep information received within the exchange of information process as confidential. The so-called secrecy clause of article 26 paragraph 2 OECD Model DTT—which was moved from paragraph 1 to paragraph 2 in 2005—has a role in ensuring that both administrations ‘will treat with proper confidence the information which it will receive in the course of their cooperation’.40 The obligation of secrecy has a double component.41 First, the requesting state is obliged to keep the information received secret under the same conditions as information obtained under domestic law. Secondly, the provision will describe exactly to whom the information obtained may be transferred. The exact content of the secrecy requirement has evolved over time.42 The maintenance of secrecy in the receiving state, as well as sanctions for violation of the rule, are a matter of domestic law.43 In a recent report,44 the OECD suggested some basic principles that states should apply under the secrecy rule, in order to ensure a minimum common threshold of protection. It follows that information received may be disclosed only to the persons or authorities listed in provisions similar to article 26 paragraph 2 OECD Model DTT. The possibility of informing the courts was only added in 1977. In 2005, a further change of the text of this provision included persons concerned with the oversight of the tax administration and the courts. The information may also be communicated to the taxpayer, their proxy, or to witnesses. In addition, such persons or authorities must use the information only for the purposes mentioned in paragraph 2. They may, however, disclose the information in public court proceedings or in judicial decisions.45 This rule corresponds to the principle of specialty, also characterized as the ‘purpose limitation principle’, which provides, in

40

OECD Commentary, on art. 26, n. 11 . P. Pistone and M. Gruber, ‘Die Möglichkeiten der Verweigerung des Informationsaustausches nach Art. 26 OECD-MA’, in M. Lang, J. Schuch, and C. Staringer, eds, Internationale Amtshilfe in Steuersachen, Vienna (Vienna: Linde, 2011), 92. 42 See Oberson, International Exchange of Information, 27. 43 OECD Commentary on art. 26, n. 11. 44 OECD, ‘Keeping it Safe— the OECD Guide to the Protection of Confidentiality of Information Exchanged for Tax Purposes’ (July 2012). 45 OECD Model DTC, art. 26 para. 2 in fine. 41

362 Xavier Oberson general, that information received may only be used for the tax purposes requested.46 The information may, however, be disclosed in public court proceeding or in judicial decisions.47 According to the Swiss Supreme Court, the specialty principle protects, on the one hand, the right of the person involved to a fair process and, on the other hand, the sovereignty of the requested state.48 In principle, information received by a contracting state may not be disclosed to a third country, unless a specific provision of a DTT allows such disclosure. On 17 July 2012, article 26 paragraph 2 OECD Model DTT was, however, completed by an additional sentence, which provides that ‘information received by a Contracting State may be used for other purposes when such information may be used for such other purposes under the laws of both States and the competent authority of the supplying State authorizes such use’. In this respect, the Supreme Court confirmed that, according to article 26 paragraph 2 of the OECD Model DTT, the information may only be transmitted to the persons or authorities listed and only for tax matters (including criminal tax purposes). The information may not be used for other purposes, notably for criminal purposes.49 A special provision of a DTT could, however, allow the use of such information for other purposes. In addition, the 2005 change provides for a new article 26 paragraph 4, under which the requested state cannot refuse to respond to an information request solely because it has no domestic interest in such information. The purpose of the clause, which appears as a clarification of an existing principle, is to prohibit the use of the so-called national interest clause, under which certain states refuse to exchange information.

21.5 Forms of EOI In general, article 26 paragraph 1 allows information to be exchanged in three different ways (on request, spontaneously, or automatically). These forms may be combined. In addition, the provision does not restrict the possibilities for exchanging information and other techniques, such as simultaneous examination, tax examination abroad, or joint audits.50 The OECD further published a Manual on the Implementation of Exchange of Information for Tax Purposes in 2006. Information upon request is the most frequent method applied so far. The requesting state requests from the other contracting state information foreseeably relevant in a 46 Opel,

‘Neuausrichtung der schweizerischen Abkommens- politik in Steuersachen’, 453; A.- P. Dourado and J. Becker, Klaus Vogel on Double Taxation Conventions, 5th ed. (Kluwer Law International, 2022), n. 266 on art. 26, OECD-MC. 47 OECD Model DTC, art. 26 para. 2 in fine. 48 FSC, 13 July 2020, 2C_537/2019, para. 3.4.1. 49 FSC, 18 December 2017, ATF 144 II 29. 50 OECD Commentary on art. 26, n. 9; see also OECD, Tax Information Exchange between OECD Member Countries: A Survey of Current Practices (Paris: OECD Publishing, 1994).

Exchange of Information and Tax Treaties 363 specific situation. Spontaneous information exchange is the provision of information to another contracting state that is foreseeably relevant to that party but has not been previously requested.51 AEOI means transmission of information, on a routine basis, at regular intervals, without any specific request from another state.52 Under article 26 paragraph 1, this type of transfer could occur, for example, when information about various categories of income having their source in one contracting state (typically dividends, interest, or royalties) is systematically transmitted to the receiving contracting state.53 This system of exchange typically covers recurring financial transactions, involving paying agents, such as interest, dividends, capital gains, or royalties. Tax examination abroad allows for the possibility of obtaining information through the presence of a representative of the requesting state in the requested state.54 For instance, to the extent allowed under domestic law, a contracting state may allow a representative of the other contracting state to enter its territory and to interview individuals or examine records, or be present during interviews or examinations carried out by the authorities of the first-mentioned state.55 This type of assistance is possible provided: (1) both contracting states agree on the interpretation of the treaty; and (2) reciprocity requirements are fulfilled.56 Under a simultaneous examination, two or more contracting states examine simultaneously, each in their own territory, the tax affairs of a taxpayer in which they have a common interest or related interest, with a view of exchanging any relevant information which they so obtain.57 A joint audit is an arrangement between two or more states joining together to form a single audit team of tax auditors from each state to examine an issue or transaction of an individual or a legal person involved in cross-border tax activities.58 Although neither article 26 OECD Model, nor its Commentary, specifically mentions it, the provision can serve as the base for a joint audit since the list of forms of exchange described there is not exhaustive. Finally, an industry-wide exchange of information is an exchange concerning a whole economic sector (e.g. oil, pharmaceuticals, or the banking sector) and not one taxpayer in particular.59 The purpose of this type of exchange is to secure comprehensive data 51 OECD,

Manual on the Implementation of Exchange of Information Provisions for Tax Purposes’ (Paris: OECD Publishing, 2006), Module 2 on Spontaneous Exchange of Information, 3. 52 Ibid, Module 3 on Automatic (or routine) Exchange of Information (2012), 3. 53 OECD Commentary on art. 26 n. 9(b). 54 Ibid., n. 9.1. 55 Ibid., n. 9.1; OECD, Manual on the Implementation of Exchange of Information Provisions for Tax Purposes, Module 6 on Conducting Tax Examinations Abroad (2006). 56 OECD Commentary on art. 26, n. 9.1. 57 Ibid., n. 9.1; OECD, Manual on the Implementation of Exchange of Information Provisions for Tax Purposes, Module 5 on Conducting Simultaneous Exchange of Information (2006). 58 OECD, Manual on the Implementation of Exchange of Information Provisions for Tax Purposes, Module 9 on Joint Audits (2010). 59 OECD Commentary on art. 26, n. 9(a).

364 Xavier Oberson on industry practices, enabling tax inspectors to conduct more knowledgeable and effective examinations.60

21.6 Limits to the EOI 21.6.1 In General OECD Model DTT, article 26, paragraph 1 as a contract of international public law, entails an obligation to exchange information.61 This obligation does not exist, however, when the conditions of the exchange are not met or if the requested state refuses to supply the information based on a ground for refusal.62 Such grounds for refusal are provided by article 26 paragraph 2 OECD Model DTT. They can be based on: (1) the domestic law of the requesting or requested state; (2) the protection of a trade or business secret; or (3) ‘ordre public’. However, these limits are restricted by article 26 paragraph 5 of the OECD Model, which can be characterized as a derogation from paragraph 3. When the grounds for refusal in article 26 paragraph 3 are met, the requested state is not obliged to supply the information, but the requested state still has the discretion to provide the information.63 The possibility of supplying information, despite a ground for refusal under the treaty, may, however, be restricted by domestic law.64 According to the domestic law of many states, the discretion to provide information, in the presence of a ground for refusal or when the conditions of exchange are not met, may have criminal law consequences.65

21.6.2 Principle of Good Faith (Notably Requests Based on ‘Stolen Data’) International assistance in tax matters, based on a treaty, is also governed by the principle of good faith, according to articles 24 and 31 of the Vienna Convention on the Law of Treaties.66 In general, according to case law, the requesting state is presumed to be 60 OECD, Manual on the Implementation of Exchange of Information Provisions for Tax Purposes, Module 4 on Industry-wide Exchange of Information (2006). 61 Engelschalk, ‘Commentary to Art. 26 OECD MC’, n. 2. 62 Pistone and Gruber, ‘Die Möglichkeiten der Verweigerung des Informationsaustausches nach Art 26 OECD-MA’ , 79. 63 Ibid., 80; Engelschalk, ‘Commentary to Art. 26 OECD MC’, n. 98. 64 Pistone and Gruber, ‘Die Möglichkeiten der Verweigerung des Informationsaustausches nach Art 26 OECD-MA’ , 80; Engelschalk, ‘Commentary to Art. 26 OECD MC, n. 98. OECD Model DTC Commentary on art. 26, n. 14. 65 Pistone and Gruber, ibid., 81. 66 Vienna Convention on the Laws of Treaties, 23 May 1969.

Exchange of Information and Tax Treaties 365 acting in good faith.67 It is only if the requesting state provides contradictory or unclear information to justify its request that respect for the principle of good faith can be challenged.68 The test case, in this respect, is a request from a foreign state based on so- called stolen or illegal data. The Swiss courts have had to deal with this issue in various cases and arrived at contrasting views. In a first case, the French authorities, based on information obtained from stolen data, sent a request concerning French taxpayers holding accounts in a Swiss bank. The Supreme Court took a restrictive interpretation of a provision of Swiss domestic rules, which forbids the exchange of information based on ‘illegal behaviour’ under Swiss law. In that case, the arguably ‘illegal’ data had been obtained in France. As a consequence, the Supreme Court admitted the request for information because the illicit behaviour could not be regarded as a crime (or a misdemeanour) under Swiss law.69 Indeed, the arguable ‘illegal behaviour’, which formed the basis of the stolen data, had not been committed on Swiss territory. Secondly, about one month later, the Supreme Court also had to deal, in the famous (‘Falciani case’), with a request stemming from illicit data, notably a list obtained from the computer site of the HSBC bank in Geneva. This time, based on the previous jurisprudence, the Supreme Court could not admit the request, precisely because the illicit behaviour had been carried out on Swiss soil, in the premises of the Geneva bank.70 We may wonder whether such a distinction, based on the place of committing the illicit behaviour, makes any sense. After all, based on the principle of double incrimination, what is more relevant is the extent to which the specific criminal behaviour, even if committed abroad, would have been characterized as a crime or a misdemeanour under Swiss law. In our view, despite the principle of sovereignty, a state should remain the guardian of its constitution and of the rights embodied therein. It appears to be highly disputable and against the principle of good faith that in order to justify foreign assistance, the use of the data must originate from illicit behaviour, even if the criminal activity was conducted abroad. Apparently, the so-called Falciani list was later transferred to various states interested in the potential existence of undisclosed banks accounts. In two cases, based on the list of ‘stolen or illicit’ data, the Supreme Court dealt with requests from the competent Indian authorities about taxpayers arguably holding bank accounts in Switzerland.71 In that instance, the Supreme Court admitted that the request for assistance was not contrary to the good faith principle, as long as it was not proven that the Indian government had acquired the information directly from the seller of the data. In other words, illicit

67

FSC, 24 September 2015, ATF 142 II 161, para. 2.1. FSC, 16 February 2017, ATF 143 II 202, paras 6.3.1 and 8.7; FSC, 13 February 2017, 2C_954/2015, paras 5.2 and 5.2. 69 FSC, 16 February 2017, ATF 143 II 202. 70 FSC, 17 March 2017, ATF 143 II 224. 71 FSC, 2 August 2018, 2C_819/2017; FSC, 17 July 2018, 2C_648/2017. 68

366 Xavier Oberson data obtained indirectly, even if it establishes the base of the request, may still serve as justification for a request for information. In another case, the Supreme Court also confirmed that despite the fact that a ‘general agreement’ (global settlement) had been reached between the tax authorities and the taxpayer, in his place of residence, the competent authorities could still maintain a request for exchange of information without violating the principle of good faith.72

21.6.3 Principle of Subsidiarity Under this rule, the requesting state must first use all regular sources of information available under its domestic law before sending a request to the other state. The exchange of information, under article 26 OECD Model DTT, is therefore secondary to the domestic information-gathering process. As such, the principle of subsidiarity is not mentioned in the text of article 26 OECD Model DTT but in the Commentary. The legal basis of the principle of subsidiarity is disputed.73 According to some commentators, the principle derives from the requirement of ‘foreseeably relevance’.74 For others, the principle of subsidiarity is more based on the principle of good faith.75 The Swiss Supreme Court confirmed that the latter opinion appears ‘more convincing’.76 In general, the principle of subsidiarity aims to ensure that the burden of obtaining the information is not placed on the requested state, to the extent that said information would have been accessible to the requesting state according to its domestic tax procedure.77 From that perspective, the principle of good faith does indeed appear to be an appropriate legal basis.

21.6.4 Principle of Reciprocity A contracting state cannot be obliged to carry out administrative measures that are at variance with the laws and administrative practice of that or the other contracting state, or to supply information that is not obtainable under the laws or in the normal course of the administration of that or the other contracting state (art. 26 para. 3 lit. a and b OECD Model). It follows that a state cannot be requested to adopt measures that it cannot implement under its domestic law. In addition, existing domestic legal rules, such as a right

72

FSC, 16 April 2018, ATF 144 II 206. See OECD Commentary on. art. 26, n. 9 Model DTC. 74 In this sense, A. Donatsch et al., eds, Internationale Rechtshilfe unter Einbezug der Amtshilfe im Steuerrecht, 2nd ed. (Zurich: Schulthess, 2015), 234. 75 Opel, ‘Neuausrichtung der schweizerischen Abkommens-politik in Steuersachen’, 365. 76 FSC, 16 April 2018, ATF 144 II 206, para 3.3.3. 77 Ibid., para 3. 73

Exchange of Information and Tax Treaties 367 of refusal for information for witnesses, information holders, or taxpayers, may prevent the information from being supplied.78 In some countries, the domestic law will include procedures for notifying the person concerned with the enquiry prior to the supply of information. According to the OECD standard, these procedures—which belong to procedural rights granted under domestic law of the requested state—should not be applied in a manner that, in the particular circumstances of the request, would frustrate the efforts of the requesting state.79 This means that the notification rules should allow exceptions from prior notification, for instance ‘in cases in which the information request is of a very urgent nature or the notification is likely to undermine the chance of success of the investigation conducted by the requesting State’.80 Under article 26 paragraph 3 lit. a and b OECD Model DTT, not only the domestic law of the requested state is relevant, but also the domestic legislation and administrative practice of the requesting state are to be considered. It follows that a state cannot be asked to implement a measure that the requesting state would not be in a position to obtain under its own domestic rules. It follows that a contracting state cannot take advantage of the system of exchange of information of the other contracting state if that system is wider than its own.81 In exchange of information, based on article 26 paragraph 3 OECD Model DTT, the lowest common denominator of the legal rules of both contracting states is therefore applicable.82

21.6.5 Trade or Business Secret A contracting state is not obliged to supply information which would disclose any trade, business, industrial, commercial, or professional secret or trade process (art. 26 para. 3 lit. c OECD Model). As a ground for refusal, article 26 paragraph 3 lit. c OECD Model DTT should only refer to information which requires special and relevant protection.83 This should be the case only and to the extent that the protection of secrecy appears not to be sufficient in view of the need for confidentiality of the information.84 A trade or business secret is generally understood ‘to mean facts and circumstances that are of considerable economic importance and that can be exploited practically and the unauthorized use of which may lead to serious damage (e.g. may lead to severe financial

78 Ibid. 79

OECD Commentary on art. 26, n. 14.1..

80 Ibid. 81 Ibid.

82 Pistone and Gruber, ‘Die Möglichkeiten der Verweigerung des Informationsaustausches nach Art 26 OECD-MA’, 100; Engelschalk, ‘Commentary to Art. 26 OECD MC’, n. 101. 83 Pistone and Gruber, ibid., 103. 84 Ibid.

368 Xavier Oberson hardship)’.85 Although the concept of ‘trade or business secret’ is not defined in article 26 paragraph 3 OECD Model, the majority of commentators contend that this concept requires an autonomous definition.86 In addition, bank data and financial information (including books and records) are by nature not characterized as business secrets, commercial secrets, or other types of secret protected according to a treaty provision similar to article 26 paragraph 3 lit. c OECD Model.87 In this context, a state may decline to provide information relating to confidential communications between attorneys, solicitors, or other admitted legal representatives in their role as such and their clients (client–attorney privilege).88 According to the OECD, the scope of protection under this rule should not be overly broad so as to hamper the effective exchange of information. In particular, communications between clients and attorneys are only confidential if such representatives act in their capacity as attorneys, solicitors, or other admitted legal representatives and not in a different capacity, such as nominee shareholders, trustees, settlors, company directors, or under a power of attorney to represent a company in its business affairs.89

21.6.6 Public Policy Finally, a state cannot be obliged to supply information the disclosure of which would be contrary to public policy (ordre public) (art. 26 para. 3 lit. c OECD Model). This means that a requested state, in a specific situation, could refuse to exchange information if such supply would infringe its ‘fundamental evaluations’ of its domestic law.90 An example of where such a ground for refusal would be justified is where a tax investigation from a requesting state was motivated by political, racial, or religious motives.91 This limitation could also apply in cases where the information constitutes a state secret, for example information held by the secret service, the disclosure of which would be contrary to the vital interests of the requested state.92 An interesting issue in this context pertains to whether an exchange of information from a requesting state based on stolen data (CD, hard drives, etc.) may be challenged

85 OECD Commentary on art. 26, n. 19.2; see also Engelschalk, ‘Commentary to Art. 26 OECD MC’, in K. Vogel, ed., Klaus Vogel on Double Taxation Conventions, 3rd ed. (Alphen aan den Rijn: Kluwer, 1997), n. 110. 86 Pistone and Gruber, ‘Die Möglichkeiten der Verweigerung des Informationsaustausches nach Art 26 OECD-MA’, 102; Engelschalk, ‘Commentary to Art. 26 OECD MC’, n. 108 on art. 26. 87 FSC, 26 March 2019, 2C_615/2018, para 5; FSC, 9 July 2019, 2C_616/2018, para. 7; FAC, 29 June 2018, A-5694/2017, para. 6.11; FAC, 25 May 2020, A-3785/2018. 88 OECD Commentary on art. 26, n. 19.3. 89 Ibid. 90 Engelschalk, ‘Commentary to Art. 26 OECD MC’, n. 112. 91 OECD Commentary on art. 26, n. 19.5. 92 Ibid.

Exchange of Information and Tax Treaties 369 under article 26 paragraph 3 lit. c (public policy).93 In practice, such requests have been more specifically analysed under the requirement of good faith from the requesting state.94 Other states have adopted contrasting views. For instance, in the so-called KB- Lux case, the use of information obtained illegally was rejected by the Belgian tax administration but was accepted by the Dutch recipient tax authorities and authorized by the Dutch courts.95

21.6.7 The Derogation of Article 26 Paragraph 5 of the OECD Model Following the changes of 2005, the limits clause of article 26 paragraph 3 OECD Model has been restricted by a new paragraph 5, which provides that the provisions of paragraph 3 cannot be construed to permit a contracting state to decline to supply information ‘solely because the information is held by a bank, other financial institution, nominee or person acting in an agency or a fiduciary capacity or because it relates to ownership interests in a person’. This new paragraph is notably the result of the evolution of the international environment towards more global transparency, as already reflected in the publication, in 2002, of the TIEA Model.96 In particular, the Swiss Supreme Court has also confirmed that the rule of a provision similar to article 26 paragraph 5 OECD Model, is ‘self-executing’.97 First, article 26 paragraph 5 OECD Model DTT overrides paragraph 3, to the extent that some states would decline to supply information based on bank secrecy provisions under domestic law.98 However, the refusal to provide information may still apply if the reasons, based on the domestic law of the requested state, are unrelated to the person’s status as a bank, financial institution, etc.99 In this context, Swiss case law has ruled that banking information about a non-resident who appears to be the beneficial owner of a Swiss bank account (typically in so-called form A in the banking documentation) is foreseeably relevant.100 In our view, as an additional justification, the court could also have relied on the text of this provision, which refers to ‘ownership interests’, a concept

93 A. Steichen, ‘Information Exchange in Tax Matters: Luxembourg’s New Tax Policy’, in Rust/Fort (The Netherlands: Wolters Kluwer Law & Business, 2012), 9. 94 See Section 21.6.2. 95 F. Debelva and I. Mosquera, ‘Privacy and Confidentiality in Exchange of Information Procedures: Some Uncertainties, Many Issues, but Few Solutions’, Intertax 45 (2017), 362, 374. 96 Oberson, International Exchange of Information, 74. 97 FSC, 24 September 2015, ATF 142 II 161, para. 4.5.2; the description of the motives and the change of law can be found in the Federal Gazette (‘Feuille fédérale’), FF 2013, 7501ff. 98 OECD Commentary on art. 26, n. 19.11. 99 Ibid., n. 19.14. 100 FSC, 16 April 2018, ATF 144 II 206, para. 4.5; see also FSC, 24 September 2015, ATF 141 II 436, para. 4.6

370 Xavier Oberson that should also include beneficial ownership.101 In a more recent case, the Supreme Court also confirmed that a Swedish request for information on bank accounts ‘held by’ residents in Sweden could also include accounts held, not only directly but also indirectly, through a structure involving a company from Panama and a trust under which the Swedish taxpayers appeared as settlors.102 Secondly, a contracting state cannot refuse to provide information held by a person acting as a fiduciary or in the capacity of an agent, because such information would be treated as a professional secret under domestic law.103 A person is said to act in a fiduciary capacity when the business which the person transacts, or the money or property which the person handles, is not their own or for their own benefit, but for the benefit of another person to whom the fiduciary stands in a relationship implying and necessitating confidence and trust on the one part and good faith on the other part, such as a trustee.104 It follows that the concept of a ‘fiduciary’ also includes persons acting as nominees or trustees. The term agency is broad and covers all forms of services providers (company-formation agents, trust companies, etc.).105 Finally, contracting states cannot decline to supply information because it relates to an ownership interest in a person, including companies and partnerships, foundations, or similar organizational structures.106

21.7 Taxpayers’ Rights The principles governing EOI in various legal instruments tend to protect the interests of the contracting states. However, we believe that the interests of the taxpayers should also be considered, notably in cases where the state has a discretionary power of decision in the process. Such rights may be divided into substantive rights (human rights, constitutional rights, or data protection), on the one hand, and procedural rights, on the other hand (rights to be notified, to be heard or to appeal, and to control the legitimacy of the process).107 There are, however, different practices among states.108 In summary, there are two different schools of thought pertaining to the application of rights for the taxpayer within the EOI framework.109 On the one hand, many states consider administrative assistance as a fact gathering process, under which no specific rights should 101 Oberson, International Exchange of Information, 41s. 102

FSC, 27 August 2020, 2C_1037/2019, para. 5.3. OECD Commentary on art. 26, n. 19.12. 104 Ibid. 105 Ibid. 106 Ibid, n. 19.13. 107 See Oberson, International Exchange of Information, 268ff. 108 P. Baker and P. Pistone, ‘General Report’, in IFA, The practical protection of taxpayers’ fundamental rights, vol. 100b (Basel Congress), The Hague (2015), 15ff. 109 Oberson, International Exchange of Information, 270. 103

Exchange of Information and Tax Treaties 371 be granted to the taxpayer at the level of the requested state (investigation stage). The taxpayer may then invoke all relevant legal rights at the contentious stage, in front of the requesting state.110 On the other hand, some states tend to view the EOI as an independent procedure, under which the persons involved should be granted the relevant procedural rights. This opinion should prevail because, notably, the secrecy rules, and the limits to the EOI, are also in the interests of the taxpayer.111 The essential rights which are then granted to the taxpayer can be found in the domestic laws of the various states.

110 This position was notably adopted by the European Court of Justice (CJEU), in the framework of Directive 77/799 on tax cooperation between Member States, in the famous Sabou case (C-276/12), as far as the position of a taxpayer under investigation is concerned. In Case C-682/15 Berlioz, however, the CJEU admitted that the information holder should at least be granted some rights because they do not participate in the second stage of the investigation. Further, in Joined Cases C-245/19 and C-246/ 19 Luxembourg v. B, the Court ruled that the right to an effective remedy, guaranteed by the Charter of Fundamental Rights of the EU, requires that the person that holds the information requested in an EOI process must be able to bring a direct action against such a request. 111 Oberson, International Exchange of Information, 270.

Chapter 22

Beneficial Ow ne rsh i p and Tax Tre at i e s Dietmar Gosch and Nadia Altenburg

22.1 Introduction The principle of beneficial ownership has to be placed in context of the introduction of modern income tax systems, the subsequent occurrence of double taxation, as well as efforts to ensure that the consequently introduced double taxation treaties are correctly applied and grant benefits only to the actual economically entitled taxpayer. The scope of the beneficial ownership principle therefore extended the idea of preventing non-entitled parties from obtaining treaty protection. The idea was developed mainly to allocate income correctly to a taxpayer in situations where the actual recipient of a payment was obviously not the person entitled to treaty protection because the payments were received by someone other than that person, such as: (1) nominees or agents; (2) conduits companies; or (3) trusts. The latter were a particular concern for common law jurisdictions. From the initial efforts to correctly allocate income streams to the ‘true recipient’, it was quickly identified that agent and conduit structures bear a high risk of abuse, especially for notably mobile income such as dividends, interests, or licence payments. Due to the particularities of its common law system, the UK1 unsurprisingly advocated 1

In the UK, a trustee or agent is taxable on the basis of receiving or being legally entitled to income, so that receiving without being beneficially entitled is sufficient. While a UK resident trustee receiving foreign income for a non-resident beneficiary is not taxable, the trustee is taxable if receiving it for a resident beneficiary and so, arguably, is a UK resident for treaty purposes in respect of all income received as trustee. The UK thus may have been concerned about reducing its withholding tax in its capacity as a source state if the treaty partner state had similar taxing rules and the income was beneficially owned by a person who was not a resident of that state. Also see R. J. Vann, ‘Beneficial Ownership: What Does History (and Maybe Policy) Tell Us’, in M. Lang et al., eds, Beneficial Ownership: Recent Trends (Amsterdam: IBFD, 2013), 267, 285; A. Meindl-Ringler, Beneficial Ownership in International Tax Law (Alpen aan den Rijn: Wolters Kluwer, 2016), 19.

374 Dietmar Gosch and Nadia Altenburg manifesting the principle of beneficial ownership into modern tax treaties as a measure for encountering potential cases of abuse.2 The protocols to the OECDs Working Party 273 show that the delegate from the UK strongly supported the introduction of a subject- to-tax test into the articles on dividends, interest, and royalties.4 Such subject-to-tax tests aim to ensure that tax relief at source should only apply if the beneficial owner of the income in question is resident in the other contracting state; otherwise, there would have been reason to fear that ‘the articles are open to abuse by taxpayers who are resident in a third country and who could, for instance, put their income into the hands of bare nominees who are resident in the other contracting State’.5 In the view of Working Party 27, ‘making relief in the state of source dependant on effective taxation in the state of residence, was contrary to the spirit and general economy of the Draft Convention and would moreover give rise to difficulties in the appreciation of the concept of effective taxation.’6 After balancing the effects and consequences of the test, Working Party 27 therefore agreed to recommend ‘that there be written into the Model Convention a provision whereby the beneficial owner text would be applied’.7 Consequently, the term ‘beneficial owner’ was introduced for the first time into articles 10(2)8, 11(2),9 and 12(1)10 of the final version of the 1977 OECD Model11 as well as into the respective passages of the Commentary on the OECD Model Convention (OECD MC).12 Later, in 2017, the term was also included in article 8(1) of the Multilateral Instrument (MLI),13 and more recently in the draft of a new article 12B of the 2017 UN Model under the heading ‘Income from Automated Digital Services’.14 The introduction of the term ‘beneficial owner’ into the 1977 OECD Model cannot, however, be understood as a remarkable change in treaty application and interpretation. 2

The principle of beneficial ownership was first discussed in an internal memo regarding the double taxation convention between the UK and the USA in 1945 in relation to granting treaty benefits for dividend payments, Meindl-Ringler, Beneficial Ownership in International Tax Law, 17 (for details see fn. 64 referring to FC/WP27(68)1 (30 Dec. 1968)), 14. 3 The OEEC and OECD documents referred to are available at https://www.taxtreatieshistory.org. 4 OECD, TFD/ FC/216 (1967), 14; Vann, ‘Beneficial Ownership: What Does History (and Maybe Policy) Tell Us’, 285. 5 OECD, TFD/FC/216 (1967), 14. 6 OECD, FC/WP27 (70)1 (1970), 13. 7 Ibid. 8 In connection with the term ‘payment’. 9 In connection with the term ‘draw’. 10 Evidently in connection with both terms. 11 See also, for Germany, ‘Basis for negotiation for agreements for the avoidance of double taxation and the prevention of fiscal evasion with respect to taxes on income and on capital’ (22 Aug. 2013). 12 OECD Committee on Fiscal Affairs, Model Double Tax Convention on Income and on Capital (Paris: OECD Publishing, 1977), paras 5 and 8 on art. 11, and para. 4 on art. 12. 13 OECD, Multilateral Convention to Implement Tax Treaties Relates Measures to Prevent Base Erosion and Profit Shifting. 14 https:// w ww.un.org/ d eve l opm e nt/ d esa/ f inanc i ng/ e ve nts/ 2 1st- s ess i on- c ommit t ee- e xpe r ts- international-cooperation-tax-matters. See Hidien, ‘Entwurf einer neuen Verteilungsnorm für die Besteuerung von automatisierten, digitalen Dienstleistungen im UN-MA 2017—Lehren für Paris?’, Internationale Steuer Rundschau (2021), 255.

Beneficial Ownership and Tax Treaties 375 As can be derived from Protocols from Working Party 27, civil law jurisdictions have always implicitly assumed that only the ‘true recipient’ should be subject to treaty benefits.15 Therefore, the explicit introduction of the term into the OECD MC did not per se change the general application of tax treaties themselves. The explicit introduction of the principle can be marked as the starting point of the discussion of whether the term is to be understood in a narrow ‘legal’ sense or based on a broad understanding, building on economic principles.

22.2 Legal vs Economic View 22.2.1 The View of the OECD It is the OECD’s understanding that the Model Tax Convention requires constant review to address new tax issues that arise in connection with the evolution of the global economy. Lead by the work of Working Party No. 1 of the OECD’s Committee on Fiscal Affairs, the OECD’s understanding of beneficial ownership has broadened over time. While the added wording in the 1977 OECD Commentary still supported a relatively narrow meaning of beneficial ownership that can be seen as a technical legal understanding, the 2003 OECD Commentary advanced to a broader interpretation ‘in the light of the object and purpose of the Convention, including avoiding double taxation and the prevention of fiscal evasion and avoidance’16 that indicates an increasingly economic view. This development is emphasized in the language of the 2003 Commentary that states in paragraph 8.1 of article 11 that ‘it would be inconsistent with the object and purpose of the Convention for the State of source to grant relief or exemption where a resident of a Contracting State, otherwise than through an agency or nominee relationship, simply acts as a conduit for another person who in fact receives the benefit of the income concerned’. Based on the work performed by the OECD on conduit companies17 in 1986, and further work on the clarification of the meaning of ‘beneficial owner’ in the OECD Model Tax Convention,18 the OECD today clearly advocates a broad approach that explicitly goes beyond the meaning ‘the term might have under domestic law’ and is therefore an economic interpretation of the term.19 The report from the Committee on Fiscal Affairs 15 OECD,

DAF/FC/69(10) (1969), 6; Vann, ‘Beneficial Ownership: What Does History (and Maybe Policy) Tell Us’, 283. 16 OECD, Model Tax Convention on Income and on Capital: Condensed Version 2017 (Paris: OECD Publishing, 2017), para. 12.1 on art. 10, para. 9 on art. 11, and para. 4 on art. 12. 17 OECD, Report on Double Taxation Conventions and the Use of Conduit Companies (27 Nov. 1986); OECD, ‘Report (6) on Double Taxation Conventions and the Use of Conduit Companies’, in Model Tax Convention on Income and on Capital: Full Version 2014 (Paris: OECD Publishing, 2015). 18 OECD CTPA, ‘Clarification of the Meaning of “Beneficial Owner” in the OECD Model Tax Convention’, Public discussion draft (Apr. 2011), 29, http://www.oecd.org/tax/treaties/47643872.pdf. 19 2017 OECD MC, para. 12.1 on art. 10, para. 9 on art. 11, and para. 4 on art. 12.

376 Dietmar Gosch and Nadia Altenburg entitled ‘Double Taxation Conventions and the Use of Conduit Companies’20 concluded that ‘a conduit company cannot normally be regarded as the beneficial owner if, though the formal owner, it has, as a practical matter, very narrow powers which render it, in relation to the income concerned, a mere fiduciary or administrator acting on account of the interested parties’.21 The OECD discussion draft led to subsequent changes in the Commentary’s language and clarified that: • beneficial owner’s rights might not only be ‘constrained by a contractual or legal obligation to pass on the payment received to another person’ but the obligation may be based on ‘facts and circumstances’ showing that the recipient’s rights are ‘in substance’ constrained; • any obligation to pass the payment received to another person ‘must be related to the payment received’ and ‘would therefore not include contractual or legal obligations unrelated to the payment received even if those obligations effectively result in the recipient using the payment to satisfy those obligations’.22 The subsequent language of the OECD Commentary therefore required that ‘this type of obligation would not include contractual or legal obligations that are not dependent on the receipt of the payment by the direct recipient such as an obligation that is not dependent on the receipt of the payment . . .’23 The addendum that the ‘domestic law meaning is applicable to the extent that it is consistent with the general guidance included in this Commentary’24 was not included in the Commentary’s phraseology.25 However, the wording of the 2017 OECD Commentary26 clearly stipulates that the OECD builds on an autonomous interpretation of beneficial ownership and that a domestic understanding is not decisive for the interpretation, even though the OECD MC itself lacks a clear definition of the term. It is, however, this lack of definition that gives rise to uncertainty when applying the beneficial ownership principle in practice and therefore continues to stir the discussion to what extent a domestic understanding can be used to construe the term. Some bilateral double taxation treaties include explicit collision-resolving regulations regarding the application of the beneficial ownership principle (and understand the beneficial owner to be the person to whom the income is attributable for tax purposes).27 20

OECD, Report (6). Ibid, para. 14 b. 22 OECD CTPA, ‘Revised Proposals concerning the Meaning of Beneficial Owner in Articles 10, 11 and 12 of the OECD Model Tax Convention’ (Oct. 2012). 23 2017 OECD MC, para. 12.4 on art. 10, para. 102. on art. 11, and para. 4.3 on art. 12. 24 OECD CTPA, ‘Clarification of the Meaning of “Beneficial Owner” ’, 3, para. 12.1; partially criticized by B. J. Arnold in ‘International Tax Treaty News’, Bulletin for International Taxation 65/10 (2011). 25 OECD CTPA, ‘Clarification of the Meaning of “Beneficial Owner” ’, para. 12.1. 26 2017 OECD MC, para. 12.1 on art. 10, para. 9 on art. 11, and para. 4 on art. 12. 27 Similarly, the protocol to the German– Italian double taxation treaty (1989), no. 9; the protocol to the German–Norwegian double taxation treaty (1992), no. 4; art. 43 para. 3 of the German–Swedish double taxation treaty; the protocol to the German–USA double taxation treaty (1992), no. 10; the 21

Beneficial Ownership and Tax Treaties 377 However, such individual agreements cannot be generalized. They therefore remain without any fundamental significance for interpretation purposes. Against this background, it would be conceivable to fall back on the ‘origins’ of the term beneficial ownership in common law as promoted by the UK and to consider it to be decisive. Such a recourse must, however, fail as there is no indication that states unfamiliar with the term would wish to adopt the Anglican law without explicitly agreeing to use the term in its original sense. Those states will frequently be more inclined to understand the term in the context of allocating income or preventing abuse. It is precisely this heterogeneous understanding of the term beneficial ownership that suggests double taxation treaties lack an autonomous definition of the term and that—despite the OECD’s guidance in its Commentary—domestic law (lex fori) is therefore decisive for its interpretation (art. 3 para. 2 OECD MC). There are further considerations that might suggest drawing on a domestic understanding of the term. These include the aspect of determining income and—above all—the aspect of allocating income as already mentioned. This corresponds to the general understanding that double taxation treaties do not contain rules on the allocation of income. Following article 4 OECD MC only the taxable person, object, and, in relation to the taxable person, the residence need to be determined as an essential feature of treaty entitlement. However, when determining the source of income and thus the allocation of such income, drawing on domestic law is inevitable. Such an understanding would be supported if the term could be substituted by the otherwise common term of the ‘recipient of the payment’ as such a determination ultimately depends on domestic law.28 Yet, if the concept of beneficial ownership is reduced to the mere attribution and determination of the taxable person, it would result in a clear risk of double taxation. It is, however, the consistent and explicit objective of the concerned articles in the OECD MC to establish concordance between the withholding taxation in the source state, on the one hand, and the crediting of taxes in the country of residence, on the other hand, thus a common understanding is even more necessary. It therefore reflects the ‘spirit’ of the double taxation treaties to rely as far as possible on the context of the treaty itself. If reduced to an instrument of income allocation, beneficial ownership would be redundant as this is a definite responsibility of domestic law. Without an independent meaning, the term would be dispensable.29

protocol to the German–Australian double taxation treaty (1972), no. 7 (esp. the differentiation between the entitled party and the economic owner of the assets from the Australian side). 28 Also

see F. Haase, ‘Nationales Steuerrecht vs. DBA— Ausgewählte terminologische und systematische Unterschiede’, Internationale Steuer Rundschau (2014), 185; see also F. Wassermeyer, ‘Seminar D: Missbräuchliche Inanspruchnahme von Doppelbesteuerungsabkommen’, Internationales Steuerrecht (2000), 505. 29 J. Wheeler, The Missing Keystone of Income Tax Treaties, vol. 23 Doctoral Series (Amsterdam: IBFD, 2012), 17; critical: C. Kaeser and F. Wassermeyer in Wassermeyer’s Doppelbesteerungsabkommen Kommentar (Munich: C. H. Beck, 2015), art. 10 MA paras 72ff.

378 Dietmar Gosch and Nadia Altenburg Based on these considerations, a separation of income allocation and beneficial ownership should be favoured along with an acknowledgement that they have different interpretations. The interpretation of the term beneficial ownership should therefore—also in accordance with article 34 of the Vienna Convention on the Law of Treaties (VCLT)—be derived from the context of the agreement rather than from the domestic law of the respective state of use.30 The OECD Commentary seeks to provide further guidance and has always endeavoured to ‘clarify’ the correct understanding of the term.

22.2.2 The View of the EU and the ECJ As with the OECD MC and the OECD Commentary, EU law does not contain a comprehensive definition of the term beneficial ownership. However, the term is used in the Interest and Royalties Directive,31 (‘the Directive’) that grants an exemption from withholding tax only where the recipient is the beneficial owner of the payment in order to prevent abusive tax avoidance, characterized by the Commission as ‘artificial conduit arrangements’.32 Article 1 of the Directive provides: 1. Interest or royalty payments arising in a Member State shall be exempt from any taxes imposed on those payments in that State, whether by deduction at source or by assessment, provided that the beneficial owner of the interest or royalties is a company of another Member State or a permanent establishment situated in another Member State of a company of a Member State. . . .

30

Also art. 31 VCLT advocating for a strict interpretation of the treaty by itself without consultation of the domestic law or supplementary documents, for more details on this, see H. J. Ault in ‘The Role of OECD Commentaries in interpreting Tax Treaties’, Intertax 4 (1994), 144. 31 Council Directive 2003/ 49/ EC of 3 June 2003. Deviating from this, a beneficial ownership requirement is unknown to the EU Parent–Subsidiary Directive Counsel Directive 2011/96/EU enacted in 1990, recast 30 November 2011. See Opinion of Kokott AG, 1 March 2018, in Case C-117/16, paras 31ff. At para. 43: In that regard, it makes perfect sense that (unlike the Interest and Royalties Directive) the Parent–Subsidiary Directive is ‘only’ predicated on the distribution of profits by a subsidiary to its parent company (which must have a certain minimum holding). Unlike interest payments, dividends do not, as a rule, represent operating expenditure which may be set against profit; therefore, it makes sense that, according to its wording, the Parent–Subsidiary Directive does not contain any further substantive criteria (such as drawing of dividends in one’s own name and on one’s own account or suchlike). 32 Report from the Commission to the Council in accordance with article 8 of Council Directive 2003/49/EC on a common system of taxation applicable to interest and royalty payments made between associated companies of different Member States, COM(2009)179 final, 8.

Beneficial Ownership and Tax Treaties 379 4. A company of a Member State shall be treated as the beneficial owner of interest or royalties only if it receives those payments for its own benefit and not as an intermediary, such as an agent, trustee or authorised signatory, for some other person.33 This wording seems to imply that in order to determine whether benefits have to be granted under the Directive, one does not necessarily have to look at the recipient of the payments itself but, rather, should identify whether the beneficial owner is a company in a member state (or has a permanent establishment in another member state to which the payment has to be allocated). This would mean that in cases where the direct recipient is not receiving the payments for their own benefit, the tax exemption under the Directive can still be claimed if the beneficial owner behind the recipient is also a company resident in any member state. In the European Court of Justice (ECJ), Advocate General Kokott is of the opinion that the beneficiary owner within the meaning of the Directive is always the person who is entitled to the civil law claim; however, that understanding is not necessarily certain.34 In any event, the Directive clarifies in article 5 that it does not preclude the application of domestic or treaty-based provisions required for the prevention of fraud or abuse and that member states (in the case of transactions for which the principal motive or one of the principal motives is tax evasion, tax avoidance, or abuse) may withdraw the benefits of the Directive or refuse its application. Therefore, benefits could be denied in cases where the beneficial owner, who is not the direct recipient, is located in a member state but the discrepancy between the beneficial owner and the direct recipient is rooted in the motivation to evade or avoid taxes and where the domestic or treaty law provides a legal basis for denying benefits on such grounds. In principle, the decisions of the ECJ35 outline the same understanding. The Court states in its Danish cases that the mere fact that the company which receives the interest in a member state is not its ‘beneficial owner’ does not necessarily mean that the withholding tax exemption provided for in the Directive is not applicable.36 The Court clarified that interest will also be exempt in the source state when the receiving company transfers the respective amount to a beneficial owner who is established in the EU,37 which means that the concept of ‘beneficial ownership of the interest’ within the meaning of the Directive must be interpreted as designating an entity which actually 33

Emphasis added. Opinion of Kokott AG, Case C-117/16, paras 40ff; on this I. Lazarov, ‘(Un)Tangling Tax Avoidance Under the Interest and Royalties Directive: the Opinion of AG Kokott in N Luxembourg 1’, Intertax 46/11 (2018), 873. 35 ECJ, Joined Cases C- 115/16, C-118/16, C-119/16 and C-299/16, ECLI:EU:C:2019:134 26, February 2019. -In the meantime, the final judgments of the Danish Supreme Court (Højesteret), 9 January 2023 - 69/2021, 79/2021, 70/2021, and 4 May 2023 -116/2021, 117/2021 are available at https://domstol.dk/hoejeste ret/soeg/?iss=1&SenderQuery=1491&st=2&q=beneficial. 36 Ibid., para. 94. 37 Ibid. 34

380 Dietmar Gosch and Nadia Altenburg benefits from the received interest.38 In this context, the ECJ explicitly clarified—as did the 2017 OECD Commentary—that the concept of ‘beneficial ownership of the interest’ as mentioned in the Directive cannot refer to concepts of national law that vary in scope.39 However, when it comes to defining and clarifying the meaning of beneficial ownership, the Court both evades and elaborates as follows: • in principle, the beneficial owner has to be identified;40 • however, in cases of an abuse of rights or an abusive practice, a member state that wishes to deny a tax exemption under the Directive does not have to identify the beneficial owner of a payment;41 • the existence of a double taxation convention that grants similar tax benefits to a recipient does not exclude an abuse of rights per se;42 • the member state has to establish the existence of the elements constituting such an abusive practice whilst taking account of all relevant factors.43 When outlining the constituent elements of an abuse of rights, the ECJ names artificial arrangements set up for reasons that do not reflect economic reality, singling out the interposition of a conduit company as a specific example.44 The Court requires as a subjective element that the conduit is interposed with the principal objective (or one of the principal objectives) of obtaining tax advantages running counter to the aim or purpose of the applicable tax law. It lists numerous obvious indicators that can support the existence of a subjective element, inter alia: • the conduit companies have to pass on the payments very soon after receipt to a company that does not fulfil the conditions of the Directive;45 • the conduit company has no activities other than receiving interests, which can be established where the company lacks appropriate substance;46 or • where the conduit company has entered into various contracts giving rise to intra- group flows without any economic activity by the conduit company itself.47 Instead of outlining the Court’s understanding of the term beneficial ownership, the ECJ focuses on identifying abusive structures. At first glance, the approach of the ECJ 38

Ibid., paras 94 and 88. Ibid., para. 84. 40 Ibid., para. 88. 41 Ibid., para. 143. 42 Ibid., para. 135. 43 Ibid., para. 142. 44 Ibid., para. 127. 45 Ibid., para. 128. 46 Ibid., para. 131. 47 Ibid., para. 132. 39

Beneficial Ownership and Tax Treaties 381 appears sensible as it spares the respective tax administration the burdensome exercise of identifying the actual beneficial owner by first analysing whether a case exhibits any elements of abusive practice, which makes the analysis of beneficial ownership obsolete. However, this leaves the member states without guidance regarding the interpretation of the term and diverts the discussion to the fact that, in its decisions, the ECJ elevates the objective of preventing abuse and tax avoidance as a general legal principle to the rank of primary law and thus to the same rank as constitutional law.48 As the EU is not mandated to harmonize direct taxes, the decision whether to tax certain transactions and, if so, in which form, lies exclusively at the discretion of each member state. European law—primary or secondary—can therefore only limit the taxation rights of a member state but cannot constitute them. An extension of taxation rights based on the refusal to grant benefits under a European directive contradicts the system of EU and domestic law and can only be based on domestic anti-avoidance provisions.49

22.3 Domestic Jurisprudence Within domestic jurisprudence, the interpretation of beneficial ownership is fragmented. Countries are neither following a clear pattern of domestic interpretation or an independent international fiscal understanding, nor one of a legal or economic approach. Countries with a civil as well as a common law background are applying either of the approaches with different emphases. As expected,50 the UK is following a broad economic interpretation based on an international fiscal meaning that can be traced back to the Indofood51 decision of the UK Court of Appeal. Even though the decision is abnormal in many aspects, it laid the ground for further guidance issued by His Majesty’s Revenue and Customs (HMRC). The circumstances in the case were rather extreme. Therefore, the comprehensive explanations by the judge on the interpretation of the term beneficial ownership appear unnecessary as beneficial ownership could also have been denied on the basis of a legal understanding. The decision did, however, anchor the economic and independent interpretation of the term that has subsequently been clarified by HMRC in further guidance.52 Such guidance, for example, stipulates that beneficial ownership will not be tested in cases in which the recipient might have an obligation to pass on the payments

48 See

W. Schön, ‘Rechtsmissbrauch im europäischen (Steuer)Recht –eine methodologische und institutionelle Perspektive—Teil 2’, Europäische Zeitschrift für Wirtschaftsrecht (2020), 685. 49 See ibid. 50 It was the UK that advocated the introduction of the beneficial ownership concept in the OECD Model, see Section 22.1 for further details. 51 Indofood International Finance Ltd. v. JP Morgan Chase Bank [2006] EWCA Civ 158. 52 HMRC guidance of 9 April 2016: INTM332050, INTM 332060, and INTM 504030.

382 Dietmar Gosch and Nadia Altenburg but where the actual beneficial owner would also be entitled to the same treaty benefits since there would be no fiscal evasion or avoidance. Canada, another example of a common law country, has adopted a rather restrictive legal approach with a narrow understanding. This was demonstrated in the Tax Court decisions such as Prévost Car53 and Velcro,54 in which the court considered domestic law via article 3 paragraph 2 of the Dutch–Canadian treaty. Based on this understanding, the Canadian court rejected the prosecution’s invitation to determine that beneficial owner ‘mean[s]‌the person who can, in fact, ultimately benefit from the dividend’ as such a proposed definition does not appear anywhere in the OECD documents and the use of the word ‘can’ would open up a myriad of possibilities, jeopardizing the relative degree of certainty and stability that a tax treaty seeks to achieve.55 Instead, the court based its reasoning on the fact that the receiving entity was not acting as an agent or nominee;56 that it could not be said that the entity receiving payments in the case at hand had ‘absolutely no discretion as to the use or application of funds put through it as a conduit’ and had not ‘agreed to act on someone else’s behalf pursuant to that person’s instructions without any right to do other than what that person;57 and that there was no predetermined or automatic flow of funds to the shareholders of the receiving entity.58 Also in the Velcro59 decision, the court did not deny the treaty benefits based on the fact that the receiving entity was legally entitled to the payments and that the payments were deposited in its own account, converted into other currencies, and earned interest as well as commingling with other funds of the entity which were then used for various purposes including bill payments. The court determined that there was no automatic flow of funds from the receiving entity to its shareholders and therefore rejected the qualification of it as a mere conduit, agent, or nominee. The court emphasized that this would only be the case where the company has absolutely no discretion. The USA as a third example of a common law country appears to follow, as with Canada, a rather narrow technical understanding60 of the term based on the domestic understanding of ‘dominion and control’ of assets via article 3 paragraph 2 of the applicable treaty as demonstrated by the decisions in Aiken Industries61 and Del Commercial.62 In both decisions, treaty benefits were denied as the receiving entities were determined 53

Prévost Car Inc. v. Canada, 2008 TCC 231, 2004-4226(IT)G,. Velcro Canada Inc. v. Canada, 2012 TCC 57, 2007-1806(IT)G,. 55 Prévost Car Inc. v. Canada, 2008 TCC 231. 56 Ibid., para. 100. 57 Ibid. 58 Ibid., para. 102. 59 Velcro Canada Inc. v. Canada, 2012 TCC 57. 60 This was also demonstrated in the US opposition to the Indofood decision, insisting on the ability to define the term themselves, see B. Gibert and Y. Oumrane, ‘Beneficial Ownership—A French Perspective’, European Tax (Jan. 2008), 2. 61 Aiken Industries, Inc. v. Commissioner of Internal Revenue, United States Tax Court, 5 August 1971, 292-69, 56 TC 925. 62 Del Commercial Properties Inc. v. Commissioner of Internal Revenue, United States Tax Court, 20 December 1999, 1887-98, TC Memo 441. 54

Beneficial Ownership and Tax Treaties 383 to be conduit companies, as the receiving entities did not have any notable commercial substance and were committed to paying out exactly what they had collected without generating any profit themselves. In Del Commercial,63 the paying entity additionally began to make payments directly to the shareholder of the conduit company and therefore, in its decision, the court did not even refer to the beneficial ownership principle. Both cases are interpreted as a roadmap on how to argue beneficial ownership in order to ensure treaty benefits. Subsequently, courts have acknowledged beneficial ownership in cases where: (1) a financial affiliate was adequately capitalized and had a valid business purpose; and (2) a back-to-back licence agreement had sufficiently separate and distinct terms that the royalties received by the taxpayer were not directly related to royalties that the taxpayer paid to another group entity. Based on the narrow understanding used by the USA that beneficial ownership and treaty benefits can only be denied where cash flows actually match, the principle was deemed ineffective and specific domestic anti-abuse rules were implemented in domestic law and limitation-on-benefits (LOB) clauses were included in its tax treaties. The Netherlands takes a very prominent position within civil law countries when it comes to cross-border tax planning. The Dutch understanding of the beneficial ownership concept is therefore of particular interest. As expected, the Hoge Raad (Dutch Supreme Court) follows—without explicitly referencing article 3 paragraph 2 of the applicable treaty—a very strict technical interpretation of the term and would deny beneficial ownership only in circumstances where the recipient of a payment is acting as an agent, a nominee, or is legally obliged to pass the receipts on to a third party.64 Even in cases where the recipient was established in a low-tax jurisdiction, was not carrying out any substantial economic activity, and was only established for tax reasons, the Dutch court would allow beneficial ownership as long as the recipient can freely dispose of the income received.65 The Netherlands applies other concepts to counter abusive tax practices, such as the fraus legis, the fraus conventionis, or the substance-over-form doctrine. Yet the Dutch courts apply these anti-abuse measures in a very conservative manner. The fraus legis doctrine is only applied where the application of the principle has been agreed with the respective treaty partner, while the fraus conventionis doctrine is understood to overrule treaty law.66 Nevertheless, Dutch courts apply the fraus conventionis and the substance- over-form doctrine only in situations where the tax evasion or avoidance is blatant, for example where structures were only established for a specific transaction.67

63 Ibid. 64

Market Marker Case, Hoge Raad, 6 June 1994, BNB 1994/217.

65 Ibid.

66 D. Smit, ‘The Concept of Beneficial Ownership and Possible Alternative Remedies in Netherlands Case Law’, in Lang et al., Beneficial Ownership, 68ff. 67 Ibid.

384 Dietmar Gosch and Nadia Altenburg Interestingly, in cases where the beneficial ownership of the direct recipient was denied, Dutch courts will apply the available identified taxpayer’s treaty benefits to the actual beneficial owner.68 While Swiss courts refer via article 3 paragraph 2 of the OECD MC (or the equivalent of the respective tax treaty) to the domestic law of the source state (as the state of application) and thereby rely on Swiss withholding tax legislation, they nevertheless adopt a broad interpretation of the term béneficiaire combined with a substance-over-form approach.69 However, the Federal Administrative Court has clarified that where reference is made to domestic law, it must be viewed in connection with the object and purpose of the tax treaty—here, the avoidance of double taxation.70 Swiss legislation includes the beneficial ownership principle based on an economic point of view and analyses whether the recipient has the authority to dispose of the payment which was subject to withholding taxes. Any fiduciary relationship or obligation to transfer the taxed payments to a third party would exclude such authority. Beneficial ownership can therefore be denied in the event that the applicant has a contractual or factual obligation to pass payments on to a third party.71 The recent decisions of the ECJ have related to cases emanating from the Danish courts,72 and confirm a broad interpretation of the beneficial ownership concept for challenging tax-motivated structures based on an independent international fiscal meaning taking into account OECD materials.73 In the HHU74 case, the Court based its decision on the facts that the transaction in question took place between related parties, that there was a certain automatism present in channelling the funds from the Danish company through a Swedish company and yet further on to the ultimate parent company, and that the structure’s purpose was to avoid Danish withholding tax. The Court further took into account that in Sweden no tax was due and no business activities or any other form of substance existed.75 Similar factors were present in the Cook76 case, where the transactions leading to the structure in question were also made between related parties. In addition, the funds were channelled through a Swedish company where the net income was not taxed and, as a consequence, the payments were passed on without any tax leakage. The Swedish 68

Hoge Raad, 28 June 1989, BNB 1990/45. Tax Appeals Commission of Switzerland, VBP 65.86 (28 Feb. 2001); Swiss Federal Administrative Court A-6053/2010, ASA 79 (2010/2011), 926; Federal Supreme Court (4 Feb. 2020), 2C_ 344/2018; Federal Supreme Court (20 Apr. 2020), 2C_354/2018, both decisions available at https://www. bger.ch/index.htm. 70 Federal Administrative Court, 7 March 2012, A-6537/2010. 71 Ibid.; Federal Administrative Court, 23 July 2012, A-1246/2011. 72 See Section 22.2.2. 73 A. Meindl-Ringler, Beneficial Ownership in International Tax Law (Alphen aan den Rijn: Wolters Kluwer, 2016), 253. 74 The HHU Case, SKM 2011.57 LSR. 75 J. Bundgaard, ‘The Notion of Beneficial Ownership in Danish Tax Law: The Creation of a new Order with Uncertainty as a Companion’, in Lang et al., Beneficial Ownership: Recent Trends, 113. 76 The Cook Case, SKM 2011,485 LSR. 69 Federal

Beneficial Ownership and Tax Treaties 385 holding company was not engaged in any activities other than holding shares and had no other business activities.77 However, in the ISS78 case, in which dividend and interest payments received by a holding company were not immediately forwarded to its parent company, the holding company was held to be the beneficial owner of the payments as the management of the holding company was ‘authorised under corporate law to manage the company and decide on the use of the received dividends’. The Bundesfinanzhof (BFH, German Federal Fiscal Court) has also had to deal with the concept of a beneficial owner on various occasions and takes a ‘formal’, restrictive view to determining beneficial ownership. It therefore grants the term beneficial owner, which translates to ‘entitled user’ (i.e. art. 10 para. 2 US–German treaty of 1989/2008), only the character of a mere income-allocation provision.79 According to the BFH, this aligns with the systematic position of the provision in article 1 of the US–German treaty 1989/2008, which defines the ‘general scope’ of the treaty and in this context (in its para. 1) interlinks the treaty entitlement of the persons concerned and the treaty benefits with the residence of those persons in one of the two contracting states. However, the provision does not actually determine to which person income has to be allocated, and it certainly does not order a so-called linkage of qualification which could bind the source country to the qualification of the legal entity in the country of residence. The wording of the provision does not provide for such a far-reaching interpretation. The questions of determining the recipient of the income and the person to whom the income should be allocated should therefore be kept strictly apart, with income allocation being subject to the legislation of the respective state. In this sense, the BFH understands the beneficial owner only as a qualified ‘payee’, as is still understood in some double taxation treaties. An interpretation that exceeds the mere term beneficial owner is ruled out according to the BFH.80 If such an interpretation were to take hold, it could only occur in the form of a treaty override.81 The court is solely guided by domestic allocation rules, on the one hand by the so- called beneficial owner, and on the other by the general abuse avoidance rule. It is worth mentioning, however, that in later decisions the BFH has increasingly detached the legal figure of the beneficial owner from the merely legal interpretation and construed the term from an economic view, and likewise in the context of abuse avoidance.82 Naturally, this indirectly affects the interpretation under treaty law. Interestingly, the BFH, in its established jurisprudence, only assigns an indicative effect to the legal understanding in 77 Bundgaard, ‘The Notion of Beneficial Ownership in Danish Tax Law: The Creation of a new Order with Uncertainty as a Companion’, 113. 78 The ISS Case, SKM 2010.268, previously SKM 2011.121. 79 BFH, 20 August 2008, I R 39/07, BFHE 222, 509. 80 BFH, 19 May 2010, I R 62/09, BFHE 230, 18. 81 As was done in Germany via § 50d para. 11 of the Income Tax Act (EStG). 82 BFH, 15 April 2014, I R 27/12, BFHE 246, 15; BFH, 18 August 2015, I R 88/12, BFHE 251, 190; following on from this, but—in relation to the avoidance of abuse—argumentatively delimiting it: BFH, 2 February 2022, I R 22/20, BStBl II 2022, 324.

386 Dietmar Gosch and Nadia Altenburg the OECD Model Commentaries when interpreting double taxation treaties, but does not understand it to be legally binding.83

22.4 Interim Conclusion While the OECD and the ECJ clearly advocate for an independent international fiscal meaning of the term beneficial ownership based on an increasingly broad economic approach, domestic jurisprudence shows a differentiated pattern. Domestic courts that apply a narrow technical understanding of the term mostly argue that the absence of a clear definition could lead to uncertainties, which is reflected in the increase in court cases dealing with the term’s interpretation. The main challenge appears to be the ominous purpose of the beneficial ownership principle. While it was once clearly applied to allocate income to the rightful recipient, it is now primarily used to address tax abuse without defining the scope of the potential abuse it is meant to tackle. Without a clear definition of the purpose of the principle, however, it is difficult to outline a clear understanding of the principle itself.

22.5 Beneficial Ownership and Anti- Abuse Mechanisms While the foregoing does not give a clear picture of whether beneficial ownership is to be construed as an independent international term or by its use in domestic law, developments in recent years clearly demonstrate that the term is increasingly connected with the prevention and combatting of abusive tax practices. Based on its report on conduit companies, the OECD Commentary clarifies that while: the concept of ‘beneficial owner’ deals with some forms of tax avoidance (i.e. those involving the interposition of a recipient who is obliged to pass on the dividend to someone else), it does not deal with other cases of abuses, such as certain forms of treaty shopping, that are addressed by [other] provisions and principles.84 The

83

BFH, 16 January 2014, I R 30/12, BFHE 244, 354: ‘Opinion of the Fisci involved’ ‘Meinungsbild der beteiligten Fisci’; see D. Gosch, ‘Über die Auslegung von Doppelbesteuerungsabkommen’, Internationale Steuer Rundschau (2013), 87. 84 Referring to art. 29, introduced as a result of the report Preventing the Granting of Treaty Benefits in Inappropriate Circumstances, Action 6—2015 Final Report, and the principles put forward in the section on ‘Improper use of the Convention’ in the Commentary on art. 1 that will apply to prevent abuses, including treaty-shopping situations where the recipient is the beneficial owner of the dividends.

Beneficial Ownership and Tax Treaties 387 beneficial ownership principle, therefore, must not be considered as restricting the application of other approaches to addressing such cases.85

The OECD Commentary explicitly acknowledges that the application of judicial doctrines or principles of interpretation, such as substance over form, economic substance, sham, business purpose, step-transaction, abuse of law, and fraus legis, do not create conflict with the application of a tax convention.86 Despite this, the OECD Commentary remains silent when it comes to determining relations between the beneficial ownership principal and (other) anti-avoidance mechanisms. Based on its design, beneficial ownership is to be understood as a specific anti-abuse instrument,87 even though it cannot be understood as a technical concept that has to be introduced in treaty language and can be applied to a specific case. As with an LOB clause, a tax advantage has to be denied or withdrawn if the main motivation, or one of the main motives, for business transactions is tax avoidance or abuse. Both legal institutions, that of the beneficial owner as well as that of the specific abuse avoidance rule, seem to stand side by side and are mutually subject to a systematic ranking. Seen in this way, the search for economically attributable ownership deserves priority and has to be determined as a first step. Consequently, and from a legal systematic perspective, addressing abuse is only considered in as a subsequent second step: therefore, even a ‘recognized’ beneficial owner could be denied treaty and directive benefits if it is interposed in an abusive manner.88 If an abusive practice cannot be determined, the beneficial owner should be entitled to treaty and directive benefits. In practice, application of the practice by the legal institutions is unfortunately not as distinct. While searching for the ‘true’ beneficial owner, at least indirectly, one might also aim to address abusive arrangements, and this is how the ECJ proceeds when determining the beneficial owner as a legal figure in order to prevent abuse. To a certain extent, economic objectives and anti-abuse objectives overlap at this point, and the beneficial owner is instrumentalized in that regard without respecting the systematic hierarchy. From a methodological point of view, the approach is somewhat reminiscent of the so-called internal theory,89 which comes from Austrian legal theory and aims to 85

OECD CTPA, ‘Clarification of the Meaning of ‘Beneficial Owner’, 4 para. 12.4. OECD MC, art. 1 paras 78ff. 87 Oliver R. Hoor and K. O’Donnel, ‘Update on BEPS Action 6: No Land in Sight’, Tax Notes International (20 Apr. 2015), 265; partially also 1. L. De Broe and J. Luts, ‘BEPS Action 6: Tax Treaty Abuse’, Intertax 43/2 (2015), 122; similarly T. Booker, ‘Beneficial Ownership’, European Taxation (Apr. 2013), 146; J. Bernstein ‘Thoughts on the OECD Discussion Paper on Beneficial Ownership’, Tax Notes International (2011), 49. 88 F. Debelva and J. Luts, ‘European Union—The General Anti-Abuse Rule of the Parent–Subsidiary Directive’, European Taxation 55/6 (2015), 223. 89 P. Fischer, ‘Geltungsanspruch des Steuergesetzes, Steuerumgehung und “wirtschaftliche o- der sonst beachtliche außersteuerliche Gründe”—Zehn Thesen zum Meinungsstreit zwischen “Außen-und Innentheorie” ’, Finanzrundschau (2001), 1212. 86

388 Dietmar Gosch and Nadia Altenburg combat misuse of design by means of a normative legal interpretation of the relevant terms. However, this understanding of the law is not universally recognized. At least the German tax law,90 and probably the tax law of most continental states, does not follow this approach. In theory, as described, there is a strict distinction between the assignment of economic ownership, on the one hand, and anti-abuse measures, on the other hand. Both the allocation of income to the economic owner and the allocation of the ‘tax benefit’ to a person other than the legal or economic owner can converge if an ‘appropriate structure’ has been chosen. The same seems to apply with beneficial owners. From that point of view, the aforementioned practice of the ECJ is in line with EU law in the so-called Denmark decisions91 as it not only reduced anti-treaty shopping as an interpretation aid but also identified an anti-abuse principle at the primary law level as an independent legal institution.92 With this interpretation, the specific anti-abuse provisions (as contained in the relevant guidelines, i.e. the Parent–Subsidiary Directive and the Interest-Licence Directive) remain redundant as they no longer convey any specific significance in relation to the term beneficial owner. Based on this understanding, the idea of a beneficial owner should apply in the same way from a treaty perspective. Similarly to the practice of the ECJ, here too reference can be made to the materials of the OECD and the OECD Model Commentary, according to which the concept of beneficial ownership has to be understood from an economic point of view, on the one hand, but also as a feature for addressing abusive practices, on the other. The recent introduction of a specific anti-abuse rule in article 29 of the OECD MC with the inclusion of a limitation on benefits provision does not contradict such an understanding. Combating abuse does not require a specific form that is firmly established according to the content of specific provisions and legal effects, instead the prevention of abuse is heterogeneous. The parties to a treaty and the domestic legislators are both free to use different measures to address abuse, which can also occur in the form of beneficial ownership. In this context, it would then apply as a special avoidance instrument that prevails over a general anti-avoidance rule (GAAR) as a matter of principle.93

90 Ibid.

91 Ibid.. 92

Schön, ‘Rechtsmissbrauch im europäischen (Steuer)Recht’, 685. is, however, noteworthy that the German legislator does not seem to take this understanding from the treaties themselves, but rather achieves that result by way of a treaty override, especially regarding conduit companies (Sec. 50d para. 3 of the Income Tax Act in the version of the draft of a so- called ‘withholding tax relief modernization law’ [Abzugsteuerentlastungsmodernisierungsgesetz]), namely for cases of mere conduit companies; critically regarding this problem, see G. Frotscher, ‘Der “Beneficial owner”—eine Missbrauchsvermeidungsvorschrift?’, in R. Ismer et al., eds, Festschrift für Moris Lehner (Cologne: Verlag Dr Otto Schmidt, 2019), 237. 93 It

Beneficial Ownership and Tax Treaties 389

22.6 Conclusion This foray has shown that the term beneficial owner is an ambiguous one. It oscillates between a mere rule of attribution, a legal attribution, an attribution according to economic standards, and even anti-abuse prevention. In some cases, attempts are made to deal with the understanding of the term in a treaty-autonomous manner, yet in other— and probably most—cases, interpretation is sought through recourse to national law, the lex fori. Thus, it remains unclear whether the law of the state of residence or the source state is decisive or whether the law of the so-called country of application takes precedence, which in turn can provoke double taxation. The OECD should be urged to create its own definition of the term in the Model Convention.94 As long as this is not the case, the binding orientation towards the legal system of one of the participating states, preferably that of the source state, has to be given preference. Apart from these preliminary questions, the following applies in terms of content: even if the term is charged with an attempt to address abusive arrangements, that does not mean that this legal figure has to comprise all the characteristics that are commonly associated with a GAAR. The opposite is the case: the determination of the ‘true’ beneficial owner is based on economic standards. Beneficial ownership is determined based on indications; subjective elements are not decisive. Motives for the choice of a certain structure are therefore insignificant for the question of beneficial ownership. The situation is different, however, when it comes to the substance requirements of the existence of the business activities of the beneficial owner. Such requirements are part of the economic characteristics that underline economic entitlement and show whether a person can be qualified as a ‘true’ market participant rather than a mere agent or nominee. However, they can only be an indication because a conduit company can also be structured as an entity with economic substance to the extent that a certain flow of payment is routed through it even though no economic value is in fact created and it has no capacity to decide whether such income is forwarded to the recipient.

94 Likewise O. Milanin, Die Bedeutung des OECD- Musterkommentars für die Auslegung von Doppelbesteuerungsabkommen am Beispiel des Nutzungsberechtigten (Beneficial Owner) (Baden- Baden: Nomos, 2021), ch. C.

Chapter 23

The Principa l P u rp o se Test u nde r Tax Treat y L aw Robert J. Danon

23.1 Introduction Further to the Base Erosion and Profit Shifting (BEPS) initiative,1 the Principal Purpose Test (PPT) is the multilateral minimum standard to combat the improper use of double taxation conventions (DTCs) and has been included in article 29(9) of the OECD and UN Model Tax Conventions (MCs). The PPT, which has been much discussed in academic literature,2 reads as follows: 1 This chapter is partly based on R. J. Danon, ‘The Principal Purposes Test Under the OECD and UN Model Tax Conventions: Impact for Taxpayers and Limits for States Under International Law’, in M. Butani and T. Jain, eds, General Anti-Avoidance Rules: The Final Tax Frontier?: Indian and International Perspectives (Toronto: Thomson Reuters, 2021), ch. 32. The ideas expressed in this chapter have also been discussed in R. J. Danon et al., ‘The Prohibition of Abuse of Rights after the ECJ Danish Cases’, Intertax 49/6/7 (2021), 482–516; R. J. Danon, ‘The PPT in Post-BEPS Tax Treaty Law: It Is a GAAR but Just a GAAR!’, Bulletin for International Taxation 74 (2020), 242, R. J. Danon, ‘Treaty Abuse in the Post-BEPS World: Analysis of the Policy Shift and Impact of the Principal Purpose Test for MNE Groups’, Bulletin for International Taxation 72 (2018) 45; R. J. Danon, ‘Intellectual Property (IP) Income and Tax Treaty Abuse: Relevance of BEPS Actions 5 and 8–10 for the Principal Purpose Test’ in G. Maisto, ed., Taxation of Intellectual Property under Domestic Law, EU Law and Tax Treaties (Amsterdam: IBFD, 2018). The author would like to thank Mr Benjamin Malek, PhD candidate on the problem of the improper use of DTCs and research associate at the Tax Policy Center of the University of Lausanne, for his research and editorial help in preparing this chapter. 2 See, among others, A. Baéz Moreno, ‘GAARs and Treaties: From the guiding principle to the Principal Purpose Test. What Have We Gained from BEPS Action 6’, Intertax 45/6/7 (2017), 432–446; V. Chand, ‘The Principal Purpose Test in the Multilateral Convention: An In Depth Analysis’, Intertax 46/1 (2018), 18–44 and ‘The Interaction of the Principal Purpose Test (and the Guiding Principle) with Treaty and Domestic Anti-Avoidance Rules’, Intertax 46/2 (2018), 115–123; L. De Broe, ‘BEPS Action 6: Tax Treaty Abuse’, Intertax 43/2 (2015), 122–146 and ‘Tax Treaty and EU Law aspects of the LOB and

392 Robert J. Danon Notwithstanding the other provisions of this Convention, a benefit under this Convention shall not be granted in respect of an item of income or capital if it is reasonable to conclude, having regard to all relevant facts and circumstances, that obtaining that benefit was one of the principal purposes of any arrangement or transaction that resulted directly or indirectly in that benefit, unless it is established that granting that benefit in these circumstances would be in accordance with the object and purpose of the relevant provisions of this Convention.

The architecture of the PPT is based on a so-called ‘two-pronged test’. That is, where ‘one of the principal purposes’ of an arrangement is to obtain treaty benefits (subjective prong), these benefits must still be granted where it is established that granting the latter is ‘in accordance with the object and purpose of the relevant provisions of this Convention’ (objective prong). In other words, the question of whether tax treaty benefits ought to be granted is ultimately settled based on a sui generis purposive interpretation of tax treaty law.3 The architecture of the PPT is not new as it is directly inspired by the so-called ‘guiding principle’ inserted in the OECD Commentaries in 2003.4 The prohibition of abuse developed by the Court of Justice of the European Union (CJEU) and recently introduced as a general anti-avoidance rule (GAAR) in European direct tax directives is equally based on such a two-pronged analysis.5 On the other hand, it is fair to say that, until recently, there was no prominent and settled tax treaty case law which could serve as guidance for the interpretation of the PPT. This has now changed with the landmark judgment delivered by the Supreme Court of Canada in Alta Energy on 26 November 2021.6 There are a number of reasons why this judgment is of global importance and PPT provision proposed by BEPS Action 6’, in R. J. Danon, ed., Base Erosion and Profit Shifting (BEPS)— Impact for European and International Tax Policy (Zurich: Schulthess, 2016), 213; D. G. Duff, ‘Tax Treaty Abuse and the Principal Purpose Test—Part I and Part II’, Canadian Tax Journal 66/3/4 (2018); C. Elliffe, ‘The Meaning of the Principal Purpose Test: One Ring to Bind Them All?’, World Tax Journal 11 (2019), 1; B. Kuźniacki, ‘The Principal Purpose Test (PPT) in BEPS Action 6 and the MLI: Exploring Challenges Arising from Its Legal Implementation and Practical Application’, World Tax Journal 10/2 (2018); M. Lang, ‘BEPS Action 6: Introducing an Anti-abuse Rule in Tax Treaties’, Tax Notes International 74/7 (2014), 655; S. van Weeghel, ‘A Deconstruction of the Principal Purposes Test’, World Tax Journal 11/1 (2019); W. Schön, ‘The Role of “Commercial Reasons” and “Economic Reality” in the “Principle Purpose Test” under Art. 29(9) OECD Model Tax Convention 2017’, Working Paper of the Max Planck Institute for Tax Law and Public Finance No. 2022-03 (7 Feb. 2022), available at https://papers.ssrn.com/sol3/pap ers.cfm?abstract_id=4029573; A. Martín Jiménez, ‘Is There an International Minimum Standard on Tax Treaty Shopping after BEPS Action 6? Some Recent Divergent Trends’, World Tax Journal 14/3 (2022). 3

This purposive interpretation is sui generis in the sense that it is different, has a scope of its own, and comes after the regular treaty interpretation based on art. 31 VCLT, see Section 23.2. 4 2003 OECD Model Commentary on art. 1, para. 9.5. The guiding principle continues to operate as a limitation to tax treaty benefits under the 2017 OECD Commentaries, see 2017 OECD Model Commentary on art. 1, para. 61. On the relation between the interpretative guiding principle and the PPT, see Section 23.2. 5 On the convergence between the PPT and the prohibition of abuse of rights under EU law, see Danon et al, The Prohibition of Abuse of Rights, 482ff. 6 Canada v. Alta Energy Luxembourg SARL [2021] SCC 49, 24 ITLR 346. This Supreme Court judgment upholds the decisions of two lower courts to which I shall also refer in this chapter (Alta

The Principal Purpose Test under Tax Treaty Law 393 of direct relevance to the interpretation of the PPT. First, as will be discussed in this chapter, the case involved the application of the Canadian GAAR which includes a two- pronged test which is substantially identical to that of the PPT. Secondly, the Supreme Court considered the most important aspect of the two-pronged analysis, namely whether a particular arrangement or transaction is in accordance with the object and purpose of the provisions of a DTC. Thirdly, because the case ultimately boiled down to the application of a treaty provision based on article 13(5) OECD MC, the Supreme Court could clarify in a straightforward way the roots and limits of the most important condition to access tax treaty benefits: treaty residence (art. 4 OECD MC). Finally, in support of its conclusion, the Supreme Court also referred to the principle of predictability and fairness as a limit to the application of a GAAR. It will be submitted that in a tax treaty context and for purposes of the PPT, this is in accordance with the principle of systemic integration (art. 31(3)(c) Vienna Convention on the Law of Treaties, VCLT) which includes the principle of legal certainty as a general principle of law (art. 38(1) Statute of the International Court of Justice, ICJ Statute)) forming part of the interpretive exercise. Against this background, the objective of this chapter is twofold. I begin by discussing the subjective (Section 23.3) and objective (Section 23.4) prongs of the PPT. I contrast, in particular, the OECD commentaries with existing tax treaty case law. Secondly, I consider the impact of Alta Energy on the interpretation of the objective prong of the PPT (Section 23.5). However, prior to moving to the core of this chapter, a preliminary question must be settled: what is the difference between a purposive interpretation under the PPT and under article 31 VCLT (Section 23.2)? I will argue that the difference is significant:7 under article 31 VCLT, any purposive interpretation is limited by the wording of the treaty. Under the PPT, which only comes into play after the proper interpretation of treaty law pursuant to article 31 VCLT, this is by contrast not the case. From a conceptual point of view, this question is important in order to properly grasp the impact of the PPT in post-BEPS DTCs as well as to understand what has changed compared to pre-BEPS DTCs which do not include such a rule. This distinction is relevant regarding ongoing treaty disputes involving pre-BEPS DTCs.

Energy Luxembourg SARL v. R, 21 ITLR 219, 239–240 and Alta Energy Luxembourg SARL v. R, 2020 FCA 43, 22 ITLR 509). 7 Contra:

M. Lang, ‘The Signalling Function of Article 29(9) of the OECD Model: The “Principal Purpose Test” ’, Bulletin for International Taxation 74/4/5 (2020), 264–268, who sees the PPT as merely clarifying the purposive interpretation under art. 31(1) VCLT. As will be discussed, I on the other hand consider that the PPT has a scope of its own which is distinct and comes after the interpretation of treaty law and permits a sui generis purposive interpretation beyond the treaty wording. Such sui generis purposive interpretation, however, only comes into play in the presence of an avoidance transaction (i.e. when the subjective element of the PPT is met).

394 Robert J. Danon

23.2 Purposive Interpretation under Article 31 VCLT and the PPT The PPT has been codified in article 29(9) OECD and UN MCs. It thus flows from this systematic insertion that the PPT only comes into play after a proper establishment of the facts relevant to the case at hand8 and the interpretation of treaty law pursuant to articles 31ff VCLT. For example, under a regular interpretation of treaty law, it must be first determined whether a person qualifies as a resident (art. 4 OECD MC) and whether the requirements imposed by the relevant distributive rule (e.g. the beneficial ownership limitation9 in the case of arts 10–12 OECD MC) are satisfied. If the outcome of the interpretation of treaty law reveals that treaty benefits should be granted, then a possible denial of these benefits pursuant to article 29(9) OECD and UN MCs finally comes into play. At the level of the interpretation of treaty law, the new question that has emerged is the impact of the new preamble to the OECD and UN MCs in this framework.10 According to article 31 VCLT, a treaty ‘shall be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose’. It is established that article 31(1) VCLT expresses the primacy of the textual approach in the interpretative process.11 This implies first of all that the principle of effectiveness must be observed.12 Therefore, any interpretation that would render a treaty provision superfluous or diminish its practical effects is to be avoided.13 For example, overlooking the beneficial ownership requirement in the interpretation of provisions modelled pursuant to articles 10–12 OECD and UN MCs would not be compatible with the principle of effectiveness. By mirrored reasoning, reading an additional beneficial ownership limitation in a distributive rule which does not incorporate such a requirement (e.g. art. 13(5) OECD MC and art. 13(6) UN MC relating 8 Technically, therefore, the PPT is not necessary to deny tax treaty benefits where the arrangement or transaction put in place is a sham or, to use civil law terminology, a simulated operation. This is because a sham or simulated arrangement may, in principle, already be disregarded at the level of the establishment of the facts. In practice, however, the distinction between sham and abuse is often blurred. 9 For a detailed discussion of the beneficial ownership limitation in tax treaty practice, see inter alia R. Danon, ‘The Beneficial Ownership Limitation in Articles 10, 11 and 12 OECD Model and Conduit Companies in Pre-and Post-BEPS Tax Treaty Policy: Do We (Still) Need It?’, in G. Maisto, ed., Current Tax Treaty Issues: 50th Anniversary of the International Tax Group (Amsterdam: IBFD, 2020), 585–661 and A. Martín Jiménez, ‘Beneficial Ownership’, in R. Vann, ed., Global Tax Treaty Commentaries (IBFD online version of July 2022). 10 Danon, ‘GAAR’, 251–254. 11 F. Engelen, Interpretation of Tax Treaties under International Law (Amsterdam: IBFD, 2004), 426; K. Vogel and A. Rust, ‘Introduction’, in Klaus Vogel on Double Taxation Conventions (The Hague: Kluwer Law Taxation Publishers, 1991), n. 86. 12 O. Dörr, ‘Article 31 VCLT’, in O. Dörr and K. Schmalenbach, eds, Vienna Convention on the Law of Treaties: A Commentary (Berlin: Springer, 2012), para. 52. 13 Ibid.

The Principal Purpose Test under Tax Treaty Law 395 to capital gains) is in breach of article 31 VCLT. At the same time, however, article 31 VLCT does not favour a purely grammatical interpretation of tax treaty law. Rather, the interpretation must be holistic and contextual.14 For this reason, treaty terms should be construed ‘in their context’.15 In other words, ‘the reality of the methods of interpretation of tax conventions should be taken into account’.16 According to article 31(2) VCLT, the context includes the preamble to the treaty. According to the new preamble to the OECD and UN MCs, contracting states now intend to eliminate ‘double taxation . . . without creating opportunities for non-taxation or reduced taxation through tax evasion or avoidance (including through treaty-shopping arrangements aimed at obtaining reliefs provided in this Convention for the indirect benefit of residents of third States)’. Undoubtedly, the new preamble is context according to article 31(2) VCLT.17 It is unclear, however, whether, and if so to what extent, the new preamble could—on a stand-alone basis—influence the interpretation of treaty provisions. The 2017 OECD and UN Commentaries do not provide any guidance in this respect. Rather, the preamble is mainly mentioned in the framework of article 29 OECD MC, which is deemed to reflect ‘the intention of the Contracting States, incorporated in the preamble of the Convention’.18 This begs two questions. The first question is whether the new preamble has a sufficient level of normative density to really make a difference on its own in the interpretation exercise. In existing treaty practice, it is fair to say that preambles have had little impact, notably because the DTCs reviewed by courts did not include reference to tax avoidance but only to tax evasion.19 An exception is perhaps the Verdannet case in which the French Conseil 14

M. E. Villiger, Commentary on the 1969 Vienna Convention on the Law of Treaties (Leiden: Martinus Nijhoff, 2009), para. 10 on art. 31 VCLT; J. F. Avery Jones, ‘Global Treaty Interpretation’, in Global Tax Treaty Commentaries (Amsterdam: IBFD, 2018), s. 3.4.10 and, specifically as regards the problem of improper use of DTCs: V. Lowe, ‘How Domestic Anti-Avoidance Rules Affect Double Taxation Conventions’, Proceedings of a Seminar held in Toronto, Canada, in 1994 during the 48th Congress of the International Fiscal Association, vol. 19c (The Hague: Kluwer Law International, 1995), 7; J. F. Avery Jones, GTTC ad Interpretation (2018), s. 3.4.10. 15 Article 31(1) VCLT. 16 Re Verdannet, 20 ITLR 832, 862. 17 See OECD Model, Introduction, para. 16 which states the obvious ‘since the title and preamble form part of the context of the Convention and constitute a general statement of the object and purpose of the Convention, they should play an important role in the interpretation of the provisions of the Convention’. 18 2017 OECD Model Commentary on art. 29, para. 1; Danon, ‘GAAR’, 252ff. 19 See, e.g., in Australia, Lamesa Holding BV v Federal Commissioner of Taxation [1997] FCA 134, 35 ATR 239: ‘The Agreement is an agreement for the avoidance of double taxation and the prevention of fiscal evasion with respect to taxes on income. Although, therefore, the Agreement has this dual object, the Agreement substantially concerns allocation of taxing power’ and ‘Save as to its operation to allocate taxing power, the Agreement is little concerned directly with fiscal evasion. However, Article 25 provides for an exchange of information between the competent authorities of each State, which exchange is vital to countering fiscal evasion’ and in Canada, Alta Energy Luxembourg SARL v. R, 21 ITLR 219, para. 77: ‘A tax treaty is a multi-purpose legal instrument. The preamble of the Treaty states that the two governments desired ‘to conclude a Convention for the avoidance of double taxation and the prevention of fiscal evasion with respect to taxes on income and on capital.’ While indicative of the general purpose of the treaty, this statement remains vague regarding the application of specific articles of the treaty.

396 Robert J. Danon d’État referred to the broad objective of the France–Luxembourg DTC and considered that:20 The States that are parties to the Franco-Luxembourg tax treaty cannot be regarded as admitting, in the distribution of the power of taxation, the application of its provisions to situations arising from artificial transactions devoid of any economic substance. It follows that in finding that the operation in question was contrary to the objectives pursued by the two signatory States, the Court did not commit any error of law in its judgment.

Verdannet concerned a clear round-tripping case (i.e. a transfer of shares by a French taxpayer to a Luxembourg entity which in turn sold those shares to a company controlled by the French taxpayer’s ex-wife). In this instance, it was found that the very essence of the DTC had been defeated. In the words of the rapporteur public: ‘the primary function of these treaties, beyond this immediate purpose, is to facilitate international economic exchanges. . . . It is, therefore, part of their very logic that they be read as not intending to apply to taxpayers who artificially create the conditions of foreignness allowing them to claim, according to a literal interpretation, the benefit of their clauses’.21 Whether, however, the new preamble to the OECD MC will on its own make a difference in treaty practice can be left open at this stage. In my opinion, the real question lies in the limits to the contextual and purposive interpretation under article 31 VCLT. It is indeed settled that the contextual and purposive interpretation finds its limits in the text of the DTC and, therefore, may not lead to a result that undermines the actual terms of the agreement.22 At the same time, however, the answer depends on the interpretation of article 31 VCLT that is favoured, on the one hand, and on the language of the relevant treaty rule, on the other hand. Several scholars23 around the globe have considered that a proper interpretation of treaty law under article 31 VLCT allows, in certain instances, treaty benefits to be denied 20

Re Verdannet, 20 ITLR 832, 856–857.

21 Ibid.

22 Engelen, Interpretation of Tax Treaties, 429.

23 Vogel was one of the first scholars to express this view (see, Introduction, para. 121 in Klaus Vogel on Double Taxation Conventions: A Commentary to the OECD-, UN-and US Model Conventions for the Avoidance of Double Taxation of Income and Capital with Particular Reference to German Treaty Practice (The Hague: Kluwer Law Taxation, 1991), which was then endorsed by several commentators: D. A. Ward, ‘Abuse of Tax Treaties’, Intertax 23 (1995), 180; R. Prokisch, ‘Artikel 1’, in K. Vogel and M. Lehner, eds., Doppelbesteuerungsabkommen (DBA), Kommentarder Bundesrepublik Deutschland auf dem Gebiet der Steuern vom Einkommen und Vermögen. Auf der Grundlage der Musterabkommen (Munich: Beck, 2008), para. 117; F. Engelen, On Values and Norms: The Principle of Good Faith in the Law of Treaties and the Law of Tax Treaties in Particular (The Hague: Kluwer Law International, 2006), 36; L. De Broe, International Tax Planning and Prevention of Abuse: A Study Under Domestic Tax Law, Tax Treaties and EC Law in Relation to Conduit and Base Companies (Amsterdam: IBFD, 2008), 374–375; including the present author (R. Danon, ‘Article 1’, in R. Danon, X. Oberson, P. Pistone, and D. Gutmann, Modèle de Convention fiscale OCDE concernant le revenu et la fortune: Commentaire détaillé (Basel: Helbing Lichtenhahn, 2014), para. 144.

The Principal Purpose Test under Tax Treaty Law 397 in cases of abuse. Since the 2003 update of its Commentary, the OECD also supports the view that treaty benefits may be denied where this results: ‘from the object and purpose of tax conventions as well as the obligation to interpret them in good faith (see Article 31 of the Vienna Convention on the Law of Treaties)’ (interpretative approach).24 As is well known, this approach is based on the so-called ‘guiding principle’.25 Courts have also relied on the 2003 Commentaries to deny treaty benefits, the most emblematic example being the Swiss Federal Tribunal in A Holding ApS26 and the Israeli District Court of Tel-Aviv in Yanko-Weiss.27 In Vodafone, by contrast, the Indian Supreme Court chose not to rely on the reasoning favoured in A Holding ApS.28 It is, however, submitted that the approach advocated by the Swiss and Israeli courts in A Holding ApS and in Yanko- Weiss can be reconciled with the sham or tax avoidance threshold used by the court in Vodafone.29 In A Holding ApS,30 the entity interposed in Denmark and receiving Swiss source dividends was so artificial that it was akin to a sham arrangement as understood in Vodafone. On the facts, it was in particular clear that it was used to channel treaty- favoured income to its offshore shareholders and its own corporate interests had been grossly bypassed.31 This artificiality was also present in the restructuring in Yanko-Weiss Holdings. In this case, a company originally incorporated in Israel migrated to Belgium and claimed the benefits of the Israel–Belgium DTC.32 From this perspective, the facts in A Holding ApS and Yanko-Weiss were thus different from the ones at issue in Vodafone and decided in favour of the taxpayer. That being said, the 2017 OECD and UN Commentaries suggest that the PPT merely confirms the guiding principle33 which could also apply independently.34 Therefore, the question arises as to whether, at the level of the interpretation of treaty law pursuant to article 31 VCLT, the guiding principle could have a wider scope (i.e. a scope equivalent to that of the PPT). The issue is relevant in practice as regards disputes involving DTCs not yet incorporating the PPT. In my view, not being an express treaty GAAR, the scope of the guiding principle is undoubtedly more limited than that of the PPT. This is because the argument here is made on the basis of article 31 VCLT and not pursuant to a genuine treaty GAAR. This is implicitly recognized by the OECD Commentaries which when

24

2017 OECD Model Commentary on art. 29, para. 59. Ibid, para. 60. 26 A Holding ApS v. Federal Tax Administration, 8 ITLR 536. 27 Yanko-Weiss Holdings 1 (1996) Ltd v. Holon Assessing Office, 10 ITLR 524. 28 Vodafone International Holdings BV v. Union of India and another, 14 ITLR 431, 514. This reasoning was also not favoured in MIL (Investments) SA v. Canada, 9 ITLR 29, Tax Court of Canada and Federal Court of Appeal, 9 ITLR 1111. 29 Vodafone International Holdings BV v. Union of India and another, 14 ITLR 431, 451. 30 A Holding ApS v. Federal Tax Administration, 8 ITLR 536. 31 Ibid. 32 Yanko-Weiss Holdings 1 (1996) Ltd v. Holon Assessing Office, 10 ITLR 524. 33 2017 OECD Model Commentary on art. 29, para. 61. 34 Ibid. 25

398 Robert J. Danon justifying the interpretative approach supporting the application of the guiding principle precisely rely on article 31 VCLT.35 Hence, as rightly observed by commentators, a denial of treaty benefits on the basis of article 31 VCLT may only come into play in the presence of a wholly artificial arrangement which completely defeats the fundamental purposes of DTCs (i.e. promoting the free movement of goods, persons, services, and capital between the contracting states by avoiding international double taxation is frustrated).36 This would be the case, for example, if the interposition of an entity in the state of residence does not (or very marginally) contribute to any international activity.37 It would appear to me that these extreme circumstances would be satisfied in the case of a circular artificial round-tripping structure which, by definition, involves no real international activity. I therefore submit that the guiding principle alone does not support the denial of treaty benefits beyond wholly artificial arrangements or round-tripping cases. There is a further argument in favour of such a restrictive interpretation: neither the 2003 OECD nor the 2011 UN Commentaries contain guidance or examples illustrating the scope and practical application of the guiding principle. Therefore, irrespective of the limits imposed by article 31 VCLT with respect to the purposive interpretation, a wider application of the guiding principle by a contracting state would, from this perspective, be in breach of the principle of legitimate expectations imposed by the principle of good faith.38 Returning to the new preamble to the OECD MC, the conclusion may be drawn that the latter will in most instances not have any impact of its own at the level of interpretation of treaty law pursuant to article 31 VCLT. This is primarily because the contextual interpretation may not produce a result that is not supported by the actual wording of the agreement. As one commentator correctly submits, this in particular holds true as regards treaty residence under article 4 OECD MC which sets the required nexus with the state of residence to access treaty benefits.39 An exception could, however, be foreseen in cases in which the treaty terms are rather open-ended and could support an interpretation geared towards the preamble (e.g. beneficial ownership40). A different question which I shall pick up when discussing the impact of Alta Energy on the interpretation of the PPT is whether, and if so to what extent, the findings of the Supreme Court would have (or should have been) different if the DTC in place had included the preamble to the 2017 OECD MC. The position which will be argued in this chapter is simple: the answer is no.41 35

Ibid., para. 59.

36 Engelen, On Values and Norms, 36; De Broe, Prevention of Abuse, 375. 37

De Broe, Prevention of Abuse, 375. Art. 26 VCLT. 39 J. Gooijer, Tax Treaty Residence of Entities (Alphen aan den Rijn: Kluwer, 2019), 77: ‘the purpose of the OECD MC to prevent tax avoidance cannot be of any help in the interpretation of the definition of resident in Article 4(1), if the provision itself does not contain any reference to the motives behind the establishment of residency in a particular case’. 40 See thereon Danon, ‘Beneficial Ownership’, s. 15.3.1. 41 See Section 23.5. 38

The Principal Purpose Test under Tax Treaty Law 399

23.3 The Subjective Prong of the PPT: ‘One of the Principal Purposes’ Moving to the core of this chapter, I begin with the subjective prong of the PPT. Article 29(9) OECD and UN MCs provide that a denial of treaty benefits by the source state may come into play where ‘one of the principal purposes’ of the arrangement or transaction was to obtain those benefits. In accordance with the ‘reasonableness’ test of the PPT, it is thus not necessary to find conclusive proof of the intent of a person concerned with an arrangement or transaction.42 The purpose of this requirement is:43 to ensure that tax conventions apply in accordance with the purpose for which they were entered into, i.e. to provide benefits in respect of bona fide exchanges of goods and services, and movements of capital and persons as opposed to arrangements whose principal objective is to secure a more favourable tax treatment.

According to the Commentaries, where an arrangement can only be reasonably explained by a benefit that arises under a treaty, it may be concluded that one of the principal purposes of that arrangement was to obtain the benefit.44 Moreover, the reference to ‘one of the principal purposes’ means that obtaining the benefit under a tax convention need not be the sole or dominant purpose of a particular arrangement or transaction.45 The Commentaries illustrate this with the following example concerning article 13(5) OECD MC:46 a person may sell a property for various reasons, but if before the sale, that person becomes a resident of one of the contracting States and one of the principal purposes for doing so is to obtain a benefit under a tax convention, art. 29(9) OECD MC could apply notwithstanding the fact that there may also be other principal purposes for changing residence, such as facilitating the sale of the property or the re-investment of the proceeds of the alienation.

In the area of capital gains, in particular, I would submit that indicators developed by international leading case law may be relied upon in order to determine whether the subjective element of the PPT is satisfied. A good example is the findings of the Indian Supreme Court in the Vodafone case. The Supreme Court considered that the distinction between ‘a preordained transaction which is created for tax avoidance purposes’ and ‘a transaction which evidences investment to participate in India’47 should be 42

2017 OECD Model Commentary on art. 29, para. 178. Ibid., para. 174. 44 Ibid., para. 178. 45 Ibid., para. 180. 46 Ibid. 47 Vodafone International Holdings BV v. Union of India and another, 14 ITLR 431, 456. 43

400 Robert J. Danon settled by looking at a transaction in a ‘holistic manner’ and on the basis of the following indicators: (1) the duration of time during which the holding structure was in place; (2) the period of business operation in India; (3) the generation of taxable revenue in India; (4) the timing of the exit; and (5) the continuity of the business in India after the exit.48 In this area, the close time lapse between the realization of the capital gain and the restructuring (transfer of residence, migration, incorporation of a new entity) is indeed particularly relevant. This is precisely why in the Alta Energy case, to which I revert later, the taxpayer had conceded that the restructuring in Luxembourg was an avoidance transaction (i.e. had not been ‘arranged primarily for a bona fide purpose other than to obtain a tax benefit’).49 This aspect was also present in the Verdannet case which concerned a circular share transfer also involving the incorporation of an entity in Luxembourg:50 the chronology was confounding: the first act of purchase of the real estate involved Mr Verdannet personally, who created the Luxembourg company on the same date, with which he was replaced a few weeks later and which ultimately purchased the property less than six months later, without ever carrying out thereafter the real estate business for which its corporate purpose had been extended, a posteriori, for the sole needs of the cause. We therefore believe that the existence of an artificial arrangement, as the court has judged, is not seriously questionable.

I submit that these decisions provide good indicators which could be relied upon when considering the subjective element of the PPT. However, a controversial point is whether the level of nexus or ‘substance’ that the taxpayer has with the state of residence matters in the PPT analysis. For some commentators, this element is relevant to both the subjective and objective components of the PPT,51 while other scholars tend to connect it to the latter.52 There are indeed numerous references in the OECD Commentary suggesting that economic substance plays a role in the PPT analysis. For instance, in an example involving a regional company providing intra-group services, the Commentaries note that:53 48

Ibid., 456–457. Alta Energy Luxembourg SARL v. R, 21 ITLR 219, 235. 50 Re Verdannet, 20 ITLR 832, 870. 51 B. Kuzniacki, ‘The Principal Purpose Test (PPT) in BEPS Action 6 and the MLI: Exploring Challenges Arising from Its Legal Implementation and Practical Application’, World Tax Journal 10 (2018). 52 van Weeghel, ‘A Deconstruction of the Principal Purposes Test’, 38ff. 53 2017 OECD Model Commentary on art. 29, para. 182, example G. See also para. 182, example K, involving an institutional investor: 49

The decision to establish the regional investment platform in State R was mainly driven by the availability of directors with knowledge of regional business practices and regulations, the existence of a skilled multilingual workforce, ... RCO employs an experienced local management team to review investment recommendations from Fund and performs various other functions which, depending on the case, may include approving and monitoring investments, carrying on

The Principal Purpose Test under Tax Treaty Law 401 Assuming that the intra-group services to be provided by RCO, including the making of decisions necessary for the conduct of its business, constitute a real business through which RCO exercises substantive economic functions, using real assets and assuming real risks, and that business is carried on by RCO through its own personnel located in State R, it would not be reasonable to deny the benefits of the treaties concluded between State R and the five States where the subsidiaries operate unless other facts would indicate that RCO has been established for other tax purposes or unless RCO enters into specific transactions to which paragraph 9 would otherwise apply.

This has led scholarly writing to question whether the PPT incorporates a substance- oriented requirement or to what extent there would be a connection between the PPT and other BEPS actions items, such as BEPS Actions 8–10 relating to transfer pricing or BEPS Action 5 concerning intellectual property (IP) regimes and the modified nexus approach.54 As I have already argued elsewhere,55 it follows from the Commentaries that the presence of significant functions and substance in the state of residence are important elements to evidence that one of the principal purposes of an arrangement (i.e. particularly the creation and maintenance of an entity in the state of residence, respectively to transfer to such entity the rights giving rise to income) is not to obtain the benefits of the treaty concluded by the state of residence with the state of source. From this perspective, I have submitted that in the area of IP income, the transfer pricing principles flowing from BEPS Actions 8–10 could, for instance, serve as guidance. The same holds true as regards IP income falling into the scope of the modified nexus approach applying to IP regimes. In such case and subject to a minor uplift of 30%, IP income is indeed deemed to be linked to substantial R&D activities exercised in the state of residence. In fact, from the perspective of tax certainty and administrability, the notion of substance under the modified nexus approach and BEPS Action 556 is even easier to apply than a transfer pricing analysis under BEPS Actions 8–10 which remains more subjective. At the same time, however, it should be borne in mind that the PPT is not a specific anti-avoidance rule (SAAR) focusing mechanically and solely on the transfer pricing

treasury functions, maintaining RCO’s books and records, and ensuring compliance with regulatory requirements in States where it invests. 54 A. Martin Jiménez, ‘Tax Avoidance and Aggressive Tax Planning as an International Standard— BEPS and the “New” Standards of (Legal and Illegal) Tax Avoidance’, in A. P. Dourado, ed., Tax Avoidance Revisited in the EU BEPS Context (Amsterdam: IBFD, 2017), 25ff. 55 Danon, ‘Treaty Abuse’, 48–49; Danon, ‘Intellectual Property’, 25. 56 I would certainly agree with van Weeghel, ‘A Deconstruction of the Principal Purposes Test’, 39ff, that the modified nexus approach could only serve as an indicator in the case of IP regimes. For this reason, this commentator has, in particular, argued in favour of relying on the ‘active business test’ of LOB clauses. However, in the area of R&D functions, the modified nexus approach provides for a much more detailed framework (distinguishing in particular between acquired IP, outsourcing, and in-house R&D) than the active business test which is much more rudimentary and seems to assume that R&D functions are always performed in one state (see 2107 OECD Model Commentary on art. 29, para. 74.

402 Robert J. Danon functions exercised in the state of residence to grant or deny treaty benefits. Such an interpretation of article 29(9) OECD MC would clearly not be supported by its clear wording. Rather, all circumstances surrounding the arrangement or event must always be considered on a case-by-case basis.57 Therefore, the level of substance in the state of residence should only be understood as a possible proxy to exclude the subjective element of the PPT. The Commentaries note indeed that:58 A purpose will not be a principal purpose when it is reasonable to conclude . . . that obtaining the benefit was not a principal consideration and would not have justified entering into any arrangement or transaction that has, alone or together with other transactions, resulted in the benefit. In particular, where an arrangement is inextricably linked to a core commercial activity, and its form has not been driven by considerations of obtaining a benefit, it is unlikely that its principal purpose will be considered to be to obtain that benefit.

Therefore, the absence of significant transfer pricing functions in the state of residence should not automatically lead to the denial of treaty benefits if other compelling factual elements reveal that the subjective element of the PPT is not satisfied (e.g. because the arrangement put in place is predominantly based on commercial and non-tax reasons). Example E of the Commentaries relating to conduit cases illustrates this. In this example, a holding company only keeps a small spread in a licensing and sub-licensing structure (thereby exercising very few functions in the state of residence in respect of that income and presumably having little organizational substance due to its purpose as a holding company) but the subjective element is not satisfied because the entity is ‘conforming to the standard commercial organization and behaviour of the group’.59 From this perspective, as I have argued, the approach taken under the PPT rule is different from a broad substance-oriented but objective (i.e. without taking into the motives of the taxpayer) interpretation of beneficial ownership favoured by several jurisdictions. Therefore, coordination and policy problems could emerge if tax treaty benefits are denied at the level of the distributive rules (for lack of beneficial ownership), whereas this would not have been the case under the PPT.60 That being said, even where the subjective element of the PPT is satisfied, treaty benefits should not be automatically denied. In accordance with the escape clause of article 29(9) OECD and UN MCs, the taxpayer should indeed still be given the opportunity 57

Ibid., para. 178. Ibid., para. 181. 59 Ibid., para. 187, example E. See also UK–US Income Tax Treaty, art. 3(1)(n), example 5 from which this example of the OECD Commentary is derived: 58

even though the specific fact pattern, as presented, meets the first part of the definition of a ‘conduit arrangement’ . . . on balance the conclusion would be that ‘the main purpose or one of the main purposes’ of the transactions was not the obtaining of UK/US treaty benefits. So the structure would not constitute a conduit arrangement. 60

See Danon, ‘Beneficial Ownership’, 628ff.

The Principal Purpose Test under Tax Treaty Law 403 to establish that granting treaty benefits would be in accordance with the ‘object and purpose of the relevant provisions of this Convention’. A similar situation occurred in the Canadian Alta Energy case to which I revert later: as part of a restructuring, a company was formed in Luxembourg in 2012 and shares in a Canadian corporation were transferred to it. In 2013, the Luxembourg company sold its Canadian shareholding, realized an important capital gain, and claimed that this capital gain was exclusively taxable in Luxembourg in accordance with article 13(5) of the Canada–Luxembourg DTC.61 The restructuring qualified as an avoidance transaction under the Canadian GAAR because the taxpayer had conceded that the use of the Luxembourg entity had not been ‘arranged primarily for a bona fide purpose other than to obtain a tax benefit’.62 However, the taxpayer won—before the Tax Court63, the Federal Court of Appeal,64 and the Supreme Court65—because the Crown had failed to demonstrate that the object and purpose of the provisions at stake (arts 1, 4, and 13 DTC) had been frustrated. Therefore, the transaction was not regarded as abusive.66 In the same vein, in the foregoing example given by the Commentaries, treaty benefits would still have to be granted if the taxpayer manages to demonstrate that the object and purpose of articles 4 and 13 OECD MC are not defeated. As shall now be seen, the Commentaries on the MCs provide little concrete guidance concerning the interpretation of this second, but nonetheless very important, element. Moreover, the shift of the burden proof to the taxpayer is debatable.

23.4 The Objective Prong of the PPT: ‘Object and Purpose of the Relevant Provisions of this Convention’ As discussed, where the subjective element of the PPT is satisfied, treaty benefits are denied ‘unless it is established that granting that benefit in these circumstances would be in accordance with the object and purpose of the relevant provisions of this Convention’.67 In effect, therefore, the onus of proof is controversially shifted to the taxpayer (‘unless it is established’).68 From a policy perspective, it is questionable whether this approach is correct. As pointed out, for instance, by Canadian case law:69 61

Alta Energy Luxembourg SARL v. R, 21 ITLR 219; Alta Energy Luxembourg SARL v. R, 22 ITLR 509. Alta Energy Luxembourg SARL v. R, 21 ITLR 219, 235. 63 Ibid. 64 Alta Energy Luxembourg SARL v. R, 22 ITLR 509. 65 Canada v. Alta Energy Luxembourg SARL [2021] SCC 49, 24 ITLR 346. 66 See later. 67 Art. 29(9) OECD and UN MCs. 68 See De Broe, ‘LOB and PPT’, 213; Baéz Moreno, ‘GAARs and Treaties’, 435. 69 Canada Trustco Mortgage Co. v. Canada, 2005 SCC 54, [2005] 2 SCR 601, para. 39. 62

404 Robert J. Danon The taxpayer, once he or she has shown compliance with the wording of a provision, should not be required to disprove that he or she has thereby violated the object, spirit or purpose of the provision. It is for the Minister who seeks to rely on the GAAR to identify the object, spirit or purpose of the provisions that are claimed to have been frustrated or defeated, when the provisions of the Act are interpreted in a textual, contextual and purposive manner. The Minister is in a better position than the taxpayer to make submissions on legislative intent with a view to interpreting the provisions harmoniously within the broader statutory scheme that is relevant to the transaction at issue.

For this reason, under the Canadian GAAR it is, by contrast, the tax authorities that are to establish that a benefit that a taxpayer would otherwise obtain would contravene the object and purpose of the relevant provisions.70 Nonetheless, based on a proper interpretation of the escape clause of the PPT, the following observations may be formulated. First, the ‘relevant provisions’ are those on which the requested treaty benefits are based. This will typically be the treaty residence article (art. 4 OECD MC) in combination with the applicable distributive rule (arts 6–22 OECD MC) and, as the case may be, the relief provisions (art. 23A or 23B OECD MC). Secondly, within the framework of article 29(9) OECD and UN MCs, the purposive interpretation is not limited by the wording of the applicable provisions. The Commentaries state that article 29(9) OECD and UN MCs: ‘must be read in the context of paragraphs 1 to 7 and of the rest of the Convention, including its preamble. This is particularly important for the purposes of determining the object and purpose of the relevant provisions of the Convention’.71 However, because the interpretation of article 29(9) OECD and UN MCs is itself subject to the principles embodied in article 31 VCLT, the access to or denial of treaty benefits under the escape clause must be rooted in the object and purpose of the relevant provisions and not simply in the object and purpose of the convention in general.72 Such latter reading of article 29(9) OECD and UN MCs would indeed be erroneous and conflict with the clear wording of the provision. This conclusion also makes sense from a policy perspective. While the general objective of the 2017 OECD and UN MCs, as reflected in their preamble, is ‘the elimination of double taxation without creating opportunities for non-taxation or reduced taxation through tax evasion or avoidance (including through treaty-shopping arrangements aimed at obtaining reliefs provided in this Convention for the indirect benefit of residents of third States)’73 and, of course, treaty provisions (in particular, the treaty residence article and the distributives rules) put this objective into effect, access to tax treaty benefits is not, however, granted under the same conditions by all provisions. I indeed submit that some rules are subject to particular conditions (e.g. a beneficial ownership limitation or a holding threshold and period in the case of dividends74), while others are not (e.g. 70 Duff, ‘Tax Treaty’, 986. 71

2017 OECD Model Commentary on art. 29, para. 173, example E. Danon, ‘Treaty Abuse’, 45; Baéz Moreno, ‘GAARs and Treaties’, 437; De Broe, ‘LOB and PPT’, 213. 73 Preamble to the 2017 OECD MC. 74 Art. 10(2) OECD MC. 72

The Principal Purpose Test under Tax Treaty Law 405 as regards the default exclusive allocation of the right to tax capital gains to the state of residence75). These differences reflect various policy considerations which have guided the drafters of the OECD and UN MCs and DTCs patterned upon the latter. Therefore, merely considering the general object and purpose of the convention would not only contradict the wording of article 29(9) OECD MC but also erroneously be based on the premises that treaty provisions either have no specific object and purpose of their own or that the latter is the same for all rules. In some instances, the Commentaries rightly and rigorously focus on the object and purpose of the applicable provision. A good illustration is example J relating to article 5(3) OECD and UN MCs which provides that ‘a building site or construction or installation project constitutes a permanent establishment only if it lasts more than twelve months’. In this example, RC a company resident of state R has successfully submitted a bid for the construction of a power plant for SCO, an independent company resident of state S. That construction project is expected to last twenty-two months. During the negotiation of the contract, however, the project is divided into two different contracts, each lasting eleven months. The first contract is concluded with RCO and the second contract is concluded with SUBCO, a recently incorporated wholly owned subsidiary of RCO, a resident of state R. At the request of SCO, which wanted to ensure that RCO would be contractually liable for the performance of the two contracts, the contractual arrangements are such that RCO is jointly and severally liable with SUBCO for the performance of SUBCO’s contractual obligations under the SUBCO–SCO contract. The Commentaries correctly conclude that the subjective element of the PPT would be satisfied here and that granting ‘the benefit of that rule in these circumstances would be contrary to the object and purpose of that paragraph as the time limitation of that paragraph would otherwise be meaningless’.76 A confirmation that the object and purpose of article 5(3) OECD MC would be defeated by arrangements of this kind also flows from the fact that the Commentaries recommend an optional SAAR to deal with this problem.77 A second illustration is example E relating to the ownership threshold of article 10(2)(a) OECD MC. In this case, RCO is a company resident of state R and, for the last five years, has held 24% of the shares of company SCO, a resident of state S. Following the entry into force of a tax treaty between states R and S, RCO decides to increase to 25% its ownership of the shares of SCO. The facts and circumstances reveal that the decision to acquire these additional shares has been made primarily in order to obtain the benefit of the lower rate of tax provided by article 10(2)(a) of the treaty. This example is well conceived because it perfectly illustrates the articulation of the subjective and objective components of the PPT. The Commentaries note indeed that:78

75

Art. 13(5) OECD MC. 2017 OECD Model Commentary on art. 29, para. 182, example J. 77 Ibid., art. 5, paras 51 and 52. 78 Ibid., art. 29, para. 182, example E. 76

406 Robert J. Danon Although one of the principal purposes for the transaction through which the additional shares are acquired is clearly to obtain the benefit of Article 10(2) (a), paragraph 9 would not apply because it may be established that granting that benefit in these circumstances would be in accordance with the object and purpose of Article 10(2) (a). That subparagraph uses an arbitrary threshold of 25 per cent for the purposes of determining which shareholders are entitled to the benefit of the lower rate of tax on dividends and it is consistent with this approach to grant the benefits of the subparagraph to a taxpayer who genuinely increases its participation in a company in order to satisfy this requirement.

However, the Commentaries insist on the genuineness of the increase of participation. Accordingly, I believe that the benefit of article 10(2)(a) could still be denied if the increase of participation is a sham or simulated operation (in which case this conclusion would be technically based on a proper establishment of the facts) or defeats the object and purpose of this provision (i.e. to provide a lower residual tax in the state of source to facilitate international investment between in parent–subsidiary relationships79) because, for instance, the shareholding is immediately reduced below the threshold of article 10(2)(a) after a large dividend distribution. That being said, in several other examples, the object and purpose of the applicable provisions are not seriously considered. Rather, ‘the object and purpose of the tax convention’80 is referred to in order to determine whether treaty benefits should be granted. Hence, to deny treaty benefits, it is contended that ‘it would be contrary to the object and purpose of the tax convention to grant the benefit of that exemption under this treaty-shopping arrangement’81 and in cases in which the PPT rule does not apply, the fact that ‘the general objective of tax conventions is to encourage cross-border investment’82 is put forward. The problem with this approach is threefold. First, as already noted, it does not do justice to the clear wording of article 29(9) OECD MC. Secondly, from a policy perspective, what constitutes an encouragement of cross-border investment is highly subjective and depends by essence on various national policy considerations. Probably the only situation in which most countries would agree that there is no encouragement of cross-border investment is the one in which a DTC is used for a circular arrangement (‘round-tripping’) in which by essence the cross-border dimension is completely artificial. Finally, in various situations and as correctly pointed out, for example, by the Full Federal Court of Australia in the Lamesa case, the object and purpose of the DTC will not throw light on the interpretation to be adopted with respect to a particular treaty provision.83

79

Ibid., art. 10, para. 10. Ibid., art. 29, para. 182, examples A, B, C. and D. 81 Ibid., example A. 82 Ibid., example C. 83 Federal Court of Australia, 20 August 1997, 36 ATR 589 (1997), 77 FCR 597, 608: 80

At this point one may have regard in considering Art 13(2)(a)(iii), both to the policy of double taxation treaties generally and to the specific policy revealed in Art 13. There will be cases, and

The Principal Purpose Test under Tax Treaty Law 407 It is therefore submitted that it must always be determined whether the relevant provisions have their own specific purpose.84 In our opinion, the optional SAARs which have been included in the 2017 OECD and UN MCs may play a useful role in this respect. Two situations should be distinguished in this regard. The first situation is the case in which the applicable DTC includes the relevant SAAR. Assume, for instance, that articles 10(2)(a) and 13(4) DTC include the ‘365 day’ period test designed to prevent abusive dividend transfer transactions and the circumvention of the look-through principle applying to gains realized on the disposal of interests in real estate entities. Using the language of the Commentaries in relation to the 25% holding threshold of article 10(2)(a), it is submitted that if the taxpayer ‘genuinely’ satisfies this arbitrary 365-day test, granting tax treaty benefits is in accordance with these relevant provisions. That being said, the argument is often made that a SAAR does not exclude the application of a GAAR because a SAAR may, by its essence, not cover all cases of abuse.85 For example, in a circular letter the Indian Department of Revenue has considered that ‘[i]‌t is internationally accepted that specific anti avoidance provisions may not address all situations of abuse and there is need for general anti-abuse provisions in the domestic legislation. The provisions of GAAR and SAAR can coexist and are applicable, as may be necessary, in the facts and circumstances of the case.’86 The language of article 29(9) OECD and UN MCs, ‘Notwithstanding the other provisions of this Convention’, seems to suggest that the principle lex specialis derogat generali is no longer applicable so that the PPT could be applied ‘on top’ of any SAAR. As I have argued, there is some confusion on this point. The PPT can certainly apply alongside a SAAR but only to test an additional factual element that is not covered by the SAAR itself.87 Therefore, where, for example, the taxpayer satisfies a 365-day holding period within the meaning of article 10(2)(a) OECD MC, treaty benefits may not be denied because the tax administration nevertheless finds a particular restructuring abusive on the mere ground of the duration of the holding period. Rather, an additional factual element (unrelated to the holding

Thiel was one, where resort to the purpose of the double tax treaty to be found in the words ‘for the avoidance of double taxation with respect to taxes on income’ may throw light on the interpretation to be adopted with respect to a particular Article. But it is difficult to see that that is the case here. If Art 13 applies, then profit from the alienation is authorised to be taxed in the place where the realty referred to in the Article is. If the alienation falls outside Art 13, then any profit falls to be taxed under Art 4, in this case in the Netherlands. 84

In the same vein, De Broe, Prevention of Abuse, 344; E. Furuseth, The Interpretation of Tax Treaties in Relation to Domestic GAARs (Amsterdam: IBFD, 2018), 147. 85 See, e.g., Government of India, Ministry of Finance, Department of Revenue Central Board of Direct Taxes (F. No. 500/43/2016-FT& TR-IV), Clarifications on implementation of GAAR provisions under the Income-tax Act, 1961, 27 January 2017: ‘It is internationally accepted that specific anti avoidance provisions may not address all situations of abuse and there is need for general anti-abuse provisions in the domestic legislation. The provisions of GAAR and SAAR can coexist and are applicable, as may be necessary, in the facts and circumstances of the case.’ 86 Government of lndia, ibid. 87 Danon, ‘Treaty Abuse’, 53.

408 Robert J. Danon period itself, e.g. a conduit arrangement) would need to be present and, assuming that the subjective element of the PPT is satisfied, it would then need to be demonstrated that this additional factual element defeats the object and purpose of article 10 OECD MC. The statement of the Indian Department of Revenue that ‘[i]f a case of avoidance is sufficiently addressed by LOB in the treaty, there shall not be an occasion to invoke GAAR’88 should in our opinion be understood and put into effect in this manner. It is submitted that the 2017 OECD and UN Commentaries clarify the relationship of the PPT rule with treaty SAARs in a way that corresponds to the foregoing suggested interpretation. First of all, the Commentaries state that a benefit that is denied in accordance with a limitation-on-benefits (LOB) provision is not a benefit to which the PPT rule would also apply.89 It may thus be inferred from this passage that the PPT rule does not cover a factual situation, which an LOB clause would address as a lex specialis.90 The Commentaries discuss a situation in which a public company would be regarded as a qualified person under an LOB clause, but would enter into a conduit arrangement with respect to certain items of income. On this point, the Commentaries note that ‘[a]s long as that company is a “qualified person” . . ., it is clear that the benefits of the Convention should not be denied solely on the basis of the ownership structure of that company, e.g. because a majority of the shareholders in that company are not residents of the same State’.91 Furthermore, the object and purpose of an LOB clause are to establish a threshold for the treaty entitlement of public companies whose shares are held by residents of different states. The Commentaries then go on to say:92 the fact that such a company is a qualified person does not mean, however, that benefits could not be denied under paragraph 9 for reasons that are unrelated to the ownership of the shares of that company. Assume, for instance, that such a public company is a bank that enters into a conduit financing arrangement intended to provide indirectly to a resident of a third State the benefit of lower source taxation under a tax treaty. In that case, paragraph 9 would apply to deny that benefit because subparagraph c) of paragraph 2, when read in the context of the rest of the Convention and, in particular, its preamble, cannot be considered as having the purpose, shared by the two Contracting States, of authorising treaty-shopping transactions entered into by public companies.

In our opinion, the foregoing passages of the Commentaries may be construed as meaning that the PPT rule remains applicable to the factual elements that are not

88

Government of lndia, Ministry of Finance, Department of Revenue, Central Board of Direct Taxes, Circular No. 7 of 2017, 27 January 2017, Question No. 2. 89 OECD, Preventing the Granting of Treaty Benefits in Inappropriate Circumstances, Action 6—2015 Final Report, OECD/G20 Base Erosion and Profit Shifting Project (Paris: OECD Publishing, 2015), 59– 60; 2017 OECD Model Commentary on art. 29, para. 171. 90 In the same vein, see Baéz Moreno, ‘GAARs and Treaties’, 440. 91 OECD, Action 6 Final Report, 60; 2017 OECD Model Commentary on art. 29, para. 173. 92 Ibid.

The Principal Purpose Test under Tax Treaty Law 409 covered by the relevant SAAR.93 As already mentioned, this requires the scope of the SAAR to be determined on the basis of its interpretation. Since, in this particular example, the LOB clause covers the factual elements relating to the ownership structure of the company, these elements may not be reviewed in the light of the PPT rule. By contrast, the fact that this company also enters into a conduit arrangement is a factual element that is not related to the ownership structure of the company. Therefore, to the extent that this second and distinct factual element is not addressed by the applicable SAAR, the PPT applies to it by default. The second situation is the case in which the relevant optional SAAR does not incorporate the 365-day period test in articles 10(2)(a) and 13(4). In this situation, I would argue that this 365-day test could still be used to identify the object and purpose of these provisions. The argument runs as follows: the fact that this optional SAAR has been included in the 2017 OECD and UN MCs and Commentaries is evidence that the object and purpose of these provisions would be defeated if they could be manipulated by last-minute restructurings. As regards dividends, this interpretation is confirmed by the earlier Commentaries to the MC.94 Therefore, if such an abusive restructuring takes place, the PPT may then come into play to deny treaty benefits as the taxpayer could not successfully invoke the benefit of its escape clause in these circumstances. This is also confirmed by the Commentaries in a number of instances without, again, properly discussing the object and purpose of the provisions at stake.95 Following this line of reasoning, I would then further submit that even if the 365-day test is not incorporated in the applicable treaty provisions, this test may nevertheless serve as a safe harbour to determine whether the object and purpose of these provisions have been defeated with respect to issues concerning the timing of a restructuring. This would not only increase tax certainty but also make sense from a policy perspective: given the fact that an arbitrary 365-day test has been identified and incorporated into the OECD and UN MCs, and that the timing element is key in these types of restructurings, it would be absurd if, by relying on the PPT alone, countries having chosen not to incorporate these SAARs into their DTCs look at the object and purpose of articles 10(2)(a) and 13(4) using different timing tests.96 93

94

In the same vein, see Baéz Moreno, ‘GAARs and Treaties’, 441. See 2014 OECD Model Commentary on art. 10, para. 17:

The reduction envisaged in subparagraph a) of paragraph 2 should not be granted in cases of abuse of this provision, for example, where a company with a holding of less than 25% has, shortly before the dividends become payable, increased its holding primarily for the purpose of securing the benefits of the above-mentioned provision, or otherwise, where the qualifying holding was arranged primarily in order to obtain the reduction. To counteract such maneuvers Contracting States may find it appropriate to add to subparagraph a) a provision along the following lines: provided that this holding was not acquired primarily for the purpose of taking advantage of this provision. 95 2017 OECD Model Commentary on art. 29, para. 182, example A and in relation to a rule shopping situation, para. 186. 96 In the same vein Duff, ‘Tax Treaty Abuse’, 964–965.

410 Robert J. Danon

23.5 The Limits of the Purposive Interpretation: The Example of Alta Energy The PPT, however, reaches its limits where the problem at stake concerns the issue of nexus in the state of residence. Indeed, the primary rule, which then comes into play is the definition of treaty residence under article 4 OECD and UN MCs. Of course, various tests (e.g. the active business test of LOB clauses applied by analogy97 or, as I have suggested, the BEPS Action 5 modified nexus approach in the case of IP income) may be used as safe harbours and, if these tests are satisfied, it may be contended that the general objective of the DTC (fostering the exchanges of goods and services, and the movement of capital and persons while preventing tax avoidance)98 is fulfilled and that granting treaty benefits would be in accordance with article 4 OECD MC as this provision does not, of course, contradict such objective but aims at putting the latter into effect. Nevertheless, this does not mean that granting treaty benefits in circumstances in which such threshold is not satisfied would necessarily be in breach of the object and purpose of article 4 OECD MC (even contextually construed in the light of the new preamble). That would, indeed, amount to building an additional nexus requirement in article 4 OECD MC through the purposive interpretation. However, as shown by the Alta Energy case, the purposive interpretation cannot by definition accomplish this. In this case, it was apparent that the entity claiming treaty benefits in Luxembourg did not have strong economic ties with the country. Accordingly, the Crown had submitted that granting treaty benefits in those circumstances would defeat the object and purpose of the provisions at stake:99 Articles 1, 4 and 13(4) of the Convention, together, are intended to grant a particular treaty benefit to Luxembourg investors whose investments in specific taxable Canadian property gives rise to gains for them, in Luxembourg. Those provisions are not intended to benefit entities who do not have the potential to realize income in Luxembourg, nor have any commercial or economic ties therewith. Such situations are wholly dissimilar to the relationships or transactions that are contemplated by those provisions of the Convention.

The Federal Court of Appeal clearly dismissed this reasoning noting that:100

97

See van Weeghel, ‘A Deconstruction of the Principal Purposes Test’, 38. 2017 OECD Model Commentary on art. 1, para. 56. 99 Alta Energy Luxembourg SARL v. R, 2020 FCA 43, para. 38 referring to para. 91 of the Crown’s Memorandum. 100 Ibid., para. 65. See also already in the same vein Garron and another v. R; Re Garron Family Trust; Garron v. R; St Michael Trust Corp. v. R; Re Fundy Settlement; Dunin v. R; St Michael Trust Corp. v. R; 98

The Principal Purpose Test under Tax Treaty Law 411 There is no distinction in the Luxembourg Convention between residents with strong economic or commercial ties and those with weak or no commercial or economic ties. If a person satisfies the definition of resident in Article 4, then that person is a resident for the purposes of Articles 13(4) and (5). The Crown did not provide any support for its contention that the object, spirit and purpose of Articles 13(4) and (5) was only to exclude from taxation in Canada gains arising from the disposition of shares held by Luxembourg residents with strong economic or commercial ties to Luxembourg.

In other words, according to the court, ‘While the GAAR can change the tax consequences from what they would otherwise be, the GAAR cannot be used, in this case, to justify adding a requirement for investment that is not present in the Luxembourg Convention.’101 In a long-awaited judgment delivered on 26 November 2021, the Supreme Court of Canada upheld the lower courts’ findings.102 Specifically, the Supreme Court confirmed that the share transfer did not frustrate the object and purpose of articles 1 (personal scope), 4 (treaty residence), and 13(5) (capital gains). Therefore, there was no abuse.103 The following passages of the judgment capture the essence of the court’s reasoning: there is nothing in the Treaty suggesting that a single-purpose conduit corporation resident in Luxembourg cannot avail itself of the benefits of the Treaty or should be denied these benefits due to some other consideration such as its shareholders not being themselves residents of Luxembourg. In this case, the provisions operated as they were intended to operate; there was no abuse, and, therefore, the GAAR cannot be applied to deny the tax benefit claimed.104

Re Summersby Settlement, 12 ITLR 79, 131–132: ‘The Minister also submits that the treaty exemption was not intended to apply to the trusts because they had very little connection with Barbados. It was noted that the assets, contributors and beneficiaries are all Canadian. To apply the treaty exemption in these circumstances would facilitate the avoidance of tax by Canadians’ (para. 380) and ‘The problem that I have with this argument is that, if accepted, it would result in a selective application of the treaty to residents of Barbados, depending on criteria other than residence. It seems to me that this is contrary to the object and spirit of the treaty, which is apparent in art I and art IV(1). Residents of Barbados, as defined for purposes of the treaty, are entitled to the benefits of art XIV(4) as long as they are not also residents of Canada’ (para. 381). 101

Alta Energy Luxembourg SARL v. R, 2020 FCA 46, para. 46. Alta Energy was a split 6:3 judgment in favour of the taxpayer. The minority sided with the Crown and considered in substance that the Alta share transfer was abusive because: (1) Alta Luxembourg lacked substantial economic connections in Luxembourg (Canada v. Alta Energy Luxembourg SARL [2021] SCC 49, paras 166ff); and (2) ‘the object, spirit or purpose of the relevant provisions of the Treaty is to assign taxing rights to the state with the closest economic connection to the taxpayer’s income’ (Canada v. Alta Energy Luxembourg SARL [2021] SCC 49, para. 104). In my view, and as shown in this chapter, the minority’s position cannot be reconciled with a proper application of the objective prong of the PPT in that it amounts to revisiting the conditions to access treaty benefits which, by essence, is beyond what a GAAR may accomplish. 103 Canada v. Alta Energy Luxembourg SARL [2021] SCC 49, 24 ITLR 346, para. 95. 104 Ibid., para. 94. 102

412 Robert J. Danon This result accords with the true intention of the partners to the Treaty and must be respected.105

The richness of the Supreme Court’s reasoning makes it impossible to discuss its findings exhaustively. Worth mentioning, however, is the particular emphasis placed by the court on the rules of the VCLT and on the contractual nature of DTCs. In particular, the Supreme Court made it clear that a general anti-avoidance rule cannot be used to revisit the treaty bargain agreed by the contracting states through the back door106. Therefore, in the words of the court and consistent with the principle of reciprocity, treaty benefits may not be denied where a ‘planning strategy is consistent with the compromises reached by the contracting states’.107 In other words, the bilateral nature of DTCs which also entail the obligation to comply with the principle of pacta sunt servanda108 cannot be overridden by the ‘national self-interest’ of a contracting State.109 As I have already argued elsewhere, this observation is also fully applicable to the PPT.110 More concretely and turning first to the object and purpose of articles 1 and 4, the Canadian Supreme Court made several important observations. First, the notion of treaty residence is a defined notion111 which entirely takes its stand from domestic law.112 Secondly, treaty residence only requires the existence of unlimited tax liability in a contracting state. The fact that an entity is not effectively subject to tax on a capital gain (due to, e.g., a participation exemption) is irrelevant.113 Thirdly, unlimited tax liability and treaty residence of a company in a state based on its incorporation in that state reflects ‘an international consensus’.114 Therefore, the object and purpose of article 4 does not support the conclusion that treaty residents should have additional or sufficient substantive economic connections to their country of residence.115 Finally, in dismissing this proposition, the court also held that: treaty partners do not have the unfettered liberty to alter or redefine residence as they wish for the purposes of a tax treaty. The broader context of international tax law and the law of treaties helps to understand what was within the contemplation of Canada and Luxembourg when they drafted arts. 1 and 4(1) of the Treaty. Pursuant to the

105

Ibid., para. 95. Ibid., para. 9. 107 Ibid., para. 36. 108 Art. 26 VCLT. 109 Canada v. Alta Energy Luxembourg SARL [2021] SCC 49, 24 ITLR 346, para. 37: ‘treaties must be interpreted “with a view to implementing the true intentions of the parties” . . . The national self-interest of each contracting state must be reconciled in the interpretive process in order to give full effect to the bargain codified by the treaty’. 110 See Danon, ‘GAAR’, 243. 111 Canada v. Alta Energy Luxembourg SARL [2021] SCC 49, 24 ITLR 346, para. 53. 112 Ibid., para. 56. 113 Ibid., para. 54. 114 Ibid., para. 62. 115 Ibid., para. 67. 106

The Principal Purpose Test under Tax Treaty Law 413 principle of pacta sunt servanda, parties to a treaty must keep their sides of the bargain and perform their obligations in good faith (art. 26 of the Vienna Convention). Domestic law definitions of residence should therefore broadly correspond to international norms and not have the effect of redefining residence in a way ‘that takes the words unmistakably past their accepted usage’ . . ., including the definitions of residence that were in effect in the two states at the time the Treaty was drafted.116

In the light of the foregoing, the court concluded that: In sum, the object, spirit, and purpose of arts. 1 and 4(1) are to allow all persons who are residents under the laws of one or both of the contracting states to claim benefits under the Treaty so long as their resident status could expose them to full tax liability (regardless of whether there is actual taxation). They are broadly consistent with international norms.117

With respect to the object and purpose of article 13(5) (allocating to the state of residence of the alienator of the shares the exclusive right to tax a capital gain), the Supreme Court confirmed that this provision was equally not subject to any additional condition other than treaty residence within the meaning of the DTC. I subscribe to the foregoing which I also find to be relevant for the purposes of the objective component of the PPT. Of course, the Commentaries certainly provide that the notion of ‘arrangement or transaction’ within the meaning of article 29(9) OECD and UN MC:118 also encompass arrangements concerning the establishment, acquisition or maintenance of a person who derives the income, including the qualification of that person as a resident of one of the Contracting States, and include steps that persons may take themselves in order to establish residence. An example of an ‘arrangement’ would be where steps are taken to ensure that meetings of the board of directors of a company are held in a different country in order to claim that the company has changed its residence.

Therefore, there is little doubt that the timing and nature of activities in the state of residence are relevant considerations to take into account when looking at the subjective element of the PPT. However, it is not possible to rely on the object and purpose of article 4 OECD and UN MCs—if one prefers the sui generis purposive interpretation—to build these additional requirements into the provision. That said, one question which has emerged following the Alta Energy case is whether the new preamble to the 2017 OECD MC should lead to a different conclusion. The new preamble certainly states that the purpose of DTCs is now ‘the elimination of double 116

Ibid., para. 59. Ibid., para. 62. 118 2017 OECD Model Commentary on art. 29, para. 177. 117

414 Robert J. Danon taxation without creating opportunities for non-taxation or reduced taxation through tax evasion or avoidance (including through treaty-shopping arrangements aimed at obtaining reliefs provided in this Convention for the indirect benefit of residents of third States)’.119 The question is therefore whether the new preamble—when taken into consideration as part of the objective prong analysis of the PPT—should lead to the conclusion that the mere incorporation of an entity in a country without actual significant economic ties with that country defeats the object and purpose of the notion of treaty residence. In my view, even in the context of the purposive interpretation dictated by the PPT (which is no longer limited by the treaty wording120), such a conclusion cannot be upheld. This is because the notion of treaty residence which refers exclusively to domestic law (‘who, under the laws of that State, is liable to tax’121) leaves no room for such an amplification. Moreover, there is another problem with such reasoning: the preamble is so vague that the sufficient level of nexus with the state of residence becomes impossible to determine in any given case. Therefore, such reasoning would clearly be in breach of the principle legal certainty122 which, as a general principle of law (art. 38(1) ICJ Statute), is a relevant rule of international law applicable between the parties (art. 31(3)(c) VCLT)123 which must also be taken into account for the purposes of the interpretation of the PPT. In Alta Energy, the Supreme Court of Canada rightly emphasized that compliance with the tax treaty bargain equally concerns taxpayers and their right to legal certainty and fairness when arranging their tax affairs: [T]‌he principles of predictability, certainty, and fairness and respect for the right of taxpayers to legitimate tax minimization are the bedrock of tax law. In the context of international tax treaties, respect for negotiated bargains between contracting states is fundamental to ensure tax certainty and predictability and to uphold the principle of pacta sunt servanda, pursuant to which parties to a treaty must keep their sides of the bargain.124

This also applies to the PPT which cannot be used to amplify the treaty residence test in an unpredictable way. I appreciate that the foregoing conclusion obviously has an impact where the only requirement contained in a distributive rule is treaty residence in a contracting state. This holds true, for instance, as regards article 13(5) OECD MC and article 13(6) UN MC

119

Preamble to the 2017 OECD MC. As this is the case under the regular interpretation of treaty law prescribed by art. 31 VCLT which precedes the application of the PPT, see Section 23.2. 121 Art. 4(1), 2017 OECD MC. 122 On the relevance of the principle of legal certainty in the context of art. 31(3)(c) VCLT, see recently in investment treaty practice, Cairn (PCA Case No. 2016-07), paras 1712–1715. For a discussion of the Cairn case, see inter alia R. Danon and S. Wuschka, ‘International Investment Agreements and the International Tax System: The Potential of Complementarity and Harmonious Interpretation’, Bulletin for International Taxation 75 (2021), 687–703. 123 Dörr, ‘Article 31 VCLT’, para. 95. 124 Canada v. Alta Energy Luxembourg SARL [2021] SCC 49, 24 ITLR 346, para. 1. 120

The Principal Purpose Test under Tax Treaty Law 415 (capital gains) as well as article 21(1) OECD and UN MCs (other income). For example, article 13(5) OECD MC states that ‘Gains from the alienation of any property, other than that referred to in paragraphs 1, 2, 3 and 4, shall be taxable only in the Contracting State of which the alienator is a resident.’ This provision does not, in particular, include any specific nexus requirement or, for example, a holding period during which the movable assets sold would need to be held. The fact that a 365-day period was included in article 10(2)(a) and in article 13(4) but not in article 13(5) OECD and UN MCs confirms this conclusion. As mentioned already, the Commentaries state, however, that:125 A person may sell a property for various reasons, but if before the sale, that person becomes a resident of one of the contracting States and one of the principal purposes for doing so is to obtain a benefit under a tax convention, art. 29(9) OECD MC could apply notwithstanding the fact that there may also be other principal purposes for changing residence, such as facilitating the sale of the property or the re-investment of the proceeds of the alienation.

While it is conceivable here that the subjective element of the PPT is satisfied (e.g. if the asset is sold shortly after the transfer of residence), I fail to see why granting treaty benefits in these circumstances would defeat the object and purpose of article 13(5) OECD MC.126 As already correctly observed by a commentator at the time in relation to the Lamesa case,127 the foregoing shows that the fact that a taxpayer sets up an arrangement with the aim of avoiding tax is in itself not sufficient to deny them the treaty benefits. Rather, the taxpayer should also act contrary to the objectives of the provision of which they seek an advantage as such objectives are materialized in the terms of that provision.128 In my opinion, this observation remains fully valid with respect to the PPT.

23.6 Conclusions The findings of this chapter which I have also expressed elsewhere129 may be summarized as follows: (1) Today, the PPT represents the multilateral minimum standard to combat the improper use of DTCs. The impact of the PPT on existing business models of 125

2017 OECD Model Commentary on art. 29, para. 180. Treaty benefits could, however, be denied if it were established that, as a matter of legal substance, the transaction had already been carried out before the transfer of residence and that the subsequent sale was just a sham. The denial of treaty benefits would then not be based on the PPT but rather on a proper establishment of the facts. 127 Federal Court of Australia, 20 August 1997, 36 ATR 589 (1997), 77 FCR 597, 608. 128 De Broe, Prevention of Abuse, 373. 129 The ideas expressed in this chapter have also been discussed in Danon, ‘The PPT in Post-BEPS Tax Treaty Law’, 32–34. 126

416 Robert J. Danon MNE groups has triggered much speculation. In particular, because the PPT is rooted in the BEPS initiative which has seen the emergence (or reinforcement) of concepts such as value creation and substance, the question has arisen as to whether, and if so to what extent, the PPT is capable of affecting fundamentally the conditions governing the access to tax treaty benefits (e.g. by imposing additional substance requirements in order to be regarded as a resident of a contracting state). (2) This chapter has argued that as a GAAR, the PPT is not capable of building into tax treaty law requirements that were never intended. This holds true, in particular, as regards the notion of ‘resident of a contracting state’ under article 4 OECD and UN MCs which has not been modified by the BEPS initiative and which continues to be based on the domestic law of the relevant contracting state. From this perspective, I fail to see how a purposive interpretation based on the ‘object and purpose of the relevant provisions of this Convention’130 could have any effect on a definition which is exclusively one of a domestic nature. It may be argued that the liberal notion of treaty residence under article 4 OECD and UN MCs may facilitate instances of treaty shopping and that predominance of residence over source taxation may be achieved by the mere incorporation of an entity in a contracting state as long as such incorporation triggers unlimited tax liability in that state. However, this is not something that the PPT as a GAAR may revisit. Rather, this is a matter of treaty policy which would require reconsidering the notion of treaty residence under article 4 OECD and UN MCs. On this point, the findings of the Supreme Court of Canada in the Alta Energy case are thus to be approved and are also relevant to the interpretation to the PPT. Further, the preamble to the 2017 OECD MC which explicit references that the prevention of treaty shopping cannot be used to amplify the notion of treaty residence (i.e. requiring strong economic ties with a contracting state to qualify as a resident) in an unpredictable fashion. Any other conclusion would also be in breach of the principle of legal certainty which as a general principle of law is applicable between the contracting states and thus forms part of the interpretative exercise pursuant to article 31(3)(c) VCLT. While I appreciate that as a GAAR, the PPT is naturally at tension with the principle of legal certainty, that principle cannot however be fully emptied so as to lead to completely unpredictable outcomes.

130

Art. 29(9) OECD and UN MCs.

Chapter 24

Tax Treaties a nd Huma n Rights L aw Philip Baker

24.1 Introduction: The Development of Tax Treaties and Human Rights Law This chapter considers a small range of issues relating to the interrelationship between double taxation conventions (DTCs or ‘tax treaties’) and human rights norms as set out in general human rights conventions.1 The network of over 3,500 largely bilateral tax treaties has been negotiated and developed almost entirely on the basis of model conventions whose origins predate the main developments in human rights law.2 It 1 There is only a very limited literature on the issue of tax treaties and human rights law. This is an adapted version of ’Double Taxation Conventions and Human Rights’, originally published in Tax Polymath: A Life in International Taxation. Essays in Honour of John F. Avery Jones, ed. P. Baker and C. S. Bobbett (Amsterdam: IBFD, 2011). For more information, visit https://www.ibfd.org. Adapted with permission. Other literature includes: I. Stepanova, ‘European Double Taxation Conventions and Human Rights in a Globalised World’, unpublished master’s degree dissertation from Uppsala University, available from the author; T. Georgopoulos, ‘Tax Treaties and Human/Constitutional Rights: Bridging the Gap?’, (unpublished, but available at http://www.law.nyu.edu/sites/default/files/upload_docume nts/gffgeorgopoulospaper.pdf); B. Hassan Al-Rawashdeh, ‘Taxpayers’ Rights in Mutual Agreement Procedure on the Basis of Double Taxation Treaties and the BRICS Countries’, Journal of Legal, Ethical and Regulatory Issues 22 (2019), 1–7. Additionally, there is a very full discussion of the application of human rights provisions and the participation of taxpayers in the mutual agreement procedure in K. Perrou, Taxpayer Participation in Tax Treaty Dispute Resolution (Amsterdam: IBFD, 2014). 2 The principal models are the OECD Model Tax Convention on Income and on Capital, which originated in a draft convention in 1961, and the UN Model which originated in 1980 (and was largely based on the OECD Model). Despite the fact that these Models have been updated (most recently in 2017 in both cases), neither document makes even a single reference to ‘human rights’. That is somewhat astonishing of itself, given the increasing impact of human rights law in all fields of human activity. The main developments in human rights law postdate the 1960s, and the application of human rights norms in the tax field is largely a development of the past twenty years. For an overview of human rights norms

418 Philip Baker would not be surprising, therefore, if there were a need to update the models to comply with those norms. This initial expectation must be tempered, however, by the realization that human rights instruments have, so far, had a limited impact in the field of taxation. Nevertheless, one can discuss the extent to which tax treaties may or may not be compatible with the—albeit limited—impact of human rights instruments in the tax field. This chapter considers these issues. The starting point is to recognize that tax treaties, like any part of a tax system, may be scrutinized for compatibility with human rights norms. There is no valid reason why tax treaties should be immune from scrutiny to ensure consistency with international human rights norms. Nor is there any reason in principle why taxpayers might not challenge the application to them of the provisions of tax treaties on the grounds that those provisions infringe human rights norms. Such challenges may have a limited chance of success, given the limited impact of human rights norms in the tax field so far, but there may be certain specific areas (e.g. the exchange of information and the operation of the mutual agreement procedure—both discussed later) where there are real issues to consider. Even where challenges are unlikely to succeed, those involved in amending the OECD or UN Models might wish to consider changes to ensure that no possibility exists that conventions based on the models might even appear to infringe human rights norms. There is no evidence to suggest that such a review has ever taken place. In terms of human rights norms, this chapter focuses on the provisions of the European Convention on Human Rights (ECHR) and, to a lesser extent, of the International Covenant on Civil and Political Rights (ICCPR).3 It should be recalled, however, that in many countries there will also be constitutional or other legal guarantees for the rights of taxpayers which may give wider protection than these international instruments.4 There are also regional human rights instruments, such as the American Convention on Human Rights5 or the European Union’s Charter of Fundamental Rights which contain rights similar but not identical to those in the ECHR and ICCPR. This chapter considers first the (albeit very limited) case law under the ECHR where a tax treaty has been considered by the European Court of Human Rights (ECtHR). It and taxation in general, see D. Bentley, Taxpayers’ Rights: Theory, Origin and Implementation (Alpen aan den Rijn: Kluwer, 2007); P. Baker and P. Pistone, ‘General Report: The Practical Protection of Taxpayers’ Fundamental Rights’, Cahiers de Droit Fiscal International, vol. 100B (2015); M. P. Maduro and P. Pistone, eds, Human Rights and Taxation in Europe and the World (Amsterdam: IBFD, 2011); P. Alston and N. Reisch, Tax, Inequality and Human Rights (Oxford: Oxford University Press, 2019); IBFD, ‘Observatory on the Protection of Taxpayers’ Rights’, available at https://www.ibfd.org/Academic/Observatory-Pro tection-Taxpayers-Rights; W. Nykiel and M. Sek, eds, Protection of Taxpayer’s Rights (Warsaw: Wolters Kluwer, 2009). 3 The ICCPR was adopted by the UN General Assembly by Res. 2200 A (XXI) of 16 December 1966. There are currently (February 2023) 173 states parties to the ICCPR. 4 Obviously, the position here varies from country to country, but in many cases the constitution will contain guarantees similar to those found in the ECHR and ICCPR. 5 Otherwise known as the Pact of San José, Costa Rica of 22 November 1969: there are presently 24 countries that have ratified (and not denounced) the convention.

Tax Treaties and Human Rights Law 419 then proceeds to consider, as a matter of principle, how those provisions of the ECHR most relevant to taxation (which are identified later) may interrelate to tax treaties in respect to a number of specific, identified issues (e.g. in respect of the exchange of information).

24.2 Cases on Tax Treaties Before the ECnHR and ECtHR Tax treaties (or equivalent measures) have been considered in a small number of complaints brought by individuals under the complaint mechanism of the ECHR. This mechanism previously involved consideration of the complaint by the European Commission of Human Rights (ECnHR), but the functions of the ECnHR have now been assumed by the ECtHR.6 In none of these cases was a breach of the ECHR found; nevertheless, the discussion of issues in these cases is illuminating. In Hanzmann v. Austria,7 the applicant was an Austrian civil servant living across the border in Germany. As a result of the application of the Austria–Germany tax treaty of 1954, he was subject to taxation on his salary in Austria but, because he was resident in a foreign country, was subject only to limited tax liability. The consequence was that certain tax allowances, which were available only to persons with full tax liability, were not extended to him. He complained that as a result of the refusal of these allowances, he was discriminated against both in comparison with an equivalent civil servant resident in Austria and with a civil servant, such as a diplomat, who was both resident and working in a foreign country. He based his complaint on article 14 (non-discrimination) of the ECHR8 in conjunction with article 1 of the First Protocol to the ECHR (protection of property). In line with its jurisprudence in many other tax cases, the ECnHR noted the wide margin of appreciation enjoyed by states in the field of taxation. The ECnHR found that the decision not to extend the allowances to a person subject to limited tax liability fell well within the margin of appreciation enjoyed by Austria. Given the wide acceptance of the OECD or UN Model as a basis for double taxation conventions, one can assume that it is extremely unlikely that the conclusion of a convention based upon one of these Models would be regarded as falling outside a state’s margin of appreciation. That does not mean, however, that the provisions of these Models should automatically be regarded as compliant with human rights norms. 6

The equivalent mechanism for the ICCPR is review by the Human Rights Committee. Unlike the system under the ECHR, not all contracting states have accepted the right of individual petition. At present, the author is unaware of any complaints to the Human Rights Committee which have involved double taxation conventions. 7 App. No. 12560/ 86, 60 DR 194. It is interesting to compare this case with Gilly v. Directeur des Services Fiscaux [1998] STC 1014, [1998] ECR-I 2793 (ECJ). 8 Discussed later.

420 Philip Baker In the case of H v. Sweden,9 the applicant, who was a resident of Sweden, went to work in Germany for a period of four years. He did not report his income from his employment in Germany to the Swedish authorities. Following his return, he was assessed to additional tax and penalties for his failure to report the income. He appealed against the assessment and was given an oral hearing at the first trial, but no oral hearing on the subsequent appeals. He complained both of the absence of an oral hearing and of the failure of the Swedish tribunals to take account of the tax treaty between Sweden and Germany in determining whether he was liable to tax. The ECnHR dismissed the case on the grounds, inter alia, that the Commission had no jurisdiction to review a failure of the domestic tribunals to take account of the provisions of a double taxation convention. A somewhat similar issue arose in the case of Lindkvist v. Denmark,10 where the underlying question was whether the taxpayer had been resident in Denmark or in Switzerland during the years 1985–1990. That issue turned in part on the application of the Denmark–Switzerland tax treaty of 1973. The issue of the taxpayer’s residence had been submitted to the mutual agreement procedure under that tax treaty but without any agreement being reached. The human rights issue concerned the treatment of the taxpayer at the hands of the Danish revenue authorities, and involved allegations relating to article 5 ECHR (deprivation of liberty), article 6 (based on the excessive length of the procedure), and article 1 of the First Protocol (arising from the seizure of the taxpayer’s property). The ECnHR concluded that all of the alleged infringements were manifestly ill-founded. The case of FS v. Germany11 is so far perhaps the most interesting case to involve a tax treaty and human rights norms. In that case, the applicant complained under article 8 ECHR (right to respect for private and family life) against exchange of information between the German and Dutch tax administrations under the provisions of the European Community Directive on Mutual Administrative Assistance.12 The ECnHR agreed that the exchange of information was an infringement of the right of privacy but that it could be justified within the scope of article 8(2) ECHR: the measures in question were taken in the interests of the economic well-being of the country, and were necessary in a democratic society. The ECnHR noted the current trend towards strengthening international cooperation in the administration of justice and concluded that there were relevant and sufficient reasons for the introduction of the directive.13

9 App. No. 12670/87. This case is available on the HUDOC database on the ECtHR website (https:// www.echr.coe.int). 10 App. No. 25737/94, decision of the ECnHR of 9 September 1998. 11 App. No. 30128/96, also available on HUDOC, see https://hudoc.echr.coe.int. 12 Council Directive 77/ 799/ EEC of 19th December 1977 concerning mutual assistance by the competent authorities of the member states in the fields of direct and indirect taxation. 13 Issues relating to exchange of information between states have concerned the ECtHR on several occasions, though discussion of the relevant provisions in tax treaties did not concern the Court to any significant degree. These cases include GSB v. Switzerland (App. No. 28601/11) involving the conclusion

Tax Treaties and Human Rights Law 421 In that case, the measure concerned was a European Community directive. It seems highly likely that an identical result would have been reached had the issue concerned an exchange of information under the equivalent of article 26 of the OECD Model.14 This discussion has shown that the impact of a tax treaty has been part of the background to several cases before the ECtHR. However, that Court has not, to date, been required to consider in any depth how human rights norms might apply to the conclusion or application of a tax treaty.

24.3 Application in Principle of the ECHR and ICCPR to Tax Treaty Provisions Turning from a consideration of complaints which have been made to the ECnHR or the ECtHR, one can consider this topic by examining as an issue of principle the rights guaranteed by the ECHR and ICCPR and whether provisions contained in tax treaties might infringe these rights. In this discussion, not all provisions of the ECHR and ICCPR are by any means relevant: the only provisions which are likely to have any relevance whatsoever in the context of taxation are (in the ECHR—the ICCPR will contain similar, but not always identical, provisions): Article 6 (right to a fair trial); Article 8 (right to respect for private and family life); Article 13 (right to an effective remedy); Article 14 (prohibition of discrimination); Article 1 of the First Protocol (protection of property); and Article 2 of the Fourth Protocol (freedom of movement).15

of an inter-governmental agreement between Switzerland and the USA in relation to the supply of financial information for tax purposes (i.e. one of the agreements implementing the Foreign Account Tax Compliance Act). The taxpayer argued that the agreement had retrospective effect and interfered with his right to privacy. The ECtHR concluded that there was no breach. A further case is Othymia Investments BV v. Netherlands (App. No. 75292/10) which involved the exchange of information between the Netherlands and Spain. The information was already held by the Dutch tax authorities, so the taxpayer was not given a prior warning of the exchange of information and had no way to challenge it. The ECtHR held that there was no duty to give prior notification to the taxpayer. 14 But see later for a discussion of the procedural safeguards enjoyed by taxpayers and others in relation to requests for exchange of information. 15 Or the equivalent provisions of the ICCPR, where those exist.

422 Philip Baker

24.3.1 The Right to a Fair Trial and its Application to Competent Authority Proceedings (Mutual Agreement Proceedings, MAP) One of the most interesting issues relating to tax treaty provisions and human rights norms is the question of whether the dispute-mechanism procedure in tax treaties—that is the competent authority procedure (sometimes referred to as the mutual agreement procedure, MAP)16—conforms to human rights standards (or is even required to conform to those standards). Article 6 ECHR commences as follows: Article 6—Right to a fair trial

1. In the determination of his civil rights and obligations or of any criminal charge against him, everyone is entitled to a fair and public hearing within a reasonable time by an independent and impartial tribunal established by law. This article raises issues relating to the resolution of disputes under the mutual agreement procedure. If article 6 applies to that procedure, then there is very little doubt that the procedure would not satisfy the guarantees in the article. These guarantees include ‘the right to a court’—that is, the right to take the dispute to a tribunal, including the right to an independent and impartial tribunal, and the right to make representations to the tribunal on the basis of an equality of arms. The mutual agreement procedure does not satisfy any of these guarantees. Specifically, the taxpayer has no right to initiate MAP (which may be refused by one or both of the competent authorities) and no right to participate directly in the proceedings. The competent authorities are not independent or impartial; the procedure is a classic example of an internal procedure involving the revenue authorities interested in the outcome of the dispute. Article 6 also requires a determination within a reasonable time: while efforts have been made recently to improve the time taken to resolve MAP proceedings, it is still the case that some proceedings fail to satisfy the reasonable time requirement. There is always the possibility that competent authority proceedings will fail to reach a resolution of the issue at all; there is a requirement for the competent authorities to endeavour to resolve a dispute, but no obligation to do so, and there are many recorded cases where no resolution was agreed. The taxpayer will not, therefore, necessarily obtain a determination of the issue through the proceedings, which may be the only ones that offer any prospect of an effective remedy for the taxpayer. Some of the failings in the mutual agreement procedure may be remedied by the introduction of mandatory binding arbitration through provisions based on article

16

On MAP, see art. 25 of the OECD Model and the discussion in the Commentary on that article.

Tax Treaties and Human Rights Law 423 25(5) of the OECD Model or Part VI of the Multilateral Convention to Implement Tax Treaty Related Measures To Prevent Base Erosion And Profit Shifting (the BEPS MLI). The introduction of mandatory binding arbitration at least ensures that, where a dispute is admitted to MAP, there will be a resolution of the dispute. What these measures do not guarantee, however, is any greater participation of the taxpayer in that process. Nor do they guarantee that the dispute will be accepted and admitted to MAP in the first place. Even after the introduction of mandatory binding arbitration, the MAP procedure still falls well short of international human rights standards. There are several preliminary questions, however, relating to the applicability of article 6 to competent authority proceedings. First, the ECtHR has concluded that article 6 does not apply to ordinary litigation relating to the determination of a tax liability.17 The reason for this conclusion is that article 6 refers to the ‘determination of . . . civil rights and obligations’.18 The ECnHR and ECtHR have both concluded that in the drafting of the ECHR it was intended to exclude administrative law disputes because of the essentially public law nature of those proceedings. Competent authority proceedings are quintessential, public law proceedings carried out by way of discussion and negotiation between civil servants of the two contracting states. Though the case law of the ECtHR in this context has been criticized, this has been followed in recent cases.19 It is notable, however, that the ICCPR employs slightly different wording, ‘the determination of . . . his rights and obligations in a suit at law ...’, though competent authority proceedings do not easily come within the phrase ‘a suit at law’.20 Recent jurisprudence before the Human Rights Committee may suggest a different approach to public law procedures under the ICCPR.21 At the same time, other regional human rights instruments such as article 8 of the American Convention on Human Rights expressly applies to fiscal matters. Additionally, the jurisprudence of the ECtHR has come to recognize that disputes involving any substantial, tax-geared penalty will fall within the criminal head of article 6.22 This will be true of many of the more significant disputes that arise for the competent authorities. The prime example is a transfer pricing dispute, where substantial misstatement penalties are likely to follow for the enterprises that are

17 See Ferrazzini v. Italy (App. No. 44759/98) and other cases referred to there. A number of people (the author included) consider that this case (which was an 11:6 majority decision, with the taxpayer not being represented) was wrongly decided. 18 The ECtHR has held that disputes relating to substantial tax- geared penalties may fall within the guarantees of art. 6 for ‘criminal charges’—see Bendenoun v. France (App. No. 12547/86) (1994) 18 EHRR 54. 19 See, e.g., Remy v. Germany (App. No. 70826/01) and HM v. Germany (App. No. 62512/00). 20 It is worth noting, however, that the equivalent provision in the American Convention on Human Rights expressly applies to fiscal disputes. 21 See Lederbauer v. Austria, CCPR/C/90/D/1454/2006, decision of 11 September 2007. 22 See, e.g., Jussila v. Finland (App. No. 73053/01).

424 Philip Baker involved in the dispute. The penalties would, of course, be matters of domestic law and may not even be at issue in the competent authority proceedings the purpose of which is to settle the transfer pricing issues or methodology. However, the eventual imposition of penalties may be regarded as integrally associated with the determination of the transfer pricing issues. On that basis, article 6 would apply to the MAP process. Secondly, the mutual agreement procedure is an alternative or a supplement to litigation before the courts of the contracting states for most disputes. The guarantees of article 6 would normally be satisfied by the proceedings before the national courts (which should provide a fair trial before an independent and impartial tribunal). To satisfy article 6, it is not necessary that all stages of the resolution of a dispute fulfil the guarantees, so long as there is at some point the possibility of recourse to an independent and impartial tribunal which fulfils the requirements of article 6. To that extent, it is possible to argue that MAP need not meet human rights standards because the alternative route of appeal to national courts will satisfy those requirements. This argument only goes so far, however. There are some disputes where the only effective means of resolving the dispute is recourse to competent authority proceedings, for example the application of the final leg of the tiebreaker on individual residence in article 4(2)(d) of the OECD Model. In practice, disputes involving transfer pricing are ones where the taxpayer may be largely indifferent to the outcome, so long as the same transfer pricing approach is adopted in both countries: litigation separately before the courts in two countries is likely to offer a significantly less attractive option. There are also rare cases where litigation in the contracting states has reached or will reach different results and the only way to avoid double taxation is via competent authority proceedings. In practice, though not perhaps as a matter of strict rules, MAP is the only effective method of resolving many international tax disputes: if this were not so, it is hard to see why tax treaties would contain MAP provisions. The taxpayer whose final tax liability in the two contracting states will be determined by competent authority proceedings may feel aggrieved that those proceedings do not begin to satisfy article 6. However, it still remains for the extent to which article 6 applies to those proceedings to be determined. A taxpayer may have a better argument under either the ICCPR or one of the other regional human rights instruments. So far, the author is not aware of any case before the ECtHR or similar body where the failure of MAP to satisfy human rights guarantees has been directly in issue.

24.3.2 Double Taxation Conventions and the Right to Privacy A second area where the ECHR may have relevance concerns the right to privacy in article 8 ECHR (or the equivalent provision in art. 17 ICCPR) and the exchange of information under article 26 of the OECD Model. Article 8 ECHR provides as follows:

Tax Treaties and Human Rights Law 425 Article 8—Right to respect for private and family life

1. Everyone has the right to respect for his private and family life, his home and his correspondence. 2. There shall be no interference by a public authority with the exercise of this right except such as is in accordance with the law and is necessary in a democratic society in the interests of national security, public safety or the economic well-being of the country, for the prevention of disorder or crime, for the protection of health or morals, or for the protection of the rights and freedoms of others. The issue of exchange of information and infringement of the right of privacy was considered by the ECnHR in FS v. Germany (discussed earlier). As in that case, it seems that in most cases the countries concerned with the exchange of information could justify any prima facie infringement of privacy arising from the passing on of information relating to a taxpayer by reliance on the provisions of paragraph 2 of article 8. In particular, in many cases the exchange of information will be necessary as perhaps the only effective way to combat international tax avoidance. In all cases, the information would also have been collected by one of the revenue authorities prior to exchange: if that collection of information could be justified under article 8(2), then it is hard to argue that the transmission of the information to another state under the provisions of a tax treaty would constitute an infringement of the ECHR.23 There are, however, a small number of situations where exchange of information may raise issues of possible infringement of human rights norms. One concern is where there is a well-founded fear that the recipient country will not preserve the confidentiality of the information. This issue touches more broadly on the question whether exchange of information can take place with a jurisdiction that has a history of poor observance of taxpayer confidentiality, or alternatively a jurisdiction that has strong freedom of information

23 One issue has previously arisen in France with respect to the disclosure to the taxpayer of information obtained from another revenue authority. The view of the French Administration was understood to be that provisions equivalent to art. 26 of the OECD Model provide only for information which has been exchanged to be disclosed to ‘persons or authorities (including courts and administrative bodies) concerned with the assessment or collection of, the enforcement or prosecution in respect of, or the determination of appeals in relation to, the taxes covered by the Convention [the relevant DTC]’. The taxpayer in respect of whom the information has been exchanged was not such a person, so disclosure to them was not permitted, even where the information is made available to the tax tribunal. Utilizing information obtained by administrative cooperation, but not making that information available to the taxpayer concerned, appears to be a prima facie breach of the principle of equality of arms (always assuming that art. 6 applies to the tax proceedings concerned). It is understood that French practice has subsequently changed, and that information exchanged is now made available to the taxpayer. Most countries would be expected to take the view that a duty of candour (or similar duty of disclosure) requires them to disclose information obtained by exchange with another revenue authority to the taxpayer concerned. This issue is discussed in part in the case of SA Diebold Courtage, Conseil d’État, 8th and 9th subsections, 13 October 1999, Req. No. 191 191, Revue de Droit Fiscal, 1999, No. 52, Comm. 948. The case is printed with a partial translation in (1999) 2 ITLR 365.

426 Philip Baker legislation which allows public access to confidential taxpayer information.24 Tax treaties may have been concluded with countries that have a regime that may not be regarded as benign, and that has a poor human rights record. Question may arise of the misuse of information by that regime to target opponents of the regime, or those with foreign links.25 More generally, there are broader issues relating to the protection of the right of privacy and also the question of data protection.26 It is understood that the peer-review process within the OECD for countries to receive information under the Common Reporting System for financial account information includes an assessment of a country’s record on maintaining confidentiality. However, there are concerns about the independence and effectiveness of that review, particularly in the light of the relatively frequent reports of information being inadvertently disclosed by tax authorities, or the hacking of computers of revenue authorities. The breach of article 8 would be an additional leg to an argument that exchange of information should not take place in those circumstances. There will also be an issue under human rights norms where there is no effective procedure to challenge the exchange of information under article 26 of the OECD and UN Models. For example, article 26 promises that no information will be exchanged which reveals trade, business, industrial, commercial, or professional secrets. However, for most countries this promise is illusory since the taxpayer is not informed that information about them is being exchanged and has no chance to challenge the exchange.27 Article 13 ECHR requires that everyone whose rights and freedoms under the ECHR have been violated should have an effective remedy.28 Suppose that a taxpayer were

24

A revenue official who exchanges information in those circumstances, knowing that the recipient authority does not in practice or cannot guarantee confidentiality, risks potential liability for breaching taxpayer confidentiality. It would be questionable whether the revenue official could rely upon the Double Taxation Convention as a defence where they are aware—or possibly reckless—as to the possible disclosure of information in the other contracting state. 25 A specific concern relates to the possible misuse of information obtained by way of exchange of information against human rights defenders by some regimes, e.g. by identifying and persecuting those groups that have received donations from foreign organizations, including charities that support human rights monitoring and defence work. 26 This latter issue is generally governed by specific legislation relating to data protection, e.g. the European Union General Data Protection Regulation (Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/ 46/EC). The exchange of information between different governmental authorities is a topic that has developed rapidly within the last twenty years, and most general human rights instruments predate this development. The topic raises a number of issues which go beyond the scope of this short chapter. 27 It is interesting to note that the Keith Committee on the powers of the UK revenue authorities recommended that taxpayers should be informed and have a right to challenge an exchange of information, but the Inland Revenue rejected this recommendation. It is arguable that art. 13 ECHR requires such a remedy where a potential breach of the right to privacy is at issue. 28 Art. 13 provides as follows: Everyone whose rights and freedoms as set forth in this Convention are violated shall have an effective remedy before a national authority notwithstanding that the violation has been committed by persons acting in an official capacity.

Tax Treaties and Human Rights Law 427 able to show that information had been gathered about them in breach of article 8. Or suppose that the information was to be exchanged with another country which might not be justifiable under article 8. The taxpayer might have a remedy in a national court after the event for damages (if they could show any loss—which is unlikely), but in most countries they would have no effective means of preventing the exchange. At least at first blush, there is an argument that article 13 has been violated. The taxpayer has not been given an effective remedy to challenge the breach of article 8.29 The issue of the appropriate remedies that are to be provided to parties who are the ‘target’ (in broad terms) of cross-border information requests has been the subject of recent litigation before the Court of Justice of the European Union. Though, strictly speaking, this litigation has arisen under the protection of fundamental rights guaranteed by the EU Charter of Fundamental Rights, there is a clear parallel here with the rights guaranteed under general human rights instruments, including the ECHR and the ICCPR. In particular, in the Berlioz Investment Fund SA case30 Advocate General Wathelet explicitly referred to the equivalent protections under articles 6 and 13 ECHR.31 In that case, a request was sent by the French tax administration to the tax administration of Luxembourg to seek information from a company in Luxembourg in respect of an investigation relating to the company’s French subsidiary and, potentially, the ultimate shareholders of the Luxembourg company. Thus, the Luxembourg company which was asked to supply information relating, inter alia, to the identity of its shareholders, was not itself subject to a tax investigation in France. A distinction was drawn between the taxpayers in France, who ultimately would have an opportunity to challenge their tax liability in France if an assessment were issued against them, and the Luxembourg company which would not have that opportunity. Under Luxembourg law, the company had no opportunity to challenge the notice requiring it to provide information: the only remedy would be if it refused to provide information and was subject to a consequential penalty, it could then appeal against that penalty. Consequently, the company had no effective remedy to challenge the notice to provide information about its shareholders. That was a breach of the rights enshrined in the equivalent of article 6 (right to a fair trial) and article 13 (right to an effective remedy) of the ECHR. Further litigation has followed arising from the same issue.32 The extent to which a person, who is subject to a process of exchange of information between countries based on the provisions in a tax treaty or equivalent instrument, has a right of access to an effective and independent procedure to challenge that process remains to be conclusively determined.

29 This potential position under art. 13 is wholly independent of the issue of applicability of art. 6—the issue at stake is the potential infringement of rights under the ECHR, not the determination of a tax liability. 30 Case C-682/15, ECLI:EU:C:2017:373. 31 See in particular paras 34–39 of the Advocate General’s Opinion, ECLI:EU:C:2017:2. 32 See Joined Case C-245/19 and C-246/19 State of Luxembourg, ECLI:EU:C:2020:795.

428 Philip Baker

24.3.3 Non-Discrimination Double taxation conventions and human rights instruments overlap to the extent that they both contain non-discrimination provisions. There is, of course, a wide difference in form and scope between the non-discrimination provisions in article 24 of the OECD and UN Models, and non-discrimination provisions in human rights instruments. The non-discrimination provisions in tax treaties identify a small number of specific situations where discrimination is prohibited under the domestic tax system: these include discrimination on grounds of nationality (art. 26(1) of the OECD Model), discrimination against stateless persons (art. 26(2) of the OECD Model), discrimination in the taxation of permanent establishments (art. 26(3) of the OECD Model), discrimination in respect of deductions paid to residents of the other contracting state (art. 26(4) of the OECD Model), and discrimination against enterprises owned by residents of the other contracting state (art. 26(5) of the OECD Model). At the same time, the non-discrimination provisions in article 14 ECHR and article 26 ICCPR differ significantly in their form. Article 14 ECHR33 is a non-free-standing non-discrimination provision: that is, persons are not to be discriminated against only with respect to the enjoyment of the various rights and freedoms secured by the ECHR. Article 26 ICCPR, however, is a broader, non-discrimination guarantee against discriminatory treatment more generally. Article 14 provides a level of general protection against discriminatory taxation when it is joined with the right to enjoyment of property in article 1 of the First Protocol. One can conceive of various ways in which a taxpayer might argue that the provisions of a double taxation convention discriminate against them. In virtually all situations, however, the government concerned would be able to justify any difference in treatment by pointing to the wide margin of appreciation enjoyed by states in tax matters. The most fundamental distinction in a double taxation convention is between persons who are residents of one contracting state and persons who are residents of the other state. Although in one tax case the ECtHR decided that a difference in treatment between a resident and a non-resident was not justifiable,34 the circumstances in that case were quite unusual. Given the fundamental distinction made in international tax law between the tax liability of residents and non-residents, there seems no real prospect that a double taxation convention might be faulted on this ground. There is one rather intriguing point here. Suppose that state A enters into tax treaties with both states B and C. Residents of state B are given a greater benefit under the 33

The text of art. 14 ECHR is as follows:

Article 14—Prohibition of discrimination The enjoyment of the rights and freedoms set forth in this Convention shall be secured without discrimination on any ground such as sex, race, colour, language, religion, political or other opinion, national or social origin, association with a national minority, property, birth or other status. 34

Darby v. Sweden (App. No. 11581/85), (1990) 13 EHRR 774.

Tax Treaties and Human Rights Law 429 convention than residents of state C—a lower withholding tax on dividends, for example. A resident of state C can show that they are in an objectively identical position to an equivalent resident of state B (save only for their place of residence), yet they are subject to a difference in tax treatment. Usually, the difference in treatment could be justified by pointing to the different course of the negotiations with states B and C, and the different priorities for the states arising from their different tax and economic systems. The difference is inherent in a system based on bilateral conventions. However, this is only a partial explanation for someone who is aggrieved by the failure to treat them in a fashion equivalent to their neighbour. This leaves the interesting question whether human rights instruments (or, more likely, constitutional guarantees) impose on a state the obligation to try to achieve a degree of commonality between the bilateral conventions it negotiates with other states. In the 2000s, the OECD commenced a review of the non-discrimination article in the Model. That review was overtaken, however, by a focus of work on the BEPS Project, and little progress was made on redrafting the non-discrimination article. Consideration was given to the non-discrimination principle in EU law. However, there is much to be said for comparing the non-discrimination articles in human rights instruments as a possible model for a new clause in tax treaties. For example, an article based on article 14 ECHR might look something like this: The enjoyment of the rights and freedoms set forth in this [Double Taxation] Convention shall be secured without discrimination on any ground such as sex, race, colour, language, religion, political or other opinion, national or social origin, association with a national minority, property, birth or other status.

This wording has the attraction that its scope is limited to the rights under the tax convention itself (as is the case for art. 14) so that it should be easier to foresee the exact impact of the article. This would then leave the broader issues of discrimination in the domestic tax systems to more general non-discrimination provisions. If, however, a wider scope of protection against all forms of unjustified discrimination were required, then a possible text might be based on the following: The contracting states undertake not to impose on persons resident in either contracting state any taxation which discriminates on any ground such as sex, race, colour, language, religion, political or other opinion, national or social origin, association with a national minority, property, birth or other status.

24.3.4. The Right to Property and Tax Treaties Article 1 of the First Protocol to the ECHR protects the right to property as follows:35 35

There is no provision in the ICCPR equivalent to art. 1 of the First Protocol to the ECHR.

430 Philip Baker Article 1—Protection of property Every natural or legal person is entitled to the peaceful enjoyment of his possessions. No one shall be deprived of his possessions except in the public interest and subject to the conditions provided for by law and by the general principles of international law. The preceding provisions shall not, however, in any way impair the right of a State to enforce such laws as it deems necessary to control the use of property in accordance with the general interest or to secure the payment of taxes or other contributions or penalties.

In general, tax treaties set out to relieve double taxation and so secure to taxpayers the enjoyment of property which they might have lost if they were doubly taxed. This raises the (again, rather intriguing) issue whether provisions equivalent to article 1 of the First Protocol impose on states a positive duty to avoid double taxation: that is, that unrelieved double taxation deprives a person of their property, and cannot be justified. The ECnHR and the ECtHR have said on several occasions that the imposition of an excessive and individual burden of tax on a person, such that it fundamentally interferes with that person’s financial position, may constitute an infringement of the right to enjoy one’s possessions.36 To date, however, cases where the taxpayer has shown that the provisions of any country’s tax laws infringe this principle are rare.37 Suppose, however, that the combined effect of two countries’ tax laws, including the absence of effective measures to relieve double taxation, have exactly that effect. Neither country has individually imposed an excessive burden; in combination, however, the domestic tax laws of the countries and the lack of effective means of relieving double taxation have resulted in an excessive burden. This is not to impose on states a positive duty to avoid an overlap in tax jurisdiction, but rather to ensure that their tax systems contains effective measures to relieve any double taxation which may result from claims to tax cross-border transactions. Arguably, the right to property and the prohibition of excessive and individual tax burdens impose at least some obligation on states to include unilateral provisions for the relief of double taxation in their laws or to seek to enter into a network of double taxation conventions. States would no doubt argue that it was within their margin of appreciation whether to include measures for relieving international double taxation. Given the widespread inclusion of such measures in countries’ laws,

36 See, e.g., Kaira v. Finland (App. No. 27109/95) (available on HUDOC) and Wasa Liv v. Sweden (App. No. 13013/87), 58 DR 163, 177–178. More recently, a series of cases from Hungary have involved a 98% tax on certain categories of income, which has been held to be excessive—see NKM v. Hungary (App. No. 66529/11), Gall v. Hungary (App. No. 49570/11), and RSz v. Hungary (App. No. 41838/11). 37 An unusual example was the case of Di Belmonte v. Italy (App. No. 72638/01) where a delay in payment of compensation meant that the payment was subject to a withholding tax which would not have been the case if the payment had been promptly made. The ECtHR held that the circumstances imposed an excessive and individual burden on the taxpayer concerned, and interfered with his right to property.

Tax Treaties and Human Rights Law 431 however, the position may more correctly be that states have a margin of appreciation in the approach to be taken to relieve double taxation (e.g. credit or exemption), but no discretion that some form of measures of relief are required to prevent the imposition of an excessive burden. Article 1 of the First Protocol also applies to measures for the enforcement of tax. The amendment to the OECD Model in 2005 to include new article 27 on assistance in collection of taxes raises the possibility that these measures contain inadequate protections for the rights of taxpayers. In particular, the provisions for taking protective measures or requiring the taxpayer to litigate the tax liability only in the country that imposes the tax raises the potential issue that a taxpayer may be deprived of their property without adequate protection. Much may depend on the particular circumstances of a particular taxpayer who would need to show that the measures taken in their case exceeded the margin of appreciation.

24.3.5 The Right to Leave a Country A similar point arises with respect to article 2 of the Fourth Protocol which provides in paragraph (2) that ‘Everyone shall be free to leave any country, including his own.’ Generally, tax treaties remove barriers to the free movement of persons that might otherwise arise if the person were subject to double taxation on moving country. For states that have signed this protocol, an obligation may arise to ensure that the combined effect of the tax system and any double taxation conventions is to remove any barriers to the free movement of persons. An interesting extension of this would be an argument that, for example, an exit tax might impede the exercise of the freedom to leave one’s country. A double taxation convention would not impose such an exit charge, but might relieve from such a charge by providing, for example, that the charge is deferred until disposal of an asset within, say, five years of moving countries. The right to leave one’s country might, therefore, bolster an argument for a positive measure to be taken in the Double Taxation Convention.

24.4 Concluding Comments The application of human rights conventions to tax matters in general is a relatively novel phenomenon: most of the case law of the ECtHR dates from the last thirty-five years. Many issues remain to be resolved, not least of which is the scope of application of certain key rights in a tax context, such as the right to a fair trial or the right to enjoyment of property. By contrast, tax treaties are older, with a history which now exceeds 100 years. Much of the work of developing the current Models took place before the development of the application of human rights norms to tax matters. It is appropriate to open

432 Philip Baker a discussion of the extent to which tax treaties might give rise to issues of compatibility with human rights norms, or whether changes might be made to the model conventions to ensure better conformity with those norms. More broadly, through prohibition of non-discrimination, for example, tax treaties may actively advance the practical implementation of human rights norms in the tax context.

Chapter 25

Taxation of Internat i ona l Partnersh i p s Ton Stevens

25.1 Introduction Many states divide their company law between corporations and partnerships. Partnerships have been widely used in business structures for many decades. In the literature1 at least three reasons are given why partnerships are preferred to other legal forms of business (e.g. corporations) in some countries or situations: (1) partnership law is in many countries more flexible than corporation law; (2) partnership law offers in many countries more legal forms to choose from than corporations; and (3) partnerships sometimes offer preferential tax treatment. With respect to the third reason, partnerships are quite often treated as fiscally transparent compared to corporations being separate tax subjects (opaque tax treatment). However, not all states have the same tax treatment for partnerships. Such different tax treatment by states delivers (very) complex tax issues for those partnerships that operate internationally. The taxation of such international partnerships is the subject of this chapter. For the purposes of this chapter, we define international partnerships as partnerships incorporated under a state’s company law2 that operate internationally in the sense that they operate in a

1 See

F. Haase, ‘Introduction’, in F. Haase, ed., Taxation of International Partnerships (Amsterdam: IBFD, 2014), 3–5. 2 Compare the definition from the 1999 OECD Partnership Report, para. 2: ‘the references to ‘partnerships’ in this report cover entities that qualify as such under civil or commercial law as opposed to tax law’. Other more elaborated definitions can be derived from J.-P. Le Gall, ‘General Report: International Tax Problems of Partnerships’, Cahiers de Droit Fiscal International vol. LXXXa (1995), 657 and D. Jimenez, and F. Borrego, ‘Legal Personality, Limited Liability and CIT liability’, in D.

434 Ton Stevens number of countries and/or have partners in other states than the residence state of the partnership. Unless otherwise indicated, the partnership carries on a business: for that reason, investment funds, trusts, and/or other contractual relationships are outside the scope of this chapter. Here we will first discuss the domestic tax treatment of international partnerships (Section 25.2), followed by the tax treaty treatment of international partnerships (Section 25.3). Finally, the EU law aspects of international partnerships are further analysed (Section 25.4).

25.2 Domestic Tax Treatment of International Partnerships 25.2.1 General In many states, there is a division between corporations that pay separate corporate income tax (CIT) (tax opacity) and partnerships that are treated as fiscally transparent (tax transparency). Many states also use this division to classify domestic entities as well as equivalent foreign entities. However, although in general this may be true, it is also an over-simplification of reality.3 In this section, therefore, there is a brief description of the different classification methods for domestic partnerships (Section 25.2.2) and foreign partnerships (Section 25.2.3) are discussed. In Section 25.2.4, the issue of fiscal transparency is further analysed.

25.2.2 Classification of Domestic Partnerships States use different criteria for defining CIT subjects.4 Depending on the use of different criteria, states may also treat their domestic partnerships differently (opaque or fiscally transparent). Systematically, at least four categories of tax treatment of partnerships can be distinguished: (1) states that treat all (domestic) partnerships as fiscally transparent (e.g. because (domestic) partnerships lack legal personality or limited liability); (2) states that treat all (domestic) partnerships as opaque (e.g. because (domestic) partnerships have legal personality); Gutman, ed., Corporate Income Tax Subjects, EATLP International Tax Series vol. 12 (Amsterdam: IBFD, 2015), 18. 3

4

See D. Gutman, ‘General Report’, in Gutman, Corporate Income Tax Subjects, 1. See Gutman, ibid., 2.

Taxation of International Partnerships 435 (3) states that treat (domestic) partnerships partly as fiscally transparent and partly as opaque (e.g. different treatment for the general partner(s) share(s) and for the limited partner(s) share(s)); and (4) states that grant (domestic) partnerships the possibility of opting in or opting out of CIT liability.

25.2.3 Classification of Foreign Partnerships The classification of foreign entities (including partnerships) need not necessarily use the same method as the classification method used for the classification of domestic entities (including partnerships). In this section, the different methods used for the classification of foreign entities (including partnerships) are further analysed. Generally,5 states use one of the following three methods to classify foreign entities (including partnerships): (1) The so-called similarity approach. Under this approach, normally, on the basis of the private law characteristics of the foreign entity an equivalent domestic entity is sought. If an equivalent domestic entity is found, the foreign entity receives the same tax treatment (opacity or transparency) as the domestic equivalent. A relatively large number of states (e.g. the UK, Germany, Belgium, Luxembourg, Netherlands, Austria) use this method. This method is rather time-consuming and may lead to legal uncertainty,6 certainly in situations where no domestic equivalents exist. (2) An optional system. Within optional systems, taxpayers (the entity or partnership) may themselves choose to be treated as opaque or as fiscally transparent. Although a number of states have optional elements in their classification system, only the USA uses an optional system both for domestic and foreign entities (check-the-box system). An optional system clearly provides for a very flexible system and also delivers major opportunities for tax planning. (3) The so-called fixed approach. Under this approach, all foreign entities are classified in the same manner, for example as opaque, regardless of the classification of equivalent domestic entities. This is, for instance, the classification method used in Italy for foreign entities. The method is rather simply to handle for taxpayers and provides, in that respect, for legal certainty. However, in the EU

5 See J. Barenfeld, Taxation of Cross- Border Partnerships, IBFD Doctoral Series vol. 9 (Amsterdam: IBFD, 2005), s. 3.5, 103–126 and L., Parada, Double Non-Taxation and the Use of Hybrid Entities, Wolters Kluwer Series on International Taxation vol. 66 (Alphen aan den Rijn: Kluwer Law International, 2018), s. 3.3, 109–156. 6 See B. Peeters, ‘Classification of Foreign Entities for Corporate Income Tax Purposes’, in Gutman, Corporate Income Tax Subjects, 66.

436 Ton Stevens context questions can be raised about the EU compatibility of this method because it might easily treat foreign entities differently from domestic entities.7 All three methods have in common that the classification by the other state is irrelevant under the domestic classification method. Some states, for example Sweden, Spain, and Denmark, have elements of the so-called symmetrical approach in their classification method; however, until now no state uses this symmetrical approach as its sole classification method for the classification of foreign entities. Under this approach, a state recognizes (mutual recognition) the tax classification of an entity’s foreign home state for domestic tax purposes. By using this symmetrical approach, classification conflicts are resolved; for that reason, in the literature8 this approach is quite often favoured for resolving the negative effects (double taxation or double non-taxation) of classification conflicts. The differences in classification (classification conflicts) by states is the source of double and double-non-taxation of hybrid entities. A hybrid entity9 is an entity that is classified differently by two or more states (either as opaque or as fiscally transparent). Within both the OECD and the EU, several solutions are introduced to mitigate or abandon the negative consequences of classification conflicts and other hybrid mismatch arrangements. These solutions are further discussed in Sections 25.3.3.3 and 25.4.3.

25.2.4 Tax Treatment of Fiscally Transparent Partnerships 25.2.4.1 General In the event that two (or more) states classify a partnership as fiscally transparent, mismatches can occur because not all states use the same concept of fiscal transparency. For that reason, in this section the different concepts of fiscal transparency are discussed first (Section 25.2.4.2), before the tax treatment of guaranteed payments is further analysed (Section 25.2.4.3).

25.2.4.2 The different concepts of fiscal transparency In the literature,10 at least four concepts of fiscal transparency are distinguished with two extreme positions. The first concept accepts all consequences of the tax transparency concept and treats the partnership as a mere conduit (aggregate approach). The other extreme accepts only one consequence of the tax transparency concept in the sense that 7

See ibid., 66–68 and see Section 25.2. See Sections 25.3.3.3.3 and 25.4.3.4. 9 In the literature, a division is made between regular hybrid entities (HE) and reversed hybrid entities (RHE) depending on the classification by the state of residence of the hybrid entity (HE: opaque, RHE: fiscally transparent). 10 Le Gall, ‘General Report’, 662–664. 8

Taxation of International Partnerships 437 the partnership is treated as a separate entity but the tax liability is transferred from the partnership to the partners (entity approach). In between are some intermediate systems under which the degree to which the partnership is tax transparent seems to depend on its function (intermediate approach). Sometimes, the partnership only serves as an administrative entity that books expenses and income on behalf of the partners and files the corresponding tax return (administrative function). In other cases, however, the partnership is considered to be a separate enterprise, held jointly by its partners, which declares its own profit, but the partners pay the taxes due on their share of the profits of the partnership.

1.2.5 Tax Treatment of Guaranteed Payments Some states apply special rules to payments made by a partnership to its partners independent from their profit-share (so-called guaranteed payments). In the literature,11 guaranteed payments are defined as ‘payments which a partner, as a contractor of the partnership (lessor, lender, licensor, etc.) is entitled to receive, independently from the partnership’s profit or loss’. Some states treat such guaranteed payments received by such partners as part of the partnership profit-share (e.g. as business income if the partnership carries on a business); other states treat such guaranteed payments as separate (rental, interest, royalty, etc.) income independent of the profit-share. Also in the latter case, some states (e.g. Germany) add back the separate income to the profit-share of the partnership received in a second stage (legal fiction as business income).

25.3 Tax Treaty Treatment of International Partnerships 25.3.1 General According to article 1 of the OECD Model Convention (MC)12 only persons who are residents of one or both of the contracting states can invoke the tax treaty. In relation to the tax treatment of international partnerships, it should, therefore, be further investigated in which situations partnerships can be considered to be persons and residents within the meaning of the OECD MC (Section 25.3.2). Classification conflicts13 cause international partnerships to become hybrid entities, the tax treaty treatment of which causes many complex situations, quite often resulting in cases of double taxation 11

Ibid., 703–704. Art. 1(1) OECD MC 2017. 13 See Section 25.2.3. 12

438 Ton Stevens or double non-taxation. In the past, the OECD has tried to provide some solutions both in the OECD Commentary (OECD 1999 Partnership Report14) and, recently, also in the OECD MC (the introduction by the OECD of the new art. 1(2) in the 2017 OECD MC). The tax treaty treatment of hybrid entities as well as the OECD solutions are further discussed in Section 25.3.3. Aside from classification conflicts, complex treaty qualification issues might arise in the case of international partnerships, especially regarding guaranteed payments. These issues are further discussed in Section 25.3.4.

25.3.2 Treaty Entitlement of International Partnerships Only persons who are residents are entitled to invoke the tax treaties that are modelled on the OECD MC (art. 1(1) OECD MC 2017). The first question to be answered with respect to the treaty entitlement of international partnerships is, therefore, in which situations partnerships qualify as persons and residents under the OECD MC. Both ‘persons’ and ‘residents’ are defined terms in the OECD MC. According to article 3(1) (a) OECD MC, the term ‘person’ includes an individual, a company, and any other body of persons. The term ‘company’ is defined in article 3(1)(b), whereas the term ‘any other body of persons’ is not defined in the OECD MC itself. Article 3(1)(b) defines ‘company’ as any body corporate or any entity that is treated as a body corporate for tax purposes. Although, the text of article 3(1)(b) is somewhat ambiguous, the OECD Commentary15 makes clear that only entities—body corporates or not—that are treated as a body corporate for the purposes of the tax law of the contracting state of which it is a resident, qualify as companies under the OECD MC. This means that only partnerships that are treated as opaque by the state where they are resident can be considered to be companies. Nevertheless, fiscally transparent partnerships can also qualify as persons (art. 3(1)(a)), because they qualify as ‘any other body of persons’. Although not undebated in the literature,16 the OECD clarified this position through the 2000 change to the OECD Commentary on article 3 as a result of the OECD Partnership Report. Since the 2000 change, paragraph 2 of the Commentary on article 3 (last sentence) reads as follows: Partnerships will also be considered to be ‘persons’ either because they fall within the definition of ‘company’ or, where this is not the case, because they constitute other bodies of persons. The partnership’s country of organization (i.e. the country under

14 OECD, The Application of the OECD Model Tax Convention to Partnerships, Issues in International Taxation, no. 6 (Paris: OECD Publishing, 1999). 15 2017 OECD Commentary on art. 3, para. 3. 16 For an overview of the different positions, see, e.g., B. Peeters, (2011), Fiscale transparantie: toerekening van inkomsten (Gent: Larcier, 2011), 460–461; G. Toifl, Personengesellschaften im Recht der Doppelbesteuerungsabkommen, Schriftenreihe zum Internationalen Steuerrecht Band 22 (Vienna: Linde, 2003), s.2.2, 43–54; and R. Muttaqin, ‘Hybrid Entities in the pre-BEPS Era: Impact of the OECD Partnership Report’, in S. Govind, and J.-P. Van West, eds, Hybrid Entities in Tax Treaty Law (Vienna: Linde, 2020), 39–40.

Taxation of International Partnerships 439 whose laws the partnership is organized or in which it is established) will then determine whether the partnership is a person within the meaning of the tax treaty.

Qualification as a person is not sufficient to provide international partnerships with treaty entitlement. In order to receive treaty entitlement, international partnerships should also qualify as residents within the meaning of article 4(1) OECD MC. Under article 4(1), the term ‘resident of a Contracting State’ means any person who, under the laws of that state, is liable to tax therein by reason of their domicile, residence, place of management, or any other criterion of a similar nature. With respect to international partnerships, the most relevant test seems to be whether the partnership is ‘liable to tax’ in the contracting state. This term, however, is heavily disputed in the literature.17 The prevailing view seems to be that a person who is ‘liable to tax’ means a person who is subject to full taxation (i.e. worldwide taxation: unlimited tax liability), despite the fact that taxes might or might not be effectively applied (e.g. through an exemption) in that state. Since the 2000 change to the Commentary, resulting from the 1999 OECD Partnership Report, the Commentary gives some guidance on when partnerships should be considered ‘liable to tax’. Paragraph 8.1318 of the Commentary on article 4 states: ‘Where a State disregards a partnership for tax purposes and treats it as fiscally transparent, taxing the partners on their share of the partnership income, the partnership itself is not liable to tax and may not, therefore, be considered to be a resident of that State.’ Generally speaking, from this clarification of article 4 it can be derived that partnerships that are treated as opaque by their residence state will qualify as resident, and partnerships that are treated as fiscally transparent by their residence state will not.19 However, the term ‘fiscally transparent’ is not further defined and, as we have seen (Section 25.2.4.2), states use different concepts of fiscally transparency. In the 1999 report, the OECD had already given some further clarification on how to interpretate the term ‘fiscally transparency’ in relation to the residency issue of partnerships. The Committee first decides20 between two common approaches to the taxation of partnerships21 and then states22 that there is agreement within the Committee that ‘for purposes of determining whether a partnership is liable to tax, the real question is whether the amount of tax payable on the partnership income is determined in relation to the personal characteristics of the partners’. In 2017, that clarification was added in a slightly adapted form to the OECD Commentary23. This clarification first states that the concept of ‘fiscally transparent’ used in the OECD Commentary refers to situations where, under the domestic law 17 Parada, Double Non-Taxation and the Use of Hybrid Entities, 69 and J. Gooijer, Tax Treaty Residence of Entities (Ede: GVO, 2019), s. 5.5, 127–172. 18 2000 OECD Commentary on art. 3, para. 8.3. 19 See Section 1.3.2.2 for the OECD ‘look-through’ rule in the case of fiscally transparent partnerships. 20 Partnership Report, paras 38–39. 21 In fact, being the two different variations to the intermediary approach as discussed in Section 25.2.4.2. 22 Partnership Report, para. 40. 23 2017 OECD Commentary on art. 1, para. 9.

440 Ton Stevens of a contracting state, the income (or part thereof) of the entity or arrangement is not taxed at the level of the entity or the arrangement but at the level of the persons who have an interest in that entity or arrangement. Furthermore, the Commentary mentions some further elements to verify whether or not an entity (including partnerships) can be considered fiscally transparent: (1) income legally earned by an entity is fiscally attributed to the underlying members and/or partners; (2) the income is aggregated to the overall income of the latter according to their relevant personal circumstances; (3) the relevant tax is reduced by any personal deductions and allowances; (4) the final tax must be determined as if the member and/or partner had earned that income; and (5) the character and source, as well as the timing of the realization, of the income for tax purposes will not be affected by the fact that the income that is computed at the level of the entity or arrangement before the share is allocated to the person will not affect its qualification as being ‘fiscally transparent’. Under these elements, although of course not all uncertainty has been clarified,24 the aggregate and intermediary approaches as identified in Section 25.2.4.2 seem to qualify as ‘fiscally transparent’ under the OECD Commentary description. The separate entity approach identified in Section 25.2.4.2 does not appear to qualify as ‘fiscally transparent’ under the Commentary description.

25.3.3 Tax Treaty Treatment of Hybrid Entities (Including Partnerships) 25.3.3.1 Introduction International partnerships are quite often classified (as opaque or fiscally transparent) differently by the two or more states involved (e.g. the state where the partnership is established and the state where the partners of the partnership are fiscally resident). Such classification conflicts might cause so-called allocation conflicts and/or qualification conflicts.25 Allocation conflicts are situations where two or more states allocate an item of income to different taxpayers (e.g. the partnership state allocates the income to the partnership (opaque treatment of the partnership), whereas the source state allocates the income to the partners of the partnership (fiscally transparent treatment of the partnership)). In the event of qualification conflicts, two or more states qualify an item of income differently based on their national tax treatment (e.g. in the case used in the previous example, the item of income is an interest payment from the source state to the 24

See, e.g., M. L. Brabazon, ‘Application of Tax Treaties to Fiscally Transparent Entities’, Global Tax Treaty Commentaries (Amsterdam: IBFD, 2020), s. 2, and M. Sutto, ‘Hybrid Entities: BEPS Action 2 and Article 1(2) of the 2017 OECD MTC’, in Govind and Van West, Hybrid Entities in Tax Treaty Law, 128–135. The uncertainties of the term ‘fiscally transparency’ with regard to the controlled foreign company rules, group taxation, and investment funds are outside the scope of this chapter. 25 Both allocation and qualification conflicts can also occur outside the situation of hybrid entities. In other words, allocation and qualification conflicts might have other causes than only classification conflicts.

Taxation of International Partnerships 441 partnership state and the partnership state might qualify the interest income as business income, while the source state qualifies the interest income as interest income). Both allocation and qualification conflicts might cause double taxation or double non-taxation. In the case of juridical double taxation, such a result is clearly not in accordance with the object and purpose of tax treaties. Therefore, in the 1990s the OECD had already set up a working group to study the treaty tax issues of hybrid entities (in the first instance only focusing on partnerships) and to provide for solutions to those issues. The first result was the OECD Partnership Report which was released in 1999. The findings of the report are further discussed in Section 25.3.3.2.

25.3.3.2 OECD Partnership Report 1999 25.3.3.2.1 Introduction In 1993, the OECD Committee on Fiscal Affairs formed a working group to study the application of the OECD Model to partnerships, trusts, and other non-corporate entities.26 In 1999, the OECD Committee published it first report on the application of the OECD Model to partnerships (the OECD Partnership Report).27 The report focuses on specific factual examples and its conclusions and proposed changes are reflected in the 2000 OECD Commentary. The proposed application rules are not reflected in the OECD MC, which is one of the major points of criticism about the report. Hereafter, the key points of the report (Section 25.3.3.2.2) as well as its limitations and criticism (Section 25.3.3.2.3) are discussed separately.

25.3.3.2.2 Key points of the Partnership Report The Partnership Report deals with eighteen practical examples and does not contain a separate chapter covering the principles or approaches used in it. However, from the practical examples and the proposed changes to the OECD Commentary it is possible to derive some key points.28 These points can be divided into three areas: (1) the tax treaty entitlement of partnerships; (2) the application of tax conventions by the state of source (allocation conflicts); and (3) the application of tax conventions by the state of residence (qualification conflicts and allocation conflicts). The tax treaty entitlement of partnerships With regard to the tax treaty entitlement of partnerships, the report starts with the analysis of two important questions: to what extent can partnerships be seen as ‘persons’ and ‘residents’ in the sense of the OECD MC? For this analysis, see Section 25.3.2. In the view of the report (as from 2000 laid down in the OECD Commentary), partnerships will always qualify as ‘persons’ under the OECD MC either because they are ‘companies’ (opaque partnerships) or ‘bodies of

26

For a historical overview, see e.g. Muttaqin, ‘Hybrid Entities in the pre-BEPS Era’, 28–29. The Application of the OECD Model Tax Convention to Partnerships, Issues in International Taxation no. 6 (Paris: OECD, 1999). 28 See, e.g., Muttaqin, ‘Hybrid Entities in the pre-BEPS Era’, 30–38. 27

442 Ton Stevens persons’(fiscally transparent partnerships). Fiscally transparent partnerships, however, cannot be ‘residents’ under article 4 OECD MC. The Partnership Report gives some further guidance on that point on the concept of ‘liable to tax’ with regard to (fiscally transparent) partnerships; on which, refer to Section 25.3.2. An important statement in the report in relation to the tax treaty entitlement of partnerships is that in the case where the partnership is a fiscally transparent entity, the partners are entitled to tax treaty benefits with respect to their share in the income of the partnership.29 Application of tax conventions by the state of source (allocation conflicts) In the case of allocation conflicts, there is a mismatch between the allocation of income and the treaty entitlement with respect to that income between two (or more) states. To avoid such mismatches and, therefore, to secure treaty benefits, the OECD introduces in the second part of the Partnership Report a new principle, sometimes referred to as the ‘OECD General Principle’30 or as the ‘OECD flow-through (or look-through) principle’. This principle links the obligations of the source state under the OECD MC to the allocation and treaty entitlement of a specific item of income by the residence state. The report notes that ‘the source State, in applying the Convention where partnerships are involved, should take into account, as part of the factual context in which the Convention is to be applied, the way in which an item of income arising in its jurisdiction is treated in the jurisdiction of the taxpayer claiming the benefits of the treaty as a resident’.31 In other words, the source state should take into account how the income arising in its jurisdiction is treated in the residence state. How the income is treated in the residence state will depend on how the residence state treats the partnership itself. If the residence state treats the partnership as opaque, such partnership is liable to tax and entitled to tax treaty benefits (regular hybrid entity). If the residence state treats the partnership as fiscally transparent, the income will be allocated to its partners, provided that the partners are also considered as residents of that state (reverse hybrid entity). This approach is reflected in e xamples 4–10 of the report and is also applied by the OECD in multilateral cases. Application of tax conventions by the state of residence (qualification conflicts and allocation conflicts) In addition, the application of tax treaties by the state of residence might raise difficulties where partnerships are involved (e.g. the application of art. 23 on elimination of double taxation). In the case of qualification conflicts (i.e. the situation where both the residence and source state apply different articles of the tax treaty with regard to the same item of income on the basis of differences in their domestic law), the wording of article 23 of the OECD MC ‘requires’—in the view of the Partnership Report—that the residence state in granting double tax relief (exemption or credit

29

Partnership Report, examples 1–3. See, e.g., Muttaqin, ‘Hybrid Entities in the pre-BEPS Era’, 33 and C. Bergedahl, ‘Hybrid Entities and the OECD Model (2017): The End of the Road?’, Bulletin For International Taxation (2018), 418. 31 Partnership Report, para. 53. 30

Taxation of International Partnerships 443 method) follows the qualification of the item of income by the source state (under the condition that the source state applies the treaty in accordance with the provisions of the MC correctly). Also, this new linking rule is reflected in the 2000 changes to the OECD Commentary32 and, to avoid double non-taxation, a new paragraph 4 was added to article 23A of the 2000 OECD MC. This new linking rule is illustrated by two partnership examples (dealing with guaranteed payments; on which, see Section 25.3.4); however, the changes to the 2000 Commentary and MC are general changes, also applicable beyond partnership issues. Finally, the Partnership Report deals with some allocation conflicts in the case of partnerships from the perspective of the residence state. Examples 16 and 17 deal with situations where a state is both the residence state of one of the partners or the partnership and source state. The question to be answered is than whether the OECD General Principle (see earlier) has to be followed or whether the General Principle has to be overruled by another principle that states are not restricted by tax treaties to tax their own residents on their share of the income of the partnership. The delegates of the working group that produced the report could not reach a conclusion, however, the majority was of the opinion that tax treaties should not restrict states to taxing their own residents on their share of the income of the partnership. This principle is also reflected in the 2000 Commentary on article 1,33 where it is, furthermore, added that some states may wish to include in their conventions a provision that expressly confirms a contracting state’s right to tax resident partners on their share of the income of a partnership that is treated as a resident of the other state. With respect to the elimination of double taxation resulting from allocation conflicts (caused by classification conflicts), the Partnership Report only deals with the situation of a ‘regular’ hybrid entity (example 18). The situation of a reverse hybrid entity is not discussed in the report. Example 18 only discusses the various difficulties with respect to the elimination of double taxation by the residence states, but does not provide for extended solutions. The only guidance that is also reflected in the 2000 OECD Commentary on article 2334 obliges the residence state of the partner to credit the foreign corporation tax paid by the hybrid entity (‘flow-through’ approach of creditable taxes coherent to the ‘flow-through’ taxation of the income).

25.3.3.2.3 Limitation and criticism of the Partnership Report Although the Partnership Report is clearly a major step in clarifying tax treaty issues that arise with the taxation of international partnerships, the report still received a great deal of criticism in the literature.35 The major points of criticism can be summarized as follows:

32

See OECD Commentary on art. 23, para. 32. See 2000 OECD Commentary on art. 1, para. 6.1. 34 See ibid., art. 23, para. 69.2. 35 See the reference to the literature in Muttaqin, ‘Hybrid Entities in the pre-BEPS Era’, 42–46. 33

444 Ton Stevens (1) The OECD General Principle lacks justification. It forces the source state to follow the allocation of income by the residence state for the application of the distributive rules of the OECD MC. However, there is no justification in the report for this preference of the residence state allocation compared to the source state allocation36 (2) The report is internally inconsistent and incomplete. It is considered inconsistent in two ways. First, the OECD General Principle is not applied in all cases;37 in cases where a state is both the residence state of one of the partners or the partnership and the source state, such state is not limited by the OECD General Principle to tax its residents on the income derived by or through the partnership. Secondly, under the OECD General Principle, the allocation by the residence state is favoured above the allocation by the source state. However, as a solution for qualifying conflicts caused by differences of the application of two states’ national laws, the opposite choice has been made (the source state’s qualification has to be followed by the residence state in granting double tax relief). The report is considered to be incomplete because the double taxation caused by allocation conflicts is not resolved in a considerable number of cases.38 (3) Key points of the report are not reflected in the OECD Model provisions, ‘only’ in the OECD Commentary. (4) The report is limited to partnerships and does not deal with other hybrid entities (e.g. trusts) or with investment funds. (5) The interaction between the OECD General Principle and the other distributive rules and concepts of the OECD MC (e.g. the concept of beneficial ownership) has been regarded as unclear and rather incoherent. (6) The report does not deal with the place of establishment non-discrimination rule.39 The huge amount of criticism also seems to be reflected by the relatively high number of reservations made to the Partnership Report.40 After its release, a number of countries continued or started to insert special tax treaty clauses, mainly focusing on the OECD General Principle. Quite often, these special tax treaty clauses are modelled on article 1(6) of US Model Tax Convention. In court decisions, the OECD General Principle does not seem to be widely accepted by the national courts.41 36 M. Lang, The Application of the OECD Model Tax Convention to Partnerships, Schriftenreihe zum Internationalen Steuerrecht Band 11 (Vienna: Linde, 2000), 37–39. 37 See examples 16 and 17 as discussed in Section 25.3.3.2.2. 38 e.g. the double taxation caused by the exceptions to the OECD General Principle (examples 16 and 17) and in the allocation conflicts caused by classification conflicts from a residence state perspective (example 18 and the lack of an example for reverse hybrid entity cases). 39 See, e.g., K. Jain, ‘The OECD Model (2017) and Hybrid Entities: Some Opaque Issues and Their Transparent Solutions’, Bulletin for International Taxation (Mar. 2019), s. 2.5, 134–135. 40 See Partnership Report, Annex II. 41 For an overview, see e.g. T. Meyer, ‘Tax Treaty Entitlement to Hybrid Entities: An Overview of Domestic Court Jurisprudence’, in Govind and Van West, Hybrid Entities in Tax Treaty Law, 51–70.

Taxation of International Partnerships 445

25.3.3.3 BEPS Action 2 Report: the new article 1(2) OECD MC 25.3.3.3.1 Introduction The use of hybrid entities and the resulting hybrid mismatch arrangements (HMAs) have been further analysed by the OECD in the Base Erosion and Profit Shifting (BEPS) Project. The 2015 BEPS Action 2 Report42 provides for a two-part solution to counteract the negative consequences of HMAs: in Part I a number of recommendations for domestic law are provided (the so-called ‘linking rules’43) and Part II lays out a set of recommendations on tax treaty issues. In chapter 14 of the 2015 BEPS Action 2 Report, a new treaty provision on transparent entities (art. 1(2) OECD MC) is proposed, which is one of the changes that took place in the 2017 version of the OECD MC. Although the new provision also limits treaty entitlement in cases of reverse hybrid structures, its scope is much wider than a treaty solution for some specific HMAs. In fact, the OECD General Principle of the Partnership Report as a solution for allocation conflicts from a source state perspective is now incorporated in the new treaty provision. Also, in order to change existing tax treaties a similar provision has been taken up in the Multilateral Instrument (MLI) (art. 3(1) MLI).44 Hereafter, the scope and the remaining limitations as well as criticism of the new article 1(2) OECD MC are further analysed.

25.3.3.3.2 The scope of new article 1(2) OECD MC Article 1(2) OECD MC reads as follows: For the purposes of this Convention, income derived by or through an entity or arrangement that is treated as wholly or partly fiscally transparent under the tax law of either Contracting State shall be considered to be income of a resident of a Contracting State but only to the extent that the income is treated, for purposes of taxation by that State, as the income of a resident of that State.

From the OECD Commentary on article 1(2),45 it can be derived that this new provision addresses the situation of the income of entities or arrangements that one or both contracting states treat as wholly or partly fiscally transparent for tax purposes. The provisions of the new paragraph 2 should ensure that the income of such entities or arrangements is treated, for the purposes of the convention, in accordance with the principles reflected in the 1999 report. It is not clear which of the principles of the report are meant by this reference in the Commentary, but from the wording and place

42

OECD/G20, Base Erosion and Profit Shifting Project, Neutralizing the Effects of Hybrid Mismatch Arrangements, Action 2—Final Report (2015). 43 The EU has taken over these linking rules almost completely in the ATAD; see Section 25.3.3. 44 See, e.g., K. Kajiwara, ‘The Coverage of Hybrid Entities under the MLI: Interpretation of the Provisions and Analysis of Opting Countries and Consequent Changes to Tax Treaties’, in Govind and Van West, Hybrid Entities in Tax Treaty Law, 189–213. 45 OECD Commentary on art. 1, para. 2.

446 Ton Stevens of the new provision it can be derived that the provision tries to incorporate the OECD General Principle as explained in Section 25.3.3.2.2. The Commentary continues by stating that the ‘report therefore provides guidance and examples on how the provision should be interpreted and applied in various situations’. Although not mentioned in the Commentary, the text of article 1(2) is clearly inspired by article 1(6) of the US Model Tax Convention.46 With respect to the principles laid down in the Partnership Report, the new treaty provision of article 1(2) OECD MC has a broader scope in two respects: first, the new provision applies not only to partnerships but to all ‘entities or arrangements’. Secondly, it is (further) clarified that the provision also applies in the case where one of the contracting states treats the entity or arrangement only partly as fiscally transparent. Through this broadening of its scope, the new provision is now also applicable (e.g. to trusts) and it is also clear that some partnerships that are treated partly as fiscally transparent and partly as opaque47 are covered by the provision (‘to the extent that’ they are treated as fiscally transparent). The exception to the OECD General Report as laid down in the OECD Partnership Report (examples 16 and 17) is in the 2017 OECD MC incorporated into a new article 1(3) (saving clause).48 Furthermore, the interlinkage between the saving clause and the relief of double taxation under article 23 OECD MC is further ‘clarified’ by a set of examples in the Commentary on article 23.49 However, even after these clarifications, double taxation arising out of conflicts of residence are left for the larger part unsolved.

25.3.3.3.3 Limitation and criticism of article 1(2): other solutions The criticism in the tax literature on the solutions in the Partnership Report (see Section 25.3.3.2.3) are in two respects resolved by the introduction of article 1(2) OECD MC (2017): the scope of the provision is extended and the OECD General Principle is now shifted from the Commentary to the OECD MC itself. Also, the exception to the OECD General Principle is now incorporated in the OECD MC itself (art. 1(3): saving clause) and the interplay between that exception and double tax relief is further clarified.

46

For a comparison between the two model treaty provisions, see e.g. A. Sutto, ‘The Tax Treatment of Hybrid Entities under the BEPS Project: Action Plan 2 and Article 1 (2) of the 2017 OECD Model Convention’, in Govind and Van West, Hybrid Entities in Tax Treaty Law, 123–125. 47 e.g. the Dutch so-called open CV and some countries’ partnerships limited by shares. 48 See about the interlinkage between art. 1(2) and (3) OECD MTC, e.g. T. de Mattos Marques, ‘The Saving Clause and Hybrid Entities’, in Govind and Van West, Hybrid Entities in Tax Treaty Law, 141– 160, and A. Nikolakis et al., ‘Some Reflections on the Proposed Revisions to the OECD Model and Commentaries, and on the Multilateral Instrument, with Respect to Fiscally Transparent EntitiesPart 2’, Bulletin for International Taxation (2017), s. 6. 49 2017 OECD Commentary on art. 23, paras 11.1 and 11.2. These examples are extensively discussed, e.g., by A. Nikolakis et al., ‘Some Reflections on the Proposed Revisions’, s. 6; C. Bergedahl, ‘Hybrid Entities and the OECD Model (2017): The End of the Road?’, Bulletin for International Taxation (2018), s. 3.2, 421–428 and P. Peshori, ‘Obligation to Provide Double Taxation Relief in the Case of Hybrid Entities under the 2017 OECD Model, in Govind and Van West, Hybrid Entities in Tax Treaty Law, 333–356.

Taxation of International Partnerships 447 However, the other points of criticism and unresolved issues of the report are not answered by the new treaty provision(s). To resolve the remaining issues of the report and/or the tax treaty treatment of hybrid entities in the literature,50 a number of solutions have been proposed in the past. Although an extensive discussion clearly exceeds the scope of this chapter, the solutions can be divided between two types: on the one hand, solutions that provide for changes in national law and, on the other hand, changes at the tax treaty level. With respect to changes of national law, several authors51 suggest a domestic ‘interlinkage’ rule whereby the classification of the source (or residence) country is aligned with the classification given to the entity in its ‘home country’. Clearly, these solutions have the advantage that the source of the problem (classification differences) is tackled instead of ‘curing’ the negative consequences (double or non-taxation caused by allocation conflicts). The implementation of a domestic rule on a worldwide basis, however, is not very easy to accomplish. The treaty-based solutions vary between further building on the findings of the Partnership Report (including the new art. 1(2) OECD MC)52 and the development of a whole new system for the convention to be built on (treaty entitlement is shifted from persons to income).53 From a short-term perspective, the first-mentioned tax treaty solutions that build on the existing OECD tax treaty framework seem to be the easiest to realize.

25.3.4 Tax Treaty Treatment of Guaranteed Payments The tax treaty treatment of guaranteed payments delivers some additional issues in the case of international partnerships. Even in a case where both states treat the international partnership as fiscally transparent (no hybrid entity), differences in treaty qualification and the application of the distributive rules might easily lead to situations of double taxation or double non-taxation. Cases causing such qualification conflicts are separately discussed in e xamples 13 (double non-taxation) and 15 (double taxation) 50

See for an overview of the solutions provided for in literature, e.g. B. Ferreira Liotti, ‘Hybrid Entities in Tax Treaty Law: Issues, their Source and Possible Solutions’, in Govind and Van West, Hybrid Entities in Tax Treaty Law, 16–21. 51 See, e.g., G. Fibbe, EC Law Aspects of Hybrid Entities, IBFD Doctoral Series vol. 15 (Amsterdam: IBFD, 2009); Parada, Double Non-Taxation and the Use of Hybrid Entities; and C. Kahlenberg, ‘Hybrid Entities: Problems Arising from the Attribution of Income Through Withholding Tax Relief—Can Specific Domestic Provisions be a Suitable Solution Concept?’, Intertax 2 (2016), 146–162. 52 Nikolakis et al., ‘Some Reflections on the Proposed Revisions’, s. 7, 561 (and Annex A), provide for two alternative formulations to revise art. 1(2) OECD MC. K. Jain, ‘The OECD Model (2017) and Hybrid Entities: Some Opaque Issues and Their Transparent Solutions’, Bulletin for International Taxation (Mar. 2019), s. 2.7, 137–139, also provides for additional rules to grant double tax relief in the case of residence– residence conflicts. 53 J. Wheeler, The Missing Keystone of Income Tax Treaties, IBFD Doctoral Series vol. 23 (Amsterdam: IBFD, 2012).

448 Ton Stevens of the Partnership Report. The solution that the report provides for these qualification conflicts is found in the introduction of the new rule that the residence state, in granting double tax relief, is obliged to follow the qualification of the guaranteed payment by the source state. However, the rule is only applicable in those cases where qualification conflicts are caused by differences in the national tax laws of the contracting states. In the case of a difference in interpreting the facts or the application of the various distributive rules of the OECD MC, the negative consequences (double taxation or double non-taxation) can only be resolved by a MAP (art. 25 OECD MC) or special tax treaty rules. In the literature, this solution provided by the report is heavily debated.54 The criticism seems to be twofold. On the one hand, some authors do not find sufficient reasoning for the rule to be derived from the interpretation of article 23 OECD MC. On the other hand, a number of authors point out that the report uses the wrong starting point in stating that the application of different distributive rules is caused by differences in the national tax law of the contracting states. That starting point, in fact, favours the reference by article 3(2) to the applying country’s national law above the treaty context. Within that context, in the view of those authors, there is sufficient possibility for arriving at satisfying solutions for the issue. In the examples of the Partnership Report, two situations are dealt with where a partner has granted a loan to the partnership in which they participate as a partner. Under their national tax laws, the interest is in one state qualified as business income and in the other state as interest income. Within the OECD MC framework, the starting point of the application of the distributive rules should be, in such a case, the priority rule of article 7(4) which indicates that the interest payments by the (fiscally transparent) partnership to the foreign partner should be dealt with by article 11 (interest article). Assuming that the partnership has a permanent establishment in the state where it is located, the interest seems to ‘arise’ (art. 11(5)) from the partnership’s state, which leads to a division of taxation rights between the source state (limited taxation right based on art. 11(2)) and the residence state of the partner (taxation right based on art. 11(1) with the obligation to provide for double taxation relief (credit method) under art. 23). That interpretation can be made whether or not the two contracting states have a difference in qualification of the income under their national tax laws (business income vs interest income). However, based on article 7(4), the distributive rule of article 11 is not applicable, but instead that of article 7, in the case where the foreign partner has a permanent establishment in the source state and the debt- claim in respect of which the interest is paid is ‘effectively connected’ with such permanent establishment. In the literature, there is presently no consensus on how to apply the term ‘effectively connected’ in those cases of interest payments received by a foreign partner from a (fiscally transparent) partnership. Daniels55 defends the position by 54 See for an overview of the different views, e.g., C. Staringer, ‘Leistungsbeziehungen zwischen der Personengesellschaft und den Gesellschaftern aus abkommensrechtlicher Sicht’, in W. Gassner, M. Lang, and E. Lechner, Personengesellschaften im Recht der Doppelbesteuerungsabkommen (Vienna: Linde, 2000), 101–120. 55 T. Daniels, Issues in International Partnership Taxation (The Hague: Kluwer Law International, 1991), 176.

Taxation of International Partnerships 449 stating that in the event that interest ‘arises’ from the source state due to the fact that it is paid on behalf of the permanent establishment of the partnership, such interest and the corresponding debt-claim is always ‘effectively connected’ with the permanent establishment of the foreign partner. In the German literature,56 based on several decisions by the Bundesfinanzhof (Federal Fiscal Court),57 the possibility of a debt-claim being ‘effectively connected’ with a permanent establishment of the foreign partner is denied due to the fact that the interest costs are deductible at the level of the partnership. A third, intermediary, position is defended by Lang.58 He acknowledges the ‘effectively connectedness’ only to the extent that the partner receives interest from ‘himself ’ (the part that is deducted from his profit share) (art. 7 is applicable). To the extent that the partner receives interest at the expense of the profit-share of the other partners, in his view ‘effectively connectedness’ is not possible and article 11 remains applicable.

25.4 International Partnerships and EU Tax Law 25.4.1 General In this section several EU tax law issues concerning international partnerships are discussed. In Section 25.4.2, the tax treatment of international partnerships under primary EU law is further analysed. Tax treatment of international partnerships under secondary EU law is covered in Section 25.4.3. The focus of the following section lies with the situation that international partnerships qualify as hybrid entities.

25.4.2 Tax Treatment of International Partnerships under Primary EU Law Under primary EU law, two principal questions should be discussed regarding international partnerships that qualify as hybrid entities. Hybrid entities were defined earlier59 as the situation where two or more states (in this context, EU member states) classify an entity (the international partnership) differently (as opaque or as fiscally 56

See for an overview, e.g., ‘Staringer, Leistungsbeziehungen zwischen der Personengesellschaft’ 118– 119, and O. Rosenberg, ‘Einkünftequalification im Sonderbetriebsbereich’, in F. Wassermeyer, S. Richter, and B. Schnittker, eds., Personengesellschaften im Internationalen Steuerrecht (Cologne: Otto Schmidt, 2015), 587–631. 57 BFH 27 February 1991, BStBl. II, pp. 444–448, BFH 26 February 1992, BStBl. II, pp. 937–940, BFH 30 august 1995, BStBl. 1996 II, p. 563, and BFH 23 October 1996, BStBl. 1997 II, p. 313. 58 Lang, The Application of the OECD Model Tax Convention to Partnerships, 80–81. 59 See Section 25.3.3.1.

450 Ton Stevens transparent) based on their national methods of classification. These differences in classification might cause double taxation and double non-taxation. Initially, the first principal question to be answered is whether national classification methods could be in violation of the fundamental freedoms. As such, any tax measure, including a tax classification method, can be in violation of primary EU law. However, a violation should be analysed under the same four-step method applied by the European Court of Justice (ECJ) towards any other tax measure. Under such analysis, the general view in the literature60 seems to be that most classification methods used by the EU member states61 are not in violation of the fundamental freedoms, with the possible exception of the so-called fixed methods or approaches.62 Secondly, the next principal question is whether primary EU law, more specifically the principal of mutual recognition, can provide a solution for the negative consequences of classification conflicts. The principle of mutual recognition, which is embedded in the fundamental freedoms, prohibits member states from imposing their own more stringent or differing domestic law provisions if the qualification standards are met in the member state of origin. The relevant question in the case of hybrid entities is, therefore, whether that principle of mutual recognition would also apply to the fiscal status of foreign entities, which would mean that member states would be forced to apply the classification derived from the classification method of the state of origin of the hybrid entity. In the literature, however, the general view63 seems to be that due to the decision by the ECJ in the Columbus Container Services case,64 the principle of mutual recognition cannot be applied to the fiscal status of foreign entities.

25.4.3 Tax Treatment of International Partnerships under Secondary EU Law 25.4.3.1 General In this section, the treatment of international partnerships that qualify as hybrid entities under some EU directives (the Parent–Subsidiary (P–S) Directive, the Interest and 60

See for an overview, R. Baete, ‘The Tax Treatment of Hybrid Entities under Primary EU Law and CJEU Jurisprudence’, in Govind and Van West, Hybrid Entities in Tax Treaty Law, 509–530. 61 See Section 25.2.3; Baete, ‘The Tax Treatment of Hybrid Entities’, 511–512. 62 See Section 25.2.3; Baete, ibid., 525–526. Normally, only cases of double taxation could provide for a potential restriction of the fundamental freedoms. Cases of double non-taxation could potentially provide for a violation of EU state aid law (art. 107 TFEU). See, for this issue and further literature references, Baete, 528–530. 63 See, e.g., G. K. Fibbe, EC Law Aspects of Hybrid Entities, IBFD Doctoral Series vol. 15 (Amsterdam: IBFD, 2009), 167–73; G. K. Fibbe and A. J. A. Stevens, ‘The Approach to Hybrid Entities under Primary EU Law and in the EU Tax Directives’, in G. K. Fibbe and A. J. A Stevens,. eds, Hybrid Entities and the EU Direct Tax Directives (The Hague: Kluwer Law International, 2015), 7–11; and Baete, ‘The Tax Treatment of Hybrid Entities’, 522–523. 64 CJEU, Case C-298/05 Columbus Container Services BVBA & Co., ECLI:EU:C:2007:754, paras 53–54.

Taxation of International Partnerships 451 Royalties (I&R) Directive, and the Anti-Tax Avoidance Directive) are further analysed.65 In addition, the Merger Directive66 contains some specific rules regarding the application of the directive on mergers and/or reorganizations involving hybrid entities, but an in-depth analysis of these rules is outside the scope of this chapter.67

25.4.3.2 Parent–Subsidiary Directive and Interest and Royalties Directive Since the 2003 amending directive,68 the P–S Directive has contained a special rule for hybrid entities in outbound situations (art. 4(2) of the P–S Directive). Under this rule, the member state of the parent company can uphold its classification as a subsidiary in another member state, however, it must refrain from taxing the distributed profits of the subsidiary and must grant either an exemption or allow a tax deduction for corporate tax that has been paid by the subsidiary. The member state where the subsidiary is located is not permitted to tax the profit distributions by the qualifying subsidiary to the parent company in the other member state. For inbound situations, the P–S Directive does not provide for any specific rules for hybrid entities. In the I&R Directive,69 there is no explicit guidance with respect to the application of hybrid entities. However, in the 2009 evaluation report70 in respect of the I&R Directive, the European Commission seemed to be of the opinion that there is no legal basis in the provisions of the directive that would allow a source state to ‘look-through’ a qualifying entity resident in another member state when that entity meets the requirements of article 3 of the I&R Directive in its member state of residence. In cases where the ‘look- through’ approach would be permissible, the benefits of the I&R Directive should be extended to the shareholders of the intermediary entity. This approach is relatively comparable with the OECD General Principle71 as laid down for inbound cases in article 1(2) of the 2017 OECD MC. For this reason, in the literature72 it is proposed to further implement the OECD General Principle both in the P–S and the I&R Directives.

65 The European Economic Interest Grouping as the sole EU partnership is left out of the scope of this chapter. 66 Council Directive 90/434/EEC of 23 July 1990 [1990] OJ L225/1. 67 See for an in-depth analysis, e.g., G. F. Boulogne, Shortcomings in the EU Merger Directive, Series in international taxation vol. 57 (The Hague: Kluwer Law International, 2016), 174–176 and 215–231, and P. Matthieu, ‘Hybrid Entities and their Treatment under the Merger Directive’, in Govind and Van West, Hybrid Entities in Tax Treaty Law, 555–577. 68 Council Directive 2003/123/EC of 22 December 2003 [2004] OJ L7/41. 69 Council Directive 2003/49/EC of 3 June 2003 [2003] OJ L157/49. 70 European Commission, ‘Report from the Commission to the Council in accordance with Article 8 of Council Directive 2003/49/EC on a common system of taxation applicable to interest and royalty payments made between associated companies of different Member States’, COM (2009) 179 final (17 Apr. 2009), para. 3.3.5.2. 71 See Section 25.3.3.2.2. 72 See, e.g., Fibbe and Stevens, Hybrid Entities and the EU Direct Tax Directives, 260– 265 and S. Picariello, ‘Hybrid Entities and Their Treatment under the Parent–Subsidiary Directive and the Interest and Royalty Directive, in Govind and Van West, Hybrid Entities in Tax Treaty Law, 552.

452 Ton Stevens

25.4.3.3 ATAD With the Anti-Tax Avoidance Directive (ATAD) as extended by ATAD II,73 the European Commission took over the OECD BEPS approach to tackle the negative consequences (double non-taxation) caused by hybrid mismatch arrangements within multinational enterprises (MNEs). The ATAD introduces two sets of measures: (1) a set of ‘linking rules’ for situations of double deduction (DD) and deduction without inclusion (D/NI) in article 9 ATAD; and (2) a recharacterization rule for reverse hybrid situations in article 9a ATAD. The linking rules include situations where payments are made by and/or to hybrid entities (and payments made by and/or to and/or between permanent establishments) (art. 2(9) ATAD) that might be of relevance for international partnerships. The linking rules (art. 9(1) and (2) ATAD) deny deductions and/ or force inclusion of otherwise untaxed income using the OECD BEPS Action 2 ‘primary’ and ‘defensive’ rules, thus ‘only’ eliminating the negative consequences (DD or D/NI) of HMAs and not providing a solution for the underlying source of the problems (qualification and/or classification conflicts).74 The recharacterization rule of article 9a(1), however, provides for a solution for the underlying classification conflict in the case of reversed hybrid entities (including international partnerships): the (EU member state) country of residence is forced to regard the reversed hybrid entity as a resident for tax purposes and to tax the income of that entity to the extent that that income is not otherwise taxed under the laws of the member state or any other jurisdiction where its members/shareholders are located. Although under this recharacterization rule the classification conflict is resolved (both the country where the entity is resident and the countries where the members/shareholders are resident treat the entity as opaque), from a conceptual viewpoint it can be debated whether this solution (country of residence of the entity is obliged to follow the classification of the country of its members/ shareholders) is a convincing approach.

25.4.3.4 Alternatives As Section 25.4.2 revealed, primary EU law does not provide solutions for the negative consequences (double and double non-taxation) of hybrid entities. Secondary EU law does provide some solutions,75 but mainly for double non-taxation situations (ATAD). However, to a large extent only the negative consequences are repaired by the introduction of linking rules and, therefore, no fundamental solution is found for

73 Council Directive (EU) 2017/952 of 28 May 2017 amending Directive (EU) 2016/ 1164 regarding hybrid mismatches with third countries [2017] OJ L144/1. 74 See, e.g., B. Erbetta, ‘Hybrid Mismatches under ATAD II’, in Govind and Van West, Hybrid Entities in Tax Treaty Law, 606–607. 75 The linking rules in the latest Common Consolidated Corporate Tax Base proposal of the European Commission (arts 61–63 of the European Commission, Proposal for a Council Directive on a Common Corporate Tax Base, COM (2016) 685 final) are not further discussed in this chapter. See, for a discussion of these linking rules, e.g. T. Anil, ‘A Uniform Classification System for Hybrid Entity Mismatches under EU Law’, in Govind and Van West, Hybrid Entities in Tax Treaty Law, 660–661.

Taxation of International Partnerships 453 the problem itself. For that reason, the literature76 has proposed the introduction of a new EU directive on the establishment of a uniform classification system for entities for tax purposes. Under such directive classification conflicts between EU member states would be resolved by a uniform classification system. However, even under such a new EU directive, classification conflicts would remain unresolved in relation to third states.

76 G. F. Fibbe, EC Law Aspects of Hybrid Entities (Amsterdam: IBFD, 2009), 323–359 and T. Anil, ‘A Uniform Classification System for Hybrid Entity Mismatches under EU Law’, in Govind and Van West, Hybrid Entities in Tax Treaty Law, 648–662.

Chapter 26

Re gional D ou bl e Tax Treat y Mode l s Craig West

26.1 Introduction Regional tax treaty models are often dismissed as being simple hybridizations of the ‘main’ model tax conventions, being the OECD Model Convention on Income and on Capital (OECD MC) or the United Nations Model Double Taxation Convention between Developed and Developing Countries (UN MC).1 But insufficient attention to these models may be a missed opportunity to tease out regional tax treaty position or preferences. Regional tax treaty models emerge from (usually) economic groupings of countries and therefore should be a document that reflects that regional body’s preferences. This chapter seeks to evaluate the regional tax treaty models at the point of origination against the OECD MC and UN MC at that point in time. This first stage evaluates the extent to which there are any significant deviations from the major world tax treaty models in those regional tax treaty models. The second stage of the investigation seeks to evaluate any significant deviations or positions identified and test whether these are unique to the regional grouping or more generically used in other treaty models. These stages aim to answer whether the regional tax treaty models provide any insight into tax treaty policy at a regional level.

1 For

a discussion on the ATAF Model see C. West, ‘Emerging Treaty Policies in Africa—Evidence from the African Tax Administration Forum Models’, Bulletin for International Tax 75/1 (2021).

456 Craig West

26.2 Regional Tax Treaty Models Identified The author has attempted to identify all the regional tax treaty models that have been developed after the 1963 OECD MC, but the list may remain incomplete. This may be due to a regional model being only available in a language other than English or is not a document available in the public domain (i.e. may only be available to members of that regional body as a closed group). Table 26.1 documents the regional models identified: Table 26.1 List of regional tax treaty models Regional tax treaty model

Year concluded

Global region

Andean Communitya Income and Capital Model Tax Treaty 1971 (Standard Agreement to Avoid Double Taxation between Member Countries and States outside the Subregion)

South America

Intra-ASEANb Model Double Taxation Convention on Income

1987

South-East Asia

SADCc Model Tax Agreement on Income

2011 (revised 2013)

Southern Africa

ILADTd Multilateral Model Convention for Latin

2012

Latin America

EACe Model Taxation Convention

2012

East Africa

COMESAf Model Convention

2013

Eastern and Southern Africa

ATAFg Income Model Convention

2016 (revised in 2019) Africa

America for the Avoidance of Double Taxation

aThe Andean Community members are: Bolivia, Colombia, Ecuador, and Peru b

Association of South-East Asian Nations (ASEAN) members: Brunei, Cambodia, Indonesia, Laos, Malaysia, Myanmar (Burma), Philippines, Singapore, Thailand, and Vietnam. c

The members of the Southern African Development Community are: Angola, Botswana, Comoros, Democratic Republic of Congo, Eswatini, Lesotho, Madagascar, Malawi, Mauritius, Mozambique, Namibia, Seychelles, South Africa, United Republic of Tanzania, Zambia, and Zimbabwe. dLatin American Institute of Tax Law. This model was developed by members of the Institute,

being: Addy Mazz, Antonio Hugo Figueroa, Heleno Taveira Torres, Jacques Malherbe, Natalia Quiñones Cruz, and Pasquale Pistone. eMembers of the East African Community are: Republic of Burundi, Republic of Kenya, Republic of

Rwanda, Republic of South Sudan, United Republic of Tanzania, and Republic of Uganda. fThe Common Market for Eastern and Southern Africa (COMESA) members are: Burundi, Comoros,

Djibouti, Democratic Republic of the Congo, Egypt, Eritrea, Ethiopia, Kenya, Libya, Madagascar, Malawi, Mauritius, Rwanda, Seychelles, Somalia, Sudan, Eswatini, Tunisia, Uganda, Zambia, and Zimbabwe. g African Tax Administration Forum (ATAF): Angola, Benin, Botswana, Burkina Faso, Burundi, Cameroon, Chad,

Comoros, Ivory Coast, Egypt, Eritrea, Gabon, Gambia, Ghana, Kenya, Lesotho, Liberia, Madagascar, Malawi, Mauritania, Mauritius, Morocco, Mozambique, Namibia, Niger, Nigeria, Rwanda, Senegal, Seychelles, Sierra Leone, South Africa, Sudan, Eswatini (formerly Swaziland), Tanzania, Togo, Uganda, Zambia, and Zimbabwe.

Regional Double Tax Treaty Models 457 As can be seen from the list in the table, the development of such regional models has arisen in developing countries. This, at first glance, may imply that these regional models represent the dissatisfaction by these regional bodies with the two main models (OECD MC and UN MC). But such conjecture should be tested against the available evidence. There is one outlier in the regional models identified and that is the ILADT Model which was developed by a number of experts and therefore not by a regional economic body or, in the case of ATAF, a body representing tax administrations. As a result, the ILADT Model varies significantly from the structure of the OECD and UN Models in contrast to other regional models. For this reason, analysis of this model must stand apart from the other regional models of the world. Excluded from the analysis are regional–multilateral tax treaties themselves, such as the Nordic Convention (1996)2 and the CARICOM Agreement (1994).3 With regard to such regional treaties (i.e. not models), it should be noted that only the Andean Community and the East African Community have both a double tax treaty model as well as a multilateral tax treaty.

26.3 Analysis of the Regional Tax Treaty Models 26.3.1 Introduction Baring the ATAF Model, which borrowed text from the MLI discussions for its 2016 Model, the regional models in Table 26.1 would have had access to both the OECD MC and the UN MC as at the time of their conclusion. The 2017 versions of the OECD and UN Model Conventions have therefore not influenced any, apart from ATAF, regional model treaties, at least not yet. The analysis that follows is divided along the lines of the structure of the OECD MC. This has been done to analyse each article across all regional tax models and evaluate any unique aspects. Finally, the section will conclude with any wholly unique provisions inserted in the regional models that do not appear in the OECD or UN Model Conventions.

2 Convention between the Nordic Countries for the Avoidance of Double Taxation with respect to Taxes on Income and on Capital (1996). 3 Agreement among the Governments of the Member States of the Caribbean Community for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, Profits or Gains and Capital Gains and for the Encouragement of Regional Trade and Investment (1994).

458 Craig West

26.3.2 Regional Model Articles Replicating or Deviating from the OECD and UN Model Conventions 26.3.2.1 Scope of the Convention 26.3.2.1.1 Article 1: persons covered Termed personal scope in the OECD and UN Model Conventions (until the 1997 OECD MC changes), the text appears identical in almost all the regional tax treaty models. The models also uniformly change to ‘Persons covered’ following the OECD MC shift. The only observable variations are in relation to the ATAF Model which includes the recommended text from the BEPS Action 2 Report4 in its 2016 version. The 2019 ATAF Model adopts the same wording as appears in the 2017 OECD MC.

26.3.2.1.2 Article 2: taxes covered Minor amendments to the text of article 2 were made in the OECD Model in 1977 and in 2000. The most substantive was the change to the duty of notification from annually to only in the event of significant change to the taxes, which was the amendment in the OECD MC (2000).5 The UN Model followed these changes. On a regional basis, some variations are noted. The Andean Community Model (1971) varies from the OECD MC (1963) in that the essence of paragraph 4 of the OECD MC (1963) is contained in a different textual formulation. The ASEAN Model (1987) differs only in that it excludes ‘capital’ from the wording, focusing on ‘income’ only. This is also true of the SADC Models (2011 and 2013), the EAC Model (2012), COMESA Model (2013), and the ATAF Models (2016 and 2019). The ASEAN Model (1987) refers only to the contracting state, whereas the SADC Models (2011 and 2013) only omit local authorities. For paragraph 2 of article 2, those models referring only to income in paragraph 1 omit reference to capital appreciation in paragraph 2 for consistency. The SADC Models (2011 and 2013), EAC Model (2012), and COMESA Model (2013) also omit reference to wages and salaries in this paragraph. The notification of changes to domestic laws requirement in paragraph 4 is largely consistent with the OECD and UN Model Convention of the time. Only the EAC Model (2012) differs in that it adds: ‘and if it seems desirable to amend any Article of the Agreement, without affecting the general principles thereof, the necessary amendments may be made by mutual consent by means of an Exchange of Notes’.

4 OECD,

Neutralising the Effects of Hybrid Mismatch Arrangements, Action 2—2015 Final Report, OECD/G20 Base Erosion and Profit Shifting Project (Paris: OECD Publishing, 2015), http://dx.doi.org/ 10.1787/9789264241138-en, 139. 5 OECD , Model Tax Convention on Income and on Capital 2017 (Full Version) (Paris: OECD Publishing, 2019), http://dx.doi.org/10.1787/g2g972ee-en, M-8.

Regional Double Tax Treaty Models 459

26.3.2.2 Definitions 26.3.2.2.1 Article 3(1): general definitions The general treaty definitions are largely consistent with OECD and UN Model Conventions of the time. Minor variations may be noted, but none that cause a significant change in meaning. For the African regional tax treaty models (apart from ATAF), ‘international traffic’ includes ‘road and rail’, presumably in response to the many of the member states being landlocked countries sharing borders with other member states. Similarly, these regional models include a definition for ‘business’ to encapsulate professional services for the purpose of expansion of the definition of permanent establishment (see later).

26.3.2.2.2 Article 3(2): treatment of undefined terms The undefined terms interpretative provision is generally consistent with the OECD and UN Model Conventions of the time. The Andean Community Model (1971) differs in that it defaults all undefined terms to the domestic law and does not cater for a meaning of the term in context. The EAC Model (2012) reverts back to the OECD MC 1963 wording and the ATAF Model 2019 updates the text to the OECD MC 2017 wording deviating from the UN MC 2017 wording.

26.3.2.2.3 Article 4: resident The residence article in the regional models have, for the most part, followed the OECD Models of the time. However, variations do emerge, some of which appear to relate to the region at the time. The Andean Community Model (1971) reduces the residence article to a single paragraph in which the taxation rights are linked to source and not residence. This substantive variation is relevant to the regions use of source taxation for many years. Other regional models, such as the ASEAN Model (1987), despite the 1977 OECD MC change, drop the additional descriptors of residence (e.g. domicile etc.) and merely refer to a person considered as resident for tax purposes in the relevant state. The SADC Models (2011 and 2013) both provide two options catering for reference to ‘ordinarily resident’, a term used in the tax legislation of South Africa. Finally, the EAC Model (2012) also carries a reference to partnerships. For the individual tiebreaker for residence, some variations may be observed. Not all of the hierarchical tests listed by the OECD MC are included in the regional models. The ASEAN Model (1987) moves from habitual abode to mutual agreement procedures, thus eliminating the nationality tiebreaker. The others mirror the OECD MC provision. For persons other than individuals, despite the OECD and UN Models remaining unchanged until 2017, some African regional models incorporated mutual agreement procedures where doubt as to the place of effective management arose.6 Furthermore,

6

These include the SADC Models (2011 and 2013), the COMESA Model (2013), the EAC (2012), and the ATAF (2016).

460 Craig West both the EAC (2012) and the ATAF (2016) made provision for the denial of treaty benefits until dual residence was resolved, albeit that the ATAF (2016) version had borrowed from the OECD discussions culminating in the 2017 wording. In this respect, the EAC (2012) version stands apart as unique.

26.3.2.2.4 Article 5: permanent establishment Apart from the Andean Community Model (1971) which stands apart in its framing of a form of permanent establishment blended with a form of the business profits article, the regional models uniformly follow the basic permanent establishment rule. Naturally, regional models of developing countries tend to seek to expand the scope of the permanent establishment rule to secure full taxing rights over business profits arising in their jurisdictions. The regional models also tend to expand (slightly) the forms of permanent establishment as formulated in the illustrative list in paragraph 2. The UN Model over time appears to have had a greater influence in paragraph 3 of article 5. While the ASEAN Model (1987) follows the earlier UN Model, the African regional models adopted not only the UN Model provisions but added further expansions, with the SADC (2011) and EAC (2012) adding services of individuals and the SADC (2013), COMESA (2013), and ATAF (2016 and 2019) all expanding this paragraph further with installations or structures used for exploration for natural resources. A further modification appears in these African regional models in that the article commences with the permanent establishment being ‘deemed to include’ the listed provisions stepping further than the UN Model’s suggestion that the permanent establishment provision would also encompass such a list. This stronger wording reflects the African continent’s concerns and desire to objectively identify a permanent establishment in such contexts. Juxtaposed to this position of expansion of the permanent establishment scope, the regional models (barring the ASEAN 1987) tend to follow the OECD exclusion list (which is slightly broader in scope than the UN Model). Permanent establishments (PEs) created by agents in early regional models varied from the OECD Model, being a strong variation such as exclusion or a simplified approach (Andean Community 1971) or followed the UN approach (ASEAN 1987). The later African regional models tended to follow the OECD Model of the time. A similar pattern emerges for both independent agents and controlled companies. The influence of the UN MC on African regional tax treaties can be seen in the inclusion in all these models of the insurance PE provision of the UN Model.

26.3.2.3 Taxation of income 26.3.2.3.1 Article 6: income from immovable property It would seem that regional bodies considered this a largely uncontroversial article of the treaty. In almost all cases, the regional models adopted the OECD MC wording of the time. One notable deviation is the exclusion from the definition of ‘immovable property’ of rail and road transport vehicles in both the SADC and EAC Models. This is likely to clarify the position due to the inclusion of such enterprises within article 8 of those models. There is, however, some overlap between this provision and the scope of the PE provisions of the regional models (thus creating an overlap with article 7).

Regional Double Tax Treaty Models 461

26.3.2.3.2 Article 7: business profits The business profits article represents a drawing by the regional models on both the OECD and UN MCs as well as a universal rejection of the OECD’s authorized approach (as reflected in the 2010 OECD MC). As indicated in Section 26.3.2.2.4, the Andean Community Model blends the traditional articles 5 and 7 into a single article providing for taxation in the source state. The ASEAN Model draws again mainly on the UN MC, baring the inclusion of the ‘mere purchase’ provision to which the UN MC refers only as an option. The African regional models adopt a cherry-picking approach to the business profits article, utilizing the OECD MC (pre-2010) for paragraph 1, before reverting to the UN MC, except for the EAC Model, which adopts (mainly) the UN MC wording. The ATAF Model of 2016 includes the mere purchase paragraph, but this was dropped in the 2019 version of that model.

26.3.2.3.3 Article 8: international shipping and air transport While there is some conformity to the OECD and UN MCs, the regional models do deviate from both these mainstream models. Furthermore, none of the regional models make use of the UN MC’s paragraph 2 (the ‘more than casual’ nexus rule). The Andean Community Model offers two polar opposite alternatives in this article, either exclusive taxation in the state of residence or exclusive taxation in the state of source. Clearly, this is a contentious issue for the member states to this model. The ASEAN Model deviates from both models (at the time) in that it provides the taxing right to the resident state of the enterprise. The ASEAN Model further differentiates between air transport and shipping as the tax on shipping is reduced by 50%. The only provision borrowed from the OECD or UN MC is with reference to pooled schemes. Finally, the ASEAN Model defines international transport in the article itself rather than within the general definitions in article 3. The African regional models offer a mixed bag of inclusions. Apart from the EAC Model, all refer to the residence of the enterprise rather than the place of effective management. The SADC, EAC, and COMESA Models all also refer to rail and road transportation, broadening the scope of the provision, whereas the ATAF Model follows the pattern from the 2017 OECD and UN MCs in that it refers only to shipping and air transportation. All the African regional models then define ‘international traffic’ within article 8 to include rental of boats on a bare boat basis as well as the lease of containers. The African regional models referring to rail and road include rental activity in this clarification of scope. Finally, only the EAC Model and the 2016 ATAF Model include a separate paragraph addressing boats used on inland waterways.

26.3.2.3.4 Article 9: associated enterprises The Andean Community Model has no such article. The ASEAN Model does not include the corresponding adjustment paragraph (art. 9(2)). For the African regional models, while the COMESA Model simply replicates the OECD and UN MC wording, the SADC Models both add in the corresponding

462 Craig West adjustment paragraph that such adjustment shall only be made if the state to make the adjustment ‘agrees that the adjustment made by the first mentioned State is justified both in principle and as regards the amount’. The EAC Model also includes the third paragraph from the UN Model and adds two additional paragraphs providing, in the first, for a time limit for the application of such adjustment and, in the second, removing such limit in instances such as fraud. The ATAF Models both replicate the UN MC wording.

26.3.2.3.5 Article 10: dividends The Andean Community Model simply provides that dividends and equity investments will only be taxed where the enterprise has its legal residence. The ASEAN Model follows the structure of the 1977 OECD MC, but does not provide for any qualifying interest, subjecting all dividends to a maximum withholding of 15%. The African regional models contain a number of variations. The 2011 SADC Model initially adopts the 2011 UN Model working, only simplifying the definition of dividend in article 3 in the removal of terms not in use in the SADC states, whereas the COMESA Model adopts the UN MC text. The EAC Model first stipulates the withholding rates in paragraph 2 (thus deviating from the UN MC), but sets the reduced rates for qualifying interests at holdings greater than 50%. This model also sees the inclusion of a most- favoured-nation clause as regards these withholding rates. It further continues the use of the reduced dividend definition. The 2013 SADC Model continues to include a limited anti-abuse provision.7 The ATAF Models utilize the 2017 OECD MC wording, but for a reduced definition for dividends in paragraph 3 and allowing the withholding rates in paragraph 2 to be determined by negotiation (i.e. are not stipulated).

26.3.2.3.6 Article 11: interest The Andean Community Model again demonstrates its adherence to a source basis in that it simply provides that interest may be taxed where the loan finance is used (which is the state of payment unless otherwise proven). The ASEAN Model adopts much of the UN MC wording, but sets the withholding rate cap at 15%. It adds that amounts paid to the government of one of the contracting states shall be exempt. The 2011 SADC Model adopts the OECD MC wording but omits reference to the competent authorities in paragraph 2, an issue rectified in the 2013 version. The 2013 SADC Model also adds and anti-abuse provision (which is replicated in the COMESA Model). The EAC Model provides for differentiated interest withholding rates and adds a most- favoured-nation provision and an exemption for interest payable to a contracting state.

7 ‘No relief shall be available under this Article if it was the main purpose or one of the main purposes of any person concerned with the creation or assignment of the shares or other rights in respect of which the dividend is paid to take advantage of this Article by means of that creation or assignment’, art.10(6) 2013 SADC Model.

Regional Double Tax Treaty Models 463 The ATAF Models replicate the OECD wording, but leave the withholding rate to negotiation.

26.3.2.3.7 Article 12: royalties For the Andean Community Model, source rules result in taxation of the royalties in the state in which such property is used. The ASEAN Model adopts the UN MC wording, but for a small adjustment to royalties in connection with a PE where the UN MC refers to business activities in paragraph (1)(c) of article 7. Th African regional tax treaty models universally follow the UN MC, but adjust for the absence of an article for independent personal services. The SADC 2011 Model, like the interest article, omits reference to the competent authorities in paragraph 2. Similarly, the 2013 SADC Model rectifies this position and adds an anti-abuse provision (replicated again in the 2013 COMESA Model). The EAC Model adjusts the provision as it did for the article on interest. The ATAF Models adopt the UN wording but for the exclusion of reference to fixed base and the reference to business activities in paragraph (1)(c) of article 7. This accords with the findings in practice as to the prevalence of the UN MC in tax treaties.8

26.3.2.3.8 Article 12A: technical services While an article on technical services is absent from the OECD MC and was only inserted in the 2017 UN MC, practice amongst various developing countries led to the inclusion of such an article in some of the regional tax treaty models in Africa. No such article appears in either the Andean Community Model9 or the ASEAN Model. A named technical services article appears in the SADC Models and the ATAF Models. In the case of the latter, the wording as ultimately appearing in the 2017 UN MC was utilized (but for the absence of cross-reference to independent personal service). The SADC Models simply replicated the structure of the earlier passive income articles. However, in this case, no anti-abuse provision was inserted. This appears to an inconsistency with the earlier provisions. The EAC Model includes an article titled ‘Management and professional fees’ which operates and is structured in the same manner as a technical services article. The structure of this provision operates in the same way as the SADC Model, but also includes a most-favoured- nation provision.

8 W. F. G. Wijnen and J. J. P. de Goede, ‘The UN Model in Practice 1997–2013’, Bulletin for International Taxation 68/3 (2014), journal articles and opinion pieces IBFD. 9 At least, the Andean Community Model does not have a separate technical services article, but merges a reference to ‘technical assistance companies’ with an article on professional services.

464 Craig West

26.3.2.3.9 Article 13: capital gains The Andean Community Model again generates a unique article based on a source concept. The ASEAN Model, as with most of its articles, utilizes the UN MC wording, but omits the paragraph concerning alienation of shares. The African regional tax treaty models, other than the ATAF Models, all borrow the wording directly from the 2010 OECD MC. The EAC Model, in a strange switch around, refers in paragraph 2, which usually defaults gains attaching to a PE to be taxed in the State in which the PE is situated, rather than to the enterprise’s place of effective management. Further, in paragraph 3, which matched the article 8 taxing rights to tax the capital gains in the enterprise’s place of effective management, refers rather to the state of residence. The latter is understandable based on the EAC Model article 8, but the former appears to be an error. But for reference to independent personal services, the ATAF Models utilize the 2017 UN MC wording.

26.3.2.3.10 Article 14: independent personal services The Andean Community Model, with its source focus, includes no reference to fixed base or any other threshold before the source state can tax the professional services rendered in that state. The provision simply awards the source state the taxing right. The ASEAN Model largely adopts the UN MC wording, but does not require a fixed base for the independent service provider. Rather, the provision refers simply to a time threshold or whether the amount paid is borne by a PE or fixed base in that state and exceeds a set monetary threshold. The definition of professional services is also clarified for engineers to include software and hardware engineers. The African regional models were drafted after the OECD deleted this article from its model. As may already be evident from the earlier provisions, the African regional models, baring the EAC Model, do not include this article. The EAC Model recognizes the need for a fixed base, but utilizes the time threshold as an objective (alternative) test for a fixed base. Economists are added to the list of professions offering professional services.

26.3.2.3.11 Article 15: income from employment True to its emphasis on the source state, the Andean Community Model provides for exclusive taxation of remuneration from personal services (employment) in the state in which the service was rendered. Two exceptions are services rendered by a person employed by one contracting state delivering services in the other contracting state and the other relates to the employer of crew for ships, aircraft, buses, and other vehicles engaged in international transport. At the time of the ASEAN Model, the OECD MC and the UN MC wording were the same. The ASEAN Model adopted the same wording with the only deviation being to also render the provision subject to the contents of article 20 (students). The African regional models all mirrored the OECD and UN MC wording, but for consistency with their positions on article 8 with reference to international transport (i.e. all refer to the resident state except the EAC Model which refers to the enterprise’s

Regional Double Tax Treaty Models 465 place of effective management). In addition, the EAC Model, in applying the time threshold in paragraph 2(a) refers to 183 days in a calendar year which offers opportunity for manipulation. The EAC Model also retains reference to a fixed base in line with the UN MC.

26.3.2.3.12 Article 16: directors This article finds no variation in regional models, all of which follow the OECD MC except the ATAF Models which follow the UN MC. The article is entirely absent from the Andean Community Model.

26.3.2.3.13 Article 17: entertainers and sport persons The Andean Community Model (being based on source) simply reflects that an entertainer or sportsperson will be taxed in the country of source irrespective of the time spent. The remaining regional models appear to have adopted the UN MC. The adoption of the UN Model is apparent from those regional models that retained the independent personal services article as well as the later ATAF Models retaining the reference to business profits in paragraph 1, despite the OECD dropping such reference in the 2014 OECD MC. In addition, apart from the Andean Community, all the regional models include some form of exemption for sporting activities funded by the contracting states. Finally, the EAC Model adopts the view that the indirect taxation of entertainment or sporting activities in paragraph 2 operates as a anti-avoidance provision and therefore clarifies that it does not apply where the entertainer or sportsperson does not share in the profits directly or indirectly.

26.3.2.3.14 Article 18: pensions Apart from the Andean Community Model’s unique source rule for pensions, the remaining regional models have all adopted wording from the UN MC. The ASEAN Model adopted the 1980 UN MC Alternative B for this article. The later African regional models adopted Alternative B, but included a reference to ‘annuities’ in paragraph 1, deleted the UN MC paragraph 2 from that option, and included a definition of ‘annuities’ in paragraph 2.

26.3.2.3.15 Article 19: government service The Andean Community Model does not address this form of income. The ASEAN Model follows the OECD and UN MCs of the time, but omits (b) within paragraph 2 of the article. The African regional models tend to follow the OECD and UN MCs but the SADC Models do not refer to ‘similar remuneration’ in (b) of paragraph 2 of the article and the EAC and ATAF Models only refer to pensions in paragraph 2. None of these variations bring significant change to the article.

26.3.2.3.16 Article 20: students The Andean Community Model contains no such article. The ASEAN Model follows the UN MC but expands the income categories in paragraph 2. The African regional models, despite moving some of the language around, maintain the same provision as appears in the OECD MC.

466 Craig West

26.3.2.3.17 Article 21: other income The Andean Community Model does not contain such a provision. The ASEAN Model provides for shared taxing rights over other income in a single paragraph. In contrast, the African regional models universally utilize the wording of the UN MC.

26.3.2.4 Article 22: taxation of capital Apart from the 1971 Andean Community Model, all the remaining regional models universally exclude an article on capital.

26.3.2.5 Article 23: methods for elimination of double taxation The 1971 Andean Community Model appears to have no article addressing the method for elimination of double taxation. This may be a result of the structure of the model itself, which aims to grant exclusive taxing rights to the source state. This may generally negate the need for the article (or the matter is devolved to domestic legislation). While the remaining regional tax models all adopt the credit method for the elimination of double taxation, some deviations are noted. The ASEAN Model includes a tax sparing clause. The SADC Models of 2011 and 2013 also provide for a form of a tax sparing credit, where mutually agreed by the competent authorities as qualifying schemes for economic development.

26.3.2.6 Special provisions 26.3.2.6.1 Article 24: non-discrimination Despite inclusion since the 1963 OECD MC, none of the non-discrimination articles in any regional model refer to stateless persons. The Andean Community Model does not contain a non-discrimination article at all. The 1987 ASEAN Model appears to restrict the non-discrimination clause to both the persons covered in article 1 and the taxes covered in article 2. The African regional models generally adhere to the OECD MC wording. An observable deviation is the exclusion of a reference to the deductibility of disbursements in respect of the debt of an enterprise owing to a resident of the other state. The EAC Model (2012), after applying article 24 to taxes of all types and descriptions adds an additional paragraph stating: ‘In this Article the term ‘taxation ‘means taxes which are the subject of this Agreement.’ Whether this clarifies or confuses the earlier provision cannot be said.

26.3.2.6.2 Article 25: mutual agreement procedure The Andean Community 1971 Model does not provide for a separate mutual agreement procedure article. Rather, the ability of the competent authorities to engage in discussions is merged with its highly shortened version of the exchange of information provision. The ASEAN Model 1987 largely follows the OECD 1977 and UN 1980 Models. However, it allows a national of a contracting state to request a mutual agreement procedure despite article 24. Furthermore, the sentences in the final paragraph of article 25 addressing some procedural aspects are omitted from the ASEAN Model. This model also excludes the paragraph concerning disbursements to non-residents.

Regional Double Tax Treaty Models 467 For the African regional models, the SADC Model 2011 adopts the text from the 2010 OECD MC, whereas the EAC Model 2012 utilized alternative A from the 2011 UN MC. By 2013, the SADC Model moved to Alternative B from the 2011 UN Model, but retained its original amendment to the wording of paragraph 4 from the 2010 OECD Model. The ATAF Models of 2016 and 2019 follow the OECD MC of the time.

26.3.2.6.3 Article 26: exchange of information The exchange of information articles contained in the regional models largely follow the OECD MC. The ASEAN 1987 Model (much like most of that regional model) adopts the UN version at the time. The SADC 2013 Model introduces text in paragraph 2 of the article that is taken up in the 2014 OECD MC, namely the authorization by a contracting state for other purposes. The regional models otherwise adopt the latest version of the OECD MC. The EAC 2012 Model again varies the language used slightly and includes an additional paragraph concerning the depositions of witness and authenticated copies of unedited documents.

26.3.2.6.4 Article 27: assistance in the collection of taxes As a provision only introduced by the OECD in 2003, this article does not feature in any of the regional models before that date. Apart from different wording used in the EAC 2012 Model, the remaining post-2003 regional tax models (noting ILADT’s exclusion from this sample) all mirror the OECD article wording.

26.3.2.6.5 Article 28: members of diplomatic missions and consular posts This particular article is adopted as standard text across the regional models with little variation. Only the earliest regional models adopted slightly different language that had no bearing on the substance of the article.

26.3.2.6.6 Article 29: entitlement to benefits Regional models, like the OECD and UN MCs generally, had no specific article referencing any entitlement to benefits. The EAC 2012 Model was the first to provide for an entitlement to benefits provision. The only other regional model specifying an entitlement to benefits provision are the ATAF Models. The 2016 version borrowed the principal purpose test wording from the base erosion and profit splitting (BEPS) discussions at the OECD (text that ultimately found its way into the 2017 OECD Model as article 29(9). The 2019 version of the ATAF Model expanded the entitlement to benefits article by including the wording from article 29(8) of the 2017 OECD MC.

26.3.2.6.7 Article 30: territorial extension Based on the general definitions in the regional tax treaties defining the territories of the contracting states, this article does not feature in any of the model treaties within the scope of this chapter. The one exception (outside the scope of this analysis) is the ILADT Model.

468 Craig West

26.3.2.7 Final provisions 26.3.2.7.1 Article 31: entry into force The entry into force provisions are substantively the same as the OECD MC provisions. However, the African regional models cater for start dates for taxes at source and for other taxes. The most recent deviation is the inclusion in the ATAF 2019 Model for the provisions of articles 26 and 27 to apply to any taxable year following the entry into force of the treaty (thus providing for retrospective assistance). Whether this recent deviation will appear in treaty practice remains to be seen, especially considering the proliferation of exchange of information and mutual administrative assistance treaties concluded on a bilateral and multilateral basis.

26.3.2.7.2 Article 32: termination The termination articles are substantively similar to the OECD MC provisions of the time. The main deviations arise with the specification in the African regional models that termination may only take place five years after the agreement came into effect, thus setting a minimum period of operation for the treaty. Furthermore, the African regional models specify transition periods for termination as regards taxes at source and other taxes.

26.3.2.8 Unique regional provisions 26.3.2.8.1 Professors and teachers Only the ASEAN, SADC, and EAC Models include such an article, aiming to attract researchers and educators. Such a provision can be found in a number of developing country treaties. However, the provision appears to not find favour in the regional models. Pickering suggests that such provisions can be problematic to administer and so such provisions should instead be considered under domestic law.10 It seems from the regional treaty models that consensus on this issue could generally not be obtained and so was left to the individual countries to determine whether or not such a provision would be included as a matter of specific negotiation rather than based on a model. The structure and content of the articles are very similar.

26.4 Emerging Positions on a Regional Basis The first observation is that regional tax treaty models are not as pervasive as one may anticipate. Apart from the African region, regional models developed by bodies of

10 A. Pickering, ‘Tax Treaty Policy Framework and Country Model’, Paper No. 2-N (2013), available at https://www.un.org/esa/ffd/wp-content/uploads/2013/05/20130530_Paper2N_Pickering.pdf (last accessed 1 May 2021).

Regional Double Tax Treaty Models 469 governmental agencies are dated and may need to be reviewed. The ILADT Model, as a model not developed by a governmental agency grouping represents, to some extent, a reconsideration of tax treaties as regards the Latin American region (and certainly can be viewed as a significant update on the Andean Community Model of 1971). The ASEAN Model of 1987 is in need of an update and re-evaluation against the tax policy priorities for the region. In Africa, much similarity can be observed between the SADC and COMESA Models, as the COMESA Model borrows much from the earlier SADC Model.11 The second observation is that while these regional models show some variations when compared to the OECD and UN MCs, they do not represent, apart from the Andean Community Model, any significant or unique aspects representative of the regions. This second observation is, perhaps unsurprising. A regional tax model should reflect the regional tax policy, but this may be in conflict with the national tax policies of the individual member states. This conflict will bring difficulty for the harmonization of tax policy. Further, the member states themselves may be in different levels of development (or strength) which may deepen any conflict between national tax policies. Within Africa, this is further exacerbated by the overlap of membership of the states in a number of regional bodies (which themselves may have competing policies); however, efforts are being made to coordinate between these regional groupings as well. Naturally, factors other than tax may also play a significant role (e.g. political will12), but such considerations are beyond the scope of this chapter. Additionally, it can be observed that variation from the traditional OECD and UN MCs can be a double-edged sword. The EAC Model, for example, utilizes most-favoured- nation clauses within the passive income articles. This is surprising for a model for the member states of the region to use in negotiations with third states, particularly capital- exporting states, as this has the potential to erode the tax base for these income streams

11 M.

Hearson, ‘Tax treaties in Sub-Saharan Africa: A critical review’, Tax Justice Network—Africa’ (2015), available at https://www.taxjustice.net/wp-content/uploads/2020/11/tjna_treaties.pdf (accessed 1 May 2021) and L. Charalambous, ‘COMESA Pushes For DTAs Between Member States’, Tax-News.com (2012), available at https://www.tax-news.com/news/COMESA_Pushes_For_DTAs_Between_Mem ber_States____53993.html (last accessed 1 May 2021). 12 With Africa, there seems to be a strengthening of political will to harmonize and cooperate. See the Outcome Statement of the 4th High-Level Policy Dialogue (27 Aug. 2020) in which it is stated: Of note was the need for the AU and the Regional Economic Communities (RECs) to coordinate tax policy harmonisation that should be aligned to national tax policies. Participants also called for the establishment of a tax department within the AUC to coordinate tax policy in Africa. The meeting emphasised the necessity for a continental solution rather than implementing unilateral solutions. As such, the legislative role of the Pan African Parliament (PAP) was deemed to be crucial for the advancement of continental cooperation in tax policy and administration. Also, the meeting called for enhanced partnerships in Africa, focusing on standards, statistics, and the development of multilateral instruments. See https://au.int/sites/default/files/pressreleases/39172-pr-out come_statement_-_4th_high-level_policy_dialogue.pdf (accessed 1 May 2021).

470 Craig West leading to significant loss.13 However, it can also be observed that the models do offer some protection (at least in the case of SADC and COMESA) in the inclusion of an anti- abuse provision. As such inclusion predates the principal purpose text (included in the 2017 versions of OECD and UN MC), it would seem that such provisions stem from concerns of the region. Caution should also be exercised when reviewing regional tax treaty models as, in some cases, ‘stronger’ member states may have had a hand in furthering their own national agenda/policy without full regard to the other member states’ contributions. Furthermore, it should also be understood which member states participated in the discussion and development of the model. For example, in the case of ATAF, only twenty- one of thirty-eight members states contributed to the development of the model (and recorded reservations against the model). It should be further noted that the SADC, COMESA, and ATAF Models permit member states to record reservations14 against the models. Such reservations are of particular value where many of the countries do not publish tax treaty policies or lodge non-member country positions against the OECD MC. Such reservations therefore provide insight into the particular tax treaty issues of relevance to a particular member state (and where disagreement on a coordinated approach may be found). Reservations can also serve to indicate whether the state wishes to retain a consistent position or differentiate its position relative to different economic groupings of which it is a part. This can be observed, for example, in the consistent South African position not to reserve against the inclusion of the arbitration provision in the mutual agreement procedure article in the ATAF and SADC Models, but to record such disagreement with the inclusion of a relevant clause against the OECD MC. This demonstrates the willingness of South Africa, as a stronger economic player in its region, to test arbitration before considering such a provision on a global scale.15 In some respects, participation by these countries in greater coordination instruments (e.g. the Administrative Cooperation agreements) and global coordination bodies (e.g. the Global Forum) aid coordination for the administrative articles of the model treaties.

26.5 Conclusions It is first evident from the previous analysis that the regional models borrow heavily from the OECD and UN MCs. Variation outside both these models is limited, and, where found, can lead to difficulties. Only one regional model stands apart from the

13

Hearson, ‘Tax treaties in Sub-Saharan Africa’ and West, ‘Emerging Treaty Policies in Africa’. For an analysis of the ATAF reservations, see West, ibid. 15 Ibid. 14

Regional Double Tax Treaty Models 471 others, being the Andean Community Model of 1971. However, its black-and-white allocation of taxing rights may not work for all regions and countries. The bilateralism of double tax treaties is one of the causes of the difficulties for regional models to standardize practices and achieve regional solutions. Regional tax treaty models, to serve as both inward-looking and outward-looking, would require greater domestic coordination of income tax systems to be successful. Should the region wish to negotiate with larger developed countries as a bloc on the basis of the regional model, it would only take the adjustment to domestic legislation by one member state to potentially unravel the network. Further, it would require a level of coordination of taxes not yet even observable in the EU (as the most developed coordination of a single economic market). In the West African Economic and Monetary Union (for which no regional tax treaty model could be found), it is observed by Mansour and Graziosi that even where there is regional integration on direct taxes, it led to the ‘fragmentation of policy making at the national level by providing countries with the incentive to enact special tax regimes outside their tax laws’ due to the exclusion of such incentives from the coordination.16 Thus, even if a working regional tax treaty model were to be found, it may be undermined by the failure to coordinate. However, regional models may well serve a role in relationships between members states. This is observed with respect to the Andean Community Model.17 In general, this was the purpose behind the SADC and COMESA Models; namely, to encourage treaties between the member states.18 It may therefore be questioned whether these same models (as are occasionally used by the members states for negotiation with third parties) are fit for purpose. It may be said that, at the very least, a regional tax model may serve as a base for negotiation, especially for those states that have not yet developed their own clear tax treaty wish list. Furthermore, the possibility of recording reservations against regional models for those states not represented at the OECD or having input to the UN MC enables such states to develop their tax treaty policy and reflect on it with regional trade partners. Such regional organizations may provide an opportunity for a state with no or under-developed tax treaty policy to learn from more established regional partners. Looking to the future, regional tax treaty models should further and boldly put forward the tax treaty policy of the region. Regions, where appropriate and in the climate of greater global coordination, should seek greater regional coordination. The regional model could set in motion greater coordination and facilitate regional negotiations with third states. In addition, this may have a cost-saving for both the third state and

16

M. Mansour and G.-R. Graziosi, ‘Tax Coordination, Tax Competition, and Revenue Mobilization in the West African Economic and Monetary Union’, IMF Working Paper WP/13/163, Fiscal Affairs Department (2013), 36. 17 R. J. Vann, ‘A Model Tax Treaty for the Asian- Pacific Region?’, Bulletin for International Fiscal Documentation 45 (8 Nov. 2010), 99–111; APTIRC Bulletin 8 (1991), 392–422; ‘Anti-Avoidance and Tax Treaty Policies and Practice in the Asian-Pacific Region, APTIRC, Sydney Law School Research Paper No. 10/122 (1990), 27, available at https://ssrn.com/abstract=1705765. 18 Hearson, ‘Tax treaties in Sub-Saharan Africa’.

472 Craig West the region in the negotiation stages. Further, such regional models may well lead to true multilateral tax treaties. Ultimately, while currently the regional models (apart from the Andean Community Model) do not bear significantly unique features to reflect regional priorities, the opportunity remains for regions to begin to work towards such a goal.

Chapter 27

Unil atera l i sm, Bil aterali sm, a nd Mu ltil atera l i sm i n Internationa l Tax L aw Miranda Stewart

The United States is now seeking a global agreement on a strong minimum tax through multilateral negotiations.1

27.1 Introduction International tax rules are mostly established by states in national tax law—that is, tax jurisdiction is primarily established unilaterally. However, bilateral tax treaties also play an important role in setting the scope of tax jurisdiction. Bilateral treaties are negotiated in the context of multilateral models developed during the twentieth century under the auspices of the OECD and the UN. As a result of these layers of international tax law, today, few rules addressing the taxation of international flows of people, goods, services, and capital, are solely unilateral, bilateral, or multilateral. Unilateral, bilateral, and multilateral international tax rules are a product of bargains between states, and between states and mobile capital or corporations (and sometimes mobile labour or people). Currently, the layers of unilateral, bilateral, and multilateral tax rules do not always combine in an integrated, coherent, or harmonized international

1 The White House, ‘The American Jobs Plan’ (31 Mar. 2021), available at https://www.whitehouse. gov/briefi ng-room/statements-releases/2021/03/31/fact-sheet-the-american-jobs-plan/ (accessed 31 March 2021).

474 Miranda Stewart tax system. Instead, they may generate diverse and fragmented outcomes that create complexity, double taxation, or opportunities for tax arbitrage. The form of international tax rules is also diverse. States engage with each other through legal agreements, consensus-based international standards or peer-reviewed norms, administrative cooperation, data-sharing systems, judicial analysis, and formal and informal policy transfer through bureaucratic, academic, and professional networks. Unilateral and bilateral international tax rules are administered through an increasingly multilateral web of executive agreements, institutions, and data-sharing processes, authorised by treaties or domestic law. International tax law has developed through these processes and networks, usually in an incremental way, although international tax law development is punctuated from time to time with developments that have the effect of shifting international tax to a new equilibrium. The quote that opens this chapter makes a promise of multilateralism to achieve a fairer, more effective international tax system. Whether multilateralism succeeds in achieving this promise depends ultimately on the unilateral action of states, based on their technological and cooperative capabilities, and whether this can shift the system to a new equilibrium that taxes mobile capital more than the previous regime.

27.1.1 Unilateralism, Bilateralism, and Multilateralism In public international law, the terms unilateralism, bilateralism, and multilateralism refer to the way in which states act, or decline to act, in relation to other states. It is assumed that states always act in their own self-interest and they may achieve this through unilateral or cooperative action. Usually, ‘unilateralism refers to an individualistic approach to foreign affairs . . . apart from other states’.2 Examples of unilateralism with respect to taxation include tax law-making which is based on national sovereignty, with the power to tax comprising a core element of statehood.3 It can also include enforcement of tax law; the exercise of judicial power in court cases; political acquiescence or protest by states in relation to actions of other states; or a state’s approach to international negotiations, such as the decision to put forward a reservation to a treaty, or a decision not to sign a treaty or to withdraw from a treaty. Bilateralism refers to cooperation between two states and is traditionally the dominant mode of international tax cooperation.4 Since the nineteenth century, tax treaties have been negotiated on the principle of reciprocity involving full give and take between

2 A. Nollkaemper, ‘Unilateralism/ Multilateralism’ (last updated Mar. 2011), in A. Peters and R. Wolfrum, eds, The Max Planck Encyclopedia of Public International Law (Oxford: Oxford University Press 2008), https://www.mpepil.com (accessed 7 August 2022). 3 W. Schon, ‘National Sovereignty and Taxation’, in T. Cottier and K. Schefer, eds, Edward Elgar Encyclopedia on International Economic Law (Cheltenham: EEonline, 2017), 507–510. 4 N. Bravo, A Multilateral Instrument for Updating the Tax Treaty Network (Amsterdam: IBFD, 2020).

Unilateralism, Bilateralism, and Multilateralism 475 two sovereign nations.5 The diversity of national tax systems and state economic and legal circumstances have been considered as the reason, and justification, for bilateralism in tax treaties.6 Most attention has focused on the relief of double taxation of income and capital, to remove barriers to investment and achieve fairness for taxpayers. Countries have different tax interests and the distributive conflict about residence and source tax jurisdiction depends on their economic relations, such as asymmetry in capital flows.7 Multilateralism ‘describes an approach to foreign relations that seeks cooperation with other states’.8 An institutional approach views multilateralism as a coordinated set of rules and patterns of conduct grounded in common norms. There has been a proliferation of multilateral institutions in international economic law, which may even compete with each other.9 Multilateralism does not require universal agreement but it usually refers to cooperation between a large number of states, with the potential to establish rules or institutions that have wide application. This may be contrasted with plurilateral, often regional, cooperation, in which a set of rules or standards is established between a limited number of states. Multilateralism in a particular field of law seeks to achieve a coordinated international regime that transcends the limitations of national sovereignty. Some forms of multilateralism in international tax operate as ‘hard’ and binding law, including tax treaties and administrative cooperation across borders authorized by national tax laws or by tax treaties, while other forms are ‘soft’ principles, standard or guidance. Another less well-studied way in which states engage with each other to establish international tax law is through the judicial arm of government; for example courts may cite the decisions of courts in other countries on tax treaties, leading to a growing specialist field of treaty law.10 Unilateral, bilateral, and multilateral modes of state action are often intermingled. A state may take a unilateral approach on one aspect of a topic, but a multilateral approach on other aspects of the same topic. Multilateral institutions or treaties may be ‘efficient ways of creating and regulating sets of bilateral relations’; or seemingly multilateral processes may be effectively controlled by one state or a small group of

5

S. Jogarajan, Double Taxation and the League of Nations (Cambridge: Cambridge University Press, 2018); S. Jogarajan, ‘Prelude to the International Tax Treaty Network: 1815–1914 Early Tax Treaties and the Conditions for Action’, Oxford Journal of Legal Studies 31 (2011), 679–707. 6 H. D. Rosenbloom and S. Langbein, ‘United States Tax Treaty Policy: An Overview’, Columbia Journal of Transnational Law 19 (1981), 359–406, 366. 7 T. Rixen and I. Rohlfing, ‘The Institutional Choice of Bilateralism and Multilateralism in International Trade and Taxation’, International Negotiation 12/3 (2007), 389–414. 8 Nollkaemper, ‘Unilateralism/Multilateralism’; M. Kolsky Lewis, ‘Multilateralism in International Economic Law’, in Cottier and Schefer, Elgar Encyclopedia of International Economic Law, 33–35. 9 G. Anton, ‘Substantive Multilateralism in the Context of the MLI’, in A. P. Dourado, ed., International and EU Tax Multilateralism: Challenges Raised by the MLI (Amsterdam: IBFD, 2020), 15–32. 10 E. Baistrocchi, ed., A Global Analysis of Tax Treaty Disputes (Cambridge: Cambridge University Press, 2017).

476 Miranda Stewart states.11 A multilateral framework may be combined with plurilateral negotiations.12 Fundamentally, unilateral state conduct or practice, if consistent with the practice of many other states, may reveal the existence of customary international law, which may bind all states. Unilateral tax law-making based on a common set of norms, combined with elements of the tax treaty network based on multilateral models, forms the basis of Reuven Avi-Yonah’s argument that ‘a coherent international tax regime exists, embodied in both the tax treaty network and in domestic laws, and that it forms a significant part of international law (both treaty-based and customary)’.13 An important consequence of Avi-Yonah’s thesis is that the international tax regime constrains unilateral action of states in making domestic tax laws. Avi-Yonah suggests that international tax law comprises two principles. First, the ‘single tax principle’ under which income should be taxed once, no more and no less, that is, in one jurisdiction, so it is not double-taxed, but nor is it untaxed. Secondly, the ‘benefits principle’ (active business income is taxed primarily at source and passive income is taxed primarily at residence).14 While sceptical of the single tax theory, David Rosenbloom has observed that the residence state’s worldwide jurisdiction to tax, and its responsibility to alleviate double taxation (which together should ensure a ‘single’ layer of tax), will apply because the residence state stands ‘in a position to see the entirety of the taxpayer’s income’.15 The increase in taxes levied by many countries early in the twentieth century combined with the enthusiasm for multilateralism of the League of Nations, led to its project aimed at establishing principles to divide up tax jurisdiction around the world. A series of reports produced in the 1920s led to the first model treaties which formed the basis of a small network of bilateral tax treaties, mostly in Europe.16 The League of Nations asked the Four Economists, who produced their report in 1923: Can any general principles be formulated as the basis for an international convention to remove the evil consequences of double taxation, or should conventions be made between particular countries, limited to their own immediate requirements? In the latter alternative, can such particular conventions be so framed as to be capable ultimately of being embodied in a general convention?17

11

Nollkaemper, ‘Unilateralism/Multilateralism’. Basedow, ‘The WTO and the Rise of Plurilateralism—What Lessons Can We Learn from the European Union’s Experience with Differentiated Integration?’, Journal of International Economic Law 21 (2018), 411–431. 13 R. Avi-Yonah, International Tax as International Law: An Analysis of the International Tax Regime (Cambridge: Cambridge University Press, 2007), 1. 14 Ibid. 15 D. Rosenbloom, ‘What’s Trade Got to Do With It’, Tax Law Review 49 (1994), 593, 596. For more on tax jurisdiction, see M. Stewart, Tax and Government in the 21st Century (Cambridge: Cambridge University Press, 2022), Part III. 16 S. Jogarajan, Double Taxation and the League of Nations (Cambridge: Cambridge University Press, 2018), 3. 17 .73.F.19 (Geneva: League of Nations, 1923), extracted in Jogarajan, Double Taxation and the League of Nations, 19. 12 R.

Unilateralism, Bilateralism, and Multilateralism 477 In his Lectures of 1928, Seligman criticized bilateralism but observed that a universal agreement on taxation may be impossible to achieve. He asked, ‘Why must the choice be limited to one between universal and bilateral agreements? Why can we not have multilateral agreements?’18 Seligman saw multilateralism as a way for countries with similar tax systems to cooperate. By the time the Council of the Organisation for European Economic Co-operation (OEEC) adopted its first Recommendation on double taxation, seventy bilateral tax treaties had been signed between countries which are now members of the OECD.19 Today, it is widely observed that there are about 3,000 bilateral tax treaties around the world. These have specific features negotiated between countries. However, they are based on common principles and a common structure developed multilaterally in ‘soft’ law norms through models and commentary. In some areas of international economic law, the late twentieth century saw a rise in multilateralism and increasing support for it among both states and scholars. The establishment of the WTO in 1995, now comprising more than 150 member states, was an important milestone. Regional cooperation increased in international trade, in the EU, the Southern African Customs Union, and the Asian Free Trade Area of the Association of South East Asian Nations (ASEAN), among others. These global and regional trade agreements aim to eliminate barriers to trade by incremental reduction of tariffs, often on a most-favoured-nation basis. Less visible but equally important is the underlying, essentially universal, agreement on the categorization and valuation of goods to support uniformity in international trade and customs.20 International tax law is partly affected by the multilateral or bilateral features of trade and investment regimes, which are imported into international tax at the point where these regimes intersect.21 The goal of trade law to reduce tariffs negatively affects the capability of some nations to raise tax revenues; while tax concessions to encourage investment may be prohibited subsidies under multilateral regimes.22 However, apart from these indirect effects, international tax law seemed immune to the trend towards multilateralism, at least explicitly in the form of multilateral treaties. There was debate about the causes and merits of multilateralism, and why international

18 E.

R. A. Seligman, Double Taxation and International Fiscal Cooperation (New York: Macmillan, 1928), 170. 19 E. Reimer and A. Rust, eds, Klaus Vogel on Double Taxation Conventions, 4th ed. (Amsterdam: Wolters Kluwer, 2015), 1. 20 D. Rovetta, ‘Harmonized System and Schedules of Concession’, in Cottier and Schefer, Edward Elgar Encyclopedia of International Economic Law, 350– 351; D. Rovetta and L. C. Beretta, ‘The International Law of Customs: Customs Valuation’, in Cottier and Schefer, Edward Elgar Encyclopedia of International Economic Law, 354–356. 21 UN Committee of Experts on International Cooperation in Tax Matters, ‘Secretariat Paper: The Interaction of Tax, Trade and Investment Agreements’, E/C.18/2019/CRP.14 (2019). 22 J. Cagé and L. Gadenne, ‘Tax Revenues and the Fiscal Cost of Trade Liberalization, 1792–2006’, Explorations in Economic History 70 (2018), 1–24; WTO, Agreement on Subsidies and Countervailing Measures, available at https://www.wto.org/english/docs_e/legal_e/24-scm_01_e.htm.

478 Miranda Stewart tax law appeared exceptional.23 One issue raised in the literature concerns transaction costs.24 Transaction costs of multilateralism appear at first glance to be lower: one treaty is applicable for many treaty partners, in contrast to the high transaction cost of negotiating many bilateral treaties. In international tax, it has been argued that the modification of bilateral tax treaties within a multilateral framework would reduce transaction costs and increase efficiency.25 A different view about transaction costs developed after multilateral trade negotiations faltered in the first decade of the twenty-first century.26 Multilateral institutions, such as the WTO dispute-resolution system, seemed to enter a period of ‘crisis’.27 In international investment law, multilateralism was not achieved and the dominant mode remains bilateral. The OECD attempt to establish the Multilateral Agreement on Investment was a costly failure.28 Reaching a multilateral agreement may therefore pose high transaction costs for relatively low or diffuse benefits to states and private actors. Furthermore, some suggest that ‘multilateral rule making appears to be unable to respond to the changing needs and problems of the modern international economy’.29 Yet the turn to bilateralism and plurilateralism in international trade law may deliver only modest economic or distributional gains, suggesting that persisting with multilateral negotiations may achieve better outcomes for global welfare. While bilateral tax treaties often take a long time to negotiate, they tend to be reasonably stable. It is surprising when a treaty is terminated, as in the case of the withdrawal by Sweden from its long-standing bilateral tax treaties with Greece and Portugal. The

23

K. Brooks, ‘The Potential of Multilateral Tax Treaties’ in M. Lang et al., eds, Tax Treaties: Building Bridges Between Law and Economics (Amsterdam: IBFD, 2010), 211, 212; K. Fuchi, ‘Unilateralism, Bilateralism, and Multilateralism in International Taxation’, Japanese Yearbook of International Law 59 (2016), 216–228; Dourado, International and EU Tax Multilateralism: Challenges Raised by the MLI. More broadly, see e.g. S.Woolcock and K. Heydon, Rise of Bilateralism: Comparing American, European and Asian Approaches to Preferential Trade Agreements (Tokyo: UN University Press, 2009); S. Schill, The Multilateralization of International Investment Law (Cambridge: Cambridge University Press, 2009); R. Leal-Arcas, International Trade and Investment Law: Multilateral, Regional and Bilateral Governance (Cheltenham: Edward Elgar, 2010); G. Blum, ‘Bilateralism, Multilateralism, and the Architecture of International Law’, Harvard International Law Journal 49 (2008), 323–379. 24 A. Thompson and D. Verdier, ‘Multilateralism, Bilateralism, and Regime Design’, International Studies Quarterly 58 (2014), 15–28. 25 J. F. Avery Jones and P. Baker, ‘The Multiple Amendment of Bilateral Double Taxation Conventions’, Bulletin for International Taxation 60 (2006), 1, 21. 26 A. Martin and B. Mercurio, ‘Doha Dead and Buried in Nairobi: Lessons for the WTO’, Journal of International Trade Law and Policy 16/1 (2017), 49–66. 27 T. Payosova, G. C. Hufbauer, and J. J. Schott, ‘The Dispute Settlement Crisis in the World Trade Organization: Causes and Cures’, Policy Briefs PB18-5, Peterson Institute for International Economics (2018). 28 P. T. Muchlinski, ‘The Rise and Fall of the Multilateral Agreement on Investment: Where Now?’, The International Lawyer 34 (1999), 1033–1053; UN Conference on Trade and Development, ‘Lessons from the MAI’, UNCTAD/ITE/IIT/MISC.22 (1999). 29 A. Reich, ‘Bilateralism versus Multilateralism in International Economic Law: Applying the Principle of Subsidiarity’, University of Toronto Law Journal 60 (2010), 263–287, 277.

Unilateralism, Bilateralism, and Multilateralism 479 termination of those treaties was a response to the phenomenon of expatriate pensioners who depart the cold north for the warmer south, attracted by low or ring-fenced taxes for inbound migrants, and taking pension taxing rights with them.30 The contentious issue—the residence basis for taxation—may be difficult to solve without challenging the fundamental norm of tax residence as the jurisdictional nexus, a growing issue in an era of long-lived, and mobile, populations. A further argument for multilateralism in international tax is to ensure justice.31 A justice approach could, in the suggestion of Ana Paula Dourado, involve ‘a non- discrimination principle between mobile and immobile taxpayers and investment without forgetting that redistribution is one of the main states’ budgetary functions, and in the current international tax system, redistribution is a function played by the state’.32 However, achieving fairness and legitimacy in multilateral tax rule-making requires attention to which countries set the agenda and control the decision-making process; which countries may have veto power; and, ultimately, in whose interest the multilateral bargain operates. An important criticism concerns the exclusion of the Global South, or developing countries, in relation to OECD-led international tax negotiations of all kinds, including the G20-OECD Base Erosion and Profit Shifting (BEPS) Project.33 In these negotiations, power may be located in the largest developed economies, such as the USA, or in trade blocs such as the EU. The international institutions themselves may jockey for pre-eminence in multilateral tax lawmaking processes.34 Even those who have argued against the ‘cartelization’ of international tax, as likely being unjust to low-income countries and detrimental for global welfare, accept that some multilateral regulation of the international tax framework would make us better off. Tsilly Dagan suggests that tax treaties that are bilaterally negotiated on the basis of a fundamentally embedded OECD model and the pre-existing network could be improved through establishment of an improved standard in the international tax regime.35 The current standard, ‘with its particular focus on allocating tax revenues, has not proven to be very effective. In fact, it may have even been counterproductive . . . the reason for the lack of effectiveness is that this standard fails to create an adequately robust regime, thus allowing for too many gaps and frictions between the various systems.’36 A new and better standard would be different from a ‘harmonized’ international tax regime:

30 IBFD News, ‘Sweden Proposes Terminating Tax Treaty with Greece, Sweden Proposes Terminating Tax Treaty with Portugal’ (23 Mar. 2021). 31 A. P. Dourado, ‘Introduction: International Tax Multilateralism or Reinforced Unilateralism?’, in Dourado, International and EU Tax Multilateralism: Challenges Raised by the MLI, 1. 32 Ibid. 33 A. Christians, ‘BEPS and the New International Tax Order’, BYU Law Review 6 (2016), 1603–1647. 34 D. Ring, ‘Who is Making International Tax Policy? International Organizations as Power Players in a High Stakes World’, Fordham International Law Journal 33 (2009/2010), 649–722. 35 T. Dagan, ‘Tax Treaties as a Network Product’, Brooklyn Journal of International Law 41 (2016), 1081–1106, 1081. 36 Ibid., 1090.

480 Miranda Stewart Whereas harmonization is aimed at curtailing tax competition so that countries can collect enough taxes to fund their welfare state, the standard this article envisions could streamline the tax regime and allow each country to freely (and efficiently) determine its own tax rates and packages of public services—thereby facilitating fiscal competition, while minimizing the costs of that competition.37

Developing countries, if they choose and have the resources, now have a seat at the table in the OECD-sponsored Global Forum and the Inclusive Framework, although they still have an inadequate voice in these forums. The agenda-setting and framework parameters established in multilateral processes provide little benefit for developing countries. However, the outcome of multilateral negotiations is unlikely to be worse than the current system, which provides few avenues for developing countries to increase their tax revenues, and it may be better. To support coherence, fairness and legitimacy, it will be necessary to ensure that multilateral processes can support the international tax jurisdiction, and tax collection, for developing countries.

27.2 Multilateral Process of International Tax Lawmaking 27.2.1 Model Treaties The use of model treaties is the most prominent example of ‘soft’ multilateralism in international tax. As a result of models, the ‘treaty based international tax regime already has multilateral features laced into its fundamental network-like structure, features explicitly amplified by the MLI [Multilateral Instrument]’.38 The first model tax treaty was drafted in 1928 by the Technical Experts. Further models were drafted in Mexico (1943) and London (1946) before the establishment of the OEEC and its successor, the OECD.39 The first Model Convention of the OECD was published in 1963.40 The Committee on Fiscal Affairs was established in 1971 and it produced the first OECD Model Convention with Commentary in 1977.41 In 1991, the approach to updating the

37

Ibid., 1092. Y. Brauner, ‘McBEPS: The MLI—The First Multilateral Tax Treaty That Has Never Been’, Intertax 46 (2018), 6–17, 7. 39 Reimer and Rust, Klaus Vogel on Double Taxation Conventions, 20. 40 OECD, Draft Double Taxation Convention on Income and on Capital (Paris: OECD Publishing, 1963); the Council of the OECD recommended it to all member Governments by a resolution of 30 July 1963. 41 OECD, Model Double Taxation Convention on Income and Capital (Paris: OECD Publishing, 1977). 38

Unilateralism, Bilateralism, and Multilateralism 481 Model and Commentary was made ‘ambulatory’ while, more recently, the views of non- OECD member states on the OECD Model were invited and their engagement in the model revision was encouraged. In Vogel on Double Tax Conventions, it is stated that the ‘main purpose’ of the OECD Model Convention is ‘to clarify, standardise, and confirm the fiscal situation of taxpayers who are engaged in commercial, industrial, financial, or any other activities in other countries through the application by all countries of common solutions to identical cases of double taxation’.42 The OECD, with a relatively small number of thirty- eight member states, has the capacity to make rules binding on its members, but instead makes recommendations established by consensus: ‘the choice to use its competence to make recommendations is a clear indication of the intention not to create binding rules’.43 The UN, with 193 member states, may claim a mandate for the widest representation in terms of multilateral taxation. Since 1980, the UN Committee of Experts on International Cooperation in Tax Matters for the Economic and Social Council of the UN has produced the UN Model Convention.44 The Committee was established in 2004 and was recently upgraded to meet twice a year.45 The Committee has twenty- five members including experts in tax policy and administration and appointed by governments with four-year terms. It provides non-binding recommendations and maintains and revises the UN Model, and a toolkit and manual on tax treaty negotiation between developing and developed nations. The potential for expansion of this role was indicated the UN General Assembly announced in 2023 that it would commence intergovernmental discussions ‘on ways to strengthen the inclusiveness and effectiveness of international tax cooperation through the evaluation of additional options, including the possibility of developing an international tax cooperation framework or instrument . . . taking into full consideration existing international and multilateral arrangements.’46 The UN Committee has insufficient resources to act as a global multilateral tax institution, so this development (which was opposed by most OECD member states), would require a shift in international tax politics to reorient support to the UN.

42

Reimer and Rust, Klaus Vogel on Double Taxation Conventions, 1. J. Schwarz, Schwarz on Tax Treaties, 5th ed. (Amsterdam: Kluwer Law International, 2018), 12. 44 UN Model Double Taxation Convention between Developed and Developing Countries, ST/ESA/ PAD/SER.E/213 (2017). See UN, ‘Charter of the United Nations’, 1 UNTS XVI (26 June 1945), art 681; UNECOSOC, https://www.un.org/development/desa/financing/what-we-do/ECOSOC/tax-committee/ tax-committee-home (accessed 29 March 2021). 45 UN, ECOSOC Res. 2004/69 (11 Nov. 2004). 46 Economic and Financial Committee (Second Committee) Promotion of inclusive and effective international tax cooperation at the United Nations LXXVII UN Doc A/C.2/77/L/11/Rev.1 (2022) at [2]‌. There is an ongoing debate about the role of the UN in international tax: e.g., S. Kasturirangan, ‘The United Nations Tax Committee as a Player in the International Tax Policy Discussion’, in A. Binder and V. Wohrer, eds, Special Features of the UN Model Convention (Vienna: Linde, 2019). 43

482 Miranda Stewart

27.2.2 Standards and Tax Policy Transfer Another way in which tax laws around the world may come to resemble each other based on an international norm is through policy transfer, including the adoption of multilateral standards to guide unilateral law-making. An example is the international tax rules for the VAT or Goods and Services Tax (GST). This multi-stage tax on supplies of goods or services, enacted in about 170 countries around the world, applies on a destination basis that establishes jurisdiction to tax in the country of the ultimate consumer. The spread of the VAT was at least in part as a tax reform imperative driven by the international financial institutions and developed countries to align developing country tax systems with the dominant WTO principles for free trade (by removing export taxes).47 The spread of VATs requires national tax legislation, and there remains significant diversity among ‘real’ VATs around the world, which are rather different from the benchmark ‘ideal’ VAT.48 The coordination of destination-based VAT has taken on a new life in response to global digital commerce.49 The OECD developed the Ottawa Taxation Framework (OTF) in 1998 to settle principles for global e-commerce. This enshrined the destination principle and established the goal of neutrality between domestic and foreign supplies. The OECD also established Working Party No. 9 on Consumption Taxes, which has had an increasing influence on national VAT laws and their updating for the global digital era.50 One reason was the early engagement of the OECD with countries outside its membership and with NGOs and businesses.51 The OTF requires that business-to-consumer (B2C) supplies are taxed in the place of consumption and business-to-business (B2B) supplies are taxed in the place of business presence of the recipient.52 For cross-border B2B supplies or imports, most VAT laws solve the issue of how to tax the importation with a ‘reverse charge’ mechanism that puts the obligation to pay the tax on the importer, instead of the offshore supplier. It has been observed, however, that in many countries around the world, no distinction is made between B2C and B2B supplies in the VAT law, and the place of performance or sale is

47 M. Stewart, ‘Global Trajectories of Tax Reform: The Discourse of Tax Reform in Developing and Transition Countries’, Trajectories of Tax Reform 44 (2003), 139–191. 48 K. James, The Rise of the Value-Added Tax (Cambridge: Cambridge University Press, 2015). 49 K. James and T. Ecker, ‘Relevance of the OECD International VAT/GST Guidelines for Non-OECD Countries’, Australian Tax Forum 32 (2017), 317–376, 335. 50 A. Cockfield, ‘The Rise of the OECD as an Informal “World Tax Organization” Through National Responses to E-Commerce Tax Challenges’, Yale Journal of Law & Technology 6 (2005/2006), 136, 160– 161; OECD, ‘Implementation of Ottawa Taxation Framework Conditions’ (2003); OECD Committee on Fiscal Affairs, ‘Electronic Commerce: Taxation Framework Conditions’ (1998); OECD, ‘Electronic Commerce: The Challenges to Tax Authorities and Taxpayers’ (1997). 51 D. Bentley, ‘International Constraints on National Tax Policy’, Tax Notes International 30 (2003), 1127–1140, 1140; Cockfield, ‘The Rise of the OECD as an Informal “World Tax Organization” Through National Responses to E-Commerce Tax Challenges’, 166. 52 Cockfield, ibid., 148.

Unilateralism, Bilateralism, and Multilateralism 483 the only jurisdictional rule. This means that the Guidelines may not be able to be fully implemented in many countries.53 The 2015 Global Forum on VAT hosted by the OECD included more than 100 countries.54 The OECD sees its role on VAT as an ‘inclusive standard setter’, responding dynamically to digitalization challenges through promulgation of the OECD VAT Recommendation, incorporating the Guidelines, on the application of VAT to international trade in services and intangibles.55 The OECD now aims to establish collection rules for digital platforms for the collection of VAT for multiple tax jurisdictions. The VAT Recommendation is, unlike the tax treaty Model and Commentary, described as an ‘OECD Act’ and recorded under the ‘legal instruments’ of the OECD, adopted by consensus between members (with reservations noted), and open to ‘adherence’ by non- OECD member states. However, only one non-member state, Costa Rica, has become an ‘adherent’ to the Recommendation.

27.3 Regionalism in Taxation Regional cooperation in international economic law usually arises between states that share geographic, political, ideological, or economic affinities.56 Some states have agreed to regional tax treaties, but this has not been as common in tax as in trade law. The EU has established the most comprehensive regional tax cooperation, although it is far from complete. Direct tax policy within the EU remains the ‘sole responsibility of member states’.57 The principle of subsidiarity is intended to ensure that member states retain sovereignty over direct taxation and a key protection in this regard is the requirement for unanimity in voting on tax matters. The European Council has established various binding directives for company and personal taxation and joint measures apply to prevent tax avoidance and double taxation, including relating to cross-border bank accounts; tax administration, arbitration, and transparency. In December 2022, the Council announced EU-wide agreement on a directive to implement the Pillar Two minimum tax (discussed below).

53 James

and Ecker, ‘Relevance of the OECD International VAT/GST Guidelines for Non-OECD Countries’, 337. 54 OECD, ‘Recommendation of the Council on the Application of Value Added Tax/ Goods and Services Tax to the International Trade in Services and Intangibles’ OECD/LEGAL/0430 (27 Sept. 2016), https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0430 (accessed 29 March 2021). 55 OECD, ‘International VAT/GST Guidelines’ (2017). 56 M. Forteau, ‘Regional Co-operation’ (last updated 2007), in Peters and Wolfrum, The Max Planck Encyclopedia of Public International Law, http://www.mpepil.com (accessed 7 August 2022). 57 See Publications Office of the EU, ‘Taxation’, https://eur-lex.europa.eu/summary/chapter/taxation. html?root_default=SUM_1_CODED%3D21, (accessed 29 March 2021).

484 Miranda Stewart The EU state aid regime provides a multilateral framework requiring member states to eliminate tax concessions that subsidise investment.58 However, member states of the EU have not entered into a multilateral income tax treaty, despite the significant coordination discussed earlier; instead, a network of bilateral tax treaties remains. EU indirect tax policy addressing cross-border trade in goods and services is more harmonized. The adoption by member states of a VAT that allocated tax jurisdiction on a destination basis was a requirement for membership. The framework for European VATs today is established by Council Directive 2006/112/EC.59 While there is a lot of variation between VATs in EU member states, this directive establishes fundamental parameters for the VAT base and rate in each country. The multilateral basis for VAT enabled the EU to respond quickly to tax issues arising from cross-border digital commerce. The Council Directive on e-commerce was issued in 200260 and led to early commitments to comply by digital firms such as Amazon and AOL/TimeWarner.61 The EU is a strong supporter of OECD work on VAT, discussed earlier. The EU leads in international cooperation for collection of cross-border VAT, through the One Stop Shop system; while desirable, there is no evidence so far that this will be copied elsewhere in the world.62 The Court of Justice of the European Union (CJEU) has had a significant effect on national tax laws of the member states. Hundreds of rulings by the CJEU on taxation have affected all kinds of taxes, from death and gift taxes, to personal and corporate income tax, and all kinds of taxpayers, from individuals to charities to small businesses and multinational enterprises (MNEs). The CJEU is multilateral because the court was established under a multilateral treaty; however, the impact is directly felt on national tax laws and has led to many governments being required to amend their national tax laws. The combination of CJEU rulings and EC legislation has been argued to significantly constrain unilateral taxing powers among the EU member states.63 One long-standing project of the EU, which has not reached fruition, is the establishment of a harmonized corporate tax. This project has been on foot at least since the Ruding Report of 1992.64 In 2001, the EC proposed a single consolidated corporate tax 58 EU,

‘State Aid’, https://ec.europa.eu/competition/state_aid/overview/index_en.html (accessed 21 March 2021). 59 EU Council Directive 2006/112/EC (28 Nov. 2006). 60 Council Directive 2002/38/EC amending Directive 77/338/EEC as regards the value added tax arrangements applicable to radio and television broadcasting services and electronically supplied services [2002] OJ L128/41. 61 Cockfield, ‘The Rise of the OECD as an Informal “World Tax Organization” Through National Responses to E-Commerce Tax Challenges’, 160–161. 62 See further https://ec.europa.eu/taxation_customs/business/vat/modernising-vat-cross-border- ecommerce_en, presenting all of the documents for the new VAT e-commerce rules applicable in the EU. 63 P. Genschel and M. Jachtenfuchs, ‘How the European Union Constrains the State: Multilevel Governance of Taxation’, European Journal of Political Research 50 (2011), 293–314. 64 Commission of the European Communities, ‘Company Taxation in the Internal Market’, COM(01)582 final (23 Oct. 2001), 8; O. H. Ruding, Report of the Committee of Independent Experts on Company Taxation (Commission of the European Communities, 1992); H. O. Ruding, ‘The David

Unilateralism, Bilateralism, and Multilateralism 485 base.65 It was proposed again in 2011, and the proposal was relaunched in 2016 as the Common Consolidated Corporate Tax Base (CCCTB).66 The UK was a strong opponent of the CCCTB and its departure from the EU may enable more agreement, although the Parliaments of several other EU member states including Ireland, Sweden, Denmark, Malta, Luxembourg, and the Netherlands also opposed the 2016 proposal. It has been suggested that vested interests, especially of large corporate taxpayers, prevent cross- border harmonization of the tax base, as this would reduce tax arbitrage opportunities, even though it would also reduce compliance costs.67 The most well-known regional tax agreement is the Nordic Tax Treaty.68 This includes Denmark, the Faroe Islands, Finland, Iceland, Norway, and Sweden, and is a full multilateral income tax treaty that has been updated several times. Other multilateral treaties have been agreed between the Andean Pact countries (Bolivia, Colombia, Ecuador, Peru, and Venezuela) and in the Caribbean Community (CARICOM) region.69 These tax treaties establish principles for relief of double taxation and allocation of tax jurisdiction, and may include scope for the adoption of different approaches even within the treaty. The South Asian Association for Regional Cooperation (SAARC), including Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan, and Sri Lanka, signed a more limited multilateral tax treaty in 2005, which was implemented in 2011.70 It applies to residents of the member states, and to income taxes or substantially similar taxes, but without most provisions of the OECD or UN Models. The SAARC treaty establishes a residence tiebreaker rule and includes substantive provisions to relieve tax on professors, students, and training, but mainly focuses on exchange of information and

R. Tillinghast Lecture: Tax Harmonization in Europe: The Pros and Cons’, Tax Law Review 54 (2000), 101; M. Stewart, ‘The David R Tillinghast Lecture: Commentary’, Tax Law Review 54 (2000), 111; H. W. Sinn, ‘Tax Harmonization and Tax Competition’ in Europe’, European Economic Review 34 (1990), 489. 65 EC,

‘Company Taxation: Commission Suggests Single Consolidated Tax Base’ IP/ 01/ 1468 (Oct. 2001). 66 EC, Proposal for a Council Directive on a Common Consolidated Corporate Tax Base, COM(2016) 683 final (25 Oct. 2016). 67 J. Roin, ‘Taxation without Coordination’, Journal of Legal Studies 31 (2002), 61. 68 Convention between the Nordic Countries for the Avoidance of Double Taxation with respect to Taxes on Income and Capital; see M. Helminen, The Nordic Multilateral Tax Treaty as a Model for a Multilateral EU Tax (Amsterdam: IBFD, 2014); N. Mattson, ‘Multilateral Tax Treaties: A Model for the Future?’, Intertax 28 (2000), 301–308. 69 Andean Community, ‘2004 Andean Community Income and Capital Tax Convention’, Decision 578 (5 May 2004); Agreement Among the Governments of the Member States of the Caribbean Community for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, Profits or Gains and Capital Gains and for the Encouragement of Regional Trade and Investment (1994); see discussion in Brooks, ‘The Potential of Multilateral Tax Treaties’, 211. 70 SAARC Limited Multilateral Agreement on Avoidance of Double Taxation and Mutual Administrative Assistance in Tax Matters (2005). A recent goal to expand tax and other cooperation was expressed in Press Release, ‘Secretary General of SAAR addressed the Fourteenth Informal Meeting of the SAARC Finance Ministers in Washington’ (19 Oct. 2019).

486 Miranda Stewart mutual assistance. Revenue agencies in the region aim to meet on a regular basis to cooperate in tax administration. In contrast, while the member countries in ASEAN have carried out various economic and trade integration activities including a free trade area, they have not yet established a multilateral tax treaty. Sunita Jogarajan has advocated for the adoption of a multilateral tax treaty in ASEAN to assist in removing barriers to investment, eliminate double taxation, and provide certainty.71 Jogarajan points to the success, albeit limited, of regional multilateral tax cooperation elsewhere, and suggests that this could be a useful model for the ASEAN region.

27.4 Multilateral Tax Administration The twenty-first century has seen a dramatic increase in cooperation between tax authorities in the administration of taxes and countering tax avoidance and evasion, supported by a widely adopted multilateral treaty.72 The Convention on Mutual Administrative Assistance in Tax Matters (the ‘Administrative Convention’) has been signed by 146 jurisdictions including seventeen jurisdictions covered by territorial extension and is in force in 136 jurisdictions.73 The Administrative Convention was given new life by the BEPS Project in 2011 and today it is the legal framework for a wide network supporting cooperative tax information exchange, audit, and debt collection across borders. The Administrative Convention has become a successful multilateral endeavour partly because it provides a framework under which governments may enter into bilateral arrangements to implement specific elements, while not implementing other elements of the convention. This has enabled significant data and information sharing. For example, under Article 6 of the Administrative Convention, the competent authorities of more than 110 countries have signed a multilateral agreement for automatic exchange of information based on a Common Reporting Standard (CRS), activating over 4,500 bilateral exchange relationships.74 The USA is an exception, having signed but not ratified the Administrative Convention. Governments may make reservations to the Administrative Convention. Many countries have reserved from the provision enabling assistance in collecting taxes, as they refuse to enforce the tax debts 71 S. Jogarajan, ‘A Multilateral Tax Treaty for ASEAN—Lessons from the Andean, Caribbean, Nordic and South Asian Nations’, Asian Journal of Comparative Law 6 (2011), 1–23. 72 M. Stewart, ‘Transnational Tax Information Exchange Networks: Steps Towards a Globalized, Legitimate Tax Administration’, World Tax Journal 4 (2012), 152–179. 73 OECD and Council of Europe, ‘The Multilateral Convention on Mutual Administrative in Tax Matters Amended by the 2010 Protocol’ (2011), https://www.oecd.org/tax/exchange-of-tax-information/ convention-on-mutual-administrative-assistance-in-tax-matters.htm (accessed 7 August 2022). 74 OECD, ‘CRS information’, https://www.oecd.org/tax/automatic-exchange/international-framew ork-for-the-crs/ (accessed 7 August 2022).

Unilateralism, Bilateralism, and Multilateralism 487 of other countries, retaining the so-called revenue rule which existed as customary international law but was overturned by the Administrative Convention.75 Despite the promise of multilateral tax administration, to date few developing countries, and none of the poorest countries in the world, have benefited from automatic tax information exchange. It appears that expansion of the multilateral cooperative tax network is a goal of the OECD; this requires its member states to work actively to support capability and extend cooperation including data sharing to developing countries. The BEPS Project Action 13 established country-by-country (CbC) reports for corporate profits of MNEs, now involving 91 countries.76 CbC reports are shared under over 3,000 ‘activated’ bilateral exchange agreements, adopting an automatic data exchange framework. This process is established in the Multilateral Competent Authority Agreement on the Exchange of CbC reports which supported the first automatic exchanges in 2018.77 The OECD and governments are working actively to expand these exchange networks. In contrast, the USA has applied the unilateral Foreign Account Tax Compliance Act (FATCA). FATCA compels financial institutions in foreign jurisdictions to provide information about US taxable subjects.78 The extraterritorial reach of this law would have breached the domestic laws of many countries. It required the negotiation of bilateral intergovernmental agreements (now numbering more than 100), and law reform in other countries, to implement the regime. The bilateral agreements were negotiated sometimes, but not always, on a reciprocal basis. Governments agreed to sign up to FATCA at the behest of local financial institutions, which risked losing their US customers if the reporting regime was not implemented. Although it appeared to be quite ‘aggressive’ unilateralism, FATCA rapidly set a global standard, as it was the trigger for the OECD to establish the CRS for automated sharing of financial data worldwide. Governments around the world could see how a bank account data-sharing regime could also assist them to enforce their own tax laws.

75 Council of Europe, ‘Reservations and Declarations for Treaty No. 127— Convention on Mutual Administrative Assistance in Tax Matters’ (2007). The US ‘will not provide assistance in the recovery of any tax claim . . . pursuant to Article 11 . . . of the Convention’; see https://www.coe.int/en/web/conv entions/full-list/-/conventions/treaty/127/declarations (accessed 29 March 2021). See W. Hellerstein, ‘The Rapidly Evolving Universe of US State Taxation of Cross-Border Online Sales after South Dakota v Wayfar, Inc., and its Implications for Australian Businesses’, eJournal of Tax Research 18 (2020), 320– 349, 345. 76 OECD, ‘OECD releases peer review documents for assessment of BEPS minimum standards (Actions 5 and 13)’ (2017); OECD ‘BEPS: Action 13 Country-by-Country Reporting’; OECD, ‘Transfer Pricing Documentation and Country-by-Country Reporting’ (2015). 77 OECD, ‘BEPS: Action 13 Country-by-Country Reporting’, http://www.oecd.org/tax/beps/beps-acti ons/action13/ (accessed 29 March 2021). 78 S. A. Dean and R. M. Kysar, ‘Introduction: Reconsidering the Tax Treaty’, Brooklyn Journal of International Law 41 (2016), 967–972, 967.

488 Miranda Stewart International tax cooperation can also operate by persuasion, peer review, and consensus among a group of countries.79 Such ‘soft law’ approaches may have a legal effect as they encourage governments to change their domestic tax law. The Global Forum on Transparency and Exchange of Information for Tax Purposes, established under the auspices of the BEPS Project and hosted by the OECD, now boasts 166 member states, making it the second largest intergovernmental tax organization after the UN itself. The peer review of country laws concerning exchange of information, bank secrecy, automated data sharing, and CbC reporting has been maintained in the last decade.80 The BEPS Project has also led to increased peer review of national tax laws to identify if they meet minimum standards for Action 5 (spontaneous country exchange of information on tax rulings for tax concessions) and Action 13 (CbC reporting). This peer review process is done by the Inclusive Framework for BEPS, another means by which the OECD is seeking to extend relevance and engage countries around the world. Peer review of harmful tax practices may lead to a recommendation for repeal of a tax concession in national tax law.81 Multilateral cooperation— perhaps more accurately described as plurilateral—is also carried out by tax authorities in various other ways. The most important institution is the Forum on Tax Administration (FTA) which includes representatives from over fifty revenue agencies.82 Its most recent initiative is the International Compliance Assurance Program (ICAP), which after being piloted over the previous two years, was established as a formal programme in December 2020.83 The ICAP has, so far, nineteen jurisdictions participating. It aims to leverage CbC reporting, sharing data, and managing multilateral risk assessment of MNEs ‘to determine each group’s specific international tax risks, including those related to transfer pricing and permanent establishment’.84 The participating agencies in ICAP are rather diverse, although most are 79 A.

Christians, ‘Hard Law, Soft Law, and International Taxation’, Wisconsin International Law Journal 25 (2007), 325–334; L. Philipps and M. Stewart, ‘Defining Fiscal Transparency: Transnational Norms, Domestic Laws, and the Politics of Budget Accountability’, Brooklyn Journal of International Law 34 (2009), 798–859. 80 See OECD, ‘Global Forum on Transparency and Exchange of Information for Tax Purposes’, http:// www.oecd.org/tax/transparency/ (accessed 7 August 2022); OECD, ‘Peer Review of the Automatic Exchange of Financial Account Information 2021’ (2021), https://www.oecd.org/tax/exchange-of-tax- information/peer-review-of-the-automatic-exchange-of-financial-account-information-2021-90bac5f5- en.htm (accessed 7 August 2022). 81 An example is the repeal by Australia of its Offshore Banking Unit regime which was found to be a harmful tax regime under Action 5 peer review: Australian Parliament, Treasury Laws Amendment (2021 Measures No. 2) Act No. 110, 2021 (Cth), Sch. 2, available at https://www.legislation.gov.au/Details/ C2021A00110. 82 See OECD, ‘Forum on Taxation’, http://www.oecd.org/tax/forum-on-tax-administration/about/ (accessed 29 March 2021). 83 Forum on Tax Administration, ‘2020 FTA ‘Amsterdam’ Plenary Communique’ (2020). 84 S. S. Johnston, ‘Multilateral Tax Risk assessment Program to Exit Pilot Phase’, Tax Notes International (9 Dec. 2020), available at https://www.taxnotes.com/tax-notes-today-international/trans fer-pricing/multilateral-tax-risk-assessment-program-exit-pilot-phase/2020/12/09/2d946 (accessed 7 August 2022).

Unilateralism, Bilateralism, and Multilateralism 489 sophisticated revenue agencies. They include many OECD countries; low-tax hubs of MNE activity, Luxembourg, and Singapore; and Russia. There are no countries from Latin or South America, or Africa, participating at present. Another FTA initiative, of longer standing but recently expanded in its membership, is the Joint International Task Force on Shared Intelligence and Collaboration (JITSIC) which now includes forty-two revenue agencies.85 The African Tax Administration Forum has thirty-eight member countries and engages with relevant international institutions to support capacity building and cooperation among member tax authorities, and is an active voice in international tax reform debates.86

27.5 Is the Twenty-First Century Multilateral? In the twenty-first century, governments have paid increased attention to the risk of ‘double non-taxation’, tax arbitrage, or tax avoidance through the combined application of tax treaties and national tax laws of different jurisdictions. This has mostly taken place under the G20-OECD BEPS Project which gained momentum after the global financial crisis of 2008. ‘Double non-taxation’ involves the combination of diverse country tax laws with the use of tax havens by MNEs in global tax structures. It relies on a tax treaty network that permits hybrid strategies enabling differential tax treatment in different jurisdictions, generating the phenomenon of international tax arbitrage or ‘stateless income’, resulting in low levels of taxation paid by some, mostly digital, MNEs around the world.87 The BEPS Project has been perceived as a move towards multilateralism with the potential to transform international tax.88 Apart from international tax administration, discussed previously, it has led to significant reforms of domestic tax law, for example on interest deductibility and transfer pricing. The BEPS project also produced the Multilateral Convention to Implement Tax Treaty Related Measures to Prevent

85 OECD, http://www.oecd.org/tax/forum-on-tax-administration/about/jitsic/ (accessed 7 August 2022). 86 See https://www.ataft

ax.org/(accessed 7 August 2022). H. D. Rosenbloom, ‘International Tax Arbitrage and the International Tax System’, Tax Law Review 53 (1998), 137; E. Kleinbard, ‘Stateless Income and its Remedies’, in T. Pogge and K. Mehta, eds, Global Tax Fairness (Oxford: Oxford University Press, online edn, 24 Mar. 2016), https://doi.org/10.1093/acprof:oso/ 9780198725343.003.0006 (accessed 7 August 2022). 88 See, e.g., R. Mason, ‘The Transformation of International Tax’, American Journal of International Law 114/3 (2020), 353–402; P. Pistone, ‘Coordinating the Action of Regional and Global Players during the Shift from Bilateralism to Multilateralism in International Tax Law’, World Tax Journal 6 (2014), 1, 3–9; I. Mosquera Valderrama, ‘Legitimacy and the Making of International Tax Law: The Challenges of Multilateralism’, World Tax Journal 7 (2015), 343. 87

490 Miranda Stewart Base Erosion and Profit Shifting also known as the Multilateral Instrument (MLI).89 In February 2015, the G20 provided a mandate90 under Action 15 of the BEPS Project, for the MLI. By July 2022, ninety-nine jurisdictions had signed the MLI, and it was in force in seventy-seven jurisdictions. The MLI is bolted on top of the network of bilateral income tax treaties based on the OECD and UN Models. As a matter of international law, the MLI is said to ‘modify’ all covered tax agreements (art. 1) without amending them. This was explained by the OECD as follows: ‘the Convention does not function in the same way as an amending protocol to a single existing tax treaty which would set out amendments to the text of specified provisions of the tax treaty. Instead, the MLI is applied alongside existing bilateral tax treaties.’91 The approach has been suggested as a model for reform of the similarly large network of bilateral investment treaties around the world.92 The MLI has helped to establish minimum standards for tax treaties on a multilateral basis. Most importantly, it confirms an old tax treaty principle that tax treaties should prevent tax avoidance and evasion, as well as provide relief from double taxation. The MLI includes a principal purpose pest that outlaws treaty abuse as a minimum standard. However, some suggest that this may not change business as usual for taxpayers or governments, in particular when the USA refuses to participate.93 Wei Cui has questioned the value of multilateralism in international tax, in an era when ‘unilateralism’ may be interpreted as ‘without the consent of the United States’.94 Yariv Brauner argues that far from achieving a multilaterally agreed international tax regime: the core mission of the MLI is to support the current, bilateral treaties based international tax regime and to empower its hold on the tax world. This mission is consistent with BEPS that evolved into a political project primarily designed to preserve the institutional status of the OECD as the caretaker of the international tax regime.95

89 OECD, ‘Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting’ (2017), available at https://www.oecd.org/tax/treaties/multilateral-convent ion-to-implement-tax-treaty-related-measures-to-prevent-beps.htm (accessed 7 August 2022). 90 OECD, ‘Mandate for a Multilateral Instrument on Tax Treaty Measures to Tackle BEPS’ (2015). 91 OECD Directorate for Legal Affairs, ‘Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting: Functioning under Public International Law’ (2015), paras 15–16; OECD, ‘Explanatory Statement to The Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting’ (2016), para 13. 92 W.

Alschner, ‘The OECD Multilateral Tax Instrument: A Model for Reforming the International Investment Regime?’, Brooklyn Law Journal 45/1 (2019), 1–73. 93 C. A. Brown and J. Bogle, ‘Treaty Shopping and the New Multilateral Tax Agreement— Is It Business as Usual in Canada?’, Dalhousie Law Journal 43/1 (2020), 1–32. 94 W. Cui, ‘What is Unilateralism in International Taxation?’, Symposium on Ruth Mason, ‘The Transformation of International Tax’, American Journal of International Law 114 (2020), 260–264, doi:10.1017/aju.2020.52. 95 Brauner, ‘McBEPS: The MLI—The First Multilateral Tax Treaty That Has Never Been’, 16.

Unilateralism, Bilateralism, and Multilateralism 491 A different view is presented by Ruth Mason who argues that BEPS ‘reflected—and to a significant extent operationalised—major changes in the participants, agenda, institutions, norms, and legal instruments of international tax’.96 These changes involved, she suggests, wider agreement on an international norm of ‘full taxation’ of income and the establishment of various minimum standards, for example in transfer pricing and anti-abuse. While the BEPS Project in general, and the MLI specifically, does reinforce bilateralism, this is consistent with the role of multilateral agreements in other areas of international economic law. The new multilateralism in international tax is layered with bilateralism and unilateralism, continuing to protect the interests of developed countries and of MNEs in the process.97 What should be the goal of multilateralism in international tax? Brauner has proposed a ‘partial and gradual rule-harmonization solution to current international tax challenges’98 and advocated a multilateral treaty so as to achieve such global harmonization. Dagan suggests that a ‘standardized’ international tax regime: could better determine source and residency rules. It could also set a uniform formula for determining transfer prices, incorporate anti-abuse arrangements, and coordinate the rules on flow-through and opaque entities. Detailed mechanisms and technological compatibility could be set for the collection and dissemination of information among enforcement agencies and could facilitate mutual assistance.

Some of these goals have been partly achieved in recent multilateral developments. However, other goals such as the standardization of source and residency rules have not been achieved. There remains significant diversity, and conflict, in fundamental tax jurisdictional concepts as they are enacted in national tax laws. Multilateral tax agreements do two things. First, they require countries to change national tax laws. We have already discussed examples, including the requirements for governments to end bank secrecy, abolish harmful tax practices, or enact laws to support sharing CbC reports on corporate profits. Secondly, multilateral tax agreements take diverse national tax laws as they find them and establish a rule to improve the integration of these rules. An example is BEPS Action 2 on hybrid instruments, which addresses the consequences of hybrid instruments that have differential tax treatment across countries by producing a single tax outcome where two countries approach the instrument entity differently. This is an example of ‘coordinated unilateralism’.99 In what is undoubtedly a complex process, this has the effect of interlocking two country systems

96

R. Mason, ‘The Transformation of International Tax’, American Journal of International Law 114/3 (2020), 353–402, 354. 97 R. G. Anton, ‘Multilateral Dynamics in Bilateral Settings: Back to Realpolitik’, British Tax Review 4 (2019), 462–486. 98 Y. Brauner, ‘An International Tax Regime in Crystallization’, Tax Law Review 56 (2003), 259, 262. See also Brooks, ‘The Potential of Multilateral Tax Treaties’. 99 Ibid., 374.

492 Miranda Stewart within a multilateral framework, to eliminate ‘double non-taxation’ arising from hybrid tax arbitrage. In 2021, multilateral negotiations in the BEPS Inclusive Framework of 144 countries produced the Statement on a Two-Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy.100 The two-pillar solution, if implemented, will in some respects be transformative of international tax law. Pillar One would establish an entirely new tax jurisdictional rule for the largest global MNEs, including a new tax base of global accounting profits, formulary apportionment, and nexus test to the market, or consumer country. This is proposed to be established in a multilateral convention, which may be negotiated in 2023 and take effect at the earliest in 2024. The rules would apply only to the largest MNEs, with global revenues in excess of €20 billion in 2023 (reducing to €10 billion by 2030), and excluding resource and financial corporations. If enacted in its proposed form, Pillar One would calculate 25% of residual profit of in-scope MNEs above a 10% profit threshold, determined on global financial accounting profit (adjusted for tax). This proportion of residual profit would be allocated to the market, or consumer, countries in which the MNE operates, to be taxed by that country. It is proposed that there will be a binding dispute-resolution process to address conflicts, and centralized interpretive guidance. Pillar Two will empower (and require) states which sign up to ensure a minimum level of taxation for MNEs with global revenues over the threshold of €750 million in 2023, at a 15% effective tax rate. This will apply through a set of interlocking rules to be enacted in domestic law by participating countries, consistent with Model Rules established by the OECD Inclusive Framework. The Income Inclusion Rule (IIR) will authorise the revenue authority jurisdiction where a parent or higher level group company of the MNE is located, to include and tax income of the MNE where it is subject to tax lower than the effective 15% tax rate in another jurisdiction. If there is no jurisdiction that applies the IIR to the MNE, the Under Taxed Profits Rule (UTPR) will permit the revenue authority of another jurisdiction to deny a deduction for a related outgoing payment, or otherwise to levy tax on the undertaxed profits. The IIR is intended to commence in 2023 and the UTPR in 2024. Pillar Two will also support countries to include a qualifying domestic minimum tax, to ensure that they reach the 15% effective tax threshold, and introduces a new minimum withholding tax on base-eroding payments in bilateral tax treaties, likely through a multilateral instrument similar to the MLI. The design of Pillar Two enables a country that has unilaterally enacted legislation consistent with the Model Rules to tax an amount that arises in another, lower-taxed jurisdiction; in the case of the UTPR, this can occur even if there is no nexus of the MNE in that jurisdiction. This fundamentally changes previous rules of tax jurisdiction. 100 OECD/G20

Inclusive Framework on BEPS, ‘Statement on a Two-Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy’ (8 Oct. 2021), agreed by 136 member jurisdictions on 8 Oct. 2021, further developing the work of the OECD in its Blueprint reports, https:// www.oecd.org/tax/beps/oecd-g20-inclusive-framework-on-beps-invites-public-input-on-the-reports- on-pillar-one-and-pillar-two-blueprints.htm.

Unilateralism, Bilateralism, and Multilateralism 493 The two-pillar solution is intended to operate as a compromise package, with both Pillars One and Two being adopted together. An essential element of the two-pillar consensus was intended to be the replacement of unilateral measures with the multilateral agreement. This would include the abolition of unilateral Digital Services Taxes (DSTs) that have been adopted, or proposed, by about forty countries around the world. A DST is usually applied to the gross revenues of large digital MNEs operating in market jurisdictions, for example providing social media, e-commerce, or advertising services with users or consumers in the jurisdiction. Unilateral DSTs on inbound e-commerce from large digital MNEs have raised significant revenue in some countries. Under the previous US government, they led to a reaction which argued that they are a discriminatory trade measure. The US Trade Representative under President Trump began to investigate DSTs enacted by Austria, India, Italy, Spain, Turkey, and the UK, on the basis that they excessively burdened US digital companies and were inconsistent with the principles of international taxation.101 The interaction between unilateral DSTs and the multilateral Inclusive Framework evidence a complex multilateral bargaining process about global digital taxation, involving states, MNEs, and multilateral forums such as the OECD and EU. If implemented, the consensus decision by the Inclusive Framework on the two-pillar solution may indeed be a transformation shifting international tax towards multilateralism and a new equilibrium. Despite this ambition, it is not clear that Pillar One will be achieved. Pillar Two seems more likely, but this fundamentally changes the apparent overall two-pillar bargain. Work continues on the details of the two-pillar solution in OECD technical working parties and in the Inclusive Framework.102 As at April 2023, various countries around the world have committed to Pillar Two, including some which have already enacted legislation. Committed countries include the EU member states (committed by European Council Directive), the United Kingdom, Japan, South Korea, Canada, Mexico, Australia, New Zealand, Qatar, United Arab Emirates, Thailand, Malaysia, Indonesia, Singapore and Hong Kong. Most propose enactment or commencement of legislation effective from 2024.103 Major economies including the USA, China and Brazil have not yet announced their position. Despite these significant gaps, the Pillar two rules may operate effectively with sufficient country engagement.

101 Office of the US Trade Representative, ‘USTR announces next steps of Section 301 Digital Services Taxes Investigations’ (26 Mar. 2021) https://ustr.gov/about-us/policy-offices/press-offi ce/press-releases/ 2021/march/ustr-announces-next-steps-section-301-digital-services-taxes-investigations (accessed 29 March 2021). 102 OECD, ‘Progress Report on Amount A of Pillar One, Public Consultation’ (July 2022), https:// www.oecd.org/tax/beps/progress-report-on-amount-a-of-pillar-one-july-2022.pdf (accessed 7 August 2022); OECD, ‘Tax Challenges Arising from the Digitalisation of the Economy Global Anti-Base Erosion Model Rules (Pillar Two)’ (2021), https://www.oecd.org/tax/beps/tax-challenges-arising-from- the-digitalisation-of-the-economy-global-anti-base-erosion-model-rules-pillar-two.pdf (accessed 7 August 2022) 103 KPMG, BEPS 2.0 State of Play (April 2023), https://assets.kpmg.com/content/dam/kpmg/xx/pdf/ 2023/04/beps-2-0-state-of-play-april-2023.pdf

494 Miranda Stewart In particular, the USA seems unlikely to legislate domestic rules for Pillar Two, or to sign the multilateral treaty for Pillar One. The lack of progress on Pillar One in particular means that there will be ongoing instability in taxation of MNEs, as countries continue to implement DSTs. Meanwhile, many countries continue with other unilateral solutions, such as tightening up domestic anti-avoidance rules.

27.6 Conclusion Multilateralism has a chequered history in international tax. Yet despite challenges, this author suggests that the two-pillar solution and other developments show that international tax is becoming increasingly multilateral. The formalization through treaty negotiation of multilateral rules, and the harmonization of domestic tax law applicable to global taxpayers, continues to expand, in terms of both the number of countries involved and the topics addressed. Increasing agitation by developing countries for full participation to support their own interests in multilateral tax forums, although scarcely achieved, contributes to legitimating multilateralism as a process for establishing international tax law. In the two-pillar negotiations, and through the more successful measures aimed at levying VAT on global digital commerce, states are engaging in cooperative action to address tax issues that go beyond the competence of any one state. In future, this kind of multilateral action may address taxation in respect of other extra- sovereign contexts, such as outer space.104 The administration of international taxation is also becoming increasingly multilateral. However, developing countries benefit little from established multilateral tax administrative systems and forums; effectiveness and legitimacy of this system requires governments, especially OECD member states, to continue to scale up international cooperation. This leads to the aspiration for a genuinely global tax organization. The League of Nations had originally considered that ‘tricky’ issues in international tax were to be dealt with by a permanent international organization.105 The OECD, as collector of tax statistics, tax researcher, rule designer, and secretariat for the Global Forum and Inclusive Framework, has become an agenda-setter because of the powerful developed country members that actively engage in it. This makes it a ‘de facto’ world tax organization.106 However, many would resist this idea, seeing it as unrepresentative and with vested interests of a small membership of rich countries. Recent developments indicate that the UN may obtain renewed credibility as a forum for international tax law-making in future. This may only be feasible through collaboration between the UN and the 104 G.

Zeyen, ‘Taxation of Outer Space Income Resulting from Air Transport or Employment Activities: Is the OECD Model Convention an Appropriate Tool?’, Intertax 49 (2021), 333. 105 Jogarajan, Double Taxation and the League of Nations, 99. 106 See, e.g., Brauner, ‘An International Tax Regime in Crystallization’; Cockfield, ‘The Rise of the OECD as an Informal “World Tax Organization” ’.

Unilateralism, Bilateralism, and Multilateralism 495 OECD; alternatively, to develop into a truly legitimate and effective world tax organization, the OECD itself must change through genuine expanded multilateralism. Multilateralism in international tax does not replace, but rather combines with unilateralism, as taxation remains tied to national sovereignty. National tax systems will continue to differ in terms of the level and types of tax, and the details of rates and bases. The transfer of policy ideas and approaches in tax law and policy continues to increase but the adoption and implementation of these ideas will remain diverse and partial, responding to the specific political and economic situation of states. This diversity can be accommodated in an increasingly multilateral framework of interlocking international tax laws.

Chapter 28

Age nts in Inte rnat i ona l Tax Treat i e s Sunita Jogarajan

28.1 Introduction One of the most common ways through which an enterprise may undertake business in another jurisdiction is through the provision of services by an agent. Where the business activities are undertaken by a dependent agent, rather than an independent agent, the enterprise may be subject to tax in the other jurisdiction on the profits arising from the dependent agent’s activities. This is on the basis that the dependent agent constitutes a ‘permanent establishment’ of the enterprise. Under most, if not all, of the more than 3,000 bilateral tax treaties currently in existence, a foreign-resident enterprise is not subject to tax in a jurisdiction on business profits arising in that jurisdiction unless it has a ‘permanent establishment’ there. The inclusion of dependent agents as permanent establishments has a long history dating back to 1885.1 However, the determination as to whether an agent is a dependent agent and thereby a permanent establishment can be a tricky matter.2 One issue that may arise is whether the agent is a dependent agent or an

1 B. Walker, ‘The Evolution of the Agency Permanent Establishment Concept’, Australasian Tax Teachers Conference, Melbourne (2018). 2 See, e.g., G. Persico, ‘Agency Permanent Establishment under Article 5 of the OECD Model Convention’, Intertax 28 (2000), 66; C. Panayi, ‘Agency Permanent Establishments in Securitization Transactions’, Intertax 33 (2005), 286; R. Vann, ‘Travellers, Tax Policy and Agency Permanent Establishments’, BTR 6 (2010), 538; C. Dunahoo, ‘Contract Conclusion and Agency Permanent Establishments: Here, There, and Everywhere?’, in P. Baker and C. Bobbett, eds, Tax Polymath: A Life in International Taxation: Essays in Honour of John F. Avery Jones (Amsterdam: IBFD, 2010); H. Pijl, ‘Agency Permanent Establishments: In the Name of and the Relationship between Article 5(5) and (6): Part 1’, BFIT 67/1 (2013), 3 and ‘Agency Permanent Establishments: In the Name of and the Relationship between Article 5(5) and (6): Part 2’, ibid, 62; J. F. Avery Jones and J. Lüdicke, ‘The Origins of Article 5(5) and 5(6) of the OECD Model’, WTJ (2014), 203, 204–206 (hereafter Jones and Lüdicke, ‘OECD Model’).

498 Sunita Jogarajan independent agent. The distinction is, of course, significant as in the case of the former, the foreign enterprise is subject to tax in the agent’s jurisdiction, whereas in the latter case, the agent, and not the foreign enterprise, is subject to tax in that jurisdiction. This chapter examines one particular issue in relation to agents in international tax treaties—the treatment of commissionaire arrangements. The question here is whether the commissionaire constitutes a ‘dependent agent’. The issue has arisen due to different conceptions of agents under civil law and common law. The potential for differing tax treatment arising from the different concepts of agents was recognized during the drafting of the first model tax treaties in 1928 but the issue was left unresolved. However, the potential to avoid taxation or minimize taxes through the use of commissionaire arrangements has recently been addressed through the coordinated global effort to target base erosion and profit shifting (BEPS). The next section provides a brief history of the agency permanent establishment provision in model tax treaties prior to the latest amendments in response to BEPS. Section 28.3 explains what a commissionaire arrangement is, and the tax consequences of such arrangements. Section 28.4 discusses the reforms to address commissionaire arrangements arising out of the work on BEPS. Finally, Section 28.5 concludes the chapter by suggesting that, despite the efforts to address commissionaire arrangements, the problems in relation to dependent agency permanent establishments are likely to persist.

28.2 Model Tax Treaties This section examines the inclusion of the concept of an agency permanent establishment in selected model tax treaties.3 Modern-day tax treaties are largely based on model treaties and the model treaties have developed along a fairly continuous path and, thus, understanding the historical development of the model treaties is crucial in understanding current tax treaty practice and in guiding future reform.4 As will be seen, the different views of agency under civil law and common law which give rise to the problem with the commissionaire arrangements discussed later, was already apparent at the outset. Further, the agency provision in the OECD Model was developed following principles established by the League of Nations and has not changed substantially since (prior to the latest amendments to address BEPS).5 3 For

a comprehensive account of the history of the agency permanent establishment provision in international tax treaties, see Walker, ‘The Evolution of the Agency Permanent Establishment Concept’. 4 R. Vann, ‘Writing Tax Treaty History’, in G. Maisto, ed., Current Tax Treaty Issues: 50th Anniversary of the International Tax Group (Amsterdam: IBFD, 2020), 15–40. Vann notes that tax treaty history is not only relevant in interpreting current tax treaties, but is also relevant in developing tax policy, matching historical experience with theoretical approaches, and in understanding how things happened and why. See also L. Friedlander and S. Wilkie, ‘Policy Forum: The History of Tax Treaty Provisions—And Why It Is Important To Know About It’, Canadian Tax Journal 54 (2006), 907. 5 T. Sokolov, ‘The History of Article 5 of the OECD Model Convention’, in F. Brugger and P. Plansky, eds, Permanent Establishments in International and EU Tax Law (Amsterdam: IBFD, 2020), 28.

Agents in International Tax Treaties 499 The very first model international tax treaties, published by the League of Nations in 1928, included agents as one of the activities potentially giving rise to a permanent establishment.6 The three model treaties published by the League of Nations in 1928 were titled Draft Convention Ia, Draft Convention Ib, and Draft Convention Ic. Draft Convention Ia distinguished between personal and impersonal taxes while Draft Conventions Ib and Ic did not.7 All three Draft Conventions included a provision along the following lines:8 Income derived from any industrial, commercial or agricultural undertaking and from any other trades or professions . . . shall be taxable in the State in which the permanent establishments are situated. The real centres of management, branches, mining and oil fields, factories, workshops, agencies, warehouses, offices, depots, shall be regarded as permanent establishments. The fact that an undertaking has business dealings with a foreign country through a bona-fide agent of independent status (broker, commission agent, etc) shall not be held to meant that the undertaking in question has a permanent establishment in that country.

The discussion of the agency provision in the 1928 model treaties was rather limited, as by that stage the government experts were tasked with finalizing the model treaties and were instructed that the time for academic or theoretical discussion had passed.9 As such, it is necessary to examine the work of the government experts convened by the League which preceded the publication of the 1928 model treaties.10 The first of these was

6

League of Nations, ‘Double Taxation and Tax Evasion: Report presented by the General Meeting of Government Experts on Double Taxation and Tax Evasion’(League of Nations 1928) (1928 Report). The 1928 model treaties were developed at a general meeting of experts from twenty-seven countries. Walker traces the establishment of the concept of an agency as a permanent establishment to the German states in the late nineteenth century: Walker, ‘The Evolution of the Agency Permanent Establishment Concept’. 7 On the background as to why the League of Nations published three model treaties in 1928, see S. Jogarajan, Double Taxation and the League of Nations (Cambridge: Cambridge University Press, 2018), 161–164, 182–183. Draft Convention Ic, developed by the French and German representatives, is in fact the model that has prevailed as ‘the model’: J. Avery Jones, ‘Categorising Income for the OECD Model’, in L. Hinnekens and P. Hinnekens, eds, A Vision of Taxes within and Outside European Borders (Alpen aan den Rijn: Wolters Kluwer, 2008), 99. 8 Art. 5, Draft Convention Ia: 1928 Report, 8; art. 2, Draft Convention Ib: 1928 Report, 16; art. 3, Draft Convention Ic: 1928 Report, 19. 9 League of Nations, ‘Minutes of the First Meeting of the General Meeting of Government Experts on Double Taxation and Tax Evasion’, D.T./Reunion/P.V.1 (1928). 10 The work of the government experts was preceded by a theoretical examination of the issue of international double taxation by four eminent economists from Italy, the Netherlands, UK, and the USA: G. Bruins et al., Report on Double Taxation Submitted to the Financial Committee (Geneva: League of Nations, 1923). This study considered the appropriate methods of addressing international double taxation, and the allocation of taxing rights in the case of shared jurisdiction to tax. The use of agencies was contemplated in the report but not examined in detail. On the publication of the 1923 Report, see S. Jogarajan, ‘Stamp, Seligman and the Drafting of the 1923 Experts’ Report on Double Taxation’, WTJ (2013), 368, 368–392.

500 Sunita Jogarajan the development of a series of resolutions on double taxation and tax evasion by seven government experts.11 Of relevance here is the resolution relating to the imposition of impersonal or schedular taxes in relation to industrial and commercial establishments which stated:12 If the enterprise has its head office in one of the States and in another has a branch, an agency, an establishment, a stable commercial or industrial organisation, or a permanent representative, each one of the contracting States shall tax that portion of the net income produced in its own territory.

When discussing the proposed resolution, the experts noted that a distinction needed to be drawn between an agent acting in the name of the company and a commission agent.13 In the case of the former, the company would be taxed, whereas in the case of the latter, the agent would be taxed. The publication of the resolutions on double taxation and tax evasion in 1925 was followed by the development of a draft model treaty by government experts from thirteen countries.14 The 1927 draft model tax treaty included a provision on agency permanent establishment similar to the provision in the final model treaties published in 1928. The main issue discussed by the experts when drafting the provision was the distinction between dependent agents (treated as permanent establishments of the company or business) and independent agents (taxed in their own right and not treated as permanent establishments of the company or business). This resulted in the text regarding a bona fide agent of independent status not being a permanent establishment being added to the provision.15 The discussions in 1927 also highlighted the difference in the tax treatment of agents as between common law and civil law countries.16 The experts discussed the example of a person working in an office in France to sell goods

11 Technical Experts to the Financial Committee of the League of Nations, ‘Double Taxation and Tax Evasion: Report and Resolutions’ (League of Nations, 1925;). The seven government experts were from Belgium, Czechoslovakia, France, Great Britain, Italy, Netherlands, and Switzerland. 12 Ibid., 31 (emphasis added). On impersonal or schedular taxes versus personal or general income tax, see J. Avery Jones, ‘Avoiding Double Taxation: Credit versus Exemption—The Origins’, BFIT 66 (2012), 67, 67–70. 13 League of Nations, ‘Minutes of the Fifth Meeting of the First Session of the Committee of Government Experts on Double Taxation and the Evasion of Taxation’, E.F.S./D.T./1–4 Session/P.V; E.F.S./D.T./P.V.5.(1) (1923). Note that ‘commission agent’ is not the same as a ‘commissionaire’ (discussed later): D. Ward and J. Avery Jones, ‘Agents as Permanent Establishments Under the OECD Model Tax Convention’, BTR 5 (1993), 341, 343. 14 League of Nations, Committee of Technical Experts on Double Taxation and Tax Evasion, ‘Double Taxation and Tax Evasion’, C.216.M.85.1927.II (1927). The experts from the original seven countries were joined by experts from Argentina, Germany, Japan, Poland, the USA, and Venezuela. 15 Jogarajan, Double Taxation, 142. The experts initially wanted to include the clarification in the commentary to the draft model treaty, but the issue was considered too significant to be relegated to the commentary and was eventually included in the text of the provision. 16 League of Nations, ‘Minutes of the Fourth Meeting of the Seventh Session of the Committee on Double Taxation and Fiscal Evasion’, D.T./7th Session/P.V.4(1) (1927).

Agents in International Tax Treaties 501 from a British firm in London. The French expert noted that if the person had the power to sign contracts on behalf of the British firm, the French office would be considered a permanent establishment of the British firm. However, if the person’s role were limited to connecting the British seller with French buyers for a fee, the French authorities would have no grounds to tax the British firm. The British expert explained that there was no such distinction in England. The League’s 1928 model treaties are generally considered one of the few successes of the League’s work in commercial policy during the interwar period and more than a hundred bilateral tax treaties, based on those models, were concluded between 1929 and 1939.17 The League published a further two model treaties in 1943 and 1946 (the Mexico and London Models respectively). For political as well as other reasons, these models were of limited practical impact although the London Model did serve as the basis for the development of the first OECD Model in 1963.18 The League’s successor, the United Nations, continued some of the League’s work in international taxation,19 but did not publish a model tax treaty until 1980 (discussed later). The OECD Model was first published in 1963 after seven years of work by the Fiscal Committee.20 The OECD Model has had a significant influence on bilateral tax treaties with research showing that bilateral tax treaties demonstrate a clear trend of convergence towards the language of the OECD Model.21 Of relevance to this chapter, paragraphs 5 and 6 of article 5 of the OECD Model prior to the amendments to address BEPS (discussed later) stated:22 Notwithstanding the provisions of paragraphs 1 and 2, where a person—other than an agent of an independent status to whom paragraph 6 applies—is acting on behalf of an enterprise and has, and habitually exercises, in a Contracting State an authority

17 A. Loveday, ‘The Economic and Financial Activities of the League’, International Affairs 17/6 (1938), 788, 790; A. Sweetser, ‘The Non-Political Achievements of the League’, Foreign Affairs 19/1 (1940), 179, 183; League of Nations, Commercial Policy in the Interwar Period: International Proposals and National Policies (London: Allen and Unwin, 1942), 30; ‘The Progressive Development of International Law by the League of Nations’ AJIL 41 (1947), 49, 55. However, Spitaler argues that the 1928 models were of limited use and the early German treaties were more influential: A. Spitaler, Das Doppelbesteuerungsproblem: Bei Den Direckten Steuern (Reichenberg: Stiepel, 1936), 32–46. 18 M. Carroll, ‘International Tax Law— Benefits for American Investors and Enterprises Abroad’, International Lawyer 2 (1968), 692, 707–709. 19 e.g. the UN continued to publish compilations of all international agreements and domestic provisions addressing double taxation and fiscal evasion, work that had been undertaken by the League between 1928 and 1936: UN Department of Technical Cooperation for Development, International Tax Agreements (Geneva: United Nations, 1948). 20 OECD, Model Convention on Income and on Capital (Paris: OECD Publishing 2017), 10 (OECD Model (2017)). 21 E. Ash and O. Y. Marian, ‘The Making of International Tax Law: Empirical Evidence from Natural Language Processing’, UC Irvine School of Law Research Paper No 2019-02 (2019), https://ssrn.com/ abstract=3314310 (accessed 1 April 2021). 22 OECD, Model Convention on Income and on Capital (Paris: OECD Publishing, 2014), 27 (OECD Model (2014)).

502 Sunita Jogarajan to conclude contracts in the name of the enterprise, that enterprise shall be deemed to have a permanent establishment in that State in respect of any activities which that person undertakes for the enterprise, unless the activities of such person are limited to those mentioned in paragraph 4 which, if exercised through a fixed place of business, would not make this fixed place of business a permanent establishment under the provisions of that paragraph. An enterprise shall not be deemed to have a permanent establishment in a Contracting State merely because it carries on business in that State through a broker, general commission agent or any other agent of an independent status, provided that such persons are acting in the ordinary course of their business.

The UN Model was first published in 1980 as it was recognized that, in order to promote greater investment into developing countries, it was necessary to prevent or eliminate international double taxation arising in relation to such investments.23 As had already been recognized since the period of the League of Nations, bilateral tax treaties are perhaps the best mechanism of preventing or eliminating international double taxation. However, existing treaties at the time, largely based on the OECD Model, were not considered suitable for encouraging the conclusion of tax treaties between developed and developing countries and the UN Model was developed for this purpose. The UN Model used the OECD Model and Commentary as its main reference but adapted it to meet the needs of developing countries.24 Although the UN Model continues the compromise between the source and residence principles established in the 1920s, the UN Model generally gives more weight to the source principle as compared to the OECD Model.25 Of relevance to this chapter, article 5 of the UN Model (prior to the amendments to address BEPS) is similar to article 5 of the OECD Model mentioned earlier; but, since its inception, the UN Model has included a stock agent provision. Specifically, paragraph 5(b) of article 5 of the UN Model deems a permanent establishment to exist where a person in one jurisdiction, other than an independent agent, habitually maintains a stock of goods or merchandise from which they regularly deliver goods or merchandise on behalf of an enterprise in the other jurisdiction, even where they do not have the authority to conclude contracts in the name of the enterprise. The accompanying commentary notes that this paragraph represents a significant departure 23 Department

of Economic & Social Affairs, United Nations Model Double Taxation Convention Between Developed and Developing Countries (Geneva: United Nations, 1980), 1–3 (UN Model (1980)). 24 The OECD Model and Commentary were used as the basis for the UN Model for two reasons. First, it was recognized that the OECD Model and Commentary embodied decades of useful technical expertise. Secondly, from a practical perspective, the OECD Model and Commentary were already being used by OECD members not just in negotiations with each other but also with developing countries and it was thought that familiarity would promote the use of the UN Model: UN Model (1980), 4. This is consistent with studies of other international regimes which found that countries are bound by a certain level of path dependence and more likely to comply or follow with the familiar: A. Chayes and A. Chayes, The New Sovereignty: Compliance with International Regulatory Agreements (Cambridge, MA: Harvard University Press, 1995), 10. 25 Department of Economic & Social Affairs, Model Double Taxation Convention between Developed and Developing Countries (Geneva: United Nations, 2017), iii, vi (UN Model (2017)).

Agents in International Tax Treaties 503 from the OECD Model and considerably broadened the scope of the provision.26 It was thought that the OECD Model provision was too narrow and potentially restricted the type of agent that would constitute a permanent establishment thereby potentially limiting source-country taxation. Despite the inclusion of the stock agent provision, the Committee of Experts on International Cooperation in Tax Matters decided to modify paragraph 5 of article 5 of the UN Model to incorporate the OECD Model amendments to address the artificial avoidance of permanent establishment status through commissionaire arrangements and similar strategies (discussed later).27 However, paragraph 5 of article 5 of the UN Model continues to be broader in scope than paragraph 5 of article 5 of the OECD Model through the inclusion of the stock agent provision. Some countries were of the view that the narrower formulation in the OECD Model could potentially lead to a dependent agent presenting themselves as acting on their own behalf.28

28.3 Commissionaire Arrangements The next best thing to not paying any taxes at all.29

The quotation above has been attributed to tax advisors promoting the use of commissionaire arrangements, one of the key issues arising in relation to agency permanent establishments. The problem of commissionaire arrangements in matters of international taxation has been extensively examined elsewhere,30 and only a brief discussion is included here. The issue stems from the different approaches to agency under civil law and common law. As noted previously, the different approaches to agencies and the potential for different outcomes under the agency permanent establishment rules was already apparent in the discussions at the League on the first model treaties but was left unresolved. One typical arrangement for the sale of goods by an enterprise into another jurisdiction (i.e. where it is not resident) is for the enterprise to set up a subsidiary company in the other jurisdiction. The subsidiary company acquires the goods from the parent company and sells the goods to third parties in the other jurisdiction. In this arrangement, the ownership and risk in relation to the goods is with the subsidiary company and the subsidiary is responsible for the marketing and sale of the goods. Under this arrangement, the subsidiary company is subject to tax in its country of residence on the profits arising from the sale of the goods. An alternative arrangement which potentially 26

UN Model (1980), 71. UN Model (2017), 186. 28 Ibid., 194. 29 M. Swanick, M. Mudrick, and E. Bouwman, ‘Tax and Practical Issues in Commissionaire Structures’, Tax Notes International (1997), 137. 30 L. Parada, ‘Agents vs Commissionaires: A Comparison in Light of the OECD Model Convention’, Tax Notes International 72 (2013), 59; A. Pleijsier, ‘The Agency Permanent Establishment—Practical Applications (Part Two)’, Intertax 29 (2001), 218. 27

504 Sunita Jogarajan reduces the overall tax burden is for the subsidiary company to sell the parent company’s goods under a commissionaire relationship. In this case, the parent company continues to own the goods and the subsidiary company enters into sale contracts with customers in its own name. Although the parent company owns the goods, it is not bound by the sale contracts. Under this arrangement, the profits on the sale of the goods are attributable to the parent company and not the subsidiary because the subsidiary is not considered to be acting as an ‘agent’ of the parent company. In civil law jurisdictions, a party which contracts as an agent must not only have the authority to do so (i.e. to enter into contracts which bind the principal), it must also disclose the principal to the third party during the contractual proceedings.31 Where the principal is not disclosed, the concluded contract does not bind the principal. In common law jurisdictions, the contract is treated as between the principal and the third party, through the agent, whether or not the principal is disclosed to the third party. The fact that commissionaire arrangements are not treated as an agency permanent establishment means that the profits on the sale of the goods are not taxed in the jurisdiction where the sales occur. The jurisdiction where the sales take place (i.e. the subsidiary company’s country of residence) can only tax the remuneration received by the subsidiary company for its services (typically a commission).32 In addition to being used to avoid the creation of a permanent establishment, commissionaire arrangements have sometimes also been used for other tax advantages, such as to avoid domestic-controlled foreign company rules.33 Two cases in particular have confirmed that commissionaire arrangements do not fall within the scope of the agency permanent establishment provision of the OECD Model.34 In Zimmer,35 the parent company, Zimmer Ltd, was a UK-resident company which sold orthopaedic products. Until 1995, the orthopaedic products were sold in France through

31

Avery Jones and Lüdicke ‘The Origins of Article 5(5) and 5(6) of the OECD Model’, 204–206. commission paid to the commissionaire is typically low as compared to the sales in the jurisdiction, but it is difficult for tax authorities to challenge the commission as it involves the balancing of several factors and the availability of comparable criteria: P. Luquet and E. Llinares, ‘The Remuneration of the Commissionaire’, ITR 16 (2005), 60, 60–62. In the Dell case discussed later, a commission of less than 1% of the company’s turnover was not challenged. 33 M. McClintock and S. Ward, ‘International Tax and Business Planning Opportunities through Commissionaire Arrangements’, International Tax Journal 22/ 3 (1996), 55, 57; J. H. Momsen, ‘Commissionaire Arrangements: An Australian Perspective of the Preferred United States Method of Foreign Distribution and Sale of Goods’, Taxation in Australia 5/4 (1997), 185; Swanick, Mudrick and Bouwman, ‘Tax and Practical Issues in Commissionaire Structures’, 137–138; M. Picat and E. Resler, ‘Between Commercial Agency and Distribution—The Commissionaire Structure’, IBL 30 (2002), 211; D. Bass, D. Hryck, and R. Rothman, ‘Reducing Global Tax Rates Using Commissionaires and Limited Risk Distributors International Tax Strategies’, Taxes 84/4 (2006), 13. 34 Case summaries adapted from B. Arnold, ‘Article 5: Permanent Establishment’, in Global Tax Treaty Commentaries (Amsterdam: IBFD, 2020), para. 3.5.2. 35 P.- J. Douvier and X. Lordkipanidze, ‘Zimmer Case: The Issue of the Deemed Existence of a Permanent Establishment Based on Status as a Commissionaire’, ITPJ 17/4 (2010), 266. On the impact of the Zimmer case and the treatment of commissionaire arrangements in Belgium, Canada, Denmark, Italy, Germany, Netherlands, Norway, Portugal, Spain, Sweden, the UK, and the USA, see International Transfer Pricing Journal 17/5–6 (2010) and also 18/1 (2011). 32 The

Agents in International Tax Treaties 505 a wholly owned French subsidiary company, Zimmer SAS. The subsidiary was responsible for the distribution and marketing of the products in France. In 1995, Zimmer SAS sold its products to Zimmer Ltd but continued to market and distribute the products in France under a commissionaire arrangement. Under the arrangement, Zimmer SAS could ‘accept orders, make offers, negotiate prices and terms of payment, grant discounts, and conclude contracts with both new and existing clients without the prior approval of its UK parent, Zimmer Limited’.36 In essence, the subsidiary’s roles and responsibilities in relation to its customers remained the same under the commissionaire arrangement as they did prior to that, but the risks in relation to the goods were transferred from the French subsidiary to the UK parent. The French Conseil d’État (Supreme Administrative Court) held that, under the commissionaire arrangement, the French subsidiary did not constitute an agency permanent establishment of the UK parent. This was because, under French law, a commissionaire acts in its own name and does not bind its principal. As such, the profits from the sale of the orthopaedic goods were not taxable in France, only the fees or commission received by the French subsidiary for its services. In Dell Products v. Government of Norway,37 the products—computers—were manufactured in Ireland by Dell Products (Europe), an Irish-resident company, which sold them to its wholly owned subsidiary, Dell Products (Ireland), also an Irish-resident company. Dell Products (Ireland) was responsible for the sale of the computers in various countries through national sales companies. In Norway, the sales were undertaken by Dell Norway under a commissionaire arrangement with Dell Products (Ireland).38 Dell Norway was a wholly owned subsidiary of the ultimate US parent company of the Dell group. It had no employees during the relevant years and used the services of other Dell companies. It was remunerated by commission and its taxable income was less than 1% of its turnover. The terms and conditions were that of a typical commissionaire arrangement and the Høyesterett (Supreme Court of Norway) held that Dell Products (Ireland) did not have an agency permanent establishment in Norway.

28.4 The BEPS Response In February 2013, the OECD released its report on addressing BEPS.39 The OECD 2013 Report recognized that base erosion posed a significant risk to the tax revenues, tax 36

Arnold, ‘Article 5’, para. 5.2.2.1. Leegaard, ‘Supreme Court Holds That Commissionaire Structure Does Not Amount to a Permanent Establishment’, European Taxation 52 (2012), 317; N. Pearson-Woodd and H.-M. Jørgensen, ‘Norway’s Supreme Court Rules That a Commissionaire Company is Not a PE of the Principal Company’, ITR 23/1 (2012), 67. 38 The same commissionaire structure was used in fifteen other European countries, apparently without issue. 39 OECD, Addressing Base Erosion and Profit Shifting (Paris: OECD Publishing, 2013) http://www. oecd.org/tax/addressing-base-erosion-and-profit-shifting-9789264192744-en.htm (accessed 1 April 2021) (OECD 2013 Report). 37 T.

506 Sunita Jogarajan sovereignty, and tax fairness of all countries, not just members of the OECD.40 The report also recognized that there were several ways in which domestic tax bases could be eroded but suggested that profit shifting was perhaps the most concerning cause of base erosion. Although the data on corporate income tax revenues and effective corporate income tax rates was not conclusive, the research into foreign direct investment flows provided some evidence that profit shifting was a problem that needed to be addressed.41 For example, in 2010, Barbados, Bermuda, and the British Virgin Islands (combined GDP of approximately 12 billion in 2010 in current US$) received 5.11% of global foreign direct investment—more than Germany (4.77%) (GDP of 3.4 trillion in 2010 in current US$) and Japan (3.76%) (GDP of 5.7 trillion in 2010 in current US$).42 The research also indicated significant investments through the use of special-purpose entities.43 The OECD 2013 Report identified a need for increased transparency regarding the effective tax rates of multinational entities as well as six key areas for action—international mismatches in entity and instrument characterization, the application of tax treaty concepts to profits earned through the delivery of digital goods and services, the tax treatment of related party debt-financing, captive insurance and other inter-group financial transactions, transfer pricing, the effectiveness of anti-avoidance measures, and the availability of harmful preferential regimes.44 The OECD noted that a lack of collaboration on these issues would likely result in unilateral measures to address BEPS and potentially undermine the existing collaborative approach to international taxation issues. The OECD 2013 Report was endorsed by the G20 at its ministerial-level meeting held in Moscow in February 2013.45 The OECD 2013 Report was followed by the release of an Action Plan by the OECD on 19 July 2013.46 The OECD BEPS Action Plan was immediately endorsed by the G20 at the minister level in July 2013 and also at its Russian Summit in St Petersburg in September 2013.47

40

Ibid., 5. Ibid., 15–21. 42 Ibid., 17. GDP data from World Bank, ‘World Bank Open Data’, https://data.worldbank.org (accessed 1 April 2021). 43 Special Purpose Entities have ‘no or few employees, little or no physical presence in the host economy, whose assets and liabilities represent investments in or from other countries, and whose core business consists of group financing or holding activities’: OECD 2013 Report, 18. 44 OECD 2013 Report, 47–48. 45 G20, ‘Communiqué of Meeting of G20 Finance Ministers and Central Bank Governors’ (16 Feb. 2013), para. 20, http://www.g20.utoronto.ca/2013/2013-0216-finance.html (accessed 1 April 2021). The G20 had previously declared its support for the OECD’s work on BEPS: G20, ‘G20 Leaders Declaration’, Declaration (18– 19 June 2012), para. 48, https://www.consilium.europa.eu/uedocs/cms_Data/docs/ pressdata/en/ec/131069.pdf (accessed 1 April 2021). 46 OECD, ‘Action Plan on Base Erosion and Profit Shifting’ (19 July 2013), https://www.oecd.org/ctp/ BEPSActionPlan.pdf (OECD BEPS Action Plan). 47 G20, ‘Communiqué of Meeting of G20 Finance Ministers and Central Bank Governors’ (19–20 July 2013), paras 18– 20, http://www.g20.utoronto.ca/2013/2013_Final_Communique_FM_July_ENG.pdf (accessed 1 April 2021). G20, ‘G20 Leaders’ Declaration’ (6 Sept. 2013), paras 50–52, http://www.oecd.org/ g20/summits/saint-petersburg/Saint-Petersburg-Declaration.pdf (accessed 1 April 2021). 41

Agents in International Tax Treaties 507 The OECD BEPS Action Plan contained fifteen action items to address BEPS. Of relevance to this chapter, Action Item 7 of the OECD Action Plan called for changes to the definition of permanent establishment to prevent the artificial avoidance of permanent establishment status in relation to BEPS, including through the use of commissionaire arrangements and specific activity exemptions.48 In developing this Action Item, the OECD recognized that the interpretation of existing treaty rules in many countries permitted the negotiation and conclusion of contracts for the sale of goods belonging to a foreign enterprise by a local subsidiary of that foreign enterprise without the profits on the sale being taxed in the sale country to the same extent that they would be if the sales were made by a distributor. As such, traditional distributor arrangements were being replaced with ‘commissionaire arrangements’ resulting in profits being shifted out of the country where the sales were taking place but with no substantive change in operations. Also of relevance to this chapter is Action Item 15, which proposed the development of a multilateral instrument to enable countries to implement measures developed to target BEPS in a timely manner in existing bilateral treaties.49

28.4.1 Addressing Commissionaire Arrangements In relation to BEPS Action 7, the publication of the OECD BEPS Action Plan was followed by the publication of a Discussion Draft in October 2014.50 The Discussion Draft acknowledged the interpretation issues with the existing text (discussed earlier) and proposed that the policy objective should be that an enterprise is considered to have a permanent establishment in a country where it uses an intermediary to regularly conclude contracts unless the intermediary does so through an independent business. The Discussion Draft set out four alternative proposals for amending the text of paragraphs 5 and 6 of article 5 of the OECD Model to achieve this objective. Option B, set out below, formed the basis for the eventual amendments to article 5.51 5. Notwithstanding the provisions of paragraphs 1 and 2 but subject to the provisions of paragraph 6, where a person is acting in a Contracting State on behalf of an enterprise and, in doing so, habitually concludes contracts, or negotiates the material elements of contracts, that are a) in the name of the enterprise, or b) for the transfer of the ownership of, or for the granting of the right to use, property owned by that enterprise or that the enterprise has the right to use, or c) for the provision of services

48

OECD BEPS Action Plan, 19. Ibid., 24. 50 OECD, ‘BEPS Action 7: Preventing the Artificial Avoidance of PE Status’, Discussion Draft (2014), https://www.oecd.org/tax/treaties/action-7-pe-status-public-discussion-draft.pdf (accessed 1 April 2021) (OECD BEPS Action 7). 51 Ibid., 13. 49

508 Sunita Jogarajan by that enterprise, that enterprise shall be deemed to have a permanent establishment in that State in respect of any activities which that person undertakes for the enterprise, unless the activities of such person are limited to those mentioned in paragraph 4 which, if exercised through a fixed place of business, would not make this fixed place of business a permanent establishment under the provisions of that paragraph. 6. Paragraph 5 shall not apply where the person acting in a Contracting State on behalf of an enterprise of the other Contracting State carries on business in the first- mentioned State as an independent agent acting on behalf of various persons and acts for the enterprise in the ordinary course of that business. Where, however, a person acts exclusively or almost exclusively on behalf of one enterprise or associated enterprises, that person shall not be considered to be an independent agent within the meaning of this paragraph with respect to these enterprises.

The publication of the Discussion Draft was followed by the publication of a Revised Discussion Draft in 2015.52 The Revised Discussion Draft noted that more than 850 pages of comments were received in relation to the Discussion Draft.53 A public consultation in relation to the proposals in the Discussion Draft was also held on 21 January 2015. With regard to the four options for reform to address commissionaire arrangements, the Revised Discussion Draft noted that Option B received by far the most support, but not without some criticism and suggestions for amendments. The concerns were mainly in relation to the concept of ‘negotiates material elements of contracts’ which was considered to be quite uncertain and the wording ‘the supply of goods owned by the enterprise or for the provision of services by the enterprise’ which was considered to be ‘unclear and overreaching’.54 Strong objections were also raised in relation to the scope of the independent agent exception. The first two objections were not addressed in the Revised Discussion Draft but the concept of ‘associated enterprises’ in the independent agent exception was narrowed through the introduction of paragraph 6(b) as follows: For the purpose of this Article, a person shall be connected to an enterprise if one possesses at least 50 per cent of the beneficial interests in the other (or, in the case of a company, at least 50 per cent of the aggregate vote and value of the company’s shares or of the beneficial equity interest in the company) or if another person possesses at least 50 per cent of the beneficial interest (or, in the case of a company, at least 50 per cent of the aggregate voting power and value of the company’s shares or of the beneficial equity interest in the company) in the person and the enterprise. In any case, a

52

OECD, ‘BEPS Action 7: Preventing the Artificial Avoidance of PE Status’, Revised Discussion Draft (2015), http://www.oecd.org/tax/treaties/revised-discussion-draft-beps-action-7-pe-status.pdf (OECD Revised Discussion Draft). 53 Ibid., 10. The OECD published approximately 100 submissions on 15 January 2015: OECD, ‘Comments Received on Public Discussion Draft, BEPS Action 7: Prevent the Artificial avoidance of the PE Status’ (2015), https://www.oecd.org/ctp/treaties/public-comments-action-7-prevent-artificial- avoidance-pe-status.pdf (accessed 1 April 2021). 54 OECD Revised Discussion Draft, 11.

Agents in International Tax Treaties 509 person shall be considered to be connected to an enterprise if, based on all the relevant facts and circumstances, one has control of the other or both are under the control of the same persons or enterprises.

The revised proposal was accompanied by proposed changes to the OECD Commentary. The publication of the Revised Discussion Draft was followed by the publication of the Action 7 Final Report in October 2015.55 The Final Report set out the amendments to the text of article 5 as put forward in the Revised Discussion Draft and further amended to address the concerns regarding ‘negotiates material elements of contracts’ and further clarify the operation of the independent agent exception. The Commentary accompanying the new text notes that the activities described in the new version of paragraph 5 of article 5 go beyond mere promotion or advertising and result in the conclusion of contracts and, as such, constitute a permanent establishment.56 The final proposed amendments to article 5 is set out below (amendments to the existing OECD Model text are highlighted with additions identified in italic and deletions in strikethrough):57 5. Notwithstanding the provisions of paragraphs 1 and 2 but subject to the provisions of paragraph 6, where a person other than an agent of an independent status to whom paragraph 6 applies is acting in a Contracting State on behalf of an enterprise and has, and habitually exercises, in a Contracting State, an authority to conclude contracts, in doing so, habitually concludes contracts, or habitually plays the principal role leading to the conclusion of contracts that are routinely concluded without material modification by the enterprise, and these contracts are a) in the name of the enterprise, or b) for the transfer of the ownership of, or for the granting of the right to use, property owned by that enterprise or that the enterprise has the right to use, or c) for the provision of services by that enterprise, that enterprise shall be deemed to have a permanent establishment in that State in respect of any activities which that person undertakes for the enterprise, unless the activities of such person are limited to those mentioned in paragraph 4 which, if exercised through a fixed place of business, would not make this fixed place of business a permanent establishment under the provisions of that paragraph. 6. An enterprise shall not be deemed to have a permanent establishment in a Contracting State merely because it carries on business in that State through a broker, general commission agent or any other agent of an independent status, provided that such persons are acting in the ordinary course of their business.

55 OECD, ‘Preventing the Artificial Avoidance of Permanent Establishment Status, Action 7— 2015 Final Report’ (5 Oct. 2015), https://www.oecd.org/tax/preventing-the-artificial-avoidance-of- permanent-establishment-status-action-7-2015-final-report-9789264241220-en.htm (accessed 1 April 2021) (OECD Action 7 Final Report). 56 Ibid., 19. 57 Ibid., 16.

510 Sunita Jogarajan

a) Paragraph 5 shall not apply where the person acting in a Contracting State on behalf of an enterprise of the other Contracting State carries on business in the first-mentioned State as an independent agent and acts for the enterprise in the ordinary course of that business. Where, however, a person acts exclusively or almost exclusively on behalf of one or more enterprises to which it is closely related, that person shall not be considered to be an independent agent within the meaning of this paragraph with respect to any such enterprise. b) For the purposes of this Article, a person is closely related to an enterprise if, based on all the relevant facts and circumstances, one has control of the other or both are under the control of the same persons or enterprises. In any case, a person shall be considered to be closely related to an enterprise if one possesses directly or indirectly more than 50 per cent of the beneficial interest in the other (or, in the case of a company, more than 50 per cent of the aggregate vote and value of the company’s shares or of the beneficial equity interest in the company) or if another person possesses directly or indirectly more than 50 per cent of the aggregate vote and value of the company’s shares or of the beneficial equity interest in the company) in the person and the enterprise.

The text of paragraphs 5 and 6 of article 5 as set out in the OECD Action 7 Final Report are incorporated in the latest version of the OECD Model and the Multilateral Convention to Implement Tax Treaty Related Measures to Prevent BEPS (discussed further later).58

28.4.2 Attribution of Profits to Permanent Establishments The OECD Action 7 Final Report recognized that in addition to introducing new rules to address commissionaire arrangements, it was also necessary to develop additional guidance on how profits would be attributed to permanent establishments arising out of the new rules.59 However, it was also recognized that the work on attribution of profits could only be undertaken after the work on Action 7 and Actions 8–10 (on transfer pricing). A detailed discussion of the OECD’s work on attribution of profits in relation to commissionaire arrangements is beyond the scope of this chapter and only a brief summary is included here. The OECD released a discussion draft on the

58 OECD,

OECD Model Tax Convention on Income and on Capital: Condensed Version (Paris: OECD Publishing, 2017), 32; Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting (opened for signature 24 November 2016, entered into force 1 July 2018), https://www.oecd.org/tax/treaties/multilateral-convention-to-implement-tax-treaty-related-measures- to-prevent-BEPS.pdf (accessed 1 April 2021), 19. 59 OECD Action 7 Final Report, 45. The difficulty in attributing profits in relation to commissionaire arrangements is not a new issue: A. Pleijsier, ‘The Agency Permanent Establishment—Allocation of Profits (Part Three)’, Intertax 29 (2001), 275.

Agents in International Tax Treaties 511 attribution of profits to permanent establishments in July 2016.60 The 2016 Discussion Draft raised questions for comment in relation to the attribution of profits to permanent establishments outside the financial sector. Of relevance to this chapter, the 2016 Discussion Draft provided guidance on specific fact-patterns related to dependent agent permanent establishments and sought input on the application of different approaches, such as the authorized OECD approach.61 Following comments on the 2016 Discussion Draft and country positions, the relevant committee (Working Party No. 6) developed a new discussion draft to replace the 2016 Discussion Draft, which was released in June 2017.62 The 2017 Discussion Draft provided further examples but did not provide the actual calculations for determining profits, as had been done in the 2016 Discussion Draft. The OECD published a Final Report containing the additional guidance on attribution of profits to permanent establishments which was released in March 2018.63 The 2018 Final Report builds on the examples in the 2017 Discussion Draft to illustrate the application of the underlying principles of profit attribution to commissionaire structures but does not provide details on the actual calculations of those profits. The fundamental principle is that the profits attributable to a permanent establishment would be the profits that would have been derived by the permanent establishment if it were a separate and independent enterprise engaged in the same or similar activities under the same or similar conditions. The starting point in each of the examples is the authorized OECD approach but it is recognized that other approaches are possible which would lead to different outcomes. The report provides some useful guidance but is arguably of limited utility given its high-level approach which leaves a number of issues unresolved.64

28.4.3 Implementing Change As mentioned previously, Action Item 15 of the BEPS Action Plan proposed the development of a multilateral instrument (MLI) to enable countries to implement

60 OECD,

‘BEPS Action 7 Additional Guidance on the Attribution of Profits to Permanent Establishments’ (2016), https://www.oecd.org/tax/transfer-pricing/BEPS-discussion-draft-on-the-attr ibution-of-profi ts-to-permanent-establishments.pdf (accessed 1 April 2021) (2016 Discussion Draft). 61 On the authorized OECD approach, see OECD, ‘2010 Report on the Attribution of Profits to Permanent Establishments’ (CTPA 2010), 13, https://www.oecd.org/ctp/transfer-pricing/45689524.pdf. 62 OECD, ‘BEPS Action 7: Additional Guidance on Attribution of Profits to Permanent Establishments’ (2017), https://www.oecd.org/tax/transfer-pricing/beps-discussion-draft-additional- guidance-attribution-of-profi ts-to-permanent-establishments.pdf (2017 Discussion Draft). 63 OECD, ‘Additional Guidance on the Attribution of Profits to a Permanent Establishment’ (2018), https://www.oecd.org/tax/beps/additional-guidance-attribution-of-profi ts-to-a-permanent-establishm ent-under-beps-action7.htm (2018 Final Report). 64 R. Collier and J. Vella, ‘Five Core Problems in the Attribution of Profits to Permanent Establishments’, WTJ 11 (2019), 159.

512 Sunita Jogarajan measures developed to target BEPS in a timely manner in existing bilateral treaties. The text of the MLI was released on 24 November 2016.65 The MLI is a one-off ‘moment in time’ instrument designed to update the existing bilateral treaty network with BEPS minimum standards. There are currently ninety-five signatories to the MLI which entered into force on 1 July 2018. The MLI contains some mandatory articles—the minimum standards (Action 5 on countering harmful tax practices, Action 6 on countering treaty abuse, Action 13 on transfer pricing documentation and country-by-country reporting, and Action 14 on improving dispute-resolution mechanisms)— but most are optional. The proposed reforms on commissionaire arrangements are contained in article 12 of the MLI and is not one of the minimum standards (also of relevance is art. 15 in relation to closely related entities). It is questionable whether the work on Action 7 and resultant reforms will prove successful in practice as only forty-five of the ninety-five current signatories to the MLI have agreed to adopt article 12. While some key jurisdictions such as France, the Netherlands, and Spain have agreed to adopt article 12, others such as Germany, Italy, Sweden, Switzerland, and the UK have not.66 Further, it should be noted that the USA is not currently a signatory to the MLI. In 2017, the UN Committee of Experts on International Cooperation in Tax Matters also decided to modify paragraphs 5 and 7 of article 5 of the UN Model in line with the recommendations in the Action 7 Final Report.67 As noted earlier, the scope of the UN Model in capturing agency permanent establishments continues to be broader than the OECD Model through the inclusion of the stock agent provision.

65 OECD, ‘Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting’ (2016), http://www.oecd.org/tax/beps/multilateral-convention-to-implem ent-tax-treaty-related-measures-to-prevent-BEPS.pdf (accessed 1 April 2021). 66 The countries which have agreed to adopt art. 12 are: Albania, Argentina, Armenia, Bosnia and Herzegovina, Burkina Faso, Cameroon, Chile, Colombia, Costa Rica, Ivory Coast, Croatia, Egypt, Fiji, France, Gabon, India, Indonesia, Israel, Jamaica, Japan, Jordan, Kazakhstan, Kenya, Lithuania, Malaysia, Mexico, Netherlands, New Zealand, Nigeria, North Macedonia, Norway, Papua New Guinea, Peru, Romania, Russia, Saudi Arabia, Senegal, Serbia, Slovak Republic, Slovenia, Spain, Tunisia, Turkey, Ukraine, and Uruguay. The countries which have not adopted art. 12 by making a reservation pursuant to art. 12(4) are: Andorra, Australia, Austria, Bahrain, Barbados, Belgium, Belize, Bulgaria, Canada, China, Curacao, Cyprus, Czech Republic, Denmark, Estonia, Finland, Georgia, Germany, Greece, Guernsey, Hong Kong (SAR), Hungary, Iceland, Ireland, Isle of Man, Italy, Jersey, Korea, Kuwait, Latvia, Liechtenstein, Luxembourg, Malta, Mauritius, Monaco, Morocco, Oman, Pakistan, Panama, Poland, Portugal, Qatar, San Marino, Seychelles, Singapore, South Africa, Sweden, Switzerland, United Arab Emirates, and UK. 67 Economic and Social Council, Committee of Experts on International Cooperation in Tax Matters, ‘Report on the Fourteenth Session’, E/2017/45-E/C.18/2017/3 (2017); Committee of Experts on International Cooperation in Tax Matters, ‘Proposed Base Erosion and Profit-Shifting Related Changes to the United Nations Model Double Taxation Convention between Developed and Developing Countries’, E/C.18/2017/CRP.7 (2017).

Agents in International Tax Treaties 513

28.4.4 Ongoing Work to Address Tax Challenges Arising from the Digitalization of the Economy In 2016, the OECD/G20 Inclusive Framework on BEPS was established to enable interested countries and jurisdictions to participate in the development of standards on BEPS-related issues as well as reviewing and monitoring the implementation of the OECD/G20 BEPS Project.68 The Inclusive Framework is currently working on developing consensus-based, long-term solutions to the international tax challenges arising from the digitalization of the economy. The Inclusive Framework has published two reports in this regard. The Report on Pillar One Blueprint seeks to establish a new tax framework on nexus and profit allocation in relation to business profits.69 The goal of the new tax framework is to expand the taxing rights of market jurisdictions (which in some cases may be the jurisdictions where users are located) where ‘there is an active and sustained participation of a business in the economy of that jurisdiction through activities in, or remotely directed at, that jurisdiction’.70 This is coupled with the introduction of new dispute-prevention and resolution mechanisms to improve tax certainty. Pillar One involves the determination of two components—a new taxing right for market jurisdictions over a share of residual profit calculated at an MNE group (Amount A) and a fixed return for certain baseline marketing and distribution activities taking place physically in a market jurisdiction, in line with the arm’s-length principle (Amount B). Of relevance to this chapter, the application of the new framework to commissionaire arrangements (in particular, whether they may be captured within the scope of Amount B in some cases) is still the subject of disagreement and additional work.71 The Report on Pillar Two Blueprint focuses on a global minimum tax intended to address remaining BEPS issues.72 Of relevance to this chapter, the report notes that further technical work is necessary to determine whether a targeted rule to address commissionaire arrangements is necessary so that the appropriate outcomes are achieved without the imposition of undue compliance or administrative burdens.73 To date, 137 countries and jurisdictions have agreed to the two-pillar solution and work is ongoing in this regard.74 68 OECD, ‘About the Inclusive Framework on BEPS’, https://www.oecd.org/tax/beps/beps-about.htm/ (accessed 1 April 2021). Almost 140 countries have joined the Inclusive Framework. 69 OECD/G20 Base Erosion and Profit Shifting Project, ‘Tax Challenges Arising from Digitalisation— Report on Pillar One Blueprint’ (2020), https://doi.org/10.1787/beba0634-en (accessed 1 April 2021). 70 Ibid. 11. 71 Ibid. 168–69. 72 OECD/G20 Base Erosion and Profit Shifting Project, ‘Tax Challenges Arising from Digitalisation— Report on the Pillar Two Blueprint’ (OECD 2020) https://doi.org/10.1787/abb4c3d1-en (accessed 1 April 2021) . 73 Ibid., 132. 74 OECD/ G20 Base Erosion and Profit Shifting Project, ‘Statement on a Two-Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy’ (2021), https://www.oecd. org/tax/beps/statement-on-a-two-pillar-solution-to-address-the-tax-challenges-arising-from-the-dig italisation-of-the-economy-october-2021.pdf (accessed 15 August 2022); OECD, ‘Progress Report on

514 Sunita Jogarajan

28.5 Conclusion The problems with commissionaire arrangements have long been recognized and the amendments to the agency permanent establishment threshold incorporated in the MLI and the 2017 OECD Model are to be welcomed. However, it remains to be seen whether these amendments result in a practical difference. First, as noted earlier, only forty-five countries have thus far agreed to the agency permanent establishment provision in the MLI. In order for a provision in a bilateral treaty to be amended by the MLI, both parties to the treaty must have signed the MLI and have agreed to the relevant provision. Given the limited acceptance of the agency provision, the previous agency permanent establishment provision, which did not capture commissionaire arrangements, is likely to continue to apply in many cases. Further, if the new agency provision is only included in some bilateral tax treaties, the possibility of treaty shopping exists despite efforts to curb such practice.75 Secondly, it is possible that multinational enterprises may develop alternative distribution strategies which are not captured by the new agency provision. For example, low-risk distributor arrangements are one possible alternative to commissionaire arrangements and are unlikely to create a permanent establishment for the enterprise in the jurisdiction where the sales occur.76 Thirdly, the determination of a dependent agent still involves significant interpretational issues which are likely to be the subject of disagreement between taxpayers and tax authorities as well as between different tax authorities.77 Fourthly, even if commissionaire arrangements are captured within the agency provision and constitute a permanent establishment, there is still the problem of attribution of profits which is likely to remain unresolved despite the OECD’s ongoing work in this regard.78 For these reasons, it is considered that the problem of commissionaire arrangements specifically, or practices that avoid permanent establishments more generally, are likely to persist. It would be beneficial for future research to examine the actual impact of the measures to address commissionaire arrangements on taxpayer behaviour.

Amount A of Pillar One, Two-Pillar Solution to the Tax Challenges of the Digitalisation of the Economy, OECD/G20 Base Erosion and Profit Shifting Project’ (2022), https://www.oecd.org/tax/beps/progress- report-on-amount-aof-pillar-one-july-2022.pdf. 75 V. Arel- Bundock, ‘The Unintended Consequences of Bilateralism: Treaty Shopping and International Tax Policy’, Intern’l Org 71 (2017), 349. 76 OECD Revised Discussion Draft, 5. 77 J. Bellemare, ‘Evolution of the Permanent Establishment Concept’, Canadian Tax Journal 65 (2017), 725, 740– 741; C. Garbarino, Taxation of Bilateral Investments: Tax Treaties after BEPS (Cheltenham: Edward Elgar, 2019), paras 1.48–1.71. 78 J. Eisenbeiss, ‘BEPS Action 7: Evaluation of the Agency Permanent Establishment’, Intertax 44 (2016), 481, 497–501; M. Levey, I. Gerdes, and A. Mansfield, ‘The Key BEPS Action Items Causing Discussion in the United States’, Intertax 44 (2016), 399, 402; M. Floris de Wilde, ‘Lowering the Permanent Establishment Threshold via the Anti-BEPS Convention: Much Ado About Nothing?’, Intertax 45 (2017), 556, 563–566.

Pa rt I V

L E G A L A SP E C T S OF I N T E R NAT IONA L T R A N SF E R P R IC I N G

Chapter 29

The Role of A rt i c l e 9 OECD MC Miguel Teixeira de Abreu

29.1 Introduction Article 9(1) MC1 reads as follows: 1. Where (a) an enterprise of a Contracting State participates directly or indirectly in the management, control or capital of an enterprise of the other Contracting State, or b) the same persons participate directly or indirectly in the management, control or capital of an enterprise of a Contracting State and an enterprise of the other Contracting State, and in either case conditions are made or imposed between the two enterprises in their commercial or financial relations which differ from those which would be made between independent enterprises, then any profits which would, but for those conditions, have accrued to one of the enterprises, but, by reason of those conditions, have not so accrued, may be included in the profits of that enterprise and taxed accordingly.

Article 9(1), in its current text, was introduced by the 1963 version of the Model Convention (MC). It grants a contracting state the right to include in the profits of a resident enterprise profits which were registered not to its benefit, but to the benefit of a non-resident associated enterprise. The idea of adjusting profits declared by an enterprise through an adjustment to the price of transactions can be dated back to 1917, when the US War Excess Profits Tax permitted that the Internal Revenue Service be given authority to ‘compute the amount

1

OECD Model Tax Convention on Income and on Capital (1963 version, unchanged by subsequent revisions). When referring to MC or to a specific article, reference is to the 2017 version.

518 Miguel Teixeira de Abreu of the tax properly due from each corporation on the basis of an equitable and lawful accounting’.2 Later, a provision was inserted into the 1921 US Revenue Act stating that, in the case of transactions between related businesses, ‘the Commissioner may consolidate [their] accounts . . . for the purpose of making an accurate distribution . . . of gains, profits, income, deductions, or capital’.3 This provision was slightly amended in the 1928 US Revenue Act, which established that, in such cases, ‘the Commissioner is authorized to . . . allocate gross income or deductions between . . . such . . . businesses, if he determines that such . . . allocation is necessary in order to prevent evasion of taxes or clearly to reflect the income of any of such . . . businesses’.4 The possibility for the tax authorities to reallocate income was adopted by the League of Nations: following a report presented in 1927, four drafts for a Convention on the Prevention of Double Taxation were proposed in 1928, the draft on Direct Taxes stating that ‘[S]‌hould the undertaking possess permanent establishments in both Contracting States, each of the two States shall tax the portion of the income produced in its territory’.5 Although the 1928 drafts already called for some apportionment criteria, it was not until the 1933 Draft Convention on the Allocation of Business Income between States for the purposes of Taxation that the application of the arm’s-length principle (ALP) to transactions carried out between two different enterprises was expressly recognized. Article 5 stated that: [W]‌hen an enterprise of one contracting State has a dominant participation in the management or capital of an enterprise of another contracting State, or when both enterprises are owned or controlled by the same interests, and as the result of such situation there exists, in their commercial or financial relations, conditions different from those which would have been made between independent enterprises, any item of profit or loss which should normally have appeared in the accounts of one enterprise, but which has been, in this manner, diverted to the other enterprise, shall be entered in the accounts of such former enterprise.

Article 9, as it reads today, can thus set its origins on article 5 of the 1933 draft.6 However, when comparing both provisions, some relevant differences can be detected.

2

Art. 77, reg. 41. Cf. M. Lehner, ‘Article 9 –Associated Companies’, in T. Ecker and G. Ressler, eds, History of Tax Treaties: The Relevance of the OECD Documents for the Interpretation of Tax Treaties“ (1971), 387. http://www.steuerrecht.jku.at/ml/de/elemente_site/Kommentare-Sammelwerke/Leh ner%20-%20Art%209%20OECD%20MC.pdf. 3 S. 240(d). 4 S. 45. 5 Art. 5. 6 Cf. also, London and Mexico Model Tax Conventions (1946).

The Role of Article 9 OECD MC 519 First, while the 1933 draft required a dominant participation in the management or capital of another enterprise, article 9 applies whenever an enterprise directly or indirectly participates in the management, control, or capital of another enterprise. Secondly, the term used in the 1933 draft that profits or losses had been ‘diverted’ from one enterprise to another, was replaced by the simple reference to profits having not accrued to the accounts of an enterprise. To divert something requires an act of intent, while the term ‘accrual’ is more factual and notes an objective result, that is, one that does not value intent. Thirdly, the reallocation of an item of ‘profit’ or ‘loss’ in the 1933 draft was limited to items of profit. While the 1933 draft used the term ‘shall’, the current draft uses the term ‘may’. The word may reflects an option by the contracting state as the article only deals with positive corrections. The Commentary on the 1963 version of the MC was very laconic as it merely stated that article 9 called ‘for very little comment’. Paragraph 16 of the general remarks to the MC, however, stated that ‘more detailed rules and procedures . . . will be considered by the Fiscal Committee in its future work’. This said, the commentary to article 7 already set a route for the determination of profits which could be used in rectifying the accounts of an enterprise under article 9, particularly by developing the method of separate accounting (paras 10–12) or the method of fractional apportionment (paras 22–25).7 The 1977 draft of the MC maintained the 1963 draft but incorporated a second paragraph to deal with corresponding adjustments.8 Subsequent revisions to the MC9 did not materially change the Commentary on article 9(1), save in respect of two issues. While clarifying that the application of domestic thin capitalization rules is not prevented by article 9, paragraph 3 of the respective Commentary states that the ‘application of rules designed to deal with thin capitalization should normally not have the effect of increasing the taxable profits of the relevant domestic enterprise to more than the arm’s length profit, and that this principle should be followed in applying existing tax treaties’.10 Also, paragraph 74 of the Commentary on article 24 requires that domestic 7

The 1933 draft already called for the possibility of applying different methods; preferably, the method of separate accounting or, alternatively, the turnover method or, if none would work, the fractional apportionment method. 8 Although absent from the 1963 MC, corresponding adjustments had already been recommended in 1946 under art. IV of the London and Mexico Models (‘a rectification made in the accounts of an establishment situated in one country should not, however, result in double taxation of the enterprise concerned. A rectification made in one country may therefore call for a corresponding adjustment in the accounts of the establishment in the other country’). 9 Revisions were made in 1992, 1994, 1995, 1997, 2000, 2002, 2005, 2008, 2010, 2014, and 2017. Future updates of the MC are expected to contemplate OECD, ‘Proposed Changes to Commentaries in the OECD Model Tax Convention on Article 9 and Related Article’ (2021). Hopefully, such updates will help clarify the role of art. 9, especially when confronting its provisions with domestic (non-transfer pricing provisions) aimed at adjusting the profits of resident enterprises. 10 Cf. J. Wittendorff, ‘The Transactional Ghost of Article 9(1) of the OECD Model’, Bulletin for International Taxation (March 2009), 119.

520 Miguel Teixeira de Abreu thin capitalization rules may be applied ‘insofar as these are compatible with paragraph 1 of Article 9’. Although acknowledging that some countries consider that article 9 does not prevent them from adjusting profits ‘under conditions that differ from those of the Article’, paragraph 4 of its Commentary also states that where the application of domestic law results in adjustments ‘at variance with the principles of the Article’, contracting states may ‘deal with such situations by means of corresponding adjustments . . . and under mutual agreement procedures’. Despite the fact that the current Commentary on article 9 did not change much from the 1963 draft, the problems raised by this provision have always been substantial and, in 1979, the OECD introduced its Transfer Pricing and Multinational Enterprises Report (the 1979 Report). The 1979 Report was repealed in 1995 when the OECD introduced its Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations (the TP Guidelines). These Guidelines were amended in 2010 and 2017.11 As article 9 incorporated the ALP approach that was first outlined in the 1920s, the OECD TP Guidelines focus on finding a common approach for tax administrations and ‘associated enterprises’ on how to apply such principle—thereby ‘avoiding unilateral responses to multilateral problems’.12

29.2 Interpretation of the Term ‘Associated Enterprises’ In order for article 9(1) to apply, two conditions must be met. First, two enterprises must be ‘associated’. Simply put, either two enterprises are ‘associated’ and a contracting state is entitled to adjust the profits accrued to its resident enterprise, or they are not and article 9 cannot apply.13 Secondly, if they are ‘associated’, their profits may be adjusted but only to the extent required to meet the ALP.14 11 When referring to the TP Guidelines, reference is to the 2017 version. References to previous versions will be preceded by the respective year. 12 TP Guidelines, para. 7. Also, TP Guidelines, para. 12: ‘an international consensus is required on how to establish for tax purposes transfer prices on cross-border transactions’. 13 Cf. Commentary on art. 9, para. 2: the basis for an adjustment must result from the fact that ‘the true taxable profits’ of the resident enterprise are not shown ‘as a result of the special relations between the enterprises’. 14 The application of the ALP can also be found in art. 7(2). Although art. 7 deals with juridical double taxation and art. 9 deals with economic double taxation, art. 7 operates under the fiction that the permanent establishment must be treated as a ‘separate and independent enterprise’. Therefore, the basis for adjustments under the ALP must be identical under art. 7 and art. 9, meaning that ‘the word may, as used in Art. 9(1) in connection with “any profits” must be taken to refer to “any profits and only these” ’: K. Vogel, Double Taxation Conventions -A Commentary to the OECD-, UN-and US-Model Conventions for the Avoidance of Double Taxation of Income and Capital (Kluwer, 1991), art 9, para. 17.

The Role of Article 9 OECD MC 521 Defining the term ‘associated enterprises’ is therefore of enormous relevance in determining the role of article 9. In order to carry out this exercise, we need to call upon the rules of interpretation of article 3, as complemented by the Commentaries on the MC and the TP Guidelines.15 Article 3(1) contains a list of defined terms, which apply regardless of their interpretation in the domestic laws of each contracting state. In addition to the terms listed under article 3(1), other terms are defined under other provisions of the MC (e.g. ‘resident’, ‘permanent establishment’, ‘associated enterprises’). The definition of the term ‘associated enterprises’ is given to us by article 9(1): it includes two enterprises where either: (1) one participates in the ‘management, control or capital’ of another; or (2) they are both under the ‘management, control or capital’ of other ‘persons’.16 Although the term ‘enterprise’ is defined under the MC as the ‘carrying on of any business’, paragraph 1 of the Commentary on article 9 makes it clear that it intends to limit its application to companies;17 that is, ‘parent and subsidiary companies and companies under common control’. While the term ‘parent and subsidiary companies’ is self-explanatory, the term ‘under common control’ raises more difficulties as it is not defined in the MC.18 For an undefined term, article 3(2) states that it shall ‘unless the context otherwise requires or the competent authorities agree to a different meaning pursuant to the provisions of Article 25, have the meaning that it has at that time under the law of that State’. So, the question is whether or not the term ‘under common control’ or, more particularly, the term ‘control’, should be interpreted exclusively by reference to domestic law, or if an international meaning deriving from the ‘context’ of the convention should be imposed on both contracting states. 15 Art. 31 of the Vienna Convention on the Law of Treaties (VCLT) states that the terms used in a treaty should be interpreted ‘in the context and in light of its object and purpose’. On the role of the commentaries on the MC and their relationship with arts 31 and 32 VCLT, cf. K. Vogel, ‘The Influence of the OECD Commentaries on Treaty Interpretation’, IBFD Bulletin –Tax Treaty Monitor (December 2000), 614. Cf. also M. J. Ellis, ‘The Influence of the OECD Commentaries on Treaty Interpretation – Response to Prof. Dr Klaus Vogel’, IBFD Bulletin –Tax Treaty Monitor (December 2000); M. Lang and F. Brugger, ‘The Role of the OECD Commentary in Tax Treaty Interpretation’ (2008). https://www.wu.ac. at/fileadmin/wu/d/i/taxlaw/institute/staff/publications/langbrugger_australiantaxforum_95ff.pdf. 16 Cf. art. 3(1)(a). 17 Cf. art. 3(1)(b). This rules out any association based on relationships of blood, marriage, or any other community of interests. However, two companies under the control of the same person will be associated for the purposes of art. 9. Cf. K. Vogel, Double Taxation Conventions -A Commentary to the OECD-, UN-and US-Model Conventions for the Avoidance of Double Taxation of Income and Capital, art. 9, paras 24–26. Also, M. Lehner, ‘Article 9 –Associated Companies’, in T. Ecker and G. Ressler, eds, 404. 18 Art. 5(8) links the concept of ‘closely related enterprises’ to the capacity of one controlling the other, and then connects control to one enterprise possessing 50% or more of the beneficial interest of the other. However, para. 119 of the Commentary on art. 5 is clear in stating that this concept ‘is to be distinguished from the concept of associated enterprises which is used for purposes of Article 9; although the two concepts overlap to certain extent, they are not intended to be equivalent’.

522 Miguel Teixeira de Abreu It might be argued that paragraph 4 of the Commentary on article 9 settles the question, as some countries consider the function of article 9 to be limited to raising ‘the arm’s length principle at treaty level’ and, thus, are not prevented from adjusting profits ‘under conditions that differ from those of the Article’. The USA had taken a similar position when it introduced in its MC a provision clarifying that the ability of a contracting state to adjust profits in accordance with its domestic provisions was not affected by the wording of article 9, ‘when necessary in order to prevent evasion of taxes or clearly to reflect the income of any such persons’.19 This provision was later removed on the argument that it was ‘confusing’ and that the ability of a contracting state to apply its domestic provisions was not dependent on a provision like the one inserted in 1981.20 It might also be argued that the MC is aimed at restricting the right of a contracting state to tax non-resident companies or giving a credit for taxes paid by its resident companies in the other contracting state, but not to restrict the right of a contracting state to tax its own residents. This argument was reinforced by the introduction of article 1(3),21 which stated that a double taxation convention (DTC) ‘shall not affect the taxation, by a Contracting State, of its residents except with respect to benefits granted under . . . paragraph 2 of Article 9’. All these arguments imply that article 9 does not prevent a contracting state from applying its own concept of ‘control’, no matter how wide such concept may be. However, no argument can be made against the fact that article 3(2) imposes that a domestic interpretation must concede if a different interpretation is imposed by the ‘context’ of the convention. Paragraph 12 of the Commentary on article 3 states that ‘the context is determined by the intention of the Contracting States when signing the Convention as well as the meaning given to the term in question in the legislation of the other Contracting State (an implicit reference to the principle of reciprocity on which the Convention is based)’. In our attempt to understand if the ‘context’ of the MC imposes that article 9 be interpreted as containing a set of rules that are binding on both contracting states or if it merely allows any definition of ‘control’ adopted by their respective domestic laws, let us start with a simple exercise, and assume that article 9, together with article 1(3), gives a contracting state the right to adjust profits of a resident enterprise in accordance with its own definition of ‘control’. Let us imagine that state A adjusts the profits of a resident enterprise on the argument that 20% of its capital is owned by an enterprise resident of the other contracting state (B) and its domestic law considers that such a holding is sufficient to grant it ‘control’, regardless of the fact that the remaining 80% is in the hands of one independent enterprise.

19

Cf. art. 9(3) US MC (1981). Cf. Commentary on art. 9(1) US MC (1996). 21 Introduced in the (2017) MC as a result of BEPS, Action 6 (cf. art. 11 of the MLI). 20

The Role of Article 9 OECD MC 523 Taken as the starting point of our exercise, such adjustment would have to be accepted as having been made ‘in accordance with the Convention’.22 Article 9(2), left outside the savings provision of article 1(3), would then operate to impose on state B the obligation to grant a corresponding adjustment. However, state B might respond by reverting to paragraph 6 of the Commentary on article 9, which states that ‘State B is . . . committed to make an adjustment of the profits of the affiliated company only if it considers that the adjustment made in State A is justified both in principle and as regards the amount’.23 Such a position by state B would trigger the application of article 25 and the two contracting states would then enter into a mutual agreement procedure (MAP) aimed at avoiding economic double taxation. In such MAP, the contracting states would first have to agree on the difference between a ‘controlled transaction’ and an ‘uncontrolled transaction’, as the TP Guidelines would not work properly if one were not to be distinguished from the other. That would force the two contracting states to agree on a common definition of ‘associated enterprises’, as only upon such a definition would they be able to agree that transactions between such enterprises were, in fact, ‘controlled transactions’. As state B would consider that the two enterprises were not ‘associated’, it would consider that any transactions carried out between them would be ‘uncontrolled transactions’ and would, by definition, be carried out at arm’s length. Failing to reach a common understanding under the MAP, arbitration would ensue. The arbitration award would then decide on the interpretation of the term ‘control’ for the purposes of article 9(1), and such interpretation would ‘be binding on both Contracting States’.24 22 Cf. G. Kofler and I. Verlinden, ‘Unlimited Adjustments: Some Reflections on Transfer Pricing, General Anti-Avoidance and Controlled Foreign Company Rules, and the “Saving Clause” ’ Bulletin for International Taxation (April/May 2020), 277. https://koflerge.files.wordpress.com/2020/09/unlimited- adjustments.pdf: ‘[I]‌f . . . article 9(1) . . . had no restrictive force, how could a primary adjustment beyond arm’s length ever “not [be] in accordance” with the tax treaty?’ 23 This statement is, however, complemented by the TP Guidelines’ statement at para. 4.35: while acknowledging ‘the non-mandatory nature of corresponding adjustments’ (a mechanism necessary to ensure that ‘a tax administration is not forced to accept the consequences of an arbitrary or capricious adjustment by another State’), it then clarifies that it will be so only ‘in the absence of an arbitration decision arrived at pursuant to an arbitration procedure comparable to that provided for under paragraph 5 of Article 25 which provides for a corresponding adjustment’. 24 Art. 25(5): The exception being that if a court of either contracting state has rendered a decision on such disputed issue, then there will be no submission to arbitration. Arbitration was introduced in the MC in 2010 with a footnote that it ‘should only be included in the Convention where each State concludes that it would be appropriate to do so’. The footnote was deleted in the 2017 version (Commentary on art. 25, para. 65): ‘in recognition of the importance of including an arbitration mechanism that ensures the resolution of disputes between the competent authorities where these disputes would otherwise prevent the mutual agreement procedure from playing its role’. Arbitration proceedings to settle tax disputes (in particular, those involving transfer pricing) have also been introduced in the EU, originally under the Convention for the Elimination of Double Taxation in Connection with the Adjustment of Profits of Associated Enterprises, Convention 90/436/EEC, and more recently through the introduction of Council Directive (EU) 2017/1852.

524 Miguel Teixeira de Abreu This exercise shows how the ‘context’ operates to devoid of any practical effect an interpretation of the term ‘control’ that would fail to be accepted by both contracting states. Unfortunately, more often than not, one needs to discover such ‘context’ amidst conflicting remarks of the commentaries to the MC and of the TP Guidelines. One point seems certain though: the ‘context’ requires the elimination of double taxation and the arbitration provision of paragraph 5 of article 25 secures that effect.25 However, in order to get to that point, the arbitrators have to sort out an interpretation of the term ‘control’ that respects the text of article 9. In such a task, arbitrators would probably seek to find an international meaning for the term ‘control’, one that could go beyond the domestic interpretation of both contracting states and one that could generate a ‘broad basis of common understanding’.26 And what interpretation might that be? Looking at the MC and the TP Guidelines, it seems clear that article 9 is targeted at multinational enterprises (MNEs). The 1979 Report had recollected the following notion in a footnote:27 ‘[MNEs] usually comprise companies or other entities whose ownership is private, State or mixed, established in different countries and so linked that one or more of them may be able to exercise a significant influence over the activities of others’. This raises a question: where the participation in management or capital exists without control, are the two enterprises still associated for the purposes of article 9? Imagine a company that has the right to appoint one member of a board of eleven in another company, or takes a participation of 5% in the capital of another company. Such participation would not grant control but would, in literal terms, provide for a participation in the management or in the capital of the other company. By focusing on MNEs, the TP Guidelines imply that the relevant connection in article 9(1) is ‘control’, and any participation in the capital or management is valid only in as much as it provides ‘control’.28 Naturally, there are other forms of exercising ‘control’, for example through voting rights, subordination agreements, management agreements, or even through joint ventures. Any such form of ‘control’ should be admitted as long as based on law or contract. This brings us to another question: where an enterprise exercises a dominant position in the market, and imposes non-market prices on other unrelated enterprises,

25 Against this view, it may be argued that prior to the 2010 version of the MC, arbitration was not required and, if the contracting states failed to reach an agreement as to the exact definition of a term like ‘control’, double taxation would not be avoided. However, even in the case of DTCs without an arbitration provision, the context of the convention would, in our view, disallow a domestic interpretation of the term ‘control’ which would defraud the ‘object and purpose’ of the convention (which is the elimination of double taxation). 26 Cf. 1979 Report, para. 7: ‘[I]‌t was not thought necessary for the purpose of the report to define such expressions as “associated enterprises” or “under common control”. A broad basis of common understanding of what is meant is assumed.’ 27 Para. 7, 11. 28 Cf. M. Lehner, ‘Article 9 –Associated Companies’, in T. Ecker and G. Ressler, eds, 404.

The Role of Article 9 OECD MC 525 would that constitute ‘control’ for the purposes of article 9(1)? We think not. The reference under the Commentary on article 9(1) implies that only legally or contractually established situations are relevant (‘parent/subsidiary companies or companies under common control’).29 Also, considering that article 9 includes situations of de facto control would create enormous uncertainty, as such situations exist in many relationships between independent enterprises.30 We agree that, for the sake of clarity, an international meaning should be given to this term.31 Such international meaning should start from the fact that, under the ‘context’ of the convention, there can be no ‘control’ unless one enterprise exercises a ‘significant influence’ over the decisions of another. Although determining what amounts to a ‘significant influence’ calls for very subjective interpretations, one must conclude that, in the context of the MC, that should always require that such participation gives one enterprise the legal or contractual right to force another enterprise to accept what it would not accept were it an independent enterprise. It seems, then, evident that for an enterprise to be in a position to make or impose such conditions, such participation must be relevant. Although relevance cannot be determined by reference to the concept of ‘closely related enterprises’ of article 5(8), it must be based on an enterprise lacking the legal or contractual capacity to oppose to a transfer price that is detrimental to its interests.32 Where one enterprise is in no position to exert such influence in the management, capital, or control of another enterprise, then it is not ‘associated’ and its transactions are ‘uncontrolled transactions’. Such being the case, ‘no re-writing of the accounts . . . is authorised’.33 29 Cf.

R. Dwarkasing, ‘Comments on the Revised Discussion Draft of Transfer Pricing Aspects of Intangibles’ (2013). https://www.oecd.org/ctp/transfer-pricing/dwarkasing-maastricht-university.pdf. Differently under art. 9(1) US MC, ‘Technical Explanations’ (2006). Also cf. K. Vogel, H. A. Shannon, III and R. L. Doernberg: United States Income Tax Treaties (Kluwer, 1994), 428. 30 e.g. franchise agreements; monopolistic, duopolistic, or oligopolistic markets. 31 Cf. P. Baker, Double Taxation Agreements and International Tax Law (Sweet & Maxwell, 1991), 155: it seems equally arguable that an international meaning should be given to ‘associated enterprises’ and ‘control’: any domestic legislation which departs from these international meanings would then be incompatible with Art. 9(1) . . . This issue will depend very much on whether ‘associated enterprises’ and ‘under common control’ are undefined terms to be defined under the domestic law of each state (under Art. 3(2)), or whether the context prevents such domestic definition and an international meaning should be attributed. 32 In

our opinion, where capital or voting rights are spread over, or management is appointed by, different unrelated shareholders, it would be extremely difficult for an enterprise to impose transfer prices that are not arm’s length, as such other shareholders (or managers) would certainly react against such prices. Control should therefore be connected to a lack of checks and balances within a group of enterprises. Cf. W. Schön, ‘Transfer Pricing –Business Incentives, International Taxation and Corporate Law’. Max Planck Institute for Tax Law and Public Finance Working Paper (2011), https://www.tax.mpg. de/RePEc/mpi/wpaper/Tax-MPG-RPS-2011-05.pdf, 11. 33 Commentary on art. 9, para. 2. If the 2021 OECD proposed changes to the Commentaries on art. 9 are introduced, this statement should be interpreted as reading ‘no re-writing of the accounts . . . is authorized’ under the provision of art. 9.

526 Miguel Teixeira de Abreu There is another argument worth mentioning which is addressed at the right of a contracting state to tax its own residents. In a transfer pricing situation, as a result of the corresponding adjustment in article 9(2) and of the arbitration proceedings in article 25(5), the adjustment of the price of a transaction is equivalent to a transfer, for tax purposes, of the profits of the enterprise resident in state B to the enterprise resident in state A.34 Because of this effect (the increase in the tax revenue of state A is equivalent to the decrease in revenue in state B), the right of state A to adjust transfer prices under article 9 must be restricted by the wording of article 9. If the domestic law of state A were to prevail, then state B’s domestic law would have to give in to the definition of ‘control’ adopted by state A. That would restrict its right to tax its own residents. This reasoning is also important when looking at the scope of article 1(3). The wording of article 1(3) impacts on the very core of article 9, as if requires one to take a position on whether article 9 is permissive (i.e. a mere idealistic statement without any impact on the scope of domestic transfer pricing provisions) or restrictive (i.e. a provision aiming at defining the boundaries within which domestic transfer pricing provisions may operate in a Decenary).35 If article 1(3) imposes a permissive interpretation of article 9(1), then the effect is that article 9(1) will be devoid of any role. As Vogel explained: ‘If these opinions were correct Art. 9 MC and treaty rules corresponding to it would be superfluous They would neither be a legal basis for profit adjustments nor would they have any impact on domestic law.’36 In order to understand how these two provisions may interact, we need to call on how article 1(3) came to be part of the MC. Article 1(3) was introduced by the Base Erosion and Profit Shifting (BEPS) Project as a means to ‘prevent misinterpretations intended to circumvent the application of a Contracting State’s domestic anti-abuse rules (as illustrated by the example of controlled foreign companies rules)’.37 The Multilateral Instrument (MLI) then welcomed this proposed provision under article 11, dealing with treaty abuse. However, the purpose of article 9(1) is not to deal with treaty abuse, but rather to adequately price intra-group transactions, so that profits are taxed in the state where they 34

Reverting to our exercise earlier. G. Kofler and I. Verlinden, ‘Unlimited Adjustments: Some Reflections on Transfer Pricing, General Anti- Avoidance and Controlled Foreign Company Rules, and the “Saving Clause” ’, 279; K. Vogel, H. A. Shannon, III and R. L. Doernberg: United States Income Tax Treaties, 420; BEPS Action Plan 6, paras 61ff. 36 DTCs, art. 9, para. 16: ‘[T]‌he profit adjustment provisions make sense only if they are taken to have the same significance as DTC rules generally have, viz. that they restrict domestic law. The only way that business profits can be assured to be taxed only in the State where they originate economically . . . is by guaranteeing that both contracting States are bound by firm adjustment criteria under the conditions set out in Art. 9 MC’. Cf, also K. Vogel, H. A. Shannon, III and R. L. Doernberg: United States Income Tax Treaties, 420: ‘[I]t would be inconsistent to interpret Article 7 to bind the contracting states with respect to the taxation of a permanent establishment and its head office while interpreting Article 9 not to bind the contracting states with respect to the taxation of associated enterprises’. Cf. also M. Lehner, M. Lehner, ‘Article 9 –Associated Companies’, in T. Ecker and G. Ressler, eds, 130. 37 Action Plan 6, 61. 35 Cf.

The Role of Article 9 OECD MC 527 originate, much as is intended under article 7(2) in respect of an enterprise carrying on business in the other contracting state through a permanent establishment. By carrying out a primary adjustment beyond the limits of article 9(1), a contracting state would in fact be enlarging its taxing rights at the expense of the taxing rights granted to the other contracting state. So, article 1(3) must be interpreted as allowing a contracting state to set aside article 9(1) when it considers that a domestic anti-avoidance provision should be applied instead.38 In such situations, if article 9(1) is not the applicable provision, then no corresponding adjustment is due by the other contracting state. But, where article 9(1) is applied, and until such time as an international meaning of the term ‘control’ is defined in the MC, or one is developed by jurisprudence, one point seems clear to us: that however distinct the domestic definition of this term may be between two contracting states, it results from the ‘context’ of the convention, and from the ‘principle of reciprocity’ to which paragraph 12 of the Commentary on article 3 MC so rightly alludes, that it must have a meaning acceptable to both contracting states.

29.3 Domestic Law and Treaty Law Looking at the role of article 9 is, thus, not only a matter of interpretation, but also a matter of how domestic law and treaty law interact in general. The question is one of determining if a DTC has a positive function of granting a contracting state a tax claim that is not otherwise granted under domestic law, or if it has a mere negative function of restricting a tax claim that is also granted under domestic law.39 The VCLT has two provisions that are relevant in answering these questions: article 26 (pacta sunt servanda) establishes that a treaty ‘is binding upon the parties to it and must be performed by them in good faith’ and article 27 (international law and observance of treaties) establishes that ‘a Party may not invoke the provisions of its internal law as justification for its failure to perform a treaty’. Thus, international treaty rules determine that a domestic provision which is in conflict with a treaty provision must be restricted in its scope to the limits established by the latter.40 38 e.g. if domestic transfer pricing provisions do not allow recharacterization but a GAAR/SAAR may be applied. 39 A. Xavier, Direito Tributário Internacional (Almedina, Coimbra, 2007), 111. 40 Cf. A. Knechtle, Basic Problems in International Fiscal Law (HFL Publishers, 1979), 65: ‘the conflict rules have only a negative effect in the sense that they leave the fiscal jurisdiction of the State unchanged, or restrict it, but they cannot extend it’. Also, K. Vogel, Double Taxation Conventions -A Commentary to the OECD-, UN-and US-Model Conventions for the Avoidance of Double Taxation of Income and Capital, Introduction, para. 46: ‘a tax treaty neither generates a tax claim that does not otherwise exist under domestic law nor expands the scope . . . of an existing claim’.

528 Miguel Teixeira de Abreu Similarly, principles of non-discrimination and equality also impose that a treaty provision cannot create a tax claim unless such claim is supported by a domestic provision.41 We cannot but agree with these positions. In fact, it seems clear to us that article 9 has a double negative effect; that is, on the one hand, it applies only if there is a domestic provision that allows transfer pricing adjustments and, on the other hand, it restricts the application of a domestic transfer pricing provision to those situations covered by article 9.42 The double negative effect of a DTC is also relevant when dealing with issues of recharacterization.

29.4 Article 9(1) and Recharacterization In order to determine if recharacterization is within the scope of article 9, one must first understand that when seeking to price a transaction in accordance with the ALP, a tax administration may take one of three possible approaches. The tax administration: (1) may consider that the contractual terms of a transaction do not conform to the actual conduct of the parties; or (2) may accept that the contractual terms of a transaction conform with the actual conduct of the parties but it considers that an independent party would have carried out such transaction under a different contractual set-up; or (3) may accept that the contractual terms of a transaction conform with the actual conduct of the parties but considers that an independent party would never enter into such a transaction.43 The TP Guidelines are clear in respect of the first approach, paragraph 1.4644 stating that a tax administration should carefully examine ‘whether the arrangements reflected 41 Cf.

K. Vogel, Double Taxation Conventions -A Commentary to the OECD-, UN-and US-Model Conventions for the Avoidance of Double Taxation of Income and Capital, art. 9, para. 12: ‘the legal basis for an adjustment . . . is exclusively domestic law—Art. 24 (5) prohibits discrimination . . . of an enterprise of one contracting State which is controlled by an enterprise of the other contracting State compared with enterprises controlled by a resident of the first mentioned State’. 42 Cf. K. Vogel, Double Taxation Conventions -A Commentary to the OECD-, UN-and US-Model Conventions for the Avoidance of Double Taxation of Income and Capital, art. 9, para. 12: ‘adjustments of profits of enterprises controlled by an enterprise of the other contracting State are subject to a dual restriction, viz. first by the arm’s length criterion under Art. 9 and secondly, within those limits, by possibly more restrictive adjustment rules applied to domestically controlled enterprises’. 43 Jens Wittendorff mentions three forms of transactional adjustments: (1) a transaction is totally disregarded; (2) the existence of the transaction is recognized but its form is disregarded and recharacterized for tax purposes; or (3) the existence and form of the transaction are recognized but price- sensitive terms are recharacterized. Cf. also A. Bullen, Arm’s Length Transaction Structures: Recognising and Restructuring Controlled Transactions in Transfer Pricing (IBFD Doctoral Series, 2011), (Summary), 8ff. Cf. https://ec.europa.eu/taxation_customs/sites/taxation/files/docs/body/ andreas_bullen_summary.pdf. 44 Reaffirmed in para. 1.120

The Role of Article 9 OECD MC 529 in the actual conduct of the parties substantially conform to the terms of any written contract, or whether the associated enterprises’ actual conduct indicates that the contractual terms have not been followed’. Examining the actual conduct of the parties (in terms of the assets they use, the risks they assume, and the functions they exercise) enables the tax administration to ‘determine the factual substance and accurately delineate the actual transaction’. The actual conduct of the parties would then be subject to the arm’s-length test. In these cases, however, the transaction the parties executed is different from that which they have written. However, when dealing with the second and third approaches, the TP Guidelines are not as clear. The 1979 Report, as well as previous TP Guidelines, have been quite adamant that recharacterization is not generally acceptable under article 9(1).45 The (2017) Guidelines reaffirmed this general rule in paragraph 1.121: ‘[A]‌tax administration should not disregard the actual transaction or substitute other transactions for it unless the exceptional circumstances described in the following paragraphs 1.122–1.225 apply’.46 However, what is generally unacceptable seems to be exceptionally tolerable. The 1979 Report and all TP Guidelines allowed exceptions to this general rule. However, while the 1979 Report was quite restrictive,47 the TP Guidelines were far more comprehensive.48 The (2017) TP Guidelines, influenced by BEPS Reports on Actions 8—10, allow recharacterization when a transaction between associated enterprises, viewed in its totality, differs from that ‘which would have been adopted by independent enterprises behaving in a commercially rational manner in comparable circumstances . . . taking into account . . . the options realistically available to each of them’.49 Examples provided under paragraph 1.125 include: (1) the full disregard of a transaction where the transaction is shown to be ‘commercially irrational’ and independent entities would simply not enter into such a transaction, or into a transaction with similar effects; or (2) the replacement of the transaction carried out between associated enterprises by a different transaction where the executed transaction is ‘commercially irrational’ but a recharacterized transaction is possible and should be priced. These examples, reflecting an exception to the general non-recharacterization principle, should be interpreted in a restrictive manner and the tax administration should not lose sight of the fact that BEPS Action 8–10 Reports, while aiming at aligning transfer 45

1979 Report, para. 23; 1995 TP Guidelines, para. 1.36; 2010 TP Guidelines, para. 1.64. This principle is reaffirmed in para. 1.122: ‘[B]‌ecause non-recognition can be contentious and a source of double taxation, every effort should be made to determine the actual nature of the transaction and apply arm’s length pricing to the accurately delineated transaction, and to ensure that non- recognition is not used simply because determining an arm’s length price is difficult’; and para. 1.123: ‘[R] estructuring of legitimate business transactions would be a wholly arbitrary exercise the inequity of which could be compounded by double taxation created where the other tax administration does not share the same views as to how the transaction should be structured’. 47 1979 Report, para. 23: ‘[T]‌ he report does, however, recognize that it may be important in considering, for example, what is ostensibly interest on a loan to decide whether it is an interest payment or, in reality, a dividend or other distribution or profit.’ 48 Cf. 1995 and 2010 TP Guidelines (respectively para. 1.37 and para. 1.65). 49 Para. 1.122. 46

530 Miguel Teixeira de Abreu pricing outcomes with value creation within a MNE group, had as its purpose the allocation of profits according to the functions exercised in each contracting state, not on the basis of what the enterprises eventually establish, but on their actual behaviour.50 Factoring all this into account, we take the view that in transactions where evidence is shown that the actual conduct of the parties differs from the contractual terms agreed between them, the role of article 9 is to tax the actual transaction undertaken by the parties (first approach earlier). This would include the approach to business reorganizations, as they involve transactions, some of which are underlying the reorganization itself, but transactions nonetheless. In such cases, the role of article 9 is about identifying actual transactions (e.g. transfer of intangibles, agreements reached on the termination of existing agreements and the execution of alternative agreements/arrangements, pricing the value of clientele)51 and pricing them in accordance with the ALP. Disregarding the contractual terms of a transaction (second and third approaches discussed earlier) is not the role of article 9 but rather the role of a general anti-avoidance rule (GAAR)/specific anti-avoidance rule (SAAR). The reason is that recharacterization implies the production of evidence with respect to the economic reasons and tax-driven motives behind a transaction, as designed by the parties. The requirements needed to meet the burden of proof in such situations go beyond the mere application of the ALP and transfer pricing provisions are not equipped to deal with such burden of proof.52 In these cases, we would suggest that, but for the exceptional circumstances stated under the TP Guidelines, interpreted in a restrictive manner, a GAAR/SAAR should be applied.53 One must also noted that the restrictive effect of article 9, as well as the negative effect of a DTC, operate so as to: (1) disallow the application of a domestic transfer pricing provision which permitted a more extensive recharacterization of transactions than admitted under the exceptional circumstances mentioned in the TP Guidelines; and (2) disallow the application of the exceptional recharacterization situations admitted under the TP Guidelines when a domestic transfer pricing provision fails to allow recharacterization.

50 Cf.

TP Guidelines, para. 1.124: ‘[T]‌he structure that for transfer pricing purposes, replaces that actually adopted by the taxpayers should comport as closely as possible with the facts of the actual transaction undertaken whilst achieving a commercially rational expected result that would have enabled the parties to come to a price acceptable to both of them at the time the arrangement was entered into’. 51 Cf. ibid., paras 9.15ff. 52 Commentary on art. 9, para. 4, states that ‘almost all member countries consider that . . . even a reversal of the burden of proof, would not constitute discrimination within the meaning of Article 24’. While the reversal of the burden of proof might be acceptable when applying the ALP to a transaction, as designed by the parties, it would be unthinkable for a tax authority to recharacterize a transaction and then force the taxpayer to prove otherwise. 53 If, upon applying a GAAR/ SAAR, a transaction is recharacterized, that (recharacterized) transaction needs to be priced and the ALP would then apply thereto.

The Role of Article 9 OECD MC 531

29.5 Article 9(1) and Tax Avoidance The prevention of tax-avoidance is not the role of article 9. In fact, there are many reasons—none of which related to tax avoidance—that justify a provision like article 9, including: • securing, through the separate entity approach, an appropriate tax base in each contracting state; • putting associated enterprises and independent enterprises on an equal footing for tax and competition purposes, ensuring neutrality between them; • serving general principles of equality; • promoting international trade and investment by removing tax considerations from economic decisions.54 Also, from the perspective of associated enterprises, there are a considerable number of non-tax reasons that may trigger the establishment of non-market prices, such as: (1) restrictions on the flow of capitals and exchange controls; (2) customs valuations and duties; (3) government restrictions; (4) political instability; (5) exchange rate risks; or (6) pressure to keep profits at the level of the parent company in non-consolidation scenarios.55 Article 9 has, therefore, the ‘dual objective of securing the appropriate tax base in each jurisdiction and avoiding double taxation, thereby minimising conflict between tax administrations and promoting international trade and investment’56 and ‘[T]‌ax administrations should not automatically assume that associated enterprises have sought to manipulate their profits’.57 Evidently, transfer prices may be misused for tax avoidance purposes and the simple application of the ALP will help combat, if not prevent all, such misuses. However, that does not justify characterizing article 9 as an anti-avoidance provision. The TP Guidelines are sensitive to this point. The 1979 Report, while accepting that transfer prices might be adopted to minimize taxes, was clear in that ‘the need to adjust the actual price to an arm’s length price . . . arises irrespective of . . . any intention of the parties to minimise tax. Hence, the consideration of transfer pricing problems should not be confused with the consideration of problems of tax fraud or tax avoidance, even though transfer pricing policies may be used for such purposes.’58 In fact, the principle behind transfer pricing is that ‘in taxing profits of an enterprise which engages in transactions with associated enterprises . . . profits should be calculated 54 Cf. R. Dwarkasing, ‘Comments on the Revised Discussion Draft of Transfer Pricing Aspects of Intangibles’. 55 Cf. TP Guidelines, para. 1.4. 56 Cf. ibid., Preface, para. 7. 57 Ibid., para. 1.2. 58 Para. 3, reaffirmed by para. 1.2 of the (2017) TP Guidelines.

532 Miguel Teixeira de Abreu on the assumption that the prices charged in these transactions are arm’s length prices. This is the underlying assumption in Article 9(1) of the OECD MC’.59 Furthermore, in a transfer pricing exercise, ‘[T]‌ax administrations are encouraged to take into account the taxpayer’s commercial judgement about the application of the arm’s length principle in their examination practices and to undertake their analysis of transfer pricing from that perspective.’60 The fact is that, in transfer pricing, ‘even the best intentioned taxpayer may make an honest mistake’ and ‘even the best-intentioned tax examiner may draw the wrong conclusion from the facts’.61 Characterizing article 9 as a tax avoidance provision would bear a considerable impact in view of other features of a contracting state’s tax system, including those related to inspection proceedings, administrative claims and judicial processes, burden of proof, statutes of limitation, applicable sanctions or penalties, and other consequences related to the treatment and graduation of tax offences. It would also be at variance with the effect caused by the operation of corresponding adjustments under article 9(2). If the primary adjustment were to reflect a reaction to a tax avoidance scheme, there would be no reason why the other contracting state should respond to such scheme by granting the resident enterprise a corresponding adjustment.62 As a matter of fact, while article 9 proposes a reallocation of taxing rights between two contracting states (thus, the corresponding adjustment), anti-avoidance provisions do not share such effect.63 It would also contaminate the very concept of ‘transfer prices’, a concept that should remain neutral, related only to ‘a system of pricing the transfer of goods, services and intangibles between entities of one MNE’.64

59

1979 Report, para. 3. TP Guidelines, Preface, para. 16. 61 Ibid., para. 4.9. 62 Note art. 9(3) UN MC, whereby: ‘[T]‌he provisions of paragraph 2 shall not apply where judicial, administrative or other legal proceedings have resulted in a final ruling that by actions giving rise to an adjustment of profits under paragraph 1, one of the enterprises concerned is liable to penalty with respect to fraud, gross negligence or wilful default’. The lack of such a provision in art. 9 reinforces the conclusion that art. 9 should not be used as an anti-avoidance provision. The same conclusion could be reached by calling upon para. 70 of the 1984 Report of the OECD Committee on Fiscal Affairs, Transfer Pricing and Multinational Enterprises: Three Taxation Issues (Paris: OECD Publishing, 1984), which states that ‘Article 9(2) did not in fact impose any obligation to match a corresponding adjustment where the original adjustment was made to the deliberate manipulation of a transfer price by the enterprise for the purpose of gaining a tax advantage’. A contrario, such a primary adjustment would not be ‘in accordance with’ art. 9(1) and thus, it would release the other contracting state from its obligation to make a corresponding adjustment. This also reinforces the argument that art. 1(3) is aimed at allowing a contracting state to apply a domestic anti-avoidance provision instead of art. 9(1) but not to apply art. 9(1) as a replacement for such anti-avoidance provision. 63 Cf. G. Kofler and I. Verlinden, ‘Unlimited Adjustments: Some Reflections on Transfer Pricing, General Anti-Avoidance and Controlled Foreign Company Rules, and the “Saving Clause” ’, 64. 64 Cf. H. Hamaekers, ‘An Introduction to Transfer Pricing The American Versus the European Approach’ (1995). https://repositorio.cepal.org/bitstream/handle/11362/34228/S9500513_en.pdf?seque nce=1&isAllowed=y 3. 60

The Role of Article 9 OECD MC 533 One must also note that often a transfer pricing conflict is not one between a taxpayer and its administration, but one between the tax administrations of the two contracting states. It is often the case that the taxpayer is in agreement with its tax administration as to the transfer prices it applies on intra-group transactions and it is the other tax administration that is in disagreement. It is not uncommon, taken the subjective nature surrounding the application of transfer prices, that a tax administration, in its pursuit of additional tax revenue, challenges the prices practised between its resident enterprises and related enterprises in other jurisdictions. Accepting that article 9 is not an anti-avoidance provision leads us to two conclusions: The first is that under article 9, intent is irrelevant.65 While the different motivations behind a certain transfer price would be extremely relevant in determining intent in a tax avoidance case,66 they are irrelevant when it comes to article 9.67 Secondly, while anti-avoidance schemes should be met with more severe penalties (to deter such behaviours) and more stringent procedural rules (to protect the taxpayer), transfer pricing adjustments do not share such concerns, and should be aimed only at the repositioning of the arm’s-length price and the reparation of damages suffered by a contracting state as a result of an adequate adjustment (e.g. compensatory interest for late payment of any additional taxes or penalties for failure to comply with legal proceedings, such as disclosure of a transfer pricing file).

65

Cf. TP Guidelines, para. 1.2 Cf. Cahiers de Droit Fiscal International, ‘Anti-Avoidance Measures of General Nature and Scope – GAAR and Other Rules’. General Report by Paulo Rosenblatt and Manuel E. Tron, vol. 103A (2018), 23. 67 Actually, the term ‘divert’ was present in the 1933 draft and removed from the text of the 1963 Draft MC. 66

Chapter 30

OECD Transfe r Pri c i ng Gu idelines a nd Internationa l Tax L aw Yuri Matsubara and Clémence Garcia

30.1 Introduction In the international tax system, permanent establishments belonging to multinational groups are usually taxed as separate entities in the jurisdiction where they operate. For that purpose, assessing the flow of goods and services transferred in cross-border- related party transactions is necessary when allocating taxable profits. Transfer pricing is the area of tax practice dealing with intra-group transaction prices in multinational companies, addressing the assessment of related party transactions. In the last thirty years, many jurisdictions have adopted some specific rules in order to protect their tax base against profit shifting and other manipulations.1 This chapter concerns the OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administration,2 which is the main international standard in relation to transfer pricing.3 Since transfer pricing is a complex and volatile subject, the Guidelines have frequently been revised after the prototype was issued in 1979. Similarly, valuation methods have, in practice, grown dramatically both in complexity and diversity. These particularities explain the importance given to economic substance in the area of transfer pricing.

1

Collier and Andrus, Transfer Pricing and the Arm’s Length Principle After BEPS (2017), 85.

2 OECD, Transfer Pricing Guidelines for Multinational Enterprises and Tax Administration (2022).

3 Other notable legislation include the UN Transfer Pricing Guidelines, along with some domestic legislation, e.g. s. 482 of the US Internal Revenue Code. See Rocha, ‘General Report: The Future of Transfer Pricing’, Cahiers de Droit Fiscal International vol. 102B (2017).

536 Yuri Matsubara and Clémence Garcia As far as tax avoidance is concerned, manipulating intra-group transfer prices is an easy way to shift profits and avoid taxation in high-tax jurisdictions.4 Consequently, after the Base Erosion and Profit Shifting (BEPS) Project, the version of the OECD Guidelines issued in July 2017 increased the substance criteria after BEPS Actions 8– 10 on the arm’s-length principle (ALP). Another important change was induced by BEPS Action 13 on country-by-country reporting (CbCR), aimed at improving the exchange of information among jurisdictions. Further guidance regarding the transactional profit split method and hard-to-value intangibles were issued June 2018 to complete the Guidelines. Most recently, some supplementary provisions were issued on the treatment of financial transactions and the implications of the Covid-19 pandemic on transfer pricing issues.5 In January 2022, a consolidated version of the Transfer Pricing Guidelines was issued, including all the aforementioned guidance. Later in the chapter, dispute resolutions on international taxation (i.e. tax treaty arbitration) will be covered.

30.2 The Development of Transfer Pricing Before discussing modern issues concerning the OECD Guidelines, it is interesting to consider the historical development of transfer pricing legislation. In the field of international taxation, the birth of transfer pricing is usually associated with the Carroll Report in 1933.6 Its author, Mitchell B. Carroll, was commissioned by the Fiscal Committee of the League of Nations to investigate the tax treatment of cross-border transactions in thirty-five countries. As can be seen from the Carroll Report, the foundations of modern transfer pricing rules were already known and enumerated at that time.7 Depending on the jurisdiction, the tax base would be allocated based on the accounts of local establishments (separate accounting method), based on a comparison with similar businesses using publicly available information (empirical methods), or based on the apportionment of group income (fractional apportionment method).8 The ALP was also referred to in the report for some jurisdictions that had adopted the separate accounting method. The principle substantially provides that transaction prices between related parties should be assessed

4

See Avi-Yonah, ‘Rise and Fall of Arm’s Length Standard’, Virginia Tax Review 15/89 (1995), 95. Transfer Pricing Guidance on Financial Transactions: Inclusive Framework on BEPS (2020), Actions 4, 8–10; OECD, ‘Guidance on the transfer pricing implications of the Covid-19 pandemic’ (2020). 6 Carroll, Methods of Allocating Taxable Income, Taxation of Foreign and National Enterprises vol. 4 (1933). 7 Wittendorff, Transfer Pricing and the Arm’s Length Principle in International Tax Law (2010), 89–91. 8 Petruzzi, ‘The Arm’s Length Principle: Between Legal Fiction and Economic Reality’, in Lang et al., eds, Transfer Pricing in a Post-BEPS World (2016), 6. 5 OECD,

OECD Transfer Pricing Guidelines and International Tax Law 537 at ‘arm’s length’, which was originally a substitute for market prices. Assessment should approach ‘prices that we obtain at arm’s length from independent purchasers’.9 The conclusions of the Carroll Report favoured the separate entity approach and the ALP and that tradition has remained for almost a century, and was most recently reconfirmed in the post-BEPS developments of the OECD Guidelines. A second important aspect to highlight here is that transfer pricing was originally addressed as an issue of profit shifting between entities located in the same jurisdiction.10 For example, Carroll’s earlier investigations in the US context addressed the allocation of profits among branches of a business which could be taxed as separate entities.11 Further developments in the USA tend to confirm the nature of transfer pricing rules as domestic anti-abuse legislation. Their scope enlarged to cross-border transactions when their economic importance subsequently increased. In 1968, three transaction-based transfer pricing methods—namely, the comparable uncontrolled price (CUP), cost- plus, and retail price methods—were formally introduced in the US Internal Revenue Code.12 Coming back to the international rules, the first predecessor of the modern OECD Guidelines, entitled Report on Transfer Pricing and Multinational Enterprises, was issued in 1979.13 True to the original stance of the Carroll Report, the Guidelines adopted the separate entity approach with the ALP. The same measurement methods as the US transfer pricing legislation were also included in the OECD Report. In 1986, a second important issue, that remains controversial today, was first addressed in the US transfer pricing rules. In order to prevent abusive profit shifting using intangible investments (i.e. provisions known as a cost-sharing regime), the ‘commensurate with income’ standard was introduced.14 Initially, these rules raised criticism from internal constituencies, as well as from OECD members, because they departed 9 Carroll, ‘Observations on Report of the Committee of the National Tax Association on Uniformity and Reciprocity in State Tax Legislation, Especially with regard to the Allocation of Income of International Enterprises’, in Proceedings of the Annual Conference on Taxation under the Auspices of the National Tax Association, vol. 24 (1931), 343. 10 On the influence of branches as a unit of transfer pricing, see Vann, ‘Reflections on Business Profits and the Arm’s Length Principle’, in Arnold, Sasseville, and Zolt, eds, The Taxation of Business Profits Under Tax Treaties (2003), 133–169; on the first TP rules, see Petruzzi, ‘The Arm’s Length Principle’, 1–32. 11 Carroll began his interviews with public administrations as early as 1913, prior to the development of US domestic transfer pricing rules in 1921 and, indeed, twenty years before his famous report to the League of Nations. His early investigations, based on collections of business practices and interviews, also explained the lack of a clear definition of arm’s-length price. From its origins, an arm’s-length price was designed as an objective to pursue by means of several pricing methods, of which he acknowledged the limitations. See Carroll, ‘Observations on Report of the Committee of the National Tax Association’, 343. 12 See Avi-Yonah, ‘Rise and Fall of Arm’s Length Standard’, also Vann, ‘Reflections on Business Profits and the Arm’s Length Principle’, in Arnold, Sasseville, and Zolt, eds, The Taxation of Business Profits Under Tax Treaties (2003), 133–169. 13 OECD, Report on Transfer Pricing and Multinational Enterprises (1979). 14 Brauner, ‘Transfer Pricing Aspects of Intangibles: the Cost Contribution Arrangement Model’, in Lang et al., Transfer Pricing in a Post-BEPS World, 111.

538 Yuri Matsubara and Clémence Garcia from the traditional arm’s-length transactional methods. A series of reforms followed in the USA, trying either to soften or to annihilate the anti-abuse rule; however, it eventually prevailed. In 1995, the OECD Transfer Pricing Guidelines finally adopted similar provisions. However, the issue of intangibles is far from settled; it remains central in discussions of digital businesses and business restructuring. Looking at the global landscape, only a few jurisdictions enacted detailed provisions on transfer pricing prior to the 1990s. Such provisions spread rapidly after the 1995 OECD Guidelines as a response to the globalization of economic transactions and due to the added complexity of the rules in the USA. Since 1995, the OECD Guidelines, along with most domestic legislation on transfer pricing, have provided for transfer pricing methods based on the functional analysis of transactions; in other words, the economic substance of arrangements between the parties. In the 2000s, the most important aspect in relation to transfer pricing was the permanent establishment. The Permanent Establishment Report in 2010 meant that international profit allocation became the subject for identifying a permanent establishment before applying the transfer pricing rules to assess a company’s local profits. In 2010 again, additional guidance on business restructuring introduced new provisions for intangible assets, including the analysis of risks and expected returns when allocating income in cross-border transactions. As mentioned earlier, the 2017 update of the OECD Guidelines was based on BEPS Actions 8–10 concerning the ALP and Action 13 concerning CbCR. As a result, the formal criteria used in the pre-BEPS Guidelines -for example a taxpayer’s legal structure or some contractual elements of transactions -were weakened and partially replaced with substance criteria. Globally, the 2022 version of the OECD Guidelines even further emphasizes substance over form by means of new economic criteria, such as control over risk and the specific provisions on the functional analysis for hard-to-value intangibles.

30.3 The ALP The precise definition of the ALP has not yet reached consensus. In its traditional sense, arm’s length refers to the methods used to determine transfer prices between related parties using comparables.15 However, the modern definition of arm’s length encompasses some methods that are not strictly based on comparables, such as profit splits or transfer pricing for intangibles and intra-group financing. In a broader definition, arm’s length refers to any method used to determine transfer prices between related parties which results in prices similar to those of unrelated transactions.16 15 See, e.g. Avi-Yonah, ‘Rise and Fall of Arm’s Length Standard’, 94; Wittendorff, Transfer Pricing and the Arm’s Length Principle in International Tax Law, 6–7. 16 Avi-Yonah, ‘Rise and Fall of Arm’s Length Standard’, 94.

OECD Transfer Pricing Guidelines and International Tax Law 539 With regard to the OECD Guidelines, the ALP is derived from the legal provision in article 9.1 concerning associated enterprises in the OECD Model Convention:17 Conditions are made or imposed between the two enterprises in their commercial or financial relations which differ from those which would be made between independent enterprises, then any profits which would, but for those conditions, have accrued to one of the enterprises, but, by reason of those conditions, have not so accrued, may be included in the profits of that enterprise and taxed accordingly.

From this provision, the main rationale of the ALP is the equal tax treatment of entities belonging to a multinational group compared with that of independent taxpayers. In the OECD Guidelines, the ALP is not defined as an independent concept: it is delineated by the provisions on its implementation. It is also worth noting that article 9.1 is not a general provision to assess transfer pricing as an assessment issue. Instead, it addresses transfer pricing from the viewpoint of base adjustments, as an anti-abuse rule.18 Except for specific cases where controlled transactions are not comparable with uncontrolled transactions, transfer pricing is implemented using a comparability analysis. The objective is to find the most reliable comparable to determine the arm’s-length price for each transaction.19 A comparable uncontrolled transaction is defined as ‘a transaction between two independent parties that is comparable to the controlled transaction under examination’.20 It can be either an ‘internal comparable’ (i.e. transactions between the taxpayer and an independent party) or an ‘external comparable’ (i.e. transactions involving two independent enterprises unrelated to the taxpayer). The comparability analysis is a complex process, involving tedious information- gathering to conduct functional analysis and search for comparables. Functional analysis in step 3 in Table 30.1 is a qualitative assessment of operations that seek ‘to identify the economically significant activities and responsibilities undertaken, assets used or contributed, and risks assumed by the parties to the transactions’.21 Functional analysis relies on a qualitative assessment of each party’s role in the controlled transaction. The post-BEPS rules place particular emphasis on the repartition of risks between the parties, which include the expectation of risks and mitigation functions: ‘The assumption of a risk should be compensated with an appropriate anticipated return, and risk mitigation should be appropriately remunerated.’22 The 17 OECD, Model Convention on Income and on Capital (2017).

18 Although the anti-avoidance role of transfer pricing is consensual in the tax literature, some view transfer pricing rules more as an issue of profit shifting than anti-abuse. See Koomen, ‘Transfer Pricing in a BEPS Era: Rethinking the Arm’s Length Principle, Part II’, International Transfer Pricing Journal (July/Aug. 2015), 231; Navarro, ‘Transfer Pricing’, in Brauner, ed., Research Handbook on International Taxation (2020), 97. 19 2022 OECD Guidelines, 3.2. 20 Ibid., 3.24 21 Ibid., 1.51. 22 Ibid., 1–100.

540 Yuri Matsubara and Clémence Garcia Table 30.1 Typical comparability analysis process 1

Determination of years to be covered.

2

Broad-based analysis of the taxpayer’s circumstances.

3

Understanding the controlled transaction under examination, particularly based on a functional analysis, to choose the tested party, the most appropriate transfer pricing method, and significant comparability factors to consider.

4

Review of existing internal comparables, if any.

5

Determination of available sources of information on external comparables and their reliability.

6

Selection of the most appropriate transfer pricing method and the relevant financial indicator, if applicable.

7

Identification of potential comparables and the key characteristics of comparability.

8

Comparability adjustments, if applicable.

9

Interpretation and use of the data collected, determination of the arm’s-length remuneration.

Source: Adapted with permission from 2022 OECD Guidelines, 3.4.

assessment of risks associated with the functions performed is a key element regarding the transfer pricing of intangibles (see 30.5 below).23 In summary, implementing transfer pricing method results in high compliance costs for sometimes uncertain results. Hence, maintaining the separate entity approach for profit allocation or replacing it with formulary apportionment has been a long-standing debate since the time of the League of Nations.24 Most recently, Pillar One of the Global Anti-Base Erosion (GloBE) approach has provided a formulary apportionment of residual profit, contrasting with the post-BEPS transfer pricing rules introduced earlier. Why is ALP still preferred in OECD guidelines? The main reason for the justification by the OECD is the equal treatment of taxpayers: groups and independent companies should be on an equal footing for tax purposes.25 Additionally, comparables can be obtained at reasonable cost: an arm’s-length price may be found from comparable transactions between independent parties in many cases. Cases that cannot be resolved using comparable transactions can instead be addressed with a transactional profit split (discussed later).26 Lastly, international consensus on the ALP makes this approach more legitimate than formulary apportionment.27

23 This point is developed in Section 30.5 concerning intangibles. See also Gonnet, ‘Risks Redefined in Transfer pricing Post-BEPS’, in Lang et al., Transfer Pricing in a Post-BEPS World, 45. 24 See, e.g., Petruzzi, ‘The Arm’s Length Principle’, 4. 25 2022 OECD Guidelines, 1.8. 26 Ibid., 1.9. 27 Ibid., 1.15.

OECD Transfer Pricing Guidelines and International Tax Law 541 From the opposite view, the detractors of the ALP prefer formulary apportionment due to the uncertainty of correctly implementing transfer price methods, which often results in legal dispute.28 Additionally, formulary apportionment may result in a better allocation of profits between developing and developed countries.29 In the view of the OECD, formulary apportionment is not a valid alternative to the ALP due to, inter alia, its high compliance costs, lack of consideration for country-specific performance, and lack of consistency with existing tax treaties.30 Judging from the position adopted in Pillar One of the GloBE approach, the debate between supporters of the separate entity approach and formulary apportionment of profits is likely to continue.

30.4 Transfer Pricing Methods As described in Section 30.2, the methods used to assess the ALP have become increasingly sophisticated over time, mirroring, to some extent, the development of international transactions. As far as the OECD Guidelines are concerned, traditional transaction methods— the CUP, resale price, and cost- plus methods— have been supplemented with two transactional profit methods: transactional net margin and transactional profit split. Indeed, taxpayers are given the choice of the most appropriate method based on the strengths and weaknesses of each alternative, their relevance in terms of functional analysis, the availability of information, and the reliability of comparability adjustments where applicable.31

30.4.1 Comparable Uncontrolled Price (CUP) Method The CUP method compares the price charged in a controlled transaction to that of an uncontrolled transaction in comparable circumstances.32 In a case in which those prices differ significantly, the taxpayer should substitute the uncontrolled price with the control price. The CUP method is preferred in cases where uncontrolled transaction prices can be reliably obtained, for example in cases where an independent company sells the same products. With regard to the limitations of this method, ‘it may be difficult to find a transaction between independent enterprises that is similar enough to a controlled

28

See Avi-Yonah, ‘Rise and Fall of Arm’s Length Standard’, 151.

29 Ibid. 30

2022 OECD Guidelines, 1.29. Ibid., 2.2. 32 Ibid., 2.14. 31

542 Yuri Matsubara and Clémence Garcia transaction such that no differences have a material effect on price’.33 The results of comparability analysis are a key aspect of whether this method should be applied.

30.4.2 Resale Price (RP) Method The RP method provides an assessment of the exit value (resale market price) that can be obtained from an independent party in a sales transaction, after subtracting an appropriate gross margin called the resale price margin.34 The RP method is deemed to be appropriate for marketing operations or businesses in which the resale price margin can be reliably assessed. When the resale price margin used is based on the margin of an independent company in a comparable transaction, both companies must be assessed to ensure that they perform the same functions.35

30.4.3 Cost-Plus Method (CPM) This method consists of adjusting the production cost of goods or services exchanged in a controlled transaction with an appropriate mark-up, considering the functions performed by the controlled entity. The cost-plus mark-up of the supplier in a controlled transaction should ideally be established by reference to the mark-up that the same supplier earns in comparable uncontrolled transactions (internal comparables). In addition, the mark-up that would have been earned in comparable transactions by an independent enterprise can serve as a guide (external comparables).36 This method is interesting because even though it is primarily based on internal information, it is still considered reliable for services or intermediary products that do not have a market price at the early stages of the value chain. The resale and cost-plus methods examine the net profit derived from an appropriate base such as sales, assets, or costs.37 The most appropriate profitability indicator must be selected based on a functional analysis and other elements relevant in order to select a transfer pricing method.38

33

Ibid., 2.16. Ibid., 2.27. 35 Ibid., 2.33. 36 Ibid., 2.46. 37 Ibid., 2.64. 38 Ibid., 2.82 34

OECD Transfer Pricing Guidelines and International Tax Law 543

30.4.4 Transactional Net Margin (TNM) The TNM method relies on profit indicators that are not obtained from a comparison with a competitor for a particular transaction; therefore, the assessment of the tax base is based on the general performance of the business39 and not on a single comparable transaction. Hence, the TNM method may be preferred in cases when controlled transactions cannot be reliably compared with uncontrolled ones. Some critics point out the lack of comparability of profit indicators based on ‘blended profits’; in other words, the profits of diversified businesses.40

30.4.5 Transactional Profit Split (TPM) The TPM consists of identifying profits to be split for the associated enterprises and then allocating part of the combined profits to entities based on ‘an economically valid basis that approximates the division of profits that would have been anticipated and reflected in an agreement made at arm’s length’.41 This method is appropriate in cases when associated enterprises are integrated, in particular, cases ‘where both parties to a transaction make unique and valuable contributions’.42 However, the profit split implies adjusting and combining the books of both parties which incurs a significant administrative burden.43 Beyond these five methods, the OECD Guidelines allow the use of any other methods that provide a relevant arm’s-length price. In this regard, transactions associated with intra- group financing and business restructuring sometimes require specific approximation due to their intrinsic lack of comparability with independent party transactions.44 In 2020, the OECD issued some specific rules for financial transactions to address the issue of arm’s-length interest between related parties, the use of credit rating, and the treatment of cash pooling.45 Lastly, the 2022 consolidated version of the OECD Transfer Pricing Guidelines contains separate chapters on intangibles, intra- group services, cost contribution arrangements (CCAs), and business restructuring.

39

Ibid., 2. 68 mentions, e.g., the return on assets or the operating profit margin as such indicators. Transfer Pricing and Intangibles, US and OECD Arm’s Length Distribution of Operating Profits from IP Value Chains (2018), 270–271. 41 2022 OECD Guidelines, 2.114. 42 Ibid., 2.119. 43 Ibid., 2.123. 44 e.g. on cash pooling arrangements, see Chand, ‘Transfer Pricing Aspects of Cash Pooling Arrangements in Light of the BEPS Action Plan’, International Transfer Pricing Journal 38 (2016). 45 OECD, ‘Transfer Pricing Guidance on Financial Transactions: Inclusive Framework on BEPS’ (2020), Actions 4, 8–10. 40 Torvik,

544 Yuri Matsubara and Clémence Garcia

30.5 Intangibles Transfer pricing of intangibles is a central concern of the BEPS Project due to the numerous possibilities for tax avoidance that they offer.46 Indeed, the assignment of the ownership of intangibles within a group of companies affects the allocation of income between jurisdictions.47 In 2012 and 2013, the OECD had already published two reports on the matter prior to the start of the BEPS Project.48 Unsurprisingly, Actions 8–10 on the ALP and Action 13 on CbCR gave special importance to the valuation and reporting of intangibles. Action 8 introduced some rules to prevent groups from moving intangibles from one establishment to another, including notably hard-to-value intangibles (see later). Action 9 addressed the allocation of risks and the remuneration of capital necessary to finance intangibles, both areas in which comparables are typically unavailable. Action 10 addressed some concrete measures to implement Action 9. Lastly, Action 13 provided for a new format of reporting, CbCR, which allows for better exchange of information among jurisdictions. The 2022 consolidated version of OECD Guidelines reflects these changes. It defines intangibles broadly as ‘something which is not a physical asset or a financial asset, which is capable of being owned or controlled for use in commercial activities, and whose use or transfer would be compensated had it occurred in a transaction between independent parties in comparable circumstances’.49 Indeed, the definition includes some intangibles that may not be legally protected or recognized in financial accounts. After identifying the intangible elements involved in a transaction, a functional analysis must be performed to correctly delineate the transaction and the entitlement of the parties to intangible income.50 The treatment of intangible income can be summarized as follows: to the extent that one or more members of the MNE group other than the legal owner performs functions, uses assets, or assumes risks related to the development, enhancement, maintenance, protection, and exploitation of the intangible,51

46 Aside from the issues relating to profit shifting through measurement and location of intangibles, many high-tax countries offer advantageous regimes for investment in intangible assets. E.g. the cost- sharing regime in the US allows significant tax advantages for intangible investments. See Brauner, ‘Transfer Pricing Aspects of Intangibles: the Cost Contribution Arrangement Model’, 97–129. 47 In particular, the transfer of intangibles in the early development stages has been scrutinized by tax administrations in the last decades. See Torvik, Transfer Pricing and Intangibles, 63ff. 48 Kofler, The BEPS Action Plan and Transfer Pricing: The Arm’s Length Standard Under Pressure?, British Tax Review 5 (2013), 646–665. 49 2022 OECD Guidelines, 6.6. 50 Ibid., 6.34. 51 Development, enhancement, maintenance, protection, and exploitation (DEMPE). Some countries, e.g. India and China, have increased the number of functions to better represent the value

OECD Transfer Pricing Guidelines and International Tax Law 545 such associated enterprises must be compensated on an arm’s length basis for their contributions. This compensation may, depending on the facts and circumstances, constitute all or a substantial part of the return anticipated to be derived from the exploitation of the intangible.52

For transfer pricing methods, most intangibles are not separable or directly marketable to independent parties. Transfer pricing methods based on expected or realized income are preferred to cost-based methods and market value, because ‘there is rarely any correlation between the cost of developing intangibles and their value or transfer price once developed’.53 Hard-to-value intangibles (HTVIs) are the intangibles for which predicting future cash flows is difficult. In other words, HTVIs are intangibles that are subject to contingent risks. In such cases, the measurement of intangible income ex ante is not reliable or verifiable by the tax administration since most assumptions depend on the taxpayer’s internal information sources. The new rules provide that ‘the tax administration can consider ex post outcomes as presumptive evidence about the appropriateness of the ex-ante pricing arrangements’.54 To summarise the differences in these methods, the risks of intangible investments vary over time and the economic outcome is likely to differ from the expectations of a good-faith taxpayer and there is obviously a tax certainty issue arising from the ex post recharacterization of transactions.55 Finally, CCAs were also granted a specific regime.56 A CCA is a contractual agreement between related parties that transfers a share of the development risks in an early-stage development project to a related party in order to reduce subsequent transfer prices when its value appreciates. The rules addressing CCAs are aimed at preventing the separation of intangible income and economic activities that produce that income.57 To summarize, the post-BEPS changes on intangibles and the new transfer pricing rules show significant departures from the transactional view of the arm’s-length principal, reflecting a broader approach to comparability.

creation of decentralized establishments. See Sim, ‘The Arm’s Length Principle: Stretching the Tent to Bring Asia under the Broad Church’, in Sim and Soo, eds, Asian Voices: BEPS and Beyond (2017), 641–650. 52

OECD Guidelines, 6.71. Ibid., 6.142. 54 Ibid., 6.192. 55 Wittendorff, Transfer Pricing and the Arm’s Length Principle in International Tax Law, 689. 56 See Brauner, ‘Transfer Pricing Aspects of Intangibles: the Cost Contribution Arrangement Model’, 97–129, for a comparison with the US cost-sharing regime. 57 Collier and Andrus, Transfer Pricing and the Arm’s Length Principle After BEPS, 217. 53

546 Yuri Matsubara and Clémence Garcia

30.6 Recent Transfer Pricing Disputes around the Globe Due to the development of multinational enterprises and international labour mobility, the volume of overseas corporate transactions has grown rapidly in the last two decades.58 Consistently, the number of transfer pricing disputes around the world has also drastically increased. According to the TPcases.com database,59 from January 2010 until December 2020 more than 580 transfer pricing disputes were settled through legal means worldwide. The legal issues in those cases are varied and complex. For instance, the burden of proof for the eligibility of transfer pricing methods, business restructuring (especially hidden distributions of profit, non-recognition, and recharacterization), valuation methods for financial transactions (e.g. DCF valuation), commodity transactions (value chain analysis, controlled transactions), evaluation of intangibles (i.e. goodwill, know-how patents), the application of general anti-avoidance rule (GAAR), and the characterization of permanent establishments have been the most numerous matters in dispute. Furthermore, the same issues are duplicated in several jurisdictions because multinational groups sometimes file petitions in several countries on transfer pricing issues. It should be noted that the aforesaid transfer pricing judgments played an important role in the development of transfer pricing. Some of the most important decisions have been reflected in the interpretation and implementation of international rules, such as the OECD Transfer Pricing Guidelines, and also when amending existing domestic legislation (e.g. IRC, s. 428 in the USA) along with the relevant tax treaties.60 For example, GlaxoSmithKline, a UK-headquartered multinational pharmaceutical group, stated to the Canadian tax court that the transfer price for transactions related to the medical substance ‘ranitidine’ between its Swiss subsidiary and Glaxo Canada was reasonable and was in accordance with the Canadian domestic transfer pricing provision.61 The parent company, Glaxo Holdings UK, had decided that local distributors should retain a gross profit of 60% of revenue, while the remaining 40% would belong to Glaxo Holdings UK as royalties or a different means of transfer pricing. Thus, Glaxo Canada relied on a specific application of the resale price method. However, the Canadian tax authorities referred to the CUP method when reassessing the arm’s-length 58 According to the statistic of the CPB World Trade Monitor in the Netherlands (retrieved 23 March 2021), except for the periods concerning (1) the 9/11 attacks in 2001, (2) the 2008 financial crisis, and (3) the Covid-19 crisis in 2020, the volume of world trade has been increasing; namely, in comparison with the figures for January 2000, the volume of world trade in January 2020 had more than doubled. 59 ‘Transfer Pricing Case Law and Guidelines from around the World’, tpcases.com, (accessed 23 March 2021). As at August 2022, approximately 600 cases were recorded in the database. 60 See further, e.g., Nutt, ‘Glaxo, U.S. Settle Transfer Pricing Dispute’, Tax Notes International 43 (2006), 956; Torvik, Transfer Pricing and Intangibles, 221–236. 61 CA: s. 69(2) of Canadian Income Tax Act (1985 version).

OECD Transfer Pricing Guidelines and International Tax Law 547 price. It was assessed based on a comparison with transactions between unrelated generic companies and other pharmaceutical companies in Canada. The Tax Court of Canada at first instance, and the Federal Court of Canada at second instance, upheld the position of tax the authorities, adding that OECD Commentaries should provide information on the interpretation of the Canadian domestic transfer provisions.62 In October 2012, in its first ruling on a transfer pricing case, the Supreme Court of Canada (SCC) reversed the decisions of the courts of lower instance. The SCC found that the 1995 OECD Guidelines required economically relevant characteristics between the controlled and uncontrolled transactions to be sufficiently comparable.63 Further, the SCC went on to comment that the generic comparables did not reflect the economic and business reality of Glaxo Canada, nor did they indicate the arm’s-length price for ranitidine. Thus, the Canadian tax authority lost the case. Recently, several important transfer pricing judgments were rendered in the USA. First, the US Internal Revenue Service (IRS) succeeded in a case against the Altera Corp. in the Ninth Federal Circuit Court (2d) in 2017.64 One of the pioneers of the specialized electronic components in programmable logic devices, Altera Corp., entered into a cost- sharing arrangement (CSA) for R&D with its foreign subsidiary (Altera Cayman) in May 1997. According to that arrangement, Altera Corp. retained the right to exploit any fruits of the R&D in the USA and Canada while permitting Altera Cayman to do the same in the rest of the world. For the period 2004–2007, federal income tax returns included Altera Cayman payments to Altera Corp. as part of the CSA. However, the amount of compensation that Altera Cayman paid to Altera USA excluded the amount that Altera Corp. paid for employees through stock options or other stock-based compensation.65 As a consequence, Altera Corp. saved the withholding tax burden in the USA for its resident employees, through stock-based compensations that had been promulgated in the US Treasury Department’s 2003 final rule.66 The main issues were: (1) whether ‘intangibles’, such as stock-based compensation, needed to be included in the CSA; and (2) if so, whether that new rule applied retrospectively to a CSA concluded in 1997. The US tax court decided in favour of Altera Corp.67 and the IRS appealed to the Ninth Circuit Court, arguing that including stock-based compensation in the costs was appropriate on the basis of the arm’s-length standard.68 Further, the IRS added that since there were no equivalent uncontrolled arrangements, the commensurate-with-income method could apply to allocate costs and create an arm’s-length result. Altera Corp. responded that the arm’s-length standard required nothing other than a comparative

62

CA: TC ruling, paras 86–96. CA: SCC ruling, para. 42. 64 US: Altera Corp. v. Commissioner (Altera II-1), No. 16-70497 (2018). 65 Brian, ‘Altera Corp. v Commissioner and Transfer Pricing’, University of Cincinnati Law Review 87/ 4 (2019), 1123. 66 US: 68 Fed. Reg. 51,171, 51,177, 51,179 (Aug. 2003 version). 67 US: Altera Corp. v. Commissioner (Altera I), 145 TC (2015). 68 US: Altera Corp. v. Commissioner (Altera II-1), No. 16-70497 (2018) (Altera I: 145 TC (2015)). 63

548 Yuri Matsubara and Clémence Garcia analysis.69 The Ninth Circuit Court upheld the position of the IRS in July 2018. Even though its opinion was withdrawn at one stage due to the death of Judge Reinhart, the opinion issued in June 2019 confirmed the judgment in favour of the IRS.70 Altera Corp. appealed to the US Supreme Court, but its request for review was denied. Another recent court decision upheld the position of the taxpayer. The Coca-Cola Co. and its Mexican subsidiary were involved in a dispute concerning the deductibility of foreign tax credits.71 The Coca-Cola Co. produces concentrate and sells it to licensed bottlers who convert the concentrate into soft drinks. During the years 2007–2009, Coca-Cola’s tax return reported income from foreign supply points using a formulary apportionment method: ‘10–50–50’. First, a fixed profit margin was calculated as 10% of revenue, then the residual profit was allocated equally (50–50) between the US parent of Coca-Cola (licensor) and its foreign licensees. Initially, the formulary apportionment was accepted by the IRS in an advance pricing agreement in 1996. However, the IRS later changed its position because it had been based on the tax returns from 1987 to 1995. The IRS assessed that by choosing the ‘10–50–50’ method, Coca-Cola had undercharged its foreign affiliates; therefore, Coca-Cola should have used the CPM to determine arm’s- length prices. In this instance, the US tax court allowed the deduction of foreign tax credits to Coca-Cola, since Coca-Cola and its Mexican subsidiary had agreed with the Mexican tax authority to continue using the already expired CSA with the US IRS on the advice of their Mexican counsel. Subsequently, Coca-Cola filed a second lawsuit against the IRS in relation to transfer prices of intellectual property. In November 2020, the court upheld the IRS position,72 which is rare in recent US transfer pricing disputes and is therefore noteworthy. In this second case, the tax court agreed with the IRS that Coca-Cola’s US-based income should be restated by US$9 billion in a dispute over royalties from its foreign licensees. Despite Coca-Cola filing a motion to reconsider the opinion in June 2021 (183 days after the tax court had rendered its opinion),73 the court confirmed its judgment in favour of the IRS.74 In Europe, another important transfer pricing case on a cross-border financial transaction (in this case an in-bound loan) was decided in May 2021 before the German Federal Tax Court75 which directly reflected the updating of chapter X of the 2022

69

US: Altera II-1, ibid., 7.

70 US: Altera Corp. v. Commissioner (Altera II-2), Nos 6-70496 and 70497. 71 US: Coca-Cola Co. & Sub. v. Commissioner, 149, TC No. 21 (2017).

72 US: Opinion of Coca-Cola Co. & Sub. v. Commissioner, 115 TC No. 10, Docket No. 31183-15 (Nov. 2020); Coca-Cola v. Commissioner, TC Docket No. 31183-15. 73 US: Order of Coca-Cola v. Commissioner, Docket No. 31183-15 (TC 2021). 74 As mentioned earlier, the US Tax Court, in its opinion of 18 November 2020, stated that the CPM method was the most appropriate for determining arm’s length in that case. The opinion was upheld in the order of October 2021. 75 Germany: Bundesfinanzhof (Federal Tax Court), BFHE 18 May 2021, https://www.bundesfinanz hof.de/de/entscheidung/entscheidungen-online/detail/STRE202110196/ (in German only, accessed 15 August 2022).

OECD Transfer Pricing Guidelines and International Tax Law 549 OECD Guidelines. A German company established a Dutch financial subsidiary and a Dutch holding company that controlled both of them. That German company concluded a loan agreement with its Dutch financial company. In addition, the German parent received an external loan from a German bank that was guaranteed by the Dutch holding company. While the German tax authority held the position that CPM should be applied, the taxpayer maintained that CUP should be applicable. While the lower instance Tax Court of Munster decided in favour of the tax authority, the German Federal Tax Court overruled the decision because the court had not sufficiently examined whether the CUP method could be applied to determine the arm’s-length interest rate of the transaction. Finally, it should also be noted that some geographical tendencies appear when selecting transfer pricing methods. While the tax authorities in many Western countries tend to prefer traditional methods such as CUP, RPM, and CPM, Asian taxpayers and tax authorities are more favourable to the TNM method.76

30.7 Impact of CbCR A new and important change to transfer pricing taxation in the post-BEPS era concerns the standardization of documentation within multinational groups. A domestic parent company (i.e. ‘ultimate parent entity’) will prepare for its group a master file, while its foreign subsidiaries, affiliates, or related companies (i.e. ‘immediate or surrogate parent entity’) will submit local files to their parent company. BEPS Action 13 calls this Country by Country Reporting (CbCR). Under the CbCR mechanism, the total number of transactions in the local files should match the amount in the master file. Thus, it is expected that tax inspectors will be able to easily trace cross-border the transactions of targeted multinationals through the exchange of information with their foreign counterparts. Initially, such transfer pricing documentation had been provided for in Europe in the Code of Conduct on transfer pricing documentation (EU TPD) in 2006.77 Even though the EU TPD merely ‘recommended’ that companies present their standardized and partially centralized TPD to the tax authorities in EU member states, this worked efficiently well to capture intra-group transactions between EU member states. The OECD has enlarged this idea and made it mandatory. When BEPS Action 13 on CbCR was released in 2015, it became one of four BEPS minimum standards. As mentioned earlier, the aim of CbCR is to promote international transparency by means of information exchange

76

See Section 30.4. EU Tax Law (2011), 239– 240. Marino, ‘The Burden of Proof in Cross- Border Situations’, in G. Meussen, ed., The Burden of Proof in Tax Law (2011), 39–45. 77 Heinemann,

550 Yuri Matsubara and Clémence Garcia between tax authorities. To ensure the effectiveness of CbCR, the OECD provided for a peer-review mechanism, which became a minimum standard in February 2017. Meanwhile, the OECD Transfer Pricing Guidelines were also amended in July 2017.78 In October 2017, the OECD released its standardized electronic format (called CbCR XML) for sharing CbCR between jurisdictions.79 In practice, the implementation of CbCR has increased the administrative burden for multinationals, especially the reconciliation of figures in the master file with those in the local files.80 Additionally, a large volume of information was transferred at one time from taxpayers to the tax administrations. Therefore, it makes sense for the relevant tax authorities to sort the entire CbCR documents/information using artificial intelligence.

30.8 Dispute Resolution Mechanisms (APA/MAP/T TA) As a means of reducing transfer pricing disputes, especially in the last two decades, two administrative dispute-resolution mechanisms have been developed on top of the domestic mechanisms.81 First, advance pricing arrangements or agreements (APA) spread during that period.82 An APA is an administrative procedure based on ex ante confirmation of the transfer pricing method to be used between one or more taxpayers and the tax authorities (including the competent authority) before they enter into international intra-group transactions. APAs were officially introduced in the USA in 2011; however, they initially appeared in Japan as a pre-confirmation system (PCS) in 1987. In the late 1990s, the idea of a PCS reached the US IRS from the Japanese tax authorities. The main function of modern APAs can also be traced to a similar system in the USA, known as the private letter ruling (PLR). Unlike the Japanese PCS, the US PLR was not free of charge, and due to several other reasons (e.g. it being time-consuming, the increasing risk of being targeted for tax audits, or the risk of leaking trade secrets), US taxpayers did not frequently request PLRs.83 In international tax disputes, the APA mechanism appears efficient to limit double taxation issues. From the perspective of multinational groups, global APAs are expected

78

See 2022 OECD Guidelines, ch. V (documentation). It was updated in January 2020. 80 Tavares, Romero J.S., ‘Country-by-Country Over-Reporting? National Sovereignty, International Tax Transparency, and the Inclusive Framework on BEPS’, S. A. Rocha and A. Christians, eds, Tax Sovereignty in the BEPS Era, WK Law (2017), 201. 81 See 2022 OECD Guidelines, ch. IV (administrative dispute resolution). 82 In the USA, it is called a transfer pricing agreement (TPA). 83 Givati, ‘Resolving Legal Uncertainly: The Unfulfilled Promise of Advance Tax Ruling’, Virginia Tax Review 29 (2009), 137. 79

OECD Transfer Pricing Guidelines and International Tax Law 551 to considerably reduce the risk of unexpected double taxation via the validation of arm’s- length pricing of cross-border transactions. Additionally, an APA may provide rules for subsequent transfer pricing adjustments by the tax authorities. Thus, refined APAs were adopted in Canada (1994), Australia (1995), South Korea (1996), and the Republic of China (1998). Currently, over forty jurisdictions use APAs which has built a successful global network. Many US groups repeatedly use global APAs; usually, these are multi- period agreements between taxpayers and one or more tax authorities. APAs can be classified into three types: (1) unilateral APAs (UAPA) involving only one taxpayer and the tax authority where its headquarter is located; (2) bilateral APAs (BAPA) which is an APA involving the taxpayer, its foreign associate enterprise (AE i.e. either subsidiary or affiliate company), and the tax authorities where the taxpayer and the foreign AE are located; and (3) multilateral APAs (MAPA) involving the taxpayer, two or more AEs, and the competent tax authorities of foreign AEs. In comparison with UAPAs and BAPAs, a MAPA is more complex and is therefore less used in practice by multinational groups. However, due to increasing transfer pricing disputes worldwide, the role of MAPAs is expected to broaden. While the APA mechanism is an ex ante solution to limit tax disputes, tax authorities also provide for an ex post solution for resolving transfer pricing disputes. A mutual agreement procedure (MAP) is an administrative bilateral negotiation between the tax authorities of two jurisdictions. If taxpayers (e.g. MNE groups) are not satisfied with the result of the reassessment of the taxable amount after an administrative tax investigation (not criminal investigation), those taxpayers are entitled to request a MAP from their tax authority, which subsequently negotiates the allocation of the tax base with its foreign counterpart. This mechanism has spread due to BEPS Action 14 that has made it a minimum standard, so that it has been implemented in all the jurisdictions of the Inclusive Framework. In October 2016, the OECD released some key documents in BEPS Action 14, which formed the basis of the MAP peer-review and monitoring process. Today, many countries have disclosed their MAP-related statistics and mechanisms directly on the OECD’s website.84 Currently, MAP is provided for in article 25(1) of the OECD Model Convention.85 It states as follows: Where a person considers that the actions of one or both of the Contracting States result or will result for him in taxation not in accordance with the provisions of this Convention, he may, irrespective of the remedies provided by the domestic law of those States, present his case to the competent authority of either Contracting State. The case must be presented within three years from the notification of the action resulting in taxation not accordance with the provisions of the Convention.

84 85

OECD, ‘Dispute Resolution (retrieved 30 April 2021). It was amended in 2017.

552 Yuri Matsubara and Clémence Garcia Over the years, MAP has been recognized by taxpayers as a means of avoiding double taxation. Conversely, the mechanism places tax inspectors in the challenging role of negotiating on behalf of taxpayers to reduce their tax burden in other state. It should be noted that—unlike international commercial arbitration—multinational groups cannot participate in person in tax treaty arbitration (TTA). The negotiation takes place between the tax authority of the residence state and its foreign counterpart which negotiates on behalf of the taxpayers. Thus, during a MAP, taxpayers neither know the details of the negotiations between the tax authorities nor the expected outcome (i.e. the amount of tax liability), the duration of the process, or the content of the transfer pricing adjustment. According to the OECD Model Convention, the MAP will be terminated between both competent authorities within two years. This was prescribed in article 25(5)(b) OECD MC in 2008. [Where] the competent authorities are unable to reach an agreement to resolve that case pursuant to paragraph 2 within two years from the date when all the information required by the competent authorities in order to address the case has been provided to both competent authorities.

The time frame appears to be enforced in many EU countries, but in the rest of the world, especially in Asian countries, the MAP deadlines are not always met.86 Additionally, if both competent authorities have failed to settle a MAP, taxpayers can either: (1) sue their government in a domestic court; or (2) engage in arbitration by applying a TTA. As mentioned earlier, in 2008 the OECD Model Convention, article 25(5), provided for an arbitration clause in the event that MAP was unsuccessful in that time frame.87 In addition to the previously mentioned traditional dispute-resolution mechanisms, arbitration is considered a possible tool for international transfer pricing matters. However, the history of TTA is much shorter than that of international commercial arbitration.88 In Europe, the EU Arbitration Convention (90/436) was released in 1990. Even if it did not relate directly to non-EU tax disputes, several European countries included some TTA clauses when concluding their bilateral tax treaties after the amendment of the OECD Model Convention in 2008. Namely, Austria, the Netherlands, and Switzerland have such a treaty policy.89 Although Switzerland is not a part of the EU, the Swiss tax authority has been involved in numerous international tax disputes, more precisely BAPAs and MAPs with the US IRS. Later, the EU Dispute Resolution Directive

86 e.g. according to the statistics of the annual report of the NTA (Japan), the number of unsettled MAPs increased almost three times from 2005 to 2019. 87 van Herksen, ‘International Developments’, A. Bakker and M. Levey, eds, Transfer Pricing and Dispute Resolution, Aligning Strategy and Execution (2011) 39–41. 88 Pistone, ‘General Report’, M. Lang et al., eds, Tax Treaty Arbitration (2020), 1. 89 Pistone, ibid., 7.

OECD Transfer Pricing Guidelines and International Tax Law 553 (2017/1852), released in 2017, also reflected the introduction of the TTA clause in the OECD Model Convention. The arbitration clause in the OECD Model Convention also appears to have influenced the US model of bilateral tax treaties, as well as the UN Model.90 However, unlike international commercial arbitration, the TTA mechanism is still in its infancy, and several issues (e.g. the lack of skilled tax arbitrator and experts witness or the place of the arbitration court) remain pending. In comparison with the EU and the US, only a few jurisdictions in the rest of the world are interested in the TTA mechanism and have a tax treaty policy that includes the TTA clause.91 Consequently, neither the 2022 OECD Transfer Pricing Guidelines nor the BEPS Action Plan refer directly to TTA or emphasize arbitration. However, this could change if the USA became part of the OECD Multilateral Instrument to engage in future BEPS 2.0 negotiations. TTA will also be in the spotlight as a new means of dispute resolution in transfer pricing matters.

90 Arnold, ‘The Scope of Arbitration under Tax Treaties’, in M. Lang and J. Owens, eds, International Arbitration in Tax Matters (2015), 113–137. 91 e.g. Canada, Australia, and Japan support TTA.

Chapter 31

C orresp ondi ng Adjustm e nts Matthias Hofacker

31.1 Introduction Corresponding adjustments are reports due to a previous initial correction of a company’s tax base. The aim is to avoid economic double taxation. Economic double taxation (taxation of the same income in the case of different persons) occurs when a company in state A whose profits are increased becomes liable to tax on an amount of that profit that has already been taxed through a company associated with it in state B.1 The adjustment requires that, in these cases, state B make an appropriate adjustment to avoid double taxation.2 This measure to avoid economic double taxation is provided for in article 9 paragraph 2 OECD Model Taxation Convention (OECD MC). Insofar as no corresponding provision has been agreed upon in the specific double taxation treaty between the contracting states, the provision is intended to serve as a model for the legislator to take precautions in the respective domestic law in order to prevent economic double taxation.3 However, the purpose of article 9 OECD MC is not primarily the avoidance of economic double taxation, but rather the allocation of taxation rights according to economic aspects (art. 9 para. 1 OECD MC) and, concomitantly, the avoidance of economic double taxation.4 This avoidance of economic double taxation is to be ensured in particular by article 9 paragraph 2 OECD MC, provided that the other state shares the initial correction of the first state on the merits and in the amount; consequently, the avoidance 1

Art. 9 no. 5 OECD Commentary. Art. 9 no. 5 OECD Commentary. 3 Para. 4.32 OECD Transfer Pricing Guidelines (OECD TP Guidelines) 2017. 4 Carroll, Methods of Allocating Taxable Income, in League of Nations, Taxation of Foreign and National Enterprises, vol. IV (Geneva: League of Nations, 1933). 2

556 Matthias Hofacker of double taxation represents an interplay of article 9 paragraph 1 OECD MC and article 9 paragraph 2 OECD MC.5 However, if a corresponding adjustment is not made via article 9 paragraph 2 OECD MC, because the other state does not share the first correction of the first state on the merits or in the amount, double taxation of the same tax object may occur. The same applies if the provision of article 9 paragraph 2 OECD MC is not agreed upon in the specific double taxation treaty. Then, at least according to article 25 paragraphs 1 and 2 OECD MC, the states shall endeavour to eliminate this double taxation by mutual consultations in the mutual agreement procedure. In international tax law, the corresponding adjustment is located in the area of transfer pricing. The application of the corresponding adjustment is based on the OECD’s idea that a correction (initial adjustment) is made to the tax base in a state because the company concerned did not observe the arm’s-length principle when agreeing on the price (transfer price) arising from a business relationship with a person related to that company, for example a sister company. Often, too low a price is agreed, so that the initial correction leads to an increase in the tax base of the company concerned.6 Without the correction, related companies could reduce the group tax rate in this way.7 This not only leads to a reduction in the tax revenue of the state concerned. From an economic point of view, the companies concerned achieve unintended tax advantages that result in a distortion of competition.8 Therefore, it is necessary to perform the initial correction. The logical consequence is that the tax base of the other company is reduced by the amount of the increase due to the initial correction.9 After all, the correction amount has been earned only once and can therefore be taxed only once. Furthermore, this corresponds to the system of profit distribution in article 7 paragraph 1 sentence 1 HS 1 OECD MC. Article 9 paragraph 2 OECD MC expressly provides that the competent authorities shall consult each other.10 This is to ensure the appropriate corresponding adjustment with the state implementing the initial correction.11 This confirms that the mutual agreement procedure under article 25 OECD MC can be used.12

31.2 Regulation The corresponding adjustment covers the cases in which the economic double taxation of two interrelated companies arises due to the attribution of an adjustment to the profit 5

Art. 9 no. 5 OECD Commentary; Li, Canadian Tax Journal (2002), 836. Para. 4.32 OECD TP Guidelines 2017. 7 OECD, ‘Addressing Base Erosion and Profit Shifting’ (2013), 6, 48. 8 Kumar et al., Journal of Business Research 134 (Sept. 2021), 275. 9 Para. 4.32 OECD TP Guidelines 2017. 10 Jacobs, Endres, and Spengel, Internationales Unternehmensbesteuerung, 8th ed. (Munich: C. H. Beck, 2016), 667. 11 Para. 4.33 OECD TP Guidelines 2017. 12 Para. 4.33 OECD TP Guidelines 2017. 6

Corresponding Adjustments 557 and its taxation in one contracting state. The corresponding adjustment in the other contracting state is therefore the consequence of the assumption of an incorrect profit allocation on the part of the affiliated companies. Thus, the corresponding adjustment becomes necessary if the result of the first adjustment remains. If this only happens because the first-adjusting state maintains its measure contrary to the opinion of the other contracting state, the other state will refuse to make the corresponding adjustment. The result is economic double taxation. If both contracting states agree that the first adjustment was based on an inappropriate arm’s-length price, then one contracting state will make a correction to the amount of the adjustment from the first adjustment, while the other contracting state will grant relief in the amount of the corrected amount of the adjustment with the corresponding adjustment. In this case, the economic double taxation is eliminated by the cooperation of both contracting states. For an application of article 9 paragraph 1 OECD MC as a basis for the corresponding adjustment, it is not decisive whether the double taxation was triggered by the taxpayer, for example because they changed the tax return, or whether the double taxation was triggered by the tax administration. What is decisive is only that an increase in the tax base has occurred as a result of the correction of the income and that a state has exercised its right of taxation under article 9 OECD MC.13 If both contracting states agree that the first adjustment in one state was not justified even on the merits, then it must be reversed by one contracting state. This eliminates economic double taxation without a corresponding adjustment under article 9 paragraph 2 OECD MC. Only if both contracting states agree that the first adjustment in one state was justified both on the merits and in terms of the amount, is the other contracting state obliged to perform the corresponding adjustment in accordance with article 9 paragraph 2 OECD MC.14 Thus, the prerequisites for the adjustment of profits are that: • in one contracting state, profits are attributed to the profits of an enterprise of that state and taxed accordingly; • an enterprise of the other contracting state has been taxed in that state; and • the attribution of profits complies with the arm’s-length principle. If these conditions exist in addition to the conditions of article 9 paragraph 1 OECD MC, whereby the assessment is made by the other contracting state, that other state shall make a corresponding change to the taxes levied there on those profits. Here, the arm’s-length principle must be emphasized once again. This forms the benchmark for the intra-group terms and conditions and compares them with those that would have been agreed upon by independent companies. Specifically, the 13 Lehner

and Eigelshoven, Doppelbesteuerungsabkommen, 6th ed. (Munich: C. H. Beck, 2015), art. 9 m.no. 129a; Rosenberger et al., Doppelbesteuerungsabkommen, 2nd ed. (Vienna: Linde, 2019), art. 9 m.no. 222 14 Wassermeyer and Baumhoff, Verrechnungspreise international verbundener Unternehmen (Cologne: Otto Schmidt Verlag, 2014), 151.

558 Matthias Hofacker arm’s-length principle means that the affiliated companies must be bound in their commercial or financial relationships by agreed or imposed terms and conditions that independent companies would agree with each other.15 In this context, the arm’s-length price is determined from empirically traceable prices or margins realized between unrelated firms.16 If the other state considers the initial correction of the first state to be appropriate in terms of reason and amount, it is obliged to make the corresponding adjustment from article 9 paragraph 2 OECD MC,17 without carrying out a mutual agreement procedure within the meaning of article 25 OECD MC.18 Of course, the contracting states will consult. Otherwise, in practice, there will be no consensus among the contracting states on the applicable correction in one state. The obligation to make a corresponding adjustment is unavoidable in the interest of the taxpayer.19 However, this also concludes the legal consequence. In order to avoid economic double taxation in the relationship between two interrelated companies, the provision in article 9 paragraph 2 OECD MC constitutes a separate rule for the avoidance of double taxation.20 Articles 23A and 23B OECD MC are not applicable. It is also irrelevant which method was used by one contracting state to adjust the profit. Therefore, it is also irrelevant whether the correction of the profit has been made within a tax balance sheet (according to accounting principles) or outside the balance sheet when determining the final tax base. Also covered by the correction of profit are, for example, cases of the reduction of a loss or cases of lower depreciation or when determining a loss carry-forwards, which only have an effect in later years.21 The profits attributed to the affiliated company are, therefore, ‘taxed’ even if the profit adjustment has not led to an immediate tax payment obligation, but to measures that will or may still have an effect on future tax payments. The same applies if a contracting state has not subjected the profits to taxation, for example because of a special tax exemption for individual companies or because a state does not tax corporate profits in general. The profit allocation and corresponding taxation according to article 9 paragraph 1 OECD MC is the connecting factor and, in principle, also the benchmark for the scope of the corresponding adjustment according to article 9 paragraph 2 OECD MC. The wording profit attribution ‘and’ corresponding taxation expresses that economic double taxation must have actually occurred. In this respect, the corresponding adjustment is

15 Rogers and Oats, ‘Transfer Pricing: Changing Views in Changing Times, Accounting Forum (2021), 83. 16 Para. 2.10 OECD TP Guidelines 2017. 17 Art. 9 no. 6 OECD Commentary; Schaumburg and Häck, Internationales Steuerrecht, 4th ed. (Munich: C. H. Beck, 2017), 882; Schwenke and Greil in Wassermeyer, ed., Doppelbesteuerungsabkommen (Munich: C. H. Beck, 2021), art. 9 m.no. 29. 18 Greil, ifst-Schrift No. 512, 39ff. 19 Jacobs, Endres, and Spengel, Internationales Unternehmensbesteuerung, 8th ed. (Munich: C. H. Beck, 2016), 667. 20 Ditz, Internationale Gewinnabgrenzung bei Betriebsstätten (Berlin: ESV, 2003), 64. 21 Schwenke and Greil, in Wassermeyer, Doppelbesteuerungsabkommen, art. 9 m.no. 191.

Corresponding Adjustments 559 directed only to the avoidance of effective (economic) double taxation.22 Assuming that the adjustment of the profit in one contracting state was carried out with the correct application of the arm’s-length principle, the principle must also be observed for the corresponding adjustment, even if no specific method of adjustment is therefore given to the other contracting state. Article 9 paragraph 2 sentence 1 OECD MC assumes that the ‘taxes levied on such profits’ of the other contracting state are to be changed. In order to determine the appropriate tax, the other contracting state must generally apply the arm’s-length principle when it is not considering crediting the taxes of one contracting state.23 However, it is sufficient for the initiation of a mutual agreement procedure pursuant to article 9 paragraph 2 OECD MC in conjunction with article 25 OECD MC, if there is a threat of taxation that does not comply with the specific double taxation agreement.24 As a connecting factor for the corresponding adjustment, only a profit entitlement between associated enterprises within the meaning of article 9 paragraph 1 OECD MC can be considered. This results from the repetition of the fiction of independence of article 9 paragraph 1 OECD MC in article 9 paragraph 2 OECD MC. The OECD MC does not contain any statement on whether article 9 paragraph 2 OECD MC is also applicable to triangular constellations: If, for example, a foreign permanent establishment in state A of a company resident in state B supplies discounted goods to a company in state C, it must be examined whether article 9 OECD MC applies. The double taxation treaty between state A and state C would not be applicable in the present case, since the company is resident in state B and the permanent establishment in state A is not a treaty-entitled person within the meaning of article 3 paragraph 1 no. d) OECD MC. Therefore, only the double taxation treaty between state B and state C could be applicable. For application, consider whether state B’s profits have been reduced by arm’s-length conditions. In this regard, it is argued that due to the world income principle, the income of the permanent establishment in state A is subject to taxation in state B.25 For the question of the applicability of article 9 OECD MC, it is therefore irrelevant whether state B has waived its right of taxation within the scope of the application of its double taxation treaty with state A. State C would therefore be obliged to make a corresponding adjustment in the double taxation agreement between state B and state C on the basis of the article comparable to article 9 OECD MC. State A would be entitled to an increase in profits on the basis of article 7 of the double taxation agreement between state A and state B, since the profits from the supply of goods are attributable to the permanent establishment in state A.

22

Eigelshoven, in Vogel and Lehner, eds, Doppelbesteuerungsabkommen (Munich: C. H. Beck, 2015), art. 9 m.no. 163; Schwenke and Greil, in Wassermeyer, Doppelbesteuerungsabkommen, art. 9 m.no. 212; Andresen, Konzernverrechnungspreise für multinationale Unternehmen: Paradigmenwechsel bei der steuerlichen Einkünfteabgrenzung (Deutscher Universitäts-Verlag, 1999). 23 Schwenke and Greil in Wassermeyer, Doppelbesteuerungsabkommen, art. 9 m.no. 212. 24 Eigelshoven in Vogel and Lehner, Doppelbesteuerungsabkommen, art. 9 m.no. 163. 25 Ibid., m.no. 164a.

560 Matthias Hofacker The OECD26 also sees the case within the scope of article 9 paragraph 2 OECD MC that under the domestic law of some countries, a taxpayer may be permitted to make a correction under appropriate circumstances and thus amend previously filed tax return.

31.2.1 Assessment by the Other Contracting State The corresponding adjustment pursuant to article 9 paragraph 2 OECD MC presupposes that, as a result of the adjustment of profits made in one contracting state, economic double taxation has occurred in the relationship between two associated enterprises. This requires the taxation of the tax object at the other associated enterprise in the other contracting state. The occurrence of economic double taxation (=identity of the tax object) is assessed by the other contracting party from the perspective of its domestic tax law.27 Therefore, the legal consequences of article 9 paragraph 2 OECD MC must also be distinguished from those arising from articles 23A and 23B OECD MC. The latter provisions only concern cases of legal double taxation; that is, the subject of the levy must have an identity. From this point of view, the legal consequences of the provisions are mutually exclusive. The taxation of profits in the first state must be appropriate for a corresponding adjustment. The corresponding adjustment must factually correspond with the correction of the profit in the one contracting state. Therefore, it does not need to match the amount of the correction made to the profit. It also does not need to be recognized in the same tax year as the adjustment to the gain.28 Nevertheless, the scope of the corresponding adjustment is limited. The corresponding adjustment only covers those cases of economic double taxation in which this is a direct consequence of the initial adjustment. If, on the other hand, the economic double taxation arises, for example, from qualification conflicts or from different national income determination rules, a corresponding adjustment is excluded.29 The inclusion of article 9 paragraph 2 OECD MC in a specific double taxation treaty does not mean that the other state will be forced to adjust the profit in every case.30 A claim under treaty law for a corresponding adjustment only exists if, from an economic point of view, the two contracting states tax the same object in two different interrelated companies. Whether this is the case is ultimately decided solely from the perspective of the domestic law of the other contracting state. In doing so, the other contracting state judges autonomously (i.e. independently of the taxation in the contracting state of the first adjustment). The other contracting state checks whether the contracting state of the first adjustment has made the adjustment of the profit in 26

Art. 9 no. 6.1 OECD Commentary. Foley and Dhanuka, ‘Secondary Transfer-Pricing Adjustments’, The Tax Adviser (June 2022). 28 Art. 9 no. 10 OECD Commentary. 29 Eigelshoven in Vogel and Lehner, Doppelbesteuerungsabkommen, art. 9 m.no. 161. 30 Art. 9 no. 5 OECD Commentary. 27

Corresponding Adjustments 561 accordance with the arm’s-length principle (based on the merits and the amount). Thus, an obligation to make a corresponding adjustment can only exist to the extent that there is a consensus among states on the application of the arm’s-length principle.31 According to the OECD, the state of primary adjustment bears the burden of proof in a competent authority proceeding to demonstrate to the other state that the adjustment is justified both in principle and in amount.32 The OECD expects the competent authorities of both parties to take a cooperative approach in resolving cases of mutual agreement.33 The other state will additionally consider whether the entities concerned are also related entities in its view (under its domestic tax law) and whether the adjustment of the profit concerns commercial or financial relations of the related entities.34 In addition, the other state will track whether the arm’s-length violation occurred in the transaction between the related entities that the first state identified when making the adjustment or to a third unrelated entity. If the results of the two contracting states differ, the other contracting state will not see its obligation to make the corresponding adjustment from article 9 paragraph 2 OECD MC in the specific double taxation agreement. The other contracting state will therefore examine whether, in the opinion of the other state, the first state of the initial adjustment has correctly applied the provision of article 9 paragraph 1 in the specific DTA. In this context, the arm’s-length principle is not limited to the amount, but also extends to the reason and, consequently, to the amount of the profit: if domestic law allows, for example, debt capital to be reclassified as equity, article 9 OECD MC does not prevent this to the extent that ‘its effect is to adjust the profits of the borrower to an amount equal to the profits that would have accrued in an arm’s length situation’.35 In other words, article 9 OECD MC seeks to ensure that a state only takes into account for taxation purposes profits that would have arisen in comparable arm’s-length transactions.36 However, a corresponding adjustment of the other state can only be considered if that state is of the opinion that by making the adjustment to the profit, the profits are now reflected that would also have arisen between unrelated third parties.37 Thus, article 9 no. 3b OECD Commentary accordingly also states that article 9 OECD MC is not only relevant to determining the fair price, but also whether a loan is to be regarded as a loan or as another type of payment, in particular as an equity contribution. This is because if an affiliated company has more borrowed capital than a comparable company, the profit will be reduced due to the increased interest burden compared with the comparable company.

31

Art. 9 no. 6 OECD Commentary; Greil, ifst-Schrift No. 512, 40. Para. 4.17 OECD TP Guidelines 2017. 33 Para. 4.17 OECD TP Guidelines 2017. 34 Schwenke and Greil in Wassermeyer, Doppelbesteuerungsabkommen, art. 9 m.no. 197. 35 Art. 9 No. 3a OECD Commentary. 36 Brauner, Intertax (2014), 615. 37 Avi-Yonah, World Tax Journal (Feb. 2010), 3. 32

562 Matthias Hofacker The other state may, of course, provide a provision in its domestic tax law for the corresponding adjustment, which obliges the other state to do so irrespective of the application of a provision under treaty law.

31.2.2 Legal Basis for the Corresponding Adjustment With regard to the implementation of the corresponding adjustment by the other contracting state, the question arises on the basis of which substantive and procedural provision it has to implement it.38 With regard to the question of the substantive regulation, in my opinion article 9 paragraph 2 OECD MC represents the legal basis for the corresponding adjustment. In this respect, article 9 paragraph 2 OECD MC has a self- executing effect. An additional legal basis in national law is therefore not required.39 Therefore, the required corresponding adjustment must be made even if the national tax law of the other state does not regulate a legal basis in this regard. If one insists on the existence of a domestic correction provision, a corresponding adjustment is left to the discretion of the contracting states not to create a correction provision in domestic law, contrary to the commitment entered into in a specific double taxation treaty. This is unacceptable in the light of the principles of international law. The objective of article 9 paragraph 2 OECD MC to avoid economic double taxation can only be effectively achieved if the provision itself is the authoritative correction provision. The OECD takes the same view in the Transfer Pricing Guidelines as a result.40 An obligation to make a corresponding adjustment by the other state only exists if an award from arbitration proceedings corresponding to article 25 paragraph 5 OECD MC is available. Further, the OECD states: Such arbitration shall provide for the appropriate adjustment. This is in line with the principle that, in principle, there is no obligation for tax administrations to reach an agreement. Based on Art. 9 para. 2 OECD [MC], a tax administration shall only make a corresponding adjustment if it considers the primary adjustment to be justified in terms of reason and amount. The non-mandatory nature of corresponding adjustments is necessary so that the tax administration of one state is not forced to accept the consequences of an arbitrary or unmotivated adjustment by the other state. This is also important for maintaining the fiscal sovereignty of each member country of the OECD.41

The OECD’s explanatory statement is not convincing. Double taxation agreements essentially consist of the restriction of a right of taxation of one of the contracting states in 38

Foley and Dhanuka, ‘Secondary Transfer-Pricing Adjustments’. Ditz in Schönfeld and Ditz, eds, Doppelbesteuerungsabkommen, 2nd ed. (Cologne: Otto Schmidt, 2019), art. 9 m.no. 148. 40 Para. 4.35 OECD TP Guidelines 2017. 41 Ibid. 39

Corresponding Adjustments 563 order to avoid double taxation. This restriction was agreed to by both contracting states and, by ratifying it into national law, they effectively deferred sovereignty.42 Furthermore, in this view, the provision in article 9 paragraph 2 OECD MC is superfluous, or at least not comprehensible. If this view were correct, the corresponding adjustment would have been regulated in article 25 OECD MC and not in article 9 paragraph 2 OECD MC. It is also argued in the literature that article 9 paragraph 2 OECD MC must be supplemented by the domestic correction of the profit rule.43 Otherwise, an arm’s- length standard could be applied in the domestic law of a contracting state, which would conflict with the provision in article 9 paragraph 2 OECD MC, which is comparable in the specific agreement. Only if there is a single arm’s-length standard—and this is also consistent with the dealing at arm’s-length principle of article 9 paragraph 1 OECD MC—is there a chance of arriving at uniform results under domestic and treaty law. Insofar as domestic law permits adjustments to profit that go beyond what is permitted by the dealing at arm’s-length principle, article 9 paragraph 1 OECD MC restricts the possibilities for adjustment under the conditions set out in this provision. What applies to article 9 paragraph 1 OECD MC in this respect must also apply accordingly to article 9 paragraph 2 OECD MC. This provision only opens the possibility for the contracting states of a specific double taxation treaty to apply their domestic rules on a versus adjustment. If the domestic tax law of one of the contracting states involved lacks a corresponding provision to adjust the profit, the corresponding adjustment to avoid economic double taxation can only be made on the basis of a mutual agreement procedure in which the other state agrees to make the corresponding adjustment for reasons of equity. Whether the taxpayer is entitled to an equitable decision is left to the respective domestic legal systems of the states involved. This view is to be followed insofar as a distinction is to be made between article 9 paragraph 1 OECD MC and article 9 paragraph 2 OECD MC. Since double taxation treaties are aimed at imposing barriers on the respective domestic tax law in the bilateral relationship, they have a direct effect on the respective tax law relationship. They are directly applicable after transposition into national law, and as such have a self-executing character.44 This character cannot be assumed for all the provisions in the OECD MC. There are provisions that are merely enabling standards (i.e. that still require a national statutory provision in order to be implemented). Whether a provision in the double taxation treaty has a self-executing effect is to be inferred from the wording and meaning of the individual provision.45 Formulations such as ‘may’ or ‘can’ speak for enabling 42

Kroppen and Rasch, Handbuch internationale Verrechnungspreise (Cologne: Otto Schmidt Verlag, 2020), OECD-Kapitel IV m.no. 144. 43 Wassermeyer and Baumhoff, Verrechnungspreise internationaler Unternehmen (Cologne: Otto Schmidt Verlag, 2014), 151; Schwenke and Greil in Wassermeyer, Doppelbesteuerungsabkommen, art. 9 m.no. 207; Jacobs, Endres, and Spengel Internationales Unternehmensbesteuerung, 8th ed. (Munich: C. H. Beck, 2016), 667. 44 Schaumburg and Häck, Internationales Steuerrecht, 759. 45 Ibid.

564 Matthias Hofacker norms, whereas formulations such as ‘are’ speak for directly applicable norms.46 Moreover, domestic applicability in such a way that the international treaty also creates rights and obligations for the individual beyond an obligation of the contracting states depends on the provisions of the double taxation treaty being specific enough (i.e. applicable). In German tax law, this corresponds to article 2 of the German Fiscal Code (Abgabenordnung). No self-executing effect can be assumed for article 9 paragraph 1 OECD MC. The wording of the provision already indicates the character of an enabling rule, because the corresponding profits ‘may’ be adjusted.47 Also the sense of the regulation demonstrates the necessity that article 9 paragraph 1 OECD MC cannot by itself represent an independent legal basis for a result correction, but rather presupposes corresponding regulations in the national tax law. This is because double taxation treaties do not create a tax liability or increase a tax above the domestic tax laws. Without a corresponding provision in national tax law, therefore, no domestic tax claim can be established or extended on the basis of article 9 paragraph 1 OECD MC. The provision in article 9 paragraph 2 OECD MC is to be understood differently. According to the wording, the other state is not authorized to make the corresponding adjustment. The other state is expressly required ‘so the other state shall make a corresponding adjustment to the tax imposed thereon on such profits’.48 This is consistent and underscores the intent of the provision. This is also how the OECD can be understood when it expressly assumes corresponding adjustments on the basis of article 9 paragraph 2 OECD MC and only refers to the domestic law of the other state in the absence of a corresponding provision in the specific agreement.49 Article 9 paragraph 2 OECD MC only triggers the obligation of the other state to make the corresponding adjustment if the initial correction is correct. In other words, the obligation to adjust the other state accordingly does not, of course, exist if, in the opinion of the other state, the initial adjustment is not compatible with article 9 paragraph 1 OECD MC in terms of reason and amount.50 Accordingly, in the case of a partial agreement between the contracting states, there is also only an obligation to make a partial adjustment in accordance with article 9 paragraph 2 OECD MC.51 The corresponding adjustment then relieves the taxpayer in the other state from avoiding economic double taxation of the tax object. In other words, the other state excludes from actual taxation any substrate to which it is not substantively entitled. It is not understandable why the other state, as a result of its obligation in a specific double

46 Ibid. 47

Aigner et al., Doppelbesteuerungsabkommen, 2nd ed. (Vienna: Linde, 2019), art. 9 m.no. 237.

48 Ibid.

49 OECD,

‘BEPS Action 14, 2015 Final Report: Making Dispute Resolution Mechanisms More Effective’ (2015), para. 11. 50 Art. 9 no. 6 OECD Commentary; para. 4.35 OECD TP Guidelines 2017; Aigner et al., Doppelbesteuerungsabkommen, art. 9 m.no. 230. 51 Ibid.

Corresponding Adjustments 565 taxation agreement in accordance with article 9 paragraph 2 OECD MC, should regulate in its national tax law not to tax something to which it has no right. Only in formal terms, must a correction provision be regulated in national law if the respective national legal system is divided between formal and substantive law. If such a separation is foreseen, there must be regulations in order to be able to break through the regularly occurring validity of the affected tax assessments.

31.3 Method of Corresponding Adjustment Article 9 paragraph 2 OECD MC does not specify a particular method for making an adjustment.52 The procedural treatment of the corresponding adjustment is governed by the adjustment rules provided for in the domestic tax laws of the contracting states if the contracting states do not agree on another method of adjustment within the framework of a mutual agreement procedure within the meaning of article 25 OECD MC. The OECD Commentary identifies two options for such an adjustment.53 On the one hand, the assessment of the company can be changed and thereby the reduction of the taxable profit can be achieved or the corresponding adjustment can be made by a corresponding application of article 23A or 23B OECD MC. The first variant leads to the fact that it is a mirror-image measure of the initial adjustment and differences in the amount of tax in both states have no influence.54 The OECD envisions this method as the usual one.55 It may be relevant whether the corrective measures are to be taken within the tax balance sheet according to domestic law or outside the tax balance sheet when determining the taxable income of the company concerned. According to accounting principles, corrections within the tax balance sheet can have either a negative or a positive effect in later assessment periods due to reversal effects. It is also considered difficult if there is no domestic norm for the corresponding adaptation and if one rejects the self-executing character of article 9 paragraph 2 OECD MC. In these cases, a tax remission or tax reduction for reasons of equity would have to be examined in the contracting state concerned. With regard to the second possibility of basing the corresponding adjustment on a corresponding application of article 23A or 23B OECD MC, it should be noted that only a corresponding application of these provisions is conceivable. Both provisions presuppose subject identity; moreover, they only cover cases in which the income ‘may be taxed under this convention’ in the other contracting state. However, this is not the case

52

Schwenke and Greil in Wassermeyer, Doppelbesteuerungsabkommen, art. 9 m.no. 29. Art. 9 no. 7 OECD Commentary. 54 Eigelshoven, Doppelbesteuerungsabkommen, 6th ed. (Munich: C. H. Beck, 2015), art. 9 m.no. 177. 55 Para. 4.34 OECD TP Guidelines 2017. 53

566 Matthias Hofacker for the profits of an enterprise in a contracting state that, according to article 7 paragraph 1 sentence 1 OECD MC, can only be taxed in that state.56

31.4 Mutual Agreement Procedure The dispute-resolution mechanism described in article 25 OECD MC is a central and indispensable part of the obligations undertaken by a contracting state when concluding a double taxation treaty. Article 25 OECD MC must be fully implemented in good faith, in accordance with its provisions and taking into account the object and purpose of a double taxation treaty. The OECD aims at the full implementation of agreement commitments regarding the mutual agreement procedure and expects that cases to be settled by mutual agreement procedures will be ensured expeditiously.57 As a minimum standard, states have defined that paragraphs 1–3 of article 25 OECD MC shall be included in their double taxation treaties in accordance with the interpretation in the OECD Commentary and subject to the derogations provided for in points 3.1 and 3.3 of the minimum standard. They should allow the use of the mutual agreement procedure for transfer pricing cases and implement the resulting mutual agreements (e.g. by adjusting the assessed tax accordingly). Therefore, the OECD opposes states denying companies access to the mutual agreement procedure after a transfer pricing adjustment in the other state.58 States should publish rules, guidelines, and procedures for initiating and conducting the mutual agreement procedure and take reasonable steps to make this information available to taxpayers. States should ensure that their guidelines for plea-bargaining procedures are descriptive and readily available to the public.59 Article 9 paragraph 2 OECD MC is applicable according to its wording without the prior implementation of a mutual agreement procedure. On the other hand, article 9 paragraph 2 OECD MC does not prohibit the implementation of a mutual agreement procedure, consultations, or a formal agreement by the contracting states. However, the corresponding adjustment according to article 9 paragraph 2 OECD MC presupposes that there is an agreement between both contracting states on the qualification of the performance payment (e.g. as licence fee, dividend, interest, etc.) and on its appropriate amount.60 Although the initial adjustment forms the standard for the scope of the corresponding adjustment, this initial adjustment is readily possible by one contracting

56 Eigelshoven, Doppelbesteuerungsabkommen, art. 9 m.no. 177a.

57 OECD, Improving the Efficiency of Dispute Resolution Mechanisms, Action Item 14—Final Report 2015, OECD/G20 Profit Shifting and Shortage Project (Paris: OECD Publishing, 2018), http://dx.doi.org/ 10.1787/9789264190122-de, 13 I.A.1, Rz. 9. 58 Para. 4.43 OECD TP Guidelines 2017. 59 Para. 4.44 OECD TP Guidelines 2017. 60 Schaumburg and Häck, Internationales Steuerrecht, 881.

Corresponding Adjustments 567 state, but this alone does not automatically give rise to an adjustment obligation in the course of a corresponding adjustment for the other contracting state.61 The agreement between the contracting states on the reason for the initial adjustment presupposes that both contracting states agree on a set of facts justifying the entitlement to profits.62 On the other hand, it is not necessary for the contracting state called upon to make the corresponding adjustment to follow the classification of the facts under one of the domestic legal bases of the contracting state making the first adjustment. The only thing that matters is that, in the opinion of both contracting states, the entitlement to profits was carried out in accordance with article 9 paragraph 1 OECD MC.63 In this context, particular attention should be paid to the fact that the matching qualification of performance pay is an important prerequisite for the comparability of initial and matching adjustments. The agreement on the amount of the adjustment requires both states to reach the same conclusions in determining the appropriate (i.e. customary) fee for services among independent third parties. This includes the consistent application of benefit-sharing principles.64 If, on the other hand, the other contracting state assumes that the initial reporting does not comply with the arm’s-length principle, article 9 paragraph 2 sentence 2 OECD MC provides for the possibility for the competent authorities of the contracting states to consult each other. The mutual agreement procedure is a proven means by which competent authorities negotiate to resolve disputes and eliminate double taxation from transfer pricing adjustments.65 The consultation is a mutual agreement procedure according to article 25 OECD MC between the contracting states involved.66 Even before the inclusion of article 9 paragraph 2 OECD MC in the 1977 OECD MC, it was assumed that the procedure could also be carried out for the elimination of economic double taxation, which was derived from the Commentary on article 25 OECD MC.67 However, the mutual agreement procedure is not a mandatory requirement for the contracting states to reach the agreement. Rather, it is a process for reaching an agreement, to be found on the basis of substantive treaty and domestic law.68 It is natural for states to provide for a compulsory arbitration procedure69 in the specific double taxation agreement or to apply the EU Arbitration Convention.70

61

Eigelshoven in Vogel and Lehner, Doppelbesteuerungsabkommen, art. 9 m.no. 164b. Art. 9 no. 6 OECD Commentary; Ditz in Schönfeld and Ditz, Doppelbesteuerungsabkommen, art 9 m.no. 146. 63 Ditz, ibid., m.no. 143. 64 Ibid., m.no. 145. 65 Para. 4.29 OECD TP Guidelines. 2017. 66 Para. 4.33 OECD TP Guidelines; art. 25 nos 10ff OECD Commentary. 67 Art. 25 nos 11 and 12 OECD Commentary. 68 Art. 25 nos 72ff OECD Commentary; OECD, ‘Transfer Pricing, Corresponding Adjustment and the Mutual Agreement Procedure’ (1982). 69 Art. 25 para. 5 OECD MC 2008. 70 Convention No. 90/ 436/EEC on the elimination of double taxation in connection with the adjustment of profits of associated enterprises of 23 July 1990. 62

568 Matthias Hofacker Since the bilateral adjustment of profits clauses permits a first adjustment and thus an increase in profits by one contracting state without regard to a second adjustment (reduction in profits) by the other contracting state, in international taxation practice, the first adjustment usually remains. The reasons for this are, on the one hand, that many double taxation treaties do not provide for adjustments,71 there is no obligation for consultations and mutual agreement procedures, and, finally, as far as such consultations and mutual agreement procedures are carried out, the compulsory agreement of the involved contracting states is not provided for.72 Thus, the bilateral correction of the profit clause is consistently to the detriment of companies domiciled in both contracting states, with the result that internationally operating corporations in particular are exposed to economic double taxation without effective protection, which cannot be avoided even through advanced pricing agreements (APAs). Effective protection can therefore only be achieved by means of an arbitration clause anchored in the double taxation agreement. (Economic) double taxation can then still be avoided through the EU Arbitration Convention, which provides for a mutual agreement procedure as the first stage and an arbitration procedure as the second stage.

31.5 Arbitration Convention A corresponding profit adjustment has been achieved with the so-called Arbitration Convention, at least within the EU. If the profits of a company are simultaneously included in the profits of a company in the other contracting state on the basis of a profit adjustment in accordance with article 9 paragraph 1 OECD MC, a three-stage procedure is initiated (art. 4 no. 1 Arbitration Convention), which leads to a binding agreement on a mandatory basis. Comparable to this, the OECD has also anchored mandatory arbitration in the text of the agreement (art. 25 para. 5 OECD MC). According to this provision, a case that has been unsuccessfully tried by the competent authorities in a mutual agreement procedure shall be submitted to an arbitration court, which shall decide the case in a binding manner. In contrast to the provision in the Arbitration Convention, taxpayers cannot directly invoke the OECD MC. The prerequisite here is that the arbitration procedure has found its way into the text of the specific double taxation agreement.

31.6 The Secondary Correction The OECD distinguishes the ‘secondary adjustment’ from the corresponding adjustment according to article 9 paragraph 2 OECD MC.73 It is not regulated in article 71

Eigelshoven in Vogel and Lehner, Doppelbesteuerungsabkommen, art. 9 m.no. 145. Schaumburg and Häck, Internationales Steuerrecht, 882. 73 Art. 9 no. 7 OECD Commentary; IFA, Adjustments 19b (1996), Rz. 1. 72

Corresponding Adjustments 569 9 OECD MC, but if a secondary adjustment is made, a tax administration is thereby bound by the wording of article 9 paragraph 2 sentence 2 OECD MC ‘in making this change, the other provisions of the Agreement shall be taken into account’ with respect to the manner of its implementation to the substantive agreement law. By way of initial adjustment (as defined in art. 9 para. 1 OECD MC), the profit share shifted in favour of a company is taxed at the company to which it is attributable according to the arm’s-length principle. The corresponding adjustment results in a mirror-image relief for the other company of the tax with which it was burdened only as a consequence of the arm’s-length transaction. Both measures are limited to the fictitious allocation of the profit without necessarily reversing the benefit actually granted. The amounts on which the taxation was based as adjusted profits are therefore generally still held by the enterprise of the contracting state that made the corresponding adjustment. The secondary profit adjustment pursues the goal of taxing those amounts in accordance with the treaty.74 If the transferred profit is deemed to be a hidden profit distribution (i.e. a ‘dividend’) within the meaning of the double taxation agreement, the contracting state that made the initial adjustment in accordance with article 9 paragraph 1 OECD MC may levy a dividend tax in addition to the corporate income tax on the profit used for that distribution. The other contracting state may also subject these dividends to its taxation with credit for the dividend tax, unless it has to grant a nesting privilege in accordance with the double taxation agreement.75 However, the OECD implies that withholding taxes on dividends should be avoided whenever possible. Cases where the taxpayer acted intentionally to avoid withholding taxes are excluded.76 There is no room for a secondary profit adjustment if the company actually holding the benefit returns it to the company subject to the initial adjustment (e.g. by posting a receivable), on consideration of expense in a secondary adjustment. The OECD points out that an off-balance sheet adjustment due to inappropriate transfer prices between sister companies may also have consequential effects on the common parent company. In Germany, an adjustment of profit due to inappropriate transfer pricing at a subsidiary triggers a hidden profit distribution via the chain to the common parent company. The parent then makes hidden contributions to the other sister company involved in the transaction.77 The example makes it clear that consequential effects arise for companies in countries that are not involved in the transaction. The OECD points out that states are encouraged to limit the impact of secondary adjustments.78 Some states also apply the concept of a disguised loan. The state making the initial authorization therefore assumes that the company has made a loan to the company in the other state in the amount of the arm’s-length remuneration. Some states continue to assume that interest must then be paid on that loan accordingly. However, this

74

Para. 4.67 OECD TP Guidelines 2017. Para. 4.69 OECD TP Guidelines 2017. 76 Para. 4.72 OECD TP Guidelines 2017. 77 Para. 4.71 OECD TP Guidelines 2017. 78 Para. 4.72 OECD TP Guidelines 2017. 75

570 Matthias Hofacker assumption only makes sense from an economic point of view if one assumes that the company would have reinvested the supposedly increased profit at interest and that a profit distribution in favour of the shareholder would not have been made. The OECD takes a sceptical view of the acceptance of hidden loans, since the acceptance of a loan with an accompanying interest payment leads to further consequential problems, which would in turn be the subject of article 9 OECD MC.79

79

Paras 4.68 and 4.71 OECD TP Guidelines 2017.

Chapter 32

Transfer Pri c i ng versus For mu l a ry App ortionme nt Georgios Matsos

32.1 Introduction Although formulary apportionment (FA) has been a relatively old operational notion in subnational corporate taxation, especially in the USA1 and Canada,2 it was not generally used in international taxation, other than in theoretical exercises. This is bound to change, as FA is currently the basis of major reform proposals aiming to reregulate cross-border allocation of taxing powers. The EU was the first to launch twice, in 2011 and again in 2016, a widely discussed formal proposal for a common consolidated corporate tax base (CCCTB), which would allocate profits among its member states not according to the usual residence/source scheme, but according to a formula granting equal weight to three allocation factors, namely sales, (tangible) assets, and payroll/number of employees.3 Only a few years later, the EU was followed on a global level by the OECD’s Inclusive Framework, which

1 For a thorough presentation of the formulary apportionment in the USA, see Ch. McLure Jr, ‘Understanding Uniformity and Diversity in State Corporate Income Taxes’, National Tax Journal (2008), 141. 2 For the Canadian formulary apportionment, see M. Smart and F. Vaillancourt, ‘Formulary Apportionment in Canada and Taxation of Corporate Income in 2019: Current Practice, Origins and Evaluation’, in R. Krever and V. Vaillancourt, eds, The Allocation of Multinational Business Income: Reassessing the Formulary Apportionment Option (Alphen aan den Rijn: Wolters Kluwer, 2020), 67. 3 See art. 28 of Proposal of the European Commission for a Council Directive on a Common Consolidated Corporate Tax Base, COM(2016) 683 final (25 October 2016), and art. 86 of the Proposal for a Council Directive on a CCCTB, COM(2011) 121/4 (16 March, 2011).

572 Georgios Matsos suggested in Pillar One a FA of what is defined there as ‘residual profits’ of either certain (‘Amount A’) or quasi all (‘Amount B’) multinational enterprises (MNEs).4 As of 16 December 2022, the suggestion had been adopted by 138 jurisdictions, including all OECD and G20 members. As the title of this chapter itself reveals, FA is usually juxtaposed with transfer pricing and with the arm’s-length principle (ALP). However, the idea of a full or even partial FA goes much further than simply replacing transfer pricing and the ALP: FA challenges the very essence of the allocation rules of the current double taxation convention (DTC) network (i.e. source and residence as allocation principles). While formally the constitutive elements of a formula are no nexus, FA modifies the consequences of source and residence in a way that effectively replaces the means used by the source/residence principles and by the DTCs in order to allocate profits. This is done by substituting the world income principle and the attraction force of residence, by profoundly altering the rules deriving from the source principle and, finally, by rendering current DTCs obsolete.5 To give an example: an MNE’s presence in a jurisdiction which consists solely of an R&D department—either in the form of a subsidiary or in the form of a PE—which conducts only basic research exclusively for the rest of the group and not producing marketable know-how, would likely not be expected to create taxable profits as an independent entity. Yet, the CCCTB labour criterion would directly allocate a significant portion of the MNE’s profits to the jurisdiction hosting such a non-market-oriented R&D department. Such a result would not be directly possible either under the source or under the residence rules and it would in principle not fall within any of the current DTC allocation rules. However, the same example illustrates why FA is usually juxtaposed with transfer pricing: According to the ALP, the MNE would have to find a correct profit to allocate to such an R&D department, whether it is a subsidiary or a permanent establishment (PE). Profit could be calculated, for example, according to the price that the MNE would alternatively have to pay to an unrelated third party in order to conduct such basic research. This illustrates why, although formal allocation rules such as source and residence would not directly allocate income to the R&D department in question, as a matter of business and tax practice the ALP would trigger such allocation. There will undoubtedly be situations in which FA such as the one in the CCCTB Proposal would still function like nexus (e.g. one could imagine the example of a temporarily unused asset, which an 4 See OECD, Tax Challenges Arising from Digitalisation—Report on Pillar One Blueprint: Inclusive Framework on BEPS, OECD/G20 Base Erosion and Profit Shifting Project (Paris: OECD Publishing, 2020), https://doi.org/10.1787/beba0634-en (accessed 17 July 2022). For a detailed analysis of Pillar One, see ‘The Allocation of Taxing Rights under Pillar One of the OECD Proposal’ in this volume. 5 Existing DTCs are planned to be significantly modified en bloc by a set of new provisions to be added to the Multilateral Convention (MLC). Hence, new nexus rules will apply for enterprises falling within the scope of Pillar One. A draft MLC amendment on such nexus rules was published on 4 February 2022 (https://www.oecd.org/tax/beps/public-consultation-document-pillar-one-amount-a- nexus-revenue-sourcing.pdf (accessed 17 July 2022)).

Transfer Pricing versus Formulary Apportionment 573 MNE has acquired in a country where is has no activity of its own), simply as a means of preventing a competitor from taking advantage of the asset. The major practical consequences of this would be a profound change of approach that MNEs apply today in everyday business life with respect to transfer pricing and the ALP. In any case and despite the practical importance of transfer pricing for the day-to- day allocation of profits, FA as an international profit allocation rule goes much further than simply replacing transfer pricing and the ALP: it tends to replace the whole system of international taxation as it has been known since the early twentieth century. A fully applicable FA would mean that source and residence individually would have no significance for profit allocation, while the very concept of current DTCs would become completely outdated. In a world dominated by FA, rules such as CFC legislation, exit taxation, thin cap rules, and the full Base Erosion and Profit Shifting (BEPS) Project would become obsolete. FA would thus become a trend for a fundamental and wholesale reform of international taxation.

32.2 Institutional Background Formulary apportionment is the brain of federal jurisdictions, whose federative entities impose their own corporate tax on corporate profits, responding to the necessity to delimit the taxing powers of each of their sub-federal taxing authorities and, possibly, to avoid overlap and/or gaps between subnational tax bases. FA was therefore developed in federal jurisdictions whose federative entities sought to draw tax boundaries between each other and thus regulate their tax competition.6 While sovereign countries had the classical tool of international Double Tax Conventions in order to set boundaries between them, sub-federal entities had to rely either on federal legislation7 or on their own—more or less—free will8 for allocating profits among them.9 As such, FA was the result of a need to find an immediate practicable solution to the obvious necessity of allocating business profits for the purposes of subnational corporate taxation.

6 By contrast, Switzerland has an old culture of open tax competition among its cantons, so that it did not have to develop any FA concept as a means to regulate and possibly reduce such competition. For details on the Swiss intercantonal tax allocation system—a system heavily based on residence—see I. Mitroyanni, Integration Approaches to Group Taxation in the European Internal Market (Alphen aan den Rijn: Wolters Kluwer, 2008), 68. 7 This is the case in Canada, where, however, the application of said federal legislation is not obligatory; for details, see Smart and Vaillancourt, ibid., 75. 8 This is the case in the USA; for details, see Ch. McLure Jr, ‘Understanding Diversity and Uniformity’, National Tax Journal LXI/1 (2008), 145. 9 In the USA, it is possible for states to conclude interstate compacts with each other, which require the approval of the Congress; however, the formula used by US states for subnational allocation of profits never formed part of a legally binding compact. For details, see McLure Jr, ibid., 151.

574 Georgios Matsos The coexistence of sub-federal entities in a federal scheme resulted not only in the necessity for FA but also allowed it: sub-federal entities had a level of integration with each other that reduced their internal antagonisms to a level that FA could still operate, even in an imperfect way: US states had not made efforts to coordinate their formulas until threatened with federal legislation to that end in 1964,10 ultimately widely adopting the so-called Massachusetts formula, which was equally weighing sales, payroll, and property as apportionment criteria. Soon thereafter, however, they started abandoning it following the US Supreme Court sanctioning the use of different formulas in each state11 in 1978 and in particular the sole use of the sales criterion. The latter gradually became either the main or the exclusive apportionment factor in most US states.12 Canada, too, did not impose a formula, despite federal legislation being in place, but rather provided a federal model legislation which provinces were free not to adopt.13 Thus, history proves that FA has been neither strictly imposed, nor perfect in its application. It was a natural development, in a way similarly natural to the (different) tools of the ALP and of the DTCs. If added to the first two differences mentioned earlier (lack of the DTCs tool and genuine federalism), imperfectionism is a third significant difference of the current international pursuit towards a strict, obligatory, and perfectionist formula that has been put in place by supranational bodies such as the EU and the OECD. The imperfect world of the FA seems in its current actual implementation even more imperfect than the world of transfer pricing.14 Notwithstanding the current imperfections, it would not be a choice for, for example, the EU to achieve FA without an extensively

10

See McLure Jr, ibid., 147. US Supreme Court, Moorman Manufacturing Co. v. Bair, 437 US 267 (1978). The US Supreme Court implicitly accepted the necessity of a uniform formula, but rejected the notion that it should be the work of the judiciary to impose such a uniform formula. See at 278 and 280 of the judgment. 12 See I. Mitroyanni, Integration Approaches to Group Taxation in the European Internal Market, 74 and 186. 13 Smart and Vaillancourt, ‘Formulary Apportionment in Canada’, 75. 14 In the USA, the only obligatory rule to be taken into account is the one enacted by the US Supreme Court in 1983 in Container Corp., 463 US at 169–70: 11

Such an apportionment formula must, under both the Due Process and Commerce Clauses, be fair . . . The first, and again obvious, component of fairness in an apportionment formula is what might be called internal consistency—that is, the formula must be such that, if applied by every jurisdiction, it would result in no more than all of the unitary business’ income’s being taxed. The second and more difficult requirement is what might be called external consistency—the factor or factors used in the apportionment formula must actually reflect a reasonable sense of how income is generated. The Constitution does not ‘invalidat[e]‌an apportionment formula whenever it may result in taxation of some income that did not have its source in the taxing State . . .’ . . . Nevertheless, we will strike down the application of an apportionment formula if the taxpayer can prove ‘by “clear and cogent evidence” that the income attributed to the State is, in fact, “out of all appropriate proportions to the business transacted . . . in that State” . . . or has “led to a grossly distorted result” . . . ’.

Transfer Pricing versus Formulary Apportionment 575 detailed, uniform, and above all obligatory set of rules allocating corporate profits to each member state in an accurate way.15 The imperfect nature of the FA subnational world, especially in the USA, is not only due to the limited cultural differences among US states, but also to the fact that US companies pay only a relatively small part of their corporate tax to these sub-federal taxing jurisdictions.16 For that reason, companies have more limited financial grounds to complain about double taxation,17 which could, in any event, often be compensated by double non-taxation. Imperfect FA has been a situation generally tolerated both by US states and by US taxpayers. In contrast, both CCCTB and Pillar One clearly aim for a perfectionist formula. What is peculiar in this perfectionist pursuit, is that such perfectionism is sought in order to replace the imperfect allocation currently in place through transfer pricing and ALP: no one is prepared to reduce uniformity and clarity of rules below their current level. One might wonder, however, whether an adequate integrational culture higher than that of the US states could be found elsewhere. Even the EU member states, despite lengthy harmonization efforts, continue until today to pursue very different policy goals with their income and corporate taxation systems. In a global environment with more divergence than exists within the EU, individual countries would be even less willing to tolerate a profit allocation scheme that is in practice more imperfect than the current one. The FA approach and the ALP/transfer pricing approach have naturally developed in different institutional and cultural environments. Both of them are imperfect in the way they are respectively applied today. The international tax community must bear in mind both these parameters before FA is labelled as a desired perfectionist system. Tax imperfectionism, as natural as it has been until today, might actually be a necessary evil.

32.3 Net vs Gross Income Apportionment Transfer pricing is criticized by supporters of FA on the basis that it cannot create parity between market and non-market transactions.18 Though this critique is in principle 15 The 2016 Proposal of the European Commission for a CCCTB, contains for the time being eighty- two articles in thirty-three pages. 16 For 2022, US states impose corporate tax rates amounting from 0% to 11.5%, while the federal rate is 21%. See the detailed presentation in: https://taxfoundation.org/publications/state-corporate-income- tax-rates-and-brackets/ (accessed 17 July 2022). 17 See S. Sicilian and J. Huddleston, ‘The US States’ Experience with Formulary Apportionment’, in Krever and Vaillancourt, ibid., 45, who point out that multiple taxation arises because the condition set by the US Supreme Court in Container Corp. can exclude double or multiple taxation only if the same formula enjoys uniform application across the USA. 18 See Y. Brauner, ‘Between Arm’s Length and Formulary Apportionment’, in Krever and Vaillancourt, ibid., 198.

576 Georgios Matsos correct, the critical question for deciding between ALP and FA is a fundamentally different one: whether profits should be apportioned according to the net profits criterion or rather according to sales or gross income criteria. Transfer pricing rules try—often unsuccessfully—to assimilate market transactions with non-market transactions. Nonetheless, the purpose of such pursuit is not to achieve parity between these two different types of transaction, but to calculate the net profits that are to be allocated in each jurisdiction. Net profits are a complicated mixture of the ups and downs in assets and liabilities over a fiscal year. Should this happen on a jurisdictional basis or rather on a unitary basis? The answer given by the international tax system currently in place is that net profits should be calculated on a jurisdictional basis. But the main argument against this type of jurisdictional calculation is that it is highly manipulable. What is the alternative offered by FA? There is no answer, unless the formula is known. While the tradition set by the Massachusetts formula (equal weight of sales, payroll, and assets) continued to shape the ideas for a FA at least until the emergence of the CCCTB in 2011 and 2016,19 both real-life evolution in the USA as well as the evolution of international plans under Pillar One clearly show that formulas used for FA cannot withstand the force of using the sole sales criterion. Indeed, US states increasingly started using the sole sales criterion as soon as they realized that they had the freedom to do so.20 However, even Pillar One deviates significantly from the gratifying mathematical formula of the CCCTB,21 and ultimately uses the sales criterion. And—most importantly— not simply sales, but mainly the place of destination of sales (i.e. where the consumer is located). The use of the sole sales criterion as the allocation criterion is clear in the formulas of both Amount A and Amount B of Pillar One. For Amount A, the allocation criterion is clearly defined as the in-scope revenue derived from each eligible market jurisdiction as sole allocation criterion,22 while sourcing rules allocate 50% of revenue from automated digital services (ADS) to the jurisdiction where the seller is based only in the case of intermediation of tangible23 or intangible24 goods and services through ADS. All other ADS revenue is allocated solely according to the place of the final consumer.25 The full revenue from consumer-facing services is allocated with no obvious exception to the 19

See Section 32.2. Ibid. 21 Art. 28 of the 2016 EU Proposal and art. 86 of the 2011 EU Proposal. 22 See Report on Pillar One Blueprint, paras 514–516. 23 Ibid., paras 251 and 256. 24 Ibid., para. 261. 25 In the case of Automated Digital Services, this includes online advertising services, with sole allocation criterion the location or the residence of the viewer of services (ibid., paras 238 and 240), sale or other alienation of user data, with sole criterion the real-time location or the residence of the user that is the subject of the data being transmitted—i.e. again of the ‘consumer’ who generates such data—(ibid., paras 243 and 245), digital content services, with only criterion the ordinary residence of the purchaser (ibid., para. 267), cloud computing services, with sole criterion the ordinary residence of the individual purchaser or the place where the service is used for business customers (ibid., paras 273 and 275). 20

Transfer Pricing versus Formulary Apportionment 577 jurisdiction of final delivery26. In the OECD ‘Draft Model Rules for Nexus and Revenue Sourcing under Pillar One—Amount A’ of 4 February, 2022, all revenue is deemed to be sourced in the jurisdiction of the customer, except online intermediation services, for which half of the revenue shall be deemed to be sourced in the jurisdiction of the seller or of the service provider.27 Equally important is that that the Amount B of Pillar One is by definition a rule allocating profits to the place where distributors are based—that is, to the jurisdiction of the consumer. Pillar One authors claim, indeed, that Amount B ‘is intended to approximate results determined in accordance with the ALP’28 and ‘in a manner that is aligned with the ALP’,29 by explicitly comparing its methodology to the Transactional Net Margin (TNM) method. However, the ALP is not one single method. It is precisely the use of multiple methods making the ALP and transfer pricing no exact science, but rather a permanent attempt to approximate market conditions. On the contrary, Amount B does not attempt to approximate market conditions and chooses only one method out of many. The chosen method is by no coincidence precisely a formula: from the sales (i.e. the revenue) that resident companies or PEs perform in the state of residence or of PE location,30 Pillar One suggests using a return on sales ‘as a fixed return for the transactions in scope’31 in order to define the appropriate profit level. Despite suggested or discussed variations of that central rule in the Pillar One report, the rule itself is nothing else but a formula and there is no reason to deny it. As a result, both components of Pillar One tend to evolve the basic framework of international taxation towards a sales-based FA method and, in fact, towards a destination- based sales formula. According to this concept, individual jurisdictions will not be able to calculate locally taxed profits as net profits but will have to tax part of global net profits of MNEs along with the gross revenue generated in principle through local consumers. Would this make corporate income taxation evolve towards becoming a sales tax or a turnover tax? Though the current answer is in principle still ‘no’, it should not escape attention that turnover and sales taxes have always been less easy to manipulate. It is therefore somewhat natural that governments might at some point attempt, if not to convert corporate taxation to turnover taxation, to at least make its interjurisdictional allocation dependent upon locally generated turnover. And although in theory (unitary) profit would still be the net profit, the taxable base will not be determined as net profit for each local jurisdiction, but as an extensively foreign notion (global profit), 26

This is the explicit rule when delivery of goods occurs through such services (ibid., paras 277 and 279), while revenue from Consumer-Facing Services is allocated to the place of ‘enjoyment or of use of the service’ (ibid., para. 281). Similarly allocated is revenue from the franchising or the licensing of goods or services (ibid., paras 283 and 285). 27 https://www.oecd.org/tax/beps/public-consultation-document-pillar-one-amount-a-nexus-reve nue-sourcing.pdf, 5–7 (accessed 17 July 2022). 28 Ibid., para. 14. 29 Ibid., para. 649. 30 Ibid., paras 659 and 660. 31 Ibid., para. 686.

578 Georgios Matsos which tends to be apportioned according to local turnover. For the recipient of tax (i.e. for each government), the local turnover will be of higher importance than (unitary) profit because only the former will fall under its full authority and control. This fundamental change from the pursuit of a net income to the pursuit of a gross income related to a global unitary profit, which would usually be for the biggest part separated from the recipient of tax, is not free of problems. The facts alone that Pillar One Blueprints devote thirty-one dense pages to tax certainty32 and that the 2022 OECD draft ‘Pillar One—A Tax Certainty Framework for Amount A’ amounts eighty-seven pages,33 show that no certainty exists regarding the application of this radically new approach in international taxation. In reality, such certainty is still being sought. As a conclusion, differentiation between ALP and FA is indeed not the achievement of parity between market or non-market conditions, but rather the focus on net versus gross country income. Both gross and net income are in principle market figures, but it is ultimately the figure of expenses that is less taken into account under a formula tending towards destination-based sales.

32.4 Arm’s-L ength Range: An Unacceptable Bias? FA is generally regarded as being more exact and allowing less leeway for manipulation of profits than transfer pricing.34 Though this view is in principle correct, the critical question is, however, a different one: is the avoidance of manipulation of profits the only or the main criterion to be taken into account when allocating business profits to the individual companies of an MNE group? As has been analysed earlier, the objective of transfer pricing is to determine the net profit of an individual company on an individual basis with no reference to a unitary group net profit and to apportionment factors such as gross revenue, payroll, and assets. To achieve this with an acceptable level of tax certainty and to the extent that transfer pricing is indeed not an ‘exact science’,35 the OECD transfer pricing guidelines have recognized an acceptable ‘arm’s-length range’, which critics of the ALP consider as an 32

Ibid., 168–198, paras 703–804 or around 12% of the Report. at https://www.oecd.org/tax/beps/public-consultation-document-pillar-one-amount-a- tax-certainty-framework.pdf (accessed 17 July 2022). Next biggest OECD public consultation document is the one on nexus (ibid., cf. with earlier), amounting to only thirty-six pages! 34 R. S. Avi-Yonah and I. Benshalom, ‘Formulary Apportionment—Myths and Prospects’), University of Michigan Law & Economics, Empirical Legal Studies Center Paper No. 10-029, (16 Oct. 2010), University of Michigan Public Law Working Paper No. 221 (2010), 14, SSRN: https://ssrn.com/abstract= 1693105 (accessed 17 July 2022). 35 See paras 1.13, 3.55 and 4.8, OECD, OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations (Paris: OECD Publishing, 2022), https://doi.org/10.1787/0e655865-en (accessed 17 July 2022). 33 Available

Transfer Pricing versus Formulary Apportionment 579 unacceptable tax bias in favour of MNEs,36 in the sense that the taxpayer would tend to choose an acceptable price that would offer the highest tax advantage. This view regards tax as an instrument solely devoted to replenishment of the government treasury. Certainly, taxes are the main source of government revenue, and this is the origin of their existence but, once taxes exist, their function in the economy is considerably more far-reaching than simply providing government revenue. They are, in effect, the basic tool of economic policy. Environment and innovation are, for example, just two factors influencing government decision making to increase or decrease tax revenue, in order to achieve policy goals. Within this framework, a business-friendly tax system is a major tool for achieving economic development. This goes far beyond what is usually understood as tax certainty. Consequently, the true question is whether it is ultimately better for governments to pursue every last penny of taxes that could escape through manipulation or, rather, to give businesses some leeway and allow them to become more productive. The arm’s-length range should not be understood as a rule allowing any transfer pricing method that can accommodate tax avoidance. On the contrary, although the OECD guidelines state that any price within the arm’s-length range is acceptable,37 the arm’s-length range is nevertheless allowed only ‘for difficult cases, where no one approach is conclusive’. In such cases, ‘a flexible approach would allow the evidence of various methods to be used in conjunction’.38 By adhering to the logic of an acceptable range of prices for ‘difficult cases’, transfer pricing simply adapts to the countless variations of business reality, which cannot always be precisely identified through legislative or administrative standardization. The policy options are either to accept such imperfectionism whilst allowing at the same time some room for manipulation of profit allocation, or to completely reject imperfectionism and let the regulatory framework evolve towards over-standardization. But does formulary apportionment resolve the problem of imperfectionism in ‘correctly’ allocating profits? As a matter of fact, a global application of FA would do exactly the opposite: the institutional consolidation of imperfectionism. While ALP seeks an acceptable range of prices instead of an exact price only in cases of difficult-to-assess non-market transactions, FA treats even market transactions as non-market ones and thus introduces the natural arbitrariness of a one-size-fits-all formula even in cases where the ALP leaves no real leeway for manipulation. Moreover, it is exactly through the tool of the arm’s-length range that the ALP can adapt to economic and financial reality, whilst even allowing—under the classical in dubio pro libertate concept—some tax bias in favour of the taxpayer. A formula, on the contrary, is incapable of adapting to any economic or financial reality. Currently, the global tax community possesses no practice and no experience of how to apportion profits in ways other than the adjustment of local net profits according to the ALP. Tax 36

See Brauner, Between Arm’s Length and Formulary Apportionment, 199. OECD TP Guidelines (2022), para. 3.60. 38 Ibid., para. 2.12. 37

580 Georgios Matsos administrations outline the ALP either in advance through APAs or ex post through transfer pricing audits. It is so far impossible to recognize and assess all potential problems that a non-flexible apportionment formula might generate, before FA is really applied.39 Since the initiation of the BEPS Project, the tax world has been dominated by a pursuit to secure government tax revenue. While this is certainly a legitimate goal, it is not the only legitimate goal of international tax law. If the use of an inflexible formula produced results unfair for the taxpayer, then no ways would be available in order to correct such unfairness. But is it ultimately better to allow unfairness towards government revenue or towards taxpayers? Taking into account that the principle is in dubio pro libertate and not in dubio pro administratione, one might wonder how the aim of securing government revenue has become the ultimate principle in international taxation in recent years. The same remarks apply in the case of intangibles, which are usually most difficult to assess under the ALP. While FA is supposed to address the major concern of the valuation of intangibles by making it simply irrelevant for the allocation of taxing powers, the ALP still allows some manipulation of profits under the rule of arm’s-length range. Is such manipulation—which is in any event extremely limited under the new transfer pricing rules introduced especially through the BEPS Project40—an evil of such proportion that its desired elimination necessitates reform of the whole system of international taxation? One should take into account that intangible-heavy companies are by definition innovative companies which produce breakthrough technology in their industries and that innovation is usually acknowledged as eligible to enjoy preferential tax treatment. Some limited leeway for manipulation of technology profits would ultimately constitute a preferential tax treatment that is not in principle undesirable. The ALP thus offers, through the rule of the arm’s-length range, a flexibility necessary in order to resolve situations unfair for the taxpayer, while situations of a limited manipulation of profit allowed by the same rule should not be regarded as an unwelcome bias. On the contrary, conventional wisdom suggests that fairness towards the taxpayer is ultimately more important than fairness towards governments not only from a legal point of view (in dubio pro libertate), but also from an economic point view, as higher economic growth leads to higher tax revenue.

32.5 Procedural Issues The OECD itself, in presenting a number of reasons which would render the application of global FA system unwelcome,41 considers as first major obstacle procedural issues, 39

See Section 32.5 for some potential problems. See esp. the ex post valuation rules on the hard-to-value intangibles in the OECD TP Guidelines (2022), paras 6.186—6.195. 41 Ibid., paras 1.16–1.32, esp. para. 1.21. 40

Transfer Pricing versus Formulary Apportionment 581 namely: (1) procedural difficulties of implementing a global FA system in a way simultaneously preventing double taxation and ensuring single taxation; and (2) the ‘enormous political and administrative complexity’ that a transition to FA would require.42 Although procedural issues would be unlikely to be the most important, they are on a practical level the most immediate obstacles preventing an FA system. Such procedural issues had already been recognized on a sub-federal level as a major reason to avoid a uniform obligatory formula: it was the US Supreme Court that denied in 1978 the ‘extensive judicial law-making’ that it would have imposed on itself ‘if the Constitution were read to mandate a prohibition against any overlap in the computation of taxable income by the States’, ‘without implementing legislation by Congress’.43 The US Supreme Court correctly realized that it would be impossible to resolve the issues of a judicially made uniform formula if such judicial lawmaking were feasible at all. But, as a matter of fact, even the US Congress never implemented the intention it declared in 1964 to issue federal legislation about a uniform and obligatory formula, even though it was that declared intention which had led then to an almost uniform, ‘voluntary’ adoption of the Massachusetts formula. It is remarkable that although uniformity was abandoned after the Supreme Court allowed non-uniform formulas, Congress never took action to harmonize FA, allowing states’ formulas to evolve towards more extensive or even exclusive consideration of the sales criterion. Even in the EU, where the European Commission, unlike the US Congress, has presented two detailed suggestions for a CCCTB (i.e. a formula), one optional in 2011 and one mandatory (for MNEs, but optional for small and medium-sized enterprises) in 2016, such fully harmonized FA has not had better luck until now. At the same time, other legislative proposals aiming to limit tax avoidance (the Anti-Tax Avoidance Directives (ATAD I and II) and the newer EU Directives on administrative cooperation in the field of taxation) were adopted with impressive speed, despite older EU lawmaking difficulties in the area of direct taxes. If the USA, a jurisdiction with a strong tradition in applying FA, has not been able to present an obligatory formula and if the detailed, double attempt by the EU to introduce one has not been (and may not become) successful, it seems questionable whether the detailed FA proposal contained in the Report on Pillar One Blueprint could ever become successful, irrespective of its adoption by politicians. A major cause for this would be certain procedural problems that the suggestions in Pillar One might generate, additional to those already presented by the OECD in the Transfer Pricing Guidelines: (1) The FA under Pillar One will apply only on part of the overall profits. This would be, for Amount A, the so-called ‘residual profit’, a concept presented academically for the first time in 201044 and, for Amount B, a fixed return on the sales of local intra-group 42

Ibid., paras 1.22–1.24. US Supreme Court, Moorman Manufacturing Co. v. Bair, 437 US 267, 277–281 (1978). 44 Avi-Yonah and Benshalom, ‘Formulary Apportionment—Myths and Prospects’, 3, who obviously refer to what the Report on Pillar One would describe in 2020 as non-routine profits. The similarity 43

582 Georgios Matsos distributors. For the rest of the profits, no FA is provided and, thus, classical transfer pricing will have to continue to apply. The simultaneous application of FA and transfer pricing for the same taxable base would be a task not only extremely difficult to surmount, but would also offer grounds for contradiction. An example of such contradiction would be the financing of taxes for loss-making MNE group members which would, however, have to pay taxes under FA. Profit-making group members would then have to finance loss-making but tax-paying group members. Such financing should result in additional tax and financial burden, as interest would have to be calculated according to the ALP. (2) Business of MNEs is or tends to become global. This means that in real life, MNEs will usually have to interact for their annual FA not with three or five national tax administrations, as the examples of the Report on Pillar One present,45 but with at least one hundred of them. By co-defining the taxable base of an MNE with all other tax administrations, each national tax administration will have a decisive word not only on its own taxable base, but also indirectly on the taxable base of all other countries. This is likely to render the whole project unworkable in real life. (3) Standard practice in most countries is that a company, whether or not part of an MNE group, is taxed in principle according to data provided to the tax authorities through the process of self-assessment, unless an audit performed by the tax authorities results in the correction of the taxable base and/or of the payable tax. the Report on Pillar One Blueprints provides for a ‘voluntary mechanism’ for MNE groups, which would be triggered by a ‘request for (early) tax certainty’.46 While even the fact of the introduction of such a mechanism reveals the level of tax uncertainty that a global FA would generate, one should consider whether this mechanism could effectively coordinate up to nearly 200 worldwide national tax administrations. Many tax administrations, either out of bureaucratic mentality or out of a legitimate will to exercise their duties appropriately, would not just let the applicant MNE escape unaudited, but would rather prompt some tax audit of companies falling under their jurisdiction. Apart from the natural unwillingness of companies to undergo tax audits, one could imagine the level of compliance costs that annual and potentially global tax audits would require. Many companies might risk tax uncertainty—unlike the assumption of the OECD draft on tax certainty47—rather than invite tax auditors to their doorsteps. Even for national tax administrations, the amount of resources that would be devoted in order to ensure that no MNE would escape unaudited would be immense.48 between the approach of Avi-Yonah and Benshalom and the approach of Pillar One on Amount A is striking, so that one could consider that article as the original source of the ideas that evolved to Amount A. 45

Report on Pillar One Blueprint, Annexes B and C. Ibid., paras 727–780. 47 OECD, Draft on Tax Certainty regarding Amount A (Paris: OECD Publishing. 2022), 10. 48 These disadvantages are not resolved by the OECD Draft on Tax Certainty about Amount A, ibid., which suggested a multilevel, rather complicated system of what is basically a global audit. Through a simple document search, sixteen instances can be counted where ‘all Affected Parties’ (i.e. all involved tax 46

Transfer Pricing versus Formulary Apportionment 583 (4) Pillar One underestimates the procedural problems that will almost certainly occur from the adoption of a unique set of tax accounting rules for the entire group. The Pillar One Blueprints suggest generally accepted accounting principles that produce ‘equivalent or comparable outcomes’ to the International Financial Reporting Standards.49 Though it is explicitly recognized that it is practically impossible to harmonize the tax base between each jurisdiction,50 no solutions are suggested for problems regarding coordination of the taxable base.51 The above- mentioned problems deriving from Pillar One suggestions will be problems for any global FA project. The first problem suggests that the simultaneous application of FA and of the ALP might ultimately prove too difficult in practice and, if FA is applied, it will have to be applied as a unique apportionment method. This would make the transition towards FA even more difficult. The second problem (involvement of potentially all tax administrations worldwide) is inherent in any global FA scheme. Theoretically, a sort of unitary global tax administration, like the Panels suggested in the 2022 OECD draft on tax certainty,52 could resolve such issues. Nevertheless, real-life examples from federal or quasi-federal jurisdictions suggest otherwise: the USA shows explicit (by its judiciary) or implicit (by its lawmakers) reluctance to work on a uniform and obligatory formula. And a strong transnational organization such as the EU does not, in reality, seem willing to adopt and apply FA. No higher efficiency should be expected on a global level, where the OECD is only a forum of legally non-binding coordination, with no authority to issue mandatory rules or to resolve unavoidable intergovernmental disputes. The third problem (global tax audits) would only increase the problems of tax uncertainty, which are implicitly admitted by the Pillar One Blueprints. Self-assessment seems easier, more reliable, and more efficient. Transfer pricing audits are not pleasant either, but, first, they do not always occur and, above all, MNEs can consistently avoid them by relying on good old APAs. APAs are an ex ante method acceptable both by tax administrations and by taxpayers. Gaining, on the contrary, the consent of all affected national tax administrations—which is necessary de facto—on the domestically produced taxable base of a company group member is something that cannot be achieved ex ante. As long as FA cannot create tools offering ex ante tax certainty administrations, ibid., 80) are requested to provide their own input, hence rendering ‘tax certainty’ to a simultaneous global audit requested by the taxpayer MNE. 49

Ibid., para. 407. Ibid., para. 408. 51 OECD, ‘Public Consultation Document on Pillar One—Amount A: Draft Model Rules for Tax Base Determinations’ (published 18 February 2022), available at https://www.oecd.org/tax/beps/public-consu ltation-document-pillar-one-amount-a-tax-base-determinations.pdf (accessed 17 July 2022), shows to what extent such problems have been underestimated in Pillar One Blueprints: while it acknowledges the necessity to present model rules, it does not, however, present such rules at all; it contains only a Title 5 and a Title 9 with no model rules, while titles containing model rules should still be drafted. 52 Ibid. 50

584 Georgios Matsos pertinent to the certainty currently achieved through APAs, both taxpayers and tax administrations will unlikely to be willing to abandon the ‘evil’ (i.e. transfer pricing) they already know. Regarding the fourth problem (determination of a unitary tax base for a unitary profit), the European Commission has recognized that the adoption of a common corporate tax base (CCTB) has to precede the adoption of a CCCTB,53 thus admitting that FA is not feasible, unless the corporate tax base is harmonized. Given the current applicability of FA only in the sub-federal context, where the corporate tax base is already harmonized, this conclusion by the European Commission seems correct: iIt is not possible to apportion something that has not been previously defined. Last but not least, an inherent procedural obstacle for the application of a global FA would be the existing antagonism between tax administrations: if the main driving force for a global FA is the wish of some countries to gain more taxes at the expense of other countries, it is not to be expected that FA will alter such a deeply rooted antagonistic culture. An established culture of tax antagonism is the exact opposite of the cultural and societal integration that is necessary for successfully applying FA. History shows that, if at all, only sub-federal entities could achieve such integration to an adequate degree. No matter how brilliant new procedural ideas might sound, political integration must be clearly considered as a conditio sine qua non for the elementary functioning of a global formulary apportionment.

32.6 Formulary Apportionment vs BEPS Project The BEPS Project had been in substance—at least in its initial phase—an effort to enhance and to generalize the use of classical tools of international tax law, aiming to abolish concrete gaps and to fix mismatches in cross-border taxation. Enhancing transfer pricing had been a major aim of those original BEPS efforts. It was no coincidence that four out of the original fifteen BEPS actions54 had transfer pricing explicitly mentioned in their title. The BEPS Project was essentially an attempt to impose anti-avoidance rules on those countries, which for whatever reason—often not unintentionally—had failed to introduce such rules in their domestic legislation and had wanted or at had at least tolerated greater leeway for tax avoidance. Tax avoidance gaps in domestic legislation and of DTCs had been including, among other things, a less strict application of transfer

53 See I. Chelyadina, Harmonization of Corporate Tax Base in the EU: An Idea Whose Time Has Come? (Bruges: College of Europe, 2019), 3. 54 Namely, BEPS Actions 8, 9, and 10, which had as a common header ‘Assure that transfer pricing outcomes are in line with value creation’ and Action 13 (‘Transfer Pricing Documentation’).

Transfer Pricing versus Formulary Apportionment 585 pricing rules, both on a procedural and on a substantial level. In the area of transfer pricing, the BEPS Project introduced numerous innovative ideas,55 including a two- step application of the ALP according to a thorough risk analysis,56 the introduction of the DEMPE (development, enhancement, maintenance, protection, and exploitation) standard about intangibles,57 and the ex post valuation of hard-to-value intangibles.58 FA, on the other hand, has never until today been an international taxation tool. Moreover, FA is structurally not a means against tax avoidance, but rather a fundamentally different concept of allocating taxing powers among jurisdictions. The idea of including FA as part of BEPS is obviously due to the fact that it is regarded as less prone to manipulation than the current allocation of taxing powers,59 but still does not directly aim at tax avoidance per se. If Pillar One thus tends to at least partially replace all reforms introduced by the (initial) BEPS Project, it is incorrect to consider it as part of the very same BEPS Project,60 despite its current title suggesting the opposite. Even on the issue of combating tax avoidance, the comparison between FA and the ALP does not take into consideration a major evolution triggered by the BEPS Project: the ALP after the initial BEPS Project is much less manipulable. Moreover, transfer pricing and the ALP have not yet been evaluated under the light of the BEPS reforms.61 Contemporary FA projects like the CCCTB and Pillar One pose, thus, numerous questions: why should the ALP be abandoned just after new reforms have been implemented and before they are evaluated? Has the initial BEPS Project not achieved anything, especially on the level of transfer pricing? What really is the purpose of an allegedly unified programme whose individual parts contradict each other? Unlike Pillar One, Pillar Two62 continues to pursue the basic idea of the original BEPS Project, namely to prevent countries from taxing less. The initial BEPS Project aimed to combat unseen tax competition (wanted or unwanted tax gaps), while Pillar Two aims to reduce, if not to eliminate altogether, tax competition by imposing an effective minimum corporate taxation. The question of how the taxable base is to distribute among

55 See

a comprehensive summary of these main substantial innovations brought to the OECD transfer pricing guidelines by A. Eigelshoven and D. Retzer in H.-K. Kroppen and S. Rasch, Handbuch Internationale Verrechnungspreise, Vorbemerkungen, 32nd update (Cologne: Otto Schmidt, 2020), para. 13. 56 OECD TP Guidelines (2022), para. 1.60. 57 Ibid., para. 6.71. 58 Ibid., paras 6.168 and 6.181ff. 59 See Section 32.3, esp. regarding the tendency of formulas to become sales-based and the reduced manipulability of sales taxes. 60 Correctly pointed out by L. Eden and O. Treidler, ‘Insight: Taxing the Digital Economy—Pillar One Is Not BEPS 2 (Part I)’ (8 Nov. 2019), https://news.bloombergtax.com/daily-tax-report/insight-taxing- the-digital-economy-pillar-one-is-not-beps-2-part-1 (accessed 17 July 2022). 61 Ibid.: ‘It is also important to recognize that, at this point in time, there has been little empirical work evaluating the impact of the BEPS reforms implemented since 2017’. 62 See for details, ‘Minimum Taxation under Pillar Two of the OECD Proposal (‘GloBE’)’ in this volume.

586 Georgios Matsos taxing jurisdictions is a different one and, as a matter of fact, it has little to do with combating tax avoidance—other than the extent to which it would theoretically be less manipulable than the ALP. FA thus shifts the focus from the elimination of base erosion and profit shifting, to the reallocation of tax revenue to countries other than those provided by the net profits/ ALP system currently in place. It aims to create a fundamentally different framework of international allocation of taxing powers. In what direction is such reallocation moving? Why has such reallocation, if it does not combat BEPS and tax avoidance, been alleged to be part of an anti-BEPS Project? And, finally, how is formulary apportionment linked to the reforms currently suggested by the Inclusive Framework?

32.7 Formulary Apportionment as Neo-P rotectionism? The answers to these questions are found in the destination-based sales formula towards which FA systems tend to evolve:63 a destination-based sales formula allocates tax revenue to jurisdictions where consumers are located. This signifies that a destination- based sales formula mostly favours consumer countries with greater purchasing power. On the contrary, producer countries—including countries producing technology— will benefit less from global corporate tax revenue. As a matter of fact, a destination- based sales formula will render their production potential ultimately irrelevant for the purposes of corporate taxation: in the extreme case that no one buys in the production country, no tax revenue will be then allocated to that jurisdiction. This means that consumer countries—which also happen to be the most powerful ones politically—would then impose themselves as recipients of corporate tax revenue over and above producer countries. This would amount to a form of neo-protectionism from the side of the currently politically powerful consumer countries, against less powerful producer countries. Such an approach takes for granted that FA is based on an exclusively destination- based sales formula. If, for example, a formula were more heavily inclined towards payroll, then producer countries would benefit more. In theory, different formulas would indeed cause different results. But, in reality, both the free-floating of the US formula as well as the plans of the international community of tax administrations have led to sales- based formulas, the latter being in fact heavily destination-based. A destination-based sales formula would be, as shown earlier, the least manipulable formula—and this is the main argument presented in its favour and the only reason to consider FA as part of an anti-BEPS Project. 63

See Section 32.3.

Transfer Pricing versus Formulary Apportionment 587 As such, it is reality which suggests that formulary apportionment should nowadays be evaluated on the central basis of a destination-based sales apportionment criterion.64 Under this explicit perspective, including formulary apportionment in the BEPS Project reveals the underlying purposes of the whole BEPS Project: while combating tax avoidance originally seemed like a principally legitimate purpose, the critical detail was that the programme aimed to oblige all participating countries to combat tax avoidance. However, if a country wanted to tax less, it could still do so by simply keeping low tax rates. Tax competition was, thus, still possible. Under Pillar Two, which introduces a minimum effective taxation, tax competition is formally declared as unwanted and is targeted as a practice that has to be eliminated. This is still a legitimate goal. But why, then, after tax competition is pretty much eliminated, should the international community still aim at reallocation of global corporate tax revenue? Moreover, why should such reallocation, after all ‘Base Erosion and Profit Shifting’ is eliminated, still be part of an alleged anti-BEPS approach? It is obvious from this evolution that the whole BEPS Project had from the very beginning neo-protectionism as its core idea. Richer countries wanted not only to preserve their high tax environment, but ultimately to attract in all possible ways as much tax revenue as they could to their treasuries. History has shown, however, that protectionism cannot reverse the currents of history: mostly dynamic countries with technology potential and higher production will ultimately prevail over countries with less economic potential but with currently higher purchasing power. Leading technology countries will lead the field even in terms of purchasing power and thus also in terms of business profits apportioned according to destination-based sales formulas. Consequently, the well-established global system which currently allocates taxing powers on the basis of net profits produced in each country should not be destroyed purely in favour of the covert neo-protectionism of formulary apportionment. Transfer pricing and the ALP as cornerstones of a net-profits-based international tax allocation system are therefore much more appropriate for a world dominated, for the foreseeable future, by antagonism among national tax administrations. This could change only if global cultural and societal integration actually reduces such antagonism.

64 See

also R. S. Avi-Yonah and K. Clausing, ‘A Proposal to Adopt Formulary Apportionment for Corporate Income Taxation: The Hamilton Project’, Public Law and Legal Theory Working Paper Series, Working Paper No. 85 (June 2007), 11. This article was among the first ones to suggest FA for international use, though at that time limited to the USA; it suggested the exclusive use of a destination- based sales formula.

Pa rt V

THE E U ROP E A N I Z AT ION OF I N T E R NAT IONA L TAX L AW

Chapter 33

The Role of t h e E C J i n t he Devel opme nt of Internationa l Tax L aw Adrian Cloer

33.1 Introduction The European Court of Justice (ECJ) is the authoritative institution for the interpretation and application of European tax law. The case law of the ECJ and the interpretation of legal provisions shaped by the Court transcends the borders of the EU member states when the member states apply the law and thereby creating sensitivity for the interpretation of terms in relation to third countries. The mode of operation of Union law is of fundamental importance, as it constitutes binding law for the member states. In terms of the binding nature of international tax law, European tax law has a broader impact (Section 33.2). On the question of the legally binding nature of EU law, a distinction must be made between the position of the ECJ intra-EU (Section 33.2.1) and extra-EU (Section 33.2.2). Within the EU, on the one hand, the focus is on effective legal protection and integration and, on the other hand. outside the EU, integration efforts and a departure from Union law. Union law has a strong position through its legally binding nature and thus also the interpretation by the institution ECJ. The role of the ECJ is subject to change and interplay with the needs of the member states (Section 33.3). The member states have shown through different directives in the area of direct taxation that there is a desire for standardization even without a comprehensive harmonization mandate from the Treaties. With the required unanimity, the member states have established the law, which is interpreted by the ECJ in the event of disputes. Beyond directive law, the ECJ has found a way to expand the content of its case law in primary law with the prevention of abuse of the benefits of Union law (Section 33.4). The ECJ has not acted arbitrarily in expanding the scope of its interpretation of the law.

592 Adrian Cloer This is because the abuse proviso is a generally recognized principle of Union law that affects the exercise of every fundamental freedom. Nevertheless, the limits of the ECJ’s competences have been extended, since the concept of abuse is universally applicable and has no definite outlines without a written basis. The expansion of competences is accompanied by the question of the legitimacy of a far-reaching interpretation of Union law (Section 33.5). This clearly results from the Treaties. With regard to the scope of legitimacy, the ECJ finds itself in an interplay between the member states’ desire for clear interpretative guidelines and, in contrast, tax sovereignty, which the member states do not want to lose in the area of direct taxation. Tax sovereignty remains with the member states, as they also transfer only parts of their competences to the EU and with it to the ECJ according to the principle of conferral of powers. Within the granted competence, the influence of the ECJ can begin at an early stage. This is because through the case law of the ECJ on the interpretation and application of Union law, the guidelines developed from this can have an influence on the norm-setting process (Section 33.6). Thus, the role of the ECJ does not begin only after legislation has been made, but already has a clear influence on the member states in the legislative process. Apart from the ECJ, no other EU institution is available for the application and interpretation of Union law. The interpretation of the ECJ is the only one to be followed. This could lead to a development in which the ECJ’s case law assumes a monopoly position (Section 33.7). Once a specification on the content of a legal norm has been made, it can only be changed after it has been resubmitted and reviewed. The extent to which the role of the ECJ in European and international tax law has changed or will change from the interpretation of Union law to a restriction of the freedom of the national legislator to shape tax law must be clarified in conclusion (Section 33.8). When speaking of the freedom to design tax laws by the member states, this must be seen in relation to the observance of the prerequisites and consequences of a tax law norm within the scope of application of Union law.

33.2 Classification of International and European Legal Binding Force The binding nature of legislation in European and international tax law is based on fundamentally different pillars. In order to understand the mode of operation and the role of the individual institutions, it is relevant to classify the binding nature of the respective legal system. Within the framework of the OECD, countries meet to discuss the structure of international tax law. There is no legal binding, so-called soft law is formed.1 This is because 1

Lehner in Vogel and Lehner, para. 124b;

The Role of the ECJ in Development of International Tax Law 593 the OECD has no genuine legislative competence. International tax law is shaped and driven forward in particular by double taxation agreements (DTAs). The DTAs in Europe are based on the OECD Model Convention (OECD MC), which represents a recommendation for action for the uniform application and consistent understanding of the respective treaty articles. The OECD MC has created a widespread, albeit non- binding, model for the conclusion of bilateral tax treaties, which aims to bring about common regulations on a uniform basis for problematic issues in the area of international (double) taxation identified at the OECD level. If the power of taxation were divided throughout Europe according to a uniform system, double taxation of the same income by two states could be completely ruled out. However, the conclusion of a Europe-wide multilateral double taxation treaty or even the enactment of comprehensive secondary conflict-of-law rules seems to be some way off.2 A more global consensus through the Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting (Multilateral Convention) is available and is used in different ways.3 The innovations agreed by the OECD within the framework of the Base Erosion and Profit Shifting (BEPS) Project are then incorporated into the revisions of the OECD MC, which is continually updated.4 Neither the BEPS measures nor the resulting implementation proposals of the OECD in the conclusions on the action points are legally binding. Certain measures, such as Action Item 14 on improving the mutual agreement procedure, were highlighted as a minimum standard, meaning that failure to implement these items would have a negative impact on other countries in terms of reduced competitiveness.5 The minimum standards are thus accompanied by political pressure for implementation. A political weighting of the OECD measures is evident not least in the fact that the EU is waiting for the OECD’s proposals on the taxation of the digital economy. Despite its own approaches6 and national go-it-alones,7 the EU is reverently awaiting the global consensus of the OECD. In the area of Union law, its supranationalism is of a different decisional rank. Union law is binding law. It has priority of application over national law.8 European law was codified in the Treaty on European Union (TEU) and the Treaty on the Functioning of the European Union (TFEU) on the basis of the principle of subsidiarity and seeks to unify the internal market through the effective exercise of EU competences.9 The objective of achieving a single internal market is clearly emphasized by the Treaties: article 2

Kokott and Henze, BetriebsBerater (2007), 914, 915. OECD, Multilateral Convention (24 Nov. 2016). 4 OECD, Model Tax Convention on Income and on Capital (21 Nov. 2017), Introduction, para. 11.2. 5 OECD (2016), ‘BEPS Project Explanatory Statement: 2015 Final Reports’, 6 para. 11, 18. 6 COM (2018) 148 final (2018); COM (2018) 147 final. 7 Austria: Digitalsteuergesetz 2020 of 20 November 2019, Federal Law Gazette I No. 91/ 2019; Spain: Impuesto sobre determinados servicios Digitales of 15 October 2020, Boletin oficial del Estado, No. 274, s. I., 88569; Great Britain: Digital Service Tax of 22 July 2020, Finance Act 2020. 8 Schaumburg in Schaumburg and Englisch, paras 4.18ff. 9 Calliess in Calliess and Ruffert, art. 5 TEU, para. 1. 3

594 Adrian Cloer 3 paragraph 3 TEU, article 4 paragraph 2 lit. a TFEU. By transferring its own sovereign rights to the EU according to the principle of conferral, the EU is empowered to create a supranational legal order.10 The binding nature of EU law is also evident insofar as measures of the OECD BEPS Project have been taken up by the EU and implemented by means of binding directive law. The EU participates in the OECD BEPS Project and promotes the work at OECD level. The implementation of the findings of the BEPS Project can be found, among other things, in the Directive laying down rules against tax avoidance practices that directly affect the functioning of the internal market (so-called Anti-Tax Avoidance Directive, ATAD),11 the Directive on Tax Dispute resolution mechanisms in the European Union,12 and in the area of administrative cooperation with the Directive on Administrative Cooperation (DAC).13 In this context, the position of the ECJ is prominent as it acts as a guardian of Union law, shapes it in particular through preliminary ruling procedures (art. 267 TFEU), and promotes the internal market as a major objective.14 This position is clearly defined in the Treaties in that article 19 paragraph 1 s. 2 TEU provides that the ECJ shall have the task of ensuring that the law is observed in the interpretation and application of the Treaties. The ECJ’s remit includes, in particular, the interpretation of Treaties or other Union law, the review of the legal acts of the Union institutions for their compatibility with higher ranking law, and the review of the conduct of the member states against the yardstick of Union law.15 If Union law is decisive, this leads in part to the inapplicability of bilateral agreements. In the two categories of legal bindingness, the primacy of application of Union law prevails over the ‘soft law’ of the OECD in international tax law.

33.2.1 Role of the ECJ Intra-EU: Effective Legal Protection and Integration As a community based on law, the EU must ensure comprehensive and effective legal protection through which Union law can be enforced. This is a general principle of Union law which derives from the constitutional traditions common to the member states and which individuals have a right to expect to be respected.16 Legal protection

10

Ibid., paras 6ff. Council Directive (EU) 2016/1164 of 12 June 2016 [2016] OJ L193/1; see further, Govind and Zolles in Lang et al. (Pub.), 217. 12 Council Directive (EU) 2017/1852 of 10 October 2017 [2017] OJ L265/1. 13 Council Directive (EU) 2018/822 of 25 May 2018 [2018] OJ L139/1. 14 Art. 3 para. 3 lit. b) TEU; Herzfeld, TNI 2014, 107. 15 Schaumburg in Schaumburg and Englisch, para. 2.25. 16 ECJ: Cases C-23/04, C-24/04, and C-25/04 Sfakianakis, ECLI:EU:C:2006:92. 11

The Role of the ECJ in Development of International Tax Law 595 is to be guaranteed first and foremost by the courts of the member states.17 The courts of the member states can by no means assume this task alone. As a supranational community of law, the EU is dependent on its own courts. Uniform effectiveness of the law and the avoidance of legal fragmentation can only be guaranteed by the European Courts.18 Given the volume of EU law, the large number of members, and the numerous EU institutions, it is inevitable that the authenticity of EU law must be determined uniformly and with effect for all by the ECJ.19 The preliminary ruling procedure under article 267 TFEU plays a decisive role. This procedure serves to ensure the uniform interpretation of Union law. Beyond the specific case, the preliminary ruling procedure has a broad effect, as it provides the framework for the interpretation of Union law and the further development of law in the EU.20 Although the judgments of the ECJ dealing with the interpretation of Union law in principle only have inter partes effect, the actual effects go beyond the individual case decided. The interpretation of Union law involves the clarification of abstract legal questions in the judgment. Since Union law and the interpretation given by the judgment take precedence over national law, this has an erga omnes effect; that is, it also binds all courts and administrative authorities in the application of the provision that was the subject of the preliminary ruling. The binding effect also affects the respective legislator with the consequence that the law must be amended on the basis of the requirements of the interpretation by the ECJ.21 This results not least from the duty of loyalty under article 4 paragraph 3 subpara. 2 TEU. Due to its power to interpret and apply Union law, the ECJ is referred to in particular as the ‘engine of integration’.22 This integration takes place, among other things, through the creation of generally binding legal principles by the ECJ. For example, the ECJ has shaped the primacy of European law over national law and the functioning of the fundamental freedoms not only as prohibitions of discrimination, but also as prohibitions of restrictions. As an interim conclusion, it can be stated that the ECJ continues to form the law within the EU. Through the further development of the law, the ECJ serves the effective protection of the law and ensures integration by means of a uniform interpretation of Union law.

17

ECJ: Case C-432/05 Unibet, ECLI:EU:C:2006:755. Oellerich in Schaumburg and Englisch, para. 5.1. 19 Weber-Grellet in Musil and Weber-Grellet, art. 19 TEU, para. 8. 20 Schaumburg in Schaumburg and Englisch, paras 24.3ff. 21 Ibid., para. 24.8. 22 Weber-Grellet in Musil and Weber-Grellet, art. 19 TEU, para. 9. 18

596 Adrian Cloer

33.3 Role of the ECJ Extra-EU: Integration Drive and Disengagement There are two main perspectives on the role of the ECJ outside the EU. On the one hand, from the point of view of the states aspiring to join the EU and, on the other hand, from the point of view of the state that has left the EU and is turning away from the EU and its legal structure. First, the states that aspire to join the community of member states. States wishing to join the EU must prepare for the status of member state by meeting certain criteria. Current candidates for EU membership are Albania, Bosnia and Herzegovina, Kosovo, Montenegro, Northern Macedonia, Serbia, and Turkey. The so-called Copenhagen criteria23 to be fulfilled by the accession candidates are essentially subdivided into a political criterion, according to which institutional stability, democratic order and the rule of law, respect for human rights, as well as respect for and protection of minorities are required. The economic criterion aims at a functioning market economy and the ability to withstand competitive pressure within the EU internal market. In addition, there is the acquis criterion. This is understood to mean the ability to make the obligations and goals arising from EU membership one’s own. This includes the adoption of the entire body of Community law (acquis communautaire). The criteria are specified by individual accession agreements within the meaning of article 49 TEU.24 The acquis criterion is of particular importance for the position of the ECJ. This criterion is intended to ensure that there is the capacity to implement Union law. To this end, a personal, organizational, and procedural safeguard is required, including in particular the independence of the courts and the effectiveness of judicial protection.25 This requirement provides the basis for national courts to refer cases to the ECJ for a preliminary ruling under article 267 TFEU. The application and interpretation by the ECJ are safeguarded by the domestic functioning of the rule of law.26 From the day of accession, all Union law is binding, so its effective enforcement must be ensured.27 The preparation for entry into the EU and the recognition of the rule of law mechanisms as well as the binding nature of Union law have been perfected over the 23 Criteria

Copenhagen, Consolidated version of the Treaty of European Union, Title VI, Final provisions, Art. 49 (ex. art. 49 TEU) [2016] OJ C202/43. 24 Commission Staff Working Document, Albania 2019 Report, Analytical Report (Bosnia and Herzegovina), Montenegro 2019 Report, North Macedonia 2019 Report, Serbia 2019 Report, Turkey 2019 Report. 25 Ohler in Grabitz, Hilf, and Nettesheim, art. 49 TEU, para. 20; see also Communication on EU enlargement policy, COM(2020) 660 final (6 Oct. 2020). 26 ECJ: Case C-11/7 7 Patrick, ECLI:EU:C:1977:133; Case C-43/75 Defrenne, ECLI:EU:C:1976:56; Case C-258/81 Metallurgiki Halyps, ECLI:EU:C1982:422. 27 See e.g. art. 2 Accession Act [2002] OJ L236/33.

The Role of the ECJ in Development of International Tax Law 597 years. The act of enlargement is contrasted with the departure from the binding nature of Union law, which is characterized by far fewer historical precedents. The withdrawal of Greenland from the European Community (EC) in 1982 took place before the explicit provision of a withdrawal clause with article 50 TEU, which came into force on 1 December 2009.28 The withdrawal of the UK from the EU under article 50 TEU is thus a novelty. According to article 50 paragraph 3 TEU, the Treaties no longer apply from the date of entry into force of the withdrawal agreement. The withdrawal from the EU requirements is regulated in detail by the withdrawal agreement.29 Nevertheless, the question arises as to how the UK will deal with domestic standards that have been shaped by the interpretation of the ECJ. The UK will not be able to immediately reverse the existing interpretation of its domestic tax laws that has been practised for years in conformity with EU law. Thus, Union law shaped by the ECJ continues to exist in domestic law without Union law still being binding. As a consequence, the UK refuses to apply Union law which, however, remains implemented in domestic law. Detachment from the case law of the ECJ requires its own efforts.

33.4 Strengthening the Role of the ECJ through the EU (Member States’) Will to Standardize The position of the ECJ is strengthened by an increasing will of the EU to harmonize in the area of direct taxation and has an effect beyond the borders of the internal market. The member states show a will to harmonize direct taxes. The EU Commission is striving to remove internal market barriers and thus achieve de facto standardization. In this way, the existing disparities among the member states will gradually be smoothed out. The problem is that the EU Commission does not have the authority and legal means to impose binding measures on the member states. The EU Commission is only endowed with a right of initiative.30 It is therefore up to the member states to harmonize. The extent of the coordination among the member states and what this means for the ECJ will be explored in more detail. In order to prevent impairment of the internal market, harmonization is carried out to the extent that the member states’ powers of competence extend. However, harmonization of laws in the internal market is not provided for in principle for taxes: article 114 paragraph 2 TFEU. Only for the harmonization of indirect taxes is there an explicit 28

Consolidated Versions of the Treaty on European Union and the Treaty on the Functioning of the European Union, 2010/C 83/01, 30 March 2010. 29 Agreement on the withdrawal of the United Kingdom of Great Britain and Northern Ireland from the European Union and the European Atomic Energy Community [2020] OJ L29/7. 30 Musil in Musil and Weber-Grellet, Einführung, para. 34

598 Adrian Cloer mandate according to article 113 TFEU, which is based on the premise that harmonization is necessary for the functioning of the internal market and the avoidance of distortions of competition. There is no explicit harmonization mandate for the area of direct taxation. An approximation of the law on direct taxation is possible within the framework of article 115 TFEU insofar as it directly affects the establishment or functioning of the internal market.31 In this respect, article 116 TFEU is not a sufficient basis for harmonization, as this provision provides for a repressive procedure. As a result of consultations with the member state, laws, regulations, and administrative provisions of the member states which distort conditions of competition in the internal market and thereby create a distortion are approximated. The scope of the standard is narrowly defined by the characteristic of distortion of competition. While articles 114 and 115 TFEU concern general harmonization, measures aimed at the functioning of the internal market—articles 116 and 117 TFEU— contain additional authorizations for harmonization in specific situations.32 Distortions of competition in the tax area could, for example, lie in the granting of a special depreciation for certain taxpayers as an economic promotion. However, the consequence of an identified distortion is only the harmonization of the provision in question. A new impulse for harmonization of an area of direct taxation does not arise from this. If a measure in the area of direct taxation were to be based on article 116 TFEU,33 this would lead to a circumvention of the unanimity principle and should be critically questioned. The importance of the standard is reflected in the fact that consultation procedures have been carried out in a few cases, but the Council has not yet adopted a directive pursuant to article 116 paragraph 2 TFEU.34 Article 115 TFEU is therefore the primary harmonization provision for direct taxation. The decision to harmonize the legislative or administrative provision under article 115 TFEU must be taken unanimously, so that the member states are protected from a sprawling loss of competence despite the possibility of issuing directives.35 The containment of a loss of competence is contrasted by the fact that projects aimed at with the unanimity requirement cannot be implemented if states make use of the veto competence. This was demonstrated by the long-standing discussion on the Common (Consolidated) Corporate Tax Base (CC(C)TB), which sometimes met with no consensus.36 A move away from the unanimity requirement in tax matters has been called for at times and has so far always been rejected.37 The ATAD,38 on the other hand, is an example of rapid consensus-building in the EU. With the ATAD, extensive use was 31

Tietje in Grabitz, Hilf, and Nettesheim, art. 115 TFEU, para. 16. Ibid., art. 116 TFEU, para. 1. 33 COM (2020) 312 final (15 July 2020). 34 Classen in von der Groeben, Schwarze, and Hatje, art. 116 TFEU, paras 35ff. 35 Korte in Calliess and Ruffert, art. 115 TFEU, para. 9. 36 COM (2016) 683 final (25 Oct. 2016). 37 BDI: EU-Steuerpolitik of 7 July 2020, 18; https://ec.europa.eu/commission/presscorner/detail/en/ IP_19_225; Sprackland, TNI 2018, 1350. 38 Council Directive (EU) 2016/1164 of 12 June 2016 [2016] OJ L193/1. 32

The Role of the ECJ in Development of International Tax Law 599 recently made of the directive competence in the area of direct taxation. In this context, it remains debatable to what extent article 115 TFEU constitutes a sufficient authorization basis for such far-reaching directives, since the direct effect on the establishment or functioning of the internal market cannot be ascertained for all measures. In particular, individual measures of the ATAD, such as article 4 ATAD with the interest barrier and article 6 ATAD as a general abuse avoidance provision, have a scope of application that can also extend to domestic law. The scope of the directive competence is disputed in detail. If directives are adopted unanimously, the ECJ is authorized to rule on the interpretation of the provisions of the directives. The ECJ’s competence to interpret Union law is clearly derived from the Treaties (art. 19 TEU, arts 251–281 TFEU, Statute of the ECJ). If the ECJ finds that a provision is contrary to European law, it must resolve this. If it did not interpret the provision on the basis of Union law, it would not be fulfilling its mandate. The Treaties attribute to the EU institutions their competences, which cannot be withdrawn to that extent. The unification of direct taxation through directives has strengthened the role of the ECJ. The member states have the backing of the ECJ, but they still retain a large part of their tax sovereignty. As a procedural consequence, the preliminary ruling procedures of the member states also strengthen the role of the ECJ. The referrals provide the ECJ with the legal questions on which it can then rule. The courts of some member states are more willing to make referrals than others. In the period from 2015 to 2019, there was an overall increase in the number of preliminary rulings. The current statistics are from 2019. While in 2015 a total of 436 references were submitted to the ECJ, in 2016 there were 470 references, in 2017 there were 533, in 2018 this reached 568, and finally in 2019 there were 641 references for a preliminary ruling.39 Of these, sixty-seven referrals in 2019 were in the area of tax law.40 Germany is by far the member state with the most references for a preliminary ruling overall (i.e. irrespective of the subject matter of the proceedings, with 114 references in 2019), followed by Italy (70 references in 2019), and Spain (64 references in 2019).41 The propensity of these member states to make referrals may be related to strong foreign trade activity.

33.5 Extension of Jurisdiction by the ECJ: Preventing Abuse of Rights The common internal market is an undisputed common objective of the member states. In order to be consistent with the functioning of the internal market, this implies that 39

ECJ, Annual Report 2019, judicial activity, 170. Ibid., 172. 41 Ibid., 173. 40

600 Adrian Cloer the advantages of the internal market are not granted indefinitely. The fundamental freedoms which apply in the internal market and which guarantee the protection of activities and persons working in the internal market cannot be invoked by those who make improper use of them. The fundamental freedoms require a coherent and non- discriminatory design of the tax systems and contribute to their convergence through these structural requirements.42 The ECJ had already established the requirement of preventing abuse at an early stage of the realization of the internal market. The ECJ refers to a general principle of Community law according to which abuse of rights is prohibited.43 The defence of abusive reliance on the fundamental freedoms is a doctrine of the ECJ that has been developed over the years. The ECJ goes a long way in defending against tax abuse: in its case law, the ECJ has derived a tax abuse avoidance rule from unwritten EU primary law, which must be respected in the member states.44 The unwritten abuse provisio is concretized by the ECJ itself and made applicable through its case law.45 With the referrals to the ECJ in cases N Luxembourg et al. and T Denmark et al., the interpretation of the opening clause of the Parent–Subsidiary Directive (PSD),46 former edition (art. 1 para. 2 of Directive 90/435/EEC, now art. 1 para. 4 PSD), as well as the enabling provision for combating abuse (art. 5 para. 2 of the Interest and Royalties Directive, IRD) enshrined in the IRD47 have become subject of dispute.48 The central statement of the ECJ rulings is the dispensability of an explicit national, general anti- abuse provision in order to prevent abuse via the directive opening clause or enabling provision. For the denial of benefits from directives due to abuse, it is not important whether the member state concerned has formally included a provision specifically intended for this purpose in its legal system. Rather, the principle of prohibition of abuse, which is universally applicable in Union law, obliges the member states, if abusive practices are identified, to compulsorily deny the benefits accruing to the taxpayer.49 For cases in which there is no written general abuse provision in national—in the submission, Danish50—tax law, general legal principles of national law are sufficient, according to the ECJ case law, to be able to combat abuse within the scope of application 42

Kokott and Henze, BetriebsBerater (2007), 914. ECJ: Case C-321/05 Kofoed, ECLI:EU:C:2007:408; Case C-212/97 Centros, ECLI:EU:C:1999:126. 44 ECJ: Case C-115/16 N Luxembourg, ECLI:EU:C:2019:134; Cases C-116/16–C-117/16 T Danmark, ECLI:EU:C:2019:135. 45 Schade and van Lück, StuW (2020), 347. 46 Council Directive 2011/96/EU of 30 November 2011 (recast) [2011] OJ L345/8 amending Directive (EU) 2015/121 of 27 January 2015 [2015] OJ L21/1; see further Tenore in Lang et al. (Pub.), 143. 47 Council Directive 2003/49/EC of 3 June 2003 [2003] OJ L157/49. 48 ECJ: Case C-115/16 N Luxembourg, ECLI:EU:C:2019:134; Cases C-116/16–C-117/16 T Danmark, ECLI:EU:C:2019:135. 49 ECJ: Case C-115/16 N Luxembourg, ECLI:EU:C:2019:134, e.g. paras 110 and 120. The Grand Chamber of the ECJ thus continues its previous case law, see Case C‑359/16 Altun, ECLI:EU:C:2018:63, e.g. para. 49. 50 ECJ: Case C-115/16 N Luxembourg, ECLI:EU:C:2019:134, e.g. para. 24. On the new anti-abuse rule in relation to the Parent–Subsidiary Directive, see Bundgaard et al., Intertax (2018), 716ff. 43

The Role of the ECJ in Development of International Tax Law 601 of Union law.51 If such legal principles are not anchored in national law, the member states must directly resort to the Union law principle of the prohibition of abuse.52 In this decision, the ECJ, with recourse to decisions on company law, value-added tax, and social policy,53 among others, now forces a requirement of combating abuse that applies to all Union law, to which the member states and their institutions must directly refer in order to preserve the practical effectiveness of Union law—the effet utile concept—and to protect the internal market.54 This obligation of the member states results from article 4 paragraph 3 TEU.55 It becomes clear that written regulations against abuse are obsolete. This is because the general abuse proviso is inherent in Union law.56 By giving the anti-abuse rule a high rank in unwritten EU primary law, the ECJ extends its competence. In the absence of written law, the ECJ can always invoke the abuse avoidance rule and give the concept its own contours without having to adhere to written guidelines. As a result, and as a goal, due to the primacy of application of European law over national law, there is no cross-border legislation that stands in the way of the greatest possible effectiveness of European law (effet utile).

33.6 Legitimation of the ECJ for the (Far-R eaching) Interpretation of Union Law Discussions about the powers of the ECJ revolve not least around the question of the legitimacy of the ECJ for such a far-reaching interpretation of unwritten primary law. The ECJ shapes the interpretation of Union law without parliamentary legislative competence. This means that the case law of the ECJ has an effect with a quasi-legislative character. Since the ECJ has the sole competence to interpret Union law, this is also mandatory. The fundamental freedoms are not limited in their application, since there is precisely no area exception for tax law in the case of the fundamental freedoms. This was not taken into account when the EU was founded and when competences were allocated in this scope. A feedback in the sense of a continuous confirmation of the ECJ’s interpretative power to the member states is found in the composition of the ECJ with the secondment of one judge from each member state. The composition with at least one judge per 51

Case C-115/16 N Luxembourg, ibid., e.g. para. 116; see too Kofler in Schaumburg and Englisch, para. 13.20. 52 Case C-115/16 N Luxembourg, ibid., e.g. para. 111; see too Buckler, EuR (2018), 377. 53 Case C-115/16 N Luxembourg, ibid., e.g. paras 96 and 100. 54 Case C-115/16 N Luxembourg, ibid., e.g. paras 104ff. 55 Case C‑323/18 Tesco-Global Áruházak, ECLI:EU:C:2019:567 AG Kokott, Opinion of 4 July 2019, para. 84. 56 See Gebhardt in Festschrift Lüdicke (2019), 176.

602 Adrian Cloer member state—regardless of the size of the member state—ensures a link between the EU and the member states.57 The ECJ has far-reaching powers of interpretation in the area of abuse defence. This could have been left aside insofar as the defence against abuse is also an integral part of turnover tax law.58 As early as 2000, the ECJ provided the first reason for anchoring the concept of abuse in the area of direct taxation.59 The anchoring of the defence against abuse in income tax law and the clear position of the ECJ is also driven by increased efforts towards standardization by means of Union law. This is because Union directives have already inserted abuse prevention. Examples include the Savings Interest Directive,60 which has now been abolished, or the correspondence provision in the Parent–Subsidiary Directive61 to avoid abuse as a directive in favour of taxpayers. One innovation is the insertion of anti-abuse provisions to the detriment of taxpayers, as provided for in the ATAD.62 In this way, the legislator of the directive intervenes extensively in the area of abuse prevention. The clear positioning of the ECJ on the abuse defence enshrined in unwritten primary law by the N Luxembourg and T Denmark63 cases is not inferior to the directive legislation. The timing of the decision only in 2019 can be attributed to the fact that, in income tax law, the distortion in the context of the fundamental freedoms must first be recognized and then challenged, with increased uncertainty as to the outcome of the judgment due to the unwritten nature of the legal principle. In interpreting the concept of abuse, the ECJ is merely following what is already standard practice in other areas of law with a Union dimension. The presentation of the general principle of abuse avoidance in the area of direct taxation can be seen as a late consequence of the lack of harmonization, since the founders of the internal market and the fundamental freedoms could not yet imagine that EU law would have an impact on tax law to this extent. Furthermore, the separation of the legal system into specialized jurisdictions can be cited for the comparatively late presentation of the general principle to prevent abuse. In the absence of a harmonization mandate in the area of direct taxation, the tax courts in the member states have immunized themselves from European law. This immunization has only been broken through with progressive understanding of the effect of European law on national tax systems. For a long time, the area of direct taxation was regarded as a genuinely domestic area. Furthermore, harmonization in the area of indirect taxes has clearly and definitely progressed. In relation to income taxes,

57

Weber-Grellet in Musil and Weber-Grellet, art. 19 TEU, para.13. ECJ: Case C-255/02 Halifax, ECLI:EU:C:2006:121. 59 ECJ: Case C-110/99 Emsland Stärke, ECLI:EU:C:2000:695. 60 Council Directive 2003/48/EC of 3 June 2003 [2003] OJ L157/38. 61 Council Directive 2011/96/EU of 30 November 2011 (recast) [2011] OJ L345/8 amending Directive (EU) 2015/121 [2015] OJ L21/1. 62 Art. 6 of Council Directive (EU) 2016/1164 of 12 June 2016 [2016] OJ L193/1. 63 ECJ: Case C-115/16 N Luxembourg, ECLI:EU:C:2019:134; Cases C-116/16–C-117/16 T Danmark, ECLI:EU:C:2019:135. 58

The Role of the ECJ in Development of International Tax Law 603 the area of indirect taxes has mostly been demarcated and legal developments have not followed in step. The ECJ is also legitimised in its far-reaching interpretation by the fact that the member states virtually demand a more stringent interpretation in sub-areas. In this context, the Cadbury Schweppes case64 should be mentioned, in which the member states wanted the requirement of abuse to be clearly defined by the ECJ. In essence, the member states wanted clear guidance from the ECJ because of doubts about the interpretation. The same is required by the ECJ’s interpretation of the ATAD when questions of doubt must be bindingly clarified by the ECJ. Directive legislation gives the ECJ more powers of interpretation. This is contrasted by the fact that the member states try to relinquish their competences only to a limited extent. An interplay can be seen between the ECJ and the member states, which on the one hand oppose far-reaching interpretation on the basis of tax sovereignty, but on the other hand want to achieve clear interpretations by the ECJ. The influence of the ECJ is not a matter of course. It is strengthened and demanded by the member states’ questions. The competence and legitimacy of the ECJ to interpret are recognized and implemented. In the field of tension of increasing integration, the resistance of critical states bounces off the ECJ insofar as it has the allocation of competence on its side and this cannot be reversed. The influence may once have been misjudged, but is confirmed by projects such as the ATAD in unanimous decision making.

33.7 Influence of ECJ Case Law on Norm-Setting The competence and legitimacy of the ECJ arises expressly for the interpretation of Union law. However, the influence of the ECJ does not stop at the interpretation of already enacted law. The influence goes beyond interpretation into the area of lawmaking and thus begins much earlier. The cases decided by the ECJ must be taken into account in the legislative process of the national legislator as well as the legislator of directives. Through the guidelines that the ECJ sets out in its case law, the legislator is bound by the ECJ’s requirements when creating a tax provision. The role of the ECJ thus does not begin solely after the law is made in the interpretation process, but has an influence on the legislative process. On the one hand, this can be seen in the example of abuse defence. Through article 6 ATAD, the Union legislator has cast the case law of the ECJ into directive law. The case law that preceded the enactment of article 6 ATAD is reflected in the concrete formulation of the provision. Beyond the scope of application of the ATAD, the prevention of abuse is ensured by the unwritten reservation of the right to abuse in primary law. 64

ECJ: Case C-196/04 Cadbury Schweppes, ECLI:EU:C:2006:544.

604 Adrian Cloer According to this, no transposition into national law is required. A written domestic abuse standard is obsolete. This is because even without a written provision, the abuse proviso ‘hovers’ over national law. If abuse provisions are created for clarification or beyond the scope of original Union law, the interpretation of the ECJ must be observed. Through this ‘negative integration’,65 the ECJ has the competence to directly influence the design of a national tax norm. At first glance, this direct influence seems alien. The ECJ’s jurisdiction relates to the upholding of Union law and not to its creation. The interpretation of Union law is almost directly reflected in the formulation of new domestic law by the national legislature. From a German perspective, an example of this development is the anchoring of the motive test in German domestic law (see s. 50d para. 3 German Income Tax Act), which goes back to the ECJ’s demand and requirements.66 Through its requirements, the ECJ directly influences the design of the national tax norm, which is adapted to the ECJ case law after attempts to interpret it in conformity with the directive.67 The trick is that this abuse-avoidance standard takes precedence over the provisions of the double taxation agreements, even if these contain their own abuse clauses. The bilateral agreement takes a back seat and the ECJ asserts its interpretative sovereignty across European borders.

33.8 The Development: Union Monopoly of Interpretation or Even More? The influence of the ECJ has an impact on the formulation of norms and, in the wake of the adoption of norms, on the interpretation of national and Union regulations. If a tax provision is shaped by Union law, the case law of the ECJ must be observed; more far- reaching national considerations take a back seat. The ECJ thus has sovereignty over the interpretation. Due to the primacy of application of Union law, contrary international tax law from the DTAs or in national tax law must be disregarded. The interpretation of the ECJ has monopoly status.

33.8.1 Interpretation in and by Directive Law The ECJ shapes the drive for unification in the EU and takes on the task of unifying the law. Directive law is binding and failure to transpose it threatens infringement proceedings under article 258 TFEU. This is also enforced against the member states, as

65 Van Crombrugge, TNI (2009), 848: beginning negative integration of direct taxes in the second half of the 1980s through ECJ jurisprudence. 66 ECJ: Case C-440/17 GS, ECLI:EU:C:2018:437. 67 Grotherr, NWB (2021), 262.

The Role of the ECJ in Development of International Tax Law 605 has recently been shown against Belgium,68 Romania,69 Portugal,70 Latvia,71 Spain,72 Germany,73 and Luxembourg74 because of the inadequate transposition of the ATAD. Directives and their interpretation by the ECJ build the bridge from unification within the EU to third countries. The ECJ has been given a regulatory framework, not least through the ATAD, within which to extend uniform EU standards in relation to third countries.75 An example of the reach of an EU directive beyond the EU sphere can be seen in the referral to the ECJ for the implementation of the controlled foreign companies’ taxation from article 7 ATAD.76 In German tax law, there is no proof of substance for third countries in the applicable law on attribution taxation. In this context, the German taxation of additional income is on a collision course with the European fundamental freedoms—namely, the freedom of establishment (art. 49, 54 TFEU) and, in third- country situations, with the freedom of movement of capital that has an effect there (art. 63 TFEU). In 2019, the ECJ ruled in the X-GmbH77 case—after Cadbury Schweppes78 and Columbus Container79—now for the third time on the taxation of additions. In the X- GmbH decision, the German additional taxation of interim income of an investment nature, which also applies to micro-participations of less than 1%, was put to the test. The ECJ found that this was fundamentally a prohibited restriction on the free movement of capital. The infringement could be justified on the basis of the general interest and, in particular, to prevent tax evasion and tax avoidance. Ultimately, however, the additional taxation was only compatible with EU law in the case of (purely) artificial arrangements—this, however, also applied within the scope of the free movement of capital. For reasons of proportionality, the taxpayer—similar to the line of case law in the Cadbury Schweppes ruling—must be given the possibility of relief through an (economic) substance or motive test, even in the case of stricter addition taxation in third-country constellations. Nevertheless, opening up the proof of substance to third countries on the basis of the free movement of capital is only required under EU law if (contractual) administrative and regulatory measures exist with the third country in the respective individual case that enable the German tax authorities to carry out a precise review of the taxpayer’s information. It should be noted that if a legal framework exists for the exchange of information within the meaning of ECJ case law with the respective 68

COM of 2 July 2020, INFR(2020)2215 (closed). COM of 24 January 2020, INFR(2020)0106. 70 COM of 24 January 2020, INFR(2020)0096. 71 COM of 24 January 2020, IFRG(2020)0067 (closed). 72 COM of 24 January 2020, INFR(2020)0045. 73 COM of 24 January 2020, INFR(2020)0024. 74 COM of 14 May 2020, INFR(2020)2183. 75 See Council Directive (EU) 2016/1164 of 12 July 2016, recital 12: ‘It is desirable to address situations both in third countries and within the Union’. 76 Haase, DStR (2019), 827; Heckerodt and van Lück, IStR (2020), 857. 77 ECJ: Case C-135/17 X-GmbH, ECLI:EU:C:2019:136. 78 ECJ: Case C-196/04 Cadbury Schweppes, ECLI:EU:C:2006:544. 79 ECJ: Case C-298/05 Columbus Container, ECLI:EU:C:2007:754. 69

606 Adrian Cloer third country, the taxpayer must certainly be put in a position to explain any economic reasons for their investment in the third country concerned in the case of (stricter) addition taxation without being subjected to excessive administrative constraints. The requirements for counter-evidence as such are then found in the ATAD and are to be implemented domestically.

33.8.2 ECJ’s Interpretation of DTAs The ECJ’s interpretative jurisdiction also arises in concrete DTA cases. In principle, DTAs can provide for mutual agreement procedures to resolve double taxation disputes arising from divergences of interpretation in individual cases. Article 25 paragraphs 1 and 2 OECD MC 2017 provides a model for the procedure of the mutual agreement procedure. Although the theory holds out the prospect of a resolution of the dispute, the mutual agreement procedure faces significant limitations. The conflict in the application of the agreement is not necessarily resolved by a mutual agreement procedure; in this respect, there is no obligation to reach an agreement. Thus, the objective of the agreement to prevent double taxation cannot always be achieved through a mutual agreement procedure. Within the procedure, there are also limited procedural rights for the taxpayer. In addition, the time factor of the mutual agreement limits the effectiveness of the procedure. Although there is a deadline for the taxpayer to submit the case to the competent authorities, there is no deadline for the competent authorities to reach an agreement to resolve the dispute. In the absence of a fixed deadline, there is no pressure on the authorities of the contracting states to reach the non- mandatory agreement. The inadequacies of the mutual agreement procedure have not gone undetected by the OECD and, as a consequence, arbitration has been introduced in article 25 paragraph 5 OECD MC. This arbitration procedure is intended to compensate for the limitations of the mutual agreement procedure. To this end, a time limit was added after which the mutual agreement procedure can be followed by arbitration upon request and a binding arbitral award clarifying the dispute is rendered. However, in comparison to mutual agreement procedures, arbitration procedures have found a more reserved entry into the respective DTAs. In addition to the EU Arbitration Convention80 aimed at transfer pricing, another measure was initiated at the EU level to achieve efficient dispute resolution in the form of the Directive on procedures for the settlement of tax disputes in the EU.81 The Multilateral Convention contains an arbitration clause in article 18, which, however, is not mandatory as an optional part.

80 Convention on the elimination of double taxation in connection with adjustment of profits of associated enterprises (90/463/EEC) [1990] OJ L225/10. 81 Council Directive (EU) 2017/1852 [2017] OJ L265/1; see further Van West and Zöhrer in Lang et al. (Pub.), 311.

The Role of the ECJ in Development of International Tax Law 607 A mandatory arbitration clause is contained in article 25 paragraph 5 DTA Germany– Austria,82 according to which the ECJ is assigned jurisdiction to rule on the dispute after a failed mutual agreement procedure on the basis of article 273 TFEU. By conferring jurisdiction under article 273 TFEU, the ECJ resolved a double taxation dispute by way of arbitration for the first time in Case C-648/15.83 The dispute concerned the assessment of a hybrid financial instrument in the light of article 11 of the German–Austrian DTA. Since the conventional mutual agreement procedure failed, Austria appealed to the ECJ. The arbitration clause contained in this DTA is unusual in that it gives the ECJ the power to decide. The legal certainty associated with the appointment of the ECJ as an arbitral body, which occurs not least through reasoned and published decisions, must be emphasized. In addition, the jurisdiction of the ECJ can also have a certain deterrent effect on the member states to reach an agreement independently in the preceding mutual agreement procedure, without having to give the procedure out of their hands afterwards.84The agreement policy of the member states could, as it were, follow the example and make arbitration proceedings pending before the ECJ by means of a corresponding arbitration clause.

33.9 Assessment and Conclusion The interpretation of the ECJ goes over to a restriction of the national legislator’s freedom of design as well as of the directive legislation. Beyond the defence against abuse, the requirement for non-discriminatory legislation must also be considered. The ECJ requires equal treatment of nationals and non-nationals. Insofar as a norm of international tax law is designed to disadvantage foreigners, it will be rejected by the ECJ. This case law is characterized by decisions such as the Lankhorst-Hohorst85 case. The effect, however, is the decisive one: if national provisions of international tax law are overturned on the grounds of discrimination against limited taxpayers, the norm is also applied to nationals. This leads to levelling by annulling the foreign-specific effect and ordering domestic application in order to meet the requirements of a non- discriminatory norm. The ECJ case law leads to increased requirements specifically for cross-border standards. This leads to the finding that there are no longer any special provisions for foreigners in national tax law with a cross-border reference, but the requirements apply equally to domestic and foreigners and are entered into national law. These requirements restrict the national legislator’s freedom to design national standards in international

82

DTA Germany–Austria, 24 August 2000, BGBl. II 2012, 146. ECJ: Case C-648/15 Österreich/Deutschland, ECLI:EU:C:2017:664. 84 Cloer and Niemeyer, FR (2018), 674. 85 ECJ: Case C-324/00 Lankhorst-Hohorst, ECLI:EU:C:2002:749. 83

608 Adrian Cloer tax law. The ECJ’s competence to interpret is transformed into a competence to influence national legislators. This is because already in the preliminary step of interpretation, the norm-setter—the national legislator—takes up the requirements of the ECJ and implements them. The ECJ sets the limits that are to be observed in the legislation of the member states.

Chapter 34

Tax Treatie s a nd E U F undamental Fre e d oms Marjaana Helminen

34.1 Introduction The member states of the European Union have concluded a large number of tax treaties with each other and with states that are not members of the Union. European Union membership does not prevent the member states from concluding bilateral or multilateral tax treaties.1 The right to conclude tax treaties for the purpose of eliminating double taxation within the Union was expressly mentioned in article 293 of the Treaty Establishing the European Community (EC Treaty). A similar provision was no longer included in the Treaty on European Union (TEU)2 or the Treaty on the Functioning of the European Union (TFEU),3 which replaced the EC Treaty. The EU member states, however, still have the right to conclude tax treaties in order to eliminate double taxation and to prevent tax avoidance.4

1 See e.g. Gilly (C-336/96), para. 23. In addition to bilateral tax treaties, the EU member states have concluded a multilateral convention on the elimination of double taxation in connection with the adjustment of profits of associated enterprises (90/436/ETY). See M. Helminen, EU Tax Law—Direct Taxation (Amsterdam: IBFD, 2022), s. 5.4 on this Arbitration Convention. 2 TEU is used to refer to the Treaty on European Union in the form as amended by the Treaty of Lisbon of 13 December 2007, which entered into force on 1 December 2009. 3 The TFEU consists of the EC Treaty as amended and renamed by the Treaty of Lisbon, which was signed on 13 December 2007 and which entered into force on 1 December 2009. 4 According to art. 4(2)(a) TFEU, the Union and the member states share competence in the area of the internal market. See also, e.g., Imfeld and Garcet (C-303/12), para. 41; Sauvage and Lejeune (C- 602/17), para. 22; E. Kemmeren, ‘After Repeal of Article 293 EC Treaty under the Lisbon Treaty: The EU Objective of Eliminating Double Taxation Can Be Applied More Widely’, EC Tax Review 4 (2008), 156– 158; and E. Kemmeren, ‘Double Tax Conventions on Income and Capital and the EU: Past, Present and Future’, EC Tax Review 3 (2012), 167–168.

610 Marjaana Helminen The right to conclude tax treaties covers both tax treaties between the EU member states and tax treaties with non-EU member states.5 Tax treaties concluded by the EU member states are accepted to the extent that they do not prevent or restrict the effective attainment of the objectives of the EU founding treaties, EU directives, or other EU law provisions and principles.6 In addition to the domestic laws of the member states, the primacy of EU law also applies to the tax treaties concluded by each member state.7 The member states must not conclude tax treaties or apply existing tax treaties that are in conflict with EU law,8 and tax treaties must be applied and interpreted in line with EU law.9 The tax treaties concluded between two or more member states and the provisions included in them have validity only insofar as they do not conflict with the provisions or principles of EU law.10 This limited scope of validity applies in the case of old member states to tax treaties concluded both before and after the conclusion of the EU founding treaties. In relation to the new member states, the limited scope of validity applies to tax treaties concluded both before and after they became member states.11 Only old tax treaties concluded between a member state and a non-member state may be effective despite a conflict with EU law.12 Even in these cases, article 351 TFEU obliges the member states to take the appropriate steps to eliminate the incompatibilities and to 5

It is not completely clear as to where the line is, after which the member states’ right to conclude tax treaties shifts exclusively to the Union. See, e.g., G. Kofler, ‘EU Power to Tax: Competence in the Area of Direct Taxation’, in C. Panayi, W. Haslehner, and E. Traversa, eds, Research Handbook on European Union Taxation Law (Cheltenham: Edward Elgar, 2020), 34–49; and B. Terra and P. Wattel, European Tax Law, vol. I, 7th ed. (Alphen aan den Rijn: Kluwer, 2019), s. 2.2.4. Due to the small degree of harmonization or coordination of direct taxes, member states, in addition to the Union, have the right to conclude tax treaties with non-member states on most matters concerning direct taxes. See, e.g., Joined Cases Open Skies (C-446–469/98, C-471–472/98 and C-475–476/98). See also, e.g., Helminen, EU Tax Law—Direct Taxation, s. 1.5.3.1 on the external treaty-making powers of the member states. 6 On the relationship between EU law and tax treaties, see also, e.g., T. O’Shea, EU Tax Law and Double Tax Conventions (London: Avoir Fiscal, 2008); C. H. J. I. Panayi, ‘The Relationship between EU and International Tax Law’, in Panayi, Haslehner, and Traversa, Research Handbook on European Union Taxation Law, 119–141; N. Mattsson, ‘Gränsgångare i Norden. Några synpunkter på skatteavtal och EG- rätt—In memoriam Kari S. Tikka 1944–2006 (Helsinki: Suomalainen lakimiesyhdistys, 2007), 221– 234; M. Hilling, ‘Free Movement and Tax Treaties in the Internal Market’, JIBS Dissertation Series No. 026, Jönköping International Business School (2005), 213–252; Kemmeren, ‘Double Tax Conventions on Income and Capital and the EU’, 157–177; P. Pistone, The Impact of Community Law on Tax Treaties (Alphen aan den Rijn: Kluwer, 2002); H. van den Hurk, ‘Is the Ability of the Member States to Conclude Tax Treaties Chained Up?’, EC Tax Review 1 (2004), 17–30; P. Pistone, ‘Towards European International Tax Law’, EC Tax Review 1 (2005), 4–9; P. Pistone, ‘Tax Treaties and the Internal Market in the New European Scenario’, Intertax 2 (2007), 75–81; and Panzeri (2021), 147–155. 7 See, e.g., Matteucci (235/87); Deserbais (286/86); and Bouanich (C-265/04), para. 51. 8 See, e.g., Commission v. France (avoir fiscal) (270/83), para. 26. 9 Saint-Gobain (C-307/97), para. 57 and Schumacker (C-279/93), para. 21. 10 See, e.g., Saint-Gobain (C-307/97), para. 57. 11 Matteucci (235/87) and Deserbais (286/86). 12 According to art. 351 TFEU, the treaties of the member states concluded with non-EU member states before 1 January 1958 (the entry into force of the EC Treaty), or before the state concerned joined the European Union, are effective despite any conflict with the founding treaties of the European Union.

Tax Treaties and EU Fundamental Freedoms 611 amend the treaties. If an EU member state applies a treaty in conflict with EU law, it may be found liable for damages unless it succeeds in renegotiating the treaty.13 The loyalty principle of article 4 TEU requires that tax treaties be applied in compliance with EU law. Member states are obliged to abstain from any measure that could jeopardize the attainment of the objectives of the founding treaties and are also obliged to resolve any possible conflict in accordance with EU law. The member states must renegotiate the bilateral and multilateral tax treaties that are in conflict with EU law or apply them in accordance with EU law.14

34.2 The Fundamental Freedoms 34.2.1 Relevant Provisions The EU internal market means the abolition of obstacles to the free movement of goods, persons, services, and capital between the EU member states.15 Tax treatment in a member state based on national tax laws or tax treaties may be in conflict with EU law if the treatment entails a restriction on the following fundamental freedoms:

• free movement of goods (art. 34 TFEU); • free movement of persons, including: ⚬ free movement of EU citizens (art. 21 TFEU); ⚬ free movement of workers (art. 45 TFEU); • freedom of establishment (art. 49 TFEU); • freedom to provide or receive services (art. 56 TFEU); and • free movement of capital and payments (art. 63 TFEU).

The fundamental freedoms are considered to be part of the general legal principles of EU law protected by the Court of Justice of the European Union.16 The TFEU articles on the fundamental freedoms may be directly referred to before national courts if national tax treatment is considered to be in conflict with these articles. The same fundamental freedoms also apply in the case of European Economic Area (EEA) states on the basis of the Agreement Creating the European Economic Area (EEA Agreement).17 13

See Helminen, EU Tax Law—Direct Taxation, s. 7.4.2. on the liability for damages. See also, e.g., C. H. J. I. Panayi, ‘The Effect of Community Law on Pre-Accession Tax Treaties’, EC Tax Review 3 (2007), 121–132. See also Helminen, EU Tax Law—Direct Taxation, s. 1.5 on the right of the EU member states to conclude and apply tax treaties. 15 Arts 3–6 TEU. 16 Stauder (29/69). 17 The EEA Agreement provides nationals of the EEA states (Iceland, Liechtenstein, and Norway) with the same fundamental freedoms as the TFEU does with EU nationals. Art. 4 of the EEA Agreement provides for the free movement of goods, art. 28 for the free movement of workers, art. 31 for the freedom 14

612 Marjaana Helminen Principles similar to the fundamental freedoms do not apply in relation to non-EU/ EEA member states unless similar principles are included in separate agreements with these states.18 The article of the TFEU on free movement of capital and payments is an exception and also applies (in addition to intra-EU situations) to the movement of capital or payments between EU member states and non-member states. The other fundamental freedoms may only have indirect relevance in relation to tax treaties concluded with non-member states, for example via a permanent establishment situated in the European Union.19

34.2.2 Discrimination and Restrictions The TFEU articles on the fundamental freedoms provide for the right to free movement from one EU member state to another and for equal treatment regardless of nationality or origin. Both direct and indirect discrimination on the basis of nationality are prohibited. EU nationals have the right to be treated on an equal footing with nationals in each member state. To a certain extent, in addition to vertical discrimination, horizontal discrimination may also be prohibited. For example, enterprises have the right to choose whether to establish themselves in other EU member states in the form of a subsidiary or in the form of a branch.20 The extent to which the EU Court considers the TFEU to prohibit horizontal discrimination, however, is somewhat unclear. The Court, for example, has not required the same tax treaty benefits to apply to all EU nationals irrespective of their state of residence.21 A tax provision or tax practice based on the national tax system of an EU member state, or on a tax treaty concluded by an EU member state that subjects domestic taxpayers and taxpayers with the nationality of another member state to different tax treatment, may be discriminatory in a way prohibited by the fundamental freedoms. For example, tax treaties must not include provisions that make tax treaty benefits available only to the nationals of the contracting EU member state and not to the nationals of the other member states.22 of establishment, art. 36 for the freedom to provide services, and art. 40 for the free movement of capital. The EEA Agreement, however, does not include an article similar to art. 21 TFEU on the free movement of EU citizens. This shortcoming may be relevant, e.g., in the case of students or retired persons who are not present in another state in order to work or do business there. See also, e.g., Terra and Wattel, European Tax Law, s. 3.2 and Helminen, EU Tax Law—Direct Taxation, ch. 2 on the fundamental freedoms. 18

See, e.g., Wächtler (C-581/17); Ettwein (C-425/11); Radgen (C-478/15); Picart (C-355/16). See also, e.g., R. Fontana, ‘Conference Report: The EU and Third Countries: Direct Taxation, 13–14 October 2006, Vienna’, Intertax 10 (2007), 590–592. 20 Commission v. France (avoir fiscal) (270/83) and Compagnie de Saint-Gobain (C-307/97). 21 See D (C-376/03) and Helminen, EU Tax Law—Direct Taxation, s. 1.5.2.3, and s. 2.1.2. on the TFEU concept of discrimination. 22 Compare to Gottardo (C-55/00) and Joined Cases Open Skies (C-446–469/98, C-471–472/98 and C- 475–476/98) for cases on areas of law other than tax law. 19

Tax Treaties and EU Fundamental Freedoms 613 The type of discrimination prohibited by the fundamental freedoms can arise through the application of different rules to nationals of different EU member states in a comparable situation, or through the application of the same rules to nationals of different member states in a different situation, with the consequence that one of the nationals is subject to worse treatment.23 In addition to direct and overt discrimination by reason of nationality, covert and indirect discrimination on the basis of nationality is also prohibited. Discrimination which, by the application of other criteria of differentiation, such as the state of residence of a taxpayer, effectively leads to the same result as discrimination based on nationality is also prohibited.24 Article 65 TFEU expressly mentions that different tax treatment must not result in arbitrary discrimination or a disguised restriction on the free movement of capital or payments. According to the EU Court, indirect discrimination or a disguised restriction, however, may also be prohibited in the case of the other fundamental freedoms. The wording of articles 45 and 49 TFEU seems to refer only to discrimination based on different treatment on the grounds of nationality. Based on the case law of the EU Court, it is clear, however, that all of the TFEU articles on the fundamental freedoms prohibit not only discrimination but also such restrictions on the freedoms that do not constitute discrimination.25 As a result of the fundamental freedoms, taxation that restricts the use of these freedoms may be in conflict with the TFEU even where the treatment would not constitute discrimination.26 The tax treatment in a member state may be in conflict with the fundamental freedoms even if the treatment does not distinguish between a purely internal situation and a cross-border situation, if the treatment entails a restriction on a cross-border activity in some other way. All measures that prohibit, impede, or render less attractive the exercise of the fundamental freedoms are regarded as restrictions.27 For example, a tax treatment

23

See, e.g., Schumacker (C-279/93), para. 30; Gschwind (C-391/97), para. 21; Wielockx (C-80/94), para. 17; Asscher (C-107/94), para. 40; Talotta (C-383/05), para. 18; and Royal Bank of Scotland (C-311/97), para. 26. On the concept of discrimination, see also, e.g., A. Cordewener, ‘The Prohibitions of Discrimination and Restriction within the Framework of the Fully Integrated Internal Market’, in F. Vanistendael, ed., EU Freedoms and Taxation (Amsterdam: IBFD, 2006), 1–46. 24 See, e.g., Biehl (175/88), paras 13 and 14; Commission v. Greece (C-155/09); Schumacker (C-279/93), para. 26; and Talotta (C-383/05), para. 17. 25 See, e.g., Daily Mail (81/87); Bosman (C-415/93); Gebhard (C-55/94); and Safir (C-118/96). 26 See, e.g., Safir (C-118/96); Baars (C-251/98); and Daily Mail (81/87). In this respect, the scope of application of the TFEU articles on the fundamental freedoms is much broader than the scope of application of art. 18 TFEU, prohibiting only discrimination based on nationality. 27 Gebhard (C-55/94), para. 37 and Krankenheim Ruhesitz am Wannsee-Seniorenheimstatt (C-157/07), para. 30.

614 Marjaana Helminen that prevents or restricts a taxpayer from moving to or establishing in another member state may be in conflict with the TFEU.28 The treatment may be in conflict with the fundamental freedoms even if the tax burden itself is no higher but the cross-border situation is subject to more burdensome procedural or administrative requirements.29 Moreover, a tax provision that applies without distinction to domestic and cross-border situations may be in conflict with the fundamental freedoms if it imposes conditions that are difficult or impossible to fulfil in cross- border situations.30 The TFEU articles on the fundamental freedoms prohibit not only discrimination against nationals of the other EU member states compared to a country’s own nationals, but also a country’s tax treatment that results in a restriction on the use of the fundamental freedoms by its own nationals. For example, the general provision on the free movement of EU citizens, the article on the free movement of workers, and the article on the freedom of establishment give an EU national the right to leave their country of nationality and move to another member state. The tax systems of member states, including tax treaties, must not prevent or restrict EU nationals from moving to other member states.31 There are a large number of cases in which the EU Court has taken a stand on the compatibility of different tax treaty provisions with the TFEU fundamental freedoms. The following sections discuss selected issues related to the compatibility between the two.32

28 See

also, e.g., A. Zalasinski, ‘The Limits of the EC Concept of “Direct Tax Restriction on Free Movement Rights”, the Principles of Equality and Ability to Pay, and the Interstate Fiscal Equity’, Intertax 5 (2009), 282–297 and F. Vanistendael, ‘The Compatibility of the Basic Economic Freedoms with the Sovereign National Tax Systems of the Member States’, EC Tax Review 3 (2003), 136–143 on the difference between discriminatory and restrictive treatment. 29 See, e.g., Schumacker (C-279/93), para. 49; Futura (C-250/95); and Vestergaard C-55/98), para. 21. See also NN (C-48/15), which concerned a penalty in the case of a failure to file an annual declaration or the failure to pay certain tax. 30 See, e.g., Van der Weegen and Pot (C-580/15), in which the remuneration of a savings account must consist of both basic interest and a fidelity premium. See also Zalasinski, ‘Tax Rules Applicable without Distinction and the EU Internal Market Freedoms: An Analysis of Recent Case Law Regarding Taxation of Investment Income’, European Taxation 12 (2017), 533–543 for how a tax provision may be in conflict with the fundamental freedoms even if it applies without distinction. 31 See, e.g., Baars (C-251/98), para. 28; Daily Mail (81/87); Überseering (C-208/00); Cartesio (C-210/ 06), paras 112–113; Schilling (C-209/01); X AB and Y AB (C-200/98); Asscher (C-107/94); ICI (C-264/ 96); Pusa (C-224/02), para. 19; De Baeck (C-268/03); and Turpeinen (C-520/04), para. 22, and, e.g., D. Weber, ‘Exit Taxes on the Transfer of Seat and the Applicability of the Freedom of Establishment after Überseering’, European Taxation 10 (2003), 350–354. 32 For examples of the types of tax treaty provisions that are problematic from the perspective of EU law, see also, e.g., A. de Graaf, ‘Designing an Anti-Treaty Shopping Provision: An Alternative Approach’, EC Tax Review 1 (2008), 12–23.

Tax Treaties and EU Fundamental Freedoms 615

34.3 Selected Issues Concerning Tax Treaties 34.3.1 Allocation of Taxing Rights EU member states have the right to conclude tax treaties for the purposes of allocating taxing rights between and among them.33Allocation of taxing rights between the member states in tax treaties is in accordance with EU law even where the tax treatment in one of the contracting states would be more burdensome than in the other.34 It is not in conflict with the TFEU fundamental freedoms whereby, because of a tax treaty provision, a taxpayer falls under the scope of the taxing powers of the country with the more burdensome tax treatment. A member state is not required to draw up its tax rules on the basis of those of another member state in order to ensure in all circumstances taxation that removes disparities arising from national tax rules.35 The use of tax treaties for the purposes of eliminating or abolishing international double taxation and tax avoidance is desirable from the perspective of the EU internal market. Member states have the right to decide the criterion upon which the allocation is made.36 According to the EU Court, the member states must take the necessary actions in order to prevent international double taxation, particularly applying the allocation rules based on international tax practice, including, for example, the allocation principles based on the OECD Model Tax Convention.37 This right may include, for example, the application of the territoriality principle of international tax law.38

33 Gilly (C-336/96); Saint-Gobain (C- 307/ 97), para. 56; and Kerckhaert and Morres (C-513/04), para. 23. 34 See Gilly (C-336/96), paras 24, 30, and 46; Bukovansky (C-241/14), para. 44; HB and IC (Joined Cases C-168/19 and C-169/19), para. 17; N (C-470/04), para. 44; Kerckhaert and Morres (C-513/04), paras 22 and 23; Oy AA (C-231/05), para. 52; and Test Claimants in the Thin Cap Group Litigation (C-524/04), para. 49. 35 Deutsche Shell (C- 293/06), para. 43 and Krankenheim Ruhesitz am Wannsee-Seniorenheimstatt (C-157/07), para. 50. See also Sauvage and Lejeune (C-602/17), para. 28 on how disparities arising from national rules and division of taxing rights in tax treaties as such do not mean discrimination or forbidden different treatment. 36 See, e.g., Gilly (C-336/96), paras 24 and 30; Bouanich (C-265/04), para. 49; Arens-Sikken (C-43/07), para. 62; Saint-Gobain (C-307/97), paras 56–57; Scorpio (C-290/04), para. 54; Test Claimants in Class IV of the ACT Group Litigation (C-374/04), para. 52; Oy AA (C-231/05), para. 52; Krankenheim Ruhesitz am Wannsee-Seniorenheimstatt (C-157/07), para. 48; and van Hilten (C-513/03), para. 47. 37 Gilly (C-336/96), para. 31; N (C-470/04), paras 45 and 49; Kerckhaert and Morres (C-513/04), para. 23; and Test Claimants in the Thin Cap Group Litigation (C-524/04), para. 49. 38 See, e.g., N (C-470/04), paras 41 and 42.

616 Marjaana Helminen

34.3.2 Most-Favoured-Nation Principle The EU founding treaties do not expressly include a most-favoured-nation clause that would require member states to make the benefits of a bilateral tax treaty available to residents of the member states that are not parties to the treaty.39 Some authors, however, stress the relevance of horizontal discrimination.40 Opinions have been expressed according to which there may be a conflict with the TFEU non-discrimination concept when the nationals of a member state residing in certain member states have the right to better tax treaty benefits than the nationals of the same state residing in another member state.41 The EU Court, however, held in D that the free movement of capital does not oblige member states to make the benefits based on a bilateral tax treaty available to the residents of member states that are not parties to the treaty.42 The obligation to make the benefits of a bilateral tax treaty available to residents of other member states was also dealt with in Test Claimants in Class IV of the ACT Group Litigation.43 In this case, the EU Court held that the fact that the benefits of a bilateral tax treaty are not made available to residents of other member states is not in conflict with the founding treaties. The dividend imputation credit based on a bilateral tax treaty did not need to be made available to the residents of such member states that had not concluded a tax treaty with a provision providing for the credit.44 The different treatment, however, must not be based on nationality. The tax benefits based on a tax treaty must be available to the nationals of all member states under the same circumstances.45

39 On the most-favoured-nation principle, see also, e.g., Hilling, ‘Free Movement and Tax Treaties in the Internal Market’, 271–290; A. Cordewener and E. Reimer, ‘The Future of Most-Favoured-Nation Treatment in EC Tax Law: Did the ECJ Pull the Emergency Brake without Real Need?, Part 1’, European Taxation 6 (2006), 239–249; S. van Thiel, ‘Why the ECJ Should Interpret Directly Applicable European Law as a Right to Intra-Community Most-Favoured-Nation Treatment, Part 1’, European Taxation 6 (2007), 263–269, Van Thiel, ‘Why the ECJ Should Interpret Directly Applicable European Law as a Right to Intra-Community Most-Favoured-Nation Treatment, Part 2’, European Taxation 7 (2007), 314–327; S. van Thiel, ‘Why the European Court of Justice should Interpret Directly Applicable Community Law as a Right to Most-Favoured Nation Treatment and a Prohibition of Double Taxation. The Influence of European Law on Direct Taxation’, in D. Weber, ed., Recent and Future Developments (Alphen aan den Rijn: Kluwer, 2007), 75–138; and A. Cordewener and E. Reimer, ‘The Future of Most-Favoured-Nation Treatment in EC Tax Law: Did the ECJ Pull the Emergency Brake without Real Need?, Part 2’, European Taxation 7 (2006), 291–306. 40 See, e.g., Cordewener (2007), 210–212 on the different dimensions of non-discrimination. 41 See, e.g., van Thiel, ‘Why the ECJ Should Interpret Directly Applicable European Law as a Right, Part 1’, 263–269 and van Thiel, ‘Why the ECJ Should Interpret Directly Applicable European Law as a Right, Part 2’, 314– 327. See also, e.g., Joined Cases Roders (C-367/93–377/93); Matteucci (235/87); Kortmann (32/80); and Gottardo (C-55/00) on areas of law other than direct taxation. 42 See esp. D (C-376/03), paras 61–63. 43 (Case C-374/04). 44 See also Orange European Smallcap Fund (C-194/06), paras 49–51. 45 See, e.g., Gottardo (C-55/00) on social security contributions.

Tax Treaties and EU Fundamental Freedoms 617 The case Riskin and Timmermans concerned a situation where different bilateral tax treaties of a state led to a situation in which a resident of the state could be subject to less favourable tax treatment when investing in certain EU member states compared to an investment in a non-member state. The Court concluded that such different treatment based on different tax treaties means a restriction of the free movement of capital.46 This restriction, however, is not in conflict with the free movement of capital because the situation of the person investing in the other EU member state and the person investing in a third country is not comparable in the light of the tax provisions concerned.47 Based on the existing judgments, it seems that the EU Court does not consider the fundamental freedoms to include a most-favoured-nation principle that would hinder member states from concluding different bilateral tax treaties with different member states.48 This legal situation is somewhat unsatisfactory because the EU Court seems to have applied the most-favoured-nation principle in cases concerning treaties other than tax treaties.49

34.3.3 Methods for the Elimination of Double Taxation Regarding situations not harmonized by EU law, the member states are not obliged to eliminate international double taxation caused by simultaneous taxation in two or more states.50 If the member state, however, chooses to eliminate double taxation, it must be done in compliance with the TFEU fundamental freedoms. According to the EU Court, it is the task of the member states to take the measures necessary to prevent double taxation by applying, in particular, the apportionment criteria followed in international tax practice, including the OECD Model Tax Convention.51 Member states have the right to choose whether to apply the credit method or the exemption method for the purpose of eliminating international double taxation.52 The method chosen, however, must be applied in such a way that it does not lead to worse treatment of cross-border situations compared to pure domestic situations in the state 46 See Riskin and Timmermans (C-176/15), para. 25. 47

See ibid., paras 32–35. however, L. M. De Groot, ‘Member States Must Apply Most Favoured Nation Treatment under EU Law’, Intertax 6/7 (2014), 405–415, who differentiates between bilateral most-favoured-nation treatment and unilateral most-favoured-nation treatment, and considers that the EU Court case law seems to oblige member states to enact unilateral most-favoured-nation treatment. 49 e.g. Joined Cases Roders (C-367/93–377/93); Matteucci (235/87); Kortmann (32/80); and Gottardo (C-55/00). 50 Columbus Container Services (C- 298/05), para. 51; Block (C-67/08), para. 31; Orange European Smallcap Fund (C-194/06); para. 42, Damseaux (C-128/08), para. 27; and CIBA (C-96/08), paras 27–29. 51 Gilly (C-336/96), para. 31; N (C-470/04), para. 45; Kerckhaert and Morres (C-513/04), para. 23; and Test Claimants in the Thin Cap Group Litigation (C-524/04), para. 49. 52 Gilly (C- 336/96); Joined Cases Haribo (C-436/08 and C-437/08), para. 104; Test Claimants in Class IV of the ACT Group Litigation (C-374/04); and Kronos (C-47/12), paras 66–68. See also van Thiel, ‘Why the ECJ Should Interpret Directly Applicable European Law as a Right, Part 1’, 268. 48 See,

618 Marjaana Helminen concerned. Other neutrality deviations that are a consequence of the differences in the tax systems of member states can be eliminated only through a broader scope of harmonization of the different tax systems.53 Some authors, however, have expressed opinions according to which not even the credit method should be accepted, even where applied in a non-discriminatory manner, but rather that the exemption method should be applied to eliminate international double taxation.54 The EU Court, however, has not taken a stand on the question of whether the elimination of double taxation should be carried out in accordance with the principle of capital import neutrality or in accordance with the principle of capital export neutrality.55 The EU Court has held that even though the application of the credit method may have the effect of rendering the pursuit of the activities of a taxpayer more expensive than the application of the exemption method, this does not necessarily mean that the application of the credit method is in conflict with the TFEU.56 If the application of the credit method does not lead to any tax disadvantage to non-residents compared to residents, there is no discrimination in conflict with the TFEU.57 The EU Court has pointed out that, for example in the case of dividends, the credit method may have to include the possibility of a carry-forward of the credit in a loss year in order for the credit method to be acceptable in a cross-border situation if the exemption is applied in a domestic situation.58 There are also other situations in which the credit method leads to worse treatment of a cross-border situation when the exemption method is applied in a comparable 53

The EU Court has, e.g., accepted the application of the exemption method with a progression clause provided that the application does not lead to worse treatment in cross-border situations compared to a pure domestic situation. See, e.g., Schulz-Delzers (C-240/10), paras 35–43 and Verest and Gerards (C-489/ 13). 54 See Terra and Wattel, European Tax Law, s. 16.2.5. 55 From the perspective of capital export neutrality, tax treatment is neutral if it does not have an effect on whether a taxpayer makes investments in their state of residence or abroad. The application of the credit method for the purposes of eliminating international double taxation reflects capital export neutrality. From the perspective of capital import neutrality, tax treatment is neutral if it is irrelevant from a tax point of view whether an income recipient or an investor is in the source state or in any other state. Capital import neutrality is achieved if the tax burden on the capital invested in a certain state and the income from the capital is the same regardless of whether the investor is a resident or a non- resident. The application of the exemption method for the purposes of eliminating international double taxation in the residence state reflects capital import neutrality. See F. Vanistendael, ‘The Functioning of Fundamental Freedoms and Tax Neutrality in the Internal Market’, in Panayi, Haslehner, and Traversa, Research Handbook on European Union Taxation Law, 142–161 for the neutrality concept and the right of establishment in the European Union. 56 Columbus Container Services (C-298/05), para. 38. See also Decision of the German Federal Tax Court (BFH) I R 114/08 issued on 21 October 2009 and published on 13 January 2010 for the follow-up decision based on the Columbus Container Services judgment. According to the German Federal Tax Court, the German switchover clause was in conflict with EU law because it did not provide for an economic substance test. 57 Columbus Container Services (C-298/05), para. 40. 58 Joined Cases Haribo (C-436/08 and C-437/08), para. 173.

Tax Treaties and EU Fundamental Freedoms 619 domestic situation. The EU Court has pointed out that in certain situations the application of the credit method will mean higher taxes for the recipient of foreign-source income compared to the recipient of domestic-source income subject to the exemption system. In these situations, the credit method and the exemption method are not equivalent.59 The credit method may lead to worse treatment of a cross-border situation, for example in the case of the dividend imputation credit system, if the company distributing the dividend is subject to lower taxation of its profits, inter alia because of a lower nominal rate of tax compared to the tax of the dividend recipient.60 In such a situation, the lower tax applied to the dividend distributor will be the final tax consequence in the case of the exemption method, while the higher tax rate applied to the dividend recipient is the final tax consequence in the credit method. Even though the nominal tax rate applied to the dividend distributing company and the dividend recipient would be the same, the effective level of taxation of the dividend distributing entity may be lower. The effective rate of taxation may be lower because of different reliefs reducing the tax base. Unlike the exemption method, the credit method does not enable the benefit of the corporation tax reductions granted at an earlier stage to the company paying dividends to be passed on to the shareholder.61 In many cases, the application of the credit method to cross-border situations, while applying the exemption method to domestic situations, leads to taxation in conflict with the TFEU fundamental freedoms.62 Furthermore, however, the exemption method has to be applied in a manner that does not lead to a conflict with the fundamental freedoms. The fact that costs related to exempt income may not be deductible may lead to a conflict, especially if the exemption method is applied with a progression clause. Even though the exemption method with a progression clause may be applied instead of the credit method, it must not be applied in a manner that leads to a conflict with the fundamental freedoms.63 In conclusion, the TFEU fundamental freedoms do not require the member states to choose a certain method for the elimination of double taxation in their bilateral tax treaties. The member states are free to decide whether to apply the credit method or the

59

Test Claimants in the FII Group Litigation 2 (C-35/11), para. 43. Ibid., para. 44. 61 See ibid., paras 46–49. 62 See ibid., para. 52. 63 The EU Court has, e.g., accepted the application of the exemption method with a progression clause, provided that the application does not lead to worse treatment in cross-border situations compared to a comparable domestic situation. See, e.g., Schulz-Delzers (C-240/10), paras 35–43; and Verest and Gerards (C-489/13). E.g. in Bechtel (C-20/16), pension contributions paid in the employment state had to be deductible in the residence state where the residence state applied the exemption method with a progression clause and where similar payments made to the residence state would have been deductible. See Bechtel (C-20/16), para. 80. 60

620 Marjaana Helminen exemption method in their tax treaties. The chosen method, however, must be applied in compliance with the TFEU fundamental freedoms.64

34.3.4 Tax Treaties and Justifications for Discrimination or Restrictions The evaluation under the TFEU fundamental freedoms articles requires examining first whether the two relevant situations are comparable in order for different treatment not to be allowed. After it has been established that there are two comparable situations, which are treated differently65 and that forbidden discrimination or restriction exists, it must be evaluated whether or not the treatment is justified. Discrimination, or restriction of the fundamental freedoms, may be allowed on certain justification grounds. Based on EU Court case law on the so-called rule of reason principle, indirect and covert discrimination or a national tax provision resulting in a restriction on one or more of the basic freedoms may be justified provided that the provision is not applied in a discriminatory manner. Tax treatment that constitutes a restriction on the basic freedoms may be justified on the basis of the rule-of-reason principle if the tax treatment has an objective that is in accordance with the TFEU, and which is justified by an overriding reason in the public interest.66 For the purposes of the rule-of-reason principle justifying restrictive tax treatment, it is necessary that the tax treatment is appropriate for ensuring the attainment of the objective in question, and that the tax treatment does not go beyond that necessary to attain the objective. Tax provisions that constitute a restriction on the fundamental freedoms are not accepted if there is a measure available to reach the same objective in a less restrictive manner. In any case, the restrictive tax provision must be in accordance with the proportionality principle, namely the restriction must not be unreasonable in comparison with the objective reached with the measure.67 Tax treaty provisions may have relevance in analysing the existence of a rule-of-reason justification for tax treatment amounting to discrimination or a restriction on the fundamental freedoms. The EU Court has, for example, accepted the need to safeguard the

64 It is important to note that in the case of dividends, interest income, and royalties, the EU Corporate Tax Directives may require a certain method to be applied. 65 A situation in which two different situations are treated in the same way may also constitute discrimination. See, e.g., Schumacker (C-279/93), para. 30. 66 Cassis de Dijon (120/78). 67 See, e.g., Dassonville (8/74); Cassis de Dijon (120/78); Gebhard (C-55/94); Futura (C-250/95), para. 26; Verkooijen (C-35/98), para. 43; Lankhorst-Hohorst (C-324/00), para. 33; X and Y (C-436/00), para. 49; de Lasteyrie du Saillant (C-9/02), para. 49; Marks & Spencer (C-446/03), para. 35; Cadbury Schweppes (C-196/04), para. 47; Test Claimants in the Thin Cap Group Litigation (C-524/04), para. 64; Krankenheim Ruhesitz am Wannsee-Seniorenheimstatt (C-157/07), para. 40; Lidl (C-414/06), para. 27; and Oy AA (C- 231/05), para. 44. See Google (C-482/18), para. 49 on how the amount of the fines incurred in the event of failure to comply with the obligation to submit a tax declaration may be disproportionate.

Tax Treaties and EU Fundamental Freedoms 621 balanced allocation of taxing rights between the member states as a justification for tax treatment constituting a restriction on the fundamental freedoms.68 A tax treaty, however, may prevent a state from relying on the balanced allocation of taxing rights justification in a situation covered by the treaty. The EU Court has mentioned that if the issue of the allocation of powers of taxation between two states is governed by a double taxation convention which, in accordance with the OECD Model, determines the exclusive right to tax for each of the contracting states, the states cannot rely on the justification. As a state in such a situation has abandoned its powers of taxation concerning a certain situation, the state cannot rely on a balanced allocation of powers of taxation in order to levy a specific tax on the basis that the taxpayers are not subject to its tax jurisdiction in the situation.69 On the other hand, a tax treaty may create the required coherence based on which the balanced allocation of taxing rights provides for a justification. For example, the exemption system based on a tax treaty may allow the justification based on the symmetry that losses are not deductible in situations in which the corresponding profits cannot be taxed.70 In this way, the balanced allocation of taxing rights justification is directly linked to symmetry and the coherence of the tax system justification. The requirements of the balanced allocation of powers of taxation and coherence of the tax system justification coincide.71 The need to safeguard the cohesion or coherence of the national tax system was first accepted as a justification for restrictive income tax treatment in Bachmann.72 The case concerned the Belgian tax system, under which only insurance premiums paid in Belgium and not in other member states were deductible. However, at the same time, the insurance premium payments were always deductible if the income related to the payment was taxable in Belgium and the payments were non-deductible if the income related to the payments was tax exempt in Belgium. The right to a tax deduction and the taxability of the income had a direct connection. Instead, a restrictive tax burden cannot be justified by a tax benefit that has no direct link with the tax burden.73 Since the Bachmann judgment, the EU Court has been very reluctant to accept the cohesion argument as a justification for tax treatment constituting a restriction on the basic 68 See, e.g., Marks & Spencer (Case C-446/03); Oy AA (Case C-231/05); Lidl (Case C-414/06); Glaxo Wellcome (Case C-182/08); SGI (Case C-311/08); and X Holding (Case C-337/08). 69 See, e.g., Familienprivatstiftung Eisenstadt (C-589/13), para. 71. 70 See Timac Agro Deutschland (C- 388/ 14), para. 38, which concerned recapture of previously deducted losses of a permanent establishment in a situation in which a bilateral treaty prevented the state of residence of the company from taxing profits related to the permanent establishment. 71 See National Grid Indus (C-371/10), para. 80 and Timac Agro Deutschland (C-388/14), para. 47. See also, e.g., Helminen, EU Tax Law—Direct Taxation, s. 2.3.4 for the balanced allocation of taxing rights justification. 72 See Bachmann (C-204/90), paras 21–28. See also Commission v. Belgium (C-300/90), para. 21. See also Helminen, EU Tax Law—Direct Taxation, s. 2.3.5 on the coherence justification. 73 See, e.g., Commerzbank (C-330/91); Asscher (C-107/94); and Danner (C-136/00) concerning the tax system of Finland under which a direct link was not considered to exist, for which the cohesion of the tax system did not justify restrictive tax measures. See also Skandia and Ramstedt (C-422/01), paras 32–36.

622 Marjaana Helminen freedoms.74 The cohesion justification, however, has been accepted provided that the direct link has been considered to exist.75 In a tax treaty situation, it is not sufficient to evaluate only the coherence of a tax system from the perspective of national legislation—the effect of the applicable tax treaty must also be taken into account.76 The denial of a tax deduction only in the case of non-resident taxpayers may be in conflict with the fundamental freedoms if the member state concerned has voluntarily given up its taxing rights on the corresponding income in a tax treaty.77 On the other hand, a tax treaty may create the required coherence. For example, the exemption system based on a tax treaty may allow the coherence justification based on the fact that losses are not deductible in situations in which the corresponding profits cannot be taxed.78

34.3.5 Anti-Tax-Avoidance Provisions The purpose of tax treaties is not only to eliminate international double taxation but— increasingly—also to prevent tax avoidance. The 2017 update of the OECD Model Tax Convention and the Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting (MLI) established a minimum standard regarding measures against treaty shopping. Combating tax avoidance is also in the interest of the EU internal market. The EU member states, however, must not apply national or treaty-based anti-tax-avoidance provisions in conflict with the fundamental freedoms. The fundamental freedoms limit the scope of anti-tax-avoidance provisions that may be included in tax treaties of the EU member states. It is not always clear, however, as to where the line is drawn between acceptable tax planning and tax avoidance that cannot be accepted from the perspective of the internal market and the fundamental freedoms.79 74 See, e.g., Wielockx (C-80/94); Baars (C-251/98); Verkooijen (C-35/98); Bosal Holding (C-168/01); and Manninen (C-319/02). 75 See, e.g., Commission v. Hungary (C-253/09), paras 70–85 and Commission v. Belgium (C-250/08), paras 70–82. See also D. Weber, ‘An Analysis of the Past, Current and Future of the Coherence of the Tax System Justification’, EC Tax Review 1 (2015), 43–54 for an analysis of the past, current and future of the coherence of the tax system as justification. 76 See X and Y (C-436/00), para. 53. 77 See Wielockx (C-80/94), para. 24. 78 See Timac Agro Deutschland (C- 388/ 14), para. 41, which concerned recapture of previously deducted losses of a permanent establishment in a situation in which a bilateral treaty prevented the state of residence of the company from taxing profits related to the permanent establishment. 79 See also, e.g., N. Vinther and E. Werlauff, ‘Tax Motives are Legal Motives: The Borderline between the Use and Abuse of the Freedom of Establishment with Reference to the Cadbury Schweppes Case’, European Taxation 8 (2006), 383–386; L. De Broe, ‘Some Observations on the 2007 Communication from the Commission: “The Application of Anti-Abuse Measures in the Area of Direct Taxation within the EU and in Relation to Third Countries” ’, EC Tax Review 3 (2008), 142–148; A. Zalasinski, ‘Some Basic Aspects of the Concept of Abuse in the Tax Case Law of the European Court of Justice’, Intertax 4 (2008), 156–167; J. Schwarz, ‘Abuse and EU Tax Law’, Bulletin for International Fiscal Documentation 7 (2008), 289–293; R. Karimeri, ‘A Critical Review of the Definition of Tax Avoidance in the Case Law

Tax Treaties and EU Fundamental Freedoms 623 EU nationals are free to choose the country for their activities that is the most beneficial from a tax perspective, including tax treaty benefits.80 This type of tax planning is acceptable because the tax systems of the member states have not been harmonized. The fact that the taxes in one state are lower than in another does not give the state with the higher taxes the right to adopt measures that constitute a restriction on the TFEU basic freedoms.81 The loss of tax revenue has not been regarded as constituting a reason that could justify a restriction on the fundamental freedoms.82 The member states, for example, do not have the right to tax away the benefit enjoyed by a subsidiary in a low-tax member state by taxing the parent company more burdensomely.83 The member states do not have the right to attack harmful tax competition on their own initiative by denying the rights based on the fundamental freedoms.84 In Halifax, concerning value-added tax, the EU Court held that a transaction, or a series of transactions, the essential aim of which is to evade taxes, constitutes abuse and is contrary to EU law.85 In such a situation, tax can be levied based on the situation that would have existed if the transaction had not taken place.86 A member state must refuse to grant the benefits of provisions of EU law if they are relied upon not with a view to achieving the objectives of those provisions but with the aim of benefiting from an advantage in EU law, although the conditions for benefiting from that advantage are fulfilled only formally.87 In such a situation, a tax benefit based on EU law must be denied even if the domestic law of the state concerned does not include an express provision allowing the denial.88 The principle that abusive practices are

of the European Court of Justice’, Intertax 6 (2011), 296–316; and M. Gregoriou, ‘Tax Avoidance with A Fragmentated Single Market’, EC Tax Journal 3 (2008), 33–47 on the difference. 80 See, e.g., ICI (C-264/96), para. 26; Barbier (C-364/01); Centros (C-212/97); Inspire Art (C-167/01); Lankhorst-Hohorst (C-324/00); and Eurowings (C-294/97). 81 See, e.g., Eurowings (C-294/97), para. 44; Cadbury Schweppes (C-196/04), paras 36–38; and Lenz (C-315/02). See, e.g., C. H. J. I. Panayi, ‘Treaty Shopping and Other Tax Arbitrage Opportunities in the European Union: A Reassessment, Part 1’, European Taxation 4 (2006), 104–110 and C. H. J. I. Panayi, ‘Treaty Shopping and Other Tax Arbitrage Opportunities in the European Union: A Reassessment, Part 2’, European Taxation 4 (2006), 139–155 on tax planning based on choosing the right EU member state, and on the extent to which such planning is acceptable. 82 See, e.g., Danner (C-136/00) and Skandia and Ramstedt (C-422/01). 83 See Commission v. France (avoir fiscal) (270/83); Eurowings (C-294/97); and Skandia and Ramstedt (C-422/01). 84 See Commission, Communication from the Commission to the Council, the European Parliament and the European Economic and Social Committee: The application of anti-abuse measures in the area of direct taxation—within the EU and in relation to third countries, COM(2007)785 (2007). 85 Halifax (C-255/02), para. 68. 86 See also Centros (C-212/97), para. 24 (a corporate law case, not a tax case). 87 Joined Cases N Luxembourg et al. (C-115/16, C-118/16, C-119/16 and C-299/16), paras 98 and 120, and Joined Cases T Danmark et al. (C-116/16 and C-117/16), paras 72 and 92. 88 Joined Cases N Luxembourg et al. (C-115/16, C-118/16, C-119/16 and C-299/16), paras 102, 111 and 122, and Joined Cases T Danmark et al. (C-116/16 and C-117/16), paras 76, 83, and 95.

624 Marjaana Helminen prohibited applies, in tax matters, where the accrual of a tax advantage constitutes the essential aim of the transactions at issue.89 The risk of tax avoidance may justify tax treatment constituting a restriction on the fundamental freedoms.90 The risk of tax avoidance is often referred to by member states before the EU Court as a reason that should justify restrictive tax treatment. Particularly in cases of direct taxation, the EU Court, however, has been very reluctant to accept tax avoidance or loss of tax revenues as a justification for tax treatment constituting a restriction on the fundamental freedoms.91 Nonetheless, the EU Court seems to have become more understanding towards the member states in this respect.92 A general anti-tax-avoidance provision that does not apply only to wholly artificial tax- avoidance arrangements is not in accordance with EU law if it constitutes a restriction on the fundamental freedoms.93 Anti-tax-avoidance provisions should be applied only to wholly artificial arrangements with the purpose of avoiding the tax normally due.94 In order for legislation to be regarded as seeking to prevent tax evasion or abuse, its specific objective must be to prevent conduct involving the criterion of wholly artificial arrangements that do not reflect economic reality, and having the purpose of unduly obtaining a tax advantage.95 Wholly artificial arrangements are arrangements that lack any economic reality and that have the purpose of avoiding taxes.96 For example, a mere letterbox company established in the European Union may be a subject of anti-tax-avoidance provisions despite the fundamental freedoms.97 According to the EU Court, the concept, however, is also capable of covering, in the context of the free movement of capital, any scheme that has as its primary objective or one of its primary objectives the artificial transfer of the profits made by way of activities carried out in the territory of a member state to third countries with a low tax rate.98 Proof of an abusive practice requires both objective and subjective analysis. It requires a combination of objective circumstances in which, despite formal observance of the conditions laid down by the EU rules, the purpose of those rules has not been achieved. Secondly, a subjective element is required, consisting in the intention to obtain an

89 Ibid.. 90

See, e.g., Lenz (C-315/02), para. 27 and Lankhorst-Hohorst (C-324/00), para. 37. See, e.g., Commission v. France (avoir fiscal) (270/83), para. 25; ICI (C-264/96), para. 26; X and Y (C- 436/00), para. 41; Eurowings (C-294/97), Joined Cases Metallgesellschaft (C-397/98 and C-410/98); and Lankhorst-Hohorst (C-324/00), para. 37. 92 See, e.g., Marks & Spencer (C-446/03). 93 See ICI (C-264/96), para. 26. 94 See, e.g., ICI (C-264/96); Lankhorst-Hohorst (C-324/00); de Lasteyrie du Saillant (C-9/02); Marks & Spencer (C-446/03); Cadbury Schweppes (C-196/04); and Thin Capitalization Group Litigation Order (C- 524/04), para. 81. See also Emsland-Stärke (C-110/99), paras 52–53 and Halifax (C-255/02), paras 74–75. 95 Eqiom and Enka (C-6/16), para. 30 and Cadbury Schweppes (C-196/04), para. 55. 96 Cadbury Schweppes (C-196/04), para. 59. 97 Ibid., paras 67–68. 98 X GmbH (C-135/17), para. 84. 91

Tax Treaties and EU Fundamental Freedoms 625 advantage from the rules by artificially creating the conditions laid down for obtaining it.99 Normally, it is not in accordance with EU law to restrict the fundamental freedoms with an anti-avoidance provision if it is proved on the basis of objective factors that are ascertainable by third parties that, despite the existence of tax motives, the arrangement concerned corresponds to economic reality.100 The taxpayer must be given the possibility to show that the arrangement concerned is real and that there are also business reasons besides tax reasons for the arrangement.101 Minimizing taxes should not be the only motive for the arrangement. Anti-tax-avoidance provisions must always meet the requirements of the EU law proportionality test.102 Such a tax-avoidance provision is not proportional, which can be brought to bear without any objective criterion, verifiable by a third party, being applied to test for the existence of a wholly artificial arrangement which does not reflect economic reality, and which has been made with the aim of escaping the tax normally due. A rule framed in such terms does not make it possible, at the outset, to determine its scope with sufficient precision and its applicability remains a matter of uncertainty. Such a rule does not meet the requirements of the principle of legal certainty, in accordance with which rules of law must be clear, precise, and predictable as regards their effects, particularly where they may have unfavourable consequences for individuals or entities.103 In order to determine whether an operation pursues an objective of fraud and abuse, national tax authorities may not confine themselves to applying predetermined general criteria. Instead, they must carry out an individual examination of the whole operation at issue. The imposition of a general tax measure automatically excluding certain categories of taxpayers from the tax advantage, without the tax authorities being obliged to provide even prima facie evidence of fraud and abuse, would go further than is necessary for preventing fraud and abuse.104 If a tax authority seeks to refuse a tax benefit on grounds relating to the existence of an abusive practice, it has the task of establishing the existence of elements constituting such an abusive practice while taking all the relevant factors into account.105

99 Joined Cases N Luxembourg et al. (C-115/16, C-118/16, C-119/16 and C-299/16), paras 124 and 139, and Joined Cases T Danmark et al. (C-116/16 and C-117/16), paras 97 and 114. 100 Cadbury Schweppes (C-196/04), paras 55 and 75. 101 Test Claimants in the Thin Cap Group Litigation (C- 524/04). See also Commission v. United Kingdom (C-112/14), para. 28, and Joined Cases N Luxembourg et al. (C-115/16, C-118/16, C-119/16 and C- 299/16), para. 126, and Joined Cases T Danmark et al. (C-116/16 and C-117/16), para. 99. 102 ICI (C-264/96) and Cadbury Schweppes (C-196/04). See also, e.g., A. Zalasinski, ‘Proportionality of Anti-Avoidance and Anti-Abuse Measures in the ECJ’s Direct Tax Case Law’, Intertax 5 (2007), 310–321. 103 SIAT (C-318/10), paras 56–58 and Itelcar (C-282/12), para. 44. 104 Euro Park Service (C-14/16), paras 55–56, and Eqiom and Enka (C-6/16), para. 32. 105 Joined Cases N Luxembourg et al. (C-115/16, C-118/16, C-119/16 and C-299/16), para. 142, and Joined Cases T Danmark et al. (C-116/16 and C-117/16), para. 117.

626 Marjaana Helminen The EU member states must not apply tax treaty anti-avoidance provisions if the application leads to tax treatment which constitutes discrimination or a restriction on the fundamental freedoms without an acceptable rule-of-reason justification. Tax treaty anti-avoidance provisions, thus, must be scrutinized against the EU Court case law concerning tax avoidance and the fundamental freedoms. For example, tax treaty limitation-on-benefits (LOB) clauses aimed at combating treaty shopping are problematic from the perspective of the fundamental freedoms.106 The MLI LOB ‘ownership clauses’107 may lead to different treatment between EU companies controlled by residents of a contracting member state and those controlled by residents of a non-contracting member state. The LOB may deprive an EU member state resident company, which is controlled by residents of another member state, of entitlement to tax treaty benefits. The LOB may lead to a conflict with the fundamental freedoms when it does not exclude genuine practices.108 Due to the EU law issues, many EU member states would rather include a principal purpose test (PPT) provision than an LOB clause in their tax treaties.109 The principal purpose test, however, is also applicable only in line with the EU Court case law on tax avoidance and the fundamental freedoms. In its 2016 recommendation,110 the European Commission encouraged the member states to include a modification to the principal purpose test based on the OECD Model when they include such provisions in their tax treaties among themselves or with third countries. The suggested modification stresses that the principal purpose test provision does not apply if it is established that the arrangement concerned reflects a genuine economic activity.111 Such a modification 106 On

EU law compatibility of LOB clauses, see, e.g., Test Claimants in Class IV of the ACT Group Litigation (C-374/04), paras 89 and 90, which concerned an intra-EU situation (the UK–Netherlands treaty), and P. Plansky and H. Schneeweiss, ‘Proportionality of Anti- Avoidance and Anti- Abuse Measures in the ECJ’s Direct Tax Case Law’, Intertax 5 (2007), 310–321; A. Greter, ‘Limitation of Benefits Clauses and European Community Law: Legitimacy and Consequences’, EC Tax Journal 3 (2008), 1–15; O’Shea, EU Tax Law and Double Tax Conventions, 192–216; T. O’Shea, ‘European Tax Controversies: A British-Dutch Debate: Back to Basics and IS the ECJ Consistent?’, World Tax Journal 1 (2013), 109; E. Osterweil, ‘Are LOB Provisions in Double Taxation Conventions Contrary to EC Treaty Freedoms?’, EC Tax Review 5 (2009), 236–248; P. Gyllenstierna, ‘Dividend Taxation and the LOB Clauses in the US– Sweden DTC’, EC Tax Journal 3 (2008), 17–32, and CFE ECJ Task Force, ‘Opinion Statement ECJ-TF 1/ 2018 on the Compatibility of Limitation-on-Benefits (LoB) Clauses with the EU Fundamental Freedoms, European Taxation’ (2018), 419–425. O’Shea, ibid., at 109 points out that residence-based LOB clauses may be compatible with EU law while nationality-based clauses usually mean a conflict. 107 See art. 7(6)–(13) of the MLI and art. 29(1)–(7) of the OECD Model. 108 See also A. Garcia Prats et al., ‘EU report, Reconstructing the treaty network’, Cahiers de droit fiscal international (2020), 53–80. 109 Art. 7(1) of the MLI and art. 29(9) of the OECD Model. 110 Commission Recommendation (EU) 2016/ 136 of 28 January 2016 on the implementation of measures against tax treaty abuse. 111 Member states are encouraged to insert the following modification: Notwithstanding the other provisions of this Convention, a benefit under this Convention shall not be granted in respect of an item of income or capital if it is reasonable to conclude, having regard to all relevant facts and circumstances, that obtaining that benefit was one of the principal purposes of any arrangement or transaction that resulted directly or indirectly in that benefit,

Tax Treaties and EU Fundamental Freedoms 627 clearly restricts the scope of application of a tax treaty PPT provision compared to the OECD principal purpose test and thus does not comply with the OECD and MLI minimum standard. This fact may explain why the EU member states have not included the recommended modification in their tax treaties.112 The EU member states, however, must not apply the principal purpose test in conflict with the fundamental freedoms.

34.4 Concluding Remarks The elimination of double taxation and the prevention of tax avoidance are important objectives of both tax treaties and EU law. Allocation of taxing rights in tax treaties is in line with the EU internal market. The EU member states, however, must not conclude or apply tax treaty provisions in conflict with the TFEU fundamental freedoms. Tax treaties must not amount to unjustified discrimination or unjustified restriction on the fundamental freedoms. On the other hand, tax treaty benefits must not lead to selective tax benefits for certain taxpayers in conflict with the TFEU articles on prohibited state aid.113 While concluding and applying tax treaties, the EU member states must balance between these fundamental EU law provisions. There are still many unclear EU law issues concerning tax treaties and the TFEU fundamental freedoms. Hence, the EU Court will have to keep clarifying the way in which EU law should be interpreted in this regard.

unless it is established that it reflects a genuine economic activity or that granting that benefit in these circumstances would be in accordance with the object and purpose of the relevant provisions of this Convention. See the 2016 Recommendation, s. 2. 112

See Garcia Prats et al., 'EU report, Reconstructing the treaty network’, 65–66. 107 TFEU. See, e.g., Helminen, EU Tax Law—Direct Taxation, s. 1.6.2 on the state aid prohibition and e.g. Garcia Prats et al., ‘EU report, Reconstructing the treaty network’, 77–80, specifically on the relevance in relation to tax treaties. 113 Art.

Chapter 35

State Aid a nd International Tax at i on Patricia Lampreave Márquez

35.1 Introduction The European Commission recognizes that, in the context of free competition, member states of the EU may intervene with public resources to promote certain economic activities or protect national sectors. However, this aid could jeopardize competition in the European internal market by favouring certain companies over their competitors. The European Court of Justice (ECJ) has reiterated that member states are free to choose the economic policy they deem most appropriate and, specifically, to distribute the tax burden imposed on the different factors of production. Member states must, however, exercise these powers in accordance with EU law.1 The Treaty on the Functioning of the European Union (TFEU)2 regulates public aid in Title VII, relating to ‘Common rules on competition, taxation and approximation of laws’, specifically in the first chapter on competition rules. In section II of that chapter (i.e. arts 107–109 TFEU), aid granted by the states is regulated as another mechanism that can distort competition in the internal market. In particular, article 107(1) stipulates: Save as otherwise provided in the Treaties, any aid granted by a Member State or through State resources in any form whatsoever which distorts or threatens to distort competition by favouring certain undertakings or the production of certain goods shall, in so far as it affects trade between Member States, be incompatible with the internal market.

1 Case C-487/06 British Aggregates v. European Commission [2008] ECR I-10515 or Case C-182/08 Glaxo Wellcome [2009] ECR I-08591. 2 Consolidated versions of the Treaty on European Union and the Treaty on the Functioning of the European Union [2010] OJ C 83/47, 91–93.

630 Patricia Lampreave Márquez Article 108(1) states: The Commission shall, in cooperation with Member States, keep under constant review all systems of aid existing in those States. It shall propose to the latter any appropriate measures required by the progressive development or by the functioning of the internal market.

The lack of specificity of these articles has been supplemented through secondary law (mainly by Commission communications) and by resolutions of the ECJ and the General Court of the European Union (GC, former EU Court of First Instance). The Commission published, in 1998,3 a notice on the application of state aid rules, emphasizing its determination to apply these rules rigorously with respect to equality of treatment. Furthermore, the Commission has been issuing, during the last decades, a significant number of regulations,4 guidelines, frameworks, and communications5 with the aim of clarifying state aid rules. State aid is defined as an advantage in any form whatsoever conferred on a selective basis to undertakings by national public authorities. To be considered unlawful state aid, a measure needs to have these cumulative features: (1) Imputability and state resources: there must have been an intervention by the state (central, regional, local authorities,6 or public/private bodies designated by the state) or through state resources, which can take a variety of forms7 (e.g. grants, interest and tax relief, guarantees, government holdings of all or part of a company, or the provision of goods and services on preferential terms). (2) Economic advantage and selectivity: the intervention must give the recipient an economic advantage on a selective basis,8 for example in respect of specific

3 European Commission Notice of 10 December 1998 on the Application of the State aid Rules to Measures relating to Direct Business Taxation [1998] OJ C 384/3. 4 European Council Regulation (EU) 2015/ 1589 of 13 July 2015 laying down detailed rules for the application of Article 108 of the Treaty on the Functioning of the European Union (Text with EEA relevance) [2015] OJ L248. 5 European Commission Notice of 19 July 2016 on the notion of State aid as referred to in article 107(1) of the Treaty on the Functioning of the European Union [2016] OJ C 262/1. 6 The constitutional structure of a member state and the allocation of fiscal powers between national and peripheral authorities is not relevant for the application of the state aid rules. See Case C-88/03 Portuguese Republic v. Commission of the European Communities, ECLI:EU:C:2006:511. 7 Case C-143/99 Adria- Wien Pipeline GmbH and Wietersdorfer & Peggauer Zementwerke GmbH v. Finanzlandesdirektion für Kärnten [2001] ECR I-08365. 8 P. Lampreave Márquez, ‘Fiscal State Aid: a supra-national control of member states regulation and administrative practice’, Working Paper 4/2021, ed. IEF, Spanish Ministry of Finance (2021). A positive measure, such as a grant, confers an advantage in and of itself. A tax measure, in isolation, does not constitute a benefit to a taxpayer. The advantage results from an exception to the normal application of the tax system.

State Aid and International Taxation 631 companies or industry sectors, legal forms, sizes, or companies located in specific regions. (3) Undertaking and economic activity: the aid must be granted to an undertaking performing an economic activity,9 regardless of the legal status of the entity and the way in which it is financed.10 (4) Effect on trade and distortion of competition: competition might have been distorted11 and the intervention is likely to affect trade between member states. Both criteria are linked. It is not necessary to define the market or to investigate in detail the impact of the measure. In parallel to the Commission Notice on the notion of state aid, the ECJ stated, in Paint- Graphos, a three-step analysis12 to assess whether a fiscal measure is selective: (1) First, it is necessary to determine the tax reference framework. The system of reference normally constitutes the framework against which the selectivity of a measure is assessed.13 For fiscal measures, the reference system is the ordinary tax rule generally applicable to all undertakings falling within its scope (e.g. corporate income tax, the standard depreciation method recognized in the corporate income tax law, VAT system, anti-tax avoidance rules). (2) The next step is to determine whether the tax system provides any derogation leading to a different treatment for economic operators which, in the light of the objective of the system, are in a comparable legal and factual situation. (3) Finally, it should be recalled that a measure that creates an exception to the application of the general tax system may be justified if it results directly from the basic or guiding principles of that tax system (e.g. redistribution of income, economic capacity, avoidance of double taxation, tax neutrality, administrative manageability).14 A relevant issue in the justification analysis is whether or not these measures are proportional and do not exceed what is necessary to achieve the objective pursued.15 9

Economic activity vs non-economic activity. See European Commission 2016 Notice on state aid, 44, para. 201. 10 Classification of a particular entity as an undertaking thus depends entirely on the nature of its activities. To clarify the distinction between economic and non-economic activities, the ECJ has consistently held that any activity consisting in the offering of goods and services on a market is an economic activity: Case C-41/90 Klaus Höfner and Fritz Elser v. Macrotron GmbH [1991] ECR I-01979. 11 Case C-102/87 France v. European Commission [1999] ECR I- 03671; Case C- 142/ 87 Belgium v. European Commission [1990] ECR I-00959; Joined Cases C-278/92 and C-280/92 Spain v. European Commission [1994] ECR I-04103. 12 Joined Cases C-78–80/08 Paint Graphos and Others [2011] ECR I-07611. 13 Case C-203/16 P Heitkamp BauHolding v. Commission, ECLI:EU:C:2018:505. The ECJ considered that ‘An error in the determination of the reference framework against which the selectivity of the measure should be assessed necessarily vitiates the whole of the analysis of the condition relating to selectivity’, para.107. 14 Case C-88/03 Portugal v. European Commission [2006] ECR I-07115. 15 The positive effects must outweigh the negative effects.

632 Patricia Lampreave Márquez The focus, of the Commission’s assessment in fiscal aid cases is based on whether or not the tax measure grants an economic advantage and if it is a selective measure. An economic advantage will be gained where, as a result of the measure, the beneficiary’s net financial position is improved.16 As regards fiscal state aid, an economic advantage may be granted through different forms (fiscal holidays, tax credits, reduced tax base, reduced tax rates, accelerated depreciation, etc.)17 and, in particular, through a reduction in the tax base, the tax rate applied, or the amount of tax due.18 Selectivity means that a member state applies more lenient tax rules to a company or to companies of a specific sector, size, legal form, or located in a particular area, in comparison with other companies in a similar legal and factual situation. The discriminatory treatment between these comparable companies is established in a way that is not justified by the nature or general structure of the system. The advantage element is as relevant as the selectivity element, but the latter is much more controversial as it is a more subjective and changeable. The 2016 Notice on the notion of state aid distinguishes between ‘de jure’ and ‘de facto’ selectivity.19 De jure selectivity, results directly from the legal criteria for granting a measure that is formally reserved for certain undertakings. De facto selectivity can be established in situations in which, although the formal criteria for the application of the measure are formulated in general and objective terms, the structure of the measure is such that its effects significantly favour a particular group of undertakings.20 De facto selectivity may be the result of conditions or barriers imposed by member states to avoid certain undertakings benefiting from the tax measure. For example, applying a tax measure only to investments exceeding a certain threshold (other than a minor threshold for reasons of administrative expediency) may mean that the measure is de facto reserved for undertakings with significant financial resources. The use of discretionary powers by national tax administrations can also lead to de facto selectivity. Tax authorities have discretionary power in applying a measure where the criteria for granting the aid are formulated in a very general or vague manner that necessarily involves a margin of discretion in the assessment.21 16

P. Lampreave Márquez, ‘An Assessment of the Anti-Tax Avoidance Doctrines in the United States and the European Union’, Bulletin International Taxation 66 (2012), 3, Journals IBFD. 17 Case C-241/94 French Republic v. European Commission [1996] ECR I-04551 18 European Commission 1998 Notice on state aid, 3, para. 9. 19 European Commission 2016 Notice on state aid, 28, para. 120. 20 Joined Cases C- 106/09 P and C-107/09 P European Commission (C-106/09 P) and Kingdom of Spain (C-107/09 P) v. Government of Gibraltar and United Kingdom of Great Britain and Northern Ireland [2011] ECR I-11113. Three aspects of the tax system introduced by Gibraltar conferred selective advantages on the companies benefiting therefrom and may, therefore, have constituted state aid; namely (1) the requirement that a company make a profit before it becomes liable to payroll tax and Business Property Occupation Tax (BPOT); (2) the cap-limiting liability to payroll tax and BPOT to 15% of profits; and (3) the inherent nature of the payroll tax and BPOT. The Court found that such a combination of taxes excluded from the outset any taxation of offshore companies, as they had no taxable basis due to the lack of employees and lack of business property in Gibraltar. 21 Case C-241/94 France v. European Commission (Kimberly Clark Sopalin) [1996] ECR I-04551.

State Aid and International Taxation 633 A general tax measure, regardless of how harmful it is, will not be considered unlawful aid if it applies to all companies operating in that state without exception. This is particularly relevant in the analysis of the tax ruling practice of member states or their territories. In the current context of the Covid-19 crisis, EU member states have granted significant amounts of state aid to support individual undertakings and the EU economy as a whole. This aid has been considered as compatible—under article 107(3)(b) TFEU—as it is a situation in which the aid granted is indispensable for stabilizing the EU economy and to accelerate research into the coronavirus. However, even in the current situation, public support continues to be subject to EU state aid control in order to ensure the proportionality of the aid granted and to minimize the potentially distortive effect on competition. Moreover, the current framework is temporary and limited in scope to Covid-19 crisis measures.22

35.2 State Aid and Tax Competition One of the consequences of globalization is the increased openness of economies and the interdependence of states. Each national tax system is now unavoidably conditioned by other tax sovereignties. In this context, tax competition between states is inevitable. In 1997, the member states agreed to adopt a package of measures in respect of EU fiscal policy. The package was signed in 1999 by all member states and included proposals to adopt a Code of Conduct on business taxation.23 The compromises accepted by the member states included, on the one hand, a standstill clause that prevented the member states from introducing new harmful regimes and, on the other, a rollback clause that obliged the member states to abolish existing harmful measures. The EU Code provided for the establishment of a Code of Conduct Working Group to assess tax regimes that may fall within the EU Code. The Commission’s criteria in identifying potentially harmful tax measure are the following.24 The regime applies an effective level of taxation that is considerably lower than the general level of taxation fixed in the country concerned; tax advantages are granted only to non-residents or in respect of activities that are isolated from the domestic economy; those advantages are granted without any real economic activity or

22 European

Commission Communication (C/ 2020/ 1863) of 19 March 2020 on Temporary Framework for State aid measures to support the economy in the current COVID-19 outbreak [2020] OJ C 91I. This has been modified several times. 23 Conclusions of the ECOFIN Council meeting on 1 December 1997 concerning taxation policy, Doc 98/C2/01 [1998] OJ C 2/1 (EU Code of Conduct). 24 P. Lampreave Márquez, ‘Fiscal Competitiveness versus Harmful Tax Competition in the European Union’, Bulletin International Taxation 65/6 (June 2011), 1–18, https://www.ibfd.org/shop/journal/europ ean-union-fiscal-competitiveness-versus-harmful-tax-competition-european-union.

634 Patricia Lampreave Márquez substantial economic presence; and there is a lack of transparency and exchange of information with the tax authorities of other member states. An essential point to note at this stage is the connection between the EU Code and the EU state aid rules. The Commission has clearly stated that the tax systems in member states should not contain harmful tax measure, but also has to be in line with state aid rules. It is arguable that there is an overlap between the work of the Code of Conduct Working Group and state aid investigations. There are, however, differences in the nature of fiscal state aid and tax competition that should be taken into account. These divergences are essentially as follows:25 (1) The EU Code essentially reflects political consensus among the member states to take action against harmful regimes. If a regime is identified as harmful, this consideration will not be legally binding; however, member states, by committing themselves to the EU Code, de facto are obliged to eliminate the harmful elements of the regime. In contrast, state aid is regulated under the TFEU. If a tax regime or measure is considered to be state aid, the decision adopted by the Commission will be binding on the member state. State aid decisions can be enforced directly and should provide results within a precise time frame and usually entails that the unlawful aid should be recovered by the tax administration from the companies which have benefited from the preferential regime. (2) One of the main practical differences lies in the fact that the EU Code looks at measures that may affect the location of mobile capital within the Union, while state aid focuses on the impact on trade between member states. In certain scenarios, state aid may even encourage the location of certain activities in less favoured regions as part of a regional development policy. (3) The EU Code and state aid regulation have different objectives. State aid aims to eradicate the distortion in ‘competition between companies’ caused by the fact that the states, through their tax policies, favour certain economic operators over others in a comparable legal and factual situation. The EU Code aims to root out the distortion of ‘competition between Member States’ which has the consequence that a harmful regime can erode the tax base of other member states. (4) There are differences regarding the criteria that lead to a regime being considered as harmful or as illegal state aid. For instance, with respect to the criterion of selectivity, this seems to be absent in the EU Code and is one of the essential elements for determining that state aid is illegal. (5) Finally, harmful tax competition normally takes the form of reverse discrimination, as it tends to treat foreign undertakings more favourably than domestic undertakings, unlike typical state aid, but in fiscal cases the discrimination may

25 P.

Lampreave Márquez, ‘Harmful Tax Competition and Fiscal State Aid: Two Sides of the Same Coin?’, European Taxation 59 (2019), 5, Journal Articles & Opinion Pieces IBFD. p. 198.

State Aid and International Taxation 635 favour foreign undertakings investing in the territory or domestic undertakings acquiring shareholdings abroad. To conclude, state aid rules and the EU Code are complementary tools but should never overlap. Both actions strive to ensure a level playing field for business, the former by removing distortions in the internal market by favouring certain undertakings in relation to others in a similar legal and factual situation and the latter by countering harmful tax practices between member states.

35.3 Specific Issues Concerning Tax Measures under the Scrutiny of the EU State Aid Regulation 35.3.1 Tax Rulings and State Aid In order to contextualize the tax ruling issue within the state aid framework, it should be considered that after the financial crisis of 2007 there was a certain degree of sensitivity regarding the manner in which multinational groups (MNEs) operated on a cross-border level, in terms of having access to sophisticated tax expertise and profiting from base erosion and profit shifting (BEPS)26 opportunities not available to enterprises operating mostly at the domestic level. The problem was tackled by the OECD through the BEPS Project.27 The key question is whether the state aid rules truly provide an appropriate instrument for the control of profit shifting by MNEs. According to the 2016 Commission Notice,28 the function of a tax ruling is to establish, in advance, the application of the ordinary tax system to a particular case in view of the specific facts and circumstances. Therefore, it can create certainty and predictability regarding the application of general tax rules, which is relevant for taxpayers. Such rulings, however, must respect state aid rules. The Commission considers that an economic advantage exists in respect of rulings when, in the absence of state intervention, an undertaking would have not obtained, under normal market conditions, what is recognized in the tax ruling. 26 BEPS

refers to tax planning strategies used by multinational enterprises that exploit gaps and mismatches in tax rules to avoid paying tax. 27 Developed in the context of the OECD/ G20 BEPS Project, the fifteen actions set out equip governments with domestic and international rules and instruments to address tax avoidance, ensuring that profits are taxed where economic activities generating the profits are performed and where value is created (https://www.oecd.org/tax/beps/beps-actions/). 28 European Commission 2016 Notice on state aid, para 5.4.4 and 170. European Commission 1998 Notice on state aid, art. 22.

636 Patricia Lampreave Márquez The relevant issue is to determine if a tax ruling simply applies a general tax rule or misapplies it. A tax ruling may confer a selective advantage on the addressee if it endorses a result that does not reflect, in a reliable manner, what should be expected by the normal application of the ordinary tax system, insofar as that selective treatment lowers the addressee’s tax liability as compared to companies in a similar factual and legal situation which have not received such tax ruling. It must be noted that the selective requirement differs depending on whether the measure in question is envisaged as individual aid or a general scheme of aid. The GC held in the MOL state aid case29 that in cases of individual aid, the identification of the economic advantage is, in principle, sufficient to support the presumption that the aid is selective, therefore no comparison between economic operators is needed. By contrast, when examining a general scheme of aid, it is necessary to identify whether the measure in question, notwithstanding the finding that it confers an advantage of general application, does so to the exclusive benefit of certain undertakings. According to article 1(d) of Regulation 2015/1589,30 an aid scheme means: any act on the basis of which, without further implementing measures being required, individual aid awards may be made to undertakings defined within the act in a general and abstract manner and any act on the basis of which aid which is not linked to a specific project may be awarded to one or several undertakings for an indefinite period of time and/or for an indefinite amount.

Therefore, three cumulative conditions must be satisfied in order for a state aid scheme to be found to exist: first, further implementing measures should not be required, which means that the essential elements of the aid scheme in question must necessarily emerge from the provisions identified as the basis of the scheme; secondly, national tax authorities cannot have any margin of discretion as regards the aforementioned essential elements of the aid and whether it should be awarded and ,thirdly, the undertakings to which the aid is granted must be defined in the act in a general and abstract manner. In the Belgian excess profits case, the GC31 annulled the Commission aid scheme decision adopted in 2016,32 finding that the decision did not succeed in demonstrating that

29 GC: Case C- 15/ 14 P MOL Magyar Olaj- és Gázipari Nyrt v. European Commission, ECLI:EU:C:2015:362. 30 European Council Regulation (EU) 2015/1589. 31 GC: Joined Cases T- 131/ 16 and 263/ 16 Magnetrol International and Belgium v. European Commission, ECLI:EU:T:2019:91. 32 European Commission Decision of 21 October 2015 on excess profit tax ruling system in Belgium, No. 2016/1699/EC [2016] OJ L260; the Commission determined that Belgium’s excess profit tax scheme, applied since 2005, allowed Belgian tax-resident companies which were part of a multinational group to pay substantially less tax in Belgium on the basis of tax rulings. The excess profit tax scheme was marketed by the tax authority under the logo ‘Only in Belgium’. The scheme reduced the corporate tax base of the companies by between 50% and 90% to discount for so-called excess profits that allegedly resulted from being part of a multinational group.

State Aid and International Taxation 637 the approach that the Commission had identified met the requirements set out in article 1(d) of the regulation. However, Advocate General Kokott proposed that the ECJ should set aside the GC judgment on the ground that the Commission had sufficiently demonstrated in its decision that the Belgian practice of making downward adjustments to profits of undertakings forming part of MNEs met all the requirements for the existence of an aid scheme.33 In 2021, the ECJ34 concluded that the three conditions for an aid scheme to exist were met and, consequently, the Commission not longer had to investigate each ruling individually. The ECJ referred the case back to the GC, which will have to decide on the remaining open issues, such as the existence of a selective advantage and the identification of the beneficiaries of the alleged aid. The final decision may take a number of years.

35.4 APAs and State Aid Regulation The OECD has provided Transfer Pricing Guidelines35 which are non-binding but represent the consensus on the estimation of an arm’s-length price.36 Transfer prices refer to prices charged for commercial transactions between the separate entities of the same corporate group. To avoid profit shifting problems, tax administrations should only accept transfer prices between intra-group companies that are remunerated as if they were agreed by independent companies negotiating under comparable circumstances at arm’s length. Since 2014, the Commission has been criticized for initiating state aid proceedings in respect of advantageous tax rulings on transfer pricing granted to certain renowned MNEs (Apple, Starbucks, Fiat, among others). Given the widespread public perception that these groups have not paid their fair share of tax in the EU, the decisions attracted a certain amount of attention.37 Nevertheless, it is not a novel approach. These proceedings are simply a further step in a long development of case law and decision-making practice which began in the late

33 Opinion of AG Kokott, Case C- 337/ 19 P Magnetrol International and Belgium v. European Commission, ECLI:EU:C:2020:990. 34 Case C-337/19 P Magnetrol International and Belgium v. European Commission and France v. European Commission, ECLI:EU:C:2021:741. 35 OECD, Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations (Paris: OECD Publishing, 2017). The OECD TP Guidelines were adopted in their original version on 27 June 1995 by the OECD’s Committee on Fiscal Affairs. The 1995 Guidelines were substantially updated in July 2010. 36 The purpose of the arm’s-length principle is to ensure that transactions between group companies are treated for tax purposes by reference to the amount of profit that would have arisen if the same transactions had been executed by independent companies. 37 Lampreave Márquez, supra 15, 205.

638 Patricia Lampreave Márquez 1990s. In 2003, the Commission issued a number of decisions concerning tax schemes on transfer pricing involving individual rulings in favour of companies. Examples of these are the Commission decision relating to the Belgian tax ruling scheme for the foreign sales corporations of US companies38 and the Belgian coordination centres regime. The latter is a reference case on the issue of a state aid investigation of transfer pricing methods validated in an advance pricing arrangement (APA). On 17 February 2003,39 the Commission adopted a final negative decision in its state aid investigation into the Belgian coordination centres regime. Among other elements of the regime considered to be state aid, the way its taxable revenues were determined according to the so-called cost-plus method was questioned.40 The Commission did not question the cost-plus method per se but the incorrect form in which the method was applied. According with the Commission, the Belgian authorities systematically used a default margin of 8% for the services provided by a coordination centre without verifying if the rate reflected the economic reality of the underlying transactions. Significant operating expenses were excluded from the total costs of the centre when calculating the taxable income. The coordination centre’s status was obtained through a renewable ten-year tax ruling granted by the Belgian tax authorities. The ECJ41 stipulated in this particular case that in order to decide whether a method of assessment of taxable income (e.g. as laid down under the regime for coordination centres) conferred an advantage, it was necessary to compare the regime with the ordinary tax system, based on the difference between profit and outgoings of an undertaking carrying on activities in conditions of free competition. The ECJ then held that the effect of the exclusion of staff and financial costs from the expenditure, which served to determine the taxable income of the coordination centres, did not resemble the transfer prices which would have been charged in conditions of free competition. 38

European Commission Decision of 24 June 2003, on the aid scheme implemented by Belgium— Tax ruling system for United States foreign sales corporations, No. 2004/77/EC [2004] OJ L23. In this case, the aid scheme implemented by Belgium consisted of a tax ruling system for transfer pricing purposes, granting to the Belgian permanent establishment of a US-resident foreign sales corporation, a forfeit determination of its tax base in Belgium, based on a cost-plus system with fixed costs of 8%. The ruling was granted specifically to exclude activities already exempted in the USA from taxation in Belgium. 39 European Commission Decision of 17 February 2003, on the aid scheme implemented by Belgium for coordination centres established in Belgium, No. 2003/757/EC [2003] OJ L282. 40 The cost-plus method is one of the five common transfer pricing methods provided by the OECD Guidelines. The cost-plus method is a traditional transaction method which compares gross profit to the cost of sales, a pricing strategy in which the selling price is determined by adding a specific mark- up to a product’s unit cost. The mark-up on costs that the manufacturer or service provider earns from the controlled transaction is compared with the mark-up on costs from comparable uncontrolled transaction. OECD, ‘Transfer Pricing Methods’ (July 2010), https://www.oecd.org/ctp/transfer-pricing/ 45765701.pdf. 41 Joined Cases C-182/03 and C-217/03 Kingdom of Belgium v. European Commission [2006] ECR I-05479.

State Aid and International Taxation 639 Finally, the ECJ concluded that a tax measure which resulted in a group company charging transfer prices that did not reflect those which were usually charged in conditions of free competition, conferred a selective advantage on the coordination centres, insofar as it resulted in a reduction of its taxable base and, thus, its tax liability under the ordinary corporate income tax system. This case law provided the intellectual foundation of the Commission’s position that the fiscal burden of integrated companies should be the same as that of independent companies resulting from the application of the standard tax rules. The dismantling of this regime was extremely controversial as these centres were frequently used for tax planning. The regime was also considered harmful by the EU Code. As already noted, since 2014, the Commission has issued several highly debatable decisions. Among others, the Starbucks case is a good example of a Commission decision in which it analysed different methods recognized in an APA under state aid regulation. Briefly, in 2001 the Dutch tax authorities had concluded an APA with Starbucks Manufacturing EMEA BV (SMBV), part of the Starbucks group, which, inter alia, roasted coffee. The objective of that arrangement was to determine SMBV’s remuneration for its production and distribution activities within the group. Thereafter, SMBV’s remuneration served to determine annually its taxable profit based on Dutch corporate income tax. In addition, the APA endorsed the amount of royalties paid by SMBV to Alki (another entity in the same group) for the use of Starbucks’s intellectual property for roasting. More specifically, the APA provided that the amount of royalties to be paid to Alki corresponded to SMBV’s residual profit.42 In 2015, the Commission found that the APA constituted an unlawful aid incompatible with the internal market and ordered the recovery of that aid. The Netherlands and Starbucks brought an action before the GC against the Commission’s decision.43 The GC’s44 judgment in the Starbucks case is highly significant as it concluded with several crucial and controversial claims. The first claim was that the Commission had violated the member states’ fiscal autonomy as it was not entitled to question whether the methodology validated in an APA could be a deviation of the arm’s-length principle leading to the grant of a selective advantage in favour of entities benefiting from that APA. The GC held that, according to the Belgian coordination centres case, when examining a fiscal measure granted to an integrated company, the Commission is entitled to compare the fiscal burden of such an integrated undertaking with one resulting from

42 The amount was determined by deducting SMBV’s remuneration, calculated in accordance with the APA, from SMBV’s operating profit. 43 European Commission Decision of 21 October 2015, Case SA.38374, on state aid SA.38374 (2014/C ex 2014/NN) implemented by the Netherlands to Starbucks, No. 2017/502/UE [2017] OJ L83. 44 GC: Joined Cases T- 760/ 15 and T- 636/ 16 The Netherlands v. European Commission, ECLI:EU:T:2019:669.

640 Patricia Lampreave Márquez the application of the normal tax rules for independent companies carrying on their activities under market conditions. The GC made clear that the arm’s-length principle is a tool that allows the Commission to check that intra-group transactions are remunerated as if they were negotiated between independent companies. The GC also accepted the Commission’s position that this examination could be made independently of whether a member state would incorporate the arm’s-length principle in its national legal system if it could be intrinsically derived from the national law. The second claim presented was related to the transactional net margin method (TNMM45) for determining SMBV’s remuneration validated in the APA. The Commission found that the transfer pricing report, on the basis of which the APA was concluded, did not contain an analysis of the royalties which SMBV paid to Alki. The GC argued that mere non-compliance with the methodological requirements did not necessarily lead to a reduction of the tax burden and that the Commission would have had to demonstrate that the methodological errors identified in the APA did not allow a reliable approximation of an arm’s-length outcome leading to a reduction in the tax burden. The third claim was linked with the application by the Commission of the comparable uncontrolled price (CUP) method46 in examining the arm’s-length nature of the royalties. As a result of that analysis, the Commission considered that the amount of the royalties should have been zero. The GC stated that the mere finding by the Commission that the APA did not analyse the royalties did not sufficiently demonstrate that those royalties were not in conformity with the arm’s-length principle or that they resulted in an advantage within the meaning of the TFEU. Finally, it concluded that the Commission was required to justify its choice of methodology and it was not entitled to find that the CUP method had to be given priority, in principle, over the TNMM or any other method recognized by the OECD Guidelines. The Commission could have appealed before the ECJ, but an appeal before the ECJ against a GC judgment can only be lodged on points of law and, in this particular case, the GC appears to have validated some of the core legal principles underlying the Commission’s approach for the state aid assessment of tax rulings. On the same day, the Starbucks’s decision was published, the GC confirmed the Commission’s decision47 on the aid measure granted by Luxembourg to Fiat Chrysler Finance Europe (FFT). In that case, the GC found that the Commission had correctly

45

The TNMM examines a net profit indicator, i.e. a ratio of net profit relative to an appropriate base (e.g. costs, sales, assets) that a taxpayer realizes from a controlled transaction (or from transactions that are appropriate to aggregate) with the net profit earned in a comparable uncontrolled transaction. 46 The CUP method compares the price charged for property or services transferred in a controlled transaction to the price charged for property or services transferred in a comparable uncontrolled transaction in comparable circumstances. 47 GC: Joined Cases T-755/15 and T-759/15 Grand Duchy of Luxembourg and Fiat Chrysler Finance Europe v. European Commission, ECLI:EU:T:2019:670.

State Aid and International Taxation 641 proved that the arrangements for the application of the TNMM endorsed by the tax ruling at issue were incorrect. It was inappropriate for Fiat to have segregated its capital by reference to its different activities for the purposes of its TNMM analysis. The GC considered that the Commission correctly found that the whole of FFT’s capital should have been taken into account and a single rate should have been applied. The Court confirmed the Commission’s view that the methodology approved by the ruling resulted in a lowering of FFT’s tax liability as compared to the tax that it would have had to pay under Luxembourg tax law. Finally, in another controversial case, the GC annulled48 the Commission’s decision on the APAs issued by the Irish tax authorities in favour of the Apple group.49 The Commission considered that the contested APAs granted in favour of two companies of the Apple group (ASI and AOE50) reduced the amount of tax for which they were liable in Ireland during the period when those rulings were in force. Ireland and ASI/AOE contended that the arm’s-length principle was not part of Irish tax law and that no free-standing obligation to apply that principle emerges from article 107 TFEU. The GC repeated the same reasoning used in the Starbucks case and stated that the Commission did not err when it used the arm’s-length principle as a tool (even if the principle was not part of national law) in order to verify that the profit allocated to the branches corresponded to the profit that could be earned under market conditions. Although the GC noted, obiter dictum, the incomplete and occasionally inconsistent nature of the contested APAs, it considered, as in the Starbucks case, that the Commission had not succeeded in demonstrating methodological errors in the tax ruling which would have led to a reduction of chargeable profits in Ireland. The defects identified by the Commission were not, in themselves, sufficient to prove the existence of an advantage for the purposes of article 107(1) TFEU. Nevertheless, the Commission decided to appeal the GC’s judgment on the Apple case before the ECJ.51 The Commission has considered that the GC’s judgment raised important legal issues that were of relevance of the institution in its application of state aid rules to pending or future tax planning investigations. Also, the Commission stated that the GC had made a number of errors of law. Therefore, the ECJ will have the last word on this crucial case. It must be considered that the ECJ has always applied a broader

48 GC:

Joint Cases T- 778/ 16 and T- 892/ 16 Ireland v. European Commission and Apple Sales International/Apple Operations Europe v. European Commission, ECLI:EU:T:2020:338. 49 European Commission Decision of 30 August 2016 on state aid, SA.38373 (2014/C), No. 2017/1283 [2017] OJ L187. 50 Apple Operations Europe (AOE) and Apple Sales International (ASI) subsidiaries of US Apple Operations International. ASI and AOE are both companies incorporated in Ireland, but are not tax- resident in Ireland. 51 Appeal brought on 25 September 2020 by the European Commission against the judgment of the General Court (Seventh Chamber, Extended Composition) delivered on 15 July 2020 in Joined Cases T- 778/16 and T-892/16 Ireland and Others v Commission (Case C-465/20 P, 2021/C 35/33.

642 Patricia Lampreave Márquez approach (in comparison to the narrow approach of the GC) on the notion of selectivity, more in line with the Commission’s interpretation of the criterion. However, recently, the ECJ52 concluded that the GC had erred in law by validating the Commission’s view in the Fiat case that Luxembourg national law was not the reference system but the OECD arm´s length principle. The ECJ considered that “(..)parameters and rules external to the national tax system at issue cannot therefore be taken into account in the examination of the existence of a selective tax advantage”. If the ECJ maintain the same argument in all tax rulings pending cases, this could constitute a major blow to the Commission tax ruling cruzade initiated by M. Vestager. A decisive argument used against the Apple Commission’s decision is that if the use of state aid could have been proved, the aid was granted by the US tax authorities and not by the Irish tax authorities. The underlying driver in all these cases is that EU state aid rules help to preserve a level playing field, among undertakings, with regard to subsidies provided by the EU authorities. However, there are no such rules for subsidies that non-EU authorities grant to undertakings operating in the EU. In order to protect the internal market from those subsidies, the Commission adopted a ‘White Paper on levelling the playing field as regards foreign subsidies’53 and proposed solutions and new tools for addressing the regulatory gap. The White Paper intended to launch a broad discussion with member states, other European institutions, all stakeholders, including industry, social partners, civil society organizations, researchers, the public in general, and any other interested party on the best way to effectively address the challenges identified. The results of the consultation on the White Paper prepared the ground for choosing the most appropriate way to address the distortions created by foreign subsidies, including appropriate proposals for legal instruments such as the Proposal for a Regulation on foreign subsidies distorting the internal market.54 As a consequence of this proposal on 12 January 2023, Regulation on foreign subsidies distorting the internal market entered into force.55 The lessons to be drawn from all this case law (i.e. the Belgian coordination centres, Starbucks, Fiat, and Apple cases) can be summarized as follows: in matters relating to tax rulings that include a transfer pricing agreement, it is questionable whether or not tax authorities, in recognizing the application of a specific transfer pricing method instead of the arm’s-length standard, do so based on their wide margin of discretion and

52 ECJ, 8 Nov. 2022, Joined Cases C-885/19 P and C898/19 P, Fiat Chrysler Finance Europe, Ireland, v. European Commission, ECLI:EU:C:2022:859. 53 White Paper on levelling the playing field as regards foreign subsidies, COM(2020) 253 final (17 June 2020), https://ec.europa.eu/competition/international/overview/foreign_subsidies_white_pa per.pdf. 54 Proposal for a Regulation on foreign subsidies distorting the internal market, COM/2021/223 final (2021). 55 European Union Regulation of 14 December 2022, on foreign subsidies distorting the internal market, No. 2022/2560, OJ L 330 (2022).

State Aid and International Taxation 643 that this automatically results in a selective advantage and, hence, should be considered as illegal state aid. Consequently, a taxpayer is free to choose a transfer pricing method as long as the method selected leads to an arm’s-length outcome for the transaction in question. Even though, the taxpayer is expected to take into account its reliability when making their choice, they are not obliged to assess all methods and then justify how the method they have chosen produces the best result under the present market conditions. It is not for the Commission to police the application of tax rules, nor to substitute its own idea of a ideal tax system for that of a member state or to determine which is the best transfer pricing method to be applied. But, according with the mentioned case law, the Commission is permitted to check that the transfer pricing methods applied by a member state do not create an automatic advantage in favour of certain undertakings. The Commission has reiterated that the OECD Guidelines are used as a reference when an APA is analysed under state aid regulations. If a transfer pricing arrangement complies with the OECD Guidelines and the ALP, it is unlikely to give rise to state aid. Nevertheless, the Commission has underlined that the Guidelines are simply a reference and are not legally binding for state aid investigations, as the OECD has never analysed transfer pricing under article 107(1) TFEU.

35.5 Conclusions In conclusion, it can be assumed that invasive actions that jeopardize the tax sovereignty of the member states are not, a priori, well received. The protectionism of states with regard to their fiscal policy and the complexity of common progress on this matter in the European Union are evidence of this reality. The EU Code of Conduct Working Group has been working since 1997 to eradicate harmful regimes established in national tax laws due to the risk of distortion of competition between member states. In parallel, and since it is evident that a clear catalyst for the distortion of competition between economic operators acting in the internal market has been certain tax subsidies granted by member states, the state aid rules have helped to root out the aforementioned distortions. State aid rules and the EU Code are complementary tools but should never overlap. Both actions strive to ensure a level playing field for business but are divergent in many respects. With regard to the role and potential limits of the Commission in reviewing tax rulings granted by tax authorities of member states, it could be concluded that it is difficult to agree with the idea that the task of state aid control is to ensure that taxes are paid in the proper place. It is also challenging that the Commission could be considered as a supranational tax authority that can establish a new notion of arm’s length or condemn certain methods recognized by the OECD Guidelines.

644 Patricia Lampreave Márquez Nevertheless, as has been stated by the EU Court, the determination of the Commission to investigate the tax ruling practice of the member states or advance pricing arrangements granted by national tax authorities are in line with the powers conferred by article 107(1) TFEU. As a final reflection, EU policies that tackles harmful tax practices and selective advantages that discriminate against economic operators in similar situations in the internal market make the EU a reference point for the global economy. Common interest should prevail over the interests of individual member states or a certain group of undertakings.

Chapter 36

Internationa l Tax L aw and the EE A / E FTA Patrick Knörzer

36.1 Introduction The European Free Trade Association (EFTA) was founded in 1960 as a free-trade zone and was intended to be an alternative economic integration to the European Economic Community. The European Economic Area (EEA) was established based on the EEA Agreement which was concluded in Porto on 2 May 1992 by the then seven EFTA member states and the then twelve member states of the European Community (EC) and was designed as a mechanism to allow EFTA member states to actively participate in the process of European economic integration by having full access to the internal market without transferring any political competences to the EU bodies. The EEA Agreement entered into force at the beginning of 1994 except for Liechtenstein. The Principality of Liechtenstein joined the EEA on 1 May 1995 after the approval of the EEA accession in two referenda. Switzerland did not become a member of the EEA because of a negative referendum in 1992. On 1 January 1995, the three EFTA member states Austria, Finland, and Sweden joined the EU. Hence, since May 1995 the EFTA only consists of Iceland, Liechtenstein, Norway, and Switzerland. For new EU candidate states, an accession to the EEA is obligatory. However, they do not become EEA members automatically or immediately after obtaining candidate status, but they are obliged to apply. Therefore, the five currently recognized candidates for EU membership (Albania, North Macedonia, Montenegro, Serbia, Turkey) are not already EEA members, neither are the potential candidates Bosnia and Herzegovina and Kosovo. Currently, the EEA comprises twenty-seven EU member states and the three EEA EFTA states—Iceland, Liechtenstein, and Norway. This follows from article 126 of the EEA Agreement, which states that the EEA Agreement applies to the territory of the EU and of the three EEA EFTA states.

646 Patrick Knörzer Although the three non-EU EEA states are part of the European single market in goods and services, including financial services, they are not part of the EU’s customs union. EEA EFTA states are entitled to negotiate free trade agreements with third countries, either independently or through EFTA. Furthermore, the EEA Agreement does not seek to establish a customs union and does not include common rules with regard to third countries. The EEA EFTA states have agreed to enact legislation similar to that passed in the EU in the areas of social policy, consumer protection, environment, company law, and statistics. The relevant EU directives, for example the directives on the European single financial market, are not directly effective in EFTA states. Instead, they have to be implemented by local law and regulation in every EEA EFTA state.

36.2 Principles of Taxation in the EEA According to the established case law of the European Court of Justice (ECJ) and the EFTA Court, direct taxation, in principle, is a matter for the EEA states’ national competence. The EEA Agreement was not meant to cover matters of taxation, as the EFTA Court points out: ‘The Court notes that, as a general rule, the tax system of an EEA State is not covered by the EEA Agreement.’1 The harmonizing measures taken within the EU in the field of direct taxation are not extended to the EFTA states, above all the Parent–Subsidiary Directive, the Interest and Royalties Directive, the Merger Directive, and the Anti-Tax Avoidance Directive, but also the Mutual Assistance Directive and the Recovery Assistance Directive. The European Commission noted that ‘Secondary Community legislation in the area of taxation, however, has not been incorporated in the EEA Agreement’.2 As a consequence, EEA EFTA states are not obliged to implement secondary Community legislation in the area of taxation. Indirect taxes are also an issue of the EEA states’ national law as VAT directives and the Capital Duty Directive do not apply to non-EU states. The same principle goes for excise duties. Goods exported to the three non-EU EEA states are subject to excise duties at the rates set by their national tax authorities. Those indirect taxes which are EU taxes—principally, VAT and customs duty—are not applicable in EFTA EEA states. Actually, Liechtenstein is in a customs union with Switzerland and basically levies the same indirect taxes as Switzerland.

1

EFTA Court 22 February 2002, E-1/01 Einarsson, para. 17; 12 December 2003, E-1/03 ESA v. Iceland, para. 26; and 20 May 1999 E-6/98, Norway v. ESA, para. 34. 2 Communication from the Commission to the Council, the European Parliament and the European Economic and Social Committee—Exit taxation and the need for co-ordination of member states’ tax policies, COM(2006) 825 final (19 Dec. 2006).

International Tax Law and the EEA/EFTA 647 However, the EEA Agreement also applies on taxes insofar as EEA EFTA states ‘must, however, exercise their taxation power consistent with EEA law’3. The structure of the EEA Agreement is similar to the TFEU, as at the beginning a prohibition of discrimination on grounds of nationality is stipulated. Further provisions include the fundamental freedoms and rules on competition as well as state aid law. In the well-known Ospelt case, the ECJ held that ‘one of the principal aims of the EEA Agreement is to provide for the fullest possible realisation of the free movement of goods, persons, services and capital within the whole European Economic Area, so that the internal market established within the European Union is extended to the EFTA States’.4 According to article 105 EEA Agreement, it is an objective of the contracting parties of the EEA to arrive at as uniform an interpretation as possible of the provisions of the EEA Agreement and those provisions of Community legislation which are substantially reproduced in the EEA Agreement. The ECJ consequently held that the provisions of the EEA Agreement should be interpreted in a similar way as the corresponding provisions of the Treaty on the Functioning of the European Union (TFEU) to ensure a uniform interpretation of the rules of the EEA Agreement and the EC Treaty, which are identical in substance, thereof throughout the EEA.5 The EFTA Court also presumed that identically worded provisions are to be interpreted in a homogeneous way.6 It cannot be ruled out that differences in the scope and purpose of the EEA and TFEU may, under specific circumstances, lead to differences in interpretation.7 But where parallel provisions are to be interpreted without any such specific circumstances being present, homogeneity should prevail.8

36.2.1 Non-Discrimination Principle The principle of non-discrimination is included in article 4 EEA Agreement which reproduces the wording of article 18(1) TFEU. According to the ECJ, both rules are virtually ‘identical’.9 The EFTA Court has described article 4 EEA Agreement as a general principle.10 The general principle of non-discrimination is repeated in article 28 EEA Agreement as regards the free movement of workers, article 34 provides for the

3 EFTA Court 4 April 2013, E-1/04 Fokus Bank, para. 20; see also ECJ 13 December 2005, C-446/03 Marks & Spencer, para. 29. 4 ECJ 23 September 2003, C-452/01 Ospelt und Schlössle Weissenberg, para. 29. 5 Opinion of the ECJ 10 April 1992, 1/92 [1992] ECR I-2821. 6 EFTA Court 26 June 2007, E- 2/ 06 EFTA Surveillance Authority v. Norway, para. 59; see C. Baudenbacher, ‘The Goal of Homogeneous Interpretation of the Law in the European Economic Area’, The European Legal Forum (E) 1 (2008), 22. 7 EFTA Court 3 December 1997, E-2/97 Mag Instrument Inc. v. California Trading Co. Norway. 8 EFTA Court 10 December 1998, E-3/98 Herbert Rainford-Towning, paras 21ff. 9 ECJ 20 January 2011, C-155/09 Commission v. Greece, para. 74 and ECJ 11 September 2014, Joined Cases C-204–208/12 Essent Belgium, para. 123. 10 EFTA Court 28 January 2013, E-16/11 ESA v. Iceland, para. 204.

648 Patrick Knörzer non-discrimination of companies, article 36 prohibits restrictions on the freedom to provide services, and the non-discrimination of service providers is ensured through article 37. The non-discrimination rule seeks to ensure equal treatment for nationals of any EEA state inside any other EEA state. In the field of taxation, this means that non-resident taxpayers cannot be discriminated against when in a situation comparable to that of resident taxpayers. Resident and non-resident taxpayers should receive equal tax treatment if they are in a comparable situation. In most cases, however, the situations of resident and non-resident taxpayers in relation to direct taxes are different. In the well-known Schumacher11 case, as the major part of the income of the non-resident taxpayer was not derived in the state of residence this led to a comparable situation to a resident that could not justify different treatment.

36.2.2 State Aid Provisions and Taxes The state aid rules under the EEA Agreement12 broadly correspond to those in the EU. The definition of state aid in article 61(1) EEA Agreement matches the definition of state aid in article 107 TFEU expanding the scope to include the EEA EFTA states. State aid comprises ‘any aid granted by EC Member States, EFTA States or through State resources in any form whatsoever’ which is incompatible with the functioning of the EEA Agreement. This scope also encompasses aid granted by regional or local bodies in the EFTA states.13 State aid can have many forms, including grants and loans and may also include certain tax measures such as tax exemptions if the advantage is granted by the state or through state resources. Consequently, state aid may be provided through tax provisions of a legislative, regulatory, or administrative nature as through the practices of the tax authorities. As a rule, state aid is prohibited to prevent market distortion and negative effects on trade. Exemptions are made where public interventions might be necessary for a well-functioning and equitable economy, in areas such as research and development, environmental protection, regional development, etc. In such cases, aid might be considered compatible with the functioning of the internal market. The tax measure must affect competition and trade between EFTA states and place the persons to whom the tax exemption applies in a more favourable financial situation than others. According to consistent case law, the trade is already affected if the company trades within the EEA. It is sufficient that the aid strengthens the position of the company compared with those of their competitors in intra-EEA

11

ECJ 14 February 1995, C-279/93 Schumacher. Annex XV and Protocols 26–27 of the EEA Agreement, arts 49, 59, 61–64 EEA Agreement. 13 ECJ 14 October 1987, C-248/84 Germany v Commission. 12

International Tax Law and the EEA/EFTA 649 trade.14 More than forty state aid schemes which have been approved by the EFTA Surveillance Authority (ESA) encompass relief from taxes in one of the three EEA EFTA states. The EFTA Court rules that the definition of state aid is more general than that of a subsidy. The concept of aid not only includes positive benefits, such as subsidies themselves, but also measures which, in various forms, mitigate the charges which are normally included in the budget of an undertaking and which, thus, without being subsidies in the strict sense of the word, are similar in character and have the same effect.15 A tax exemption or tax reduction may give rise to state aid, as a loss of tax revenue is equivalent to consumption of state resources in the form of fiscal expenditure. The advantage may be provided through a reduction in the firm’s tax burden in various ways, including: • a reduction in the tax base (e.g. special deductions, special or accelerated depreciation arrangements, or the entering of reserves on the balance sheet); • a total or partial reduction in the amount of tax (e.g. exemption or a tax credit); • deferment, cancellation, or even special rescheduling of tax debt.16 In general, tax measures of a purely technical nature such as depreciation or amortization rules, such as accelerated depreciation, do not constitute state aid as all sectors and all undertakings are subject to them under the same conditions. The method of calculating asset depreciation varies from one EEA state to another, but such methods may be inherent to the respective tax system. However, as a depreciation incentive is only applicable for certain assets or certain branches, it may give rise to state aid.17 A reduction of or an exemption from an excise tax can grant a selective advantage to the undertakings which use the product in question as an input or sells it on the market as, it is equivalent to a waiver of the fiscal revenue which would otherwise have to be paid by the producer.18 State aid can also arise if the amount is relatively small, with the exception of aid that meets the de minimis rule or if the recipient of the aid is moderate in size or its share of the EEA market is very small.19 Irrelevant for the classification as state aid is the fact that 14

EFTA Surveillance Authority Decision No. 149/99/COL of 30 June 1999 introducing guidelines on the application of State aid rules to measures relating to direct business taxation and amending for the 19th time the Procedural and Substantive Rules in the field of State aid [2000] OJ L137, ch. 17B.3.(4). 15 see EFTA Court 8 April 2014, Joined Cases E- 4/10, E-6/10 and E-7/10 Liechtenstein and Others v. ESA, para. 69; and, for comparison, ECJ 15 November 2011, Joined Cases C-106/09 P and C-107/09 P, para. 71 and case law cited therein. 16 EFTA Surveillance Authority Decision No. 149/ 99/COL of 30 June 1999 introducing guidelines on the application of State aid rules to measures relating to direct business taxation and amending for the 19th time the Procedural and Substantive Rules in the field of State aid [2000] OJ L137, ch. 17B.3. Application of art. 61(1) EEA Agreement to tax measures. 17 ESA Decision No. 3/17/COL of 18 January 2017, paras 177ff. 18 European Commission Decision 1999/7 79/EC of 3 February 1999 [1999] OJ L305/27. 19 ECJ, Joined Cases C-278/92, C-279/92 and C-280/92 Spain v. Commission.

650 Patrick Knörzer the recipient does not carry out exports20 or exports almost all of its production outside the EEA.21 The wording of article 61 EEA Agreement requires that a measure must be selective and therefore favour certain undertakings or the production of certain goods in order to be classified as state aid. The selective application of a measure therefore constitutes one of the criteria inherent in the notion of state aid.22 On the other hand, advantages resulting from a general measure applicable without distinction to all economic operators do not constitute state aid within the meaning of article 61 EEA Agreement. The ESA enforces restrictions on state aid, assessing their compatibility with the functioning of the internal market. All new state aid measures must be notified to and approved by ESA before their implementation (the ‘standstill clause’). Any new state aid put into effect by the member states without the prior authorization of the ESA is unlawful. The ESA has the power to order remedies against violations of the standstill obligation, such as the repayment of unlawful state aid. The EFTA states are required to submit to the Authority every year reports on their existing state aid systems. In the case of tax relief or full or partial tax exemption, the reports must provide an estimate of budgetary revenue loss.23

36.2.3 Fundamental Freedoms in EEA Tax Law The EEA Agreement contains the same fundamental freedoms granted in the TFEU but with some deviations. Even if a positive integration in the direct tax field does not take place in the EEA EFTA states, a kind of negative integration by the fundamental freedoms still applies. For cases concerning taxation, the freedom of establishment and the freedom of capital movements are those most relevant.

36.2.3.1 Freedom of establishment In the Keller Holding24 case, the ECJ applied the fundamental freedoms of the EEA Agreement to a direct tax case for the first time when it pointed out that the German national dividend legislation was precluded by the freedom of establishment according to article 31 EEA Agreement. The case treated the non-deductibility of expenses relating to dividends from Austria at the level of the German parent company in 1994 and 1995.

20

ECJ 13 July 1988, C-102/87 France v. Commission. ECJ 21 March 1990, C-142/87 Belgium v. Commission. 22 EFTA Court 10 May 2011, Joined Cases E-4/10, E-6/10 and E-7/10 Liechtenstein and Others v. ESA, para. 72; and EFTA Court 30 March 2012, Joined Cases E-17/10 and E-6/11, para. 52. 23 EFTA Surveillance Authority Decision No. 149/ 99/COL of 30 June 1999 introducing guidelines on the application of State aid rules to measures relating to direct business taxation and amending for the 19th time the Procedural and Substantive Rules in the field of State aid [2000] OJ L137, ch. 17B.5 Procedures. 24 ECJ 23 February 2006, C-471/04 Keller Holding. 21

International Tax Law and the EEA/EFTA 651 In another decision, the EFTA Court allowed cross-border group contributions with tax effect in the case Yara International ASA v. Norwegian Government25 concerning the Norwegian intra-group contribution rules. In accordance with these rules, the group contribution reduced the transferor’s taxable income and was included in the recipient’s taxable income regardless of whether the recipient made a loss or a profit for tax purposes. However, the rules required that both the transferor and the recipient were liable to taxation in Norway. Under that condition, Yara International ASA was denied a tax deduction for group contributions to its Lithuanian subsidiary that was later liquidated. The EFTA Court concluded that the Norwegian rules infringed the freedom of establishment protected by article 31 EEA Agreement when they denied Norwegian companies a tax deduction for group contributions to a group company in an EEA state with a tax loss in that state, and the loss incurred by the foreign group company was final. The parties agreed that the condition constituted a restriction of the freedom of establishment, but that it could be justified by overriding reasons in the public interest and that it was appropriate to attain a legitimate objective. The issue at stake was, however, the extent to which that condition was proportionate. The EFTA Court found that as for the proportionality test, there was no reason to distinguish between the different schemes of tax consolidation, such as loss transfers (group relief) under UK tax law and profit transfers (group contribution schemes) as in Norwegian and Finnish law. The decisive point was whether the restriction was appropriate to ensure the attainment of a legitimate objective, such as safeguarding the balanced allocation of taxing powers between EEA states, and that it did not go beyond that necessary to attain that objective. When the intra-group contribution was made to cover a ‘final loss’ in a foreign group company, the purpose of maintaining the balanced allocation of taxing power could not justify the restriction. It was for the national court to determine whether the loss was final, the situation could be considered a wholly artificial arrangement, and a deduction could be denied on those grounds.26 Also related to the Norwegian rules on intra-group tax deduction was the case PRA Group Europe AS v Staten v/Skatteetaten,27 where the EFTA Court found that the freedom of establishment was infringed by the fact that only companies with Norwegian-resident group members could use the Norwegian group contribution rules to mitigate the effects of the national limited interest deduction of 30% of the earnings before interest, taxes, depreciation, and amortization. In contrast, the Norwegian group contribution rules did not apply to groups with foreign companies. The Court held that Norwegian-based companies, which form part of a group with companies of other EEA states, were placed at a disadvantage vis-à-vis companies in entirely Norwegian-based groups because only the latter were able to soften the impact of the limited interest deduction.

25

EFTA Court 13 September 2017, E-15/16 Yara International ASA v. Norwegian Government. Ibid., para. 55. 27 EFTA Court 1 June 2022, E-3/21 PRA Group Europe AS v. Staten v/Skatteetaten. 26

652 Patrick Knörzer In the Seabrokers28 case, the EFTA Court rendered its decision on the compatibility of the Norwegian legislation on maximum credit allowance for tax paid in a foreign state with article 4 EEA Agreement on the prohibition of discrimination on grounds of nationality and with article 31 EEA Agreement on the freedom of establishment. In the Olsen29 case, the Norwegian controlled foreign corporation rules were applied to the members of a family for whose benefit a trust had been established in Liechtenstein as a holding entity for shares in several companies. The EFTA Court stated, first, that the right of establishment in articles 31–34 EEA Agreement was granted both to natural persons who were nationals of an EEA state and to legal entities (‘companies or firms’), no matter whether they had legal personality or not, provided they had been formed in accordance with the law of an EU or EFTA state and had their registered office, central administration, or principal place of business within the territory of the contracting parties. Secondly, the EFTA Court also recognized that the prevention of tax avoidance may provide a justification, but only where the measures taken targeted wholly artificial arrangements which did not reflect economic reality. The assessment of the facts in that respect was a matter for the national court. For this, the national court must be enabled to carry out a case-by-case examination, taking into account the particular features of each case, based on objective elements, in order to assess the abusive or fraudulent conduct of the persons concerned. Accordingly, the discriminatory tax measure must not be applied where it is proven, on the basis of objective factors, that a taxpayer was actually established in an EEA state and carried on genuine economic activities, which took effect in the EEA.

36.2.3.2 Free movement of capital Article 40 EEA Agreement prohibits any restriction and discrimination of the free movement of capital amongst the contracting parties of the EEA. Following this article, ‘there shall be no restrictions between the Contracting Parties on the movement of capital belonging to persons resident in EC member states or EFTA states and no discrimination based on the nationality or on the place of residence of the parties or on the place where such capital is invested’. Annex XII contains the provisions necessary to implement this article. This annex declares the EU Capital Liberalization Directive30 and its Annex I as applicable for the EEA. Article 1(1) of the directive obliges the EEA states to abolish restrictions on movements of capital taking place between persons resident in the EEA states. The article refers to a non-exhaustive nomenclature in Annex I to the directive, in which capital movements are classified. Point (a) under Heading III of the nomenclature classifies operations in shares and other securities of a participating nature as capital movements. 28

EFTA Court 7 May 2008, E-7/07 Seabrokers v. Staten v/Skattedirektoratet. EFTA Court 9 July 2014, Joined Cases E-3/13 and E-20/13 Fred. Olsen and Others v. Norwegian State, para. 173. 30 Council Directive 88/361/EEC of 24 June 1988 for the implementation of Article 67 of the Treaty [1988] OJ L178/5. 29

International Tax Law and the EEA/EFTA 653 The ECJ ruled that the substantial content of article 40 EEA Agreement is identical in substance to the relevant provision in the TFEU31 but still has some remarkable differences. Contrary to article 40 EEA Agreement, article 65(1)(a) TFEU explicitly allows EU member states to distinguish between resident and non-resident taxpayers and apply tax law provisions which distinguish according to the place where capital is invested. Besides, article 40 EEA Agreement does not contain an explicit justification based on public policy and public security, as does article 65(1)(b) TFEU. Furthermore, it has to be emphasized that article 40 EEA Agreement—in contrast to article 63 TFEU—does not cover cases outside the EEA. Instead, the EEA Agreement only provides for free movement of capital between residents of the EU member states and the EEA EFTA states. As a consequence, residents of an EEA EFTA state may claim free movement of capital with the twenty-seven EU member states under article 63 TFEU and with the thirty EEA member states under article 40 EEA Agreement.

36.2.4 Taxation of Dividends The EU Parent–Subsidiary Directive is not applicable for EEA states, as companies generally must be incorporated under the law of an EU member state whereas companies with their seat in an EEA EFTA state are not enumerated in the directive. For the purposes of the Parent–Subsidiary Directive, EEA EFTA states are treated as third countries which leads to the fact that EEA/EFTA-incorporated companies are not eligible for relief from withholding taxes under the directive. However, relief from domestic withholding taxes can be based on the fundamental freedoms if certain conditions are met.

36.2.4.1 Outbound dividends Regarding dividend taxation, one of the earlier and certainly most important decisions of the EFTA Court is the Fokus Bank case. The Norwegian imputation credit system allowed credits to resident, but not non-resident, individual shareholders. The credit was computed using the shareholder’s tax rate, so dividends received by domestic shareholders were effectively exempt. In addition, dividends to foreign, but not domestic, individual shareholders were subject to a final withholding tax. The EFTA Court found that the Norwegian imputation tax credit system was a restriction within the meaning of article 40 EEA Agreement which provides for the free movement of capital belonging to persons resident in EU or EFTA states. The court noted that the Norwegian tax legislation at issue could adversely affect the profit of non-resident shareholders and could thereby have the effect of deterring them from investing capital in companies based in Norway. Further, the court stated that the provisions could impede Norwegian companies from raising capital outside Norway. The Fokus Bank

31

ECJ 23 September 2003, C-452/01 Ospelt und Schlössle Weissenberg.

654 Patrick Knörzer case set a precedent specifying that EEA member states cannot tax investors from other member states at a higher rate without justification. Outbound dividends must not be treated differently from inbound dividends (i.e. dividends that non-resident companies pay to resident shareholders). A comparison can be made with the ECJ’s decision in the Manninen32 case.

36.2.4.2 Inbound dividends The source state is not allowed to levy a withholding tax on dividends paid to another EEA state when dividends paid to resident companies are effectively tax-exempt. In Amurta,33 the ECJ stated that the less favourable treatment of dividends paid to non- residents when compared to similar payments to residents constituted a restriction on treaty freedoms. In accordance with the ECJ, the freedom of establishment must be interpreted as precluding the national legislation of a member state when the national legislation exempts dividends distributed by a subsidiary resident in that state but charges withholding tax on similar dividends paid to a parent company which is resident in another member state.34 Another landmark decision addressing the freedom of establishment was made by the ECJ in Denkavit.35 In this case, the ECJ obliged the source state to give up the withholding tax levied on outbound dividends payments when allowing resident parent companies almost full exemption from such tax. Both cases, Denkavit and Amurta, are highly relevant for EEA EFTA states because they did not fall within the scope of the Parent–Subsidiary Directive that grants an exemption from withholding taxation under the conditions stated therein. In Denkavit, the ECJ decided on dividends which had been received even before the directive came into force. The Amurta case pertained to shareholders with a small participation for which the directive was not applicable.36 In Commission v. Italy,37 the ECJ ascertained that the less favourable treatment which the Italian legislation accorded to dividends distributed to companies established in EEA states constituted a restriction on the free movement of capital for the purposes of article 40 EEA Agreement. A slightly different treatment of the dividend happened in Commission v. Germany.38 The German rules were structured in such a way that parent companies domiciled in Germany could credit the withholding tax against the final corporate income tax, or if the corporate income tax was lower than the withholding tax, obtain a refund of the exceeding withholding tax. For foreign parent companies, the German withholding tax was final. The ECJ found that the free movement of capital under the EEA Agreement

32

ECJ 7 September 2004, C-319/02 Manninen. ECJ 8 November 2007, C-379/05 Amurta. 34 ECJ 18 June 2009, C-303/07 Aberdeen. 35 ECJ 14 December 2006, C-170/05 Denkavit. 36 See M. Quaghebeur, ‘ECJ Condemns Belgian Thin Cap Rule’, Tax Notes International (2008), 372. 37 ECJ 19 November 2009, C-540/07 Commission v. Italy. 38 ECJ 20 October 2011, C-284/09 Commission v. Germany. 33

International Tax Law and the EEA/EFTA 655 had been infringed, since the discrimination also affected companies established in EEA states.

36.2.5 Lack of Information Exchange as Potential Justification for Discrimination? It is crucial whether a different view on restrictions of the fundamental freedoms in relation to EEA EFTA states can be taken if a comprehensive exchange of information does not exist. In some cases, the ECJ accepted the justification of discriminatory measures by lacking assistance in tax matters. Such exchange of information procedures are established within the EU by directives, not being applicable under the EEA Agreement, or by bilateral agreements providing for an equivalent procedure. In Rimbaud,39 a Liechtenstein-resident company owned immovable property in France and—in contrast to French companies—had to pay French estate tax of 3% of the market value of the real estate. The ECJ found this restriction of the freedom of capital movements under the EEA Agreement to be justified as no information exchange between Liechtenstein and France was in place at that time. An EU member state is entitled to maintain the restriction if the non-EU member state is not bound under an agreement to provide information, because of the lack of a general system for the exchange of information comparable to the directive. The problem is no longer relevant for Liechtenstein since the Tax Information Exchange Agreement between France and Liechtenstein and the Multilateral Convention on Mutual Administrative Assistance and Exchange of Information (MAC) are now applicable40 and will most probably prevent an identical assessment of the facts as happened in the Rimbaud case. A similar question was crucial in the A41 case when a Swedish-resident individual received dividends from a Swiss company in the form of shares in a subsidiary. Sweden had concluded a tax treaty with Switzerland which did not contain an exchange of information clause and refused to exempt the dividends for that reason. The ECJ found the restriction of the free movement of capital to be justified. In Elisa,42 based on almost identical facts to Rimbaud but involving a Luxembourg company, the ECJ deemed the French legislation as infringing the free movement of capital as laid down in the TFEU. The ECJ held that taxpayers themselves must have the possibility to produce evidence in order to benefit from a certain advantage even if there is no obligation to exchange information due to a limitation of the directive, and therefore there is no possibility for the EU member state to verify the data provided. The difference between the situation in Elisa and in Rimbaud is that the information

39

ECJ 28 October 2010, C-72/09 Établissements Rimbaud SA v. Directeur général des impôts. See A. Ess, ‘Liechtenstein’, Cahiers de Droit Fiscal International vol. 105B (2020), 527. 41 ECJ 18 December 2007, C-101/05 Skatteverket v. A. 42 ECJ 11 October 2007, C-451/05 ELISA, paras 90ff. 40

656 Patrick Knörzer provided by the taxpayer could be verified by using the Mutual Assistance Directive in Elisa, whereas it was not possible in a third-country setting in the absence of a bilateral agreement between the two states involved. In Commission v. Netherlands,43 the ECJ held that dividend payments from the Netherlands to the EEA states Iceland and Norway must not be treated less favourably than dividend payments within the Netherlands or from the Netherlands to EU member states. The Netherlands argued that the Mutual Assistance Directive did not apply in relation to Iceland and Norway and that it was harder to enforce an obligation based on international law than on an EU directive. According to the Netherlands, it was not possible to verify whether the legal requirements for the exemption were fulfilled in this case. However, there were bilateral agreements in place to avoid double taxation between the Netherlands and the two EEA states which contained an information exchange clause. In A Oy,44 concerning a Finnish dividend paid to a Norwegian parent company, the Finnish court had to confirm that the exchange of information under the Nordic Convention on mutual administrative assistance in tax matters45 was as effective as that provided for in EU directives on administrative cooperation. For EEA EFTA states, the question whether a lack of administrative assistance in tax matters can justify discriminatory measures, is relevant for the admissibility of withholding taxes and seems to be especially important since the pertinent EU directives are not applicable on them. With each expansion of these directives to enhance administrative cooperation within the EU, the regulatory gap between the EEA EFTA states on the one side and the EU member states on the other side is deepened. Conversely, all EEA EFTA states are parties to the Convention on Multilateral Assistance in Tax Matters and therefore apply identical provisions.

36.2.6 Exit Taxation For some decades, an extensive case law on exit taxation in European tax law has been developed. The ECJ held that the freedom of establishment, according to article 31 EEA Agreement should be interpreted in the same way as the freedom of establishment in article 49 TFEU and, therefore, the exit tax of an EU state could potentially be justified on the basis of ensuring the effective collection of taxes.46 With regard to EEA EFTA states, the exit taxation of legal persons under the Anti-Tax Avoidance Directive (ATAD)47 is to be considered, too. Although the ATAD is addressed to the EU member states and

43

ECJ 11 June 2009, C-521/07 Commission v Netherlands. ECJ 19 July 2012, C-48/11 A Oy, para. 37. 45 Treaty No. 37/1991 which is applied in Iceland, Norway, Sweden, Finland, Denmark, and the Faroe Islands. 46 ECJ 18 July 2013, C-261/11 Commission v. Denmark, para. 41. 47 Council Directive (EU) 2016/1164 of 12 July 2016 laying down rules against tax avoidance practices that directly affect the functioning of the internal market. 44

International Tax Law and the EEA/EFTA 657 not applicable for EEA EFTA states, it is the only one of the tax directives that includes provisions especially directed at EEA EFTA states. Specific rules for relocations to non- EU states which are party to the EEA agreement must be observed that deviate from the rules for relocations to third countries. When the taxpayer makes a transfer of assets or the transfer of their tax residence from an EU state to an EEA EFTA state, the exit tax can be paid in instalments over a five-year period (without interest) on application if the member state of the taxpayer or the EU has concluded an agreement on mutual assistance for the recovery of tax claims, equivalent to the mutual assistance provided for in Directive 2010/24/EU, with the respective EEA EFTA state. If the EEA EFTA state grants assistance in the recovery of foreign taxes, the deferral of exit taxes in respect of EEA states is no different from the deferral of exit taxes in regard to EU states. For transfers to EEA EFTA states without such an agreement, or to any third jurisdictions, tax deferrals are not permitted under the ATAD. If the EEA state only offers little or no assistance in the recovery of the exit tax, the deferral of the exit tax entails an increased risk of non-recovery for the EU state which loses its taxing right with the transfer.48 Here, the European Commission is of the opinion that the EU member states should be allowed to safeguard their tax claims at the moment of transfer of the assets if there is no adequate information exchange relationship with the EEA state concerned.49 As pointed out earlier, exit taxation in the EEA EFTA states themselves is not determined by the ATAD. However, their exit tax provisions have to be in line with the fundamental freedoms and have been at issue in some cases. The right to exit is protected within the EEA, since any restriction against free movement must comply with the proportionality principle in accordance with the well-known ECJ cases of Centros,50 Cartesio,51 and National Grid Indus.52 When the Norwegian company Arcade Drilling AS transferred its central administration from Stavanger, Norway to Aberdeen, Scotland, the Norwegian tax authorities assessed a liquidation tax although the company was not deregistered. The EFTA Court found that a general anti-avoidance rule would be liable to deter a company incorporated under Norwegian law from relocating its place of management to another EEA state and that such a restriction would be in breach of the proportionality principle, if any possibilities of deferring the payment were denied. By noticing that a deferment implied a risk of non-recovery, the EFTA Court recognized that Norway retained the right to take measures to prevent tax avoidance. This risk could nonetheless be hedged by less restrictive means. The EFTA Court held that ‘[a]‌bank guarantee might even be unnecessary if the risk of non-recovery is covered by the personal liability 48 See

M. Tell, ‘Exit Taxation within the European Union/ European Economic Area: After Commission v. Denmark (C-261/11)’, European Taxation 54/2–3 (2014), 47ff. 49 Communication from the Commission to the Council, the European Parliament and the European Economic and Social Committee—Exit taxation and the need for co-ordination of member states’ tax policies, COM(2006) 825 final (19 Dec. 2006). 50 ECJ 9 March 1999, C-212/97 Centros. 51 ECJ 16 December 2008, C-210/06 Cartesio. 52 ECJ 29 November 2011, C-371/10 National Grid Indus.

658 Patrick Knörzer of shareholders for outstanding tax debts of the company’.53 It was up to the Oslo District Court to apply the proportionality test.54 The exit taxation rules of Iceland which had provided for the immediate taxation of companies and shareholders on the cross-border mergers of Icelandic companies with companies from other EEA states, without the possibility of a deferral, were a breach of the freedom of establishment and the free movement of capital according to the EFTA Court.55 As a reaction to this ruling, Iceland changed its legislation in a way that allows companies that intend to merge cross-border (within the EEA) to choose between two options: either they can immediately pay tax on the unrealized capital gains relating to assets and shares or opt for a deferral of the tax payment for up to five years.

36.2.7 Indirect Taxes Articles 14 and 15 EEA Agreement, which are identical to articles 110 and 111 TFEU, stipulate certain restrictions on the indirect taxation of imports and exports of goods by EEA states which are a key element of the internal market. Article 14 EEA Agreement prohibits any internal taxation of any kind, directly or indirectly, on the products of other EEA states in excess of those imposed directly or indirectly on similar domestic products. On the interpretation of this provision, a rich and varied case law has developed in the last decades. The Einarsson case relates to the higher VAT rate levied on books in languages other than Icelandic. In its advisory opinion on the case, the EFTA Court came to the conclusion that such different tax rates were a breach of article 14 EEA Agreement. It held that the Icelandic legislative measure was incompatible with the EEA provisions and could not be justified ‘on grounds relating to the public interest of enhancing the position of the national language’.56 Furthermore, EEA states must not impose VAT on the importation of products from another EEA state supplied by a private person where no such tax is levied on domestic supplies of similar products for private purposes in the state of importation, if the VAT on the importation does not take account of tax paid in the exporting EEA state which is still contained in the value of the goods at the time of the importation.57 The concept of taxation within the meaning of article 14 EEA Agreement must be interpreted in a wide sense. The phrase ‘directly or indirectly’ include ‘all taxation which is actually and specifically imposed on the domestic product at all earlier stages of its 53

EFTA Court 3 October 2012, E-15/11 Arcade Drilling, para. 105. the Norwegian court decisions, see E. Matei, ‘Dual residence and the right to migrate under EEA law (Arcade Drilling)’, Kluwer International Tax Blog (18 August 2015), http://kluwertaxblog.com/ 2015/08/18/dual-residence-and-the-right-to-migrate-under-eea-law-arcade-drilling/. 55 EFTA Court 2 December 2013, E-14/13 The EFTA Surveillance Authority v. Iceland. 56 EFTA Court 22 February 2002, E-1/01 Einarsson, para. 46. 57 ECJ 5 May 1982, C- 15/ 81 Gaston Schul Douane- expediteur BV and 21 May 1985, C- 47/ 84 Staatssecretaris van Financiën .v Gaston Schul Douane-Expediteur BV. 54 For

International Tax Law and the EEA/EFTA 659 manufacture and marketing or which corresponds to the stage at which the product is imported from other member states’.58 The further back one goes in the manufacturing chain, the weaker the effect of the imposed tax on the price of the final product will be.59 Payment for services provided by a public entity constitutes taxation for the purposes of article 14 EEA Agreement when customers have no option but to rely on the services of that provider.60 In this context, it cannot matter whether the price has been set according to purely commercial criteria, focusing on profit maximization, or according to considerations of the general good usually associated with the exercise of public authority. According to article 14(2) EEA Agreement, no EEA state shall impose on the products of other EEA states any internal taxation of such a nature as to afford indirect protection to other products. These rules must guarantee that internal taxation is neutral for the purposes of competition between exported and imported products.61 The decisive factor is the effect, on the market in question, of reducing the potential consumption of imported products to the advantage of competing domestic products. For this to be the case, it is not sufficient that the relevant products are in competition with one another. It must further be demonstrated that the higher tax rate applies chiefly to the imported products, and that the difference in tax burden caused by the charge in question would have an effect on the cross-elasticity of demand. In making this assessment, one must take into account, inter alia, the discrepancy in price which may exist between the products independently of that difference.62 Another case before the EFTA Court concerned flights from Iceland to other EEA states that were subject to a higher tax than that for domestic flights and flights to Greenland and the Faroe Islands. The contested tax levied per passenger travelling on intra-EEA flights was seven times higher than the tax levied per passenger travelling on domestic flights. This clearly constituted an unjustified restriction on the freedom to provide services according to the EFTA Court.63 As a complementary rule, article 15 EEA Agreement bans any repayment of internal taxation that exceeds the internal taxation imposed on products which are exported to other EEA states, whether directly or indirectly. Excessive tax refunds to exporting companies may subsidize exports and can distort competition on the EEA market. Behind this view stands the destination principle under which goods are supposed to be taxed in the state of importation. The ECJ pointed out that ‘direct’ taxation according to article 15 EEA Agreement refers to tax on the finished product itself, whereas ‘indirect’ taxation refers to taxes during the various stages in the production chain.64 58 ECJ 3 April 1968, C- 28/ 67 Firma Molkerei- Zentrale Westfalen/ Lippe GmbH v. Hauptzollamt Paderborn. 59 See P. C. Müller-Graff, ‘Free Movement of Goods’, in C. Baudenbacher, ed., The Handbook of EEA Law (Cham: Springer, 2016), 425. 60 EFTA Court 5 March 2008, E-6/07 HOB vín ehf and Faxaflóahafnir sf, para. 34. 61 ECJ 8 April 2008, C-167/05 Commission v. Sweden. 62 EFTA Court 5 March 2008, E-6/07 HOB vín ehf and Faxaflóahafnir sf, para. 52. 63 EFTA Court 12 December 2003, E-1/03 ESA v. Ireland. 64 ECJ 19 November 1969, C-45/64 Commission of the European Communities v. Italian Republic.

660 Patrick Knörzer

36.2.8 Insurance Premium Tax The EU insurance directives are also relevant for EEA states and includes rules concerning taxation of insurance premiums. Article 157 Solvency Directive II determines in which EEA member state insurance premiums will be taxed without imposing an obligation on that state for actual taxation. Taxation of the premium is the competence of the member state in which the risk is situated65 or the member state of the commitment. For life insurance contracts, the habitual residence of the policyholder or, if the policyholder is a legal person, the policyholder’s establishment to which the contract relates. If a policyholder moves to a different state within the EEA, the ECJ has stated that the insurance premium tax must be paid in the state where the policyholder is habitually resident at the moment the premium is paid.66 Of all thirty-one EEA jurisdictions (including the UK), twenty-two currently levy an insurance premium tax or stamp duty on insurances written under the freedom of services regime. Among the EEA EFTA states, Norway does not levy any insurance tax at all. Iceland implemented a parafiscal charge on property insurance where a fee of 0.00021% is charged on the assessed sum insured value on a compulsory fire insurance. Only Liechtenstein levies a classic insurance premium tax under the Swiss Stamp Duty Act but had to adapt it to the EEA rules in the major tax reform in 2011 and added premium payments for life insurance contracts concluded by a domestic policyholder with a foreign non-supervised insurer as subject to insurance premium tax.

36.3 The Future of Taxation in EEA EFTA States The development of taxation in the EEA EFTA states during the next decades is not easy to foresee. A harmonization of direct and indirect taxes within the EEA EFTA states or between EFTA and the EU has not taken place and seems not to be intended at present. As the EEA EFTA members are reduced to only four states with different political goals in the field of taxation and also in other policy areas, they do not pursue an active uniform tax policy. Their internal relationships in taxation are still based on bilateral and multilateral agreements and the relationship with the EU differs from state to state. Switzerland, in particular, has a special relationship with the EU which is solely based on bilateral agreements. As the only EFTA state which is not an EEA state, Switzerland has regulated its tax relationship with the EU largely on a bilateral basis. The EEA is more than just the ‘waiting room’ for full EU membership, as the current EEA EFTA states have held that status for more than twenty-five years now and an 65

66

ECJ 17 January 2019, C-74/18. ECJ 21 February 2013, C-243/11 RVS Levensverzekeringen NV v. Belgische Staat.

International Tax Law and the EEA/EFTA 661 accession of those states to the EU does not seem to be intended not even in the long term. Although taxation in Iceland, Liechtenstein, Norway, and Switzerland is strongly influenced by taxation in the EU, there are very few EU tax provisions applicable to EEA EFTA states, as these states refrain from harmonization of their tax laws. Furthermore, the EEA Agreement neither contains a common fiscal policy nor a customs union. In contrast to the EU, each EEA EFTA state has its own external tariffs system, with the exception of Switzerland and Liechtenstein which form a customs union based on a bilateral treaty. Moreover, the EEA EFTA states predominantly adopt their tax provisions from OECD standards and not from the EU, for example the Multilateral Instrument, the MAC, and the common reporting standards were all developed at the OECD level. The role of the EU is to ensure that there is no discrimination caused by the application of the national tax laws of the EU states and the EEA states. The EEA Agreement stipulates that EEA states must encompass some EU legislation, including all the fundamental freedoms, and covering certain other complementary areas, such as competition and state aid rules, consumer protection, company law, environment, and social policy statistics, but they do not need to adopt the EU’s direct or indirect taxation rules. Furthermore, EEA EFTA states do not have to transpose the EU directives on administrative cooperation in the field of taxation and in mutual assistance for the recovery of claims relating to taxes, duties, and other measures. In that respect, the limitations of the EEA in respect to exchange of information may lead to a situation where companies resident in EEA EFTA states are treated differently from companies with their residence in the EU. A further integration of the EEA EFTA member states into the EU is a highly political matter. On the one hand, such integration would be gradually strengthened because many EU directives are already transposed into EEA law. On the other hand, the deepening of EU integration and harmonization of the national tax provisions tends to leave the EEA EFTA states lagging as, at present, there seems to be no political intention towards ‘full’ EU membership. Nevertheless, the EU autonomously tries to define international tax standards in its code of conduct in business taxation complemented by a list of non-cooperative jurisdictions which is also relevant for EEA EFTA states.

Chapter 37

T w ent y -F irst C e nt u ry Tax Challenge s of t h e E U Candidate C ou nt ri e s Savina Mihaylova-G oleminova

37.1 Introduction The Treaty on European Union (TEU) sets out the conditions (art. 49) and principles (art. 6(1)) to which any country wishing to become an EU member must conform. Certain admission criteria must be met. These criteria (i.e. the Copenhagen criteria) were established by the Copenhagen European Council in 1993 and strengthened by the Madrid European Council in 1995. They are: stability of institutions guaranteeing democracy, the rule of law, human rights, and respect for and protection of minorities; a functioning market economy and the ability to cope with competitive pressure and market forces within the EU; ability to take on the obligations of membership, including the capacity to effectively implement the rules, standards, and policies that make up the body of EU law (the acquis), and adherence to the aims of political, economic, and monetary union. For example, for the Republic of Bulgaria—in the field of taxation and tax law—the guiding principle is set out in article 4 of the Constitution of the Republic of Bulgaria, namely that the Republic of Bulgaria is governed by the rule of law. The state is governed by the constitution and national laws. The rule of law principle is consistently applied in Bulgaria\n legislation, finding expression in the principle of legality, established in the two codes regulating the activities of the administrative bodies and the financial administration—the Tax and Social Security Procedure Code (TSSPC) and the Administrative Procedure Code (APC), and, respectively, in substantive taxation law. The rule-of-law principle is weaved into the main principles of tax and administrative

664 Savina Mihaylova-Goleminova legislation. The rule-of-law requirement1 is among the main requirements for EU accession and one of the fundamental values of the EU.2 The Republic of Albania, Bosnia and Herzegovina, Iceland, Kosovo, Montenegro, the Republic of North Macedonia, the Republic of Serbia, and the Republic of Turkey are candidate countries. The European Council has decided to grant the status of candidate country to Ukraine and to the Republic of Moldova. The European Council is ready to grant the status of candidate country to Georgia once the priorities specified in the Commission’s opinion on Georgia’s membership application have been addressed.3 Negotiations are held with each candidate country to determine their ability to apply EU legislation (acquis) and to examine their possible request for a transition period. Potential candidate countries are Bosnia and Herzegovina and Kosovo.4 Taxation is especially complicated in the context of accession negotiations of EU candidate countries, given the fact that the European Union is built on a constitutionalized legal order, which aims to sustain effective policing of agreed rules not only at supra- state but also at national level while also protecting individual rights according to international treaty law. The EU operates on the basis of the competence conferred on it by its member states pursuant to the founding treaties. Taxes are a symbol of the sovereignty of the state. Member states have broad sovereignty in the area of direct taxation, but within the rules of the EU Treaties. In accordance with articles 4 and 5 TEU, the European Union acts only within the limits of the competences conferred on it by the member states in the founding treaties to attain the objectives set out therein. All other competences belong to the member states. This scope of competence corresponds to the legality principle and the principle of legal certainty. The competence of the European Union based on the founding treaties is, however, broad. In accordance with article 352 of the Treaty on the Functioning of the European Union (TFEU), the Council can take any action that is necessary for the attainment of the Union objectives in the functioning of the common market. Direct taxation falls under the scope of divided competence. Both the Union and the member states have competence. Positive harmonization of direct taxation in the member states is possible, provided that it affects the realization of the internal market, that harmonization actions are carried out unanimously, and that

1 For more information on the rule of law, see: S. Weatherill, Law and Values in the European Union (2016); P. Craig, ‘Formal and Substantive Conceptions of the Rule of Law: An Analytical Framework’ in The Rule of Law and the Separation of Powers (1997); H. L. A Hart et al., The Concept of Law (2012). 2 See Rule of Law Regulation applying from 1 January 2021, Regulation (EU, Euratom) 2020/2092 of the European Parliament and of the Council of 16 December 2020 on a general regime of conditionality for the protection of the Union budget [2020] OJ L433I/1. 3 See European Council conclusions on Ukraine, the membership applications of Ukraine, the Republic of Moldova and Georgia, Western Balkans and external relations (23 June 2022), https://www. consilium.europa.eu/en/press/press-releases/2022/06/23/european-council-conclusions-on-ukraine- the-membership-applications-of-ukraine-the-republic-of-moldova-and-georgia-western-balkans-and- external-relations-23-june-2022/ (accessed 27 June 2022). 4 See https://ec.europa.eu/environment/enlarg/candidates.htm (accessed 30 March 2021).

21st Century Tax Challenges of EU Candidate Countries 665 the subsidiarity principle is observed. There is no exclusive competence (art. 3 TFEU— customs union, the establishment of competition rules necessary for functioning of the internal market, monetary policy for those member states whose currency is the euro, the conservation of marine biological resources under the common fisheries policy, and the common commercial policy) in the area of taxation, but shared competence in the area of indirect taxation (art. 4 TFEU—shared competence applicable to the internal market, social policy, environment, consumer protection, energy, the areas of freedom, security, and justice; cohesion policy implementation, etc.).5 Article 6 TFEU identifies seven areas in which the Union enjoys supporting, coordinating, or supplementary competence and applies it in the context of taxation. Unlike the member states, the European Union does not exercise its competences in the field of taxation having primarily a revenue objective in mind. The rules governing the financing of the EU budget are, indeed, adopted on a different legal basis and by different institutional bodies. These differences are reflected in the EU Treaty (TFEU— also referred to as the EU Treaty) by the distinction drawn between ‘tax provisions’ (arts 110–113 TFEU (former arts 90–93 EC)) under Part III (Common Policies) and ‘Financial Provisions’ (arts 310–325 TFEU (268–280 EC)). Therefore, European tax law exists despite the absence of a genuine European tax system. As a consequence, those few EU Treaty articles which explicitly or implicitly refer to taxation find their justification in their contribution to the Union policies, and in particular to the objective of the achievement of the internal market. In order to further the internal market, the EU Treaty provides for two types of tax provisions which aim at removing obstacles to intra-Community/Union trade that result from the exercise of the taxation powers by the member states. The first type of EU Treaty provision enables the Council (and only the Council) to adopt harmonization directives in the field of taxation. The second type regards general prohibitions for member states to establish or maintain obstacles to intra-Community movement and trade. From the taxpayers’ perspective, such prohibitions create individual rights and freedoms, directly enforceable before national and European courts. In respect of indirect taxation, a distinction between empowerment provisions and—directly applicable—tax prohibitions, is clearly drawn in the EU Treaty. On the one hand, article 113 TFEU (art. 93 EC) empowers ‘the Council ... acting unanimously [in accordance with a special legislative procedure and after consulting the European] Parliament and the Economic and Social Committee, [to] adopt provisions for the harmonisation of legislation ... of indirect taxation to the extent that such harmonisation is necessary to ensure the establishment and the functioning of the Internal Market [and to avoid distortion of competition]’. This legislative power in the area of indirect taxation has been exercised as regards value-added tax (VAT), excise duties, and indirect taxes on raising capital. On the other hand, article 110 TFEU (art. 90 EC) prohibits discriminatory internal taxation. Together with article 30 TFEU (art.

5

According to art. 4(2)(a) TFEU, the Union and the Member States share competence in the area of internal market.

666 Savina Mihaylova-Goleminova 25 EC), prohibiting customs duties and charges having an equivalent effect, these tax prohibitions aim at ensuring the free movement of goods in the Union and the effectiveness of the customs union. As regards direct taxation, the above- mentioned two types of provisions— empowerment and prohibitions—are to be found in the EU Treaty, although their wording does not explicitly refer to taxation. Concerning Treaty articles founding the power to adopt regulations or directives in direct tax matters, it must be emphasized that the EU Treaty does not explicitly grant legislative competence to the Council in the area of direct taxation, neither alone nor jointly with the European Parliament. Moreover, article 114 TFEU (art. 95 EC) explicitly excludes taxation from its scope of application. This does not mean, however, that legislative acts regarding direct taxation cannot be adopted, but rather that such provisions can only be adopted on the basis of general clauses such as article 115 or 352 TFEU (art. 94 or 308 EC), and only to the extent that these acts serve Community objectives. Moreover, and independently of the provisions on taxation, the EU Treaty confers upon European citizens general rights and freedoms aiming at guaranteeing non-discrimination and freedom to circulate and to undertake economic activities throughout the Union. These rights and freedoms are the free movement of workers (arts 45–48 TFEU (arts 39–42 EC)), the right of establishment (arts 49–55 TFEU (arts 43–48 EC)), the freedom to provide and to receive services (arts 56–62 TFEU (arts 49–55 EC)), the free movement of capital and payments (arts 63–66 and 75 TFEU (arts 56–60 EC)), and, since the Treaty of Maastricht, the right to move and reside freely within the territory of the EU (art. 21 TFEU (art. 18 EC)). Since the scope of application of these rights and freedoms is not limited to the extent of the Union’s legislative competence, it encompasses the direct tax provisions of the member states. According to settled case law, ‘although, as Community law stands at present, direct taxation does not as such fall within the purview of the Community, the powers retained by the member states must nevertheless be exercised consistently with Community law’.6 In general, member states have broad sovereignty in the area of direct taxation within the limits of the Treaties.7 The acquis on taxation extensively covers the area of indirect taxation: VAT8 and excise duties. Currently, Chapter 16 Negotiation covers the following:9

6 See J. Malherbe, Challenges Facing Member States and Candidate Countries of the European Union in the Field of Taxation (2019), 31–44. 7 On EU tax law and direct taxation, see: M. Lang et al., Introduction to European Tax Law on Direct Taxation (2020); H. Marjaana, EU Tax Law—Direct Taxation (2017). 8 For more information about the VAT system, see: B. Terra and J. Kajus, A Guide to the European VAT Directives 2018 (2018); EU VAT Compass 2018/2019 (Amsterdam: IBFD, 2018); R. de la Feria, The EU VAT System and the Internal Market (2009); M. Lamensch, European Value Added Tax in the Digital Era. A Critical Analysis and Proposals for Reform (2015); R, de la Feria, VAT Exemptions: Consequences and Design Alternatives (2013). 9 See https://ec.europa.eu/neighbourhood-enlargement/policy/conditions-membership/chapters-of- the-acquis_en (accessed 30 March 2021).

21st Century Tax Challenges of EU Candidate Countries 667 The acquis on taxation covers extensively the area of indirect taxation, namely value-added tax (VAT) and excise duties. It lays down the scope, definitions and principles of VAT.10 Excise duties on tobacco products, alcoholic beverages and energy products are also subject to EU legislation. As concerns direct taxation, the acquis covers some aspects of taxing income from savings of individuals11 and of corporate taxes12. Furthermore, Member States are committed to complying with the principles of the Code of Conduct for Business Taxation, aimed at the elimination of harmful tax measures. Administrative co-operation and mutual assistance between Member States is aimed at ensuring a smooth functioning of the internal market as concerns taxation and provides tools to prevent intra-Community tax evasion and tax avoidance. Member States must ensure that the necessary implementing and enforcement capacities, including links to the relevant EU computerised taxation systems, are in place.

The main principles of EU law are direct effect and the supremacy13 of the EU acquis. An EU legal provision has direct effect when it confers, immediately and directly, to natural and legal entities within the EU, subjective rights which must be recognized and protected by national courts. In this respect, the challenges facing EU candidate countries in this field will be presented briefly, focusing on the present main issues.

37.2 System of Own Resources of the Budget of the European Union and Taxation Currently, the EU does not use direct taxes for its own resource-collecting purposes. According to Council Decision 2020/2053 on the system of own resources of the EU,14 we have a roadmap towards the introduction of new own resources as follows: first, 1 January 2023: own resources based on the taxation of digital services (for the purpose of fair taxation of the digital economy); and secondly, 1 January 2026: own resources based 10 For

more information on VAT, see https://ec.europa.eu/taxation_customs/business/vat_en (accessed 30 March 2021). 11 For more information on taxation of individuals, see https://ec.europa.eu/taxation_customs/ind ividuals_en (accessed 30 March 2021). 12 For more information on company taxation in the EU, see: https://ec.europa.eu/taxation_customs/ business/company-tax_en (accessed 30 March 2021). 13 This principle was enunciated in Case 6/64 Costa v. Enel case (1964) where it was held: ‘By creating a Community of unlimited duration, having . . . powers stemming from a limitation of sovereignty, or a transfer of powers from the States to the Community, the Member States have limited their sovereign rights, albeit within limited fields, and thus have created a body of law which binds both their nationals and themselves’. 14 Council Decision (EU, Euratom) 2020/2053 of 14 December 2020 on the system of own resources of the European Union and repealing Decision 2014/335/EU, Euratom [2020] OJ L424/1.

668 Savina Mihaylova-Goleminova on the common consolidated corporate tax base (share of corporate tax taxes calculated according to uniform rules in the EU).15 As regards the first point, many EU member states are following their own unilateral steps regarding the taxation of the digital economy.16 Austria, France, Hungary, Italy, Poland, Spain, and Turkey have implemented a digital services tax (DST). Belgium, the Czech Republic, and Slovakia have published proposals to enact a DST. Latvia, Norway, and Slovenia have officially announced or shown intentions to implement such a tax. In the opinion of Advocate General Kokott in European Commission v Republic of Poland,17 she highlights that: Even though the planned digital services tax at EU level and the tax on the retail sector in Poland are similar in this respect, the Commission considers the Polish tax to constitute State aid for smaller undertakings, which are ‘taxed at too low a level’. The Commission had therefore barred the implementation of that law in advance, pending the conclusion of the investigation procedure, which Poland—like Hungary in a parallel case—considers to be an infringement of its fiscal autonomy.18 Contrary to the submission made by the Commission, a profit-based income tax is also not unquestionably preferable (in the words of the Commission ‘appropriate’). On the contrary, around the world turnover-based income taxes are on the rise, as is shown by the Commission’s proposed digital services tax. This uses annual turnover as the basis for the taxation of undertakings. The Polish tax on the retail sector and the planned EU digital services tax are no different in this respect.19

As regards the second point, European harmonization efforts are also to create a harmonious corporate tax policy within the European Union, focusing on a common consolidated corporate tax base (CCCTB), first announced for the end of 2008. This ambitious project had already been suggested in 2001 in line with the Lisbon Strategy. Since 2004, working groups of member state experts and Commission delegates have been clearing the ground and the Parliament has issued resolutions to support the project. A new impulse was given to the project in 2010 and the Commission presented a proposal in 2011 and relaunched two proposals in 2016, distinguishing the common tax base and consolidation. The CCCTB is expected to provide a comprehensive and sustainable solution for removing numerous existing tax obstacles faced by European undertakings operating in more than one member state. More precisely, the objectives are the adoption of common rules defining the tax base—and not the tax rate—of companies, in order to reduce the compliance costs arising from the differences 15 See

‘Reform of the EU own resources’, Policy Department for Budgetary Affairs Directorate- General for Internal Policies PE 690.963 (March 2021), https://www.europarl.europa.eu/RegData/etu des/IDAN/2021/690963/IPOL_IDA(2021)690963_EN.pdf (accessed 30 March 2021). 16 Source: KPMG, Taxation of the digitalized economy, https://tax.kpmg.us/content/dam/tax/en/ pdfs/2020/digitalized-economy-taxation-developments-summary.pdf (accessed 30 March 2021). 17 Case C-562/19 P, 15 October 2020. 18 At 2. 19 At 54.

21st Century Tax Challenges of EU Candidate Countries 669 between the twenty-seven national corporate tax systems, and the creation of a consolidation mechanism at the European level, in order to permit cross-border compensation of losses and to avoid transfer pricing disputes. This latter goal inevitably implies the setting up of a—fair, equitable, and simple—sharing mechanism (‘apportionment’) of the consolidated tax base between the member states concerned, mainly in order to avoid artificial profit shifting between member states and to mitigate harmful tax competition. Since the project only concerns the tax base, each member state would then remain competent to apply its own tax rate to the portion of the company’s pan-European tax base attributed to its jurisdiction. Of course, such a thorough reform raises a number of issues. Some of them are more technical, such as, among others, the relation between the rules for the determination of the tax base and the existing accounting rules— national or international, the perimeter of the consolidation group, or the optional or compulsory character of the CCCTB. Other issues are more political, that is, the willingness to accept further integration in (direct) tax matters, the abandonment of the member states’ powers to grant tax incentives in the form of a reduction of the tax base, not to mention the necessity of improving cooperation between the member states. Indeed, one cannot underestimate the administrative and judicial apparatus that needs to be put in place to make the system work. At the moment, since various member states have clearly declared that they will not participate in such a project, the possibility of enhanced cooperation has already been discussed, although such an option would be, according to the Commission, a ‘last resort approach’. Moreover, enhanced cooperation would certainly add further complexity to the already sensitive issues to be resolved. Furthermore, when talking about protection of the financial interests of the EU in the field of own resources of the EU budget, it must be said that member states and candidate countries currently face a further challenge; namely, the fight against irregularities and fraud related to public funds from the EU budget, both in terms of revenue and expenditure. VAT20 is part of the own resources of the EU budget (art. 311 TFEU and Council Decision 2020/205321). In the context of the EU budget, article 325 TFEU applies: 1. The Union and the Member States shall counter fraud and any other illegal activities affecting the financial interests of the Union through measures to be 20 The

fight against tax fraud, tax evasion, and abusive practices is an objective recognized and supported by the Sixth Directive (see, inter alia, the judgments in Cases C- 255/ 02 Halifax and Others, ECLIEU:C:2006:121, para.. 71; C- 439/ 04 and C- 440/ 04 Kittel and Recolta Recycling, ECLI:EU:C:2006:446, para. 54; and C-80/11 and C-142/11 Mahagében and Dávid, ECLI:EU:C:2012:373, para. 41). In this light, the Court has repeatedly highlighted that legal entities cannot rely on EU law for abusive or fraudulent ends (see, inter alia, the judgements in Cases C-439/04 Kittel and Recolta Recycling, ECLI:EU:C:2006:446, para. 54; C-32/03 Fini H, ECLI:EU:C:2005:128, para. 32; and C-18/13 Maks Pen, ECLI:EU:C:2014:69, para. 26). 21 See art. 2, para 1(b), the application of a uniform call rate of 0.30% for all member states to the total amount of VAT receipts collected in respect of all taxable supplies divided by the weighted average VAT rate calculated for the relevant calendar year as stipulated in Council Regulation (EEC, Euratom) No. 1553/89. For each member state, the VAT base to be taken into account for this purpose shall not exceed 50% of GNI.

670 Savina Mihaylova-Goleminova taken in accordance with this Article, which shall act as a deterrent and be such as to afford effective protection in the Member States, and in all the Union’s institutions, bodies, offices and agencies. 2. Member States shall take the same measures to counter fraud affecting the financial interests of the Union as they take to counter fraud affecting their own financial interests. The main regulation providing the legal definition of the term is Council Regulation 2988/95.22 In this respect, the Financial Regulation—Regulation 2018/1046 on the fight against fraud23 is also of significance in the field of the EU budget. The evolution of corpus juris in the European Prosecutor’s Office will allow the investigation of crimes against EU financial interests and more specifically, those set out in Directive (EU) 2017/1371.24 In respect of revenue arising from VAT own resources, the directive will apply only in cases of serious offences against the common VAT system. For the purposes of the directive, offences against the common VAT system will be considered to be serious where the intentional acts or omissions defined in point (d) of article 3(2) are connected with the territory of two or more member states of the Union and involve a total damage of at least €10 million. For the purposes of this directive (art. 2 para 2, p. (c) and (d)), the following will be regarded as fraud affecting the Union’s financial interests in respect of revenue other than revenue arising from VAT own resources: any act or omission relating to the use or presentation of false, incorrect, or incomplete statements or documents, which has as its effect the illegal diminution of the resources of the Union budget or budgets managed by the Union, or on its behalf, non-disclosure of information in violation of a specific obligation, with the same effect, or non-disclosure of information in violation of a specific obligation, with the same effect. In respect of revenue arising from VAT own resources, any act or omission committed in cross-border fraudulent schemes in relation to the use or presentation of false, incorrect, or incomplete VAT-related statements or documents, which has as an effect the diminution of the resources of the Union budget, non-disclosure of VAT- related information in violation of a specific obligation, with the same effect, the presentation of correct VAT-related statements for the purposes of fraudulently disguising the non-payment or wrongful creation of rights to VAT refunds. The complexity is further exacerbated by the new challenges in the field of taxation at a global level and the ongoing process of modification of the rules of international 22 Council Decision (EU, Euratom) 2020/2053 of 14 December 2020 on the system of own resources of the European Union and repealing Decision 2014/335/EU, Euratom [1995] OJ L312/1. 23 Regulation (EU, Euratom) 2018/1046 of the European Parliament and of the Council of 18 July 2018 on the financial rules applicable to the general budget of the Union, amending Regulations (EU) No 1296/2013, (EU) No 1301/2013, (EU) No 1303/2013, (EU) No 1304/2013, (EU) No 1309/2013, (EU) No 1316/ 2013, (EU) No 223/2014, (EU) No 283/2014, and Decision No 541/2014/EU and repealing Regulation (EU, Euratom) No 966/2012 [2018] OJ L193/1. 24 Directive (EU) 2017/1371 of the European Parliament and of the Council of 5 July 2017 on the fight against fraud to the Union’s financial interests by means of criminal law [2017] OJ L198/29/

21st Century Tax Challenges of EU Candidate Countries 671 taxation (the Base Erosion and Profit Shifting (BEPS) Project). To conclude, candidate countries should take into account the ongoing reform in the field of protection of financial interests of the European Union.

37.3 Administrative Capacity of Tax Authorities, Cooperation, and Mutual Assistance between Member States Administrative cooperation and mutual assistance between member states is aimed at ensuring the smooth functioning of the internal market as concerns taxation and provides tools to prevent intra-Community tax evasion and tax avoidance. Professor Pistone differentiates between these two fundamental notions: The difference between tax evasion and avoidance is generally clear. While tax evasion is an open violation of tax law (which can evolve into its more serious form, generally called tax fraud, in the presence of manoeuvres aimed at hiding the violation), tax avoidance arises in connection with the exploitation of the friction between form and substance for the purpose of circumventing the scope of a tax provision. Aggressive tax planning consists of the exploitation of cross-border tax disparities to the advantage of taxpayers, which shift profits out of the country of value creation, often towards low-tax jurisdictions by making use of loopholes and technicalities in the international tax rules and mismatches between the different tax systems. Anti- avoidance rules are often scarcely effective to counter this phenomenon, especially since aggressive tax planning achieves an undue tax advantage across two different tax jurisdictions.25

With regard to direct taxation, the current challenge at global, European, and national levels is the prevention of tax fraud, circumvention of tax law, and the restriction of practices associated with ‘aggressive tax planning’, requiring designated measures to reduce loss to national budgets through the improvement of tax legislation and information exchange between the tax administrations of the member states. This is a challenge for candidate countries as well. Member states must comply with the principles of the Code of Conduct for Business Taxation,26 aimed at the elimination of harmful tax practices. An example can be given with the recent judgments in Cases T-778/16 Ireland v Commission and T-892/16 Apple Sales International and Apple Operations Europe v Commission, taking into account that we can mainly speak of tax aid in the classic sense in the field of direct taxes, as they are the ones falling within the scope of sovereignty of 25

Lang et al., Introduction to European Tax Law on Direct Taxation, 53–54 more information, see https://ec.europa.eu/taxation_customs/business/company-tax/harm ful-tax-competition_en (accessed 11 June 2018). 26 For

672 Savina Mihaylova-Goleminova the member state, while the EU has a coordinating function with regard to direct taxation through the prohibition of state aid pursuant to article 107(1) TFEU. It is precisely the negative tax integration that has actively been implemented by the EU over the last few years, with its interpretations on the prohibition for state aid—and state aid is part of the overall issue of competition.27 It must be noted that in the last few years, the EU has developed a package of taxation transparency measures as part of its ambitious programme for combating tax evasion by undertakings and harmful (unfair) tax competition in the EU. A main element of this package of measures is the introduction of automatic data exchange between member states relating to their tax rulings. The consolidated version of Council Directive 2011/ 16/EU28 on administrative cooperation in the field of taxation addresses precisely these issues. Another example is Council Directive (EU) 2017/1852 on tax dispute resolution mechanisms in the European Union.29 In the context of the Copenhagen criteria on the development of strong state bodies having the capacity to effectively implement the rules, standards, and policies that make up the body of EU law, we must mention control as an element of the management of tax administration on behalf of the revenue administrations—tax and customs authorities. Speaking of financial control, this is also closely related to the public finances of the European Union—EU budget revenue and control over these activities30 (arts 310–325 TFEU). The Financial Regulation also provides a legal definition for the concept of control for its purposes (art. 2, item 19). Financial control in the context of EU public finances and national public finances is at the basis of the functions and rights of tax and customs authorities having competency and rights in the field of public funds (both national and European) management and control (the stability of such bodies and their administrative capacity to effectively implement the acquis are elements of the Copenhagen criteria), including in the field of tax and customs administration. For example, in Bulgaria these are: first, tax and customs authorities—the National Revenue Agency (NRA), the National Customs Agency (which, among others, administrate EU budget revenue), and local municipal revenue administration (local fees and taxes); secondly, the bodies exercising control and audit over the preceding activities at: the Ministry of Finance (state aid, etc.), the Protection of the European Union Financial Interests Directorate (AFCOS) with the Ministry of Interior, the State Financial Inspection Agency (ex post control for legality and protection of public

27

I. Stoynev, EU Competition Protection Law, vol. I: Anti-Cartel Law (2018). Directive 2011/16/EU of 15 February 2011 on administrative cooperation in the field of taxation and repealing Directive 77/799/EEC [2011] OJ L64/1, https://eur-lex.europa.eu/legal-content/ EN/TXT/?uri=CELEX:02011L0016-20200701 (accessed 30 March 2021). 29 Council Directive (EU) 2017/1852 of 10 October 2017 on tax dispute resolution mechanisms in the European Union. For more information, see https://ec.europa.eu/taxation_customs/business/company- tax/resolution-double-taxation-disputes_en_en#heading (accessed 30 March 2021). 30 On EU public finances, see the 5th edition of European Union Public Finance, http://ec.europa.eu/ budget/news/article_en..cfm?id=201501061636 (accessed 30 March 2021). 28 Council

21st Century Tax Challenges of EU Candidate Countries 673 financial interests), etc. Furthermore, Chapter 32 Financial control is closely related to Chapter 16. Taxation and it is imperative to have synergies between the two chapters. Under the conditions of EU membership, the following apply to the above bodies: the general, internationally adopted principles of sound financial management (art. 33 of the Financial Regulation) and internal financial control applied by the state; external (the Audit Office and the European Audit Office), and internal audit; control measures relating to own funds and the expenditure portion of the EU budget and the EU multiannual financial framework; and the measures for the protection of the financial interests of the EU31 in the field of own revenue and expenditure of the Union. The latter have already been reviewed. The rights of the revenue administration are set out in the instruments of EU law. There is an understanding, in doctrine, that financial bodies participate in financial relations in their capacity as state bodies, with their public powers being expressed in two main directions: the first being the fact that they issue public documents which have a predominantly declarative effect as the bodies act within circumscribed powers, given the fact that taxes are usually set out by law; the second is the public power to impose penalties. Competence is determined by the functions ascribed to the respective body, while powers are granted for the implementation of those functions. For this reason, in the context of the Copenhagen criteria, we can use the term ‘right to good tax administration’ because the revenue administrations of EU candidates will administrate revenue comprising EU budget own revenue (VAT and custom duties), whereby the bodies of the future member state will act within the shared competence with EU institutions. The right to good administration generally arises from the EU Charter of Fundamental Rights32 and is one of the main administrative and procedural principles derived from EU Court case law which is set out in article 41 of the Charter with reference to article 51(1) of the same.33 This right should not be solely applied in cases where state bodies enforce EU law, as it is entwined with the fundamental rights and principles and should comprise an integral part of all constitutional principles. Good tax administration is a prerequisite for EU membership in accordance with the Copenhagen criteria, so that a candidate may have the capacity to implement EU laws. Another very important aspect should be highlighted in relation to the activities of the revenue administration. In Bulgaria, this administration is a budgetary organization within the meaning of the Public Finance Act and, as such,

31 Council Regulation (EC, Euratom) No 2988/ 95 of 18 December 1995 on the protection of the European Communities financial interests [1995] OJ L312/1. 32 [2016] OJ C202/389. 33 Along with the right to good administration, the following fundamental rights are also of significance in the field of taxation: respect for private and family life (art. 7); protection of personal data (art. 8); freedom of expression and information (art. 11); freedom to conduct a business (art. 16); right to property (art. 17); right to good administration (art. 41); right of access to documents (art. 42); freedom of movement and of residence (art. 45); right to an effective remedy and to a fair trial (art. 47); presumption of innocence and right of defence (art. 48); principles of legality and proportionality of criminal offences and penalties (art. 49); right not to be tried or punished twice in criminal proceedings for the same criminal offence (art. 50).

674 Savina Mihaylova-Goleminova is obligated to observe the ‘sound financial government’ principle which requires that public funds are spent and managed economically, efficiently, and effectively. Supporters of the doctrine note that the administration must observe the principles of good administration, namely: lawfulness, suitability of procedures, participation, transparency, rationality, suitability of purposes, reasonableness, justified expectations, legal certainty, and proportionality.34 A further interpretation of the principle is that where the state does not have a sound financial administration in place, it may be liable for damages including damages arising from violation of EU law. The state may be held liable for any damages caused by unlawful acts or actions on the part of its agencies and officials,35 thus the fundamental principle of EU law regarding the liability of the state for damages in the application of article 4(3) TEU applies.36 In addition, the Court of Justice of the European Union may declare national rules to be incompatible with EU law and in this case preliminary rulings admit merely indirect control of national legislation. In fact, in a preliminary decision, the Court have interpreted Community law to the extent that it may affect the specific legal provisions at stake in particular proceedings before a national judge. Several Bulgarian tax cases and ECJ preliminary rulings serve as examples: Cases С-2/09; С-203/10; С-621/10 and С-129/11; С-284/11; С-553/16; С-142/12; С-552/16; С-118/11; С-132/16 СЕС; C-550/11; С- 549/11; С-642/11; С-234/11; С-107/13; С-76/15; С-285/11; С-78/12; С-18/13; С-492/13; С- 138/12; С-111/14; С-242/18. Also, those pending: Cases С-4/20; С-257/20; С-1/21. Member states are obliged to accept all the consequences of the Court’s rulings and to implement them in their national law, in accordance with the general principles forming part of the Community’s legal order, such as effectiveness, equivalence, and legal certainty. According to the Court, when a national tax measure is found to infringe European law, taxpayers can obtain a refund of unduly paid taxes by claiming before the national jurisdictions according to the national procedural rules, which can lead to serious financial repercussions for the budget of a member state.

37.4 Bulgarian Experience and Ongoing Practices It should be noted that the start of Bulgarian accession negotiations was 15 February 2000. Bulgaria’s position with regard to Chapter 10 Taxation was presented at an

34 For

further information on the main principles, see P. Craig, Administrative Law, 5th ed. (2016), 560–731. 35 Constitution of the Republic of Bulgaria, art. 7. 36 See also Craig, Administrative Law, 293–294 and P. Pistone, ‘Part Five. Liability for Damages in European Law and Taxation’ in Legal Remedies in European Tax Law (2009).

21st Century Tax Challenges of EU Candidate Countries 675 intergovernmental conference held on 30 April 2001 and the chapter was temporarily closed on 10 June 2002. The National Revenue Agency was established towards the end of 2002. Since 1 January 2007, Bulgaria has been a member state of the European Union. The conditions and arrangements for admission were set out in the Protocol annexed to the Treaty concerning the accession of the Republic of Bulgaria and Romania. With the Treaty of Accession of the Republic of Bulgaria to the European Union, ratified by an act passed by the National Assembly on 11 May 2005 (State Gazette issue 40/12 May 2005, in force as of 1 January 2007), the Republic of Bulgaria acceded to the European Union. According to Article 2 of the Act Concerning the Conditions of Accession of the Republic of Bulgaria and the Adjustments to the Treaties on which the European Union is Founded, applicable pursuant to Article 2 of the Treaty of Accession of the Republic of Bulgaria to the European Union as of the date of accession, the acts adopted by the institution prior to the accession are binding for Bulgaria and are applied under the conditions laid down in those Treaties and the Act itself. According to Article 19 of the said Act, the acts listed in Annex III thereto must be adapted as specified in that Annex. Item 4 ‘Taxation’ of Annex III to Article 19 of the Act of Accession lists Sixth Council Directive and its various amendments. According to Article 53 of the Act of Accession, Bulgaria must put into effect the measures necessary for it to comply, from the date of accession, with the provisions of directives and decisions within the meaning of Article 249 of the EC Treaty unless another time limit is provided for in the Act. These measures must be communicated to the Commission at the latest by the date of accession or, where appropriate, by the time limit provided for in the Act. The Act does not set out a time limit for the transposition of Sixth Council Directive which means that this should be effected on 1 January 2007, given the text of Article 53 of the Act of Accession.37

The TSSPC38 of 29 December 2005 provides general rules on tax liabilities, liable persons, tax control, refund of overpaid or unduly paid tax amounts, related parties, methods of determining market prices, limitation periods, collection, and execution. The Republic of Bulgaria adopted, and to a considerable extent implemented, acquis in the field of taxation and did not envisage any major problems with the entry into force of the relevant legislation or its implementation by the date of accession. There are, however, some cases where transitional periods or derogations were achieved (e.g. the transitional period until 31 December 2007 for implementation of minimum excise rate on cigarettes).

37 Interpretative Decision No. 3/ 06.06.2008 under Interpretative Case No. 2/2008 of the General Meeting of Judges from 1st and 2nd Division of the Supreme Administrative Court, http://www.sac. government.bg/TD_VAS.nsf/d038edcf49190344c2256b7600367606/fd0e2a1652305ecbc2257e500027f d5b?OpenDocument (accessed 11 June 2018). 38 Promulgated in State Gazette Issue 105/29 December 2005, supplemented and amended SG Issue 105/11, December 2020.

676 Savina Mihaylova-Goleminova According to the Accession Agreement: Permanent derogation from the requirements of Article 24 of Council Directive 77/388/EEI on the VAT registration and exemption threshold—fixed at the level of 50,000 BGN (approx. EUR 25,000) achieved by 31 December 2006; Bulgaria and the Czech Republic apply a reduced rate of excise duty, of not less than 50 % of the standard national rate of excise duty on ethyl alcohol, to ethyl alcohol produced by fruit growers’ distilleries producing, on an annual basis, more than 10 hectolitres of ethyl alcohol from fruit supplied to them by fruit growers’ households. The application of the reduced rate shall be limited to 30 litres of fruit spirits per producing fruit growers’ household per year, destined exclusively for their personal consumption.

Bulgaria accepted Community acquis under Chapter 10 Taxation at that time of negotiations and applied it as of 1 January 2007 except in the cases where transition periods and derogations were granted. In accordance with the agreed transition periods, the Republic of Bulgaria submitted to the EU a schedule for the gradual increase of excise rates up to the Community minimum level of taxation. Commitments made with regard to strengthening and developing administrative capacity were as follows: creation of a National Revenue Agency (NRA); improving control over compliance with tax legislation—enhancing the audits function and introducing a computerized system; training team heads for international audit implementation; creation of a designated unit for tax fraud investigation and development of methods for tax fraud detection and prevention; enhancing the VAT refund process; improving the collectability of public receivables; efficient taxpayer services with a view to increasing voluntary compliance; regular provision of consultancy services; human resource development and elaboration of a Qualifications and Career Growth Plan for tax employees within the tax/revenue administration; development and implementation of a Code of Ethics; development of information technologies and introduction of an integrated information system; administrative cooperation and information exchange with the tax administrations of EU member states; building institutional capacity for the creation of a National Educational Centre; establishing a call centre; enhancing electronic audit capacities; and improving and expanding the scope of electronic services offered to taxpayers by the revenue administration. Two derogations were granted in the field of VAT: Bulgaria may grant a VAT exemption to taxable persons whose annual turnover is less than the equivalent in national currency of €25,600 (BGN 50,000). The requirement of the EU is for no more €5,000 (art. 24 of Directive 77/388/ЕЕС); international carriage of passengers may continue to be treated as a zero-rated taxable supply in accordance with article 28(3)(b) and Annex F(17) of Council Directive 77/388/EEC. This derogation will be applied until the implementation of the conditions set out in article 28(4) of the directive (i.e. until the Council has unanimously determined to abolish all existing derogations).

21st Century Tax Challenges of EU Candidate Countries 677 In the field of excise duty, one derogation and four transition periods were granted, as follows: derogations for the implementation of reduced excise duty rate on home-made rakia—this is limited per household to 30 litres of fruit spirit per annum destined exclusively for their own consumption; transition periods: primarily in order to achieve the overall minimum excise duty on cigarettes—by derogation from article 2, paragraph 1 of Directive 92/79/EEC, Bulgaria may postpone application of the overall minimum excise duty on retail selling price (inclusive of all taxes) for cigarettes of the price category most in demand until 31 December 2009, provided that during this period Bulgaria gradually harmonizes its excise duty rates with the overall minimum excise duty set out in the directive. Without prejudice to article 8 of Council Directive 92/12/EEC on the general arrangements for products subject to excise duty and on the holding, movement, and monitoring of such products, and having informed the Commission, member states may, as long as the above transition period applies, maintain the same quantitative limits for cigarettes which may be brought into their territories from Bulgaria without further excise duty payment as those applied to imports from third countries. Secondly, in order to achieve the minimum excise duty on fuels—by derogation from article 7 of Directive 2003/96/EC—Bulgaria may apply the following transition periods. Up to 1 January 2011 for unleaded petrol used as motor fuel, with a minimum level of taxation set at €359 per 1,000 where, as from 1 January 2008, the effective tax rate applied to unleaded petrol used as motor fuel must not be lower than €323 per 1,000 litres. Up to 1 January 2010 for gas oil and kerosene used as motor oil, with a minimum level set at €302 per 1,000 and to reach the minimum level of €330 per 1,000 litre up to 1 January 2013, and, as from 1 January 2008 the effective tax rate applied to gas oil and kerosene used as motor fuel must not be lower than €274 per 1,000 litres. Thirdly, in order to achieve the minimum excise duty on coal and coke—by derogation from article 9 of Directive 2003/96/EC—Bulgaria may apply the following transition periods: up to 1 January 2010 for coal and coke used for central heating purposes with a minimum level of taxation for business purposes at €0.15 per gigajoule and for other purposes—at €0.30 per gigajoule; up to 1 January 2009 for coal and coke used for purposes other than central heating with a minimum tax rate for business purposes set at €0.15 per gigajoule and for other purposes—at €0.30 per gigajoule. As from 1 January 2007 the effective tax rate applied to these energy products must not be lower than 50% of the respective minimum Community rate. Fourthly, in order to achieve the minimum excise duty on electrical energy—by derogation from article 10 of Directive 2003/96/EC—Bulgaria may apply a transition period until 1 January 2010 for electrical energy with a minimum tax rate for business purposes set at €0.5 per MWh and for other purposes—at €1.00 per MWh. The effective tax rate applied to electrical energy must not be lower than 50% of the respective Community minimum rate. In the field of direct taxation, a transition period was granted until 31 December 2014 for non-implementation of article 1 of Directive 2003/49/EC on a common system of taxation applicable to interest and royalty payments made between associated

678 Savina Mihaylova-Goleminova companies of different member states. During the transition period, Bulgaria was allowed to impose a 10% tax at source on royalties for the first four years and 5% for the remaining period.

37.5 Conclusions Clearly, the twenty-first-century tax challenges of the EU candidate countries are huge, especially so in a pandemic, and require additional efforts and mobilization. The Western Balkans have their historic moment for acceleration of the accession process, bearing in mind the Russian war of aggression against Ukraine and its different dimensions and also the latest developments on discussions between Bulgaria and North Macedonia. To this end, the financial instruments of the EU should be used to the maximum extent, for example IPA III,39 which supports eight beneficiaries (Republic of Albania, Bosnia and Herzegovina, Iceland, Kosovo, Montenegro, the Republic of North Macedonia, the Republic of Serbia, the Republic of Turkey) in adopting and implementing the reforms required for EU membership, including taxation, as pre-accession assistance not only in terms of implementing the acquis but also in terms of the socio-economic development and cohesion in all countries of the Western Balkans and the rest of the candidate countries. Allow me to conclude with the following words: States exercise sovereignty through membership of the EU. States give up a degree of power to act unilaterally so that they may participate in the development of a collective problem-solving capacity that is a great deal more effective. Resources of power are not finite: acting through the EU expands the sum of State powers so it becomes greater than its parts. The EU ‘adds value’ to its Member States . . . The EU aims to supplement the claims of its Member States to be effective democratic and legitimate actors in an interdependent world, but it does not and should not pretend to suppress the diversity that is Europe’s most cherished richness. The EU aims to accommodate that diversity within a managed framework. It seeks not to replace States, but rather to achieve a better management of their interdependence.40

39 Regulation (EU) 2021/1529 of the European Parliament and of the Council of 15 September 2021 establishing the Instrument for Pre-Accession assistance (IPA III) [2021] OJ L330/1. 40 See: S. Weatherill, Law and Values in the European Union (2016), vi.

Chapter 38

Eu ropean A nt i -Tax - Avoidance Re g i me s Paloma Schwarz Martínez

38.1 Introduction The global initiative against tax avoidance and profit shifting has entailed profound changes in EU direct tax law. Whereas in the past, the adoption of anti-avoidance rules was ultimately left to the discretion of the member states, one can observe that the OECD Base Erosion and Profit Shifting (BEPS) Action Plan1 has ushered in a new era of direct tax policy that has urged member states to reform their tax legislation at an unprecedented speed. This chapter intends to dissemble the EU toolbox against tax avoidance and to address some challenges that follow from the implementation of the Anti-Tax Avoidance Directive2—the most significant EU legislative measure to counter abusive tax practices.

38.2 Towards an EU Toolbox against Tax Avoidance 38.2.1 Anti-Avoidance Rules in EU Tax Directives In the past, anti-tax-avoidance rules grounded on EU secondary law were a rather scarce phenomenon and would traditionally take one of the following forms: (1) rules 1

OECD, Action Plan on Base Erosion and Profit Shifting (2013). Directive (EU) 2016/ 1164 of 12 July 2016 laying down rules against tax avoidance practices that directly affect the functioning of the internal market [2016] OJ L193/1 (ATAD I) and Council Directive (EU) 2017/952 of 29 May 2017 amending Directive (EU) 2016/1164 as regards hybrid mismatches with third countries [2017] OJ L144/1 (ATAD II, together with ATAD I, the ATAD). 2 Council

680 Paloma Schwarz Martínez referring to domestic or agreement-based anti-avoidance provisions;3 and (2) rules denying the tax benefit of a directive, when the principal or one of the principal motives of a transaction is tax evasion or tax avoidance.4 According to the jurisprudence of the Court of Justice of the European Union (CJEU), both types of rules are subject to the pre-eminence of the fundamental freedoms enshrined in the Treaty on the Functioning of the European Union (TFEU)5 and are not considered to be proportionate and thus not permissible if they establish a general presumption of fraud or abuse.6 Although it was originally intended by the European legislator to put the adoption of these anti- avoidance rules at the sole discretion of the member states, the CJEU has meanwhile transformed that option into a mandate to do so. In the so-called Danish Beneficial Ownership cases,7 the CJEU picked up its recurrent statement that EU law cannot be relied on for abusive or fraudulent ends8 and declared it a general principle of European law that creates an obligation for member states to fight abusive practices. As a consequence, neither the elective character of an anti-abuse provision under a directive nor the evident lack of specific or general anti-abuse clauses under national law can affect the national authorities’ obligation to refuse to grant entitlement to rights provided for by a directive where they are invoked for fraudulent or abusive ends.9

3 Art. 1(2) Council Directive 90/435/EEC of 23 July 1990 on the common system of taxation applicable in the case of parent companies and subsidiaries of different Member States [1990] OJ L225/6 (repealed) and its mirror rule, art. 1(2) Council Directive 2011/96/EU of 30 November 2011 on the common system of taxation applicable in the case of parent companies and subsidiaries of different Member States [2011] OJ L345/8 (recast Parent–Subsidiary Directive). An identical provision had also been included in art. 5(1) Council Directive 2003/49/EC of 3 June 2003 on a common system of taxation applicable to interest and royalty payments made between associated companies of different Member States [2003] OJ L157/49 (Interest and Royalties Directive). 4 Art. 5(2) Interest and Royalties Directive and art. 11(1)(a) of Council Directive 90/434/EEC of 23 July 1990 on the common system of taxation applicable to mergers, divisions, transfers of assets and exchanges of shares concerning companies of different Member States [1990] OJ L225/1 (repealed) and its mirror rule, art. 15(1)(a) of Council Directive 2009/133/EC of 19 October 2009 on the common system of taxation applicable to mergers, divisions, partial divisions, transfers of assets and exchanges of shares concerning companies of different Member States and to the transfer of the registered office of an SE or SCE between Member States [2009] OJ L310/34. 5 Case C-6/16 Eqiom and Enka, ECLI:EU:C:2017:641, paras 19ff; Case C- 14/16 Euro Park Service, ECLI:EU:C:2017:177, paras 58ff. 6 Case C-504/16 Deister Holding, ECLI:EU:C:2017:1009, paras 19ff; Case C-14/16 Euro Park Service, ECLI:EU:C:2017:177, paras 61 and 69–70. 7 Joined Cases C-115/16 N Luxembourg 1, C-118/16 X Denmark, C-119/16 C Denmark I and C-299/ 16 Z Denmark, ECLI:EU:C:2019:134; Joined Cases C-116/16 T Denmark and C-117/16 Y Denmark, ECLI:EU:C:2019:135. 8 Case C-212/97 Centros, ECLI:EU:C:1999:126, para. 24; Case C- 196/04 Cadbury Schweppes and Cadbury Schweppes Overseas, ECLI:EU:C:2006:544, para. 35. 9 For a critical analysis of the Danish Beneficial Ownership cases, cf. W. Schön, ‘The Concept of Abuse of Law in European Taxation: A Methodological and Constitutional Perspective’, Working Paper of the Max Planck Institute for Tax Law and Public Finance No. 2019-18; L. De Broe and S. Gommers, ‘Danish Dynamite: The 26 February 2019 CJEU Judgments in the Danish Beneficial Ownership Cases’, EC Tax Review 28/6 (2019), 270–299; F. Vanistendael, ‘Tax Abuse in Europe: The CJEU’s N Luxembourg and 1 and T Danmark Judgments’, Tax Notes International 97/6 (2020), 629–634; J. Englisch, ‘The Danish Tax

European Anti-Tax-Avoidance Regimes 681 Following the initiation of the OECD BEPS Project, one can observe the beginning of a new era of tax avoidance rules in EU directives. The first signal was the introduction of an automatic special anti-avoidance rule (SAAR) targeting hybrid financial instruments via an amendment to art. 4(1)(a) of the recast Parent–Subsidiary Directive. As a consequence, member states shall only refrain from taxing foreign-source dividends ‘to the extent that such profits are not deductible by the subsidiary, and tax such profits to the extent that such profits are deductible by the subsidiary’.10 This rule applies regardless of the valid commercial objectives of the arrangement. A further amendment of the recast Parent–Subsidiary Directive followed only a few months later and entailed the introduction of a mandatory general anti-avoidance rule (GAAR) in art. 1(2) and (3) that since then prevents member states from granting the benefits of the directive to non- genuine arrangements that have been put in place to obtain a tax advantage without reflecting any economic activity.11 Both art. 4(1)(a) as well as art. 1(2) and (3) of the recast Parent–Subsidiary Directive contain an obligation to codify rules against abusive tax practices, which did not exist in the past. The most significant EU legislative measure to counter tax avoidance followed with the adoption of the first Anti-Tax Avoidance Directive (ATAD I)—the first legislative act in direct tax law that directly imposes a burden on corporate taxpayers. ATAD I was adopted as part of the Anti-Tax Avoidance Package12 that was launched in January 2016 by the European Commission (EC) in response of the OECD’s BEPS Action Plan. It intends to address tax avoidance practices in a coherent and coordinated fashion. ATAD I contains five legally binding anti-abuse measures leading to a partial harmonization of anti-avoidance rules in the member states: (1) a controlled foreign company (CFC) rule to deter profit shifting to a low/non-tax country (arts 7 and 8 ATAD); (2) an exit taxation rule to prevent companies from avoiding tax when relocating assets (art. 5 ATAD); (3) an interest-limitation rule to discourage artificial debt arrangements designed to minimize taxes (art. 4 ATAD); and (4) a GAAR to counteract aggressive tax planning when other rules do not apply (art. 6 ATAD). These rules were supplemented a few months later by new rules on hybrid mismatches (arts 9, 9a, and 9b ATAD) implemented through ATAD II.13 The provisions of the ATAD have also been used as basis for the anti-avoidance

Avoidance Cases: New Milestones in the Court’s Anti-Abuse Doctrine’, Common Market Law Review 57/ 2 (2020), 503–537 10 Council

Directive 2014/86/EU of 8 July 2014 amending Directive 2011/96/EU on the common system of taxation applicable in the case of parent companies and subsidiaries of different Member States [2014] OJ L219/40. 11 Council Directive (EU) 2015/121 of 27 January 2015 amending Directive 2011/96/EU on the common system of taxation applicable in the case of parent companies and subsidiaries of different Member States [2015] OJ L21/1. 12 Information on the Anti-Tax Avoidance Package is available at https://ec.europa.eu/taxation_cust oms/business/company-tax/anti-tax-avoidance-package_en (accessed 31 August 2022). 13 Council Directive (EU) 2017/952 of 29 May 2017 amending Directive (EU) 2016/1164 as regards hybrid mismatches with third countries [2017] OJ L144/1 (ATAD II, together with ATAD I, the ATAD).

682 Paloma Schwarz Martínez rules in the Proposal for a Common Corporate Tax Base.14 On 22 December 2021, the European Commission published a proposal for a Council directive laying down rules to circumvent the misuse of so-called ‘shell entities’ (referred to as ‘ATAD III’ as well as an ‘Unshell Directive’)15 by introducing new reporting obligations that may result in the denial of tax advantages such as the denial of tax treaty benefits, the removal of access to EU directives, and the reallocation of taxing rights to EU entities that are deemed to have no or minimal substance. To ensure a successful implementation of the measures described, the EU also had to introduce mechanisms of disclosure requirements and exchange of information among member states to ensure an efficient and well-functioning tax transparency framework. These mechanisms were implemented through different directives for administrative cooperation (DACs).16

38.2.2 Soft Law Instruments In addition to the anti-avoidance rules adopted through secondary law, EU law also provides for a series of soft law instruments which aim at tackling tax avoidance practices in those areas that are not covered by hard law. These include, in particular, the Code of Conduct for Business Taxation (hereinafter ‘CoC’) and the Recommendation on Tax Treaties. The CoC was adopted by the Council on 1 December 1997 as a non- legally binding instrument through which member states made a political commitment

14

Proposal for a Council Directive on a Common Corporate Tax Base, COM(2016) 685 final. For a comparison of the provisions of the ATAD and the CCTB, cf. B. van Raaij, ‘Where Do We Go from Here? The Steady Move towards a Common Corporate Tax Base’, in W. Haslehner et al., eds, A Guide to the Anti-Tax Avoidance Directive (Cheltenham: Edward Elgar, 2020), 304–323. 15 Proposal for a Council Directive laying down rules to prevent the misuse of shell entities for tax purposes and amending Directive 2011/16/EU, COM(2021) 565 final. 16 Council Directive 2011/ 16/EU of 15 February 2011 on administrative cooperation in the field of taxation and repealing Directive 77/799/EEC [2011] OJ L64/1 (DAC 1); Council Directive 2014/107/ EU of 9 December 2014 amending Directive 2011/16/EU as regards mandatory automatic exchange of information in the field of taxation [2016]OJ L359/1 (DAC 2); Council Directive (EU) 2015/2376 of 8 December 2015 amending Directive 2011/16/EU as regards mandatory automatic exchange of information in the field of taxation [2015] OJ L332/1 (DAC 3); Council Directive (EU) 2016/881 of 25 May 2016 amending Directive 2011/16/EU as regards mandatory automatic exchange of information in the field of taxation [2016] OJ L146/8 (DAC 4); Council Directive (EU) 2016/2258 of 6 December 2016 amending Directive 2011/ 16/ EU as regards access to anti- money- laundering information by tax authorities [2016] OJ L342/1 (DAC 5); Council Directive (EU) 2018/822 of 25 May 2018 amending Directive 2011/16/EU as regards mandatory automatic exchange of information in the field of taxation in relation to reportable cross-border arrangements [2018] OJ L139/1 (DAC 6); Council Directive (EU) 2021/514 of 22 March 2021 amending Directive 2011/16/EU on administrative cooperation in the field of taxation [2021] OJ L104/1 (DAC 7). In December 2022, the European Commission also published a proposal to extend the exchange of information requirements to crypto-assets and e-money (DAC 8), cf. Proposal for a Council Directive amending Directive 2011/16/EU on administrative cooperation in the field of taxation, COM(2022) 707 final.

European Anti-Tax-Avoidance Regimes 683 to re-examine, amend, or abolish harmful tax measures from their tax systems (rollback) and to refrain from introducing such measures (stand-still).17 On 9 March 1998, the Economic and Financial Affairs Council set up a Code of Conduct Group for business taxation (hereinafter ‘the Group’)18 to assess tax measures that may fall within the scope of the CoC.19 In recent years, the Group has elaborated overall policies on certain issues that require a coordinated implementation of certain measures among member states, for example regarding the exchange of tax rulings and the treatment of hybrid mismatches.20 Moreover, the Group is also committed to promoting the adoption of the CoC’s principles and criteria beyond European borders. Finally, the Group has also been charged with developing a coordinated and coherent EU policy towards non- cooperative third-country jurisdictions. In this context, the Group has been mandated to establish an EU list of non-cooperative jurisdictions (also known as the EU tax haven blacklist) for tax purposes to tackle tax avoidance, tax evasion, and money laundering by listing third countries that encourage abusive tax practices, which erode member states’ corporate tax revenues.21 The elaboration of this list goes back to the European Commission’s Communication on an External Strategy for Effective Taxation22 and was adopted by the EU Council on 5 December 2017.23 In this context, the Group has also been charged with continuing the dialogue with the blacklisted jurisdictions and monitoring the situation in those jurisdictions and in others that have been screened in the process. Currently, the reform and modernization of the CoC is being discussed in order to improve its effectiveness in the fight against contemporary forms of harmful tax competition.24

17 Conclusions

of the ECOFIN Council Meeting on 1 December 1997 concerning taxation policy— Resolution of the Council and the Representatives of the Governments of the Member States, meeting within the Council of 1 December 1997 on a code of conduct for business taxation—Taxation of saving, [1998] OJ C 2/1, 6.1. 18 For a study on the functioning and work and role of the Code of Conduct Group, cf. M. Nouwen, Inside the EU Code of Conduct Group: 20 Years of Tackling Harmful Tax Competition (Amsterdam: 2020). 19 Council conclusions of 9 March 1998 concerning the establishment of the Code of Conduct Group (business taxation) [1998] OJ C 99/1. 20 Council of the European Union, Note of 5 December 2019: Agreed guidance by the Code of Conduct Group (business taxation), 5814/6/18 REV 6, FISC 44 ECOFIN 75. 21 Council Conclusions of 8 November 2016: Criteria and process leading to the establishment of the EU list of non-cooperative jurisdictions for tax purpose, 14166/16, FISC 187 ECOFIN 1014. On the EU tax haven black list, cf. V. Kalloe, ‘EU Tax Haven Blacklist—Is the European Union Policing the Whole World?’, European Taxation 2/3, 58 (2018), 47–55; G. Melis and A. Persiani, ‘The EU Blacklist: A Step Forward but Still Much to Do’, EC Tax Review 28/5 (2019), 253–263; I. Mosquera Valderrama, ‘The EU Standard of Good Governance in Tax Matters for Third (Non-EU) Countries’, Intertax 47/5 (2019), 454–467. 22 Communication from the Commission to the European Parliament and the Council on an External Strategy for Effective Taxation, COM(2016) 24 final. 23 Council Conclusions 15429/17 of 5 December 2017. Since then, the list has been updated several times. From 2020, the list has been updated twice a year. 24 Communication of 15 July 2020 from the Commission to the European Parliament and the Council on Tax Good Governance in the EU and beyond, COM(2020) 313 final, 3.

684 Paloma Schwarz Martínez As part of the EU Anti-Tax Avoidance Package, the EU also issued the Recommendations on the implementation of measures against tax treaty abuse (hereinafter ‘the Recommendation’),25 which supplements the hard law instruments (i.e. the ATAD and DACs) and addresses the implementation by the member states of measures against tax treaty abuse. The Recommendation urges member states to implement the OECD BEPS proposal to address tax treaty abuse. Where member states include in tax treaties a principal purpose test-based GAAR as suggested in the OECD’s final report on BEPS Action 6,26 the European Commission recommends wording that complies with EU case law to ensure that genuine activity is not affected. The Recommendation also encourages member states to implement and make use of the proposed amendments to art. 5 of the OECD Model Tax Convention as set out in the final report on Action 7 of the BEPS Action Plan27 to address artificial avoidance of permanent establishment (PE) status. Member states should inform the European Commission of the measures taken in order to comply with the Recommendation, as well as of any changes made to such measures.

38.2.3 State Aid as a Tool to Tackle Tax Avoidance? The EU prohibition of state aid is laid down in art. 107(1) TFEU. To qualify as prohibited state aid, a tax measure needs to fulfil four cumulative conditions. First, an advantage must be present. Secondly, the advantage must be imputable to the corresponding state. Thirdly, the aid must negatively affect trade between member states and (threaten to) distort competition. Finally, the measure must favour certain undertakings or the production of certain goods. The fourth condition—described as ‘the selectivity test’—has always been the crucial criterion for determining fiscal state aid and distinguishing general tax measures pursuing legitimate economic or social policy goals from selective tax advantages benefiting only certain types of taxpayers. Over the past few years, one can observe that the European Commission has resorted numerous times to the state aid rules as a ‘backdoor rule’ to combat tax avoidance, particularly in its state aid investigations against individual tax rulings.28 When assessing 25 Commission Recommendation (EU) 2016/ 136 of 28 January 2016 on the implementation of measures against tax treaty abuse, notified under document C(2016) 271), [2016] OJ L25/67. 26 OECD, Preventing the Granting of Treaty Benefits in Inappropriate Circumstances, Action 6–2015 Final Report (Paris: OECD Publishing, 2015). 27 OECD, Preventing the Artificial Avoidance of Permanent Establishment Status, Action 7—2015 Final Report (Paris: OECD Publishing, 2015). 28 e.g. cf. Case T-760/15 Netherlands v Commission, ECLI:EU:T:2019:669; Joined Cases T-7 78/16 and T-892/16 Ireland/Commission, ECLI:EU:T:2020:338; Commission Decision (EU) 2016/1699 of 11 January 2016 on the excess profit exemption State aid scheme S.A.37667 (2015/C) (ex 2015/NN) implemented by Belgium (notified under document C(2015) 9837), OJ L260/61, 27.09.2016; Commission Decision (EU) 2018/859 of 4 October 2017 on State aid SA.38944 (2014/C) (ex 2014/NN) implemented by Luxembourg to Amazon (notified under document C(2017) 6740) [2018] OJ L153/1. The updated developments can be found at https://ec.europa.eu/competition/state_aid/tax_rulings/index_en.html (accessed 31 August 2022).

European Anti-Tax-Avoidance Regimes 685 the selective character of tax rulings, the European Commission has regularly relied on a deviation from the ‘EU arm’s-length principle’, which ‘necessarily forms part of the European Commission’s assessment under Article 107(1) of the Treaty of tax measures granted to group companies, independently of whether a Member State has incorporated this principle into its national legal system and in what form’.29 This statement also reflects the position taken by the European Commission in the Notice on the notion of State aid30 and the DG Competition Working Paper on State Aid and Tax Rulings.31 In doing so, the European Commission is overstretching and misapplying art. 107 TFEU since it ignores that it is the national tax law of the member state concerned which must serve as the benchmark for assessing the selectivity of the relevant tax measure. Although it appears tempting to rely on a uniform EU reference framework, using such a hypothetical and non-existent reference framework is impermissible as this application would clearly conflicts with the member states’ prerogative in direct tax matters to decide which events should be taxed and how to set the tax base and the tax rate. Consequently, the decisive reference framework for the analysis of fiscal measures can only be the tax legislation of the relevant country itself.32

38.2.4 Tax Avoidance: A General Principle of European Law An obligation for member states to fight tax avoidance may also result from the general principle of EU law to counter abusive tax practices. As described in Section 38.2.1, the CJEU established in its landmark Danish Beneficial Ownership cases a general principle of EU law to counter such behaviour. However, this principle only applies to the abuse of subjective rights derived from EU law. This means that the principle should not create any new, unwritten norms obliging member states to combat the abuse of other subjective rights derived outside EU law (i.e. rights derived from domestic legislation or tax treaties). Also, it should not be interpreted as establishing an obligation on

29 e.g.

Commission decision of 11 January 2016 on the excess profit state aid regime scheme implemented by Belgium, SA.37667 (2015/C) (ex 2015/NN), C(2015) 9837 final, para. 150. 30 Commission Notice on the notion of State aid as referred to in Article 107(1) of the Treaty on the Functioning of the European Union (2016/C 262/01), C 262/1, para. 172. 31 DG Competition Working Paper on State Aid and Tax Rulings, DG Competition— Internal Working Paper—Background to the High Level Forum on State Aid of 3 June 2016. 32 W. Schön, ‘Taxation and State Aid Law in the European Union’, Common Market Law Review 36/5 (1999), 911–936, 923; R. Szudoczky, The Sources of EU Law and Their Relationships: Lessons for the Field of Taxation: Primary Law, Secondary Law, Fundamental Freedoms and State Aid Rules (Amsterdam: 2014), 494; P. Schwarz, IP Box Regime im Europäischen Steuerrecht (Baden-Baden: Nomos, 2017), 75ff L. Parada, ‘Between Apples and Oranges: the EU General Court’s Decision in the “Apple Case” ’, EC Tax Review 30/ 2 (2021), 55–63, 62; AG Opinion in Joined Cases C-106/09 P and C-107/09 P Commission and Spain v Government of Gibraltar and United Kingdom, ECLI:EU:C:2011:215, para. 202.

686 Paloma Schwarz Martínez member states’ tax authorities to directly rely on the provisions of the ATAD to prevent abusive arrangements that are covered by the directive.33 In fact, the Danish Beneficial Ownership cases do not overturn the previous judgment in 3M Italia.34 In that case, the CJEU had ruled that the general (EU) principle of abuse of rights does not require national authorities and courts to combat abusive tax practices where a taxpayer does not rely on EU law. However, the current jurisprudence of the CJEU could be interpreted as also imposing a general obligation on member states to refuse a taxpayer’s right to benefit from the fundamental freedoms and to tax that taxpayer as if they had not made use of their freedoms in case they are abused for tax purposes. Should the CJEU favour this interpretation, it would also be irrelevant whether the member state in question had adopted an appropriate anti-abuse rules or not.35

38.3 The Anti-Tax Avoidance Directive 38.3.1 Preliminary Remarks From the beginning, the EU has played an active role in the global fight against tax avoidance. On 17 June 2015, the European Commission published its ‘Action Plan for fair and efficient corporate taxation in the EU’36 that linked in with the OECD BEPS Action Plan. Shortly after, in December 2015, the European Council formally endorsed the outcome of the BEPS Action Plan37 and set out the options for the implementation of the OECD BEPS Action Plan within the EU. The conclusions of the European Council considered that EU directives should be the preferred vehicle for implementing anti-BEPS measures at the European level. On this basis, the European Commission elaborated the Anti-Tax Avoidance Package that was issued in January 2016 and that included, among others, the proposal for the ATAD. In June 2016, ATAD I was enacted, establishing member states’ obligations to introduce a wide range of anti-avoidance measures. This piece of legislation was extended one year later by ATAD II to complement the existing rule on hybrid mismatches. On 22 December 2021, the European Commission published a proposal for a further amendment of the ATAD targeting the misuse of shell entities for tax purposes.38

33 W.

Haslehner, ‘The General Scope of the ATAD and Its Position in the EU Legal Order’, in Haslehner et al., A Guide to the Anti-Tax Avoidance Directive, 32–65, 56ff 34 Case C-417/10 3M Italia, ECLI:EU:C:2012:184, paras 30ff. 35 De Broe and Gommers, ‘Danish Dynamite’, 282; A. Zalasiński, ‘The ECJ’s Decisions in the Danish “Beneficial Ownership” Cases: Impact on the Reaction to Tax Avoidance in the European Union’, International Tax Studies 2/4 (2019), 16. 36 Communication from the Commission to the European Parliament and the Council: A Fair and Efficient Corporate Tax System in the European Union: 5 Key Areas for Action, COM(2015) 302 final. 37 Council conclusions on corporate taxation—base erosion and profit-shifting. 38 For an analysis of this proposal, cf. P. Pistone et al., ‘European Union/International—Abuse, Shell Entities and Right of Establishment: A Plea for Refocusing Current Proposals and Achieving Deeper

European Anti-Tax-Avoidance Regimes 687 The necessity for the adoption of the ATAD was based on the need for a consistent and coordinated implementation of the outputs of the OECD BEPS Action Plan and the goal of improving the resilience of the internal market as a whole against cross-border tax avoidance practices by setting a common minimum level of protection against these types of practices.39 The ATAD was based on art. 115 TFEU. As such, it has to comply with the principle of subsidiarity (enshrined explicitly in art. 5(3) TEU) and the principle of proportionality (art. 5(4) TEU). However, it can be questioned whether the directive is in line with these principles.40 Member states were required to transpose the provisions of the ATAD into their domestic law by 31 December 2018, with the new measures to be effective on 1 January 2019 unless a member state was granted a derogation.41

38.3.2 Brief Overview of the Provisions of the ATAD The ATAD lays down the following rules against corporate tax avoidance: (1) a CFC rule (arts 7 and 8 ATAD); (2) an exit taxation rule (art. 5 ATAD); (3) an interest-deduction limitation (art. 4 ATAD); (4) a GAAR (art. 6 ATAD); as well as (5) mismatch rules (arts 9, 9a, and 9b ATAD). The directive’s scope is in principle limited to corporate taxpayers (including PEs).42 However, some provisions also have an effect on individuals or entities that are considered to be transparent for tax purposes.43 Also, although the key objective of the ATAD is the avoidance of cross-border tax avoidance practices,44 the GAAR45 as well as the interest-limitation rule apply in cross-border and domestic situations alike.

38.3.2.1 The CFC rule (arts 7 and 8 ATAD) Articles 7 and 8 ATAD introduce detailed rules about the minimum standard for the implementation of CFC legislation and aim at discouraging multinational companies from

Coordination within the Internal Market’, World Tax Journal 14/2 (2022), 187–236; P.-R. Dukmedjia and C. Bénézet, ‘European Union—The Upcoming ATAD 3 Substance Rules: Overview and First Challenges’, Finance and Capital Markets 23/2 (2022). 39

Recitals 2, 3, and 16 ATAD. A. Graaf and K. J. Visser, ‘ATA Directive: Some Observations Regarding Formal Aspects’ EC Tax Review 25/4 (2016), 199–210; C. Brokelind, ‘The Anti-Tax Avoidance Directive under Scrutiny: A Matter of Competence?’, in J. Monsenego and J. Bjuvberg, eds, International Taxation in a Changing Landscape. Liber Amicorum in Honour of Bertil Wiman (The Hague: 2019), 45–56; I. Lazarov and S. Govind, ‘Carpet- Bombing Tax Avoidance in Europe: Examining the Validity of the ATAD under EU Law’, Intertax 47/10 (2019), 852–868. 41 Art. 11 ATAD I and art. 2 ATAD II. 42 Art. 1 ATAD I. 43 By way of example, art. 9a also applies to transparent entities. Moreover, most member states apply the GAAR to non-corporate taxpayers. 44 Recital 16 ATAD I. 45 Recital 11 ATAD I. 40

688 Paloma Schwarz Martínez shifting profits from their parent company located in a high tax country to low-taxed controlled subsidiaries located abroad to reduce the group’s tax liability.46 The CFC rule allows the inclusion of undistributed income of a foreign company or PE in the tax base of the shareholder (in proportion to its participation) provided that: (1) the shareholder exercises a certain control over the CFC;47 and (2) the CFC income is subject to low taxation.48 Member states have two options for imposing the CFC charge: they can decide to attribute either certain predefined categories of undistributed passive income of the CFC to the shareholder49 or the income of the CFC that arises from non-genuine arrangements which have been put in place for the essential purpose of obtaining a tax advantage.50 Art. 8 ATAD foresees different measures against double taxation.

38.3.2.2 The exit tax rule (art. 5 ATAD) Article 5 ATAD contains a minimum framework for a mandatory exit tax on certain cross-border transactions to prevent corporate taxpayers from moving their tax residence and/or assets from a member state to another jurisdiction, without paying taxes on unrealized gains accrued while the state of departure had taxing powers.51 The exit tax will be levied on the difference between the market value of the transferred assets (based on the arm’s-length principle) and the value of the transferred assets for tax purposes.52 Taxpayers will be given the right to either immediately pay the exit tax or defer the payment by paying it in five annual instalments where the transfer is made to another member state or a country covered by the European Economic Area.53 Where a taxpayer opts for a deferred payment, member states may decide to charge interest on the outstanding amount of exit tax or require a bank guarantee if there is a demonstrable and actual risk of non-recovery.54 Finally, the ATAD stipulates a mandatory step-up in

46 Recital 12 ATAD I. For an analysis of the CFC rule, cf. J. Schönfeld, ‘CFC Rules and Anti- Tax Avoidance Directive’, EC Tax Review 26/ 3 (2017), 145– 152; D. Blum, ‘Controlled Foreign Companies: Selected Policy Issues—or the Missing Elements of BEPS Action 3 and the Anti-Tax Avoidance Directive’, Intertax 46/4 (2018), 296–312; I. Groot and B. Larking, ‘Implementation of Controlled Foreign Company Rules under the EU Anti-Tax Avoidance Directive (2016/1164)’, European Taxation 59/6 (2019), 261–272; A. Rust, ‘Controlled Foreign Company Rule (Articles 7 and 8 ATAD)’, in Haslehner et al., A Guide to the Anti-Tax Avoidance Directive, 174–199. 47 Art. 7(1)(a) ATAD. 48 Art. 7(1)(b) ATAD. 49 Art. 7(2)(a) ATAD. 50 Art. 7(2)(b) ATAD. 51 Recital 10 ATAD. For an analysis of the exit tax rule, cf. S. Peeters, ‘Exit Taxation: from an Internal Market Barrier to a Tax Avoidance Prevention Tool’, EC Tax Review (2017), 122–132; H. Vermeule, ‘Entrepreneurial, Corporate and Asset Emigration in Exit Taxation in the ATAD’, in P. Pistone and D. Weber, eds, The Implementation of Anti-BEPS Rules in the EU: a Comprehensive Study (Amsterdam: 2018); P. Schwarz, ‘The Exit Tax Rule (Article 5 ATAD)’, in Haslehner et al., 105–126. 52 Art. 5(1) ATAD. 53 Art. 5(2) ATAD. 54 Art. 5(3) ATAD.

European Anti-Tax-Avoidance Regimes 689 the case of a transfer to another member state to avoid double taxation of the unrealized capital gains on the assets that are transferred cross-border.55

38.3.2.3 The interest limitation (art. 4 ATAD) Article 4 ATAD establishes an interest-limitation rule which provides that exceeding borrowing costs will be allowed as a deduction only up to 30% of a taxpayer’s earnings before interest, tax depreciation, and amortization (EBITDA) in the tax period in which they are incurred.56 This provision targets tax practices aimed at reducing the global tax liability of groups of companies.57 Member states are free to introduce a higher level of protection (e.g. by decreasing the ratio, placing time limits, or restricting the costs of uncompensated excessive borrowing that can be carried forward or back). Member states are also permitted to supplement the interest-limitation rule by introducing special rules against intra-group debt financing, such as thin capitalization rules. In implementing the interest-limitation rule, the ATAD provides member states with various options (e.g. a de minimis rule,58 the possibility to apply the rule at the level of the group,59 derogations in favour of stand-alone entities60 and financial undertakings,61 a grandfathering rule,62 and implementation of safe harbour clauses for members of a consolidated group).63

38.3.2.4 The GAAR (art. 6 ATAD) Article 6 ATAD contains a GAAR that tackles abusive tax practices that are not covered by other more specific anti-tax avoidance rules and covers both domestic and cross- border situations in a uniform manner.64 The GAAR comes into play where there is a (series of) arrangement(s) that has (have) been put in place for the main purpose or one of the main purposes of obtaining a tax advantage that defeats the object or purpose of the applicable tax law provided that it is (they are) not genuine having regard to all 55

Art. 5(5) ATAD. For an analysis of the interest limitation rule, cf. M. Tell, ‘Interest Limitation Rules in the Post-BEPS Era’, Intertax 11 (2017), 750ff; A. P. Dourado, ‘The Interest Limitation Rule in the Anti-Tax Avoidance Directive (ATAD) and the Net Taxation Principle’, EC Tax Review 3 (2017), 112ff; D. Gutman, ‘The Interest Limitation (Article 4 ATAD)’, in Haslehner et al.,86–104. 57 Recital 6 ATAD. 58 Art. 4(3)(b) ATAD. 59 Art. 4(1) ATAD. 60 Art. 4(3)(b) ATAD. 61 Art. 4(7) ATAD. 62 Art. 4(4) ATAD. 63 Art. 4(5) ATAD. 64 Recital 11 ATAD. For an analysis of the GAAR, cf. L. De Broe and D. Beckers, ‘The General Anti- Abuse Rule of the Anti-Tax Avoidance Directive: an Analysis against the Wider Perspective of the European Court of Justice’s Case Law on Abuse of EU Law’, EC Tax Review 26/3 (2017), 133–144; A. Baez Moreno and J. Zornoza Pérez, ‘The General Anti-Abuse Rule of the Anti-Tax Avoidance Directive’, in J. Manuel Almudí Cid et al., Combatting Tax Avoidance in the EU: Harmonization and Cooperation in Direct Taxation, Eucotax Series on European Taxation (2019), 101–112; B. Kuźniacki, ‘The GAAR (Article 6 ATAD)’. in Haslehner et al., A Guide to the Anti-Tax Avoidance Directive, 127–173. 56

690 Paloma Schwarz Martínez relevant facts and circumstances.65 If all conditions are met, the member state concerned will have to disregard the arrangement(s) and adjust the tax liability in accordance with its national tax law.

38.3.2.5 The mismatch rules (arts 9, 9a, and 9b ATAD) To counter mismatch arrangements in situations where there are differences in the legal characterization of payments/entities or allocation of payments that result in the interaction between the legal systems of two jurisdictions and lead to a double deduction or deduction without inclusion, the ATAD introduces detailed rules to neutralize hybrid mismatches (art. 9), reverse hybrid mismatches (art. 9a), and tax residency mismatches (art. 9b).66 The scope of the provisions is limited to mismatches between a taxpayer and its associated enterprises. In a nutshell, hybrid mismatch arrangements are neutralized through the disallowance of a deduction, the inclusion of income, or a limitation of tax relief at source. Most anti-hybrid provisions consist of a primary response and a secondary response (so-called linking rules).67 The primary response disallows a tax deduction for a payment expense and the secondary response requires either a deduction to be denied or taxation to be applied on the hybrid payment. The secondary response comes into play only as a defensive rule where the hybrid mismatch is not resolved through the application of the primary rule (e.g. where a third country does not allow a deduction).68 Member states have the option of excluding certain types of mismatches from the application of the secondary rule, such as payments to a hybrid entity resulting in allocation differences or payments to a disregarded PE.69

38.3.3 Implementation challenges The ATAD has the form of a minimum standard directive. As such, it allows the member states a certain degree of discretion. This, however, also implies that the responsibility

65

Art. 6(1) ATAD. an analysis of the anti-hybrid mismatch rules, see G. K. Fibbe and T. Stevens, ‘Hybrid Mismatches under the ATAD I and II’, EC Tax Review 26/3 (2017), 153–166; L. Parada, ‘Hybrid financial instruments and anti-hybrid rules in the EU ATAD (Article 9 ATAD)’, in Haslehner et al., A Guide to the Anti-Tax Avoidance Directive, 200–28; C. Spies, ‘Hybrid Entities and Anti-Hybrid Rules in the ATAD (Article 9 and 9a)’, in ibid., 229–255; B. Peeters and L. Vanneste, ‘The Hybrid Financial Instruments: The Effects of the OECD BEPS Action 2 Report and the ATAD’, Intertax 48/1 (2020), 14–45 Peeters and Vanneste, ‘The Hybrid Financial Instruments’, 14ff. 67 The only hybrid mismatch provision for hybrid entities in the ATAD that does not rely on the concept of linking rules but forces the jurisdiction where the hybrid entity is established or incorporated to follow the tax qualification of the jurisdiction of the controlling investor (so-called coordination rule) is art. 9a ATAD. 68 ATAD I only included the primary rule and was limited to hybrid mismatches within the EU. It was only through the adoption of ATAD II that anti-hybrid mismatch rules were extended to third country situations. 69 Art. 9(4) ATAD 2. 66 For

European Anti-Tax-Avoidance Regimes 691 for a consistent and coherent implementation of the ATAD provisions is ultimately put into the hands of the member states.

38.3.3.1 Assessing the need for legislative amendments The need for legislative amendments and the impact on domestic tax law very much depend on the individual member state. Whereas some countries’ tax legislation (e.g. Germany) already provided a comprehensive set of anti-tax avoidance provisions, other member states such as Luxembourg70 or Poland71 had to draft rules that were completely new to their domestic tax systems.72 The first category of states had to evaluate if and to what extent there was actually a need for (immediate) legislative changes. Indeed, it follows from the jurisprudence of the CJEU that legislative action is not needed to transpose a directive if the existing rules are sufficiently precise and clear to allow the persons concerned to know the full extent of their rights and obligations.73 This also means that the interpretation of the existing national law has to lead to the same result as that required by the ATAD.74 Legislative changes are, however, necessary if these conditions are not met. By way of example, although the Italian tax legislation already provided for an interest-limitation rule similar to the one contained in art. 4 ATAD, the Italian rules followed a definition of EBITDA that was distinct from the one of the ATAD. Whereas the ATAD explicitly excludes tax-exempt income from the EBITDA of a taxpayer, the Italian rules included qualifying foreign dividends, generally subject to the participation regime, in the calculation of the EBITDA.75 Another example are the German and Belgian exit tax rules that, prior to the adoption of the ATAD, did not include a step-up in value for assets transferred as provided for in art. 5(5) ATAD.76 The ATAD itself provides for the possibility of maintaining national targeted rules until 1 January 2024 for preventing base erosion and profit shifting that are equally effective to the interest-limitation rules contained in the ATAD.77 On 7 December 2018, the European Commission published a notice78 specifying the national provisions that it considered ‘equally effective’ to art. 4 ATAD. In addition, the notice included brief considerations of the criteria used for such assessment. Ireland and Austria, whose interest limitation provisions did not meet the criteria set out in the European Commission notice, were requested to change their provisions within a period of two

70

Prior to the implementation of the ATAD, Luxembourg did not have any CFC regulations. Before the adoption of the ATAD, Poland did not have a concept of exit taxation. 72 For a comparative analysis of the impact of the ATAD on member states’ tax legislation, cf. D. Gutman et al., ‘The Impact of the ATAD on Domestic Systems: A Comparative Survey’, European Taxation 57/1 (2017), 2–20. 73 Case C-321/05 Kofoed, ECLI:EU:C:2007:408, para. 44. 74 Haslehner, ‘The General Scope of the ATAD and Its Position in the EU Legal Order’, 52. 75 Gutman, ‘The Interest Limitation (Article 4 ATAD)’, 98. 76 Gutman et al., ‘The Impact of the ATAD on Domestic Systems: A Comparative Survey’, 7ff. 77 Art. 11(6) ATAD. 78 Commission Notice of 7 December 2018: Measures considered equally effective to art. 4 of the Anti-Tax Avoidance Directive (2018/C 441/01). 71

692 Paloma Schwarz Martínez months. Other member states included in the European Commission’s list, however, decided not to postpone the implementation of the interest-limitation rule.79 Apart from the actual legislative changes that were required in member states’ tax legislation, the adoption of the ATAD may also have an impact on how national tax courts will interpret existing anti-avoidance provisions in future. For instance, even if the wording of the existing German GAAR is not expected to be modified following the adoption of the ATAD,80 a different interpretation (at least with regard to corporate taxpayers) of the existing German GAAR may nevertheless be required in future to comply with art. 6 ATAD.81

38.3.3.2 Member states’ discretionary powers Article 3 ATAD explicitly states that ‘the directive shall not preclude the application of domestic or agreement-based provisions aimed at safeguarding a higher level of protection for domestic corporate tax bases’. This must be understood as allowing the member states to implement rules that lead to a higher corporate tax burden than required by the provisions of the ATAD.82 However, it is debatable how much room for manoeuvre is actually left to the member states since the adoption of stricter anti-abuse provisions has to comply with EU primary law and, in particular, with the fundamental freedoms and state aid law. By way of example, member states are not allowed to require a bank guarantee as a condition for deferring the payment of exit tax if there is no demonstrable and actual risk of non-recovery. A differing view would certainly conflict with the CJEU’s statement in the DMC case where the Court had explicitly ruled that the requirement of a bank guarantee constitutes a restriction of the fundamental freedoms, which cannot be justified ‘without prior assessment of the risk of non-recovery’.83 The ATAD provides member states with a number of options, such as the option of charging interest where a taxpayer elects to pay the exit tax in instalments84 or the option to exclude stand-alone entities from the scope of the interest-limitation rule85. In exercising these options, member states are in principle subject to the scrutiny of primary law and in particular to the fundamental freedoms and state aid provisions. This is notably the case where one of the options given does not entail discriminatory treatment. To the extent that these options lead to different treatment of certain

79 e.g. France decided to implement art. 4 ATAD as of 1 January 2019 following a public consultation that revealed that a majority of companies preferred not to wait until 1 January 2024, Gutman, ‘The Interest Limitation (Article 4 ATAD)’, 87. 80 The German draft implementation act does not foresee any changes to the GAAR, cf. Regierungsentwurf des ATAD-Umsetzungsgesetzes, BT-Drucks. 19/28652, Stand: 19 April 2021. 81 Gutman et al., ‘The Impact of the ATAD on Domestic Systems: A Comparative Survey’, 11. 82 Haslehner, ‘The General Scope of the ATAD and Its Position in the EU Legal Order’, 36. 83 Case C-164/12 DMC, ECLI:EU:C:2014:20, para. 67. This had already been argued by this author in their contribution to the ATAD proposal, cf. A. Navarro, L. Parada, and P. Schwarz, ‘The Proposal for an EU Anti-Avoidance Directive: Some Preliminary Thoughts’, EC Tax Review 25/3 (2016), 117–131, 121. 84 Art. 5(3) ATAD. 85 Art. 4(3)(b).

European Anti-Tax-Avoidance Regimes 693 undertakings, they may not, however, become subject to state aid scrutiny where the option is sufficiently clearly prescribed by the ATAD provided that the member state concerned remains within the limits of the directive.86 The different options provided for by the ATAD also give the member states the possibility to stand out as a business-friendly location thereby maintaining a certain degree of tax competition between member states’ tax legislations. The Luxembourg legislator has, for instance, endeavoured to preserve Luxembourg’s competitive position by implementing the provisions of the ATAD in the most lenient way. An example is the new Luxembourg CFC rules that have been introduced in a new art. 164ter of the Luxembourg Income Tax Code. As described earlier in Section 38.3.2.1, the ATAD offers two options to implement CFC rules in member states’ tax laws. Luxembourg has chosen option B, which imposes taxation on the undistributed income of a CFC arising from non-genuine arrangements implemented for the purpose of obtaining a tax advantage. The Luxembourgish legislator even decided to extend the substance test to third countries. It follows from the foregoing that the Luxembourg CFC provision only covers mere letterbox companies so that CFC taxation is almost always avoided.87 Another example is the option provided for in art. 4(7) ATAD to exclude financial undertakings from the personal scope of the interest limitations rule, which has also been chosen by Luxembourg.88

38.3.3.3 ATAD and third states The ATAD will impact third countries in various ways. This is illustrated by the following two examples. First of all, there are a number of provisions within the ATAD that may conflict with double tax treaties concluded between member states and third countries. The probably most prominent and most discussed example is the conflict that may arise from the inclusion of CFC income attributed to a foreign PE following art. 7(1) ATAD if the treaty provides for the exemption method with regard to business profits within the meaning of art. 7 OECD Model Tax Convention.89 The solution to a treaty override in third-state scenarios ultimately lies within the constitution of the member state concerned. This also seems to be the opinion of the Luxembourg State Council. In its formal opinion on the ATAD implementation law, it took the position that if a foreign PE is considered a CFC of a Luxembourg entity, the tax treaty should prevail over the CFC regime

86

Haslehner, ‘The General Scope of the ATAD and Its Position in the EU Legal Order’, 51ff. This choice was welcomed by tax practitioners, cf. E. Lebas, ‘Luxembourg/United States/European Union/OECD—Recent Amendments to EU, Luxembourg and US Tax Laws, and Their Implications for US Holding and Financing Branch Structures’, Bulletin for International Taxation 73/3 (2019), 568– 580, 571. 88 Art. 168bis(8) Luxembourg Income Tax Code. 89 D. Blum , ‘Controlled Foreign Companies: Selected Policy Issues— Or the Missing Elements of BEPS Action 3 and the Anti-Tax Avoidance Directive’, Intertax 46/4 (2018), 296–312, 311; Rust, ‘Controlled Foreign Company Rule (Articles 7 and 8 ATAD)’, 182ff. Possibilities for treaty override may also arise where a member state is required to apply the defensive rule of art. 9(2)(b) ATAD to counter a deduction-non-inclusion scenario. 87

694 Paloma Schwarz Martínez provided that the PE is located in a third state given the prevalence of international tax treaties over domestic law.90 From an EU law point view, it seems convincing to resolve this problem by an analogue application of art. 351 TFEU, meaning that tax treaties concluded before the enactment of the ATAD should not be overwritten by national provisions implementing the ATAD. Rather, member states should endeavour to change the respective tax treaty.91 Secondly, the implementation of the ATAD can also serve as an example for negative tax competition by the EU towards third states. Whereas some of the ATAD provisions, such as the interest limitation rule and the GAAR, equally apply in domestic situations, within the EU and vis-à-vis third states, others urge the member states to provide for more favourable treatment of investments made within the EU. The exit tax provision provided for in art. 5 ATAD, for instance, applies in different situations in which assets, residences, or businesses are transferred outside a member state’s tax jurisdiction. Transfers covered by this provision are taxed at an amount equal to the market value of the transferred assets, at the time of exit of the assets (less their value for tax purposes), which will be levied even though the taxpayer has not yet realized any taxable income or gain. In the case of transfers occurring from one member state to another, taxpayers are given the right to defer payment of the exit tax, whereas the directive seems to imply that such deferral possibility should not exist in the case of transfers to a third country.92 In other cases, it may, however, depend on the individual member state whether it wants to ensure equal treatment of third-state scenarios. For instance, it is up to the member state to decide whether it wishes to apply the safe harbour rule of the interest-limitation rule to a company that uses consolidated financial statements established under the rules of a third country (e.g. US GAAP).

38.4 Concluding Remarks In recent years, the fight against abusive tax practices has become one of the predominant topics in EU direct tax policy. Whereas in the past it was primarily the responsibility of the member states to protect their national corporate tax systems against aggressive tax planning, the EU has in several instances recently taken the subject up at the European level. First, following the initiation of the OECD BEPS Project, one can observe the beginning of a new era of mandatory tax avoidance rules in EU secondary law that started with the introduction of a SAAR targeting hybrid financial instruments

90

Avis, Conseil d’État no. 52.949 du 13 novembre concernant le projet de loi no. 7318, 14. Rust, ‘Controlled Foreign Company Rule (Articles 7 and 8 ATAD)’, 182ff. 92 This had already been argued by the author in W. Haslehner and P. Schwarz, ‘Harmful Tax Competition from the European Union towards Third Countries?’, in M. Jiménez, ed., The External Tax Strategy of the EU in a Post-BEPS Environment (Amsterdam: 2019), 241–266. 91

European Anti-Tax-Avoidance Regimes 695 via an amendment of the recast Parent–Subsidiary Directive and that culminated in the adoption of the ATAD. To ensure a successful implementation of these measures, the EU also introduced mechanisms of disclosure requirements and exchange of information among member states through different directives for administrative cooperation. Secondly, EU law also provides for a series of soft law instruments, which aim at tackling tax avoidance practices in those areas that are not covered by secondary law, the most significant being the output of the work undertaken by the Code of Conduct Group for business taxation. Thirdly, over the last years one can observe that the European Commission is overstretching and misapplying the state aid rules by relying on them as a ‘backdoor rule’ to combat tax avoidance, particularly in its state aid investigations against individual tax rulings. Finally, an obligation for member states to fight tax avoidance may also follow from the general principle of EU law to counter tax practices that was established by the CJEU in its landmark Danish Beneficial Ownership cases. The second part of this chapter has highlighted some challenges that may arise from the implementation of the ATAD: the implementation of the ATAD has not only entailed important changes in member states’ tax legislations but will also have an impact on how national tax courts will interpret existing anti-avoidance provisions and concepts in the future. The ATAD allows the member states to implement rules that lead to a higher corporate tax burden than required by the provisions of the ATAD. The adoption of stricter rules, however, has to comply with EU primary law and may not leave much room for manoeuvre. Also, in exercising the options given by the ATAD, member states will in particular be subject to the scrutiny of the fundamental freedoms and state aid provisions. They also provide member states with the possibility to stand- out as a business-friendly location thereby maintaining a certain degree of tax competition between member states’ tax legislation. Last but not least, the ATAD will also impact third countries in different ways. Conflicts with double tax treaties concluded between a member state and a third state will ultimately have to be resolved depending on the constitutional requirements of the member state concerned. From an EU law point of view, it seems convincing to resolve this problem by an analogue application of art. 351 TFEU, meaning that tax treaties concluded before the enactment of the ATAD should not be overwritten by national provisions implementing the ATAD. Rather, member states should endeavour to change the respective tax treaty. This chapter has also shown that the ATAD can serve as an example of negative tax competition by the EU towards third states. Whereas some provisions urge member states to provide for more favourable treatment of investments made within the EU, others are put it at the discretion of the member states.

Chapter 39

Alternative Di spu t e Resolu tion i n t h e Eu ropean U ni on Isabelle Richelle

39.1 Introduction The term ‘alternative dispute resolution’ could refer to means of resolving disputes other than the usual administrative or judicial remedies under domestic law or, considering cross-border tax disputes, other than mutual agreements between national tax administrations. Mutual agreement between competent authorities is a classic way of resolving disputes from a diplomatic point of view, while taking place directly between national tax authorities without the intervention of the diplomatic authorities themselves. More recently, a process of resolving tax disputes through arbitration has emerged, culminating in the adoption by the European Union in 2017 of a directive on tax dispute resolution, the main features of which are probably that it imposes arbitration on the member states and that the process must lead to the resolution of the dispute. However, the EU does not have a monopoly on thinking and initiatives on ‘tax arbitration’. Alternative dispute resolution may also refer to a way of preventing disputes from arising, through advanced agreements with the tax authorities; this practice has also largely developed in recent years. In the next section of this chapter, we will summarize how these alternative dispute- resolution mechanisms have developed. In the following section, we will look in more detail at the technical measures implemented in this area by the European Union.

698 Isabelle Richelle

39.2 Taxation and Cross-B order Disputes 39.2.1 Introduction From time immemorial, there have been tensions and conflicts between human beings. If all too often they have led to physical violence, there can be no doubt that there have always been people of good will who were concerned to calm the situation and to propose non-violent solutions acceptable to all parties. Thus, in Europe in particular, a judicial system1 has developed that is gradually becoming independent2 and incorporating the rules of a balanced trial between the parties.3 Conflicts can also arise in tax matters, which makes them distinct because they affect the fiscal sovereignty of states. A first form of conflict opposes taxpayers to the state4 on the very principle of taxation, which they consider too heavy, unfair, discriminatory, etc.; the conflict is then essentially political. It often leads to demonstrations, opposition, and even revolt against the state.5 Recognition of the principle of consent to taxation will generally lead to at least a momentary appeasement.6 Opposition to (or through) taxation is also used as a form of resistance in other protests. For example, the English women’s movement based its demand for the right to vote on a refusal to pay taxes based on the adage ‘no tax without representation’.78

1

Thus, e.g., a Charter of Franchise of Huy of 1066 stipulated the prohibition of private vengeance and compulsory recourse to the courts, and the right to oppose any “extraordinary charges” imposed by the seigneur. 2 The Declaration of the Rights of Man and of the Citizen of 26 August 1789 proclaims, in article 16, that ‘Any society in which no provision is made for guaranteeing rights or for the separation of powers, has no Constitution.’ 3 See art. 10 of the Universal Declaration of Human Rights; art. 6 of the European Declaration of Human Rights. 4 Understood here in the broad sense of ‘collecting power’. 5 Tax equality was one of the main demands during the French Revolution of 1789, as revealed in the “Cahiers de Doléances of the Third Estate”, challenging the privileges of the nobility and the clergy (see French archives: https://francearchives.fr/fr/article/163458854). Other movements of fiscal opposition have punctuated history, notably the Barons’ revolts in England, leading to Magna Carta (1215), the American Tea Party, and more recently the ‘Yellow Vests’ in France. 6 See, in this respect, certain founding texts, such as Magna Carta (1215): ‘Neither scutage no raid shall be imposed on our kingdom, unless by common counsel of our kingdom, except for ransoming our person . . .’; N. Delalande, ‘Le retour des révoltes fiscales?’, Pouvoirs 4/151 (2014), 15–25. 7 Women’s Tax Resistance League, at the beginning of the twentieth century. 8 See also, in the nineteenth century, the Quakers who refused to pay tithes and taxes during the campaigns for the abolition of slavery; or the tax resistance movements during the Boer War. In the early twentieth century, there were movements against progressive income tax.

Alternative Dispute Resolution in the European Union 699 In a second form of conflict, the taxpayer is opposed to the state over a particular tax that is claimed from and the debit of which it disputes. Here again, because it involves public resources, public authorities will usually frame the dispute with specific rules relating to the public law relationship between the two parties. However, recent decades have been characterized, in European countries at least, by a wider recognition of the right to a ‘fair trial’ in tax proceedings, whether at an administrative or judicial procedural stage. The rights to information, to access one’s tax file, to be heard, and, more broadly, the right to a fair procedure and trial are increasingly recognized and applied. The complexity of tax law, the increased need for legal certainty, and the need to speed up the decision-making process have led to the emergence in recent years of new practices such as advance rulings9 and national arbitration procedures.10 Also at the national level, original solutions are being developed to resolve a dispute when the taxpayer is in such a situation that it is futile to pursue the recovery of the tax or in a desire to offer a new framework for conciliation between the administrative body and the taxpayer.11 Where the tax dispute has a foreign element, domestic methods of resolving tax disputes may no longer be sufficient, as a state may not be able to resolve the difficulty alone. Over time, international practice has developed different methods of resolving problems arising from cross-border tax situations, mainly through interstate negotiations (the mutual agreement procedure) and arbitration. In this international context, we can identify three key situations in which a dispute involving a tax issue may arise. Such tax dispute may arise in the context of a commercial dispute between private persons; an arbitral tribunal may take up the issue but may not be able to resolve a possible dispute between the person and the tax authority.12 A tax dispute may also arise between an investor and a host state and could be resolved under the specific arbitration procedures provided for by treaty, such as investment treaties 9 Advance rulings are becoming more common throughout Europe, and are not restricted to the field of corporate taxation and transfer pricing. As the European Parliament notes, ‘the practice of rulings developed, in the framework of a closer and more cooperative relationship between tax administrations and taxpayers, as a tool to tackle the increasing complexity of the tax treatment of certain transactions in an increasingly complex, global and digitalised economy’ and ‘neither the OECD nor the European Commission have called for an end to the practice of tax rulings as such’ (Motion for a European Parliament Resolution on tax rulings and other measures similar in nature or effect, 2015/2066(INI) (5 Nov. 2015), points X and Y, https://www.europarl.europa.eu/doceo/document/A-8-2015-0317_EN.html. 10 e.g. Portugal introduced domestic arbitration for tax cases in 2011 with the aim of improving the speed and flexibility in the resolution of tax disputes; it created a Tax Arbitration Centre that, among other functions, conducts proceedings and manages the constitution of the arbitration courts. This arbitration is considered in accordance with the Portuguese constitution, and the CJEU stated that Tax Arbitral Court constitutes a ‘jurisdictional body’ under art. 267 TFEU (Case C-377/13 Ascendi, ECLI:EU:C:2014:1754). See S. Vasques, Tax Arbitration and VAT: The Portuguese Experience’, International VAT Monitor (2020), 267–271. 11 See, e.g., in Belgium the mechanism of indefinite postponement of recovery (Law of 25 April 2007, Belgian OJ, 8 May 2007) and the setting up of a ‘fiscal conciliation’ service (art. 413(2)–(8) Income Tax Code). 12 See, e.g., W. Park, ‘Tax and Arbitration’, Arbitration International 36 (2020) 157–220, esp. 164–166.

700 Isabelle Richelle (bilateral investment treaties, BITs).13 Finally, the tax dispute may occur between the taxpayer and one or more tax authorities concerning a cross-border tax situation. In the remainder of this chapter, we will focus on the latter two cases, where the taxpayer is directly in dispute with the tax authority. We will discuss here the fundamental issue of whether tax disputes may or may be not settled by arbitration at all,14 which we take for granted. The scope of the study is limited to income taxes.

39.2.2 Cross-Border Tax Disputes and Bilateral Investment Treaties The main purpose of BITs is to protect international investments; they usually include an arbitration mechanism allowing the international investor to bring the host state before an international arbitral tribunal. To the extent that the BIT covers tax issues, there is room for arbitration to resolve tax issues between the investor and the host state.15 These treaties are now challenged for various reasons, including the possibility of ‘treaty shopping’ by investors.16 For its part, the European Commission believes that since the Union has acquired exclusive competence in the field of foreign direct investment following the entry into force of the Lisbon Treaty on 1 December 2009, bilateral agreements between member states should be terminated. The arbitration clause in the treaty between the Netherlands and Slovakia was found by the Court of Justice to be in breach of articles 267 and 344 TFEU in the well-known Achmea case.17 Indeed: having regard to all the characteristics of the arbitral tribunal mentioned in Article 8 of the BIT . . ., it must be considered that, by concluding the BIT, the Member States 13

See, e.g., ibid., esp. 166–197. e.g., ibid., esp. 162; J.- B. Racine, L’arbitrage commercial international et l’ordre public (Paris: LGDJ, 1999), 25 sv; B. Hanotiau, ‘L’arbitrabilité’, Collected Courses of the Hague Academy of International Law 296 (online at http://dx.doi.org/10.1163/1875-8096_pplrdc_A9789041118585_02). 15 See some 100 cases identified by UNCTAD: https://investmentpolicy.unctad.org/investm ent-dispute-settlement/advanced-search. M. Lang and al., eds, The Impact of Bilateral Investment Treaties on Taxation, WU Institute for Austrian and International Tax Law, Tax Law and Policy Series 8 (Amsterdam: IBFD, 2017), 590. R. Danon and S. Wuschka, ‘International Investment Agreements and the International Tax System: The Potential of Complementarity and Harmonious Interpretation’, Bulletin for International Taxation (2021), 687–703. 16 E. Lee, ‘Treaty Shopping in International Investment Arbitration: How Often Has it Occurred and How Has it Been Perceived by Tribunals?’, Working Papers Series no. 15-167, London School of Economics and Political Science (2015) (https://www.lse.ac.uk/international-development/Assets/ Docume nts/ PDFs/ D isse r tat i on/ Prize w inn i ng- D isser t ati ons/ P WD- 2 014/ 2 014- Eun jung L ee.pdf ); Q. C. Ngo and V. A. Ly, ‘Le chalandage de traités à l’épreuve des accords d’investissement de nouvelle génération’, Revue International de droit économique T. XXXI/3 (2017), 85–108; UNCTAD Series on Issues in International Investment Agreements II (Geneva: United Nations, 2011), 163 (https://unctad.org/fr/ system/files/official-document/diaeia20102_en.pdf); K. Hober, ‘Investment Treaty Arbitration: A brief Overview’, Intertax 3 (2014), 189–193, 190. 17 Case C-284/15 Achmea, ECLI:EU:C:2018:158. 14 See,

Alternative Dispute Resolution in the European Union 701 parties to it established a mechanism for settling disputes between an investor and a Member State which could prevent those disputes from being resolved in a manner that ensures the full effectiveness of EU law, even though they might concern the interpretation or application of that law.18

The EU Commission and the doctrine concluded that the findings of the Achmea judgment are transferable to all intra-EU investment arbitrations which are based on a BIT.19 The Commission extends this solution to the Energy Charter Treaty (ECT),20 a view also shared by a broader doctrine.21 In its decision in Moldova v. Komstroy, the Court of Justice of the European Union (CJEU) ruled that the ECT’s arbitration clause was not ‘applicable to disputes between a MS [member state] and an investor of another MS concerning an investment made by the latter in the first MS’.22 With regard to investment treaties, discussions are now closed by the agreement for the termination of BITs between the EU member states which was concluded on 5 May 2020.23 The Court also ruled that the Achmea doctrine extends to contract-based investment arbitrations involving member states;24 the question remains open in the absence of an inter-state treaty.25

18 Ibid.,

para. 56. V. Korom, ‘Intra-EU BIT’s in Light of the Achmea Decision, Central European Journal of Comparative Law 1 (2022), 97—118. 19 J. Scheu and P. Nikolov, ‘The Setting Aside and Enforcement of Intra-EU Investment Arbitration Awards after Achmea’, Arbitration International 36 (2020), 253–274, 257; J. Scheu and P. Nikolov, ‘The Incompatibility of Intra-EU Investment Treaty Arbitration With European Union Law—Assessing the Scope of the ECJ’s Achmea Judgment’, German Yearbook of International Law 62 (2019), 475–503, 479 (the authors rely on the words ‘international agreement concluded between Member States, such as Article 8 of the BIT’ (Achmea, para. 60); J. Monsenego, ‘Does the Achmea Case Prevent the Resolution of Tax Treaty Disputes Through Arbitration?’, Intertax 47/8–9 (2019), 725–736. 20 Concluded between Member States, the European Union and third countries. See the above- mentioned Commission Communication of 19 July 2018, 4. In relations within the European Union, the Commission considers that investors find in European law the necessary protections, at least equivalent to those offered by the BITs. 21 Scheu and Nikolov, ‘The Setting Aside and Enforcement of Intra- EU Investment Arbitration Awards after Achmea’, 258; Scheu and Nikolov, ‘The Incompatibility of Intra-EU Investment Treaty Arbitration with European Union Law—Assessing the Scope of the ECJ’s Achmea Judgment’, s. IV, 490. 22 Case C-741/19 Republic of Moldova v. Komstroy LLC, ECLI:EU:C:2021:655, para. 66; B. Böhme, ‘The Future of the Energy Charter Treaty after Moldova v. Komstroy’, Common Market Law Review 3 (2022), 853–870. 23 [2020] OJ L169/ 1. The Agreement provides for transitional provisions concerning ongoing arbitrations. The fate of the European Energy Charter is to be settled later. 24 Case C-109/20 Republiken Polen v. PL Holdings Sàrl, ECLI:EU:C:2021:875. See also Case C- 638/19 P Commission v. European Food SA and others, ECLI:EU:C:2022:50, esp. para. 137. A certain doctrine considers these contract-based arbitrations would fall outside the scope of Achmea; Lemaire and Laazouzi note the resistance of some arbitral tribunals to European case law (M. Laazouzi and S. Lemaire, ‘Chronique de jurisprudence arbitral en droit des investissements’, Revue de l’Arbitrage 2 (2019), 553–571, esp. 561). 25 Scheu and Nikolov, ‘The Setting Aside and Enforcement of Intra- EU Investment Arbitration Awards after Achmea’, 259.

702 Isabelle Richelle

39.2.3 Tax Dispute Resolution and Double Taxation Conventions Cases of double of multiple international taxation and difficulties in the interpretation or application of double taxation conventions have significantly increased in the previous years due to the implementation of the internal market by the EU and, more broadly, by the globalization of the economy; a further increase can probably be expected following the introduction of complex anti-abuse provisions through the base erosion and profit shifting (BEPS) Project.26 International double taxation, as such, is not illegal as it merely results from the parallel exercise by states of their taxing powers.27 Nonetheless, states have sought ways to reduce such international double taxation, mainly through developing a network of bilateral conventions. Typically, these agreements include a tax dispute-resolution clause which generally provides for a mutual agreement phase supplemented, more recently, by an arbitration mechanism.28 It is said that taxpayers are overall dissatisfied with the mutual agreement procedure (MAP), which they blame for its slowness, lack of transparency, and sometimes the feeling that one cass might be traded off against another,29 while on the other hand, governments are reluctant to hand over to independent arbitrators cases that directly affect their tax revenues. Only very recently has the idea of binding arbitration been accepted by states, leading to the inclusion of Part VI in the OECD Multilateral Instrument (MLI) and the adoption of the EU Tax Dispute Resolution Directive. The idea of compulsory arbitration had, however, already been proposed in 1928 by the League of Nations, whose Draft Model Treaty on Income Tax provided in article 1430 for a mechanism to resolve ‘disputes arising between the Contracting States concerning the interpretation of this Convention’. This was an original mechanism, applicable in the absence of an amicable settlement between the states concerned, entrusting ‘with a view to an amicable settlement’ to a technical body to be set up by the Council of the League of Nations, the task of giving ‘an advisory opinion’ ‘after hearing the parties and arranging a meeting between them if necessary’. The text also specified that ‘the Contracting States may agree, prior to the opening of such procedure, to regard the advisory opinion given by the said body as 26 The BEPS Plan that led to the adoption of the MLI at the OECD level and the ATAD Directives: E. I. Richelle, ‘Atad, Double Taxation, Tax Dispute Resolution and Taxpayer Protection’, in W. Haslehner, G. Köfler, and A. Rust, eds, A Guide to the Anti-Tax Avoidance Directive (Cheltenham: Edward Elgar, 2020), 256–276. 27 See, e.g., Case C- 403/19 Société Générale, ECLI:EU:C:2021:136, paras 27–29; E. Traversa and B. Bodson, ‘Droit de l’Union européenne et conventions préventives de la double imposition: entre contradictions et complémentarités’, Cahiers de Droit Européen 1 ( 2017), 177–204, 185. 28 See art. 25.5 of the OECD Model Tax Convention. K. Koch, ‘General Report: Mutual agreement procedure and practice’, Cahiers de Droit Fiscal International, vol. 66a (1981), 93–129, 99. 29 H. Mooij, ‘Tax Treaty Arbitration’, Arbitration International 35 (2019), 195– 219, 202; K. Koch, ‘General Report’, esp. 100. 30 The same provision was also suggested for inheritance tax.

Alternative Dispute Resolution in the European Union 703 final’. States retained the option of accepting the opinion or replacing the treaty procedure with any other arbitral or judicial procedure of their choice. The Commentary stated that article 14 ‘is based upon the text inserted in other international conventions, in particular the Convention for the Simplification of Customs Formalities signed at Geneva on November 3rd, 1923’.31 The principle of an amicable conflict-resolution phase does not pose any particular difficulty for states; it fits in with their classic diplomatic practice. On the other hand, the arbitration phase entrusted to a specific body, with the possibility that an external opinion could ultimately be binding, was rejected at the time.32 The 1943 Mexico Model, the 1946 London Model,33 and the 1963 OECD Model Tax Convention only provide for a mutual agreement procedure which recognizes the right for the taxpayer to address themselves to the competent authority of their state of residence;34 when the latter is not able to arrive at an appropriate solution itself—after a ‘careful examination’,35 it will communicate with the competent authority of the other state with the view to reaching an agreement on the taxation in dispute.36 The 1977 Model introduces a time limit of three years for presenting claims under the paragraph 137 and establishes the rule that the agreement reached must be implemented notwithstanding any time limits in the domestic law of the contracting states.38 Subsequent OECD studies have shown, on the one hand, that taxpayers’ interest in an effective cross- border dispute-resolution mechanism is tempered by a feeling of arbitrariness,39 and,

31 Rapport

présenté par la Réunion générale d’Experts gouvernementaux en matière de double imposition et d’évasion fiscale, ‘Double imposition et évasion fiscale’ (Geneva: Société des Nations, 1928), Q 562.M.178.192.1928.II, esp. 18 (https://biblio-archive.unog.ch/Dateien/CouncilMSD/C-562- M-178-1928-II_FR.pdf); Report presented by the General Meeting of Government Experts on Double Taxation and Tax Evasion, ‘Double Taxation and Tax Evasion’ (Geneva: League of Nations, 1928), Q 562.M.178.192.1928.II, esp. 18 (https://biblio-archive.unog.ch/Dateien/CouncilMSD/C-562-M-178-1928- II_EN.pdf). 32 M. Markham, ‘Mandatory Binding Tax Arbitration—Is This a Pathway to a More Efficient Mutual Agreement Procedure?’, Arbitration International 35 (2019), 149–170, 151. However, it is worth noting that the UK–Irish Free State Convention of 1926 already included an arbitration clause (see Park, ‘Tax and Arbitration’, 198, fn. 165). 33 https://biblio-archive.unog.ch/Dateien/CouncilMSD/C-88-M-88-1946-II-A_FR.pdf. 34 Whether or not it has exhausted national legal remedies. 35 Com OECD(1963) 25/3. 36 Arts 25.1 and 25.2, OECD Model Tax Convention, 1963, https://read.oecd-ilibrary.org/taxation/ draft-double-taxation-convention-on-income-and-capital_9789264073241-en#page47; Koch, ‘General Report’; R. Ismer, ‘Mutual Agreement Procedure’ in E. Reimer and A. Rust, eds, Vogel on Double Taxation Conventions, 1735–1826 (Alphen aan den Rijn: Kluwer, 2015); B. Gouthiere, Les impôts dans les affaires internationales, 14th ed. (Paris: Francis Lefebvre, 2020), 1638. 37 Where the 1963 Model only suggested fixing bilateral time limits that are ‘reasonably generous’, Com OECD(1963, 25/5. 38 New art. 25.2 final sentence. Koch, ‘General Report’, esp. 97. 39 OECD, Transfer Pricing and Multinational Enterprises: Three Taxation Iissues (Paris: OECD Publishing, 1984), ch. II, 17, para. 38; OECD, lReport Improving the Process for Resolving International Tax Disputes, version released for public comment. D. R. Tillinghast noted that ‘there is always a fear of being a sacrificial lamb when two competent authorities have several cases to be resolved at the same

704 Isabelle Richelle on the other hand, a number of states are reluctant to submit to the advice of experts outside their tax administrations.40 Nevertheless, the Model Convention 2008 proposed an arbitration process for unresolved issues at the request of the person who presented the case.41 The Commentary permits contracting states that have not included the paragraph in their conventions to implement an arbitration process by mutual agreement for general application or in order to deal with a specific case.42 In the end, this option will have little success, since it is reported that in 2014 only 158 DTAs out of a network of about 3,000 agreements provided for it,43 and 200 DTAs out of about 5,000 treaties in 2019,44 while the number of MAPs has constantly increased.45 The implementation of the OECD BEPS Plan required a strengthening of the mechanisms for resolving tax disputes. Action 14 of the Plan is presented in this sense, aiming to make the mutual agreement procedure more effective by imposing a ‘minimum standard’, complemented by ‘good practices’, combined with a robust peer-review process; this set of standards should increase the protection of taxpayers.46 In addition, participating states (‘inclusive framework jurisdiction’) may still offer to the taxpayers, where the MAP is unsuccessful, the possibility of binding arbitration, the contours of which are defined in Part VI of the MLI.47 In July 2022, some thirty- three states had joined the arbitration procedure, including seven EU member states.48 The MLI imposes, as mandatory minimum standards, the guarantee of a mutual

time’, in Issues in the Implementation of the Arbitration of Disputes Arising under Income Tax Treaties (Amsterdam: IBFD, 2002), 90–99, 91. 40

See, e.g., OECD, Transfer Pricing and Multinational Enterprises, ch. III, para. 38, 25/65–66. New art. 25.5 of the Model Convention 2008. See M. de Ruiter, ‘Supplementary Dispute Resolution’, European Taxation 9 (2008), the 2008 OECD Model Tax Convention Special Issue, at 493–499; C. Garbarino and M. Lombardo, ‘Arbitration of Unresolved Issues in Mutual Agreement cases: The New Para. 5, Art. 25 of the OECD Model Convention, a Multi-Tiered Dispute Resolution Clause’ in M. Lang et al., eds, Tax Treaties: Building Bridges between Law and Economics (Amsterdam: IBFD, 2010), 459–479. 42 Com. OECD (2008), 25/69. 43 Markham, ‘Mandatory Binding Tax Arbitration’, 153, quoting H. M. Pit, ‘Arbitration under the OECD Model Convention: Follow-Up under Double Tax Conventions: An Evaluation’, Intertax 42 (2014), 445–469, 466; H. M. Pit, ‘Arbitration under the OECD MLI, Options and Choices’, Bulletin for International Taxation 71/10 (2017), 445. 44 Mooij, ‘Tax Treaty Arbitration’, 208. 45 See the statistics published by the OECD and the EU Commission. 46 BEPS Action 14, https://www.oecd.org/tax/beps/beps-actions/action14/. P. Valente, ‘Resolving Tax Cross-Border Disputes: Developments in the EU and Around the Globe’, Tax Notes International (2019), 845–852, 847. M. Markham, ‘Action 14 of the BEPS Project: Taking the Pulse of Tax Certainty and Determining the Effectiveness of the Peer Review Process Five Years On’, Bulletin for International Taxation 2 (2022), 96–107. Multilateral Instrument, adopted by virtue of BEPS Action Plan 15. J. Avery- Jones, ‘Types of Arbitration Procedures’, Intertax 8–9 (2019), 674–677. 47 Multilateral Instrument adopted by virtue of BEPS Action Plan 15. Avery- Jones, ‘Types of Arbitration Procedures’. 48 In May 2023, the OECD published thirty country profiles applying arbitration under the multilateral BEPS Convention: https://www.oecd.org/fr/fiscalite/conventions/beps-instrument-multi 41

Alternative Dispute Resolution in the European Union 705 agreement procedure and a settlement of correlative transfer pricing adjustments.49 At the same time, the Model Convention and its Commentary were adapted in the 2017 update, mainly by deleting state reservations to arbitration based on constitutional considerations and inserting in the Commentary a ‘Model Mutual Agreement on Arbitration’ specifying the scope of the mechanism and the implementation procedures.

39.3 Alternative-Dispute Resolution and the EU The EU’s objective, since the signing of the Treaty of Rome in 1957, has been to establish an internal market without frontiers in order to allow free movement and healthy economic competition. There is no doubt that this objective implies the abolition of the fiscal frontiers constituted by international double or multiple taxation. However, it is still considered that ‘double taxation is not contrary to the treaties, as long as it results from the parallel exercise of tax sovereignty by the Member States concerned’.50 As far as companies are concerned, freedom of establishment has opened the door to intra- European structuring, thus increasing the risk of double taxation, in particular in the case of adjustment of profits between associated companies or between the head office and a permanent establishment in another member state. Considering that such double taxations are ‘likely to cause distortions in the conditions of competition and in capital movements and therefore to affect the operation of the common market’, the European Commission proposed, in 1976, a procedure under which, in the first instance, the case would be submitted to the tax authorities of the two member states concerned with a view to them settling the dispute by mutual agreement and where, ‘in the absence of such agreement, the matter in dispute should be submitted to a commission, consisting both of representatives of the tax authorities concerned and of independent persons of standing, whose decision both the tax authorities and the enterprises concerned accept from the outset’.51 The proposal was rejected by the member states. However, the need for an appropriate mechanism was regularly brought back to the table.52. Finally, after

lateral-signataires-et-parties.pdf. D. Duff and D. Gutman, ‘General Report: Reconstructing the Treaty Network, Subject 1’, Cahiers de Droit Fiscal International, vol. 105A (2020), 35–36. 49 MLI, arts 16 and 17. N. Bravo, ‘Mandatory Binding Arbitration in the BEPS Multilateral Instrument’, Intertax, 47/8 (2019), 693–7 14. 50 ‘Double Taxation in the Single Market’, Communication from the Commission to the European Parliament, the Council and the European Economic and Social Committee, COM(2011) 0715 final. 51 Proposal for a Council Directive on the elimination of double taxation in connection with the adjustment of transfers of profits between associated enterprises (arbitration procedure), submitted on 29 November 1976, COM(76) 611 final [1976] OJ C301/4. 52 See the Ruding Report, ‘EC Commission, Report of the Committee of Independent Experts on Company Taxation’ (Mar. 1995), 205; Commission Communication subsequent to the conclusions of the

706 Isabelle Richelle lengthy negotiations, the proposal for a directive was transformed into an intergovernmental agreement adopted on 23 July 1990.53 Initially concluded for a period of five years, the Arbitration Convention is now extended by automatically renewable five-year terms, unless a member state objects.54 As this is a convention between member states, its legal status is debated; however, it is not disputed that the text is a multilateral convention under international law and that it does not fall within the interpretative scope of the Court of Justice.55 A number of practical details are given for its application in a Code of Conduct,56 revised in 2009 to become the Revised Code of Conduct for the effective implementation of the Arbitration Convention.57 One of the questions raised by the implementation of the Arbitration Convention is which standards to apply in assessing the adjustment of profits. While article 4 of the Convention refers to the term ‘agreed between independent enterprises’ as a framework for the Advisory Commission,58 in the absence of European harmonization of transfer pricing rules, member states and practice have referred to the OECD transfer pricing (TP) standard (mainly the arm’s-length principle). It is in this context that a Joint Transfer Pricing Forum (JTPF) was set up in 2002 on an informal basis with the objective of submitting to the Commission ‘pragmatic, non-legislative solutions to practical problems posed by transfer practices in the EU’. This informal working group was institutionalized by a Commission decision in 200659 with the main tasks of ‘advising the Commission on TP tax issues’ and assisting the Commission ‘in finding practical solutions, compatible with the OECD TP Guidelines’.60 Ruding Committee indicating guidelines on company taxation linked to the further development of the internal market, SEC(92)1118 (26 June 1992). 53

Convention 90/436/EEC on the elimination of double taxation in connection with the adjustment of profits of associated enterprises [1990] OJ L225/10. D. Schelpe, ‘The Arbitration Convention: Its Origin, Its Opportunities and Its Weaknesses, EC Tax Review 2 (1995), 68–77; D. Schelpe, ‘Opportunities and Weaknesses of the Multilateral Arbitration Convention’, International Transfer Pricing Journal 2 (1995), 119— 129; H. M. Pit, ‘Dispute Resolution in the European Union— The EU Arbitration Convention and the Dispute Resolution Directive’, IBFD Doctoral Series 42 (2018), 1729. 54 Protocol amending the Convention of 23 July 1990 on the elimination of double taxation in connection with the adjustment of profits of associated enterprises [1999] OJ C202/1. 55 See G. Kofler, ‘Tax Disputes and the EU Arbitration Convention’, in E. Baistrocchi, ed., A Global Analysis of Tax Treaty Disputes (Cambridge: Cambridge University Press, 2017), 205–236, 213; L. Hinnekens, ‘Different Interpretations of the European Tax Arbitration Convention’, EC Tax Review 4 (1998), 247–257; Pit, ‘Dispute Resolution in the European Union’, 149. 56 Code of conduct for the effective implementation of the Convention on the elimination of double taxation in connection with the adjustment of profits of associated enterprises [2006] OJ C176/8. 57 2009 Revised Code of Conduct for the effective implementation of the Convention on the elimination of double taxation in connection with the adjustment of profits of associated enterprises [2009] OJ C322/1. See D. de Carolis, ‘The European Arbitration Convention on Transfer Pricing: Legal Fossil or Instrument of New European Governance?’, Intertax 41/5 (2013), 308–312. 58 See Advisory Commission, art. 11.1. 59 Commission Decision of 22 Dececber 2006 (2007/75/EC) setting up an expert group on transfer pricing [2007] OJ L32/189, replaced by successive decisions until Commission Decision of 26 January 2015 (2015/C28/05) setting up the ‘EU Joint Transfer Pricing Forum’ expert group [2015] OJ C28/5, applicable until 31 March 2019. 60 Decision, art. 2.

Alternative Dispute Resolution in the European Union 707 The JTPF contributes to the development and dissemination of harmonized transfer pricing practices within the EU, including the documentation that multinationals must provide to tax authorities on the pricing of cross-border intra-group transactions and the organization of coordinated transfer pricing audits.61 It was also the work of the JTPF that led the Commission to propose the Guidelines for Advance Pricing Agreements within the EU in 2007.62 Finally, the implementation of the OECD BEPS recommendations in the Anti-Tax Avoidance Directive (ATAD) Guidelines required a broader tax dispute-resolution mechanism to be put in place. Indeed, in May 2017, the Ecofin Council made the statement that ‘Member States shall endeavor to explore the possibilities to further enhance the resolution of disputes among Member States relating to the interpretation and application of tax agreements and conventions by way of a permanent body, including the possibilities provided for under Article 273 TFEU.’ This resulted a few months later in the adoption of the ‘Dispute Resolution’ Directive.63

39.3.1 The Arbitration Convention The Arbitration Convention (AC) introduces a mechanism for the correction of international double taxation resulting from an upward adjustment of the profits of a related enterprise or an adjustment of the results of a permanent establishment. It provides for a multi-step procedure with the objective of preventing or correcting double taxation.64 61 EU

JTPF, ‘A coordinated approach to transfer pricing controls within the EU’, Doc. TPF/013/ 2018/ EN (Oct. 2018), https://taxation-customs.ec.europa.eu/system/files/2018-10/jtpf_report_on_a_ coordinated_approach_to_transfer_pricing_controls_within_the_eu_en.pdf. See also on simultaneous controls, art. 12 of Council Directive 2011/16/EU of 15 February 2011 on administrative cooperation in the field of taxation and repealing Directive 77/799/EEC [2011] OJ L064/1. 62 Communication from the Commission to the Council, the European Parliament and the European Economic and Social Committee on the work of the EU JTPF in the field of dispute avoidance and resolution procedures and on Guidelines for Advance Pricing Agreements within the EU, COM(2007) 71 final (26 Feb. 2007). 63 Council Directive (EU) 2017/ 1852 of 10 October 2017 on tax dispute resolution mechanisms in the European Union [2017] OJ L265/1. S. Govind, ‘The New Face of International Tax Dispute Resolution: Comparing the OECD Multilateral Instrument with the EU Dispute Resolution Directive 27(6)’, EC Tax Review (2018), 309. For an overview of national transposition provisions in some countries, see R.-A. Papotti, Tax Dispute Resolution: A Commentary on the EU Council Directive 2017/ 1852 (Alphen aan den Rijn: Kluwer International, 2020), 247. For a comparative analysis of the EU and OECD instruments, see P. Pistone, ‘The Settlement of Cross-Border Tax Disputes in the European Union’ in P. Wattel, O. Marres, and H. Vermeulen, eds, Terra and Wattel on European Tax Law, 7th ed. (Deventer: Wolters Kluwer, 2018), 331–379; J. P. Owens, ‘Mandatory Tax Arbitration: The Next Frontier Issue’, Intertax 46/8–9 (2018), 610–619. 64 On the details of this procedure, see L. Hinnekens, ‘The European Tax Arbitration Convention and its Legal Framework’, British Tax Review 2 (1996), 132–154 and 3 (1996), 272–311; Hinnekens, ‘Different Interpretations of the European Tax Arbitration Convention’; L. Hinnekens, ‘The Tax Arbitration Convention. Its Significance for EC Based Enterprise, the EC Itself and for Belgian and International Tax Law’, EC Tax Review 1 (1992), 70–105; M. Helminen, EU Tax Law—Direct Taxation (Amsterdam: IBFD, 2018), 335–362. Pit, ‘Dispute Resolution in the European Union—The EU Arbitration Convention and the Dispute Resolution Directive’, Part I.

708 Isabelle Richelle

39.3.1.1 Procedure In a first stage, the AC sets up a chain of information obligation so that the intention of a member state to make a profit adjustment is communicated to the company concerned and finally to the tax authority of the other member state concerned; the latter may accept the proposed adjustment and adjust downwards the tax base of the affiliated company, thus avoiding the occurrence of international double taxation. When double taxation is not avoided, the company may refer the case to the competent authority of its state of residence (or, where appropriate, of the state in which its permanent establishment is located) within three years ‘following the first notification of the measure which results or is likely to result in double taxation’ within the meaning of the Convention. The competent authority must then endeavour to find a solution, either unilaterally or by mutual agreement, with the other member state(s) concerned. If no mutual agreement is reached within two years, an ‘advisory commission’ is set up with the task, within a period of six months, of ‘issuing an opinion on how to eliminate the double taxation in question’. Within six months of that opinion, the competent authorities ‘shall take a decision by mutual agreement ... to ensure the elimination of the double taxation’.65 The member states concerned are therefore obliged to find a solution to the double taxation. This is an obligation of result which marks a definite innovation in the way international difficulties are resolved, since at the time the Convention was signed, the double tax conventions only proposed a mutual agreement procedure.

39.3.1.2 Scope of the Convention The AC applies to ‘enterprises’ of a contracting member state associated with an enterprise of another contracting member state or having a permanent establishment in that other state; the enterprise may be conducted as a sole proprietorship or as a partnership. The Convention remains applicable even if, subsequent to the tax years concerned, the enterprise has undergone restructuring.66 The AC imposes on member states the obligation to eliminate double taxation resulting from an adjustment of profits, on the one hand, in the relationship between affiliated enterprises where the transactions must be assessed in accordance with the arm’s-length principle and, on the other hand, in the relationship between the head office and its permanent establishment, the result of which must be determined as if it had constituted a ‘separate enterprise’. The spirit and wording of the Convention text is very close to the OECD rules in this area. The 2009 Revised Code of Conduct also clarifies that ‘the arm’s length principle will be applied, as advocated by the OECD, without regard to the immediate tax consequences for any particular Member State. 65

Convention, art. 12.1. The procedure will have a maximum duration of three years and six months, i.e. two years for the MAP phase, six months to set up the advisory committee, six months for that committee to give its opinion, and six months for the final decision of the competent authorities. 66 EU JTPF, Final Report on Improving the Functioning of the Arbitration Convention, Doc. JTPF/ 002/2015/EN (Mar. 2015), point 6.

Alternative Dispute Resolution in the European Union 709 (b) Article 4(2) of the AC should be interpreted in the spirit of the most recent version of Article 7 of the OECD Model Tax Convention . . . and the relevant Commentary’.67 It also specifies that the Convention applies even if the adjustment does not lead to the payment of a tax (a loss-making situation),68 that it covers triangular operations—in which case a third member state may be included in the process69—and cases of thin capitalization.70

39.3.1.3 Cases of exclusion The application of the Convention may be denied when a ‘serious penalty’ is imposed on the enterprise,71 terms which are specified in unilateral declarations by the member states in the annex to the AC. According to the Revised Code of Conduct, such denial should apply only in ‘exceptional cases’, such as ‘tax fraud, willful default and gross negligence’.72 The JTPF stresses the difficulty of transfer pricing cases and the need to avoid the overly rapid application of sanctions, which would then allow the application of the Convention to be refused.73

39.3.1.4 Taxpayers’ rights Although the AC gives an enterprise the right to initiate the procedure, its implementation is essentially in the hands of the tax authorities with little possibility for the taxpayer to drive forward the procedure or intervene in it;74 the taxpayer is not a party to the procedure, merely an ‘interested third party’. The Convention does not specify in particular the admissibility of the claim, the obligation to inform the taxpayer, or the means to appeal a decision of inadmissibility. However, article 10 provides for communication 67

Revised Code of Conduct, point 6. See also G. Kofler, ‘The Relationship Between the Arm’s Length Principle in the OECD Model Tax Treaty and EC Tax Law (Part 1)’, Journal of International Taxation 16 (2005), 32–45; D. de Carolis, ‘The Attribution of Profits to Permanent Establishments Within the Framework of the European Arbitration Convention: The Interpretation of Article 4(2) of the AC’, Intertax 43/8 (2015), 78–95. 68 Revised Code of Conduct, point 1.a). 69 Ibid., point 3. 70 Ibid., point 4. 71 AC, Art. 8.1. 72 Revised Code of Conduct, point 8. P. Valente, ‘Arbitration Convention 90/436/EEC: Inapplicability in Case of Serious Penalties’, Intertax 3 (2012), 220–224; Pit, ‘Dispute resolution in the European Union’, 522–546. 73 EU JTPF, Final Report on Improving the Functioning of the Arbitration Convention, Doc. JTPF/ 002/2015/EN (Mar. 2015), point 2.15; EU JTPF, Summary Report on Penalties, Commission Staff Working Document, SEC(2009) 1168 final (14 Sept. 2009). 74 L. Hinnekens, ‘European Arbitration Convention: Thoughts on its Principles, Procedures and First Experience’, EC Tax Review 19/3 (2010), 109–116; P. Adonnino, ‘Some Thoughts on the EC Arbitration Convention’, European Taxation 11 (2003), 403–408; K. Perrou, ‘Participation of the Taxpayer in MAP and Arbitration: Handicaps and Prospects in International Arbitration in Tax Matters’, in M. Lang et al., eds, International Arbitration in Tax Matters, (Amsterdam: IBFD, 2016), 287ff; K. Perrou, Taxpayer Participation in Tax Treaty Dispute Resolution, Doctoral Series vol. 28, IBFD (2014)); P. Baker and P. Pistone, ‘BEPS Action 16: The Taxpayers’ Right to an Effective Legal Remedy Under European Law in Cross-Border Situations’, EC Tax Review 25/5–6 (2016), 335–345

710 Isabelle Richelle between the taxpayer and the Advisory Commission, to which the taxpayer may ‘furnish ... any information, evidence or documents which they consider relevant to the decision’; the Advisory Commission may also make requests for ‘information, evidence or documents’.75 The Revised Code of Conduct, however, includes certain recommendations concerning tax authority information on the taxpayer’s rights under the Convention, and points out the need for the tax authority to reach a result within a reasonable period of time. There is an obligation on the taxpayer to provide any requested additional information, which implies some participation by the taxpayer in the process. This also formalizes the relational process between the tax authorities organizing, inter alia, the exchange of information and documents on the case. These practical details, which clarify the framework and facilitate the organization of the work of the tax administrations, increase the transparency of the process for taxpayers.76 The JTPF also recommends that any penalties that would have been applied should be reduced as a result of the agreement reached under the Convention.77 Finally, the Convention provides for the possibility for the competent authorities to agree on the publication of their decision, subject to the consent of the enterprises concerned. Such publication, which contributes to the transparency of the process, is generally refused by the member states concerned.78

39.3.1.5 Conclusions The AC has experienced some practical success. The number of cases submitted to MAP is increasing, reaching 547 new cases in 2017, 961 in 2020 and 670 in 2021 and with more than 2,300 pending cases at the end of 202179 (1,900 for 2017; 2.200 for 2020), of which more than 50% had been pending for more than two years. However, the number of cases submitted to arbitration is particularly low (one in 2016, two in 2017, none in 2020, three in 2021, and the first procedure was closed in 200380), which probably reflects the efforts of the tax authorities to reach agreement at the MAP level to avoid going to arbitration.

75 See for an evaluation of the functioning of the Advisory Committee, S.-O. Lodin, ‘The Arbitration Convention in Practice: Experiences of Participation as an Independent Member of Arbitration (Advisory) Commissions’, Intertax 3 (2014), 173–175. 76 See also B. Farinha Aniceto da Silva, ‘Commission Communication on work of EU JTPF with proposal for revised Code of Conduct for implementation of the Arbitration Convention’, Highlights & Insights on European Taxation 2 (2010), 9–23. 77 EU JTPF, Summary Report on Penalties, Commission Staff Working Document, SEC(2009) 1168 final (14 Sept. 2009). 78 Lodin, ‘The Arbitration Convention in Practice: Experiences of Participation as an Independent Member of Arbitration (Advisory) Commissions, 174. 79 See EU Commission, ‘Overview of numbers submitted for Statistics on Pending Mutual Agreement Procedures (MAPs) under the Arbitration Convention (AC) at the End of 2021’, https://taxation-cust oms.ec.europa.eu/system/files/2023-04/AC%20MAP_2021_FINAL.pdf 80 Adonnino, ‘Some Thoughts on the EC Arbitration Convention’, 403.

Alternative Dispute Resolution in the European Union 711

39.3.2 Advance Pricing Agreements An APA is an agreement between tax administrations over the way in which certain transfer pricing transactions between taxpayers will be taxed in the future;81 it determines, in advance of controlled transactions, an appropriate set of criteria (e.g. method, comparables, and appropriate adjustments thereto, critical assumptions as to future events) for the determination of the transfer pricing for those transactions over a fixed period of time.82 The practice of APAs has developed informally, at the instigation of the JTPF, in the wake of transfer pricing dispute-resolution procedures. On the one hand, it was useful to develop a consensual view of APAs—since tax administrations and practitioners are brought together in the JTPF—which is adapted to the characteristics of the European market. On the other hand, what better way to prevent difficulties in resolving cross- border tax disputes than to ensure that they are prevented through prior agreement; APAs are therefore more about ‘dispute avoidance’ than ‘dispute resolution’. The APA results from a multi-stage process. First, the guidelines suggest an informal approach by the taxpayer to all the tax authorities potentially involved and they then submit an informal request to the tax authorities. If this ‘prefiling stage’ seems satisfactory, the taxpayer will submit a ‘formal application’ for the involved tax administrations to decide whether or not the proposed TP treatment is acceptable. The TP treatment will be negotiated and agreed between the tax authorities involved (‘evaluation and negotiation stage’). Finally, the agreement is formally agreed in order to grant certainty to the taxpayer over the tax treatment of the transactions. Considering the growing number of APAs (by the end of 2020, about 1,300 were in force within the EU (about 1,000 in 2019), and it can be assumed that the process is viewed rather favourably by the practice. However, the statistics reveal significant imbalances between the member states, which naturally calls for further coordination efforts.83 Finally, it should be noted that since 1 January 2017, APAs are in principle subject to automatic exchange of information between the competent authorities of all other member states and with the European Commission.84

81

Communication from the Commission to the Council, the European Parliament and the European Economic and Social Committee on the work of the EU JTPF in the field of dispute avoidance and resolution procedures and on Guidelines for Advance Pricing Agreements within the EU, COM(2007) 71 final, 2.4/13 (26 Feb. 2007). 82 See definition in: https://taxation-customs.ec.europa.eu/system/files/2023-04/APA_consolidated_ 2021.pdf. 83 2019 and 2020 statistics: see PDFs at https://taxation-customs.ec.europa.eu/taxation-1/statistics- apas-and-maps-eu_en. 84 Art. 8.a.1 of Council Directive 2011/16/EU of 15 February 2011 on administrative cooperation in the field of taxation and repealing Directive 77/799/EEC [2011] OJ L064/1.

712 Isabelle Richelle

39.3.3 The Tax Dispute Resolution Directive 2017/1852 The removal of many technical barriers to free movement makes the existence of international double taxation within the single market even more troublesome, furthermore in a context where the introduction of anti-abuse measures risks increasing the number of such double or multiple taxation, making an effective corrective mechanism even more necessary. In October 2016, the EU Commission presented a draft directive85 on tax dispute-resolution mechanisms in the EU with the aim of ‘providing more certainty for businesses’.86 The proposal thus appears as a type of counterweight to the anti-avoidance measures adopted in the framework of the implementation of the BEPS initiative and the ATAD Directives; that is why the scope of the initial proposal is limited to corporate tax. The proposal was completely redrafted during the discussions in the Council, significantly changing its scope.87 The dispute-resolution procedure consists of three stages: the complaint, the mutual agreement procedure, and the arbitration procedure involving consultation of a specific Commission and a final decision by the competent authorities.88 It is also characterized by the various possibilities offered to the taxpayer to influence the procedure, thus enabling it to avoid possible blockage by the tax authorities. The directive applies to any complaint submitted from 1 July 2019 onwards relating to questions of dispute relating to income or capital earned in a tax year commencing on or after 1 January 2018.89

39.3.3.1 Scope of the directive The directive is addressed to any ‘affected person’ defined as ‘any person, including an individual, who is resident of a Member State for tax purposes, and whose taxation is

85

Communication from the Commission to the European Parliament and the Council, ‘A Fair and Efficient Corporate Tax System in the EU: 5 Key Areas for Action’, COM(2015) 302 final (17 June 2015), esp. point 3, 13. 86 Proposal for a Council Directive on Double Taxation Dispute Resolution Mechanisms in the European Union, COM(2016) 686 final; Commission Staff Working Document, Impact Assessment, SWD(2016) 343 final, esp. 34 and sv; G. Groen, ‘The Scope of the Proposed EU Arbitration Directive’, Tax Notes International 86/3 (2017), 243–254. 87 Council Directive (EU) 2017/1852 of 10 October 2017 on tax dispute resolution mechanisms in the European Union [2017] OJ L265/1. For a commentary on its country-by-country implementation, see R.- A. Papotti, ed., Tax Dispute Resolution. A Commentary on the EU Council Directive 2017/1852 (Alphen aan den Rijn, Kluwer, 2020), 247. 88 I. Richelle, ‘Dans les arcanes de la nouvelle directive sur le règlement des différends fiscaux’, in J. Wildemeersch and P. Paschalidis, eds, Liber Amicorum Melchior Wathelet (Brussels: Bruylant, 2018), 883–927; H. M. Pit, ‘The Changed Landscape of Tax Dispute Resolution Within the EU: Consideration of the Directive on Tax Dispute Resolution Mechanisms’, Intertax 47/8–9 (2019), 745–759; Pit, Dispute Resolution in the European Union—The EU Arbitration Convention and the Dispute Resolution Directive’, Part II, 1149; CFE Fiscal Committee, ‘Opinion Statement FC 4/2017 on the Proposed Directive on Double Taxation Dispute Resolution Mechanisms in the European Union’, Europen Taxation 57/ 8 (2017); G. Zeyen, ‘When Taxation Meets Arbitration: Recent Initiatives to Introduce and Promote Arbitration in International and European Taxation’, Intertax 45/11 (2017), 722. 89 Directive, art. 23.

Alternative Dispute Resolution in the European Union 713 directly affected by a question in dispute’.90 This includes natural persons, acting in the course of an economic activity or in a private capacity, as well as legal persons, whether profit-making or not, provided they are resident for tax purposes in a member state. The directive applies where a dispute arises ‘from the interpretation and application of agreements and convention that provide for the elimination of double taxation of income and, where applicable, capital’.91 It does not cover indirect taxes (VAT, inheritance taxes, gift taxes, etc.). It is necessary and sufficient that there is a dispute concerning the interpretation or application of a covered convention, without requiring that there is international double or multiple taxation.92 There is no doubt that the ‘agreements or conventions’ referred to include double taxation conventions concluded between member states. But the procedure could also apply to other conventions which contain tax provisions and for which it will be necessary to consider if they ‘provide for the elimination of double taxation’ of income or capital (e.g. air or maritime transport conventions, provisions concerning the tax status of international organizations, BITs or the European Energy Charter) and a case- by-case examination will be necessary. Is the Arbitration Convention covered? Yes, if one considers that the mechanism put in place aims at the elimination of international double taxation;93 but no if the words ‘conventions which provide for the elimination’ refer to a convention containing rules allocating taxing power between member states.94

39.3.3.2 The three stages of the procedure 39.3.3.2.1 The complaint Any affected person ‘shall be entitled to submit a complaint on a question in dispute to each of the competent authorities of each Member State concerned, requesting the resolution thereof ’.95 The complaint must contain various information, identifying the taxpayer, the subject matter of the dispute, the legal provisions involved, and any appeals that have been lodged; the relevant factual elements must be specified and the relevant supporting documents must be attached. These elements must show the existence of 90

Directive, art. 2.1, d). Directive, art. 1. 92 This is clarified in recital 2 of the directive: ‘For this reason, it is necessary that there are mechanisms in the Union that ensure the effective resolution of disputes concerning the interpretation and application of such bilateral tax treaties and the Union Arbitration Convention, in particular disputes leading to double taxation’. Contra: S. Ronco, ‘The EU Directive on Tax Dispute Resolution Mechanisms in the European Union: A Flexible But Still Perfectible Tool for Resolving International Tax Disputes’, in P. Pistone and J. Goede, eds, Flexible Multi-Tier Dispute Resolution in International Tax Disputes (Amsterdam: IBFD, 2021), ch. 17, point 17.2.2: ‘If two readings of the provision are grammatically possible, it seems to us that recital 2 clarifies the point. This is reinforced by art. 16.7 allowing a MS to deny access to the dispute resolution procedure under Article 6 on a case-by-case basis where a question in dispute does not involve double taxation’. 93 Recitals 6 and 7 of the directive seem to support this. 94 On this question, see Richelle, ‘Dans les arcanes de la nouvelle directive’, 892; Pistone, ‘The Settlement of Cross-Border Tax Disputes in the European Union’, 349. 95 Directive, art. 3. 91

714 Isabelle Richelle a dispute and will enable the national authorities to decide on the admissibility of the complaint. Since each competent authority has to receive the complaint, it may be assumed that the same text should be submitted to them. However, art. 3.3f) of the directive allows the competent authority to request ‘any specific additional information’ ‘considered necessary to undertake the substantive consideration of the particular case’. In practice, the question arises as to whether a competent authority may require additional information in the claim to be filed (which undermines the uniqueness of the complaint and makes the procedure more burdensome for the taxpayer). Furthermore, it seems that a competent authority cannot reject a claim for lack of information without first requesting the relevant additional information.96 The complaint must ‘be submitted within three years from the receipt of the first notification of the action resulting in, or that will result in, the question in dispute’.97 The starting point of the delay is therefore the ‘receipt’ of a notification, a concept not defined by the directive which will be understood by reference to the domestic law of the member state concerned, a priori the state at the source of the dispute.98. A complaint submitted out of time is inadmissible and must be rejected. The complaint must be submitted in one of the official languages of the member state receiving it, in accordance with its national law, or in any other language which that member state accepts for thatpurpose.99 Upon receipt of the complaint, the competent authority must acknowledge receipt within two months and inform the other competent authorities accordingly. A competent authority also has a period of three months from receipt of the claim to request further information from the taxpayer. Failure by the taxpayer to respond may justify the inadmissibility of the request. The competent authority must decide on the admissibility of the complaint within six months of either its receipt or the receipt of any additional information requested. The decision on admissibility must be notified to the affected person. Failure to take a decision within the time is equivalent to an admissibility decision.100 Once the admissibility of the complaint is recognized, either the competent authority may ‘decide to settle the dispute on a unilateral basis, without involving the other competent authorities of the Member States concerned’, or the procedure may be continued by consultation between the competent authorities concerned. Where the complaint is inadmissible, two situations may arise. On the one hand, all the competent authorities may have agreed on inadmissibility for lack of form, late claim, or because no decision was taken: the affected person can challenge that decision before the national courts.101 On the other hand, if the complaint is rejected by only one 96

See Richelle, ‘Dans les arcanes de la nouvelle directive, 898–899. Directive, art. 3.1. 98 See Richelle, ‘Dans les arcanes de la nouvelle directive’, 900. 99 Directive, art. 3.1. 100 Directive, art. 5.2. 101 Directive, art. 5.3. 97

Alternative Dispute Resolution in the European Union 715 of the competent authorities, the affected person can refer the matter to the ‘Advisory Commission’ set up by the directive.102 If, in the first case, a national court recognizes the admissibility of the complaint, the affected person can then refer the case to the Advisory Commission on admissibility.103 A decision on admissibility by the Commission allows the mutual agreement procedure to continue. It should be noted, however, that the directive does not provide notification of its decision to the affected person; it can be concluded that this task is the responsibility of the competent authorities, without any binding time limit being set. The procedural arrangements for this phase of the procedure, and the relationship between the Advisory Commission, the competent authorities, and the affected person are unclear, to say the least; it should probably be assumed that it is the Advisory Commission’s task to ensure that the dispute-settlement process continues to run smoothly.104

39.3.3.2.2 Mutual agreement procedure When it is not possible for one member state to resolve the difficulty unilaterally, the member states ‘shall endeavour to resolve the question’ by mutual agreement within a period of two years,105 which may be extended to three years at the request of one of the competent authorities. Formally, the affected person is not involved in this MAP and the competent authorities are not obliged to inform it of a possible extension to the deadline for a decision. Thus, mutual agreement can be reached within a maximum of four years.106 In the absence of a mutual agreement, the competent authorities are obliged to inform the affected person, indicating the general reasons why agreement could not be reached.107 The directive does not specify the form or the time limit for this information. The mutual agreement between competent authorities must be notified without delay. It is binding on the authority. It is enforceable for the affected person provided that the person accepts the decision and waives, if necessary, any further appeal (which it must also prove).

39.3.3.2.3 Dispute-resolution procedure If no amicable settlement is reached, the affected person may refer the case to the Advisory Commission,108 which must issue an opinion. This referral must be made within fifty days of the notification of the decision of the competent authorities that they

102

Directive, art. 6.1, para. 1, a. Directive, art. 6.1, para. 2; provided that there is no further appeal pending. 104 See Richelle, ‘Dans les arcanes de la nouvelle directive’, 911–912. 105 Directive, art. 4.1, para. 1. 106 Maximum three months to request additional information; three months given to the taxpayer to respond; six months to decide on admissibility; two years possibly increased by one year to reach a friendly settlement. 107 Directive, art. 4.3. 108 Directive, art. 6.1. 103

716 Isabelle Richelle have been unable to reach a mutual agreement.109 On the basis of that opinion, the competent authorities are then invited to decide on the dispute. The Advisory Commission The Advisory Commission itself must be constituted, for each case submitted, within 120 days ‘from the date of receipt of the request for constitution’. Its president must inform the affected person ‘without delay’ of the constitution.110 The Advisory Commission is composed111 of both representative(s) of each competent authority and independent persons of standing; the chairman is chosen from a list of independent persons of standing. The directive establishes mechanisms to ensure the quality of the independent persons appointed,112 the monitoring over time of the required independence, and the conditions for challenging an independent expert.113 In fact, the Advisory Commission may be called upon to give its opinion not only when a MAP has not been reached, but also when it has ruled on the admissibility of the complaint and the competent authorities have not initiated the MAP within the prescribed time frame.114 The Commission’s opinion must be delivered within six months of its establishment, a period which may be extended by three months provided that the competent authorities and the affected persons are informed. The opinion must be based on the provisions of the applicable agreement or convention and on ‘any applicable national rules’.115 It will be adopted by a simple majority of the members, the chairman having the casting vote and will be notified by the chair to the competent authorities116. Based on the opinion, the competent authorities concerned will try to ‘agree on the way to settle the dispute’,117 within six months of the notification of the opinion. The opinion of the Advisory Commission is therefore not binding and the competent authorities may depart from it. However, it does become binding if the competent authorities do not agree within the time limit on the solution of the dispute.118 The rules for the functioning of the Advisory Commission must be laid down by the competent authorities. They are communicated to the affected person, together with the date on which the opinion is to be given and references to the applicable legal provisions. The directive specifies the content of these operating rules; an EU Commission Regulation sets out the rules in detail if the competent authorities have not been able

109

Directive, art. 8. Directive, art. 6.1, para. 3. 111 See Directive, art. 8. 112 See Directive, art. 9. 113 As regards the appointment of persons of standing, see Richelle, ‘Dans les arcanes de la nouvelle directive’, 909–911; S. Piotrowski et al., ‘Towards a Standing Committee Pursuant to Article 10 of the EU Tax Dispute Resolution Directive: A Proposal for Implementation’, Intertax 47 (2019), 678, 682–684. 114 See Directive, arts 6.3 and 6.2, para. 3. 115 Directive, art. 14.2. 116 Directive, art. 14.3. 117 Directive, art. 15.1. 118 Directive, art. 15.2. 110

Alternative Dispute Resolution in the European Union 717 to specify them among themselves or if they are incomplete.119 It results from these operational rules that affected persons are brought to communicate with the Advisory Commission. They may provide it with ‘any information, evidence or documents which may be relevant to the decision’, subject to the agreement of the competent authorities.120 They may also ask to appear or be represented before the Commission, again with the agreement of the authorities concerned. The affected persons may be requested by the Commission to appear before it and it may also request information, evidence, or documents from both the persons concerned and the competent authorities. However, it cannot be considered that the rules of a ‘fair trial’ are fully respected here, especially because of the authorization to be given by the competent authorities (or by one of them?) and the imbalance between the communications from and by the taxpayers and the competent authorities.121 Final decision after the Advisory Commission’s opinion Once the opinion of the Advisory Commission has been given, the competent authorities have to decide, and, in this context, the opinion is not binding. It becomes binding if the competent authorities fail to reach agreement within the required time limit.122 The directive requires the competent authority of each member state to notify the decision without delay. In the absence of notification within thirty days of the decision, the affected person ‘may appeal in its State of residence in accordance with the applicable national rules, in order to obtain the final decision’.123 However, it is difficult to see how such an appeal could be lodged within thirty days of a decision of which the affected person has not yet been informed, considering also that the taxpayer does not receive the opinion of the Advisory Commission.124 The decision of the competent authorities is, in principle, binding on the member states concerned. However, it is up to the affected person to accept it and to waive all domestic remedies within sixty days of notification of the final decision, in order to have it implemented. It is worth noting that the dispute-settlement decision is binding, whether it results from the amicable or the arbitration phase. However, it is only enforceable if it is taken during the mutual agreement stage; if it is taken during the arbitration stage, the taxpayer has to seek enforcement, if necessary, before the competent national court. A decision at the mutual agreement stage allows the taxpayer to obtain enforcement if it waives

119 Commission implementing Regulation (EU) 2019/52 of 24 April 2019 laying down standard Rules of Functioning for the Advisory Commission or Alternative Dispute Resolution Commission and a standard form for the communication of information concerning publicity of the final decision in accordance with Council Directive (EU) 2017/1852 [2019] OJ L110/26. 120 Directive, art. 13.1. 121 Richelle, ‘Dans les arcanes de la nouvelle directive’, 913–914. 122 Directive, art. 15.2. 123 Directive, art. 15.3. 124 See Directive, art. 14.3.

718 Isabelle Richelle ‘the right to any other remedy’; at the level of arbitration, it must waive ‘the right to any domestic remedy’. The case does not necessarily end with the final decision. The directive allows member states to bring an action before a national court to challenge a ‘lack of independence’ within the meaning of article 8 of the directive.125

39.3.3.3 The role of the taxpayer One of the original features of the procedure introduced by Directive 2017/1852 is that it allows the affected person to request the initiation of the arbitration phase; in the 1990 Arbitration Convention, it was the tax authorities that initiated the arbitration. However, as we have seen, the directive gives the taxpayer several levers enabling it to influence the course of the procedure, whether at the stage of examining the admissibility of the complaint, the failure to set up the Advisory Commission without delay, or by using various appeals before the national courts or the Advisory Commission, depending on the situation. These appeal procedures, while indispensable to the taxpayer, are unnecessarily complex to implement.126.

39.3.3.3.1 Denying access to the dispute-resolution procedure Denial of access to the dispute-resolution procedure is allowed, first, when the affected person has committed tax fraud or serious misconduct and, secondly, on a case-by-case basis where the question in dispute does not involve double taxation. Access to the dispute-resolution procedure may be denied to a taxpayer on whom a member state has imposed penalties in ‘relation to the adjusted income or capital for tax fraud, willful default and gross negligence’.127 The directive does not define these terms, which should then be interpreted by reference to national law; however, one shall bear in mind that those terms are mentioned in the Revised Code of Conduct under the Arbitration Convention which could serve as a reference point.128 Tax fraud or serious misconduct must be related to the income or capital in dispute; the text of the directive does not allow the procedure to be refused for past misconduct. Since article 16.6 of the directive refers to article 6, and considering the strict interpretation of the texts, it must be concluded that the exclusion should not be assessed at the admissibility stage of the complaint by the tax authorities or the bilateral negotiations, but instead at the stage of the opinion of the Advisory Commission—be it at the time of deliberating on the admissibility of the complaint or when delivering its opinion.129 125

See Directive, art. 15.4. See Richelle, ‘Dans les arcanes de la nouvelle directive’, 917. 127 Directive, art. 16.6. 128 See the Revised Code of Conduct for the effective implementation of the Convention on the elimination of double taxation in connection with the adjustment of profits of associated enterprises [2009] OJC322/1. 129 Richelle, ‘Dans les arcanes de la nouvelle directive’, 920. Contra: Pit, ‘Dispute Resolution in the European Union’, 1333, rightly arguing that the reference to art. 6 alone is illogical and concluding, beyond the strict text of the directive, that the logic implies an extension of the exclusion to the whole procedure. 126

Alternative Dispute Resolution in the European Union 719 However, it is not clear whether or not the penalty must be final.130 The member state that imposed a sanction may refuse access to the dispute-settlement procedure. Furthermore, if national judicial or administrative proceedings have been commenced that could potentially lead to such penalties, the dispute-settlement procedure may be suspended at the request of a competent authority. Article 16.6 is a source of uncertainty for the taxpayer, because its terms will be open to interpretation; in practical terms, the proceedings would be halted, with no possibility of being resumed if the penalty were not confirmed. On a case-by-case basis, a member state is also allowed to refuse access to the dispute- settlement procedure referred to in article 6 if a dispute does not relate to double taxation.131 This exclusion was added during the negotiations within the Council and reveals the reluctance of member states towards arbitration. It would then remain for the affected person to refer its case to the national courts in order to obtain clarification on the interpretation or application of the convention or agreement or, where still possible, to use other treaty procedures.

39.3.3.3.2 Interaction with national proceedings As a rule, using domestic remedies is not an obstacle to the implementation of the dispute-resolution procedure.132 However, a national decision may be issued from which the member state may not derogate. In such a situation, the directive offers member states the possibility of providing for an obligation to inform and to terminate the dispute-resolution procedure be it at the MAP level, the request for setting-up the Advisory Commission, or before delivery of its opinion.133

39.3.3.3.3 Interpretation of the directive According to article 2 paragraph 2 of the directive, terms not defined must be interpreted in a dynamic view by reference to their meaning in the agreement or convention concerned and, in the absence of precision in those texts, by reference to national legislation; that is, to the national tax legislation relating to the taxes to which the agreement or convention applies, any meaning under the applicable tax laws of that Member State prevailing over a meaning given to the term under other laws of that member state. Such reference to domestic legislation might appear as limiting the interpretative competence of the CJEU; however, as the vocabulary of the directive and that of the conventions, especially the double taxation conventions, is essentially different, this reference to national law should, in practice, be limited in scope.134 While this interpretative provision is directly inspired by similar provisions usually found in double taxation treaties and now also in the MLI, it is questionable here

130

Pit, ibid., 1330–1335. Directive, art. 16.7. 132 Directive, art. 16. 133 Directive, art. 16.4. 134 Richelle, ‘Dans les arcanes de la nouvelle directive’, 894. 131

720 Isabelle Richelle whether the procedure used to interpret not an international agreement but a EU directive, which essentially establishes procedural rules, is the right one. In Austria v. Germany, the Court stated that an interpretative rule referring to the meaning given by the tax law of one of the contracting states ‘cannot be regarded as a rule intended to resolve differences of interpretation between the two States parties’.135

39.3.3.3.4 Varia Decision making The Advisory Commission must deliver an opinion on how to resolve the dispute, based on the provisions of the relevant agreement or convention and any applicable national rules. As this opinion is intended to help tax authorities to reach an agreement, it is reasonable to assume that it should be substantiated. The ‘last offer’ method is not applicable in this case—which differentiates this process from the one set up in a purely treaty-based framework; however, the competent authorities may agree on alternative dispute resolution through an Alternative Dispute Resolution Commission that may ‘apply, where appropriate, any dispute resolution processes or technique’.136 Alternative Dispute Resolution Commission As an alternative to the Advisory Commission, the directive offers the possibility of setting up an Alternative Dispute Resolution Commission (ADRC). The ADRC may be different in form and in its composition (except as regards the independent experts) and, with the member states’ agreement, may become permanent. Practical details for implementing the ADRC are not yet defined. The main difference between the ADRC and the Advisory Commission is the possibility for the former to use the ‘last best offer’ method. The issues at stake are significant and are clearly set out and assessed in a Working Paper published in August 2019 by the Fiscalis Project Group.137 While an ‘independent opinion’ should ‘support its conclusion taken with appropriate arguments’, ‘be reasoned’, and ‘presented in writing’, which can be more time-consuming, complex, and costly, ‘final offer arbitration’ allows, for example, ‘limiting the length of submissions’, ‘issuing opinions without the need to provide supporting rationales’, and allowing opinions not to be published.138 From the taxpayers’ point of view, the arbitration phase offers more guarantees and transparency than previous practices; ‘base-ball arbitration’ would reduce the transparency of the process and, therefore, the confidence they have in the procedure,139 while the costs might also be lower. 135

Case C-648/15, ECLI:EU:C:2017:664, paras 35–36. Directive, Aat. 10.2; Pistone, ‘The Settlement of Cross-Border Tax Disputes in the European Union’, 367; Owens, ‘Mandatory Tax Arbitration: The Next Frontier Issue’, 614. 137 Fiscalis Project Group (FPG) 093, ‘Working Paper on the Implementation of Article 10 of Directive (EU) 2017/1852 on Tax Dispute Resolution Mechanisms in the European Union’ (2019), https://taxation- customs.ec.europa.eu/system/files/2019-10/2019-tax-dispute-resolution-fiscalis-project-group-report. pdf; see also A. P. Kothat, ‘Baseball Arbitration Option under the EU Dispute Resolution Directive: How Well Does It Fare against the Objectives?’, World Tax Journal (2021), 253–281. 138 See points 6.2 and 6.3. of the Fiscalis Project Group’s Working Paper. 139 Avery-Jones, ‘Types of Arbitration Procedures’, 674. 136

Alternative Dispute Resolution in the European Union 721 Relationship with proceedings under other instruments Article 16 of the directive, relating to the ‘interaction of national procedures and derogations’, states that the introduction of the directive’s procedure, through the lodging of a complaint, has the effect of putting an end to ‘any other ongoing mutual agreement procedure or dispute settlement procedure under an agreement or convention giving rise to an interpretation or application in the context of the dispute in question’. The termination of another ongoing procedure takes effect from the date on which the complaint is first received by one of the competent authorities of the member states concerned. Special rules for individuals and smaller companies As can be seen, the procedure set up by the directive is quite complex and costly for the affected taxpayers. For this reason, a number of mitigating measures have been put in place in favour of individuals and smaller companies. Probably the most important of these is the right to lodge a complaint only with the tax authority of the member state of residence which will then act as a type of interface with the authorities of the other member state concerned. It will forward the complaint and any additional information received from the taxpayer, and so on. Thus, the complaint, replies to any request for additional information, withdrawals, and referrals to the advisory commission need only be addressed to one single state. It can be deduced from the structure of the directive that these communications are to be made in a language accepted by the state of residence and that the person concerned does not have to supply a translation.140

39.3.3.4 EU Tax Arbitration and Achmea The Achmea case law of the Court of Justice, as we have seen, condemns arbitration clauses in BITs. Could this case law be transposed to the arbitration bodies under the Arbitration Directive141? The issue is discussed in doctrine, with several authors considering that the directive shows weaknesses. Achmea is founded on the need to ensure a ‘unity of interpretation of this law, thereby ensuring its coherence, full effect and autonomy’;142 therefore, the Court could not accept that an arbitral body which may be required to hear disputes that could affect the interpretation or application of Union law,143 could not refer a question for a preliminary ruling because of its lack of

140 CFE Fiscal Committee, ‘Opinion Statement FC 4/2017 on the Proposed Directive on Double Taxation Dispute Resolution Mechanisms in the European Union’, European Taxation 57/8 (2017); Richelle, ‘Dans les arcanes de la nouvelle directive’, 849. 141 See for that discussion, e.g., Piotrowski et al., ‘Towards a Standing Committee Pursuant to Article 10’, 682–684; Monsenego, ‘Does the Achmea Case Prevent the Resolution of Tax Treaty Disputes through Arbitration?’, 733–735; K. Perrou, ‘Taxpayer Rights and Taxpayer Participation in Procedures Under the Dispute Resolution Directive’, Intertax 47 (2019), 715, 719–723. See also on this discussion, Piotrowski et al., ‘Towards a Standing Committee Pursuant to Article 10’; Monsenego, ‘Does the Achmea Case Prevent the Resolution of Tax Treaty Disputes through Arbitration?’, 733–735; Perrou, ‘Taxpayers Rights and Taxpayer Participation in Procedures under the Dispute Resolution Directive’, 719–723. 142 Achmea, paras 34–37. 143 Ibid., para. 39.

722 Isabelle Richelle independence,144 when, on the other hand, the arbitral award cannot be subject to review by a court or tribunal of a member state, ‘ensuring that questions of Union law which that court or tribunal might have to deal with can possibly be referred to the Court for a preliminary ruling’.145 This is reinforced by the subsequent case law of the Court on the matter.146 Under the Arbitration Directive, one cannot speak of functional independence as regards the amicable discussion phase147 or the final decision stage between tax authorities;148 in these cases, domestic remedies might still be open or the taxpayer may not accept the decision of the tax authorities. The question is more delicate with regard to the Advisory Commission, that will undoubtedly be called upon, at some point, to interpret or apply EU law and, in the first instance, Directive 2017/1852 itself. It may also be called upon to assess a treaty provision in the light of primary EU law. The presence of members of the member states’ administration with significant powers casts doubt on the required independence of the Commission. The fact that the Advisory Commission would be required to apply the principles of article 47 of the Charter of Fundamental Rights does not appear to be sufficient to consider the body as being independent. Two points appear to be problematic. First, the national tax authorities are involved in the decision-making process. This could be remedied by withdrawing that power from them, while ensuring their presence in the Commission, which is just as essential as that of the taxpayer who must be able to present its point of view. The required independence does not seem to be achieved by a majority decision of the independent experts, who could then be led to unite against an unshakeable position of a state, whereas their independence would naturally lead them to adopt a more nuanced position. On the other hand, the composition of the Advisory Commission is renewed for each arbitration— and, as such, it lacks any organic link to the jurisdictional system of the member states. Thus, it can be concluded that the Advisory Commission does not exhibit the required characteristics to qualify as a ‘jurisdiction’ under EU law. Furthermore, the possibility that a national court could be seized, after the Advisory Commission has given its opinion, to refer a question to the Court for a preliminary ruling is doubtful. The arbitration phase of the directive could therefore prove problematic.

144 See,

e.g., Case C-196/09 Miles, ECLI:EU:C:2011/388, para. 37, where the Court refers to the legal origin of the body, its permanence, the compulsory nature of its jurisdiction, the adversarial nature of the procedure, the body’s application of the rules of law, and its independence. 145 Achmea, para. 50. 146 See point 1.2. 147 Case C-24/92 Corbiau, ECLI:EU:C:1993:118, para. 15; D. de Carolis, ‘The EU Dispute Resolution Directive (2017/1852) and Fair Trial Protection under Article 47 of the EU Charter of Fundamental Rights’, European Taxation 58/11 (2018), 495 at 500. 148 Corbiau, ibid., para. 15; de Carolis, ‘The EU Dispute Resolution Directive (2017/1852) and Fair Trial Protection’, 500.

Alternative Dispute Resolution in the European Union 723 The same elements are transposable to the ADRC, especially since it would be composed of members of tax administrations with decision-making powers. It should also be borne in mind that the Advisory Commission is also called upon to rule on the admissibility of the complaint and that, in this context, it could be called upon to consider the scope of certain provisions in the directive: here again, there are doubts as to the possibility for the affected person to apply to a national court which could refer to the Court of Justice for a preliminary ruling. The doctrine therefore suggests different ways of providing the Advisory Commission with the characteristics required by the Court’s case law: one could be, for example, to refer to the Portuguese legislation which introduced arbitration tribunals in tax matters, whose members vary according to the case, but which are headed by a stable body specialized in the management of arbitration proceedings—a system which has been validated by the Court of Justice149—or to the Danish example allowing arbitration tribunals to operate as domestic courts and refer a matter to the Court of Justice for a preliminary ruling.150 The directive lays down rules for a mechanism to resolve disputes between member states when those disputes arise from the interpretation and application of agreements and conventions that provide for the elimination of double taxation of income and, where applicable, capital. It also lays down the rights and obligations of affected persons when such disputes arise. For the purposes of the directive, the matter giving rise to such a dispute is referred to as a ‘question in dispute’.

39.3.3.5 EU dispute resolution and taxpayers’ rights The directive undoubtedly represents an important step forward in favour of taxpayers’ rights. First of all, the directive itself recognizes that ‘it also lays down the rights and obligations of the affected persons when such disputes arise’,151 which is new in international law. Recital 9 adds: ‘In particular, this Directive seeks to ensure full respect for the right to a fair trial and the freedom to conduct a business’. Under the directive, the taxpayer has the right to request arbitration; moreover, it has various possibilities to advance the procedure and thus force the states concerned to reach a solution. It also has the right to present its arguments before the Advisory Commission, and interaction is

149 See Case C-377/13 Ascendi Beiras Litoral e Alta, ECLI:EU:C:2014:1754, paras 25–26.; Vasques, ‘Tax Arbitration and VAT: The Portuguese Experience’, 267–271. See also as regards the Benelux Court, Case C-337/95 Parfums Christian Dior, ECLI:EU:C:1997:517, para. 21; Case C-196/09 Miles, ECLI:EU:C:2011:388, para. 40. 150 Denmark allows an arbitration tribunal to ask ordinary jurisdictions to refer a question for preliminary ruling to the ECJ on their behalf: see M. Broberg and N. Fenger, ‘Arbitration Cases and Preliminary References to the European Court of Justice—An Assessment of the Danish Solution’, Arbitration International 36/1 (Mar. 2020), 147–155 (available at https://doi.org/10.1093/arbint/aiaa003). See, e.g., Monsenego, ‘Does the Achmea Case Prevent the Resolution of Tax Treaty Disputes Through Arbitration?’; Pit suggests that art. 10.1 of the directive could be a valid legal basis (Pit, ‘The Changed Landscape', 750); Piotrowski et al., ‘Towards a Standing Committee Pursuant to Article 10’, 682–684. 151 Directive, art. 1.

724 Isabelle Richelle already organized at the MAP stage by the possibility of communicating additional information.152 In this sense, the directive marks a step forward compared to tax treaties. However, the directive does not expressly establish a truly adversary process, either at the mutual agreement stage or at the arbitration stage.153 Furthermore, exclusion cases are viewed as elements of uncertainty, for which affected persons would find no remedy before the domestic courts.154 The taxpayer finds additional protection in the general principles of law which are binding on the member states, in particular the principle of good administration or the right of access to documents and, more generally, the right of defence.155 The Court has, for example, specified that the ‘respect for the rights of the defence is a general principle of EU law’, a principle which is binding on the authorities of the member states ‘when they take decisions which come within the scope of EU law, even if the applicable EU legislation does not expressly provide for such a procedural requirement’.156 Similarly, the principle of good administration applies to tax administrations when implementing EU law.157 The directive itself incorporates these fundamental principles, which it recalls in recital 9: ‘This Directive respects the fundamental rights’. These same principles are now enshrined in the European Charter of Fundamental Rights which, however, is only binding on member states when they implement Union law.158 It is reasonable to assume that by applying the tax dispute-resolution procedure of the directive, member states are ‘implementing Union law’,159 and therefore should take into

152 For more details, see Perrou, ‘Taxpayer Rights and Taxpayer Participation in Procedures Under the Dispute Resolution Directive’, point 2.3; Perrou, ‘Taxpayer Participation in Tax Treaty Dispute Resolution’; J. Kokott, ‘European Union—Taxpayer’s Rights’; European Taxation 60/1 (2020), 3–7. 153 De Carolis, ‘The EU Dispute Resolution Directive (2017/1852) and Fair Trial Protection’, 501. 154 Pit, ‘The Changed Landscape’, 755. 155 It must be noted that arts 41 and 42 of the EU Charter are formally addressed to the European institutions only. 156 Case C-298/16, Ispas, ECLI:EU:C:2017:843, para. 26; Case C-349/07 Sopropé, ECLI:EU:C:2008:746, paras 36–38; Case C-276/12 Sabou, ECLI:EU:C:2013:678, para. 38; Case C- 189/ 18 Glencore, ECLI:EU:C:2019:861, para. 39. 157 Case C-446/18 Agrobet, ECLI:EU:C:2020:369, paras 42–44 (concerning a tax audit). 158 Charter, art. 51. See also the explanations relating to art. 51 of the Charter, 2007/C303/01 (14 Dec. 2017); Case C-617/10 Fransson, ECLI:EU:C:2013:105, paras 19–23; Case C-418/11 Texdata Software, ECLI:EU:C:2013:588, paras 72–73. See ECJ, ‘Field of application of the Charter of Fundamental Rights of the European Union’ (Dec. 2017), available at https://curia.europa.eu/jcms/upload/docs/applicat ion/pdf/2018-05/fiche_thematique_-_charte_-_en.pdf; See F. Picod, ‘Article 51—Champ d’application’, in F. Picod, C. Rizcallah, and S. Van Drooghenbroeck, Charte des droits fondamentaux de l’Union européenne, Commentaire article par article, 2nd ed. (Brussels: Bruylant, 2020), esp. 1230– 1239; I. gambbardella, ‘Application de la Charte des droits fondamentaux de l’Union européenne aux Etats membres: les critères de mise en œuvre du droit de l’Union comme obstacle à son effectivité’, Cahiers de Droit Fiscal Européen vol. 5 (2021), 241–286; A. Ward, ‘Article 51 Commentary’, in S. Peers et al., eds, The EU Charter of Fundamental Rights, A Commentary (Oxford, Hart Publishing, 2021). 159 This point of view appears to be confirmed by the Court in Case C- 400/ 10- PPU J.McB, ECLI:EU:C:2010:582, para. 52 in which a European regulation with procedural scope was viewed against the Charter.

Alternative Dispute Resolution in the European Union 725 account the substance of these general principles both when transposing the directive into their national laws and when applying it in practice. However, some consider that ‘implementing Union law’ should be understood as the implementation of a substantive rule of European law:160 in this case, the substantive rule would be found in the convention or agreement, which does not fall within the scope of European law; this would justify setting aside the general protective principles. However, in the Donnellan case, the Court stated that ‘the acts taken by Member States pursuant to the system of mutual assistance established by Directive 2010/24 must be in accordance with the fundamental rights of the European Union, which include the right to an effective remedy enshrined in Article 47 of the Charter’.161 Furthermore, in the Austria v. Germany case, the Court clarified— admittedly on the question of its competence—that a preventive double taxation convention has a link with European law ‘in the light of the beneficial effect of the mitigation of double taxation on the functioning of the internal market that the EU seeks to establish in accordance with article 3(3) TEU and article 26 TFEU’.162 Thus, the conclusion should be that fundamental rights are applicable to the procedure of Directive 2017/ 1852.163 But what are these rights to be respected? At the stage of negotiation between tax administrations, only the principles of good administration apply; these principles already require a balance to be struck between the various participants in the procedure. The right to a fair trial could only be invoked before the Advisory Commission if it were considered a real ‘jurisdiction’; but that recognition, as we have seen, is still uncertain; otherwise, the principles of good administration should apply. Finally, there are still some cases in the directive where the affected person does not have a remedy provided for in the directive and where it is not certain that a remedy can be found in national law, for example where a state refuses the benefit of the directive by virtue of the cases of exclusion provided for in article 16.6 and 7. In these cases, it seems to us that the right to an effective remedy can be invoked. 160

Fiscalis Project Group (FPG) 093, ‘Working Paper on the Implementation of Article 10 of Directive (EU) 2017/1852 on Tax Dispute Resolution Mechanisms in the European Union’, 9: The rights to effective remedy and to a fair trial are accessory rights that require the existence of an (initial) substantive right enshrined in EU law in order to apply. However, the substantive law that underlies these disputes does not fall within the scope of EU law as it refers to the tax agreement or convention that provides for the elimination of double taxation between the contracting Member States. See also J. Voje, ‘The Limits to the Participation of the Taxpayer in Tax Dispute Resolution Procedure under the Dispute Resolution Directive’, Intertax 48/2 (2020), 157–176. 161 Case C-34/17 Donnollan, ECLI:EU:C:2018:282, para. 45. 162 Case C-648/15 Austria v. Germany, ECLI:EU:C:2017:664, para. 26. The Court also quoted a Communication from the EU Commission: ‘the purpose and effect of the conclusion between two Member States of a convention avoiding double taxation is to eliminate or mitigate certain consequences resulting from the uncoordinated exercise of their powers of taxation, which is, by its nature, capable of restricting, discouraging or rendering less attractive the exercise of the freedoms of movement provided for in the TFEU’. 163 See also EU Commission Explanatory Memorandum, COM(2016) 686 final, 6.

726 Isabelle Richelle

39.3.3.6 Implementation of the directive and follow-up 39.3.3.6.1 Role of the European Commission The European Commission takes an active role in the smooth functioning of the dispute-resolution procedure. It is responsible for managing the list of independent persons and competent authorities,164 which is now available online. It is also responsible for taking the necessary implementing measures and has already adopted model ‘operating rules’.165 The Commission is additionally responsible for managing a ‘central repository’ in which final decisions on regulations will be archived and made available online.166 The Commission is assisted by a ‘Committee on Dispute Resolution’167 composed of representatives of the member states and which is responsible for examining any draft implementing act that the Commission may have to take; this Committee will assist the Commission, for example in drawing up standard forms for the communication of information relating to the decisions to be published. The Commission is still responsible for evaluating the process put in place; a first evaluation report is expected in 2024.

39.3.3.6.2 Publicity of decisions The directive provides that final decisions by competent authorities are to be published in full or in summary. Publication in full is by common agreement between the competent authorities, with the consent of each of the persons concerned.168 In the absence of agreement or consent, only a summary is published. Publication is under the control of the affected person, who receives the draft publication and has sixty days to request the competent authorities ‘not to publish any information which relates to a commercial, industrial or professional secret or to a commercial process, or which is contrary to public policy’.169

39.4 Conclusions The adoption of the Arbitration Convention in 1990 marked an important step forward in the process of better protection of the taxpayer against international double taxation.

164

Directive, art. 21. Directive, art. 11.3; Commission Implementing Regulation (EU) 2019/652 of 24 April 2019 laying down standard Rules of Functioning for the Advisory Commission or Alternative Dispute Resolution Commission and a standard form for the communication of information concerning publicity of the final decision in accordance with Council Directive (EU) 2017/1852 [2019] OJ L110/26. 166 Directive, art. 19.3. 167 Directive, art. 20. 168 Directive, art. 18.2. 169 Directive, art. 18.3, para. 2. 165

Alternative Dispute Resolution in the European Union 727 This Convention has undoubtedly made it possible to experiment with an arbitration mechanism in tax matters and has contributed to its acceptance by the states. This first step has made possible the recent advances represented by the adoption of the Dispute Resolution Directive and an arbitration procedure in the MLI. If the procedure in the directive that we have examined in this chapter is not perfect, it has the merit of existing. The directive can certainly be improved: specific gaps could be filled, certain procedural points could be simplified, the status of the Advisory Commission and the ADRC will have to be clarified or adapted, and we can no doubt hope that a body specifically dedicated to the management of these tax arbitration procedures will be set up.

Pa rt V I

SE L E C T E D I S SU E S OF C RO S S -B OR DE R I N DI R E C T TAX AT ION

Chapter 40

Aspe cts of Cro s s - B orde r VAT The EU Approach and Evolving Trends Roberto Scalia

40.1 Introduction: VAT in a Cross- Border Context The essential backbone of European value-added tax (VAT), whose history dates to the inception and initial steps of the European Communities,1 was created in the ten-year period between 19672 and 19773 and later amended in several parts. Such additions and amendments were recast in the currently binding Council Directive 2006/112 on the common system of value added tax (the VAT Directive) for the purpose of achieving ‘clarity and rationalisation’.4 This evolution of the VAT rules has also affected the underlying principles of the directive and, in particular for the purpose of our analysis, the so-called ‘place-of-supply’ (POS) rules.5 As the achievement of the ‘internal market’ would have been a chimera where member states’ turnover taxes would have ‘distort[ed] conditions of competition or hinder[ed] the free movement of goods and services’,6 VAT is intended to be applied only once and 1 Echoing the outcome of the ‘Tinberghen Report’ (ECC, Committee of Experts, Report on the problem raised by the different turnover tax systems applied within the Common Market (5 Mar. 1953). 2 The First and Second VAT Directives were both issued on 11 April 1967. 3 Sixth VAT Directive (Directive 77/388/EEC). 4 Incorporating all the First VAT Directive provisions ‘still applicable’ (Preamble (1)), without ‘material changes in . . . legislation’ existing at that time (Preamble (3)). 5 POS rules are governed by arts 31–39, whereas the POS of services is laid down in art 43–59c of the VAT Directive. 6 Preamble (4) to the Sixth VAT Directive.

732 Roberto Scalia in one EU member state only and POS rules are key to ensuring the proper allocation— and taxation—of taxable supplies7 and thus avoiding double taxation (and double non- taxation) on the same supplies.8 However, since POS rules lead to the potential coexistence of more than one ‘establishment’ (i.e. establishments that an enterprise has in more than one member state), it is worth focusing on the dichotomy existing between an enterprise’s ‘main establishment’ in one state and the ‘branches’ of that enterprise based in other states.9 Such ‘branches’, in VAT jargon, are known as ‘fixed establishments’. As regards services, when such dealings were still only a tiny part of overall cross- border trade, their taxation was widely left up to the member states. Mainly for reasons of tax compliance, the Sixth VAT Directive later shifted to taxation in the country where the supplier is established.10 The aim, however, was to uphold the ‘destination principle’, while avoiding interpretation difficulties.11 The increasing importance of ‘services’ led to a rethinking of such criteria, upholding the ‘destination principle’ favouring the place where the customer is established.12 The current VAT approach—and foreseeable developments—towards cross-border supplies and, more precisely, the ‘fixed establishment’ (FE) proxy, will be analysed in the light of the VAT: (1) grounding principles; (2) origin and historical evolution; and (3) case law of the Court of Justice of the European Union (CJEU). 7 To this extent, the objective of the POS rules is to avoid conflicts of jurisdiction that could result either in double taxation or in non-taxation as stated by the CJEU based on a teleological analysis (see G. Bizioli, ‘Comparative Analysis of the Causes of Double (Non-) Taxation in the Income and VAT/ GST Contexts’, in M. Lang et al., eds, Value Added Tax and Direct Taxation—Similarities and Differences (Alphen aan den Rijn: IBFD, 2009), 406). See CJEU, Cases C-327/94 Dudda (1996), para. 20; C-291/07 Kollektivavtalsstiftelsen TRR Trygghetsrådet (2008), para. 24; C-218/10 ADV Allround Vermittlungs AG, in liquidation, v. Finanzamt Hamburg-Bergedorf (2012), para. 27; C-605/12 Welmory sp. z o.o. v. Dyrektor Izby Skarbowej w Gdańsku (2014), para. 42; and C-547/18 Dong Yang Electronics sp. z o.o. v. Dyrektor Izby Administracji Skarbowej we Wrocławiu (2020), para. 25. 8 At the international level, the OECD International VAT/ GST Guidelines (2017) have set forth ‘a number of principles for the VAT treatment of the most common types of international transactions, focusing on trade in services and intangibles, with the aim of reducing the uncertainty and risks of double taxation and unintended non-taxation that result from inconsistencies in the application of VAT in a cross-border context’. K. Spies, Permanent Establishments in Value Added Tax—The Role of Establishments in International B2B Trade in Services under VAT/GST Law (Alphen aan den Rijn: IBFD, 2019), 146–148. 9 The CJEU, Case C-388/11 Le Crédit Lyonnais v. Ministre du Budget (2013), paras 33–34, held that ‘[s]‌ince . . . the fixed establishment . . . and the principal establishment situated in another Member State constitute a single taxable person subject to VAT . . . it follows that a taxpayer is subject, in addition to the system which applies in the State of its principal establishment, to as many national systems . . . as there are Member States in which it has fixed establishments’. See Spies, Permanent Establishments in Value Added Tax, 177ff. 10 B. Terra, Sales Taxation: The Case of Value Added Tax in the European Community (Norwell: Kluwer Law and Taxation Publishers, 1998), 77–78. 11 See M. Merkx, Establishments in European VAT (Alphen aan den Rijn: Wolters Kluwer, 2013), 25. 12 See R. F. van Brederode, System of General Sales Taxation—Theory, Policy and Practice (Alphen aan den Rijn: Kluwer Law International, 2009), 230 and F. Nellen et al., Fundamentals of EU VAT Law, 2nd ed. (Alphen aan den Rijn: Wolters Kluwer, 2020) para. 4.2.

Aspects of Cross-Border VAT: EU Approach and Evolving Trends 733

40.2 The ‘Fixed Establishment’: An Enduring Dilemma in EU VAT The notion of FE plays a crucial role in this respect and was first introduced as a proxy for the services POS rules13 in article 9(1) and (2)(e) and article 26 of the Sixth VAT Directive.14 Although there is very little documentation on the rationale behind such a major amendment,15 the FE proxy shall (at least in principle): (1) be consistent with the ‘neutrality’ principle, aiming to achieve the same treatment for supplies of goods/ services resulting in the same consumption;16 and (2) be consistent with the EU freedoms. Despite its central role in the EU VAT context, there is still widespread uncertainty surrounding the FE notion. The main points at stake when addressing the issue of FE in the VAT context are: (1) the definition and essential features of the FE notion; (2) the one true nature of the FE as a ‘taxable person’ in its own right; (3) the assessment of who provides/receives the service—whether the head office (HO) or the FE; (4) which part of the enterprise (HO or FE) shall pay VAT (and can, therefore, deduct input VAT); and (5) the impact of the parent–subsidiary relationship and VAT grouping regimes (involving either or both the HO and the FE). Since 2010,17 the general POS rules for services18 can be divided into two main groups19 according to the qualification of the client, such that: (1) in business-to-business (B2B) supplies, the POS shall be the place where the client has established their business, unless 13

See B. J. M. Terra and P. J. Wattel, European Tax Law, 4th ed. (Alphen aan den Rijn: Kluwer Law International, 2005), 345 and VAT Committee Guidelines, Guidelines resulting from meetings of the VAT Committee (up until 3 November 2020) (2020), 94. 14 The Second VAT Directive did not contain a similar provision upholding the place of consumption. 15 The scant justifications for such an amendment were mainly ‘reasons of simplicity and . . . the avoidance of difficulties of interpretation’, according to the Commission of the European Communities, Proposal for a Sixth VAT Directive on the harmonisation of legislation of Member States concerning turnover taxes, COM (73) 950, 11. 16 This entails providing for the same treatment of: (1) domestic and foreign subsidiaries (see S. Pfeiffer, VAT Grouping from a European Perspective (Alphen aan den Rijn: IBFD, 2015), 43–54) and (2) branches of foreign businesses vis-à-vis local businesses (see G.-J., van Norden, ‘The Allocation of Taxing Rights to Fixed Establishments in European VAT Legislation’, in H. van Arendonk et al., eds, VAT in an EU and International Perspective—Essays in honour of Hans Kogels (Amsterdam: IBFD, 2011), 35 and 40). 17 See Council Directive 2008/ 8/ EC whose primary aim was to levy VAT ‘where the actual consumption takes place’, as stressed by A. van Doesum et al., ‘The New Rules on the Place of Supply of Services in European VAT’, EC Tax Review 17/2 (2008),, 78. 18 Special articles take precedence over the general rules mentioned earlier. See O. Henkow, Financial Activities in European VAT (Alphen aan den Rijn: Kluwer Law International, 2008), 231. 19 Arts 44 and 45 VAT Directive provide for two residual rules: the places where the person has ‘his permanent address or usually resides’. In both cases, the POS shall be the state where the FE is ‘established’. An override rule applies as per art. 59a based on where the ‘use and enjoyment of the services takes place’.

734 Roberto Scalia such services are provided to ‘a fixed establishment of [the client] located in’ another member state,20 while, conversely, (2) in business-to-consumer (B2C) supplies, the POS shall be the place where the supplier has established their business, unless services are provided from ‘a fixed establishment of the supplier located in ‘another Member State’.21 The new rules introduced in 2010 also identify new VAT liability rules when a non- resident has an FE in one member state and makes taxable supplies within the territory of that member state. That foreign person shall be regarded ‘as a person which is not established’ in the latter member state if the establishment which the supplier has within the territory of that member state ‘does not intervene in that supply’.22

40.3 The CJEU’s Key Role in Shaping a VAT Concept of FE Despite the key role of the concept of FE,23 neither the Sixth VAT Directive nor the subsequent Recast VAT Directive provided a definition of FE and, lacking a definition, the only (binding) guidance was delivered by the CJEU.24 The need to ‘ensure the uniform application of rules relating to the place of taxable transactions’25 led to the adoption of Regulation 282/2011 that introduced26 some definitions27 of FE for VAT purposes for the very first time, and echoed, to a large extent, the CJEU case law.28 However, as will be discussed later, in the race toward clarity, Regulation 282/ 2011 is not a satisfactory point of arrival and a new ‘relay’ (agreement—negotiation— arrangement) with the CJEU seems to be unavoidable.

20

Art. 44, VAT Directive. 45, VAT Directive. See J. Beerepoot, ‘About VAT Registration and Fixed or Permanent Establishment’, in H. van Arendonk et al., VAT in an EU and International Perspective, 15. 22 Art. 192a, VAT Directive. 23 As noted by the VAT Committee, ‘Clarifications of the concept of fixed establishment’, Working Paper No. 791 (2014), para. 3.1: ‘under the current transitional VAT system, the concept of [FE] appears to be irrelevant for the purposes of determining the place of supply of goods’ with few derogations. 24 Among other authors, see B. Terra, The Place of Supply in European VAT (1998) 109ff. 25 Preamble (14) to Council Implementing Regulation (EU) No. 282/2011. 26 See, among others, M. Lamensch, ‘New Implementing Regulation 282/ 2011 for the 2006 VAT Directive’, EC Tax Review 20 (2011), 4. 27 Namely, two variants of the meanings of FE (S. Cornielje and P. Slegtenhorst, ‘The Unsettled Business of the Fixed Establishment in EU VAT’, Intertax 29 (2020), 6, para. 3.2) or no less than three meanings (according to Nellen et al., Fundamentals of EU VAT Law, para. 4.5.2.3). 28 CJEU, Case C-605/12 Welmory (2014), paras 44–45 underlined that Reg. 282/2011 ‘wished to clarify certain concepts necessary for determining criteria relating to the place of taxable transactions, while taking account of the relevant case-law of the Court [and] to that extent, even though that regulation was not yet in force at the material time, it should none the less be taken into account’. 21 Art.

Aspects of Cross-Border VAT: EU Approach and Evolving Trends 735

40.3.1 Tracing the Borders between HO and PE (Berkholz, Planzer Luxembourg, Faaborg, ARO Lease, Lease Plan, Welmory, FCE Bank) The first case decided by the Court on the topic of FE in the VAT context, the Berkholz case, dates back to 1985.29 Berkholz represents a milestone in CJEU case law concerning FEs for several reasons. First, in Berkholz, the Court outlined, in a crystal-clear manner, the hierarchical relationship existing among the various ‘establishments’ of a business, stressing that the FE (i.e. any establishment other than the main establishment) is a residual criterion since ‘the place where the supplier has established his business is a primary point of reference inasmuch as regard is to be had to another establishment from which the services are supplied only if the reference to the place where the supplier has established his business does not lead to a rational result or create a conflict with another member state’.30 It can be derived from Berkholz that the elements to be considered in assessing the place where the main establishment is are crucial when applying the POS rules. The Court dealt with this issue in 2007 in the Planzer Luxembourg31 case, outlining in great detail the essential features of a main ‘establishment’ and underlining that a ‘company’32 place of establishment is where ‘the essential decisions concerning the general management . . . are adopted and where the functions of its central administration are carried out’.33 The factors to be taken into account are, according to the Court, the place where the ‘registered office, the place of its central administration, the place where . . . directors meet and . . . where the general policy of that company is determined’.34 This statement has been recalled in subsequent CJEU case law and, lastly, borrowed by article 10 Regulation 282/2011 which repeats the findings of the Court35 holding that 29

CJEU, Case C-168/84 Gunter Berkholz v. Finanzamt Hamburg Mitte-Altstadt (1985). CJEU, Case C-168/84 Berkholz (1985), para. 17. The Court’s stance was reaffirmed in subsequent cases; see, e.g., Cases C-260/95 Commissioners of Customs and Excise v. DFDS A/S (1997), para. 19; C- 190/95 ARO Lease BV and Inspecteur van de Belastingdienst Grote Ondernemingen, Amsterdam (1997), para. 15; and C-547/18 Dong Yang (2020), para. 26. Among other authors, see I. Roxan, ‘Locating the Fixed Establishment in VAT’, British Tax Review (1998) 611; H. Stensgaard, ‘Nexus for Taxpayers in Respect of VAT v. Direct Tax Treaties’, in Lang et al., Value Added Tax and Direct Taxation—Similarities and Differences, 621ff and G.-J. van Norden, ‘The Allocation of Taxing Rights to Fixed Establishments in European VAT Legislation’, 40–41. 31 Case C-73/06 Planzer Luxembourg Sàrl v. Bundeszentralamt für Steuern (2007). As fairly stated by the Court, ‘the Court’s case-law . . . (see . . . Planzer Luxembourg . . .) . . . directly inspired the wording of Article 11 of the Implementing Regulation’ (Case C-605/12 Welmory (2014), para. 58). 32 Reference shall be extended to include any form of (also unincorporated) ‘business’ according to Merkx, Establishments in European VAT, 56. 33 Case C-73/06 Planzer Luxembourg (2007), para. 60. 34 Ibid., para. 61. 35 B. Terra and J. Kajus, A Guide to European VAT Directives: vol. 1. Introduction to European VAT (Amsterdam: IBDF, 2020), 649 and R. de La Feria, ‘On the Evolving VAT Concept of Fixed Establishment’, EC Tax Review 30/5/6 (2021), 203. 30

736 Roberto Scalia ‘the place where the functions of the business’ central management are carried out’36 and the ‘place where essential decisions . . . are taken’37 take priority over other factors listed in article 10(2) Regulation 282/2011 (where the registered office is located and where management meets).38 A second major point, developed in the Berkholz case, affects the features of an FE in terms of constituting elements. Namely, the Court clarified the contours of the FE concept, stressing that (for services to be supplied by an ‘establishment’) the FE, aside from a certain degree of ‘permanence’, shall have ‘a certain minimum size and both the human and technical resources necessary for the provision of the services’.39 This is an outcome of the presumption that any ‘business establishment . . . presupposes a certain degree of organization . . . an ordered structure comprising things and persons’40 and can be deemed to be a natural corollary of the hierarchy between ‘main’ and ‘other’ establishments discussed earlier. Since Berkholz (see, e.g., Faaborg-Gelting Lininen,41 Aro Lease,42 Lease Plan,43 Welmory,44 and Titanium).45 this statement has been the backbone of the case law developed by the CJEU,46 reaffirming that an FE shall have both technical and human47 means that, jointly, should enable the FE to supply the type of services that it ought to provide.48 36 Art. 10(1) Reg. 282/ 2011 which, according to Merkx, Establishments in European VAT, 55, has a narrower scope if compared to the CJEU principles enshrined in Planzer Luxembourg. 37 Art. 10(3) Reg. 282/2011. 38 Other factors, e.g. the place of residence of the main directors, the place where general meetings are held, the place where administrative and accounting documents are kept, and the place where the company’s financial, and particularly banking, transactions mainly take place, may also need to be taken into account. 39 Case C-168/84 Berkholz (1985), para. 18. Consistent, Case C- 231/94 Faaborg-Gelting Linien A/ S and Finanzamt Flensburg (1996), paras 16–17. Among others, see R. de la Feria, The EU VAT System and the Internal Market (Amsterdam: IBDF, 2009), 182–183 and Henkow, Financial Activities in European VAT, 238. 40 See Pfeiffer, VAT Grouping from a European Perspective, 209–210. 41 Case C-231/93 Faaborg (1996). 42 Case C-190/95 ARO Lease (1997), para. 15. 43 Case C-390/96 Lease Plan Luxembourg SA v. Belgian State (1998), para. 26. 44 Case C-605/12 Welmory (2014), para. 59. 45 Case C-931/19 Titanium Ltd v. Finanzamt Österreich (2021), para. 42 holding that an FE cannot exist as ‘a structure without its own staff ’. M. Grandinetti, ‘La stabile organizzazione ai fini IVA: gli strumenti giuridici di definizione preventiva in ambito europeo e nazionale’, in Diritto e processo tributario 3 (2021), 488–489 deems that the Court stance has been arguably influenced by the question of the national judge asking whether or not ‘the letting . . . of a property . . . which constitutes only a passive tolerance of an act or situation whereby the supplier authorises a third party to do something that the latter could not do without such authorisation (Duldungsleistung) . . . even without human resources, [shall] be regarded as a “fixed establishment” ’. 46 Case C-190/95 ARO Lease (1997), paras 15–16. 47 Be it ‘its own staff or a structure which has a sufficient degree of permanence’ as clarified by Case C- 390/96 Lease Plan (1998), para. 26. 48 As suggested by AG Mancini in his Opinion in Case C-168/84 Berkholz (1985) (analysed by Pfeiffer, VAT Grouping from a European Perspective, 209–210) merely ‘engaging in . . . hiring out or leasing’ in another state does not entail the existence of an FE (Case C-390/96 Lease Plan (1998), para. 29).

Aspects of Cross-Border VAT: EU Approach and Evolving Trends 737 The ‘Berkholz’ approach has been upheld in Regulation 282/2011 providing that an FE shall have (for the so-called ‘active FEs’ in B2C supplies) a ‘sufficient degree of permanence and a suitable structure in terms of human and technical resources to enable it’ either to ‘provide’49 or (for the so-called ‘passive FEs’ in B2B supplies) to ‘receive and use . . . for its own needs’50 goods and services.51 For this purpose, substantial elements of the ‘establishment’ receiving the supply shall be assessed, such as ‘the nature and use of the service provided’ or, lacking such elements, according to formal criteria (‘the contract, . . . order form . . . VAT . . . number . . . and whether the [FE] is paying for the service’)52 while other elements, such as the value of the purchased services or the intensity of the purchasing activity, should be, per se, irrelevant in order to establish the existence of an FE.53 In 2006, the Court first addressed the topic of HO–FE internal dealings, delivering the landmark judgment in the FCE Bank case54 that became a milestone in CJEU case law. The Court stated that a supply between an HO and its FE shall be: (1) based on a legal relationship and, in addition, (2) the FE shall carry out ‘independent economic activities’.55 With respect to the relevant case, the Court upheld the Opinion of Advocate General Léger, holding that the Italian branch could not constitute an FE in its own right based on the statement that the HO (i.e. the UK bank) and not the FE (the Italian branch) ‘[bore] the economic risks associated with carrying on the business of a credit institution’ and did not have any endowment of capital.56

49 Art.

11(2) Reg. 282/ 2011. See I. Lejeune et al., ‘Fixed Establishments and Permanent Establishments: The VAT and Direct Tax Concepts Are Drifting Further Apart’, International VAT Monitor 27/9/10 (2016), 309. 50 Art. 11(1) Reg. 282/2011. 51 VAT Committee, ‘Clarifications of the concept of fixed establishment’, Working Paper No. 791 (2014), para. 3.2. stressing that a passive FE is not necessarily an active FE in its own right (‘[a]‌structure that would have the necessary human and technical resources to receive and use services supplied to it for its own needs would not necessarily have sufficient resources to provide those services on its own’). The EU approach has also been upheld in other regional contexts (see H. Hull and R. Scalia, ‘GCC VAT: International Services’, International VAT Monitor 28/5/6 (2008), 102). 52 Art. 22(1) Reg. 282/2011. The provider can consider the place where the customer has established the business as a last resort method (art. 22(2) Reg. 282/2011). See E. Kristoffersson, ‘Head Office-Branch Transactions’, in M. Lang and I. Lejeune, eds, Improving VAT/GST—Designing a Simple and Fraud-Proof Tax System (Amsterdam: IBFD, 2014), 523. 53 VAT Committee, ‘Clarifications of the concept of fixed establishment’, Working Paper No. 791 (2014), para. 3.1. 54 Case C-210/04 Ministero dell’Economia e delle Finanze, Agenzia delle Entrate v. FCE Bank plc (2006). 55 Ibid., paras 34–35. See P. Rendahl, Cross-Border Consumption Taxation of Digital Supplies (2009), 268–269. 56 Case C-210/04 FCE Bank (2006), paras 36–37. In AG Léger, Opinion on Case C-210/04 FCE Bank (2006), para. 43 states that it was ‘hardly conceivable that within the same legal entity, an [FE] could enjoy sufficient autonomy to act on its own behalf and under its own responsibility’.

738 Roberto Scalia

40.3.2 FEs in a Parent–Subsidiary Scenario (DFDS, Aro Lease, Welmory, Dong Yang, Berlin Chemie) The CJEU has identified other important elements of the overall picture of the FE notion when dealing with the specific case of subsidiaries deemed to be FEs of the foreign parent entity. In 1997, the Court first clarified the borders of the notion of ‘independence’ of the establishment, highlighting that a subsidiary (despite being an autonomous legal entity) might also be considered an FE of its parent company (DFDS)57 and, secondly, added that a sufficient degree of permanence and adequate structure, in terms of human and technical resources, of the FE shall enable it to supply the services ‘on an independent basis’ (Aro Lease).58 In DFDS, no guidance was given as to whether the human and technical means provided for by the subsidiary in its state of establishment might constitute the existence of a passive FE (in that state) of the parent company, resident in another member state. However, this issue was addressed in 2014 in the Welmory case59 and more recently in the Dong Yang60 case in 2020. In Welmory, the Court observed that in order to assess whether or not the foreign entity has an FE in the subsidiary’s state, it is immaterial: (1) that two entities (parent and subsidiary) form an ‘economic whole’; and (2) that the parent makes taxable supplies in the subsidiary’s state through an FE (i.e. acts as an ‘active’ FE).61 The Court reaffirmed the need for the foreign enterprise (a Cypriot company) to have adequate human and technical means in the relevant state (Poland) in order to assess the existence of an FE in that state.62 57

Case C-260/95 DFDS (1997). The fact that ‘the premises of the . . . subsidiary, which has its own legal personality, belong to it and not to DFDS is not sufficient in itself to establish that the subsidiary is in fact independent from DFDS’ (para. 26). The Court later narrowed the scope of the DFDS doctrine pointing out that ‘in DFDS, the independence of the status of the subsidiary was disregarded in favour of the commercial reality only to ascertain which of the parent company and the subsidiary had actually carried out the active taxable transactions . . . and, subsequently, which was the Member State of taxation for those transactions’ (Joined Cases C-318/11 and C-319/11 Daimler AG, Widex A/S v. Skatteverket (2012), para. 49). 58 Case C-190/95 ARO Lease (1997), para. 15 and, in the same vein, Daimler and Widex, ibid., para. 34 holding that ‘an [FE] is an establishment so autonomous that goods and services could be put on the market from it’. For more on this topic, see R. de la Feria, The EU VAT System and the Internal Market (Amsterdam: IBFD, 2009), 184 and J. Schafner, How Fixed is a Permanent Establishment? (Alphen aan den Rijn: Kluwer Law International, 2013), 57–58. 59 Case C-605/12 Welmory (2014). 60 See Case C-547/18 Dong Yang (2020). 61 See Case C-605/12 Welmory (2014), para. 64 and M. L. Schippers and J. M. B. Boender, ‘VAT and Fixed Establishments: Mysteries Solved?’, Intertax 43/11 (2015), 715. 62 See B. Terra, ‘Internet and the Concept of ‘Fixed Establishment’ of the Recipient of a Supply of Services: Case C-605/12 Welmory’, in M. Lamensch et al., eds, Value Added Tax and the Digital Economy—The 2015 EU Rules and Broader Issues (Alphen aan den Rijn: Kluwer Law International, 2016), 114–115.

Aspects of Cross-Border VAT: EU Approach and Evolving Trends 739 In the Dong Yang63 case, the Court declared that the assessment of the existence of an FE cannot derive, per se, from the mere fact that a foreign parent entity has a subsidiary in the state. However, it is possible that a subsidiary constitutes the FE of its parent company according to ‘substantive conditions set out in Implementing Regulation No 282/ 2011 . . . which must be assessed in the light of “economic and commercial realities” ’.64 In view of the criteria set forth in article 22, Regulation 282/2011, and according to the Court, this analysis does not imply that the supplier ‘examine[s]‌contractual relationships between a company established in a non-Member State and its subsidiary established in [a member state] in order to determine whether the former has a fixed establishment in that [member state]’ because they cannot be asked to inquire into contractual relationships between the parent company and the subsidiary given that such information is, in principle, inaccessible to them.65 The Court dismissed the Advocate General’s stance that such a (parent/subsidiary) contractual relationship shall be taken into account in cases of abusive practices,66 leaving open the question as to whether—and to what extent—the parent–subsidiary contractual ties might have any impact on the assessment of the existence of an FE of the parent company. In April 2022, in the Berlin Chemie case,67 the Court held that the existence of a subsidiary does not imply that an FE does not exist and that such assessments are conditional upon an analysis of the economic and commercial realities.68 This being said, the Court moves a step further holding that the human or technical resources69 of the subsidiary might be relevant in assessing the parent’s (FE’s) own human or technical resources, since, on the one hand, making the ‘existence of an [FE] subject to the condition that the staff . . . be bound by an employment contract to the taxable person itself and that the material resources belong to it in its own right would amount . . . to a very restrictive application of the criterion set out by the wording of Article 11(1)’ and, on the other hand, such a narrow interpretation would create a high 63

See Case C-547/18 Dong Yang (2020), para. 33 holding that ‘the existence, in the territory of a Member State, of a fixed establishment of a company established in a non-Member State may not be inferred by a supplier of services from the mere fact that that company has a subsidiary there’ (emphasis added). 64 Case C-547/18 Dong Yang (2020), para. 33. While ‘reference to economic and commercial reality is an increasingly familiar building block of recent CJEU judgements’ (as held by S. Cornielje and P. Slegtenhorst, ‘The Unsettled Business of the Fixed Establishment in EU VAT’, EC Tax Review 29/6 (2020), 289) it is worth noting () that consideration of the economic and commercial realities in VAT is problematic to legal certainty. 65 Case C-547/18 Dong Yang (2020) paras 36–37. 66 AG Opinion, ibid., paras 47ff. 67 Case Case C-333/20 Berlin Chemie A. Menarini SRL (2022). 68 Ibid., para. 41. 69 Namely, the ‘German [parent] had access to technical resources owned by the Romanian [subsidiary], such as computers, operating systems and motor vehicles’ and, in addition, the German parent’s control over sales are ‘constant and substantial [and] [i]‌t also has the right to inspect the records and premises of the Romanian [subsidiary] pursuant to a contract between those two companies’ (Case C-333/20 Berlin Chemie A. Menarini SRL (2022), paras 12 and 19).

740 Roberto Scalia level of uncertainty with respect to the POS if ‘it were sufficient for a taxable person to cover its staffing and material needs by having recourse to various service providers’.70 However, the Court considers that a ‘passive FE’ of the parent cannot be deemed to exist where the subsidiary’s human or technical resources are employed to provide services that, vice versa, cannot be received by the same persons (‘the same means cannot be used both to provide and receive the same services’).71

40.3.3 VAT Treatment of FEs in VAT Group Articulations (Skandia, Danske Bank) The topic of FEs has also been developed in another particular context, the VAT grouping regime. In such a context, many new questions arise and, namely, whether: (1) the FE is a ‘taxable person’ in its own right that can join a VAT group; (2) the FE can be deemed a taxable person in combination with other entities forming a VAT group in the same state; and (3) an HO (or an FE) joined in a VAT group and its FE (or HO) constitute two ‘independent’ entities. In 2014, when deciding the Skandia case,72 the Court dealt with an intra-group charge of costs by the USA-based HO to its FE in Sweden that73 was part of a VAT group in Sweden. The Swedish Tax Administration registered the FE for VAT purposes (alongside the VAT group) and charged the relevant VAT. The Court, in principle, confirmed the statements held in FCE Bank holding that a legal relationship shall exist between the HO and FE for the supply to be subject to VAT and the FE shall therefore carry out ‘independent economic activity’ bearing ‘economic risk arising from its business’.74 In the Court’s opinion, the FE did not fulfil such prerequisites based on the assumption that: (1) it did not operate autonomously; and (2) it did not have its own capital or assets. However, due to the existence of a VAT group, the Court paid attention to the subjective element, stressing that, according to article 11 of the VAT Directive, the creation of a VAT group erases the formerly existing entities75 belonging to the 70

Ibid., para. 45. The Court’s decision largely corresponds with the stance taken by Grandinetti, ‘La stabile organizzazione ai fini IVA’, 488 (narrowing the scope of Titanium in the light of the question put forward by the national court). For an analysis of the Titanium case in light of the the EUCJ case-law see, also, F. Cannas, EU VAT Categories and the Digital Economy (Turin: G. Giappichelli ed, 2022) 268-274. 71 Case C-333/20 Berlin Chemie A. Menarini SRL (2022), para. 54. 72 Case C-7/13 Skandia (2014). 73 The peculiarity of this case is also relevant for the purposes of the Danske Bank case decided in 2021. 74 Case C-7/13 Skandia (2014), paras 24–25. See S. Pfeiffer and K. Spies, ‘CJEU Case Law on Taxable Persons and Related Issues in 2019’, in G. Kofler et al., eds, CJEU Recent Developments in Value Added Tax 2019 (Vienna: Linde, 2020), 138 stressing that in both the FCE Bank and Skandia cases, the Court did not focus on whether the FE had the right to organize itself. 75 Notably, the Court, departing from the Opinion of AG Wathelet (in Case C-7/13 Skandia (2014), paras 44–45) did not place much importance on the linguistic differences in the English and French

Aspects of Cross-Border VAT: EU Approach and Evolving Trends 741 group.76 This implies that supplies by the HO to its FE (merged into a VAT group) are deemed to be ‘taxable supplies’ and the recipient of such supplies would be the VAT group (rather than the FE).77 On 11 March 2021, the CJEU delivered new guidance on the FE topic with respect to the VAT grouping regime in the Danske Bank case.78 The case addressed by the Court was the still dubious ‘reverse’ of the Skandia case,79 where the FE in Sweden of its Danish HO (belonging to a Danish VAT group) was charged a share of the costs of the group’s computer platform and the referring court asked whether: (1) the VAT group could be deemed to be a ‘taxable person’ separate from the foreign FE; and (2) the services supplied to the FE should be regarded as supplies of services to the Swedish FE and taxed in Sweden accordingly. Rejecting the claimants’ arguments based on an analysis of Morgan Stanley and FCE Bank,80 the CJEU held that: first, it is immaterial that either the HO (as in the Danske Bank case) or the FE (in the Skandia case) is part of a VAT group;81 secondly, the existence of a VAT group in another state must be taken into account when assessing VAT liability for the branch established in the state territory; and, thirdly, transactions between the HO and its FE cannot be placed on an equal footing, in terms of the neutrality principle, with supplies between an FE and the HO to which it belongs.82 In the light of the above, the Court concluded that the HO (part of a VAT group) and its FE should be considered separate taxable persons.83 Moreover, the Court held that the Skandia doctrine should be extended in this case too,84 since the fact that a member state’s VAT grouping regime (in this case Denmark’s) does not allow foreign FEs to be

versions of arts 9 and 11 VAT Directive that would have led to a completely different result. See M. Lamensch, ‘Taxable Supplies’, in M. Lang et al., eds, CJEU—Recent Developments in Value Added Tax 2014 (Vienna: Linde, 2015), 99–101. 76 EU Commission, ‘Vat Grouping Communication’ (2009), para. 2: ‘according to the Explanatory Memorandum, the aim of the provision on VAT grouping is to allow Member States . . . to not regard as separate taxable persons those whose “independence” is purely a legal technicality’ and, to this end, ‘a number of closely bound taxable persons are merged into a new single taxable person for VAT purposes . . . Thus . . . the group member . . . dissolves itself from any possible, simultaneously existing legal form and instead becomes part of a new separate taxable person for VAT purposes—namely, the VAT group’ (para. 3.2, emphasis added). 77 To this extent, FCE Bank and Skandia are not two contrasting judgments, but rather have two different scopes. See S. Pfeiffer, ‘Taxable Persons: VAT Grouping and Fixed Establishments’, in Lang et al., CJEU—Recent Developments in Value Added Tax 2014, 80–81. 78 Case C-812/19 Danske Bank A/S, Danmark, Sverige Filial v. Skatterverket (2021), not yet published. 79 Pfeiffer, ‘Taxable Persons: VAT Grouping and Fixed Establishments’, 74–75. 80 Namely, for a taxable supply to exist, a ‘legal relationship’ must bind two parties and, to this end, ‘it is necessary to ascertain whether the branch performs an independent activity’ that entails bearing the ‘economic risk arising from its business’. See Case C-812/19 Danske Bank (2021) paras 20–21. 81 Ibid., para. 32. 82 Ibid., para. 26. 83 Ibid., para. 35. 84 Disregarding the facts that, in Skandia: (1) the HO was in a third state; and (2) the FE (and not the HO) was part of a VAT group.

742 Roberto Scalia included implies that other member states (Sweden) should consider the group as the sole ‘taxable person’. In addition to this, the Court reaffirmed that the principle of neutrality implies that the dealings between the HO and the FE cannot be placed on a level playing field if either the HO85 or the FE86 are part of a VAT group.

40.3.4 Further Thoughts on New Challenges for the VAT Treatment of FEs . . . The VAT Recast Directive Regulation 282/2011 and the CJEU case law have still not fully clarified the scope of rules related to FEs,87 thus presenting new challenges for the CJEU. One of the aspects that is still debatable is the scope and width of the ‘intervention’ criterion as per article 192a of the VAT Directive which provides that if the FE in the customer’s state does not ‘intervene’ in the supply, then the customer shall consider that the supply is made by the foreign HO.88 Conversely, the FE’s appropriate human and technical resources shall: (1) enable the FE to make the relevant supply; and (2) be used ‘for transactions inherent in the fulfilment of the taxable supply . . . before or during this fulfilment’89 that do not comprise administrative support, accounting, invoicing, collection of debt claims, etc.90 The relevant rule under article 192a of the VAT Directive, introduced with the aim of resolving all the problems related to the allocation of the supply to either one state or another when the enterprise has an FE in the state where VAT is due, is not entirely clear and, arguably, the CJEU will be asked to clarify it. Another provision that is still open to discussion is article 11 Regulation 282/2011 insofar as it requires that a passive FE receives services ‘for its own needs’. The different treatment of FEs compared to local establishments (whose liability is not subject to such a criterion) might lead one to conclude that the scope of Regulation 282/2011 exceeds the scope of the directive.91 85

In a Danske Bank-like case. In a Skandia-like case. 87 De la Feria, The EU VAT System and the Internal Market, 189, in agreement with Ben Terra, concludes that ‘[t]‌he Court’s jurisprudence in relation to [FE] . . . creates more questions than it resolves’. 88 See P. Schilling and D. Hogan, ‘Intervention— A Problematic New Concept in EU VAT Law’, International VAT Monitor 21 (2010), 187 and W. de Witt, ‘The Fixed Establishment After the VAT Package’, in H. van Arendonk et al., VAT in an EU and International Perspective, 26–27. 89 See art. 53(1) and (2) Reg. 282/2011. In ‘Follow-up to the VAT package’, Working Paper No. 605 (2009), the VAT Committee held that where the FE does intervene in the supply before or during its fulfilment, or under the agreement, it is envisaged that it may intervene subsequently and this potential intervention does not constitute a separate supply insofar as the use of its technical and/or human resources relating to that supply is irrelevant as it will always be regarded as intervening in the supply. 90 Art. 53(2) last indent Reg. 282/2011. J. Beerepoot, ‘About VAT Registration and Fixed or Permanent Establishment’, in H. van Arendonk et al., VAT in an EU and International Perspective, 17, wonders whether appointing a tax representative would amount to fulfilling such a prerequisite. 91 See the analysis by G.-J. van Norden, G-J, ‘The Allocation of Taxing Rights to Fixed Establishments in European VAT Legislation’, 46–47. 86

Aspects of Cross-Border VAT: EU Approach and Evolving Trends 743 Another point that remains vague and obscure is how to assess the tax base in HO– FE supplies. There could be room for arguing that—unless the tax base is the consideration,92 where it is existent and is not a symbolic one—the market value could be adopted as a suitable criterion.93 However, in the light of Skandia and Danske Bank, the separation between the HO and the FE raises the question, yet unanswered, on whether the intra-group charge of costs may be considered as self-supplies (as per art. 26 VAT Directive) or supplies for the benefit of the FE (whether or not the FE is part of a VAT group).94 The latter question unveils another controversial aspect affecting the VAT regulation of a group’s internal dealings and, namely, the interaction between the VAT grouping regime and cost-sharing arrangements under article 132(1)(f) VAT Directive. The uncoordinated combination of the two regimes may lead to double non-taxation95 in an HO–FE scenario.96 A further element, that may arise in the near future, is the relationship between the permanent establishment and the FE notions in direct taxes and VAT, respectively. Notably, the use of similar but different terms has always been an element of confusion in defining the borders between the two concepts and—despite the intuitive proximity of the two concepts based on similar terms (‘fixed’ and ‘permanent’)97—the FE concept shall be considered a pure notion of EU law98 that does not, in principle, overlap with the tax treaty notion of permanent establishment.99 Since Berkholz,100 it has been apparent that the FE concept has had to run parallel but on different rails to the (direct tax) ‘permanent establishment’ notion. 92 See R. Scalia, ‘Spunti ricostruttivi sul rilievo sistematico delle note di variazione IVA alla luce di una recente pronuncia della Corte di Cassazione’, Rivista di diritto tributario XXIX/4 (2019), 160 and the case law cited therein. 93 Art. 73 VAT Directive. 94 Pfeiffer, ‘Taxable Persons: VAT Grouping and Fixed Establishments’, 79–80, argues that if art. 26 VAT Directive does not apply, input VAT would need to be adjusted. 95 See, on the relevant topic, Pfeiffer, VAT Grouping from a European Perspective, 260–261. 96 The VAT grouping and the cost-sharing agreement regime, under VAT, have different purposes (see Terra and Kajus, A Guide to European VAT Directives, 930) and two different territorial scopes (as held by Pfeiffer, VAT Grouping from a European Perspective, 259) since the cost-sharing regime is not limited to the territory of one state while it ‘includes businesses with their seat of economic activity in the territory of the Member State implementing the VAT grouping scheme but does not include those fixed establishments which are situated abroad’. See EU Commission, VAT Grouping Communication (2009), para. 3.3.2.1. and M. Gabriël and H. van Kersten, ‘VAT Group’, in M. Lang and I. Lejeune, eds, Improving VAT/GST—Designing a Simple and Fraud-Proof Tax System (Amsterdam: IBFD, 2014), 505. Notably, the EU Commission, VAT Grouping Communication (2009), para. 2 explicitly avoided dealing with it. 97 B. Larking, ‘The Importance of Being Permanent’, Bulletin for International Fiscal Documentation 52/6 (1998), 267. 98 P. Pistone, ‘Fixed Establishment and Permanent Establishment’, International VAT Monitor 10/3 (1999), 101 and Spies, Permanent Establishments in Value Added Tax, 190–191. 99 As Schafner, How Fixed is a Permanent Establishment?, 58 puts it ‘[a]‌ lthough there are similarities . . . the OECD concept is . . . much broader, so that the OECD guidelines may . . . have an impact on future interpretation . . . in the EU, but the opposite certainly does not hold true’. 100 See, e.g., Case C- 210/04 FCE Bank (2006), para. 39 affirming that ‘the OECD Convention is irrelevant since it concerns direct taxation while VAT is an indirect tax’.

744 Roberto Scalia One may wonder whether this is still the case in the light of the developments of the Base Erosion and Profit Shifting (BEPS) Project leading to the adoption of a new nexus for income taxes based on a lower threshold than in the past.101 The opinion that the adoption of new proxies in the field of direct taxation may also affect the VAT context102 may be read between the lines of the recent guidelines delivered in respect of the so-called ‘Quick Fixes’ (namely, the simplification for call-off stock warehouses as per art. 17a VAT Directive) which hold that where ‘the warehouse . . . is owned (or rented) and run directly by the [foreign] supplier with his own means present in the Member State where the warehouse is located, this warehouse shall be seen as his fixed establishment’.103 Interpreting the (VAT) FE notion vis-à-vis the (direct tax) PE concept could also be a gateway for deepening the discussion around the true scope of article 24(3) of double tax treaties104 that forbid the non-discrimination of other states’ permanent establishments.105

40.3.4.1 . . . with special regard to selected issues concerning the digitalization of society Since the last quarter of the twentieth century, the ceaseless technological evolution of society and the economy has posed new and challenging questions both at the international level106 and the EU level.107 101 See W. Hellerstein, ‘Exploring the Potential Linkages Between Income Taxes and VAT in a Digital Global Economy’, in Lang and Lejeune, VAT/GST in a Global Digital Economy, 96 and Nellen et al., Fundamentals of EU VAT Law, para. 4.5.2.3, holding that ‘there still is considerable uncertainty as to the interpretation of . . . “sufficient degree of permanence”, “suitable structure” and “human and technical resources”—especially when situations involve a high extent of digitalization and intangibility ... It is interesting to see that [BEPS] also addresses the concept of permanent (fixed) establishment in consumption taxes, clearly in an attempt to explore a proper mode of application in the context of the new economy’. For a concurring opinion, see Grandinetti, ‘La stabile organizzazione ai fini IVA’, 484–485. For the opposite perspective, see Spies, Permanent Establishments in Value Added Tax. 102 I. Lejeune and S. Claessen, ‘Conclusions: The Future of VAT in a Digital Global Economy— Innovation versus Taxation’, in Lang and Lejeune, VAT/GST in a Global Digital Economy, 206. 103 VAT Committee, ‘Guidelines resulting from the 13th Meeting’, Working Paper No. 974 (2019), para. 4. New cases decided by the CJEU (e.g. Berlin Chemie, dealt with in Section 40.3.2) may reinforce the discussion surrounding the relationship existing between the direct and the indirect tax concepts of PE and FE as stressed by J. Schwartz, ‘VAT Fixed establishment =permanent establishment? or, should direct and indirect tax practitioners talk to each other?’, Kluwer International Tax Blog (6 June 2022), http://kluwertaxblog.com/2022/06/06/vat-fixed-establishment-permanent-establishment-or-should- direct-and-indirect-tax-practitioners-talk-to-each-other/. 104 Whose scope is widened beyond art. 2. 105 Conditional upon: (1) the ‘establishment’ being a ‘permanent establishment’ for tax treaty purposes; and that (2) the two notions of ‘enterprise’ for treaty and VAT purposes overlap. See T. Ecker, A VAT/GST Model Convention (Amsterdam: IBDF, 2013), 396–400. 106 This process has affected, in the first place, e-commerce and a set of principles laid down in 1998 (see OECD, Electronic Commerce: Taxation Framework Conditions (Paris: OECD Publishing, 1998), 4) to uphold the ‘destination principle’ (see OECD, Taxation and Electronic Commerce: Implementing the Ottawa Taxation Framework Conditions (Paris: OECD Publishing, 2001), 24ff). 107 See, among others, EU Commission, Communication on the future of VAT—Towards a simpler, more robust and efficient VAT system tailored to the single market (2011), para. 4.1.

Aspects of Cross-Border VAT: EU Approach and Evolving Trends 745 The most relevant topic (see the comment on the Titanium case earlier) will be the role— if any—that the ‘human’ element of the FE will play in the future with respect to digitalized businesses.108 Some of the many other diverse topics that have recently emerged deserve careful attention. One can consider, for example, VAT issues related to crypto-assets, a wide category that also comprises crypto-currencies (CCs),109 whose functioning is based on so-called ‘smart-contracts’. VAT regulation is strictly intertwined with the ‘agreements’ according to which a supply is provided for, but such ‘contracts’ hardly fit into the traditional notion of a contract.110 Therefore, it is likely that interpretative issues will arise111 around ‘source codes’ (not contracts), which will arguably have an immediate impact on the identification of the establishment (either HO or FE) for the benefit of which supplies are carried out. It could be controversial for platforms if an FE can exist112 and the deeming provision as per article 9a Regulation 282/ 2011113 for the so- called

108

Splitting the human and the technical resources (through the conjunction ‘or’ instead of ‘and’) in Case C-333/20 Berlin Chemie A. Menarini SRL (2022), paras 37 and 41, reveals the evolution of the CJEU case law approach that is the outcome of developments in the subsequent decades after Berkholz, which have seen the progressive digitalization of the society and of the economy, which have gradually eroded the presumption of physicality on which the FE concept was based (in this vein, see for an analysis and some future perspectives, see also F. Cannas, ‘L’istituto della stabile organizzazione ai fini iva e la rilevanza della sua dotazione di “mezzi umani (e tecnici)” ’, Diritto pratica tributaria internazionale, XVIII/4(2021), 1741.). 109 Although both the VAT Committee and the CJEU have delivered some guidance on the VAT treatment of CCs, several unexplored issues remain obscure. See VAT Committee: (1) ‘Question Concerning the Application of EU VAT Provisions’, WP No. 811 (2014); (2) ‘Question Concerning the Application of EU VAT Provisions’, WP No. 854 (2015) delivered after the Hedqvist case; and (3) ‘Case- Law—Issues Arising from Recent Judgement of the Court of Justice of the European Union’, WP No. 854, (2016). For the specific issue of the treatment of so-called ‘utility tokens’ under the new VAT vouchers regime, see VAT Committee, ‘New Legislation—Matters Concerning the Implementation of Recently Adopted EU VAT Provisions’, Working Paper No. 983 (2019) and Case C-264/14 Skatteverket v. David Hedqvist (2015). 110 See, among other tax scholars, M. Merkx, ‘VAT and Blockchain: Challenges and Opportunities Ahead’, EC Tax Review 28/2 (2019), para. 3 and M. Tumpel and J. Kofler, ‘Tax Treatment of Digital Currencies’, in W. Haslehner et al., eds, Tax and the Digital Economy (Alphen aan den Rijn: Kluwer Law International, 2019), 183. 111 Also, those who consider that the hiatus between programmers and legal scholars can be reduced or eliminated (A. Bal, Taxation, Virtual Currency and Blockchain (Alphen aan den Rijn: Kluwer Law International, 2019), 19) must admit that, for this purpose, lawyers should write contracts in ‘clear and unambiguous language’. For a dissenting opinion, see R. Scalia, ‘Riflessioni su alcuni temi controversi sulla disciplina IVA delle c.d. criptovalute’, Giurisprudenza delle imposte XCIII (2020), 1, 5–7. 112 M. Lamensch, ‘The Scope of the EU VAT System: Traditional & Digital Economy Related Questions’, 124 and N. Jovanovic and M. Merkx, ‘Welmory: A Recipe for VAT Avoidance?’, Intertax 24/ 4 (2015), para. 4. 113 See M. Lamensch, ‘Taxing Digital Remote Supplies’, in Haslehner et al., Tax and the Digital Economy, 203–206 and A. Cockfield, W. Hellerstein, and M. Lamensh, Taxing Global Digital Commerce (Alphen aan den Rijn: Kluwer Law International, 2020), 288ff.

746 Roberto Scalia ‘exchange’114 and platforms operating in initial coin offering/initial token offering cases could be applied.115 Furthermore, one cannot help but consider that the classical bipartition between goods116 and services is becoming more blurred than in the past117 and may also raise challenging questions regarding the POS rules (based upon such bipartition),118 while the calculation of the VAT tax base is still far from clear.119

114 T.

Ehrkle Rabel and L. Zechner, ‘VAT Treatment of Cryptocurrencies Intermediation Services’ Intertax 48/5 (2020), paras 3.2.3.2 and 3.2.3.3. An interesting aspect is the qualification of the exchange of in-game money for legal tender if the in-game money is (virtual) remuneration for renting (virtual) land. See BFH, Urteil v. 18 November 2021. 115 R. Scalia, ‘Riflessioni su alcuni temi controversi sulla disciplina IVA delle c.d. criptovalute’, 51–55. It is worth to consider, in this respect, the stance taken by the EUCJ in C-695/20 Fenix International Ltd [2023] concerning the scope of Article 28 VAT Directive and Article 9a Reg. 282/2011. 116 Goods, in turn, could be divided into tangible and intangible goods. 117 See M. Pierro, Beni e servizi nel diritto tributario (Padua: Cedam, 2003), 291–293; Bal, Taxation, Virtual Currency and Blockchain, 199; Cockfield, Hellerstein, and Lamensh, Taxing Global Digital Commerce, 80; M. Lamensch, ‘The Treatment of “Digital Products” and Other “E-Services” under VAT’, in Lang and Lejeune, VAT/GST in a Global Digital Economy, 15–17; M. Lamensch, European Value Added Tax in the Digital Era (Amsterdam: IBDF, 2015), para. 4.2. (who underlines that ‘the characterization of online supplies as “electronically supplied services” . . . is artificial and the definition . . . is moreover too narrow to correctly tackle all the supplies . . . the primary classification of all online supplies as “services” would remain artificial’) and Scalia, ‘Riflessioni su alcuni temi controversi sulla disciplina IVA delle c.d. criptovalute’, para. 3.1. 118 J. Kollmann, Taxable Supplies and Their Consideration in European VAT—With Selected Examples of the Digital Economy (Alphen aan den Rijn: IBFD, 2019), 35–36 and M. Lamensch, ‘The Scope of the EU VAT System: Traditional & Digital Economy Related Questions’, 124. 119 If one accepts that CCs can be converted into euros (as per arts 230 and 91(2) VAT Directive), the problem would be to highlight: (1) the most representative exchange market and exchange rate (Scalia, ‘Riflessioni su alcuni temi controversi sulla disciplina IVA delle c.d. criptovalute’, 38–39); and (2) the corresponding value according to the moment at which the supply is deemed as being supplied which, due to the rapid and huge fluctuations of CCs, is a fairly crucial element (see M. Lamensch, ‘The Scope of the EU VAT System: Traditional & Digital Economy related Questions’, in M. Lang et al., eds, CJEU Recent Developments in Value Added Tax 2017 (Vienna: Linde, 2018), 130 and R. Scalia, ‘Le critto-attività nel diritto tributario europeo e internazionale’, in G. Cassano et al., eds, Il diritto di internet nell’era digitale (Milan: Giuffré, 2020), 1147).

Chapter 41

Taxation of I mp orts Thomas Bieber

41.1 Introduction The taxation of imports is a triad of customs duties, excise duties and import VAT. In this chapter, I will focus on import VAT. As imports of goods are governed by the common framework of the General Agreement on Tariffs and Trade (GATT), I will first give a short outline of the principles of the GATT regarding their VAT content (Section 41.2). Secondly, from an EU law perspective, I will give an overview of the principles of the Treaty on the Functioning of the European Union (TFEU) (Section 41.3) and will have a more detailed look at the taxation of imports according to the VAT Directive 2006/112/EC (Section 41.4). This article will not deal with the import VAT systems of individual or developing countries.1

41.2 Principles of the GATT on the Taxation of Imports The GATT allows the taxation of imports, provided that taxation does not have the effect of a customs duty or a duty equivalent to a customs duty and the import tax is part 1 See

R. Bird and P. P. Gendron, The VAT in Developing and Transitional Countries (Cambridge: Cambridge University Press, 2007). For developing countries, the taxation of imports plays an important fiscal role due to less developed administrative structures, see M. Keen, ‘VAT, Tariffs, and Withholding: Border Taxes and Informality in Developing Countries’, Journal of Public Economics 92 (2008), 1894 according to which sales tax revenue from the importation of goods in thirty-four selected developing countries ranged from 35.5% to 83% of total sales tax revenue for the years 2004 and 2005. E.g. sales tax revenue from imports was 35.5% of total sales tax revenue in Zambia, 62.2% in Ethiopia, and 83% in Guinea. However, it should be noted that no reliable data on the amount of sales tax refunds for these developing countries were available.

748 Thomas Bieber of the internal VAT system. If imports are taxed higher than the goods supplied on the domestic market, importers can rely on the principle of national treatment according to article III of the GATT. Imported goods shall not be subject, directly or indirectly, to internal taxes or other internal charges of any kind in excess of those applied, directly or indirectly, to like domestic products. Moreover, no contracting party shall otherwise apply internal taxes or other internal charges to imported or domestic products in a manner contrary to the principles set forth in paragraph 1 (art. III:2 GATT). Furthermore, the products of the territory of any contracting party imported into the territory of any other contracting party shall be accorded treatment no less favourable than that accorded to like products of national origin in respect of all laws, regulations, and requirements affecting their internal sale, offering for sale, purchase, transportation, distribution, or use (art. III:4 GATT). According to the principle of most-favoured-nation treatment (art. I GATT), WTO members have to apply any VAT benefits granted to imports from one WTO member state or from a non-WTO member state to imports from all other WTO member states. If WTO members have merged into a customs union, VAT benefits granted within the customs union need not be extended to non-members of the customs union. However, WTO members of a customs union have to extend the VAT benefits granted by one WTO member of a customs union to another non-member of the customs union to all other non-members of the customs union (art. XXIV:4–10 GATT). Besides, the GATT also defines other rules for the VAT treatment of imports. For example, the GATT precludes the taxation of the transit of goods (art. V GATT). Furthermore, article VII GATT and the Customs Valuation Agreement determine the customs value as the taxable amount for import VAT. Thus, the determination of the tax base for the importation of goods is already predetermined by customs law. Article X:1 GATT obliges the provision of comprehensive information on the VAT rates applied to the imports as well as on the methods of determination of the VAT tax base. Finally, it should be mentioned that Article XX GATT justifies violations of GATT law under certain conditions, including protection of public morals or protection of human, animal or plant life or health.

41.3 Principles of the TFEU on the Taxation of Imports 41.3.1 Prohibition of Discriminatory Internal Taxation According to article 110(1) TFEU, no member state shall impose, directly or indirectly, on the products of other member states any internal taxation of any kind in excess of that imposed directly or indirectly on similar domestic products.

Taxation of Imports 749 Furthermore, no member state shall impose on the products of other member states any internal taxation of such a nature as to afford indirect protection to other products (art. 110(2) TFEU). Article 110(1) and (2) TFEU are modelled on article III:2 GATT and are only applicable to the trade with Union goods. Non-Union goods released for free circulation in the member states have the status of Union goods and are therefore covered by article 110 TFEU.

41.3.2 Harmonization of Legislation Concerning Indirect Taxation According to article 113 TFEU, the Council shall adopt provisions for the harmonization of legislation concerning turnover taxes, excise duties, and other forms of indirect taxation to the extent that such harmonization is necessary to ensure the establishment and the functioning of the internal market and to avoid distortion of competition. Directive 2006/112/EC2 (VAT Directive) provides for several import VAT rules.3 Non- Union goods brought into the EU should be taxed the same as Union goods supplied within the EU. Indeed, non-harmonized import rules could shift imports from third countries to the member states with the most favourable rules4 leading to a distortion of competition between member states and to an undermining of the regulatory goals of article 113 TFEU.

41.4 Taxation of Imports According to VAT Directive 2006/112/EC 41.4.1 Subject Matter and Scope According to article 2(1)(d) of the VAT Directive, the ‘importation of goods’ shall be subject to VAT. The VAT Directive defines several rules for the taxation of the ‘importation of goods’.5 Goods originating in a third country (‘non-Union goods’) and brought into the EU should be taxed equally as goods originating in an EU member state (‘Union goods’) and delivered within the EU. 2

Council Directive 2006/112/EC of 28 November 2006 on the common system of value-added tax. See arts 30, 60, 61, 70, 71, 85–89, 91, 93, 94, 143–45, 168 lit. e, 178 lit. e, 201, 211, 260, 274–277, and 408– 410 of the VAT Directive. 4 See EU Commission, Tax Harmonization in the European Community (1968), 5. 5 See ibid. 3

750 Thomas Bieber ‘Non-Union goods’ can only be brought into the Union within the forms provided for by customs law to which the VAT Directive and national VAT legislation are linked at various points. The taxation of imports has thus already been preformed by customs law. The Union Customs Code (UCC),6 the Commission Delegated Regulation (UCC-DA7), the Commission Implementing Regulation (UCC-IA8), and Council Regulation (EC) 1186/ 20099 provide numerous rules for, for example, customs procedures, tariff law, customs valuation, customs debt on imports, and reliefs from customs duty. While a supply of goods shall mean the transfer of the right to dispose of tangible property as owner (art. 14(1) of the VAT Directive), according to article 30(1) of the VAT Directive, the importation of goods is caused by ‘the entry into the Community’. Article 2(1)(d) of the VAT Directive covers imports by a taxable person or a private individual, irrespective of whether the transfer is based on a legal transaction, is carried out for consideration, or is free of charge. However, import VAT is postponed if the goods are covered by a procedure or other rule within the meaning of article 61 of the VAT Directive from the time of their entry into the Community. Article 30(1) of the VAT Directive provides that ‘Importation of goods’ shall mean the entry into the Community of goods which are not in free circulation within the meaning of article 24 of the Treaty (art. 29 TFEU). According to article 201(2) of the UCC, ‘Release for Free Circulation’ shall entail the following: • the collection of any import duty due; • the collection, as appropriate, of other charges, as provided for under relevant provisions in force relating to the collection of those charges; • the application of commercial policy measures and prohibitions and restrictions insofar as they do not have to be applied at an earlier stage; and • completion of the other formalities laid down in respect to the import of the goods. Release for free circulation shall confer on non-Union goods the customs status of Union goods (art. 201(3) of the UCC).

6 Regulation (EU) No. 952/2013 of the European Parliament and of the Council of 9 October 2013 laying down the Union Customs Code. 7 Commission Delegated Regulation (EU) 2015/2446 of 28 July 2015 supplementing Regulation (EU) No. 952/2013 of the European Parliament and of the Council as regards to detailed rules concerning certain provisions of the Union Customs Code. 8 Commission Implementing Regulation (EU) 2015/2447 of 24 November 2015 laying down detailed rules for implementing certain provisions of Regulation (EU) No. 952/2013 of the European Parliament and of the Council laying down the Union Customs Code. 9 Council Regulation (EC) 1186/2009 of 16 November 2009 setting up a Community system of reliefs from customs duty.

Taxation of Imports 751

41.4.2 Place of Importation of Goods The place of importation of goods shall be the member state within whose territory the goods are located when they enter the Community, according to article 60 of the VAT Directive. The ‘non-Union good’ is considered to have entered the Union when it has physically crossed the first EU external border and is located within the territory of the EU member state of entry. The place of importation is located in the EU member state of entry if the goods brought there have been released for free circulation as defined by article 201 of the UCC. Or in other words, these goods are not transferred to a special customs procedure within the meaning of articles 210ff of the UCC after they have been transferred to the EU member state of entry. Article 60 of the VAT Directive also applies if goods are brought into the EU member state of entry but—for whatever reason—are neither released for free circulation within the meaning of article 201 of the UCC nor transferred to another special procedure or other arrangement within the meaning of article 61 of the VAT Directive. Such imports may also be relevant from a criminal tax law perspective. From a VAT perspective, it is crucial that the place of importation of a ‘non-Union good’ which is neither released for free circulation in the EU member state of entry nor placed under a special customs procedure within the meaning of articles 210ff of the UCC is located in the EU member state of entry. By way of derogation from article 60 of the VAT Directive, where, on entry into the Community, goods which are not in free circulation are placed under one of the arrangements or situations referred to in article 156 of the VAT Directive, or under temporary importation arrangements with total exemption from import duty, or under external transit arrangements, the place of importation of such goods shall be the member state within whose territory the goods cease to be covered by those arrangements or situations (art. 61(1) of the VAT Directive). Article 156 of the VAT Directive lists the following arrangements or situations: • goods which are intended to be presented to customs and, where applicable, placed in temporary storage; • goods which are intended to be placed in a free zone or in a free warehouse; • goods which are intended to be placed under customs warehousing arrangements or inward-processing arrangements; • goods which are intended to be admitted into territorial waters in order to be incorporated into drilling or production platforms, for purposes of the construction, repair, maintenance, alteration, or fitting-out of such platforms, or to link such drilling or production platforms to the mainland; and • goods which are intended to be admitted into territorial waters for the fuelling and provisioning of drilling or production platforms.

752 Thomas Bieber Article 61(1) of the VAT Directive refers to goods which have been covered by a procedure or other arrangement since their entry into the EU member state of entry. If, for example, a ‘non-Union good’ is transferred from a third country to the EU member state of entry in accordance with articles 30(1) and 61(1) of the VAT Directive, from which it is transported to another EU member state using the external Union transit procedure (T1) in accordance with articles 226 et seq. of the UCC, then the import, as defined in article 2(1)(d) of the VAT Directive, takes place in the other EU member state if the object is there released for free circulation. Whether the goods are covered by the relevant procedure or arrangement must first be determined under customs law due to the lack of own VAT provisions.10 All customs violations must therefore be reflected when determining the place of importation subject to VAT. From the date at which the customs procedure is violated, it can be assumed that the goods are no longer covered by this procedure from a VAT perspective and that the place of importation is the place where the customs procedure has been violated.11

41.4.3 Chargeable Event and Chargeability of VAT According to article 70 of the VAT Directive, the chargeable event shall occur and VAT shall become chargeable when the goods are imported. VAT becomes chargeable when the goods are released for free circulation. Articles 30(1) and 60 of the VAT Directive are applicable to imports where the ‘non-Union goods’ are released for free circulation in the EU member state of entry and where the ‘non-Union goods’ are transferred to the EU member state of entry by violating customs regulations. According to article 70 of the VAT Directive, the chargeable event shall occur and VAT shall become chargeable when the goods are released for free circulation in the EU member state of entry or when they cross the border from a third country to the EU member state of entry. According to article 71(1) subparagraph 1 of the VAT Directive, the chargeable event is postponed if the goods are covered by a procedure or other arrangements from the date of their introduction into the Union. Article 71(1) subparagraph 1 applies to transfers to a temporary warehouse, a free zone, a free warehouse, a customs warehouse, the inward- processing procedure, the external transit procedure, the internal Community transit procedure, and the temporary importation procedure with total relief from import duties. The procedures and rules listed in article 71(1) subparagraph 1 are identical to those listed in article 61 of the VAT Directive.12 If the goods are covered by one of these procedures or arrangements at the time of their transfer, the chargeable event and the tax become chargeable only at the time when the goods are no longer covered by these procedures or arrangements.

10

Case C-480/12 X (2014) para. 53 and 54. See, however, Section 41.4.4. 12 See Section 41.4.2. 11

Taxation of Imports 753 Whether the goods are subject to the customs procedure or arrangement in question must first be analysed for customs purposes. Article 71(1) subparagraph 1 of the VAT Directive does not provide for a different VAT approach. All customs violations leading to the termination of a special customs procedure must therefore be reflected for VAT purposes. From the moment at which the special customs procedure is terminated for customs purposes, it can be assumed that the goods are also no longer subject to this procedure for VAT purposes and that import VAT liability arises.13 If the imported goods are subject to customs duties, the chargeable event shall occur and VAT shall become chargeable when the chargeable event in respect of those duties occurs and those duties become chargeable. Article 71(1) subparagraph 2 of the VAT Directive defines a common date on which customs duties and VAT become chargeable. Where imported goods are not subject to any of the duties referred to in article 71(1) subparagraph 2 of the VAT Directive, member states shall, as regards the chargeable event and the moment when VAT becomes chargeable, apply the provisions in force governing customs duties. Article 71(2) is applicable to cases in which imports are not subject to import duties. In such cases, however, the rules on customs duties must be applied to determine an import VAT debt.

41.4.4 Entry of the ‘Non-Union Good’ into the Economic Network as a Requirement for Importation? According to consistent case law of the ECJ14 since 2016, the incurrence of import VAT requires that the goods in question have entered the economic network of the Union and thus could be supplied to a consumption (i.e. the transaction charged with VAT). The incurrence of the customs debt and the incurrence of the import VAT debt may differ from each other, provided that the goods have not entered the economic network of the Union. The ECJ has already emphasized in the case of Dansk Transport15 that neither the chargeable event for VAT has occurred nor that VAT has become chargeable where the imported goods are seized and simultaneously or subsequently destroyed by the competent authorities before going beyond the first customs office situated inside the customs territory of the Community. However, if the goods leave the first customs office without being subject to a procedure, they are imported into the Union and import VAT liability arises.16 Looking at the case of Federal Express,17 the ECJ further developed its concept of the ‘economic network’ and thereby again distinguished between customs and import VAT 13

See, however, Section 41.4.4. Joined Cases C-226/14 and C-228/14 Eurogate Distribution (2016); Case C-154/16 Latvijas dzelzceļš (2017); Case C-571/15 Wallenborn Transports (2017); Case C-26/18 Federal Express (2019). 15 Case C-230/08 Dansk Transport (2010), para. 93. 16 Ibid., para. 97. 17 Case C-26/18 Federal Express (2019). 14

754 Thomas Bieber in cases of customs law infringements. In this specific case, the goods were brought into the customs territory of the Union via Germany, violating customs regulations, but were subsequently transported to their destination in Greece. In its ruling, the ECJ confirmed that import VAT only occurs if the goods have entered the economic network of the Union18 and concretized its case law by a rebuttable presumption. If goods are brought into the customs territory of the Union by violating customs rules, the import VAT liability arises in the member state in which the goods have entered the economic network of the Union. In principle, it can be assumed that this is the member state in which the infringement took place. However, this presumption can be disproved. If it can be proven that the goods only entered or are to enter the economic network of the Union in the territory of another member state (country of destination), import VAT only arises in the country of destination, as import VAT can only arise once.19 The case of VS20 involved a resident of Germany who brought his passenger car from Turkey, where it was registered, into Germany, passing through Bulgaria, Serbia, Hungary, and Austria. The importation of that vehicle was discovered in Germany during a police check and the vehicle was later re-exported to Turkey where it was sold. The German authorities imposed the import duty pursuant to article 87(4) of the UCC because: (1) the German customs authorities established the infringement in Germany; and (2) the import duty amount was below €10,000. Similarly, the German authorities assessed the import VAT in Germany by an analogue application of article 87(4) of the UCC. The ECJ ruled that, similar to the case of Federal Express, the vehicle of VS physically entered the territory of the Union through Bulgaria, so that it was in that member state that there was a failure to comply with the customs obligations.21 Nevertheless, it would be apparent from that information, which it is for the referring court to determine, that even if, on its way from Turkey to Germany, the vehicle at issue first entered the customs territory of the Union in Bulgaria and, after transiting through the territory of a non- member country, namely Serbia, then re-entered the customs territory of the Union in Hungary, that vehicle was actually used in Germany, VS’s member state of residence.22 Accordingly, inasmuch as the vehicle entered the economic network of the Union in Germany, it is in that member state that the import VAT was incurred.23 In the case of VS, the ECJ does not address the (crucial) question of whether article 87(4) of the UCC can be applied to import VAT by analogy.24 Further, the ECJ does 18

Ibid., para. 44. Ibid., para. 44. 20 Case C-7/20 VS (2021). 21 Ibid., para. 34. 22 Ibid., para. 35. 23 Ibid., para. 35. 24 For the sake of completeness, it should be noted that the Hamburg Finance Court submitted the question to the ECJ of whether there is a violation of articles 30 and 60 of the VAT Directive if a member state declared article 215 para. 4 of the Customs Code (corresponds to article 87 para. 4 of the UCC) to be 19

Taxation of Imports 755 not deal with the fact that the vehicle was re-exported to Turkey where it was sold. Incidentally, it would have been obvious to deal with article 60 of the VAT Directive. Article 60 also covers cases in which the goods are brought into the EU member state of entry, but—for whatever reason—are neither released for free circulation within the meaning of article 201 of the UCC nor placed under another special procedure or arrangement within the meaning of article 61 of the VAT Directive. For VAT purposes, the place of importation of a non-Union good that is neither released for free circulation in the EU member state of entry nor placed under a special customs procedure within the meaning of article 210 of the UCC is in the EU member state of entry. If the ECJ had followed this interpretation of article 60 of the VAT Directive, the import VAT would have had to be prescribed in Bulgaria. Most recently, in the case of UB25 the ECJ decided that articles 2(1)(d) and 70 of the VAT Directive must be interpreted as meaning that the extinguishment of the customs debt on the ground provided for in article 124(1)(e) of the UCC does not lead to the extinguishment of the debt linked, respectively, to excise duty and to VAT in respect of goods unlawfully introduced into the customs territory of the European Union. Whether the UB case is an ‘individual decision’ or whether the ruling should be understood as turning away from the previously mentioned ECJ jurisprudence on the ‘entry into the economic network’ must remain open at this point.

41.4.5 Taxable Amount In terms of the importation of goods, the taxable amount shall be the value for customs purposes, determined in accordance with the Community provisions in force (art. 85 of the VAT Directive). Any value other than the customs value established under article 85 of the VAT Directive shall be excluded as the basis for assessment of importation.26 As a general rule, the customs value of imported goods is based on the transaction value, which is the price actually paid or payable for the goods when sold for export to the EU customs territory. Where the transaction value method within the meaning of article 70(1) of the UCC is not applicable, according to article 74(2) and (3) of the UCC, the customs value must be determined in the following order: transaction value of identical goods, transaction value of similar goods, deductive method, computed method, or residual method. The first method is the transaction value method according to article 70(1) of the UCC, which is applied most frequently in practice. The transaction value is the price actually paid or payable for the goods when they are sold for export to the customs

applicable to import VAT law, see the decision of the Hamburg Finance Court of 28 December 2022, 4 K 1/18 (C-791/22). 25

26

Case C-489/20 UB (2022), para. 52. Case C-62/93 BP Soupergaz (1995), para. 35.

756 Thomas Bieber territory of the Union, which must be adjusted if necessary. The transaction value method already considers numerous factors that could distort the customs value. For example, article 71(1)(c) of the UCC and article 136(4) of the UCC-IA regulate the addition of licence fees. If the customs value of affiliated companies is influenced by the transfer prices agreed (art. 70(3)(d) of the UCC, in conjunction with article 127 of the UCC-IA and article 134 of the UCC-IA), the transaction value does not apply. Income tax principles can be used as an indication of the adequacy of the arm’s-length nature but cannot be used as the sole justification for them.27 The transaction value method within the meaning of article 70 of the UCC already regularly considers the deductions listed in article 87 of the VAT Directive (cash discounts, rebates). However, these deductions have to be taken into account within the scope of the subordinate methods of customs valuation listed in article 74 of the UCC. When dealing with transport costs, the VAT approach differs from customs law. According to article 86(1)(b) of the VAT Directive, transport costs incurred after the goods have been brought into the Union and which reduce the customs value are to be added to the total insofar as they have been incurred up to the first place of destination in the member state of entry. In the case of an import within the meaning of article 2(1)(d) of the VAT Directive, only the private, final consumer should bear the VAT. Tax neutrality at the business level is ensured by the deduction of input tax for the import VAT. Inaccuracies in the determination of the customs value do not have any effect insofar as the importer is entitled to a full input tax deduction. From a VAT perspective, it is only necessary to check all application requirements as defined in article 70(3) of the UCC if the importer cannot claim the VAT paid as input tax and bears the VAT itself (e.g. imports by private individuals or imports by logistics service providers who are required to pay VAT but who are not entitled to deduct input tax for the VAT due to a lack of power of disposal).28 In these cases, the customs valuation has a direct effect on the amount of the VAT burden. In the literature, it has been argued that article 85 of the VAT Directive references the customs value by analogy and article 70 of the UCC would have to be interpreted in terms of excise duty teleology.29 The comparison with article 80 of the VAT Directive would show that the arm’s-length principle of article 70 of the UCC only applies to those cases in which the buyer and seller are linked with each other and the buyer is not entitled to a full deduction.30 A subsequent entry in the accounts in accordance with article 105(4) of the UCC would not affect VAT if the purchaser is entitled to a full deduction of the input tax. Article 27 S. Vonderbank, ‘Die Bedeutung von Verrechnungspreisen bei der Zollerhebung’, Internationales Steuerrecht 25 (2016), 331; T. Möller, ‘Verrechnungspreis und Zollwert—Weiter ein Dilemma?’, Zeitschrift für Zölle und Verbrauchsteuern 90 (2014), 150; critical C. Montag, ‘Wechselwirkungen zwischen Verrechnungspreis und Zollwert’, Der Konzern 15 (2016), 285 according to whom income tax and customs valuation rules are essentially consistent. 28 H. Jatzke in Sölch and Ringleb (Eds.), UStG Art 11 no. 12. 29 See C. Heber, ‘Verrechnungspreise im System des Mehrwertsteuerrechts’, Mehrwertsteuerrecht 3 (2015), 380. 30 See ibid., 381.

Taxation of Imports 757 105(4) of the UCC would require that the amount entered in the accounts is at a lower level than the amount payable. However, this requirement would not be met in the case of full entitlement to deduct input tax as the transaction value would have to be recognized as the basis for VAT assessment.31 In my opinion, this view is not convincing. Article 85 of the VAT Directive defines the VAT base as the amount determined as the customs value by the Community provisions in force. The customs value determined in accordance with the provisions of the UCC must therefore already be used as the basis for determining VAT. There is neither scope for implementation for the member states nor scope for an application of the customs provisions by analogy. This also applies if the customs value of affiliated companies is influenced by the transfer prices agreed (art. 70(3)(d) UCC, in conjunction with article 127 of the UCC-IA and article 134 of the UCC-IA), and the buyer is entitled to a full input tax deduction. If the customs value were not adjusted, this could, for example, affect the determination of the taxable amount for subsequent supplies of imported goods.

41.4.6 Exemptions on Importation The mandatory tax exemptions of article 143(1) of the VAT Directive are based on different objectives and can be categorized as follows: First, article 143(1)(a) of the VAT Directive provides for a mandatory tax exemption for the final importation of goods of which the supply by a taxable person would in all circumstances be exempt within their respective territory. This equality of imported goods with goods delivered in the member state of importation is due to the principle of non- discrimination of article III:2 GATT. Secondly, the purpose of the VAT Directive is to harmonize customs exemptions and VAT exemptions with regard to imports of personal luggage as defined by Directive 2007/74/EC,32 imports of goods as defined by Directive 2009/ 132/ EC,33 and imports of private, small consignments as defined by Directive 2006/79/ EC34 (art. 143(1)(b) of the VAT Directive). However, import VAT exemptions differ in some aspects from the customs exemptions laid down by Council Regulation (EC) 1186/ 2009.35

31

See ibid., 381. Council Directive 2007/74/EC of 20 December 2007 on the exemption from value-added tax and excise duty of goods imported by persons travelling from third countries. 33 Council Directive 2009/132/EC of 19 October 2009 determining the scope of Article 143(b) and (c) of Directive 2006/112/EC in regards to exemption from value-added tax on the final importation of certain goods. 34 Council Directive 2006/79/EC of 5 October 2006 on the exemption from taxes of imports of small consignments of goods of a non-commercial character from third countries. 35 Council Regulation (EC) 1186/2009 of 16 November 2009 setting up a Community system of reliefs from customs duty. 32

758 Thomas Bieber Directive 2009/132/EC was amended by Directive (EU) 2017/245536 by deleting its Title IV relating to imports of negligible value. Thus, with effect from 1 July 2021, it eliminates the VAT exemption for small consignments (worth a maximum of €22). Furthermore, Directive (EU) 2017/2455 extended the current EU one-stop shop to distance sales of goods, both intra-EU and from non-EU countries. Thirdly, imports of goods from certain agencies, bodies, or armed forces37 or imports of certain goods are exempted from VAT (fishery products, gold from central banks, gas, electricity, heating, or cooling).38 Fourthly, article 143(1)(d) of the VAT Directive provides for an exemption from VAT which must be declared in the customs declaration with the code ‘42’. The exemption under article 143(1)(d) is therefore referred to as ‘Procedure 42’. Importation into a member state with a VAT-exempt supply for another member state is exempt from VAT. As a result, an intra-Community acquisition is to be taxed in the member state of destination. ‘Procedure 42’ is characterized by a two-stage proof system. First, the exemption from VAT depends on the importer providing certain information to the competent authorities of the member state of importation at the time of importation, including various VAT identification numbers and evidence that the imported goods are intended to be transported or dispatched from the member state of importation to another member state (art. 143(2) of the VAT Directive).39 If that information cannot be provided at the time of importation or proves to be incorrect, the VAT exemption under article 143(1)(d) of the VAT Directive generally does not apply. However, in the case of Enteco Baltic, the ECJ ruled that articles 143(1)(d) and 143(2) (b) of the VAT Directive must be interpreted as precluding the competent authorities of a member state from refusing exemption from VAT on importation on the sole ground that, following a change of circumstances after the importation, the goods in question have been supplied to a taxable person other than the person whose VAT identification number was stated in the import declaration, where the importer has communicated all the information on the identity of the new purchaser to the competent authorities of the member state of import, provided that it is shown that the substantive conditions for the exemption of the subsequent intra-Community supply are actually satisfied.40 In addition to the declaration and proof obligations at the time of importation, the exemption from VAT must meet all the requirements for a tax-free, intra-Community supply as defined in article 138(1) of the VAT Directive or a tax-free, intra-Community

36 Council Directive (EU) 2017/ 2455 of 5 December 2017 amending Directive 2006/112/EC and Directive 2009/132/EC as regards certain value added tax obligations for supplies of services and distance sales of goods. 37 Art. 143(1)(f)–(i) of the VAT Directive. 38 Art. 143 (1)(j)–(l) of the VAT Directive. 39 However, Member States may provide that the evidence that the imported goods are intended to be transported or dispatched from the Member State of importation to another Member State is indicated to the competent authorities only upon request. 40 Case C‑108/17 Enteco Baltic (2018), para. 61.

Taxation of Imports 759 transfer as defined in article 138(2)(c). According to the more recent ECJ rulings41 on the relationship between the formal and the material requirements of an intra-Community supply, the VAT exemption for intra-Community supplies within the meaning of article 138(1) of the VAT Directive is granted if the material requirements are fulfilled, even if the taxable person has not fulfilled certain formal requirements. Accordingly, the exemption from VAT under article 143(1)(d) of the VAT Directive should be maintained in these cases, provided that the import-related declaration and proof obligations were fulfilled at the time of import. The ECJ has ruled in the case of Milan Božičevič Ježovnik42 that an exemption from VAT on importation is only granted if the conditions for a tax-exempt, intra- Community supply are fulfilled. A refusal to grant VAT exemption after a positive preliminary examination is only possible if the taxpayer has not acted in good faith and carefully. In the case of Vetsch, the ECJ ruled that the import followed by an intra-Community transfer and, secondly, the intra-Community supply to which the tax evasion relates, must be regarded as transactions that are independent of one another.43 Furthermore, it should be noted that article 143(1)(d) of the VAT Directive in fact entails a dual exemption: first, an exemption from VAT which, under article 201 of the VAT Directive, is normally payable on imports and, secondly, an exemption in respect of the intra- Community supply or transfer which followed such imports.44

41.4.7 Right of Deduction According to article 168(e) of the VAT Directive, insofar as the goods and services are used for the purposes of the taxed transactions of a taxable person, the taxable person shall be entitled, in the member state in which they carry out those transactions, to deduct the following from the VAT which they are liable to pay ( . . . ) the VAT due or paid in respect of the importation of goods into that member state. The right to deduct arises at the time when the deductible tax is due. On importation, this is the date on which the goods are released for free circulation or are no longer covered by any procedure or other arrangement. In order to deduct VAT, the taxable person must have an import document within the meaning of article 178(e) of the VAT Directive specifying them as the consignee or importer and stating the amount of VAT due or enabling that amount to be calculated. Where the member state of importation has introduced an electronic system for completing customs formalities, according to article 52 of the Council Implementing Regulation (EU) No. 282/2011, the term ‘import document’ shall cover electronic 41

Case C-24/15 Josef Plöckl (2016); Case C-21/16 Euro Tyre BV (2017). Case C-528/17 Milan Božičevič Ježovnik (2018). 43 Case C-531/17 Vetsch (2019), para. 38. 44 Ibid., para. 39. 42

760 Thomas Bieber versions of such documents, provided that they allow for the exercise of the right of deduction to be checked. The deduction of VAT must not be made dependent on the actual prior payment of VAT.45 Neither article 168(e) nor article 178(e) of the VAT Directive nor article 52 of the Council Implementing Regulation (EU) No. 282/2011 require that VAT must be paid in order for the deduction to be made. In its decision in Case C-187/14 of 25 June 2015, the ECJ ruled that national legislation under which a logistics service provider who is not the owner of the goods is not entitled to the deduction of import VAT is in line with EU law.46 The ECJ also ruled that the right to deduct import VAT depends on whether the value of the imported goods is part of the costs of the output sales of the taxable person. According to the ECJ, this requirement is not fulfilled in the case of a logistics service provider whose activity is limited to transportation purposes. In general, it seems questionable whether the criterion established by the ECJ for the deduction of import VAT correlates with the principle of neutrality. The importation of goods is not based on an exchange of services, but is realized independently of it. Through the importation of goods, an entrepreneur realizes a turnover without consideration. If a logistics service provider owes the VAT due on importation and fulfils all other requirements for claiming the input tax deduction, the input tax deduction should therefore be permitted. Article 168 of the VAT Directive would have to be interpreted in such a way that the importation of goods is deemed to have been carried out as a transaction for the purposes of the taxed transactions of a logistics service provider. The case of DSV Road was related to the previous import rules of the UCC. Regarding article 79 of the UCC, it must be examined whether the customs debt will be extinguished pursuant to article 124 of the UCC. For example, the customs debt was incurred pursuant to article 79 of the UCC and evidence is provided to the satisfaction of the customs authorities that the goods have not been used or consumed and have been taken out of the customs territory of the Union (art. 124(1)(k) of the UCC). According to article 124(1)(h) of the UCC, the customs debt shall be extinguished where the customs debt was incurred pursuant to article 79 or 82 of the UCC and the failure which led to the incurrence of a customs debt had no significant effect on the correct operation of the temporary storage or of the customs procedure concerned and did not constitute an attempt at deception and all of the formalities necessary to regularize the situation of the goods are subsequently carried out. If the customs debt extinguishes in one of the ways of article 124 of the UCC, the import VAT debt is also extinguished.47 The restrictive view concerning the deduction of import VAT would be obsolete in the cases of article 124 of the UCC.

45

Case C-414/10 Véleclair (2012). Case C-187/14 DSV Road (2015). 47 See, however, the latest ruling of the ECJ in Case C-489/20 UB (2022) where the ECJ declined the analogue application of art. 124(1)(e) of the UCC and the extinguishment of import VAT. 46

Taxation of Imports 761 If a logistics service provider is nevertheless prohibited from the deduction of import VAT, it must be examined whether the VAT owed by them can be refunded or waived in accordance with the conditions of articles 116 et seq. of the UCC. Such a refund claim can result from article 120(1) of the UCC, according to which import duties shall be repaid or remitted in the interest of equity where a customs debt is incurred under special circumstances in which no deception or obvious negligence may be attributed to the debtor. Obvious negligence on the part of the indirect representative can only be excluded if, in compliance with all the provisions, the indirect representative has communicated the necessary information to the competent customs office, submitted the required evidence, and taken all reasonable steps to identify or prevent possible or intended irregularities. In Austria, logistics service providers can assert a claim for reimbursement or remission for failed import clearance in procedure 42 (imports with subsequent intra-Community delivery) under the conditions of article 120 of the UCC in conjunction with article 73 of the Austrian Customs Law Implementing Act. The ECJ has ruled in the case of Weindel Logistik Service48 that a logistics service provider is not entitled to the deduction of import VAT, confirming its guiding statements from the case of DSV Road from 2015.49 With regard to the principles of neutrality and proportionality, customs authorities could—in cases of several tax debtors—impose the import VAT primarily against the import VAT debtor who has the power of disposal over the imported goods and is thus entitled to the deduction of import VAT.

41.4.8 Taxable Person For the purposes of determining the VAT debtor, a distinction must be made between regular imports, as defined by article 201 of the VAT Directive, and irregular imports, as defined by article 202 of the VAT Directive. Unlike for supplies of goods or services, the VAT Directive leaves the determination of the VAT debtor of a regular import within the meaning of article 201 entirely to the individual EU member states. Unsurprisingly, the option of article 201 has been implemented differently in the individual EU member states.50 In particular, the group of possible EU VAT debtors within the meaning of article 201 is not necessarily identical to the group of import customs debtors within the meaning of article 77(3) of the UCC. Some EU member states interpret the option of article 201 of the VAT Directive independently for VAT purposes designating direct customs representatives (art. 18 of the UCC) as VAT debtors, whereas, according to article 77(3) of the UCC—except in the case of article 77(3) subparagraph 2 of the UCC—the

48

Case C-621/19 Weindel Logistik Service (2020). a critical analysis of these rulings, see T. Bieber and L. Gläser, ‘Deduction of Import VAT: Relevance of Ownership as of Importation?’, International VAT Monitor 33 (2022), 75ff. 50 See Deloitte, ‘Information on the import VAT collection in the Member States’ (2011). 49 For

762 Thomas Bieber direct customs representative is not liable for import duties. All member states have in common that the import VAT liability within the meaning of article 201 of the VAT Directive is not limited to one person. Article 202 of the VAT Directive refers to any person who causes goods to cease to be covered by the arrangements or situations listed in Articles 156, 157, 158, 160, and 161 of the VAT Directive. The group of EU VAT debtors within the meaning of article 202 of the VAT Directive thus seems smaller than the group of import customs debtors within the meaning of article 79(3) and (4) of the UCC. While, for example, liability for import customs debt within the meaning of article 79(3)(b) of the UCC requires participation in the customs duty violation, article 202 of the VAT Directive requires that the person has ‘caused the goods to cease to be covered’. However, according to article 205 of the VAT Directive, member states can opt for that a person other than the import VAT debtor within the meaning of article 202 of the VAT Directive is jointly and severally liable to pay the tax in the cases referred to in article 202.

41.4.9 Reporting Requirements According to article 260 of the VAT Directive, the member states shall lay down detailed rules for the submission of VAT returns with respect to the importation of goods. Article 260 does not compete with Article 250 of the VAT Directive (i.e. the obligation to submit a VAT return on importation exists in addition to the obligation to submit a VAT return within the meaning of Article 250). Article 260 leaves the EU member states a wide scope for determining the specific declaration obligations. The EU member states can thus regulate the declaration obligations for the importation of goods independently. Due to the extensive scope of article 260, the member states can allow the VAT declaration obligations for imports to be fulfilled in a customs declaration. All forms of customs declaration, such as the standard customs declaration, the simplified customs declaration, or the entry in the declarant’s accounts, could be used.

41.4.10 Payment Arrangements Member states shall lay down the detailed rules for payment in respect of the importation of goods (art. 211 subpara. 1 of the VAT Directive). Like the definition of declaration obligations for imports laid down by article 260 of the VAT Directive, article 211 subparagraph 1 defines a broad scope for EU member states for establishing the relevant payment obligations. For the payment of import VAT, article 211 subparagraph 2 of the VAT Directive provides for the possibility of a tax deferral insofar as—in the case of the importation of goods by taxable persons or certain categories thereof or by persons liable for payment of VAT or certain categories thereof—the VAT due by reason of the importation need

Taxation of Imports 763 not be paid at the time of importation, on condition that it is entered as such in the VAT return to be submitted in accordance with article 250 of the VAT Directive. The taxpayer or tax debtor need not be the person who acts as the importer within the meaning of customs law. The recipient of the imported goods can also declare the import VAT in their VAT return.51 In its previous case law, the ECJ52 confirmed the admissibility of own systems for import VAT payments. Article 211 of the VAT Directive does not prevent a member state from setting different time periods and payment periods for import VAT than for VAT payments of domestic supplies.53 The EU member states may link the import VAT payment to the acquisition of a licence, provided that the modalities for obtaining that licence are compatible with the principle of neutrality.54 The ECJ did not define concrete indications of such compatibility.55 Rather, it is up to the national courts to examine the modalities of the payment of the import VAT on the basis of the principle of neutrality.56 A 2011 study by Deloitte57 showed that the import VAT payment systems of the EU member states and the implementation of article 211 of the VAT Directive differ significantly: systems vary between immediate payment of the import VAT, postponed accounting via the VAT return, deferred payment for VAT and customs duties in a similar delay, specific deferred payment for VAT purposes only, and specific deferred payment for customs duties only.

41.4.11 Final Remarks The taxation of imports is based on the fundamental principles of the GATT and the TFEU and links to the provisions of customs law at various points. The VAT Directive relies not only on the customs transport procedures for bringing goods from third countries into the Union but also on the customs valuation methods. However, customs debt and import VAT can differ from each other if the goods do not enter the ‘economic network’ of the Union. This concept should follow the principle of VAT as a consumption tax, although its interpretation and application by the ECJ (in recent cases) still leaves some questions unanswered.

51 Y. Fedchyshyn, ‘Postponed Accounting in the European Union’, International VAT Monitor 25 (2014), 13; A. Bal, ‘Import and Export in Multi-Sale Transactions—Part I’, International VAT Monitor 27 (2016), 109. 52 Case C-79/12 SC Mora IPR (2013); Case 42/83 Dansk Denkavit (1984). 53 Case 42/83 Dansk Denkavit (1984), para. 20. 54 Fedchyshyn, ‘Postponed Accounting in the European Union’. 55 Case C-79/12 SC Mora IPR (2013). 56 Ibid. 57 Deloitte, ‘Information on the import VAT collection in the Member States’ (2011).

Chapter 42

‘W h ite Suppl i e s ’ a nd D ouble Taxat i on i n Cross-B orde r VAT L aw Heidi Friedrich-Vache

42.1 Introduction There are many examples of cases where the supply of an item is unintentionally subject to value-added tax (VAT) in two or more countries—or not taxed at all. The fundamental concept is that VAT should not be applied twice because the charge is based on the principle of a consumption tax. It is, therefore, only linked to the place of supply (rather than through secondary transit routes) as a basic territoriality principle. There can only be one final place of consumption, so the VAT system should preclude double taxation since it is physically and actually impossible to supply the same item in two locations. However, in modern economies, the place of consumption may be hard to identify, for example when downloading or streaming songs, purchasing non-fungible token art(so-called NFT), the sale and issuing of vouchers or tokens for online auctions, or consuming energy in an electric vehicle during a cross-border trip. Different countries, of course, have varying legal concepts and approaches to addressing this challenge. The problem is that where those approaches are not harmonized or coordinated, double taxation, or non-taxation, is common. Each contractual party, especially the supplier, should closely evaluate the VAT treatment of each transaction. This not only applies to domestic regulations but should also include the jurisdiction of the residency location of the recipient or the country of consumption or activity. Therefore, a thorough VAT evaluation is required to avoid VAT and/or risks of penalty. Nevertheless, although this process might detect double taxation, it does not necessarily prevent it.

766 Heidi Friedrich-Vache The principle of territoriality is determined differently in various global VAT systems. Essentially, it can be linked to: (1) the supply category, with varying regulations according to the physical movement or non-movement of goods or property; (2) the presumed or actual place of activity or use; or (3) the residency of the supplier or recipient. The EU VAT system generally applies to all goods and services supplied within the bloc. Thus, goods sold for export or services rendered to non-EU recipients are not usually subject to VAT in the country of origin. For fairness and market neutrality reasons, imports are subject to EU VAT so that EU-based markets can compete equally within the internal market with external suppliers.1 The characteristic features of EU VAT are as follows: • It constitutes a consumption tax because the final consumer bears the cost. As a surcharge on the transaction price, the tax burden is visible at each stage of the production and distribution chain. • VAT-registered businesses, or the ‘taxable person’, collect the tax fractionally, deducting the amount of VAT paid (input VAT) from the VAT collected (output VAT) through business transactions. This mechanism of an input VAT deduction in specific business-to-business (B2B) cases ensures that the tax is neutral and does not burden the VAT-taxable person. • Suppliers of goods or services within the production or distribution chain pay the tax to the authorities, but the end user bears the cost as an element of the transaction price; thus, it is an indirect tax. VAT is a consumption-oriented tax and burdens the final consumer so the destination principle commands taxation at the place of consumption.2 The current European Council Directive 2006/112/EC provisions regarding the dispatch of cross-border goods seeks to realize the destination concept within the EU via the country-of-origin principle. According to that principle, VAT is payable in the EU member state of final consumption. To avoid double taxation, the supply of those goods must be previously VAT-exempt in the country of origin (where the place of taxation is determined).3 In practice, it is mainly that principle that is used

1 e.g.

see the European Commission article at https://ec.europa.eu/taxation_customs/business/vat/ what-is-vat_en (accessed 27 February 2021). 2 See the Opinion of AG Jacobs in Case C-215/94 Mohr, ECLI:EU:C:1995:405; W. Reiß in W. Reiß, J. Kraeusel, and M. Langer, eds, Umsatzsteuergesetz (UStG) (Bonn: 2015), Introduction, m.no. 10.1, requesting the country-of-destination principle to correspond with the idea of taxation at the place of consumption, see also H. Friedrich- Vache, Verbrauchsteuerkonforme Umsatzbesteuerung von Finanzdienstleistungen, Plädoyer für die Abschaffung unechter Steuerbefreiungen (Cologne: 2005), 17, 19. 3 See Opinion of AG Kokott, Case C‑401/ 18 Herst, s.r.o., ECLI:EU:C:2019:834. Currently, in a transition period, the country of origin as a general principle is applicable in the area of supply of goods for B2B business; for B2C businesses with electronically performed services, taxation switched to the residency of the recipient and will now be enlarged in the second step of the e-commerce package, see COM(2017) 2455 (5 Dec. 2017); in the second step of the digital/e-commerce package, the rules for B2C business, EU deliveries, and third-country inbound deliveries taxation switched to the country

‘White Supplies’ and Double Taxation in Cross-Border VAT Law 767 worldwide.4 Double taxation is triggered for supplies of goods and services if the country of origin does not provide an exemption for allocating VAT revenue to another country. Where the term double taxation is discussed in relation to VAT,5 double taxation is a factor if a single supply or transaction is taxed with output VAT more than once6 (i.e. in the case of more than one determined place of supply in more than one country). The Council Directive already foresees the potential for conflict between the EU member states when determining where taxable transactions occur.7 Disorderly state access to linked taxable persons and tax revenue (by personality, territoriality, or the source-state principle) leads to double taxation and thus to a higher burden on cross- border taxpayers than those who only operate domestically. This reality contradicts the objective of the European internal market and distorts competition. However, a violation of the fundamental freedoms within the EU can only be considered if and to the extent that the additional burden is only attributed to one EU member state. If, on the other hand, the disadvantage is merely a consequence of the continued existence of the autonomous, non-harmonized tax law systems of the EU member states, the Court of Justice of the European Union (CJEU) cannot provide a remedy. The elimination of double taxation is an EU law objective that can be promoted by means of double tax treaties (see ex art. 293 EC Treaty).8 To be fair and neutral, VAT systems must be applied uniformly, addressing double taxation and non-taxation. Measures should address, in particular, the problem of double taxation within cross-border transactions which can result from divergence between EU member states in applying the rules governing the place where taxable transactions are carried out.9

of destination from 1 July 2021); the VAT package of 4 December 2007 with adoption in 2010 of B2B businesses for supplies of services is covered with assumed consumption at the place of residency. 4 See H. G. Ruppe‚ ‘Internationale Probleme auf dem Gebiet der Umsatzsteuer (Generalbericht)’, CDFI 68 (1983), 15, 24ff. 5 See R. Ismer and K. Artinger, Internationale Doppelbesteuerung im Mehrwertsteuerrecht: Ursachen und mögliche Lösungen, Mehrwertsteuerrecht (Munich: 2018), 12, 14. Using almost the same definition is preferred, e.g. by A. Verstraeten in M. Lang, P. Melz, and E. Kristoffersson, eds, Value Added Tax and Direct Taxation: Similarities and Differences (Amsterdam: 2009), 1122, 1126; OECD Model Tax Convention on Income and Capital (Condensed Version 2017), Commentary, Introduction no. 1, arts 23A and 23B no. 1. 6 For further definitions, see Ismer and Artinger, Internationale Doppelbesteuerung im Mehrwertsteuerrecht, 12, 14, who prefer to focus on the economic substance of a transaction—and not a ‘tax event’; a refined definition is not required in my opinion. 7 The directive included the country- of-origin principle as a compromise to create and adopt a European single internal market between the first EU member states. During the last few decades, further rules have promoted the principle of the country of destination/residency as the place of supply/ consumption. 8 See J. Kokott, Das Steuerrecht der Europäischen Union (Munich: 2018), § 5, m.nos 14, 15. 9 See, e.g., Council Directive 2006/112/EC of 28 November 2006, Preamble, paras 17, 61, 62.

768 Heidi Friedrich-Vache The idea of VAT taxation continues to spread worldwide and international trade in goods and services has also expanded rapidly in an increasingly globalized economy. One consequence of these developments has been increased interaction between VAT systems, along with the growing risks of double taxation and unintended non-taxation in the absence of international VAT coordination, not only in the EU and its internal market but also in and between non-EU countries.10

42.2 Cases of Non-Taxation or Double Taxation In direct taxation, the terms ‘white income’ (non-taxed) and ‘double taxation’ are well known, and bilateral or multilateral agreements exist to avoid both tax consequences in an ideal economic world. According to the wording of the relevant VAT or GST,11 ‘white supplies’ with no VAT taxation can arise. In contrast, many double taxation scenarios are possible. However, EU countries should coordinate their VAT systems according to the EC Treaty and the Treaty on the Functioning of the European Union (TFEU). Non-taxation could arise, for instance, when a non-resident company provides battery-charging services for customers in Italy. This transaction constitutes a supply of service and, sui generis using different applications for charging and paying from an Italian VAT law perspective, is not a pure supply of electricity according to articles 38 and 39 Council Directive. From a German VAT perspective, purchasing and selling electricity and using ancillary administrative services leads to a taxable event linked to a German resident company at the place of consumption in the case of B2B or business-to-consumer (B2C) customers (not acting as resellers of electricity) in Italy. Therefore, from a German perspective, there is no domestic taxable supply in Germany. Nor does Italian tax law recognize a taxable domestic supply. Rather, Italy deems that for charging a car in Italy the supply takes place in Germany at the supplier’s residency in the case of B2C transactions or at the place of a German business customer.

10 See OECD, ‘International VAT/ GST Guidelines’ (2017), https://www.oecd.org/tax/consumpt ion/international-vat-gst-guidelines-9789264271401-en.htm (accessed 1 January 2021); OECD, The Application of Consumption Taxes to the International Trade in Services and Intangibles—Progress Report and Draft Principles (Paris: OECD Publishing, 2004); T. Ecker in Lang, Melz, and Kristoffersson, Value Added Tax and Direct Taxation: Similarities and Differences, 691. Neither international nor European rules forbid international double taxation, see also J. Englisch in Wettbewerbsgleichheit im grenzüberschreitenden Handel mit Schlussfolgerungen für indirekte Steuern (Tübingen: 2008), 766ff; G. Kofler, ‘Double Taxation and European Law: Analysis of the Jurisprudence’, in A. Rust, ed., Double Taxation within the European Union (Alphen aan den Rijn: 2009), 104. 11 In the following, the term VAT is used for simplicity, referring to a national tax embodying the features of VAT—it encompasses GST.

‘White Supplies’ and Double Taxation in Cross-Border VAT Law 769 For the same example but with modifications for the Polish market: in the case of a reseller of electricity that is non-resident in Poland, but in Germany, and purchasing electricity, no reverse-charge rule is implemented for purchasing the electricity according to Polish VAT law as mentioned in article 38 Council Directive. The consequence is that the respective supplier is liable for Polish VAT (by charging and using the car in Poland). The incoming invoice shows Polish VAT separately, which the reseller could normally deduct as input VAT. From a German VAT perspective, the incoming supply is taxable at the residency of the German reseller (in Germany), and the recipient (the reseller) has to declare German VAT via a reverse charge. As a result, double taxation arises because the same supply is taxed in both countries.12 Varying assumptions of permanent establishment cause similar issues. Some EU member states assume that a permanent establishment recorded for corporate income tax purposes also applies to VAT, others automatically determine that any permanent establishment applies. In contrast, yet others place less emphasis on permanent premises and consider them as described and defined in the Council Implementing Regulation13 on the common system of VAT or as decided by the CJEU.14 In practice, local tax authorities should be monitored carefully, as this could be a new field for incongruent EU taxation, for example in the Polish or Romanian jurisdictions. As an example of a simple cross-border case involving a non-EU country, car washing is treated as a supply of service according to the German classification, taxable at the place of residence of the entrepreneurial recipient (based on art. 44 Council Directive). According to Swiss VAT law, a car-washing service is the supply of a new (washed) car with a legally specified place of taxation at the point of ‘transfer of the new car’ (i.e. the Swiss petrol station). If an entrepreneurial recipient resident in Germany washes their vehicle at a Swiss petrol station by means of a German supplier (a middle entrepreneur acting in their own name and using a fuel card when purchasing and selling the supply), German and Swiss VAT must be declared and separately shown on an invoice by the supplier (in the case of a VAT registration, also in Switzerland) and paid by the recipient to them. Further, for the supply-of-service (e.g. car-washing services) reverse-charge rules are not applicable in every EU member state, such as Hungary, Spain, and Portugal. A reverse-charge mechanism is not applicable if the supplier is not resident in those countries but has a VAT registration there, because those countries qualify a non-resident supplier as resident in the case of VAT registration. That is to say, a taxable subject is created at the place of consumption to generate national tax revenue. 12

See also Case C-647/17 Srf konsulterna AB, ECLI:EU:C:2019:195, concerning the place of supply for accounting and management seminars, exploring a conflict between two EU member states applying art. 44 against art. 53 Council Directive 2006/112/EC. 13 Regulation (EU) No. 282/2011 of 15 March 2011, laying down implementing measures for Council Directive 2006/112/EC on the common system of VAT. 14 For initial ideas about the Polish VAT law, see Case C- 547/ 18 Dong Yang Electronics, ECLI:EU:C:2020:350. Further examples are given by T. Ecker, A VAT/ GST Model Convention (Amsterdam: 2013), 39.

770 Heidi Friedrich-Vache To summarize and categorize the different cases of non-or double taxation within the ‘harmonized’ EU VAT system: there may be contrasting rules when determining the place of taxation for VAT, divergences in interpretation of the law and rules, or just incompatible evaluations of the relevant business scenario. But without unified tax administrations, different implementations or interpretations of European law could result as well as different findings or assessments of facts. In a proportion of cases, double taxation arises from variations in a VAT provision contained in the Council Directive where a difference exists in classifying a ‘supply’ of goods or services under articles 31ff and 43ff of the directive. In other cases, double taxation results from different understandings of the situation by the tax administrations of the respective EU member states. For example, double taxation arises when supplies and acquisitions are made within the Community and delivered across borders. If the tax administration of the country of departure does not provide proof of their transport to another EU member state and refuses to exempt the delivery, the administration of the country of arrival considers that it has been provided and so taxes the intra-Community acquisition by the purchaser. This is also not in line with the basic understanding of the preferable-destination principle of having taxation and tax revenue at the place of consumption. From a country-of-origin principle (taxable supply at the beginning of the transport), the country of origin of the goods currently has the right of ‘sanction’ via VAT or penalties, which is considered as economic tax revenue where VAT exemptions are not available for intra-Community supplies. The territorial principle should generate compensatory payments for heavily export- oriented countries in cases of deferrals of VAT (due to, e.g., tax fraud, mistake, or non- fulfilment of VAT-exemption eligibility requirements). In other cases, double taxation results from the incomplete or incorrect implementation of EU rules. This situation can also arise because EU member states have the authority to retain different national laws or rules existing before acceding to the EU or where domestic legal descriptions vary. Language is also a central element—a core research area for philosophers of law.15 Law as a social system of order depends on coordinated behaviour and, consequently, on communication, language, and culture. Concerning the latter, the comparison of laws and their segregation into major legal families, as known from comparative law,16 and which is, for example, also based on the legal culture of countries, has not yet been thoroughly investigated for tax law purposes.17 With regard to language, there is always scope for differing interpretations, especially by local courts and tax authorities.

15 The close connection between philosophy of language and philosophy of law has been recognized for decades through the work of many influential legal philosophers; for details see A. Marmor, The Language of Law (Oxford: 2014). 16 For details, see K. Zweigert and H. Kötz, An Introduction to Comparative Law, 3rd rev. ed. (Oxford: 1998); C. Sacchetto and M. Barassi, Introduction to Comparative Tax Law (Turin: 2008), s. II. 17 See P. Chuenjit, ‘The Culture of Taxation’, Journal of Population and Social Studies 22/1 (2014), 14.

‘White Supplies’ and Double Taxation in Cross-Border VAT Law 771 This issue means that an understanding of a binding decision reached by a VAT committee is necessary for the implementation of a common VAT system (see Section 42.4). However, differences in the interpretation of VAT law are not altogether unexpected where non-EU countries are involved, even if global VAT taxation is based on the European VAT system.

42.3 The Meaning of Harmonization in Indirect Taxation In the EU internal single market with a harmonized indirect taxation system, the CJEU resolves the issue of VAT double taxation through a uniform and binding interpretation of EU law.18 To understand how to resolve the problem of double taxation in the EU (and then worldwide), the legal basis (the primary law) of the European single market in relation to VAT must be reviewed. According to the concept of the internal market, VAT double taxation or unintended non-taxation should not exist in cross-border cases. That concept is often incorrectly confused with the ‘harmonization’ of indirect taxation; however, what is meant is not the harmonization of the tax. Harmonization must be precisely clarified due to the fundamental principles of VAT, especially the neutrality of the tax (that burdens private consumption and consumers once and not with accumulated VAT) and the character of VAT as a net tax at every transaction stage (with the general right to deduct input VAT for business and economic activities). Article 113 TFEU19 (ex art. 93 EC Treaty) grants the right to harmonize VAT laws and laws on other indirect taxes to the extent necessary for the functioning of the European single market. Accordingly, the EU Council of Ministers had adopted various directives on VAT harmonization but the Council can only decide on proposals officially submitted by the European Commission. The TFEU does not permit complete harmonization of all VAT provisions; it allows harmonization only where it appears necessary. VAT harmonization in the EU began with the decision (First Council Directive 67/227/ EWG, 1967) to introduce a VAT system with input VAT deductions in all EU member states and thus to render superfluous lump-sum VAT refunds on exports, which had until then been customary.20 As developments advanced, it soon became apparent that as trade barriers were dismantled, more and more aspects of VAT legislation began to act as a ‘brake’ on further economic integration. Encouraged by the EU decision to make member state VAT revenue a basis for contribution payments, the next step towards harmonizing VAT rules, exemptions, bases, 18

See Kokott, Das Steuerrecht der Europäischen Union, § 2, m.no. 207. [2012] OJ C326/1. 20 See also translated in Kokott, Das Steuerrecht der Europäischen Union, § 2, m.no. 205. 19

772 Heidi Friedrich-Vache and rates was very quickly taken. Thus, many other VAT modalities were included (Sixth Council Directive 77/388/EWG, 1977, with subsequent amendments to Council Directive 2006/112/EC). Subsequently, tax approaches became relatively harmonized except for some politically sensitive areas (e.g. the rates of VAT are still largely selectable).21 The complexity of VAT regulations and the differences in administrative formalities in the individual countries became an increasing problem with economic integration. Efforts to harmonize VAT in recent years have concentrated on both the creation of more readable legal texts in EU law and efforts to reduce the difficulties caused by liaisons with foreign tax authorities. The latter area is also likely to be the focus of further work, seeking harmonization in substantive terms.22 As a result, the obligation to adhere to the European guidelines when drafting national VAT laws also means that the application of legal regulations must not subsequently depart from the framework set out by the EU. Otherwise, legal texts could be standardized but continue to be interpreted and applied differently by each EU member state. For that reason, the question as to what the legislator intended with certain formulations of the law (interpretation of the law) should also be asked regarding what the legislator was obliged to order in accordance with the requirements of the EU. Thus, in the case of doubt, it must be assumed that the legislator intended to impose exactly what they had to impose according to European requirements (interpretation of the respective VAT act in conformity with the directive). If, in a particular case, it is also unclear what the European specifications require with respect to a particular problem, it must be clarified in advance by a request for a preliminary ruling to the CJEU. In practice, this occurs at a stage in the tax procedure where findings or differing assessments have already been made by the local tax authorities, and taxpayers regularly find themselves involving in court proceedings after a negative appeal. The CJEU or, for example, the European Commission cannot currently be asked for an interpretation or final decision by the local tax authorities or taxpayers when discussing the place of supply, the quality of a supply ‘by the way’, or before submitting a declaration. Further, the consequences of double taxation are linked to the performing supplier as they cannot normally pass on the second VAT charge to the recipient.23 Where interpretations of the rules differ, the CJEU can make a binding decision, using the terms in the directive(s) and according to article 267 TFEU. However, double taxation already exists at this stage due to the assessment in a second country. Suppose that a CJEU decision concludes that the second state has applied the VAT Directive correctly. In that case, the taxable person is burdened with VAT in the

21 See

the proposal of the European Commission regarding VAT rates, COM(2018) 20 final (18 Jan. 2018). 22 See the Communication for an Action Plan of the European Commission, COM(2020) 312 (15 July 2020). 23 See H. Stadie in E. Rau and G. Dürrwächter, eds, Umsatzsteuergesetz (UStG), Commentary (Cologne: 20020), Introduction, m.no. 748; also see Englisch, Wettbewerbsgleichheit im grenzüberschreitenden Handel mit Schlussfolgerungen für indirekte Steuern, 765.

‘White Supplies’ and Double Taxation in Cross-Border VAT Law 773 first state due to an unappealable tax assessment. From a national procedural tax law perspective, it is unclear whether allegiance to the EU principle (art. 4 TFEU) will be implemented, so the first state will amend the tax assessment or remit the VAT payment. Where directives or VAT regulations have been applied incorrectly (or not applied), the European Commission can launch an infringement procedure. This option is impractical, given the time and effort required. For straightforward EU cases, the framework is clear: in the Implementing Regulation,24 recital 5 states that tax harmonization measures should include a common cooperation system between the EU member states. In particular, the member states should cooperate with the Commission in information exchanges. Such action will ensure the proper application of VAT on supplies of goods and services. Recital 7 already specifies that for the purposes of collecting the tax owed, the member states should cooperate to help to ensure that VAT is correctly assessed. They must therefore not only monitor the correct application of tax owed in their own territory but should also provide assistance to other member states in order to ensure the correct application of tax relating to activity carried on in their own territory but owed in another member state. Effective supervision of such transactions is therefore dependent on the member state of establishment collecting or being in a position to collect that information (recital 8). Therefore, EU member states already have a direct mandate and an obligation to cooperate with accurate information exchange to establish facts and determine the correct VAT treatment. They must work together to determine the law and the interpretations by local law courts or the tax authorities. At the requesting authority’s request, another jurisdiction should communicate the information referred to in article 1 Regulation No. 904/2010, including information relating to specific cases. The petitioned authority should arrange for administrative enquiries to be conducted and a report forwarded appropriately. Under article 28 Regulation No. 904/2010, it is possible for agreements between the two authorities to be carried out at the offices of the requested member state to exchange the information referred to in article 1 Regulation No. 904/2010. Representatives of the requesting authority may also be present during enquiries. The officials of the requested administration should carry out such investigations with access to the same premises and documents as the latter. The objective of avoiding double taxation must also be considered when interpreting the relevant provisions.25 In one case before the CJEU,26 a Polish company KrakVet with no residency for VAT purposes in Hungary (and with no business space or warehouse in Hungary), sold products via an internet platform to Hungarian customers and discussed the place of taxation with the Polish and Hungarian tax authorities. Since the

24 Council Regulation (EU) No. 904/ 2010 of 7 October 2010 on administrative cooperation and combating fraud in the field of value added tax. 25 See Case C-68/92 Commission v. French Republic, ECLI:EU:C:1993:888, para. 14; Case C-242/08 Swiss Re Germany Holding, ECLI:EU:C:2009:647, para. 32; Case C-249/84 Profant, ECLI:EU:C:1985:393, para. 25; Case C-127/86 Ledoux, ECLI:EU:C:1988:366, para. 11. 26 See Case C-276/18 KrakVet Marek Batko sp.k., ECLI:EU: C:2020:485.

774 Heidi Friedrich-Vache supplier was uncertain as to the member state responsible for collecting the VAT relating to its activities, KrakVet applied to the tax authorities of the EU member state in which it had its registered office for a ruling. By an advance tax ruling, the Polish tax authorities took the view that the place where KrakVets’s transactions were carried out was Poland and that KrakVet had to pay VAT there. The first-tier Hungarian tax authority carried out an inspection in order to verify a posteriori the VAT returns for the year 2012. In that context, KrakVet as a taxable person was assigned a technical tax identification number by that tax authority. Because of the limited information on the company and the way it operated from a tax point of view, the first-tier Hungarian tax authority was not in a position to determine whether KrakVet was established in Hungary for VAT purposes. The tax authority therefore carried out checks in relation to the activities carried on by the company. In the context of those administrative tax proceedings, the first-tier Hungarian tax authority made enquiries to the Polish tax authorities based on the rules of cooperation laid down in EU law. By a respective assessment notice and decision, the first-tier Hungarian tax authority required KrakVet to pay a sum corresponding to the difference in taxation of VAT, a penalty, and default interest, plus a fine for failure to comply with its obligations to register with the Hungarian tax authority; this was also confirmed in the Hungarian second-tier court. The CJEU decided the question of whether the Council directive and articles 7, 13, and 28–30 Regulation No. 904/2010 had to be interpreted as precluding the tax authorities of a member state from unilaterally subjecting transactions to VAT treatment different from those under which they had already been taxed in another EU member state. The CJEU referred to the specific provisions for determining the place of taxable transactions which, according to recitals 17 and 62 Council Directive, aims to avoid jurisdictional conflicts that may result in double taxation or non-taxation.27 The CJEU mentioned that ‘in particular as concerns exchange of information’ is the focus to support a correct VAT assessment. As the EU legislature acknowledges in recital 8, monitoring the proper application of VAT on cross-border transactions taxable in a different EU member state than the location where the supplier is established, depends in many cases on information which is held by the EU member state of establishment, or which can be much more easily obtained by that member state. Therefore, a submission of a request to the tax authorities of another member state is possible. Such a request may prove expedient or even necessary.28 However, as confirmed by the CJEU once again, the Regulation is confined to enabling administrative cooperation for the purpose of exchanging information that may be necessary for the tax authorities of the EU member states. The Regulation does not, therefore, govern the powers of those authorities to carry out, in the light of such information, the classification of the transactions concerned under the directive.29 27

See Case C‑111/05 Aktiebolaget NN, ECLI:EU: C:2007:195, para. 43. See to that effect Case C‑419/14 WebMindLicenses, ECLI:EU:C:2015:832, para. 57. 29 See, by analogy, Case C‑318/07 Persche, ECLI:EU:C:2009:33, paras 62, 63 and the case law cited therein. 28

‘White Supplies’ and Double Taxation in Cross-Border VAT Law 775 In conclusion, the current Regulation does not offer a fast and practical solution for handling EU disputes and double taxation cases. Regulation No. 904/2010 does not obligate the tax authorities of two EU member states to cooperate and reach a common solution for the VAT tax treatment of a transaction. There is also no requirement that the classification applied by one member state binds the tax authorities of another member state. The CJEU considers that double taxation cases only arise due to misinterpretations or misunderstandings of the directive. It believes that the outcome depends on whether national laws provide a judicial remedy to refer requests to the CJEU for a preliminary ruling.30 Thus, even if it is remarkable and astonishing,31 it is not against EU law for the tax authorities of a member state to unilaterally subject transactions to different VAT treatment than those under which they have already been taxed in another state.

42.4 Solutions to Resolve Legislative and Administrative Cooperation VAT Disputes It is possible to imagine different measures to categorize strategies to prevent VAT disputes and double (non-)taxation. As described by many authors, non-coordinated unilateral measures, coordinated unilateral measures—‘the way of the OECD’, bilateral measures (between a taxable person and the tax authority of another state), and multilateral measures (between a taxable person and two tax authorities)—or a mix of these instruments or a separate VAT/GST treaty (as a model tax convention) have already been considered.32 Ismer and Artinger describe various coordinated and binding forms as a perfect fit: unilateral and binding forms (via CJEU, European Commission, specific rules of European legislation), or non-binding forms (e.g. joint audits by different EU member states or tax authorities as a soft law approach).33 30

See Case C‑544/16 Marcandi, ECLI:EU:C:2018:540, paras 64 and 66. And also not desired according to the neutrality principle, as also stated in Case C- 50/ 88 Kühne, ECLI:EU:C:1989:262; Case C- 193/ 91 Mohsche, ECLI:EU:C:1993:203; Case C- 415/ 98 Bakcsi, ECLI:EU:C:2001:136. 32 See Ecker, A VAT/GST Model Convention, 49ff. 33 See Ismer and Artinger, Internationale Doppelbesteuerung im Mehrwertsteuerrecht, 18ff. The European Commission carried out a survey in 2018 as a pilot project for solutions to handle double taxation in the EU to apply for an EU VAT cross-border ruling, https://ec.europa.eu/taxation_customs/ business/vat/vat-cross-border-rulings-cbr_en (accessed 6March 2021). The advantage of such a ruling is that potential double taxation could be resolved before arising because different interpretations will be excluded. However, for that procedure, a taxable person must evaluate and know in detail beforehand that a different interpretation will be given under another national law or the respective opinions of the tax authorities. 31

776 Heidi Friedrich-Vache As already seen, the CJEU as a supranational court for the European taxation system could guarantee the uniformity of the law within the EU and give guidance in the case of requests by the EU member states. But the Court decides single cases and seems to be increasing its case law. Court proceedings incur time, cost, and effort,34 even if the taxable person cannot go directly to the CJEU. Court procedures are not an appropriate general method for resolving taxation- handling disputes. The German Federal Finance Court has suggested establishing rules to avoid double taxation for specific scenarios within the scope of article 267 TFEU;35 however, the CJEU36 did not follow these recommendations. When the CJEU faced a question on interpretation concerning a double taxation conflict, it first evaluated the facts in the main proceedings on which the question was based. The process resulted in the double taxation issue being omitted. As a second overarching instrument, the European Commission has the right to bring infringement proceedings according to article 258 TFEU. If the Commission considers that an EU member state has failed to fulfil a Treaty obligation, it can deliver a reasoned opinion after allowing the state to submit observations. If the state concerned does not comply with the view within the period specified by the Commission, the latter can bring the matter before the CJEU. That is not an efficient or practical solution in terms of process and timing. A flexible and more straightforward option is required to handle EU double taxation between either the EU member states or between the member states and a non- EU country. Treaties may not be the answer since only one supply or transaction will be taxed and revenues will not be divided between countries. This concept is often the reasoning behind bilateral agreements on direct taxation. The advantage of this may be consistent definitions of the supply of goods or services and the place of consumption described within a treaty. Concrete definitions and standardizations must be included in the existing Treaties, within the Council Directive, or ideally within the directly binding Council Implementation Regulation. This process was implemented previously for services supplied electronically or linked directly to immovable property. To summarize, the potential starting point to avoid double or non-taxation could be to add substantial legal provisions, such as the following. (1) Additional clear definitions of supplies and places for international and EU cross- border transactions within the Council Implementing Regulation. (2) A backup rule within the binding framework, since national tax authorities sometimes implement it stating that ‘an EU Member State will accept another country’s VAT evaluation where there are divergences’. According to the German tax authority, it will 34 See,

e.g., also European Commission, Consultation Paper of 5 January 2007, Introduction of a mechanism for eliminating double imposition of VAT in individual cases,https://ec.europa.eu/taxat ion_customs/system/files/2016-09/double_taxation_en.pdf (accessed 6 June 2022). 35 See order of the German Federal Finance Court of 22 February 2001, case no. V R 26/00. 36 See Case C-185/01 Auto Lease Holland BV, ECLI:EU: C:2003:73, para. 23.

‘White Supplies’ and Double Taxation in Cross-Border VAT Law 777 abandon the right to taxation in some scenarios; for example, in a cross-border leasing transaction, if the asset is allocated to a German-domiciled contracting party based on another EU member state’s laws, it will follow that ruling to avoid tax losses. German tax authorities will follow allocations of a taxable charge in another EU member state after being provided with proof that the transfer has been taxed elsewhere. To determine where the tax revenue exists, an additional backup clause is required if there are doubts about the destination or customer’s residency. (3) Based on the OECD Model Tax Convention, a unique treaty for VAT purposes could be agreed between the EU and non-EU member states, including issues of neutrality and reciprocity and a harmonized destination principle. The OECD has developed non-binding International Guidelines for Value Added Taxation to counteract double taxation and avoid inconsistent application of the destination principle in order to recognize the correct tax at the final consumption destination.37 Due to European regulations (e.g. art. 59a Council Directive), specific supplies of services will currently not be taxed in the EU where the non-entrepreneurial recipient is resident in a non-EU country. In that event, non-taxation is acceptable, depending on the intention at the place of consumption (and the assumption that there is no comparable VAT taxation). Article 59a Council Directive could be considered a unilateral and binding instrument (for a small segment).38 The guidelines are presented as soft law and may encourage OECD and non-OECD countries to adapt common rules to define the place of consumption, either based on the customer’s residency or at the site of consumption or utilization. This process is challenging for B2C businesses since private customers cannot declare local VAT for the supplier to produce a reverse charge.39 But this issue already exists, and policymakers in the destination country must be responsible for developing a legal and administrative framework. Although the OECD guidelines have been criticized as simple and incomplete,40 they are generally accepted. In addition, (Base Erosion and Profit Shifting (BEPS) Action Plan 141 requests rules that implement International Guidelines for Value Added Taxation; therefore, these must be pursued.42 (4) In a double taxation case, the non-EU countries concerned could reach a common understanding by referring to the Mutual Agreement Procedure direct taxation content (art. 25 OECD Model Tax Convention on Income and Capital). The agreement would

37 See OECD, ‘International VAT/GST Guidelines’ (2017), https://read.oecd-ilibrary.org/taxation/ international-vat-gst-guidelines_9789264271401-en#page10 (accessed 7 March 2021). 38 The same could be said for art. 36 para. 2 Council Directive with the aspiration to avoid intra- Community double taxation for goods supplied for installation services. 39 Note that illustrations exist, e.g. in Switzerland, related to specific transactions, including share deals. In these examples, private customers can handle taxation for the supplier via a reverse charge. 40 See, e.g., M. Lamensch, ‘International– OECD Draft Guidelines on VAT/GST on Cross-Border Services’, International VAT Monitor (2010), 271, 274ff. 41 See OECD/G20, Base Erosion and Profit Shifting Project, Addressing the Tax Challenges of the Digital Economy, Action 1—Final Report 2015. 42 See K.- H. Haydl, ‘Seminar C: Cross-border Supply of Services and VAT/GST’, Internationales Steuerrecht (2015), 587, 588.

778 Heidi Friedrich-Vache dictate in which country the VAT is levied. Currently, this is impossible as no business has the right to request the analogue application of the double tax treaty articles. The agreement is constructed with varying concepts relied on to allocate a taxation location. For indirect taxation and the territoriality principle, written rules must be based on the OECD Model since many countries with VAT systems (comparable to the EU structure) are OECD members. Ecker also presents issues to consider when developing potential VAT–GST treaties.43 Ismer and Artinger adjusted that suggestion with an obligatory arbitration procedure for final settlement.44 The solution would be that the taxable person burdened with double taxation should contact the local tax office and that authority should try to resolve the double taxation situation independently. If no solution is found, the EU member state should inform the other member states within a short time to discuss the issue bilaterally. An agreement procedure could be limited to two years, with a six-month arbitration period. That time frame is relatively long from the perspective of the taxable person but is somewhat brief for facilitating discussions between the respective tax authorities. The problem is the individual status of assessments or having ‘open cases’ from a procedural law perspective. The alternative should be (as described earlier and later) to implement a framework that enables the EU member states to manage cases independently. (5) In 2007, the European Commission launched an online public consultation to ascertain views on the possible introduction of a VAT dispute-resolution mechanism. In the meantime, it published an Action Plan for Fair and Simple Taxation supporting the Recovery Strategy.45 Specifically, the Commission proposes to start negotiations on administrative VAT cooperation agreements with relevant non-EU countries (similar to the agreement between the EU and Norway), initially with the EU’s main trading partners. In the case of disputes and to ensure legal certainty between taxpayers and tax administrations, taxpayers should still have the opportunity to correct or clarify possible misunderstandings to avoid conflict-escalation. This expedited method would keep costs for taxpayers and administrations to a minimum. In cooperation with the EU member states, the Commission should continue to work on forming a permanent dispute-resolution body, or a standing committee, for which Directive 2017/1852 has already provided the legal basis. The standing committee would contribute to effective dispute resolutions between the EU member states. Concrete solutions are preferable but have not yet begun. However, they could be implemented easily as different types of ‘mediation’ already exist in other tax areas (see later).

43 See

T. Ecker, ‘Tax Treaties— A Solution to VAT/ GST Double Taxation’, in Lang, Melz, and Kristoffersson, Value Added Tax and Direct Taxation, 691. 44 See Ismer and Artinger, Internationale Doppelbesteuerung im Mehrwertsteuerrecht, 12, 21; C. Fischer and M. Bopp in T. Ecker, M. Lang, and I. Lejeune, eds, The Future of Indirect Taxation (Alphen aan den Rijn: 2012), 512, with descriptions of the Swiss way of handling such a procedure, for South Africa in the same book see J Roeleveld and C. De Wet, 481. 45 See COM(2020) 312 (15 July 2020).

‘White Supplies’ and Double Taxation in Cross-Border VAT Law 779 (6) One suggestion is to create agreements between the respective EU member states, facilitated via the VAT Committee according to article 398 Council Directive. The VAT Committee can check (e.g. if an application is made by a member state or a taxable person) queries connected to the execution of VAT rules. The European Commission acknowledges that solution, with an interim judicial process supported via a VAT Committee.46 The EU member states must follow up on this idea. The VAT Committee’s competencies must be extended from non-binding advisory decisions to binding decisions. Currently, according to the tax sovereignty of the member states, national tax authorities could impose rulings following an application from a taxable person. Article 113 TFEU allows the Council, acting unanimously and following a special legislative procedure and after consultation with the European Parliament and Economic and Social Committee, to adopt provisions for harmonizing member states’ indirect taxation rules.47 Either the EU member states agree via the European Parliament on a transforming status with binding competence (as a ‘comitology committee’)48 or the European Commission, representing and defending the interests of the European Union, prepares new legislative proposals and is responsible for Treaty applications. It must decide on a professional dispute presentation, whether the conflict is between the EU member states or one member state and a non-EU country with an EU-based supplier. This legitimation must be complemented with respective procedural law. Creating a procedural law via a VAT Committee is complex but attainable. A definition is needed to determine who can initiate applications to the VAT Committee; this could be the taxable person (potentially in an open case) who is subject to double VAT in two member states.49 Appeal processes must be available as an administrative tool to contest a binding decision. The legal precedent could be to progress the case to the CJEU if necessary. Before this is possible, member states must cooperate with administrative dialogue between the tax authorities, and this must be increasingly extended in the future.50 The tax administrations’ dialogue (where the dispute in question follows the performance of supply of a service) and/or international cooperation measures—as a dialogue solution (via SOLVIT or a new Double Taxation Dialogue for VAT purposes)—should be the first and preferred starting point which could pre-empt further procedures.

46 SOLVIT as a further tool aims to find solutions within ten weeks—starting on the day a case is taken on by the SOLVIT Centre in the country where the problem occurred, https://ec.europa.eu/sol vit/what-is-solvit/index_en.htm (accessed February 2021). SOLVIT can help when the EU rights of a citizen or business are breached by public authorities in another EU country and the case is not taken to court (in an administrative appeal, assistance is possible). From a procedural tax-law perspective with desired binding effect and initiated by a taxable person, this should be used—in this way, tax authorities and taxable persons could contact the VAT Committee. It is an institutional solution with regard to legal disputes after a supply not beforehand and is not applicable for different assessments of facts. 47 COM(2018) 813 final (12 Dec. 2018). 48 Delegating powers should generally be possible under art. 291 TFEU. 49 It is also conceivable that the VAT Committee could act as a mediator in identifying solutions. 50 See COM(2020) 312 final (15 July 2020).

780 Heidi Friedrich-Vache That dialogue should be between the tax authorities, which should include the taxable person. European law should implement a common understanding that an EU member state will rescind the right to taxation where the customer is not a resident or final supply of services is not performed. This idea should be supported: (1) by a unique standardized confirmation letter from the country which has already taxed the supply for presentation in the other country which also wishes to tax.51 Generally, the country of the supplier’s residency should first review the case and the possible double taxation and decide about location of final consumption following its national VAT law (the supplier is the ‘VAT collector’ for the tax authorities; but if another country assesses a divergent location, that should be handled via a confirmation letter). In addition, the idea should be supported (2) by more concrete definitions of supplies and their places of taxation; (3) by rules via the Council Implementation Regulation for the EU market in the case of doubts about having taxation at the place of final consumption or the place of residency of the recipient; and (4) by European VAT Committee reviews, as mentioned earlier. Therefore, within the EU, no ‘treaties’ would be necessary. As currently, regulations for the place of supply must be better thought out52 and any respective additions must be made urgently; that is, a uniform law (via the directive or Implementation Regulations within the EU) is required as a first step. In any case, a solution must be practicable, binding, and not time-consuming and costly, otherwise entrepreneurs in acting as taxable persons will accept double taxation (depending on the amount, as they could try to forward those costs on to their customer). Even if no survey has yet been conducted—according to information orally given by the European Commission—on how often and to what extent double taxation occurs (in the EU), solutions must be found. The binding solution previously discussed could be a task of the VAT Committee: if two tax authorities assess the VAT or render contrary rulings, the European Commission (due to absence of a superior tax authority) should have the right to annul the tax assessment of an EU member state.53 Alternatively, an arbitration court (e.g. as with 51 The

German tax authorities took this practical approach when a simplification rule applied to a case of consignment stocks in another EU member state. Before legal changes brought about by the quick fixes (with a respective confirmation letter), the other EU country confirmed the application for simplification rule in the country of the consignment stock. The taxable person resident in Germany provided the letter to the tax office in charge and the tax authorities accepted the divergent taxation. To be fair, this was a case handling a VAT-exempt intra-Community supply of goods directly to the customer instead of a VAT-exempt intra-Community transfer of own goods in Germany; it was therefore not a taxed supply. But there should be no valid reason not to transfer that approach to a taxed supply. 52 E.g. the place of consumption in the case of the supply of electricity for e-cars; here, a common rule must define the place at the charging point (and not in every nation of travel thereafter). This method would simplify the assumption of the residency of a private customer after purchasing electronically supplied services. 53 See Administrative Guidance on the Binding Tariff Information Process (BTI) (21 Dec. 2018), https://taxation-customs.ec.europa.eu/system/files/2019-04/bti_guidance_en.pdf (accessed 4 August 2022) with reference to strict deadlines on actions connected to the handling of applications, consultations between member states, and issuing of binding tariff information (BTI) decisions. If member state A agrees with the argumentation of member state B and accepts that the BTI decision is incorrect, it will revoke the decision and issue a new one upon application, following the classification

‘White Supplies’ and Double Taxation in Cross-Border VAT Law 781 corporate income tax) could be possible or (as with cases of non-binding social insurance) a mediation board as a European agency.54 These alternatives should be designed as a procedure with a limited period in which to reach a result (maximum one year). Each alternative have advantages and disadvantages but should all have corresponding procedural and binding consequences for VAT purposes, which would also be legally actionable. In my opinion, this should be the solution for disputes after performing a supply—after a failed dialogue within a time period of one year—which could be described as the ‘customs tariffication solution’ with competence of the European Commission to repeal one binding decision (for a respective VAT assessment in the member state). For queries and potential double taxation before a supply of services has been performed, a solution with binding effect has already been found in the so-called VAT Cross Border Ruling (CBR);55 however, only eighteen EU member states have joined the project. The participation should be a must for all EU member states as the idea, according to the principles, is to have taxation only once in one country—therefore, a solution for pre-performance of a supply of services with potential double taxation in the EU could easily be achieved.

42.5 A Definitive EU VAT System to Offer Full Harmonization and the Prevention of Double Taxation As described earlier, the current legal framework, with its range of interpretations of supplies and their places of taxation, is not fully suitable for avoiding double or non-taxation. The legal solution within the EU is to define one place of taxation, only the place of consumption (i.e. the residency of the recipient of a service or the final destination of

suggested by member state B. All member states and the Commission should also be informed of the decision, via CIRCABC. On the other hand, if member state B accepts that the BTI decision issued by member state A is, in fact, correct, it can proceed to issue a BTI decision in conformity with the existing decision issued by member state A. However, if the two member states cannot come to an agreement, member state B should formally inform both member state A and the Commission that it wishes to submit the matter to consultation at Union level. 54 See Regulation (EC) No. 987/ 2009 of the European Parliament and of the Council of 16th September 2009 laying down the procedure for implementing Regulation (EC) No. 883/2004 on the coordination of social security system. 55 See European Commission, CBR and EU VAT Forum, https://ec.europa.eu/taxation_customs/ taxation-1/value-added-tax-vat/vat-cross-border-rulings-cbr_de and https://ec.europa.eu/taxation_cust oms/taxation-1/value-added-tax-vat/eu-vat-forum_de (accessed 6 June 2022).

782 Heidi Friedrich-Vache delivered goods). For that, definitions of supplies of goods and services should be clearly and legally described within the Council Implementing Regulation. The European Commission has consistently pressed for VAT system reform of the legal framework. The Commission’s 2016 ‘Action Plan on VAT—Towards a single EU VAT area—Time to decide’56 announced its intention to propose a definitive VAT system based on the principle of taxation at the destination (i.e. where the goods or services are consumed). A package of measures substantially modifies the rules relating to VAT, ending thirty years of a ‘transitional’ VAT regime in the single market, applicable since 1 January 1993.57 Once agreed, the amendments contained in the VAT rules will define the cross- border trade of goods as a single taxable supply which will ensure that goods are taxed in the member state where the transport of the goods ends—as it should be. The proposal finalizes this artificial split of a single commercial transaction. Technically, to make the change to the VAT rules as seamless as possible for businesses, amendments would introduce an online portal or ‘one-stop shop’ for all EU B2B traders. That provision would clarify the VAT approaches as per the European Commission’s reform proposals in October 2017.58 Following the adoption of the Action Plan, the European Commission made a series of proposals to work towards completion, as this plan will not be actionable. The first steps have been the so-called quick fixes, especially focusing on the active use of a European VAT identification number. The latter will be necessary and is, therefore, in my opinion, a preparatory measure for businesses to be as familiar as possible with its use within the European single market and to be able to allocate tax revenue. The identification number will be linked to the place of consumption (defined by the supplier’s residency or destination of the goods). In the meantime, the Commission published a study in July 202259 and a proposal, ‘VAT in the Digital Age’, in which a single VAT registration is preferred and containing a general reverse-charge mechanism within B2B businesses and with definitions and rules for the sharing and platform services. To establish the place of the supply of goods or services in B2B or B2C businesses without specific supply rules (e.g. as currently in the EU VAT system), other organizations outside the place of consumption would no longer encounter conflicts or contending powers when determining the correct location of supply. Different tax authorities or court interpretations and evaluations of facts or opinions of the national tax authorities could become obsolete since they would lead to the 56 COM(2016) 148 final (7 Apr. 2016), with the follow-up to the VAT Action Plan ‘Towards a single EU VAT area—Time to act’, COM(2017) 566 final (4 Oct. 2017). 57 See European Commission, Press Release of 25 May 2018, https://ec.europa.eu/commission/pres scorner/detail/de/IP_18_3834 (accessed 1 January 2021). 58 The second part of the e- commerce package has already been implemented, which especially concerns EU B2C traders and further supplies of services to establish taxation at the place of consumption (the residency of the private consumer or non-entrepreneur). 59 See report on the public consultation at https://taxation-customs.ec.europa.eu/package-fair-and- simple-taxation_en (accessed 24 July 2022).

‘White Supplies’ and Double Taxation in Cross-Border VAT Law 783 same place of taxation in every case. Further rules within the Council Implementing Regulations, as direct binding EU law, could pre- emptively resolve divergent classifications of goods or services, for example for leasing agreements or installation supplies.

42.6 Conclusion According to the intention of VAT, there should be only one country of taxation, which must be at the place of consumption. Therefore, the link to the country where a ‘supply’ is consumed is relevant.60 Nevertheless, the number of VAT double taxation cases in cross-border businesses has risen in the last few decades, causing a notable impact.61 Double taxation leads to issues, especially if the recipient of the supply is not entitled to receive input VAT deductions (including within B2B). Double taxation must be avoided among the EU member states as prices increase and supplies are not competitive. To prevent double taxation, non-taxation, or distortion of competition, the Council directive has already implemented some rules in one small area that determine the place of taxation outside the EU when services are performed to non-resident private customers resident outside the EU. The CJEU has also provided references and general limitations regarding solutions for discussions between the EU member states and their various claims and entitlements to VAT revenue. This structure is not enough at the current stage. Given that there is already a common legal framework within the EU, smart solutions are conceivable and attainable (i.e. adding the framework and the competence of specific European institutions, which is possible as the infrastructure of the VAT Committee, SOLVIT, and the Cross Border Ruling is available and transferable). In summary, the focus—and my objective—is to maintain the VAT system as a taxation mechanism only at the predefined place of consumption; to define it more concretely with many examples and facts; and to define more places for the supplies in the new and sharing economy.62 In the meantime, further actions are required to reach feasible and coordinated solutions.

60

See H. Stadie in Rau and Dürrwächter, Umsatzsteuergesetz (UStG), Introduction, m.no. 730. Another opinion V. Kluge, Das Internationale Steuerrecht (Munich: 2001), m.no. D 6. See further, e.g., the ECJ jurisdictions mentioned in P. Rendahl, P, ‘EU VAT and Double Taxation: A Fine Line between Interpretation and Application’, Intertax 41/8/9 (2013), 450; M. Lang et al. eds, CJEU—Recent Developments in Value Added Tax (Vienna: 2016); OECD, Consumption Tax Trends 2016, VAT/GST and excise rates, trends and policy issues (Paris: OECD Publishing, 2016), 14. 62 See in that regard, e.g., the ideas in EU, ‘VAT in the Digital Age, Final report, vol. 2: The VAT treatment of the platform economy’ (2022), https://op.europa.eu/en/publication-detail/-/publicat ion/7 181e45d-0968-11ed-b11c-01aa75ed71a1/language-en/format-PDF/source-262562115 (accessed 4 August 2022). 61

784 Heidi Friedrich-Vache The Council Implementing Regulation provides supplementary definitions of specific statutory provisions for supplies within the scope of the Regulation. EU member states must accept confirmation letters, delivered in a specific EU form, from other member states (or other states as a goodwill gesture) to simplify administrative procedures. This form can confirm whether VAT taxation has already occurred, or will occur, before other foreseen assessments. Furthermore, mediation around binding VAT Committee decisions after the point of application (outside or before court proceedings) must be permitted and (only) at the end of that procedure should the CJEU conduct a review and deliver a final decision where there is an appeal against the VAT Committee outcome. This procedure is comparable to that in relation to differences of opinion on the tariff classification of specific goods (Binding Tariff Information Process with the deadlines for that procedure); therefore such a procedure would not be completely new and is transferable for VAT, as customs duties are levies on goods imported into the EU from third countries. As the European Union is also a customs union, no customs duties are levied on goods traded between member states in the single market. So, the first step should be that the enquiring member state consults the other member state in order to seek further information on the supply and to try to reach a solution between them. If no agreement can be reached, the enquiring member state should request consultation at Union level by sending a substantiated and complete submission to the Commission. Once an opinion on the classification of a specific type of supply and place of its taxation has been rendered at Union level, no binding decision should be issued contrary to that opinion and that opinion should be respected by all EU member states.

Chapter 43

A C omparison Bet we e n EU VAT L aw and t h e OE C D International VAT /G ST Gu ideli ne s Eleonor Kristoffersson

43.1 Introduction The EU, now consisting of twenty-seven member states,1 started its value-added tax (VAT, also called GST or consumption tax) journey in 1967 when the first and the second VAT Directives were adopted.2 At that time, it was the European Economic Community with only six members, namely Belgium, France, Italy, Luxemburg, the Netherlands, and West Germany. Over the years, the EC/EU common VAT system has become increasingly sophisticated but also more and more complex. The Court of Justice of the European Union (CJEU) has delivered more than 800 VAT judgments. The most recent development is the realization of the EU Commission Action Plan on VAT, which deals with e-commerce and small and medium-sized enterprises, tackling the VAT

1

Austria, Belgium, Bulgaria, Croatia, the Republic of Cyprus, the Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, the Netherlands, Poland, Portugal, Romania, the Slovak Republic, Slovenia, Spain, and Sweden. 2 First Council Directive of 11 April 1967 on the harmonization of legislation of Member States concerning turnover taxes (67/227/EC) and the Second Council Directive of 11 April 1967 on the harmonization of legislation of Member States concerning turnover taxes. Structure and procedures for application of the common system of value added tax (67/228/EC) (Second VAT Directive).

786 Eleonor Kristoffersson gap, a definitive VAT regime for cross-border trade of goods, and more freedom for the member states regarding VAT rates.3 The OECD was founded on 30 September 1961 and thus draws experience from sixty years of economic policymaking. In the field of VAT, the OECD Committee on Fiscal Affairs launched a project to develop international VAT/GST guidelines in 2006. The project was based on the assumption that jurisdictions would benefit from an internationally agreed standard that contributed towards ensuring that VAT systems interacted consistently so that they facilitated rather than distorted international trade. The Guidelines were incorporated in the Recommendation on the Application of Value Added Tax/Goods and Services Tax to the International Trade in Services and Intangibles, which was adopted by the Council of the OECD on 27 September 2016. The OECD has thirty-eight members,4 but VAT is applied in more than 170 countries.5 There are more OECD members than EU member states; not all OECD members are EU member states and vice versa. Bulgaria, Croatia, Cyprus, Malta, and Romania are EU member states but not OECD members. Australia, Canada, Chile, Colombia, Costa Rica, Israel, Japan, the Republic of Korea, Mexico, New Zealand, Norway, Switzerland, Turkey, the UK, and the USA are OECD members but not EU members. The USA is the only OECD country that does not have VAT. The USA is thus not directly concerned by the OECD International VAT/GST Guidelines, but only indirectly due to the fact that most of its trading partners apply the principles laid down in the Guidelines in their taxation on cross-border supplies.6 Against this background, this chapter examines the relationship between EU VAT law and the OECD International VAT/GST Guidelines and makes a comparison between them. The chapter focuses mainly on business-to-business (B2B) transactions. Section 43.2 describes and compares the aim and legal status of the EU VAT Directives and OECD International VAT/GST Guidelines. Section 43.3 discusses the core features of VAT including neutrality. Section 43.4 compares the place of taxation on services and intangibles under EU law and the International VAT/GST Guidelines. The chapter ends with some concluding remarks.

3 European Commission, Communication from the Commission to the European Parliament, the Council and the European Economic and Social Committee on an action plan on VAT, ‘Towards a Single EU VAT Area—Time to Decide’, COM(2016) 148 final (7 Apr. 2016). 4 Austria, Australia, Belgium, Canada, Chile, Colombia, the Czech Republic, Costa Rica, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Israel, Italy, Japan, the Republic of Korea, Latvia, Lithuania, Luxembourg, Mexico, the Netherlands, New Zealand, Norway, Poland, Portugal, the Slovak Republic, Slovenia, Spain, Sweden, Switzerland, Turkey, the UK, and the USA. 5 OECD, ‘International VAT/GST Guidelines’ (2017), Foreword, https://www.oecd-ilibrary.org/sites/ 9789264271401-en/index.html?itemId=/content/publication/9789264271401-en&_csp_=5aeda55c6b13e 05a90961dd0a07e26eb&itemIGO=oecd&itemContentType=book#fr (accessed 20 March 2021). 6 See W. Hellerstein, ‘A Hitchhiker’s Guide to the OECD’s International VAT/GST Guidelines’, Florida Tax Review 18 (2016), 591.

EU VAT Law and the OECD VAT/GST Guidelines 787

43.2 Aim and Legal Status The EU is based on international agreements between the member states. The EU Treaties transfer authority from the member states to the EU and its institutions. In the VAT area, the division of competences emanates from article 4 of the Treaty of the Functioning of the European Union (TFEU). When the competences in an area of law are shared between the EU and its member states, as in the field of VAT, the principles of loyalty, subsidiarity, and proportionality apply.7 The fundamental provision for establishing EU VAT is article 113 TFEU, under which the EU institutions shall adopt provisions for the harmonization of legislation concerning turnover taxes to the extent that such harmonization is necessary to ensure the establishment and the functioning of the internal market and to avoid a distortion of competition. Beside these two explicit aims of the harmonization of VAT, the underlying objective is to ensure the free movement of goods and services.8 The substantive EU VAT law is harmonized through directives. The main legal act is Council Directive 2006/112/EC of 28 November 2006 on the common system of value added tax (VAT Directive). The VAT Directive has undergone several amendments since its introduction, most recently following the realization of the EU Commission Action Plan on VAT. A directive is binding, as regards the result to be achieved, for all member states but it is up to the member states to choose the form and the methods.9 The member states have implemented the VAT Directive in their national laws; thus, the national laws are the main binding legal source for VAT in the member states. Only if the member states have not done their job and correctly implemented the VAT Directive, will the taxpayers in the member states need to rely on the binding force of the VAT Directive through the principle of direct effect.10 The OECD International VAT/GST Guidelines aim at reducing the uncertainty and risks of double taxation and unintended non-taxation that result from inconsistencies in the application of VAT in a cross-border context. They do not aim at detailed prescriptions for national legislation. They intend to assist policymakers in their efforts to evaluate and develop the legal and administrative framework in their jurisdictions, taking into account their specific economic, legal, institutional, cultural, and social circumstances, and practices. The Guidelines respect the sovereignty of jurisdictions and their right to design and apply their laws.11 The OECD International VAT/GST Guidelines are non-binding soft law. They do, however, have the potential to achieve an increased coordination between OECD 7

Articles 4–5 TEU. E. Kristoffersson and P. Rendahl, Textbook on EU VAT, 3rd ed. (Uppsala: Iustus förlag, 2020), 24–25. 9 Art. 288 TFEU. 10 Case 41/74 Yvonne van Duyn v Home Office, ECLI:EU:C:1974:133. 11 VAT/GST Guidelines 2016, Preface. 8 See

788 Eleonor Kristoffersson members’ legal systems.12 When the Guidelines were adopted, it was expected that they might have an impact on the judgments of the CJEU.13 So far, the CJEU has not referred to the Guidelines in any VAT case. Since OECD members are in a minority in the global VAT community with more than 170 jurisdictions, it is relevant to ask the question whether the Guidelines have any effect for non-OECD members.14 Indeed, non-OECD members from about 100 jurisdictions were invited to give their input to the draft guidelines at four global forums on VAT.15 Their involvement resulted in the recognition that the reverse-charge mechanism is only recommended when it is consistent with the overall design of the national consumption tax system.16 It may also have had some impact on the wording of Guideline 3.5 on the place of taxation for on-the-spot supplies.17 Like OECD members, non-OECD members have no legal obligation to follow the Guidelines. As opposed to OECD members, non- OECD members are not expected to act loyally in relation to the OECD.18 For practical reasons and to facilitate the VAT treatment of cross-border transactions, the Guidelines may also assist policymakers outside the OECD. Although they are not binding for anybody, the Guidelines may be relevant for most jurisdictions with VAT.

43.3 Core Features of VAT According to the preamble to the VAT Directive, a VAT system achieves the highest degree of simplicity and neutrality when the tax is levied in as general a manner as possible and when its scope covers all the stages of production and distribution, as well as the supply of services. The core features of VAT are laid down in article 2 of the VAT Directive. Under that provision, the principle of the common system of VAT entails the application on goods and services of a general tax on consumption exactly proportional to the price of the goods and services, although many transactions take place in the production and distribution process before the stage at which the tax is charged. On each transaction, VAT, based on the price of the goods or services at the rate applicable to such goods or services, is chargeable after a deduction of the amount of VAT borne directly by the various cost components. The common system of VAT is, under article 2 of the VAT Directive, applied up to and including the retail trade stage.

12

M. Lamensch, ‘Soft Law and EU VAT’, World Journal of VAT/GST Law 5/1 (2016), 9–31.

13 Ibid. 14

K. James and T. Ecker, ‘Relevance of the OECD International VAT/GST Guidelines for Non-OECD Countries’, Australian Tax Forum 32 (2017), 320. 15 OECD Global Forum on VAT, 7– 8 November 2012, Paris; 17–18 April 2014, Tokyo; and 5–6 November 2015 and 12–14 April 2017, Paris. 16 See James and Ecker, ‘Relevance of the OECD International VAT/GST Guidelines’, 337. 17 Ibid. 18 Art. 2 of the OECD Convention.

EU VAT Law and the OECD VAT/GST Guidelines 789 Regarding cross-border trade, EU VAT law has undergone many changes since the introduction of the VAT Directive. In relation to the OECD International VAT/GST Guidelines, the most relevant source is Council Directive 2008/8/EC of 12 February 2008 amending Directive 2006/112/EC as regards the place of supply of services. This directive introduced the main rule, that B2B supplies of services should be taxed in the state of the destination and subject to reverse charge.19 The preamble to the directive states that the place of supply for all supplies of services should, in principle, be where the actual consumption takes place. To further the correct application, every taxable person identified for VAT purposes is obliged to submit a recapitulative statement of the taxable persons and the non-taxable legal persons identified for VAT purposes to whom they have supplied taxable services which fall under the reverse-charge mechanism. The core features of VAT are described in Chapter 1 of the OECD International VAT/ GST Guidelines. The overarching purpose of VAT is to impose a broad-based tax on consumption. It should be distinguished from excises targeted at specific forms of consumption such as the purchasing of alcohol. The central design feature of VAT is that it is collected through a staged process; where each business in the supply chain takes part in the process of controlling and collecting the tax and remitting the proportion of tax corresponding to its margin; that is, on the difference between the VAT imposed on its taxed inputs and the VAT imposed on its taxed outputs, the value added at each stage is taxed. In cross-border trade, the tax should ultimately be levied only on the final consumption that occurs within the taxing jurisdiction, which is an expression of the destination principle. In this respect, the destination principle is used as a jurisdictional principle allocating the taxing rights to the jurisdiction where the final consumption occurs.20 According to the Guidelines, the destination principle in VAT achieves neutrality in international trade. In the actual guideline on the place of taxation for services and intangibles in Chapter 3 of the OECD International VAT/GST Guidelines, the destination principle is used as a proxy for deciding the place of supply. Chapter 1 of the OECD Guidelines refers to the generally accepted principles for VAT policy, the so-called Ottawa Taxation Framework Conditions, that were welcomed by ministers from across the globe in 1998.21 These generally accepted principles are neutrality, efficiency, certainty and simplicity, effectiveness, and fairness and flexibility. Neutrality is further described by the OECD in Chapter 2 of the Guidelines. There are six guidelines on neutrality. Under Guideline 2.1, the burden of VAT should not lie on taxable businesses except where explicitly provided for in legislation. Guideline 19

VAT Directive, new art. 44. M. Senyk, ‘Territorial Allocation of VAT in the European Union. Alternative Approaches Towards VAT Allocation in the Internal Market’, Thesis dissertation, Lund University, School of Economics and Management, Department of Business Law (2018), 319. 21 The Ottawa Taxation Framework Conditions were welcomed by ministers at the Ministerial Conference on Electronic Commerce held in Ottawa, 7–9 October 1998, see OECD, Taxation and Electronic Commerce— Implementing the Ottawa Taxation Framework Conditions (Paris: OECD Publishing, 2001), https://www.oecd.org/tax/consumption/Taxation%20and%20eCommerce%202001. pdf (accessed 21 March 2021). 20

790 Eleonor Kristoffersson 2.2 states that businesses in similar situations carrying out similar transactions should be subject to similar levels of taxation. According to Guideline 2.3, VAT rules should be framed in such a way that they are not the primary influence of business decisions. Guidelines 2.4–2.6 are designed for cross-border trade. With respect to the level of taxation, foreign businesses should not be disadvantaged or advantaged compared to domestic businesses in the jurisdiction where the tax may be due or paid (Guideline 2.4). Furthermore, the Guidelines mention different methods for how to ensure that foreign businesses do not incur irrecoverable VAT (Guideline 2.5). Finally, where specific administrative requirements for foreign businesses are deemed necessary, they should not create a disproportionate or inappropriate compliance burden for the businesses (Guideline 2.6). The core features of VAT in the OECD International VAT/GST Guidelines are in harmony with EU VAT. There are differences in terminology, since EU VAT law identifies the place of supply, whereas the Guidelines aim at identifying the place of final consumption and in that way are able to identify the taxing jurisdiction. Since only the supply of services, and not the acquisition of services, is subject to tax under the VAT Directive, the place of supply normally coincides with the taxing jurisdiction. However, the different terminology results in two different approaches to the principle of destination: in EU law, a proxy for the place of supply and, in the Guidelines, as a jurisdictional principle. In my opinion, the Ottawa Taxation Framework Conditions are generally accepted as benchmarks for a good VAT system22 to such an extent that they hardly need to be mentioned in the context of EU VAT. Efficiency, effectiveness, fairness, and flexibility are not explicitly mentioned in the context of the VAT Directive but must still be considered as parts of EU VAT policy. The same applies to the neutrality aspects in Chapter 2 of the OECD Guidelines. In EU VAT, the principles of non-discrimination and neutrality in VAT have guided the gradual harmonization of EU VAT over the last fifty years and can be said to be inherent in the EU VAT system.

43.4 The Place of Taxation for Services and Intangibles 43.4.1 EU Law In EU VAT, the notion of services is wide. The concept of the supply of services includes any transaction which does not constitute a supply of goods.23 The supply of 22 See K. James, The Rise of the Value Added Tax (Cambridge: Cambridge University Press, 2015), 83–96. 23 VAT Directive, art. 24.

EU VAT Law and the OECD VAT/GST Guidelines 791 goods applies to the transfer of the right to dispose of tangible property as the owner.24 Consequently, intangibles are included in the concept of services under EU VAT law. Under article 44 of the VAT Directive, the place of supply of services to a taxable person acting as such is the place where that person has established their business (B2B). However, if those services are provided to a fixed establishment of the taxable person located in a place other than the place where they have established their business, the place of supply of those services is the place where that fixed establishment is located. In the absence of such a place of establishment or fixed establishment, the place of the supply of services is the place where the taxable person who receives such services has their permanent address or usually resides. There are several exemptions from article 44 of the VAT Directive. Services supplied in connection with immovable property are, under article 47 of the VAT Directive, supplied where the immovable property is located. Furthermore, cultural, artistic, sporting, scientific, educational, entertainment, and similar services such as ancillary transport services as well as evaluation and work on movable property are, under article 53 of the VAT Directive, supplied where those activities are physically carried out.25 If article 44 applies and the acquirer is established in more than one country, the service needs to be allocated to one country to decide where it is supplied. Under EU VAT law, it is not possible to allocate a service to more than one member state. The Implementing Regulation 282/201126 allocates services when the acquirer is established in several member states. Article 21 of the Implementing Regulation states that where the service is provided to a fixed establishment of the taxable person located in a place other than that where the customer has established their business, that supply shall be taxable at the place of the fixed establishment receiving that service and using it for its own needs. This method is sometimes called the restricted direct use method.27 In order to identify the customer’s fixed establishment to which the service is provided, the supplier needs to examine the nature and use of the service provided.28 If nature and use does not enable the supplier to identify the fixed establishment to which the service is provided, particular attention should be paid to the contract, the order form, and the VAT identification number attributed by the member state of the customer and which identifies the fixed establishment as the customer of the service.29 Particular attention should also be paid to whether the fixed establishment is the entity paying for the service.30 Where 24

Ibid., art. 14. See L. Hiort af Ornäs Leijon and E. Kristoffersson, ‘The OECD International VAT/GST Guidelines on Place of Supply of B2B Services and Intangibles’, World Journal of VAT/GST Law 3/1 (2014), 40. 26 Council Implementing Regulation (EU) No. 282/2011 of 15 March 2011 laying down implementing measures for Directive 2006/112/EC on the common system of value added tax (the Implementing Regulation). The Implementing Regulation has been amended by Regulations 1042/2013 and 2026/2019. 27 See M. Merkx, ‘Should EU VAT Apply the Recharge Method in the Place of Supply Rules for B2B Services’, EC Tax Review 6 (2013), 282. See also Hiort af Ornäs Leijon and Kristoffersson, ‘The OECD International VAT/GST Guidelines on Place of Supply of B2B Services and Intangibles’, 41. 28 Implementing Reg., art. 22(1) para. 1. 29 Ibid., art. 22(1) para. 2. 30 Ibid., art. 22(1) para. 2. 25

792 Eleonor Kristoffersson the acquiring entity still cannot be identified, or where services are supplied to a taxable person under a contract covering one or more services used in an unidentifiable and non-quantifiable manner, the supplier may legitimately consider that the services have been supplied at the place where the customer has established their business.31 The main principle for business-to-consumer (B2C) supplies is the country of origin. According to article 45 of the VAT Directive, the place of supply of services to a non- taxable person shall be the place where the supplier has established their business. If those services are provided from a fixed establishment of the supplier located in a place other than the place where they have established their business, the place of supply of those services shall be the place where that fixed establishment is located. In the absence of such place of establishment or fixed establishment, the place of supply of services shall be the place where the supplier has their permanent address or usually resides. There are, however, many derogations from this principle. For instance, telecommunications, broadcasting, and electronically supplied services are always taxed in the country of the customer, which is the country where the customers are registered, have their permanent addresses, or usually live. Under articles 24a, 24b, 24d, and 24f of the Implementing Regulation, some presumptions are allowed to facilitate the establishment of the location of the customer.

43.4.2 The OECD International VAT/GST Guidelines As mentioned earlier, the OECD International VAT/GST Guidelines are based on the destination principle. In Guideline 3.1, it is stated that for consumption tax purposes, internationally traded services and intangibles should be taxed according to the rules of the jurisdiction of consumption. The Guidelines explicitly attribute the taxing rights. Thus, under Guideline 3.2, the jurisdiction in which the customer is located has the taxing rights over internationally traded services or intangibles. The location of the customer is where the customer has located its permanent business presence.32 According to the commentary on Guideline 3.2, this is the general rule. The general rule is distinguished from the specific rules covered by Guidelines 3.7 and 3.8.33 Guideline 3.3–3.6 explain how the general rule should be applied. Guideline 3.3 states that the identity of the customer is normally determined by reference to the business agreement. The business agreement is adopted more as a general concept than as a term with a technical meaning.34 Business agreements have elements that identify the parties, the rights, and the obligations.35 Business agreements 31

Ibid., art. 22(2). Commentary on Guideline 3.9. 33 Ibid., 3.10. 34 Ibid., 3.13. 35 Ibid., Box 3.1. 32

EU VAT Law and the OECD VAT/GST Guidelines 793 may be oral or in written format.36 However, particularly in cases involving significant sums of money or complex matters, the parties are likely to draw up legally enforceable contracts.37 It is, however, emphasized in the Commentary on the Guidelines that contracts should not be seen as the only relevant elements of a business agreement.38 Guideline 3.4 deals with the situation when the customer has establishments in more than one jurisdiction. In such a case, the taxing rights accrue to the jurisdiction or jurisdictions where the establishment or establishments using the service or intangible is/are located. Consequently, the taxing rights of services or intangibles do not have to be allocated to only one jurisdiction but may be allocated to several jurisdictions. In the Commentary on the Guidelines, three existing broad approaches are identified: the direct use approach, the direct delivery approach, and the recharge approach.39 The direct use approach focuses directly on the establishment that uses the service or intangible. The direct delivery approach focuses directly on the establishment to which the service or intangible is delivered. The focus of the recharge method is the establishment that uses the service or intangible as determined on the basis of internal recharge arrangements within the multiple location entity (MLE), made in accordance with corporate tax, accounting, or other regulatory arrangements.40 In the Commentary on the Guidelines, it is stated that none of these approaches should have preference over another, and that they are not mutually exclusive and could be combined.41 Any approach should seek to ensure that taxation of the supply of a service or intangible to a MLE accrues where the customer’s establishments that are using the service or intangible is located.42 It should also achieve a sound balance between the interests of business and tax administrations.43 Regarding Guidelines 3.5 and 3.6 and B2C supplies, the general rule is that the jurisdiction in which the service is performed has the taxing rights. This applies for services and intangibles that are physically performed at a readily identifiable place, are ordinarily consumed at the same time as and at the same place as they are physically performed, and ordinarily require the physical presence of the person performing the supply and the person consuming the service or intangible at the same time and place as the supply of such a service or intangible is physically performed. In other cases, the jurisdiction in which the customer has its usual residence has the taxing rights over B2C supplies of services and intangibles other than those covered by the main rule. Consequently, the destination principle is also applied in B2C transactions. In Guideline 3.7, it is stated that the taxing rights over internationally traded services or intangibles supplied between businesses may be allocated by reference to a proxy 36

Ibid., 3.13. Ibid., 3.15. 38 Ibid., 3.15. 39 Ibid., 3.25. 40 Ibid., 3.25 41 Ibid., 3.37. 42 Ibid., 3.38. 43 Ibid., 3.38 37

794 Eleonor Kristoffersson other than the costumer’s location when the taxing rights by reference to the customer’s location does not lead to an appropriate result with regard to neutrality, efficiency of compliance and administration, certainty and simplicity, effectiveness, and fairness. The proxy other than the customer’s location should lead to a significantly better result when it is considered under the same criteria.44 The same also applies to B2C services.45 Finally, according to Guideline 3.8, for internationally traded supplies of services directly connected with immovable property the taxing rights may be allocated to the jurisdiction where the immovable property is located. Guideline 3.8 does not list which services should be considered as directly connected with immovable property. According to the Commentary, the following services are likely to be included: (1) the transfer, sale, lease, or the right to use, occupy, enjoy, or exploit immovable property; (2) supplies of services that are physically provided to the immovable property itself, such as constructing, altering, and maintaining the immovable property; (3) other supplies of services and intangibles where there is a very close, clear, and obvious link or association with the immovable property.46

43.4.3 Comparison Guidelines 3.1, 3.2, and 3.8 are generally in line with the EU VAT law.47 Guideline 3.3 states that the identification of the acquirer is normally determined with reference to the business agreement. Also, under the Implementing Regulation the contract plays an important role. When comparing the draft guidelines with EU Law, the Commission found that the business agreement must be seen as part of the two-step approach by which taxing rights are allocated to the jurisdiction where the establishment that uses the service is located.48 Furthermore, the Commission found that using the business agreement is at odds with the approach adopted under EU law whereby such taxing right are directly allocated and where no provision is made for taxing supplies between the different entities of a taxable person.49 This is a result of the different approaches used, where the EU provisions decide the place of supply and the Guidelines decide the taxing rights. Thus, the allocation under EU law requires that there is a supply, and a

44

Guideline 3.7.

45 Ibid. 46

Commentary on Guideline 3.173. European Commission, Group on the Future of VAT, 9 Meeting 21 October 2013, Taxud.c.l. (2013) 3384189 GFV N 036, OECD, International VAT Guidelines, 5–6. 48 Ibid., 5. 49 Ibid. 47

EU VAT Law and the OECD VAT/GST Guidelines 795 supply of services in its turn generally requires a transaction between two legally independent persons. In Guideline 3.4, a different approach to that in EU VAT law is applied, since Guideline 3.4 allocates the taxing rights to where the service is used.50 According to the Commission, this is of less concern if a service supplied to and used by the customer at its main place of establishment.51 The same applies when the service is supplied to and used by a single fixed establishment.52 In the event of a customer using services for several establishments, for example under a global contract, and it is not possible to attribute the services to a specific establishment, the place of taxation under the Implementing Regulation53 would be where the customer has established their business, rather than at the place of its use.54 A strict application of Guideline 3.4 would lead to the taxing rights being allocated to several jurisdictions, which is not possible under EU law. In B2C, the main rule under EU law applies the origin principle, whereas the OECD Guidelines also apply the destination principle for B2C transactions. In practice, there are, however, so many derogations from the EU law main rule that the differences are few. The many specific rules on place of supply in the EU VAT Directive do not have any equivalent in the Guidelines, except article 47 of the VAT Directive regarding immovable property which in B2B transactions corresponds to Guideline 3.8.55 Guideline 3.7 opens up the possibility for applying other proxies than the location of the customer when that does not lead to an appropriate result, with regard to neutrality, efficiency of compliance and administration, certainty and simplicity, effectiveness, and fairness, and the other proxy achieves those standards significantly better. This gives some flexibility to the Guidelines.56 However, the requirement of ‘significantly better’ demonstrates that a derogation from the proxy of the location of the costumer should be reserved for special situations. Thus, the specific place of supply rules in the EU VAT Directive are not automatically in accordance with the Guidelines due to Guideline 3.7.

43.5 Concluding Remarks The aim of this chapter was to examine the relationship between EU VAT law and the OECD International VAT/GST Guidelines and make a comparison between the two. EU

50 Ibid. 51 Ibid.

52 Ibid. 53

Implementing Regulation, art. 22(2). Commission, Group on the Future of VAT, 9 Meeting 21 October 2013, Taxud.c.l. (2013)3384189 GFV N 036 International VAT Guidelines, 5-6. 55 Hiort af Ornäs Leijon and Kristoffersson, ‘The OECD International VAT/GST Guidelines’, 44. 56 Ibid. 54 European

796 Eleonor Kristoffersson VAT law and the Guidelines are both important landmarks for the international coordination of VAT. The EU harmonized system of VAT is fundamental for the achievement of the internal market, whereas the Guidelines play an important role for tax coordination outside the EU. EU VAT law is law for the member states and, in its implemented form, is law for the taxpayers and tax administrations within the EU. The Guidelines do not aim at detailed prescriptions for national legislation but intend to assist policymakers in their efforts to evaluate and develop the legal and administrative framework in their jurisdictions. Because of the legal status of the Guidelines as soft law, a conflict between EU VAT law and the Guidelines is not possible. In situations where EU VAT law differs from the Guidelines, EU law prevails. Whereas the Guidelines point at the destination principle both for B2B and B2C supplies of services and intangibles, the main rule for B2C services under EU law is still origin-based. EU VAT law is, however, consistently moving towards destination-based taxation, including for B2C services. For B2B services and intangibles, the destination principle applies in both EU law and the Guidelines. The attribution of services to MLEs potentially differs, since the Guidelines provides for the recharge method, where one service may be attributed to several fixed establishments in different jurisdictions. The recharge method is, however, more in harmony with how a service would be attributed for transfer pricing purposes when applying the arm’s-length rule. The Guidelines do not provide any detailed information on how to attribute a service to MLEs. In this respect, a lesson could, on the one hand, be learnt from the OECD Transfer Pricing Guidelines which are much more instructive.57 Since, on the other hand, the aim of the OECD International VAT/GST Guidelines is not to give detailed prescriptions, such information has probably intentionally been omitted. The approaches on how to allocate services and intangibles differ between EU VAT law and the Guidelines. EU law is focused on place of supply; the Guidelines allocate the taxing rights more expressively. In practice, this may result in the taxing rights being allocated to several fixed establishments, whereas a supply of services in general requires two legally independent taxable persons. Both EU VAT law and the Guidelines are non-static and subject to gradual change. EU law is moving more towards the destination-based approach, which is in harmony with the Guidelines. If more OECD members and also non-OECD members apply the destination principle in cross-border transactions, pressure may eventually be placed on states that do not apply VAT but instead use other consumption taxes, so that they also apply destination-based taxation. Overall, the Guidelines play an important role for general consumption taxes.

57 OECD, Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations 2017 (Paris: OECD Publishing, 2017), https://read.oecd-ilibrary.org/taxation/oecd-transfer-pricing-guideli nes-for-multinational-enterprises-and-tax-administrations-2017_tpg-2017-en (accessed 23 March 2021).

Pa rt V I I

RECENT I N T E R NAT IONA L TAX T R E N D S I N M AJ OR E C ON OM I E S A N D R E G ION S

Chapter 44

The US Perspe c t i v e on Internationa l Tax L aw Kimberly A. Clausing

44.1 Introduction Prior to the 2017 tax law,1 the US system of international taxation for multinational companies was widely viewed as seriously flawed. Critics faulted it for being excessively complicated, out of line with international norms, and insufficiently attentive to the problem of corporate tax base erosion due to profit shifting. The 2017 tax law did not solve any of these problems, and it created some new ones. Still, in some respects, the 2017 tax law provides a better starting point for future international tax reform. In the short-term, the new minimum taxes can be strengthened thus combatting profit shifting and reducing tax competition pressures. In the longer term, more fundamental reforms can be considered.

44.2 The US System of International Taxation Before the 2017 Tax Act For thirty years, from the Tax Reform Act of 1986 until the 2017 Tax Act, the US international tax system was relatively stable. Although there were small changes enacted as part of a 2004 tax law, the overall principles were constant. The statutory tax rate was 34 or 35% (34% before 1993), and the tax system was purportedly worldwide, in that the foreign income of US multinational companies was taxed at the full domestic rate, after

1

2017 Tax Act (Public Law 115-97, ‘the Tax Cuts and Jobs Act’, TCJA).

800 Kimberly A. Clausing a credit for foreign taxes paid. However, tax on most foreign income was only payable upon repatriation. While some types of income (subpart F income) were taxable immediately, active foreign-business income was typically treated under the normal regime, with tax postponed until repatriation. In the years building up to the 2017 Tax Act, the 35% rate and the ‘worldwide’ system were frequently held up by critics as evidence that the US tax system was out of line with those in peer countries, where statutory tax rates were lower and foreign income was often given ‘territorial’ treatment, such that much active foreign income was exempt from domestic taxation. However, these characterizations often provided more heat than light. Despite relatively high corporate profits as a share of gross domestic product (GDP), in both historic and comparative terms, US corporate tax generated less revenue than the corporate tax systems of peer countries. Figure 44.1 shows the recent record of US corporate profits (both before and after tax) as well as corporate tax revenues as a share of GDP. Despite historically high corporate profits, corporate tax revenues showed no parallel upward trend and were instead flat as a share of GDP. In recent years, US corporate tax revenues were also substantially lower than those in peer countries, averaging 2% of GDP prior to the 2017 tax law, whereas typical peer nations raised about 3% of GDP from corporate

12.0% 11.0% 10.0%

As Share of GDP

9.0% 8.0% 7.0% 6.0% 5.0% 4.0% 3.0% 2.0% 1.0%

Corporate Profit After Tax

Corporate Tax Revenue

2019

2017

2015

2013

2011

2009

2007

2005

2003

2001

1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

0.0%

Corporate Profit Before Tax

Figure 44.1 US corporate profits, and tax revenues, 1980–2019* Source: Federal Reserve Economic Data (Profits), Congressional Budget Office (Revenue). * Note: The corporate profits data include data from some entities that are organized as corporations but may not be subject to corporate tax. I am not aware of a data series that separates out these two types of entities. Readers should be aware that those entities not subject to corporate tax (i.e. pass-through entities) have become more important over time; however, other sources (e.g. the Forbes Global 2000 series) indicate strong and rising profits for public companies.

The US Perspective on International Tax Law 801 taxes.2 More recently, since the 2017 tax law, corporate tax revenues fell to 1% of GDP in 2018 and 2019. To better understand meagre US corporate tax revenues, it is useful to distinguish the ‘label’ of the US tax system (a worldwide tax levied at 35%) from the reality of the US tax system, where effective tax rates were far lower than statutory tax rates, and the US government collected almost no revenue from taxing foreign income. Effective tax rates were lower than statutory rates for several reasons. First, the US tax base was relatively narrow. The domestic production-activities deduction reduced tax rates for companies producing in the USA, interest-financed investments were fully deductible, accelerated depreciation or even expensing was allowed (in some years), losses were treated generously, and there were various tax credits that also meaningfully reduced tax burdens for many companies. Some businesses also found it advantageous to organize their operations in pass- through form, lowering the size of the corporate sector. According to Cooper et al.,3 were such businesses taxed as corporations, revenues would have been $100 billion higher in 2011. An additional factor that undoubtedly lowered US corporate tax revenues was the aggressive profit shifting of both US and foreign multinational companies, which shifted a portion of the true corporate tax base out of the USA towards sanctuary jurisdictions.4 Using a variety of methods and data sources, Clausing5 estimates that profit shifting by multinational corporations resulted in a revenue loss to the US government of about $100 billion per year by 2017. Still, one immediate question is why such profit shifting is even worthwhile for US multinational companies, if the USA had a worldwide system at the time? Here the worldwide ‘label’ is particularly misleading. Since income could accumulate offshore without triggering US taxation, companies had an incentive to avoid repatriation, and thus only infrequently triggered US tax on foreign earnings at all. In the early years after the Tax Reform Act of 1986, companies often had sufficient foreign tax credits from operations in higher tax foreign countries to offset tax due on income from lower tax foreign countries when repatriation was desirable. As time passed, however, foreign countries lowered their tax rates, as shown in Figure 44.2. This left 2 OECD

data show that typical OECD countries raise about 3% of GDP. This ratio is quite steady in recent decades, despite declining tax rates. (See https://data.oecd.org/tax/tax-on-corporate-prof its.htm#indicator-chart). The stability of corporate tax revenues in the face of declining rates is due in part to measures that increased the tax-base breadth at the same time that rates were cut. In addition, increases in the size of corporate profits relative to GDP (as in the USA) may have played a role. 3 M. Cooper et al., ‘Business in the United States: Who Owns It, and How Much Tax Do They Pay?’, Tax Policy and the Economy 30/1 (2016), 91–128. 4 Tax sanctuaries (often referred to as havens) are often rich countries with high per-capita incomes. Of the nine most important havens for US multinational companies, four are independent European countries (Ireland, Luxembourg, Netherlands, and Switzerland), four are related to the UK or US (Caymans, Bermuda, Jersey, and Puerto Rico), and the other is Singapore. 5 K. A. Clausing, ‘Profit Shifting Before and After the Tax Cuts and Jobs Act’, National Tax Journal 73/ 4 (2020), 1233–1266.

802 Kimberly A. Clausing 45 40

Percent

35 30 25 20 15

19 88 19 90 19 92 19 94 19 96 19 98 20 00 20 02 20 04 20 06 20 08 20 10 20 12 20 14 20 16 20 18 20 20

10

United States

OECD Average

Figure 44.2 Evolution of corporate tax rates, OECD and USA, 1988–2020

US multinational companies with insufficient foreign tax credits to shelter tax due on lightly taxed sanctuary income. As a consequence, US companies accumulated foreign earnings offshore in order to avoid the tax due upon repatriation. In theory, as recognized since Hartman,6 if the repatriation tax is both unchanging and inevitable, it should not affect repatriation decisions, which should be based on where the investment would earn the highest return. Since the one-time repatriation tax must eventually be paid, it does not affect the investment location decision. However, in the case of the USA, the repatriation tax was neither viewed as unchanging nor inevitable. Indeed, companies successfully lobbied for a repatriation holiday as part of a 2004 tax law and, consequently, repatriation was allowed at a bargain rate of only 5.25%. After the holiday, companies operated under the assumption that another holiday could be arranged, or the USA would eventually transition to a territorial system with any one-time deemed repatriation tax rate much lower than 35%. Indeed, such assumptions proved correct, as the 2017 tax law made exactly such a transition, with a one-time repatriation tax rates of either 8 or 15.5%, depending on the liquidity of accumulated earnings. In the run-up to the 2017 tax law, misleading claims were frequently made about the deterrent effect of the repatriation tax on US investment. Such claims were oversold, since companies could create the equivalent of tax-free repatriation by borrowing against their offshore funds. Interest is deductible at home, yet the funds accumulated offshore earn taxable interest. If the interest rate is the same, these two tax events offset each other, and the company achieves access to whatever investment funds they desire,

6

D. G. Hartman, ‘Tax Policy and Foreign Direct Investment’, Journal of Public Economics 26/1 (1985), 107–121.

The US Perspective on International Tax Law 803 absent worries of repatriation tax. Indeed, prominent US multinational companies— with billions of dollars stockpiled offshore—frequently borrowed in order to finance domestic investments, following precisely this strategy. Further, many offshore funds were invested in US capital markets, providing investment funds to the broader economy. In sum, the label and reality of the US international tax system prior to the 2017 Tax Law were so inconsistent that the argument that the US system needed to be more favourable in order to match other countries’ systems was incomplete at best, and at times purposefully misleading. Still, that does not mean that the US system was not in desperate need of reform. Indeed, the mismatch between how the system functioned and how it was described was itself a cause for concern. The US government appeared as if it had a serious tax on corporate income, at the same time that it looked the other way on large-scale tax avoidance, or even (in some instances) directly facilitated the erosion of the US tax base. For instance, the check-the-box regulations of the late 1990s paved the way for enormous international tax avoidance, further increasing the chasm between label and reality of the US system.7 On the eve of the 2017 Tax Act, there were three main sources of discontent regarding the US international tax system. (1) Companies argued that the US system was unfair to the US-headquartered multinational companies, since they faced a worldwide tax system and a statutory rate of 35%, whereas their competitors frequently faced lower tax rates and enjoyed territorial tax systems. This left US companies at a disadvantage when competing offshore. Note that this competitiveness argument is about the competitiveness of US companies, not the USA itself as a location for doing business. Arguably, a pure territorial system would eliminate worries about the competitiveness of US companies, while exacerbating concerns about the competitiveness of the USA as a location, since foreign earnings would be (even more) tax-preferred under such a system. (2) Many non-profit actors worried about corporate tax base erosion due to multinational company profit shifting. Academic studies (discussed by Clausing8) showed profit shifting to be a large problem, with revenue losses of about $100 billion per year by 2017. There are always worthwhile uses for government tax revenue, including lowering tax rates elsewhere in the system, reducing deficits, or undertaking worthwhile public investment. (3) The rules of US international taxation were widely recognized to be mind- numbingly complex, raising serious compliance and enforcement costs, and furthering the mismatch between the label and the reality of the US system.

7 It is quite possible that the full consequences of check-the-box were incompletely understood at the time. For more on the facilitation of stateless income, see E. D. Kleinbard, ‘The Lessons of Stateless Income’, Tax Law Review 65/1 (2011), 99–171. 8 Clausing, ‘Profit Shifting Before and After the Tax Cuts and Jobs Act’.

804 Kimberly A. Clausing Consider each of these claims in turn. Claim 3 is simplest to evaluate since there is widespread agreement that US international tax rules are byzantine in their complexity, costly in terms of compliance, and very difficult to enforce. Competitiveness is more difficult to evaluate. While the advocates for corporate tax cuts frequently relied on competitiveness arguments, clear evidence of a competitiveness problem was thin at best. As noted earlier, US corporate tax revenues were low relative to both GDP and corporate profits. Even anecdotal evidence was weak, as discussed by Kleinbard.9 And, while there were a spate of corporate inversions, those were relatively few, and Treasury regulations in the late Obama years were effective in reducing inversion incentives substantially, even before the 2017 tax law. Further, US multinational companies were often singled out by US trading partners as being unfairly tax-advantaged due to their aggressive tax avoidance and special tax treatment, as in the Apple state aid case. As one measure of the competitiveness of US multinational companies, consider the Forbes ‘Global 2000’ lists of the world’s most pre-eminent multinational companies. In the 2019 list, the USA had 587 of the world’s top companies, ten more on the list than five years prior (577), and fifteen more than ten years prior (536). In 2019, the USA hosted 29% of the world’s top companies, despite having 24% of the world’s GDP in dollar terms and 16% of the world’s GDP in purchasing power parity (PPP) terms. Considering the top 2,000 companies in terms of sales, profits, or market capitalization, the US shares are even higher, at 32%, 37%, and 48%, respectively. There is little evidence of a competitiveness problem. In contrast, the evidence of tax base erosion due to profit shifting is overwhelming, regardless of the data series examined. For the year 2017, new country-by-country data allow wide country coverage, clearly demonstrating the distortion of profit towards the lowest tax rate economies.10 Figure 44.3 and Table 44.1 show the distribution of profits across countries. Of the top twenty locations where US multinational companies book profits, eleven (the darker columns of Figure 44.3) are tax sanctuaries with rock-bottom tax rates. In these very low-tax jurisdictions, US multinational companies report more than twice as much income as in the nine top non-sanctuary countries, countries that are among the largest economies in the world. The high-sanctuary profits are also out of line with their economies. These eleven sanctuaries have 1.5% of the nine big countries’ populations and only 8% of the big countries’ GDP. 9

E. D. Kleinbard, ‘Competitiveness Has Nothing to Do With It’, Tax Notes 144 (Sept. 2014), 1055–1069. These data were suggested by the OECD/G20 Action 13 recommendations. These data are new, and the USA was the first country to release a full data set. The USA also released an incomplete data set in 2016. For a more detailed description of the evidence on profit shifting using this data source, see K. A. Clausing, ‘Five Lessons on Profit Shifting from the US Country by Country Data’, Tax Notes 169 (Nov. 2020), 925–940. While there is some evidence of double-counting in these data, in 2017 that problem was relatively small for foreign income (only 7% of the total for foreign income); see J. Garcia-Bernardo, P. Jansky, and G. Zucman, ‘Did the Tax Cuts and Jobs Act Reduce Profit Shifting by US Multinational Companies?, NBER Working Paper No. 30086 (2022). 10

The US Perspective on International Tax Law 805 70 60 Billions USD

50 40 30 20 10 0

ia ey an aly any m co lia ng re nd ds co da da nd na an rg ds It m an apo erla rlan o Ri rmu ana rela Chi Jap bou gdo exi stra Ko Ind Jers f M r g n I g t e o M z C i u e m K r n t n h i i e n B A Ge e e t x S Sw a e Pu sl d Ho u I e m N L t y i Ca Un Isl

Figure 44.3 Top twenty locations for US MNC profit in 2017* * Note: dark bars are haven economies; light bars are non-havens.

Table 44.1 Characteristics of havens and non-havens in top 20 profit locations in 2017 Effective Profit of US tax rate, MNCs, in US$ pos. profit billions companies

Population, in millions

GDP, in US$ billions

Havens: 11

355

3.9%

48

2,720

Non-havens: 9

158

26.9%

3,250

32,260

Ratio: havens/non-havens

225%

1.5%

8.4%

Note: In the rest of the world (excluding the USA), US MNCs report $126 billion in profit, and an average effective tax rate (in the positive profit sample) of 22.5%, in countries with 3.9 billion people and $26.8 trillion in GDP.

In these very low-tax jurisdictions, profits per employee are one or two or even three orders of magnitude higher than they are for typical countries. For example, US multinational companies report $60 million in profit per employee in Bermuda, and $600,000 in profit per employee in Switzerland whereas, worldwide, the profit per employee averages $49,000. Due to this massive clustering of profit in these very low-tax countries, it is unsurprising that the revenue costs of profit shifting are large. As noted earlier, the lost revenues may exceed $100 billion by 2017, at 2017 tax rates. Due to the serious sources of discontent about that state of the US international tax system, there have been many proposals for reform. For example, prior to the 2017 tax

806 Kimberly A. Clausing legislation, there were multiple revenue-neutral corporate tax reform packages that addressed the problems described previously. Both the Obama administration and the Republican Ways and Means Committee leadership (under Chairman Camp) proposed reforms that coupled corporate tax rate reductions (which lose revenue) with increased revenue due to international tax measures that reduced profit shifting. Perhaps unsurprisingly, these proposals failed to get wide support from the business community.

44.3 The 2017 Tax Act (TCJA) 44.3.1 Key Provisions The 2017 tax law changed the US system of international taxation in several important and fundamental ways. The package lowered the US corporate tax rate from 35 to 21%, moved to a territorial system (with a one-time deemed repatriation tax on prior foreign earnings), and included several novel international provisions. Unlike the proposed reforms of the Obama administration and the Camp Ways and Means Committee, the 2017 tax law resulted in a large reduction in business tax revenues. According to the Joint Committee on Taxation (JCT) estimates at the time, the corporate provisions on their own lost about $650 billion, and the international provisions (excluding the one-time repatriation tax) lost a small amount of additional revenue, $14 billion.11 The one-time repatriation tax raises revenue in the ten-year window, but as explained earlier, it is a tax cut relative to prior law, since the prior US tax system required full taxation of repatriated income at a 35% rate, with credits for foreign tax payments. The logic behind a reduced rate for the one-time deemed repatriation tax was one of politics rather than economics. Indeed, it is impossible to incentivize earnings that have already occurred, nor could the repatriation tax be expected to affect anticipated future tax burdens, since only one transition to a territorial system was anticipated. The repatriation rate of 8 or 15.5% was favourable to taxpayer interests, falling far below even the new corporate tax rate of 21%. Although some taxpayers had hoped for even more generous tax treatment, this was clearly a large tax break relative to prior law. The main corporate provisions of the legislation are summarized in Table 44.2, alongside the anticipated revenue effects of the legislation (according to the JCT) as well as the qualitative effects of the provisions on the incentive to shift profits out of the US corporate tax base. The corporate tax cut (from 35 to 21%) was described at the time as a provision that was intended to both increase the attractiveness of the USA as a production location

11

In addition, pass-through business tax cuts also generated large net revenue losses.

The US Perspective on International Tax Law 807 Table 44.2 International incentives before and after the TCJA Effect on profit shifting

Ten-year JCT score, $US billions

Before TCJA

After TCJA

Statutory corporate rate

35

21

Reduced incentive to –1,349 shift out of US base (net: –654)

Tax treatment of foreign income

No tax until repatriation, then 35 less foreign tax credit*

Not taxable unless subject to minimum tax

Increased incentive to shift out of US base

–224

Global minimum tax

N/A

0 until threshold, then 10.5; up to 13.125 if blended with income from higher tax countries**

Reduced incentive to shift profits to havens; increased incentive to earn in other countries

112

Foreign-derived N/A intangible income deduction (FDII)

Tax preference for profits from export sales above threshold return on assets

Likely to have negligible effect

–64

Base erosion and anti-abuse tax (BEAT)

Add-on minimum tax when payments to foreign-related parties exceed threshold

Reduced incentive to 150 shift income out of US base

N/A

Note: Revenue numbers are from the 18 December 2017 tables provided by the JCT (JCX-67-17). * Lighter rates may apply, or be anticipated, due to holidays, anticipated holidays, or expectation of future favourable treatment upon transition to a new tax system. Permanently reinvested earnings are not taxed in the USAs, but might be expected to encounter deemed repatriation tax upon transition to a territorial system. ** These rates are scheduled to increase after 2025, to 13.125 and 16.4%. This analysis ignores interaction effects between the provisions.

and reduce the incentive to shift profits offshore. However, as discussed previously, most profit shifting is destined for jurisdictions with rock-bottom tax rates, so it was unlikely that a reduction of the US statutory rate would be sufficient to substantially reduce profit-shifting incentives. While the corporate tax rate cut alone would have lost over $1.3 trillion in revenue, other corporate tax provisions broadened the tax base in a way that reduced the total revenue cost of the domestic corporate provisions to $650 billion. These base-broadening provisions included repeal of the domestic production-activities deduction, limits on interest deductibility, and a less generous tax treatment of losses and R&D.

808 Kimberly A. Clausing Table 44.2 describes the main international tax provisions; together with more minor international provisions, they were estimated to lose $14 billion in revenue over ten years. The move to a territorial tax system is the largest revenue loser among the international provisions. Under a territorial system, foreign income is generally (with exceptions) no longer subject to US tax, so there is an increased incentive to shift profits outside the US tax base, since investors no longer fear tax upon repatriation. Under prior law, the potential repatriation tax may have provided a ‘speed limit’ on the extent of international profit shifting. However, while territorial tax treatment is the norm for foreign income, two base protection measures generate immediate US tax on foreign income in some circumstances. These two provisions should reduce the incentive to shift income offshore, relative to a version of the 2017 law without such provisions. First, there is a global minimum tax (the GILTI, for global intangible low-taxed income) that applies to returns on qualified business investment above a 10% threshold, if (on average) the income was lightly taxed abroad. Secondly, there is a minimum tax (the BEAT, for base erosion anti-abuse tax) that applies when deductions from foreign related-party payments exceed a threshold. These minimum taxes were both estimated to increase US tax revenue. The GILTI exempts the first 10% return on assets; depending on company circumstances, this can provide an incentive to increase investments in plant and equipment abroad. The GILTI also allows global averaging and a lower rate on foreign income; there is a 50% deduction for GILTI income. Together, these design features create a preference for nearly all foreign income relative to US income. For example, if the only foreign income is earned in a zero-tax rate country, the minimum tax rate of 10.5% is half the US rate of 21%. Also, if additional foreign income is earned in a country with a tax rate similar to, or even higher than, that in the USA, that income will often be tax-preferred relative to US income, since foreign tax payments generate tax credits that can be used to offset the minimum tax due on the haven income. These perverse incentives are discussed in greater detail by Clausing.12 Unlike the GILTI, the BEAT applies to both foreign and domestic multinational companies; it is designed to directly combat the shifting of income out of the US tax base due to excessive related-party payments abroad. There is some concern that implementing regulations have weakened the BEAT, and a continued rethinking of design issues for this novel tax is warranted.13 Finally, there is an export earnings subsidy in the new legislation, the FDII (foreign- derived intangible income). The FDII deduction applies for export earnings above a threshold return on US assets, so it provides a perverse incentive to reduce US physical

12

Clausing, ‘Profit Shifting Before and After the Tax Cuts and Jobs Act’. e.g., M. Sullivan, ‘Economic Analysis: Can Marked-Up Services Skip the BEAT?’, Tax Notes (Feb. 2018), 705–709 and B. Wells, ‘Get With the BEAT’, Tax Notes (Feb. 2018), 1023–1032. 13 See,

The US Perspective on International Tax Law 809 assets. There are concerns about the effectiveness of this provision as well as its compatibility with WTO obligations.14

44.3.2 Evaluating the 2017 Tax Act Did the 2017 Tax Act respond to the flaws in the prior law that were identified in Section 44.2? Regarding the excessive complexity of international tax law, the clear verdict is no. The GILTI, FDII, and BEAT layer additional labyrinths of rules and conditions on top of an already byzantine system. There is no equally offsetting source of simplification. However, the removal of tax upon repatriation was a useful step. While tax upon repatriation did not materially reduce investment in the United States for reasons already described, the repatriation tax did create distortions to the financial structure of multinational companies. Also important, it was a source of great taxpayer dissatisfaction, as shareholders felt that their cash was ‘trapped’ abroad. Is the new tax law more ‘competitive’ for US multinationals? Arguably, no. The companies that were most adept at profit shifting under the prior law may face higher tax burdens due to the GILTI and the BEAT, as argued by Dharmapala.15 However, companies without substantial competitiveness concerns, such as purely domestic US companies, experienced only large tax cuts. Given the JCT estimates of the international provisions as a whole, the revenue effects are basically a wash, implying that individual company circumstances determined whether the TCJA provided a net tax cut on foreign income. Undoubtedly, some multinational companies benefited from the new treatment of foreign income, while others did not. Still, the deemed repatriation treatment was a beneficial tax windfall relative to prior law, and any profits reported in the USA were taxed at the new, lower US rate. Thus, the corporate community was undoubtedly a net beneficiary of the new tax law, even if they had hoped for a purer version of a territorial tax system without the new, complicated base protections. In terms of other criteria, the 2017 tax law is wanting. On the eve of the 2017 tax law, there were large (about $100 billion per year) revenue losses due to profit shifting. Yet, the net effect of the international provisions of the new legislation left the profit-shifting problem largely unchallenged. (That’s better than making a bad problem worse, but that’s weak praise.) Indeed, Figure 44.4 shows that the share of US direct-investment earnings reported in tax havens was unchanged in the wake of the new tax law. The share of total foreign income in havens in 2019 (61% of after-tax income, or 1.5% of GDP) is identical to the

14 See C. Sanchirico, ‘The New US Tax Preference for “Foreign-Derived Intangible Income” ’, Tax Law Review 71/4 (2018), 625–664. 15 D. Dharmapala, The Consequences of the TCJA’s International Provisions: Lessons from Existing Research’, National Tax Journal 71/4 (2018), 707–728.

810 Kimberly A. Clausing 70%

1.8% 1.6% 1.4%

50%

1.2%

40%

1.0%

30%

0.8% 0.6%

20%

Share of US GDP

Share of Foreign Total

60%

0.4%

10%

0.2%

Share of Foreign Total

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

0.0% 2000

0%

Share of GDP

Note: Data are from the U.S. BEA. The big seven havens are: Bermuda, the Caymans, Ireland, Luxembourg, the Netherlands, Singapore, and Switzerland.

Figure 44.4 Share of US MNC income in seven big low-tax sanctuaries, 2000–2019* * Note: Data are from the US BEA. The big seven tax sanctuaries are: Bermuda, the Caymans, Ireland, Luxembourg, the Netherlands, Singapore, and Switzerland.

five-year average prior to the law (2013–2017). A similar finding is also reported by Garcia-Bernardo, Janksy, and Zucman16 using additional data sources. Beyond that failure, the legislation lost substantial revenue, increasing the budget challenges of the US government. The 2017 tax law also made the tax system substantially less progressive, the opposite of an ideal response to decades of increased income inequality. As discussed by Clausing,17 claims that the corporate tax cuts would ultimately steeply increase workers’ wages were unsupported by either subsequent or prior evidence. Further, by lightening corporate taxation, the 2017 tax law exacerbated the under- taxation of capital income within the present US tax system. About 70% of US equity income is untaxed at the individual level by the US government (see Burman, Clausing, and Austin18), leaving corporate tax as the only tool for reaching much capital income. Further, most of the corporate tax base is comprised of excess returns to capital (see Power and Frerick19), suggesting that corporate tax may be a relatively efficient tax. 16

Garcia-Bernardo, Jansky, and Zucman, ‘Did the Tax Cuts and Jobs Act Reduce Profit Shifting by US Multinational Companies?’ 17 K. A. Clausing, ‘Fixing the Five Flaws of the Tax Cuts and Jobs Act’, Columbia Journal of Tax Law 11/ 2 (2020), 31–75. 18 L. E. Burman, K. A. Clausing, and L. Austin, ‘Is US Corporate Income Double-Taxed?’, National Tax Journal 70/3 (2017), 675–706. 19 L. Power and F. Austin, ‘Have Excess Returns to Corporations Been Increasing Over Time?’, National Tax Journal 69/4 (2016), 831–846.

The US Perspective on International Tax Law 811 Thus, by weakening corporate tax, the 2017 tax law changes arguably moved the USA further away from ideal tax design principles; see Clausing20 for more detail. The tax law also has spillover effects on other countries. For example, the dramatic cut in the US corporate tax rate may encourage other countries to follow suit, further accelerating the ‘race to the bottom’ in corporate tax rates that is evident in Figure 44.2. On the other hand, as shown by Clausing,21 the GILTI helps protect the tax base of other (non-sanctuary) countries, both by reducing the marginal benefit to US multinationals of shifting income (from any country) to a low-tax sanctuary, and by reducing the negative incentive effects of non-sanctuary country taxes for US multinational companies, since the resulting foreign tax credits benefit companies by shielding them from GILTI tax.22 Indeed, despite the many flaws of the 2017 tax law, there remain grounds for optimism. First, the law may provide a better starting point for international cooperation in combating corporate tax base erosion due to profit shifting, including cooperation with OECD/G20/Inclusive Framework efforts. Secondly, by moving the US tax system away from the prior status quo starting point, it may be easier to incrementally improve the tax code.

44.3 Possible Future Reforms to US International Tax Law As Section 44.2 demonstrates, the 2017 tax law did not solve the vexing problems of prior US international tax law. The most mobile US multinational companies did not experience an increase in competitiveness, although that was not a clear problem regardless. More concerning, foreign income remained tax-preferred relative to domestic income, and the new law contained troubling new incentives that rewarded foreign investment relative to domestic investment. Increased foreign tangible-asset investment increased the share of income abroad that escaped GILTI tax, whereas increased domestic tangible-asset investment lowered the intangible return qualifying for the export

20

Clausing, ‘Fixing the Five Flaws of the Tax Cuts and Jobs Act’. Clausing, ‘Profit Shifting Before and After the Tax Cuts and Jobs Act’. 22 In contrast, a per-country minimum tax would strengthen the former effect, further reducing profit shifting to havens, but it would eliminate the second effect, since there would no longer be benefits due to tax credits from operations in higher tax countries, which offset minimum tax under a global tax. In Clausing, ‘Profit Shifting Before and After the Tax Cuts and Jobs Act’, I estimate that these two countering effects are the same size, so typical foreign countries should be indifferent about the design (global or per-country) of the US minimum tax, and they should prefer adopting a per-country minimum tax with respect to their own multinational companies. 21

812 Kimberly A. Clausing subsidy. Early evidence suggests that these incentives may have increased foreign investment relative to domestic investment in the wake of the new tax law.23 The large problem of corporate tax base erosion due to profit shifting went unsolved; it was made worse with one hand, and better with the other, leaving it roughly the same size as before. And the new law’s complexity was indisputable; even those well versed in tax law took months (if not years) to fully grasp the complexities of the new rules. What comes next? For possible reformers, there are multiple flaws embedded in current law that are relatively simple to address. Though fixing these flaws will take substantial political will, the fixes can be achieved with relatively simple tweaks to the current tax system. However, in order to tackle the full range of US international tax law problems, more fundamental reform is worth considering, with due caution that such fundamental reforms take substantial time for study, implementation, and coordination.

44.3.1 A Short-Term Reform Plan In the short term, a relatively straightforward combination of reforms would address multiple concerns associated with the post-2017 landscape: (1) paltry revenues from corporate tax; (2) the under-taxation of capital; (3) large profit-shifting incentives that result in large revenue costs; and (4) the new offshoring incentives provided by the 2017 tax law. Alongside steps forward in international cooperation (discussed in Section 44.3.2) as well as domestic backstops, these reforms will not generate substantial competitiveness concerns. However, the tax system will remain intensely complicated. The reform entails three simple steps.24 • First, raise the domestic corporate tax rate to 28%. • Secondly, strengthen the GILTI minimum tax by moving to per-country implementation, raising the rate, and removing the 10% tax-free return on assets. • Thirdly, repeal the export subsidy (FDII). In the short term, these reforms should be accompanied by strong anti-inversion provisions as well as modest parallel reforms in the BEAT. Even more important, the 23

See B. Beyer et al., ‘US Multinational Companies’ Payout and Investment Decisions in Response to the Tax Cuts and Jobs Act of 2017’, Journal of the American Taxation Association (2022). 24 I suggested this package of reforms when drafting this chapter in 2020; they are consistent with my prior writing including K. A. Clausing, ‘Taxing Multinational Companies in the 21st Century’, in J. Shambaugh and R. Nunn, eds, Tackling the Tax Code: Efficient and Equitable Ways to Raise Revenue (Washington, DC: Hamilton Project, 2020), 237–283. This package of reforms was also suggested by the Biden administration in its first budget; see US Department of the Treasury, ‘General Explanations of the Administration’s Fiscal Year 2022 Revenue Proposals’ (May 2021). A more modest package of reforms was later passed by the US House of Representatives in November 2021, which included a scaled-down version of the second step.

The US Perspective on International Tax Law 813 US government should continue working with other countries to encourage the implementation of strong minimum taxes, as recently agreed in the OECD/G20/Inclusive Framework process. This agreement, among more than 135 countries representing about 95% of the world economy, is an enormous step forward in international tax cooperation; the next step is to ensure implementation. A US corporate rate increase would address both the meagre revenues from corporate tax and the under-taxation of capital income and rents in the US economy. Analyses by the Tax Policy Center, the US Treasury, and others indicate that a rate increase to 28% will raise many hundreds of billions of dollars over the coming decade.25 Since only about 30% of US equity income is taxable at the individual level by the US government, corporate tax is an important tool for taxing capital income. Also, in the context of the current tax system, much of the normal return to capital is either tax-exempt or tax-subsidized, so the corporate tax base is mostly comprised of excess returns to capital stemming from market power, rents, or luck. A stronger minimum tax addresses the large incentives under the current law to shift income offshore. Clausing26 and the Biden Administration (see US Treasury27) both propose a 21% rate and country-by-country administration of the tax. To counter offshoring, it is also useful to remove the exemption for the first 10% return on assets.28 The higher rate reduces the tax preference for foreign income, and removing the 10% exemption removes the incentive to offshore physical assets. Country- by- country administration is particularly important. At present, global averaging converts the minimum tax into a ‘maximum tax’, as companies skilfully blend income from high-and low-tax locations offshore, paying an overall tax rate (on the part of the tax base that is not exempt) that is far lower than the US rate. This incentivizes both lightly taxed sanctuary income and higher tax-country income, since the minimum tax on sanctuary income is offset by tax credits from higher tax countries. However, if the tax is levied on a country-by-country basis at 21%, there is far less incentive to earn income in a sanctuary, since it will immediately be taxed at 21%. There is also no incentive to prefer higher tax-country income to US income, since it does not come with the side-benefit of reducing minimum tax due. Simply put, the USA would no longer have an ‘America last’ tax system. A serious minimum tax also raises serious revenues. The Tax Policy Center (see Mermin et al.29), the US Treasury,30 the US Joint Tax 25 See G. B. Mermin et al., An Updated Analysis of Former Vice President Biden’s Tax Proposals (Washington DC: Tax Policy Center, 2020); US Department of the Treasury, ‘General Explanations of the Administration’s Fiscal Year 2022 Revenue Proposals’ (May (2021); and US Department of the Treasury, ‘General Explanations of the Administration’s Fiscal Year 2023 Revenue Proposals’ (Mar. 2022). 26 Clausing, ‘Taxing Multinational Companies in the 21st Century’. 27 US Department of the Treasury, ‘General Explanations of the Administration’s Fiscal Year 2022 Revenue Proposals’. 28 For a more complete description of the rationale for these reforms, see Clausing, ‘Taxing Multinational Companies in the 21st Century’ and and US Treasury, ibid.. 29 Mermin et al., An Updated Analysis of Former Vice President Biden’s Tax Proposals. 30 US Treasury, ‘General Explanations of the Administration’s Fiscal Year 2022 Revenue Proposals’.

814 Kimberly A. Clausing Committee,31 and the American Enterprise Institute, as described by Pomerleau and Seiter,32 all calculate that the aforementioned minimum tax would raise between $450 billion and more than $650 billion over ten years (relative to the current law), quite similar to revenue estimates by Clausing33 and Clausing, Saez, and Zucman.34 Further, this sort of serious minimum tax changes the dynamic of tax competition itself, as discussed by Clausing, Saez, and Zucman.35 Now, even if other countries do not tax US multinational companies, the US government will, so they have an incentive to raise their own tax rate on mobile multinational company income. Further, with US companies no longer having an incentive to shift income out of other higher tax countries to tax sanctuaries, the US minimum tax acts as a backstop to foreign tax bases. Despite the serious advantages of per-country minimum taxes, companies often argue that a per-country minimum tax is excessively complex. One simplifying option would be to use country-by-country reporting data (already provided to the IRS) to compute the minimum tax. And, if a per-country administration is deemed too complex, the complexity of a global minimum tax can be substantially reduced if the minimum rate is set at the regular domestic rate, or very slightly below.36 Repealing the export subsidy (the FDII provision) is also a helpful step for multiple reasons. First, since the deduction is dependent on export sales, it may violate WTO obligations. Secondly, since the deduction is calculated based on the excess return (above a threshold) on US assets, it is more favourable as US assets are lower, raising concerns about new incentives to offshore physical assets. Thirdly, favourable tax treatment for excess returns is exactly the reverse of what economic theory would suggest for efficient tax design. There is an argument for lower taxes on the normal return to capital to increase investment, whereas lowering taxes on excess returns just provides windfalls for those companies that have benefited from market power or rents. While encouraging research and development in the USA is an important goal, there are far more direct ways to encourage research than relying on an export-earnings subsidy. For example, research and development tax treatment could be made more generous. (The 2017 tax law moves in the opposite direction, requiring research and experimentation expenditures to be amortized, instead of expensed, starting in 2022.) In addition, the US government can encourage intellectual property development through basic science funding, higher education support, and immigration reform. Immigrants and foreign students are an important source of entrepreneurship, innovation, and

31

US Joint Committee on Taxation, ‘Letter to Senator Sanders’ (2 Mar. 2021). Pomerleau and G. M. Seiter, ‘An Analysis of Joe Biden’s Tax Proposals, October 2020 Update’, AEI Report (2020). 33 K. A. Clausing, ‘Five Lessons on Profit Shifting from the US Country by Country Data’. 34 K. A. Clausing, E. Saez, and G. Zucman, ‘Ending Corporate Tax Avoidance and Tax Competition: A Plan to Collect the Tax Deficit of Multinationals’ (2020). 35 Ibid. 36 An additional design issue concerns the ideal extent to which foreign taxes would be creditable against minimum tax liabilities. For the purpose of international coordination, full creditability is ideal. 32 K.

The US Perspective on International Tax Law 815 scientific skill; it is vitally important to assure that the US economy again becomes open to the talents of the world. Finally, since the GILTI minimum tax only affects US multinational companies, there is a strong argument for strengthening the BEAT, raising its rate in line with the higher US corporate rate, and strengthening the BEAT’s regulatory provisions. Also, while building international cooperation on minimum taxation, anti-inversion provisions would be helpful to counter company attempts to avoid the stronger minimum tax by moving headquarters abroad. Simple anti-inversion provisions can be quite effective, including a management and control test.37 While these proposed reforms address many of the problems discussed in Sections 44.1 and 44.2, they will meet substantial opposition from the US corporate community. Corporate interests will argue that a tough minimum tax, and a higher corporate rate, disadvantage US companies when they compete abroad, ultimately harming the success of the US economy as a whole. Many of these arguments are overstated and self-serving. The US economy generates enormous corporate profits, both before tax and after tax, yet our corporate tax revenues are far smaller as a share of GDP (about 2%) than those of our trading partners (which average 3% of GDP), and particularly so in the two years immediately after the 2017 tax law, when corporate revenues fell to about 1% of GDP.38 Under the reforms suggested here, revenues would be closer to 2% of GDP, but they would remain lower than those in peer nations. And, as discussed in Section 44.2, even before the 2017 tax law, there was no evidence of a competitiveness problem. However, the tougher minimum tax suggested here, while effective at addressing profit shifting, does risk putting US multinational companies at a disadvantage. Thus, measures to encourage coordinated adoption of minimum taxes are important.

44.3.2 The Importance of International Coordination International coordination is not as quixotic as one might think. Other countries have shown a long-standing and sustained interest in addressing profit shifting, launching a multi-year coordinated effort through the OECD/G20/Inclusive Framework base erosion and profit shifting (BEPS) process; this culminated in a historic agreement by more than 135 countries encompassing about 95% of the world economy to adopt county-by-country minimum tax in autumn 2021.

37 For

example, a US resident company could be defined to include both US-incorporated firms and foreign firms with their mind and management in the United States. Foreign firms that have some managerial presence in the United States and that use the US dollar as their functional currency would face a rebuttable presumption that they are US firms. See E. D. Kleinbard, ‘The Right Tax at the Right Time’, Florida Tax Review 21 (2017), 208–388. 38 After the original downturn in the pandemic, corporate profits surged, leading to temporarily higher corporate profits in the subsequent period.

816 Kimberly A. Clausing A simple, coordinated minimum tax addresses concerns about the competitive implications of a strong US minimum tax. With most countries committing to adopt a strong minimum tax in concert, they support each others’ tax bases with their efforts. Even a subset of these countries would have a big impact. Indeed, most multinational companies are headquartered in OECD countries; in 2019, 71% of Forbes Global 2000 companies were based in the OECD. Adoption by the G20 would be even better.39 Even more important, this historic agreement does not require all countries to adopt immediately in order to tackle competitiveness pressures. The agreement includes an enforcement mechanism, the ‘undertaxed profits rule’ or UTPR, which allows adopting countries to ‘top up’ the tax burden of multinational companies based in non-adopting countries, when they operate in adopting countries. This is a powerful enforcement mechanism that should incentivize broad adoption. The agreement is enormously important in its possible impact. First, it can show citizenries that international agreements are about more than corporate interests; they can be used to resolve important international collective action problems. By focusing global agreements on the needs of citizens, counterproductive protectionist and anti- immigrant impulses will be easier to resist, countering an anti-globalization backlash.40 Importantly, such agreements also help protect country tax systems from the mobility of the capital income tax base, which exacerbates income inequality by making it difficult to implement progressive tax systems, since capital income is far more mobile (and concentrated) than either labour income or consumption. It is important to note that such agreements could provide enormous benefits to poorer countries. With strong minimum taxes in the countries that headquarter most multinational companies, poorer countries could raise their own corporate tax rate without the constant worry of losing their tax base, allowing vital revenue for public investment in health care, education, or infrastructure.

44.3.3 Long-Term Reforms Coordinated minimum taxation can be a solution to many important problems in US international tax law, and coordination eliminates the key objection to a stronger minimum tax—the concern that such a tax disadvantages US multinational companies. That said, the tax system will remain complex. It is also not ideal for tax treatment to be based on the residence of companies, when economic value has far more to do with the

39 In 2019 (the year reported in the 2020 Forbes list), only 8% of the Forbes Global 2000 companies are outside the economies of the OECD, India, China, Hong Kong, and Taiwan. The top ten headquarters economies are (in order) the USA (587 companies), China (266), Japan (217), the UK (77), Canada (61), South Korea (58), Hong Kong (58), France (57), Germany (51), and India (50). 40 For a book-length treatment of these themes, see K. A. Clausing Open: The Progressive Case for Free Trade, Immigration, and Global Capital (Cambridge, MA: Harvard University Press, 2019).

The US Perspective on International Tax Law 817 factors of production (labour, capital, entrepreneurship) as well as the market for the product (demand). To address such concerns, two reforms are worthy of longer term study: formulary apportionment of international income and a destination-based cash flow tax. (Both proposals would benefit from snappier names.) Under a formulary-based system, a multinational company would be taxed based on its worldwide income, with some fraction of the tax base assigned to each taxing jurisdiction based on a formula. In fact, the ‘Pillar One’ part of the aforementioned international agreement conceives of reassigning some of the taxing rights to the profits of the world’s most profitable corporations to market jurisdictions. Since formula factors are far more difficult to manipulate than the accounting determinations that govern where profit is booked today, this change would substantially reduce the tax elasticity of the tax base, thus dramatically reducing tax competition pressures. Formulary apportionment would also eliminate profit shifting, since there would be no way to move paper profits without moving the underlying formula factors. US states and other subnational jurisdictions have used formulary apportionment with success, and research supports the idea that formula factors are far less sensitive to tax-motivated manipulation than are paper profits.41 Further, a formulary system avoids the theoretical impossibility of assigning to particular geographic locations profit that is earned by the global enterprise as a whole. A multinational company’s very existence is premised on the notion that it will generate additional profit relative to what separate companies would earn at arms’ length, so where does this additional profit belong? (Often, according to the companies’ accountants, in a tax sanctuary!) Further, much economic value today is intangible, the result of company-specific know-how or intellectual property. It is difficult to determine where such profit is truly earned, and ambiguities make opportunities for tax avoidance. Formulary apportionment settles for rough justice. The formula factors themselves mechanically determine where profits are taxed. In the earlier US state experience, a three-factor formula was often used (based on payroll, assets, and sales). Although both labour (employee headcount or payroll) and assets (especially tangible assets) are often suggested as formula factors, including them in the formula also poses risks. Companies may be inclined to move formula factors to low-tax jurisdictions, and even if companies are not that tax-responsive, jurisdictions may be tempted to lighten the formula burden on such factors based on fears of such mobility. Indeed, over time, many US states have increased the weight on the sales factor, some even moving to sales- only formulas. In part for this reason, Avi-Yonah and Clausing42 suggest a sales-only 41

See K. A. Clausing, ‘The US State Experience Under Formulary Apportionment: Are There Lessons for International Reform?, National Tax Journal; Washington 69/2 (2016), 353–385, 257–258 (2016) and J. Mintz and M. Smart, ‘Income Shifting, Investment, and Tax Competition: Theory and Evidence from Provincial Taxation in Canada’, Journal of Public Economics 88 (2004), 1149–1168. 42 R. S. Avi- Yonah and K. A. Clausing, ‘Reforming Corporate Taxation in a Global Economy: A Proposal To Adopt Formulary Apportionment’, in Path to Prosperity: Hamilton Project Ideas on Income Security, Education, and Taxes (Washington, DC: Brookings Institution Press, 2007), 319–344 and

818 Kimberly A. Clausing formula-apportionment system, based on the destination of sales (the location of customers rather than the company originating the sale).43 Formulary apportionment would work best in a context where most economies agreed on formula choices; this reduces the possibility of double-taxation or double non-taxation due to formula variations. For this reason, formulary apportionment ideally entails some international cooperation. However, non-adopting countries will have a natural incentive to adopt if some leading countries adopt, since otherwise they will risk losing tax base through profit shifting to formulary countries. (Since the formulary country tax liability will not change with inward shifting, they will appear as if they are a giant Bermuda—or zero-tax country—with respect to non-formulary country company decisions.) A destination-based cash flow tax system provides many of the same advantages as a sales-based formulary system, but it does have a few additional wrinkles, including the necessity of a border-tax adjustment. A border tax creates particular challenges for the US economy, due to the role of the dollar in the world economy as well as the absence of VAT. These difficulties are discussed in more detail by Avi-Yonah and Clausing44 and by Clausing.45 Also, like formulary apportionment, a destination-based cash flow tax is best adopted in the context of international cooperation, for similar reasons. (Otherwise, there is the possibility of double-taxation or double non-taxation. In this case, coordination also reduces the need for offsetting exchange rate adjustment in response to the border adjustment.) One additional benefit of a destination-based cash flow tax is that it automatically includes other tax base reforms that level the tax treatment of debt and equity and that make the corporate tax base a true tax on rents. Some of these advantages are also attainable in a regular corporate tax system, by combining expensing with an elimination of interest deductibility, as suggested by Furman.46 Thus, if the technical difficulties of a destination-based cash flow tax prove insurmountable, there are incremental ways to achieve some of the same aims. Both formulary apportionment and a destination-based cash flow tax would ideally entail substantial technical work prior to adoption, including the grinding work of international coordination. At present, these new tax systems are not ready for short- term implementation, although they might prove useful ideals to strive towards.

R. S. Avi-Yonah and K. A. Clausing, ‘Towards a 21st Century International Tax Regime’, University of Michigan, Law & Economics Working Papers No. 192 (2019). 43

While there are concerns that such a system could be manipulated through the use of low-margin distributor companies, a look-through rule could help address this problem, as explained in Avi-Yonah and Clausing, ‘Towards a 21st Century International Tax Regime’. 44 R. S. Avi-Yonah and K. A. Clausing, ‘Problems with Destination-Based Corporate Taxes and the Ryan Blueprint’, Columbia Journal of Tax Law 8 (2017), 229–255. 45 Clausing, ‘Taxing Multinational Companies in the 21st Century’. 46 J. Furman, ‘How to Increase Growth While Raising Revenue: Reforming the Corporate Tax Code’, in Shambaugh and Nunn, Tackling the Tax Code: Efficient and Equitable Ways to Raise Revenue, 285–315.

The US Perspective on International Tax Law 819 In the meantime, the short-term reforms of Section 44.3.1, including the adoption of a strong minimum tax (alongside anti-inversion measures and in coordination with other countries), can address the most pressing problems of current US international tax law. While that reform path entails political difficulties, it is simpler than the larger reforms. Importantly, these straightforward reforms also improve on the status quo in multiple dimensions. A corporate tax system that is fit for purpose can raise more revenue, increase the progressivity of the tax system, address the under-taxation of capital income, and dramatically reduce the profit-shifting and offshoring incentives in current law.

Chapter 45

T he Chinese Pe rspe c t i v e on In ternationa l Tax L aw Bristar Mingxing Cao

45.1 Introduction At the current time of global changes of a magnitude not seen in the last hundred years, the rapid advancement of the digitalized economy has exacerbating the depth, breadth, and difficulty of the transition currently facing countries across the world. The digitalized economy has accelerated the imbalances already present in the globalization of the economy and has lead to problems associated with the taxation rights of market jurisdictions. Thus, international tax reform in the digitalized economy has become one of the focal issues of this transformation. The international community has proposed preliminary theoretical reasons for tax reform, but still needs to integrate innovation in the principles behind government taxation which requires further exploration. If we take a comprehensive look at the taxation reform plan for the digitalized economy represented by the OECD two-pillar solution, amongst the competing claims between countries of residence and source countries, government taxation and tax avoidance by multinationals, developed countries and developing countries, and ‘ring-fencing’ reform and complex rule-design in Pillar One, significant historical progress can be demonstrated but these issues also highlight a compromise between the EU and USA for the protection of their vested interests. At the same time, the ‘one size fits all’ rules of Pillar Two are blocking the rights of developing countries for a fair tax base in the industrial economy and suppressing the efforts by least developed regions and emerging industries to progress their economic development. The rise of the taxation rights of the market states, in the constant EU gaming with the USA, propels the current international taxation reform far beyond the original scope of global anti-avoidance, and reaches the point of seeking comprehensive global fairness in international taxation in order to rectify imbalances in the unilateral, bilateral, and

822 Bristar Mingxing Cao multilateral rules of the international economy. As the second largest country in the digitalized economy but still a developing country, China is currently confronted with many issues concerning its stance on the complex goals of tax base security, fairness, and cooperation: the continued technological innovation in the digitalized economy, the accelerated changes in the power allocation in international tax jurisdictions, the misunderstandings surrounding taxation theory, etc. Yet, the world expects to know China’s viewpoint and China itself needs a clear position because the competition in and cooperation of the taxation rights of the world’s market economies relate not only to China’s national taxation interests but also to the structure of global industry around the world which will greatly affect the future shape of the global political and economic structure.

45.2 Institutional Challenges and Theoretical Exploration Under the challenge of the digitalized economy, tax base erosion and mismatch problems between resident, source, and market countries, as well as the issue of tax havens, have aggregated the increasing imbalances in international taxation rights. The subsequent international tax reform has gone beyond the scope of global anti-tax avoidance on which governments reached consensus after the financial crisis of 2008, and is now becoming the focus of the international community. This has prompted a new understanding of the background to the international economy, the theoretical basis of national taxation, and the governance mechanisms of the international taxation.

45.2.1 The Digitalized Economy and Associated Changes The financial crisis in 2008 brought about a change in the world not seen for over a hundred year and the development of the digitalized economy was synchronized that change, thus broadening the implications and extension of the resulting changes and even leading their developmental direction. In this context, as a key part of the world economic order, the challenges faced by international taxation and its future reform will have unprecedented historical significance.

45.2.1.1 Imbalance of globalization and the increasing imbalances of the digitalized economy In recent years, President Xi Jinping has put forward the concept of ‘Major Changes Unseen in a Century’ (the Centurial Change) on many important occasions, indicating that the world is in a period of major development, major changes, and

The Chinese Perspective on International Tax Law 823 major adjustments.1 From the perspective of Marxist political economy, the Centurial Change is rooted in two aspects. One is that in the previous industrial economy, industrial productivity had reached unprecedented heights and depths but the capitalist-led relationships of global production failed to match the developments in productivity; the imbalance in distribution led to a lack of effective consumption, which resulted in global financial crisis and economic crisis. The other aspect is that the booming digitalized economy, while providing better social welfare, is increasing the imbalance of distribution caused by defects in the traditional mechanisms, and is rapidly magnifying the chronic problems of economic and social development. In detail, in the global economic cycle since modernization, under normal circumstances investment and consumption have always been in a state of imbalance. The capital-dominated distribution system of the market economy determines that there is always a shortage in effective consumer demand;2 this gap is usually manifested in insufficient consumption by the labour force. What is often overlooked and concealed is insufficient public expenditure due to insufficient national revenue. This is one of the main causes of the continuous occurrence of national and international economic crises. In the field of international taxation, aggressive tax avoidance by multinational enterprises by using tax system differences and loopholes has greatly eroded the national tax base of both the source and residence countries. If this is coupled with a lack of taxation rights for the market country, the public expenditure capability of that national government to provide public services will be more seriously weakened. Thus, the shortcomings of the traditional international taxation mechanism that have become more prominent in the digitalized economy era will further aggravate the imbalance and unsustainability of the consumption capability of some countries caused by the lack of taxation rights in market countries. This will eventually lead to continuous occurrences of global economic crisis. Only by viewing the international tax reform of the digitalized economy in such a way can we understand the essence of the profound historical background and design a system conducive to the Centurial Change.

45.2.1.2 Challenges of the digitalized economy for international tax governance mechanisms The fundamental reason why the digitalized economy poses a severe challenge to international taxation is that the digitalized economy has exacerbated the imbalance between supply and demand in the global economy; further, due to the inherent shortcomings of international tax jurisdiction, a fair and effective mechanism for the division of the international tax base cannot be provided. The digitalized economy has changed the construction of the global value chain and profit-generation mechanisms of multinational enterprises with different business models—such as value chains, value platforms, and 1 L. Deng, ‘Research on China’s International Taxation under the Great Changes Unseen in a Century’, International Taxation 2/1 (2020), 1–11. 2 T. Piketty, Capital in the Twenty-First Century (Cambridge, MA: Belknap Press, 2014).

824 Bristar Mingxing Cao value stores. This highlights the salient features of this economy which has a high degree of cross-jurisdictional mobility, an great increase in the scale of entities and a reduction in their number, an extreme reliance on intangible assets, and data becoming the key production factor,3 as well as the increasing importance of the value contribution of data and user-participation which have further induced more aggressive issues of tax base erosion and profit shifting (BEPS). In the OECD’s BEPS Action Plans, the core idea is to put forward and adhere to the basic idea that ‘profits should be taxed in the place where economic activity occurs and where value is created’, but this also bring to the fore that the two cornerstones supporting the current international taxation rules are fundamentally challenged under the conditions of the digitalized economy: one is the nexus rule that determines where to levy taxes (it is difficult to determine tax jurisdiction in the absence of a physical presence); the other is the arm’s-length principle that determines how to levy taxes (it is difficult to determine the attribution of profits when user-participation and data contribution to value creation cannot be determined). In the end, people even doubt the whole international taxation mechanism with regard to taxation subject, taxation object, taxation method, and even the nature of tax.

45.2.2 Rise of Market Jurisdiction Taxation Rights and Preliminary Reasoning In the digitalized economy, cross-border digital transactions have difficulties in distribution under the current international taxation rules, which results in tax base erosion and the emergence of market countries’ taxation rights. In some countries’ unilateral policies and in the OECD’s multilateral plans, policymakers tried to put forward a preliminary theoretical basis to justify market jurisdiction taxation rights in an effort to ensure the fairness of taxation between digital and other enterprises, and between market countries and resident countries.

45.2.2.1 The rise of market jurisdiction taxation rights Although the international community still has serious differences about international taxation reform for the digitalized economy, due to economic and fiscal pressures and for the purpose of protecting their own taxation and economic interests, some countries have taken unilateral action: the UK initiated a profit transfer tax in April 2015, also known as the ‘Google tax’ and, in 2016, India introduced an equalization tax by choosing sales as the criterion for determining whether there a company had a significant economic presence (SEP) and for confirming whether its profits were ‘deemed as sourced from India’. In France, the ‘YouTube tax’ was levied in September 2017 and then 3 C.

Pei, J. Ni, and Y. Li, ‘Political Economy: Analysis of Digitalized Economy’, Finance and Trade Economics 9/5 (2018), 5–22.

The Chinese Perspective on International Tax Law 825 changed to a ‘digital service tax’ in 2019; the European Union also announced a digital tax proposal in March 2018 which imposes a digital service tax on specific digital service revenues of multinational enterprises for the income of users who contribute to value creation, and it plans in the future to merge the proposal into the EU’s long-planned Common Consolidated Corporate Tax Base (CCCTB) programme.4 Undoubtedly, the self-interested unilateral measures of these countries can satisfy a pressing need but, in the long run, individual and regional measures will inevitably increase international tax competition and present difficulties for coordination of international tax rules, thus severely disrupting the international tax order. With the endorsement of the G20, the OECD is actively proposing a systematic response to international tax governance including a resolution for challenges to the digitalized economy. The first Action Plan of the BEPS Project released by the OECD in 2015, ‘Responding to the Tax Challenge of the Digitalized Economy’, enabled the digitalized economy to replace e-commerce as the focus of international taxation. In 2018, the OECD’s series reports proposed specific action plans—such as amending the scope and identification standards for permanent establishments, implementing withholding income tax, and adjusting value-added tax for overseas payments for digital transactions—to try to solve the new problems. However, the BEPS Action Plans focus more on resolving double non-taxation problems to fix loopholes in the current international taxation rules, rather than putting much effort into the regulatory challenges faced by economic digitization itself, it thus instead activates a desire from various countries for tax sovereignty (tax jurisdiction) in the digitalized economy. In the end, for the purpose of a multilateral long-term resolution of key issues, the G20 required the OECD to expedite the proposal for a response plan under the BEPS framework. After multiple rounds of coordination, in 2019 the OECD finally proposed the ‘two-pillar’ solution to address the digitalized economy and, in October 2020, the ‘Two- Pillar Blueprint Report’ was released. Eventually, as it was expected by many, a final consensus was reached in December 2022 under the Inclusive Framework.5

45.2.2.2 The preliminary basis for the international taxation reform of the digitalized economy Based on the principle that ‘profits should be taxed where economic activities occur and where value is created’, these theoretical propositions primarily contain three proposals: ‘User Participation’, ‘Marketing Intangibles’, and ‘Significant Economic Presence’. The User Participation proposal believes that the active and continuous participation of users is a key factor in creating value for highly digital companies, such as operating social media platforms, search engines, and online shopping malls. It advocates revising the profit distribution rules and taxation nexus rules to ensure that 4 Q. Liu, ‘Observations on the Construction of Digital Taxation: A General Analysis Framework’, International Taxation 3 (2019), 25. 5 OECD, Tax Challenges Arising from Digitalisation— Reports on Pillars Blueprint (Paris: OECD Publishing, 2020).

826 Bristar Mingxing Cao the users’ country has the right to levy taxes on companies engaged in such business models, regardless of whether or not the companies have a tangible presence in the country where their users are based. The Marketing Intangibles proposal believes that, unlike transactional intangibles, the marketing intangibles of a company have a functional internal connection with the country where its product market is located. It advocates that the current transfer pricing rules and tax treaty rules should be revised, and the marketing intangibles and related risks be allocated to the country where the market is located, so that the country where the market is located has the right to collect all or part of the unconventional or residual income that can be attributed to marketing intangibles and related risks, regardless of whether the country where the market is located can or cannot levy taxes on income related to those marketing intangibles and related risks in accordance with current tax rules. The Significant Economic Presence proposal believes that economic digitization and other technological advances allow companies to participate in the economic life of a country without having to maintain a significant tangible presence, thus invalidating the current tax nexus rules and profit distribution rules. Therefore, the proposal advocates that if relevant factors can prove that a non-resident’s income-generating business activity has established a nexus with its significant economic existence, that jurisdiction can tax the income generated by the transaction.6 These proposals provide some specific explanations on value creation and profit distribution in the digitalized economy from the perspectives of supply and demand. However, due to the limitations based on a traditional, purely microscopic, perspective, it confuses private value creation and public value consumption. Failing to correctly explain the theoretical basis of government taxation, and failing to fully reveal the legitimacy of market jurisdiction taxation rights, means that the issues still need further exploration.

45.2.3 Further Exploration of the Theoretical Basis for Market Jurisdiction Taxation Rights In fact, the increase in interest of the taxation rights of a market jurisdiction is a reaction to and correction of the traditional international taxation mechanism. The micro-marketization of the national taxation basis and global governance under the traditional mechanism has formed a one-sided and incorrect perception that ‘taxation is an encroaching of a private pocket’ and ‘international tax jurisdiction is established on the basis of profit creation by supplying sides’, leading to profit distribution by the market-dominant companies and the division of tax bases by resident governments, which continuously increase the overall lack of consumption capability. This is the root 6

Y. Zhu, ‘Game Analysis of National Interests in the Latest Proposal of OECD Digitalized Economy Tax Rules’, International Taxation 3 (2019), 5–13.

The Chinese Perspective on International Tax Law 827 cause of the imbalance in the international economy and international taxation, and the root cause of the rapid expansion of the imbalance in the digitalized economy.

45.2.3.1 Further discussion on the theoretical basis of government taxation rights The digitalized economy is a naturally international, or global, market economy. The increasingly prominent taxation and economic contradiction between producer countries and market countries, on the one hand, significantly reveals the microcosmic role of the governments in the global market, on the other hand, and also highlights and justifies government participation. If we view national taxation as compensation for government costs and as benefit sharing based on a government contributions, rather than value creation only by enterprises, we can provide a new explanation for the basis of international tax jurisdiction. In classical theory, people often discuss value creation from the perspective of micro- subjects in the market, recognizing enterprises as the source of the whole profit and tax base. But if, in the digitalized economy, we look deeply at unconventional profits and link dividends with them, we can divide value creation and profit distribution into at least three levels: (1) profit due to corporate value creation, where various contributing factors can be relied on to request the distribution right; (2) the representative government of the nation which is the overall personality of the regional market, where public input enables a request for public-cost compensation and surplus sharing; and (3) the overall surplus effect reflected in an industry or region as a whole—the supplier surplus or the producer surplus, which implies their value contribution and rights for surplus sharing. In this way, it can be observed that any transaction takes place at three levels: micro, meso, and macro. When a lump-sum price is finally reflected into profit, different entities should have the right to claim; the requesting rights of the government come from the public value generated by its public-service expenditure, which forms part of the total value through the visible micro-subjects’ cross-border market transactions. In this sense, we can at least refine Marx’s famous commodity value formula w = c +v +m as w =(c1 +c2) +(v1 +v2) +(m1 +m2), in which c1/v1/m1 are on behalf of the enterprise and c2/v2/m2 are on behalf of the government.7 Thus, taxation is the private sector’s compensation for the consumption of public goods and the corresponding surplus sharing, and taxation is not the deprivation of private wealth, it is only the exercise of government rights and interests. In addition, it should be noted that, in fact, government public expenditure occurs not only in the field of public economy but also in many other fields, such as non-economic political and social affairs. Although the price compensation for government public spending is mainly from taxation, the functions of a competent and promising government should be more complete; those functions 7 J. Du and B. M. Cao, ‘Private Exchange +Public–Private Exchange: Complementing the Panorama of Comparable Transactions from the Perspective of Government Contribution, Tax Research 7 (2020), 76–80.

828 Bristar Mingxing Cao cannot be fully compensated by narrowly defined taxation, which leaves more room for exploration in the fiscal policy and practice of moral governments.

45.2.3.2 Further discussion on the theoretical basis of market jurisdiction taxation rights The theoretical basis for the legitimacy of government taxation rights has been reimagined for the industrial economy. However, the digitalized economy is characterized by the use of intangible assets, non-physical operations, and exchanges without direct consideration. Under the traditional international taxation mechanism, which is based on the taxation rights of resident and permanent establishments, there will undoubtedly be no way to establish such a scientific justification for directly granting market countries the right to levy taxes (especially the right to levy income taxes). Whereas, if we reintroduce the concept of ‘profit from both supply and demand’ to the discussion of the distribution of international tax jurisdictions, and rectify the theory of ‘profit from supply’ that has been commonly supported by the international taxation systems in modern times, the market country and the producer country will act as both supply and demand parties; they are jointly creating profits and having the same but respective tax rights over the surplus of the cross-border transaction. This will provide a new theoretical basis for the establishment of taxation rights in the market state.8 The foundation of innovation lies in unifying the concepts of producers and consumers. First, we should realize that the transaction between producer and consumer is actually an exchange of use value between producer A and producer B; the so-called consumer is just the previous producer (or future producer), but in the case of a transaction with currency, the use value is replaced and obscured by the appearance of general equivalents of currency. Secondly, on the basis of recognizing that the transaction is an exchange between producers, we then restore the tax liability and tax base division of the two parties in their respective countries. Finally, if we recognize that state taxation is based on government contribution and its cost compensation and benefit sharing, then it is obvious that producer A and producer B will provide different use values, and will pay for their consumption of public service in their production of use value, by undertaking the taxation obligations of consumption tax and income tax in their respective countries. Thus, the problem of establishing the basis for the taxation rights of producer B, which is the market country where the consumer is located, can be effectively resolved. Further, in practice, it is common that country B only levies consumption tax on the currency value and overlooks the consumer B/producer B surplus generated under the mutually comparative advantage, thus surrendering the income tax of the corresponding tax base; moreover, in an era of shortage in the economy where the supply automatically

8 Y.

Liao, ‘Distribution of Taxation Rights for Business Profits in the Digitalized Economy Environment, Journal of Xiamen University (Philosophy and Social Sciences Edition) 4 (2017), 92–101.

The Chinese Perspective on International Tax Law 829 creates demand, profits and tax base will naturally shift to the side of the more dominant factors of production, resulting in increased losses due to position differences in the market. This is the origin of the loss of the right to tax income in market countries, and the origin of the imbalance of the rights to later international taxation. When the OECD takes the concept of permanent establishment in its model tax treaties to restrict the taxation rights of the source country, it actually confuses the allocation of taxation rights in international investment with that in international trade, and the market country’s taxation rights will be further blurred or even denied.

45.3 Interest Gaming and Proposal Evaluation In order to meet the challenges of the digitalized economy, in recent years the international community has proposed unilateral, bilateral, and multilateral tax reform plans. If we take a closer look, whether it is wanton unilateralist policies or bargaining bilateral measures, they violate the nature of the digitalized economy and cannot resolve the imbalance and insufficiency of international taxation. This has led to the urgent pursuit of global consensus on multilateralism. However, the currently proposed digital tax reform plan under the OECD BEPS framework led by the developed countries exposes three obvious characteristics of revolutionary, conservative and compromise.

45.3.1 Gaming in the Proposals for International Tax Reform Market jurisdiction taxation rights advocates by India and the European countries are represented by proposed unilateral tax reform plans such as equalization taxation or the digital service tax. Although these unilateral actions are based on legitimate theoretical purposes, they also cause many undesirable effects in practice. First, the actions are the unilateral acts of a country or region which can easily cause double taxation for enterprises. Secondly, unilateral actions may trigger economic and trade friction, such as when the USA began a ‘Section 301’ investigation of the French digital service tax. Thirdly, although in the short term levying profit transfer tax or digital service tax may increase domestic fiscal revenue, in the long run, such policy formulation deviates from the principle of tax neutrality, which is not conducive to the development of domestic digital industries. Finally, a levy such as India’s equalization tax is based on revenue, but tax calculations based on subjective factors are extremely strong which is not conducive to the healthy development of the market. Considering the continuous disputes over unilateral measures, and the intricacies of the OECD tax reform plan, no consensus on the issues has until now been reach.

830 Bristar Mingxing Cao However, the UN’s 12B plan—a proposal to insert a paragraph on digital transaction in article 12 of the UN Model Taxation Convention—provides a simpler solution.9 Although the UN platform has a wider range of representatives, and its recommendations are more conducive to developing countries, the theoretical rationality of the 12B approach and its feasibility in practice need further analysis. First, the UN still employs the income tax system in drafting article 12B; that is, levying taxes on the net income of enterprises. Secondly, the question of taxation rights in the digitalized economy requires specific answers to the following questions: what is the scope for the application of the new taxation rules? Under what conditions will taxes be imposed by market countries? What are the specific criteria for the allocation of tax rights, or how can the attribution of profits be confirmed? And many more questions. Obviously, the UN chose a solution that avoided the most important issues but solutions to the previously mentioned problems are still vague. In essence, the proposal only shifts the focus of unilateral measures and the OECD programmes of various countries to bilateral negotiations. We can predict that the endless bargaining of bilateral tax treaties will not help to promote a final consensus on the process and, at the practical level, it still depends on more experts needing to concentrate their wisdom and to coordinate the different interests of those involved. Thus, it is desirable for a reform plan with a global consensus within an inclusive framework to be reached as soon as possible. As a result, the OECD multilateral digital tax reform plan based on the G20 endorsement has become a feasible way forward. However, after many rounds of negotiation, the final Two-Pillar Blueprint Report issued by the OECD is still full of controversy. It is still therefore difficult to reach global consensus and requires further economic compromises and greater political decisions.

45.3.2 Gaming in the Establishment of Market Jurisdiction Taxation Rights In essence, international taxation reform in this new stage, with the digitalized economy as the strategic background variable, still need to coordinate taxation interests between countries. Hence, the OECD regards it as a priority issue for discussion and resolution in the first BEPS Action Plans. We then observe that the reform plan in the OECD’s Two-Pillar Blueprint Report, which is currently more affected by the EU, for the first time grants to the downstream market countries of the digitalized economy the right to levy new income taxes. But this does not mean that all countries will readily and fully agree with the OECD’s Blueprint or the EU’s digital service tax. The digitalized economy is a more advanced economic stage than the agricultural economy and the industrial economy, and will undoubtedly rewrite the global economy, 9 United

Nations, ‘New Article 12B—Income from Automated Digital Services, Tax Treatment of Payments for Digital Services’ (2020).

The Chinese Perspective on International Tax Law 831 and even the political landscape, step by step. China and the USA stand at the forefront of that economy, and the competition between the two countries is reflected in all aspects of politics and economy. In order not to lose this race, the EU has continuously introduced responsive and even adversarial rules in an effort to maintain its strategic position in the global digitalized economy. From a technical point of view, it is indeed difficult for the EU to compete with China and the USA. After all, in the UN report released in 2019, the EU accounts for only 4% of the market value of the seventy largest digital platforms in the world. Although it is difficult to compete in technology, the EU is trying to blaze its own trail by establishing rules for its digital ambitions. The most concerned measure taken by the EU is the digital service tax. Facing the dominant position of the US technology giants Google and Facebook in the European technology market, the EU proposed the imposition of a digital service tax in 2018, especially in France and Italy. Both countries enacted the new tax law and began to levy digital service tax. The US government then decided to launch a ‘Section 301 investigation’ of the French digital service tax. The trade war between the EU and the USA concerning the digital service tax over the past few years has revealed all its problems to the world. The EU also has a more subtle strategy for restraining the USA and China by passing the General Data Protection Regulation (GDPR), trying to reinvent itself as a rule maker of global digital technology, especially in the use of personal data. In EU planning, if the world eventually splits into different fields of technology and data, the GDPR will set the rules for many regions outside China and the USAs. EU policymakers presume that no matter how successful China and the USA are in technological development, the EU’s attitude to privacy and data protection will effectively balance the development of their businesses.10 Competing for the tax jurisdiction of market countries is a trend combining both subjectivity and objectivity. From the perspective of the industrial chain, value chain, and the pattern status of the global economy, the EU has become the main promoter of the OECD programme due to its digitalized economy being less developed. The USA has obvious advantages and is more resistant to digital tax. China is in the middle, having some similarities with the USA in the development of the digitalized economy to enable consumers to enjoy better benefits from technological advancement; however, China is also shares the EU reform concept in promoting tax balance in the world economy. Hence, China’s position thus far is still unclear, and how to position itself scientifically and rationally is still a huge problem.

10

Tencent News, ‘Global Digital Economy Map: China and the United States are already bipolar, how can the EU win?’ (10 March 2020), https://new.qq.com/omn/20201003/20201003A03OM600.html.

832 Bristar Mingxing Cao

45.3.3 Pillar One: Limited Fairness of the ‘Ring-Fencing’ Reform Structure Since the digital economy is an innovative economy and is in the emerging stage of development, the initial allocation of international tax jurisdiction in the digital economy is, to a large extent, the basis and key to the fairness of the division of the international tax base, so the OECD regards it as a priority issue for discussion and resolution in Pillar One. We have observed that the reform plan of the OECD, which is currently more affected by the EU, has for the first time given the market countries (downstream in the digital economy industrial chain) the right to levy new income taxes. At the same time, they have reached a compromise with the USA to secure their tax interests as resident countries in the industrial economy. The vested taxation rights of the supplying countries having being retained, OECD Pillar One can only be regarded as a reform plan of limited fairness with restrained progressiveness. Pillar One focuses on the new nexus and clarifies two important issues. The first is to grant market countries the right to levy new taxes (‘Amount A’), confirming part of the residual profits at the multinational group level (or business-line level) as Amount A, and dividing Amount A between the eligible market countries according to the agreed distribution factors using a formula distribution. The second issue is to give a fixed return (‘Amount B’) to certain basic marketing and distribution functional activities that occur in the market country in accordance with the simplified principle of independent transactions. From this point of view, the core value of the Pillar One proposal lies in the proposed new taxation right, by relinquishing persistence of the permanent establishment rule and the arm’s-length principle used in traditional international taxation, thus profit distribution and tax base division to market countries based on formula allocation is legitimately materialized.11 In fact, the formation of the concept of a ‘market country’ is an outcome of the self- revolution of international tax by the EU-led OECD. In the stage of the industrial economy, many European countries had long lived in the upper reaches of the industrial chain. International tax jurisdiction based on residential countries that the OECD strongly promotes had greatly benefited European countries. However, under the conditions of the new economy, the digital economy of China and the USA (especially the USA) has developed rapidly and is in a significant leading position in the world, while the EU as a whole is living in the lower reaches of the industrial chain, and can no longer obtain the biased benefits of international tax based on the traditional jurisdiction of the resident country. Therefore, the OECD has created the concept of a ‘market country’, transcending the arm’s-length principle, and is endowing and emphasizing tax jurisdiction and taxation rights that are not available to the traditional source countries.

11 OECD,

Tax Challenges Arising from Digitalisation—Report on Pillar One Blueprint (Paris: OECD Publishing, 2020).

The Chinese Perspective on International Tax Law 833 Based on the these points, the establishment of the taxation rights of the market state is actually a return to a proper order, which should be applied to all cross-border transactions not just digital goods and services transactions. Only in this way can the international tax jurisdiction of cross-border transactions be completely balanced. The taxation right reform of the digital economy should be reshaped according to such a rational. However, on the one hand, the OECD has actively promoted the establishment of market-state taxation rights to a limited extent without reaching a comprehensive consensus; on the other hand, it has concealed and blocked the balancing taxation rights for cross-border transactions in the current digital economy. In order to protect the existing interests in industrial chains and value chains in developed countries, digital taxation has been ‘circumvented’. Some experts have sharply pointed out the benefits and logic of the rules in the OECD Pillar One programme. First, Pillar One narrows the scope of application of Amount A from highly digitalized enterprises to automated digital services, and proposes that the new taxation right only distributes residual profits to enterprises with Amount A in market countries. Secondly, Pillar One separates the consumer-facing business from the limited risk distributors of remote marketing, and converts most of the income of the limited risk distributors of remote marketing to Amount B which is much larger than Amount A. In this way, through the exclusion rule in the scope of application, Pillar One will achieve at least the following three direct effects. First, the tax burden of multinational digital giants will be minimized. Automated digital services have prevented many multinational digital companies from being included into the scope of taxation. Even if some are included, they will only need to distribute a small part of their residual profits. Secondly, the vested interests of the USA and the European countries are well preserved. The design of the scope of application of Amount A and Amount B not only minimizes the impact on the advantageous Western service industry, but also enables the protection of remote marketing’s limited risk distributors, the practice of paying royalties abroad, and tax havens and low-tax locations. Thirdly, extra benefits are expected from the design of the rules: with the operations discussed earlier, since the USA and the EU are the world’s largest consumer markets, the distribution of consumer- facing business in Amount A and the return on basic distribution activities of Amount B may be much greater than Amount A allocated for automated digital services, their gains from this will outweigh the ‘losses’ in the new taxation rights.12 Based on this analysis, the design of the OECD Pillar One rules ensures that the USA and the European countries both benefit, but ignores and violates the interests of developing countries. In view of this, the international community needs to promote to the OECD the establishment of a fair scope of application of market-state taxation rights and a reduction of unreasonable ‘delineation’ and ‘ring-fencing’ in order to achieve comprehensive fairness of taxation rights and a simplified and unified taxation system. Only

12 Y.

Jiang and Y. Jiang, ‘Process, Key Points and Countermeasures: Analysis and Perspective of the OECD Unified Method in Pillar One’, International Taxation 12 (2020), 26–31.

834 Bristar Mingxing Cao in this way can we truly face the challenges of the digital economy and resolve more effectively the basic problems in international taxation.

45.3.4 Pillar Two: Formal Fairness of the Design of the ‘One Size Fits All’ Rule The BEPS Action Plans led by the OECD itself are an effort to deal with the tax avoidance of multinational companies. The digital economy is too complicated and the OECD’s innovation reform and rule design are relatively late. The release of the ‘Response to the Challenges of Digital Economic Taxation—Pillar One and Pillar Two Blueprint Report’, especially the design of the Pillar Two scheme, ensures that the entire anti-tax avoidance framework is complete. However, Pillar Two also implies the protection of vested interests, ignoring the need for autonomy and flexibility that are essential for least developed regions and innovative industries in the pursuit of enhancing development, thereby suppressing and denying them benefits in the application of the preferential tax system that are conducive to initiating growth and promoting innovation. The prerequisite of Pillar Two is to determine a ‘global minimum tax rate’, which will nominally generate positive benefits for all countries. However, static, formal rules often fail to reflect the actual needs of dynamic development. If the rules are idealized and abused, they will greatly restrict the development rights of developing countries. The development of emerging digital industries will also face a similar situation; namely, the legitimate autonomy of the special development needs become improper activities that are not supported and protected by rules. Pillar Two guarantees the security and fairness of the tax base of the resident country and the source country in legal form, but the premise is that the economic activities occur normally, that is to say, the rules are applicable to economically active regions and industries. If least developed regions and innovative industries need to promote growth, based on historical experience (in fact, this is also the logic of capital in the market economy) there is often an urgent need for the country and even the international community to provide financial and tax incentives for initial capital investment. No doubt there will be situations where the effective unilateral, bilateral, or even multilateral tax rate is lower than the so-called ‘global minimum tax rate’ set by the OECD. Once Pillar Two cannot exclude this situation, it will inevitably lead to interference in the fiscal autonomy of these countries and regions, which is essentially a suppression of legitimate development rights, and will form a similar ‘removing-the-ladder’ effect as that in the policy history of the developed countries to the developing countries. Therefore, the author believes that Pillar Two needs to specifically add an exclusion to the anti-tax avoidance mechanism and design, at the same time, a supporting ‘deemed tax added back rule’ (DTABR). That is, if the taxing country reaches an agreement with other countries to give tax incentives to specific regions or specific industries when calculating the ‘global minimum tax rate’, the tax amount corresponding to the

The Chinese Perspective on International Tax Law 835 tax incentives should be regarded as an effective tax payment and added to the calculation in determining the effective tax rate. This rule does not have the problem of being abused by multinational companies for tax avoidance because the application of tax- preferential conditions still requires the parties to the treaty to determine whether their respective companies are substantively participate in value creation, and whether they are truly in line with the policy goals of the rules of promoting the development of least developed regions and emerging industries.13 Of course, there is still a question here, that is, the OECD originally intended to protect the national tax base in cross-border transactions, and the DTABR approach will not only not meet the objectives of the OECD plan, but will otherwise weaken the tax rights of the parties to the transaction, especially those of the developing and least developed countries. The author believes that, on the one hand, this is the result of the paradox logic of capital and market. The initial start-up of the market economy depends on the investment of capital, but capital pursues low-cost and high-yield aims, and the initial investment risks in least developed areas or emerging industries are huge. Tax incentives have proven to be an effective incentive mechanism; in other words, if there is no capital investment to promote the initiation and development of the economy, those regions and emerging industries will not be able to generate any tax sources and tax bases. This is the actual difference between a mature economy and a developing economy. Unfortunately, developed countries often deliberately ignore and cover up such facts. On the other hand, this also involves the fiscal autonomy and integrity of developing countries. If developing countries implement tax incentives on the basis of fiscal autonomy, although they can be compensated by future taxes in the long run, they are still theoretically harming their tax base and their public service capabilities; therefore, developing countries need to strengthen their national capabilities by establishing a more complete fiscal system, and use state-owned resources and the profits of state- owned enterprises to make up for and support public expenditure on tax incentives. This involves more comprehensive issues of fiscal and political economics which are outside the scope of this chapter. From the previous discussion it can be seen that the establishment of DTABR rules is not a destruction of national taxation sovereignty and tax base, but rather the maintenance and effective cultivation of sovereignty and substantial fairness. Furthermore, we need to re-evaluate the legitimacy of tax incentives, in particular to re-understand the fifth Action Plan in the OECD BEPS framework, and also the so-called ‘harmful tax competition’ that has persisted for many years, so as to achieve a balance between the formal fairness of tax rules and the real rights for economic development, and to truly realize the sustainable development of international taxation and the world economy.

13 China

International Tax Center, International Fiscal Association China Branch, Comments on OECD Public Consultation Document Global Anti-Base Erosion Proposal (‘GloBE’)—Pillar Two (Paris: OECD Publishing, 2019).

836 Bristar Mingxing Cao

45.4 Chinese Concerns and Policy Expectation The international tax reform of the digitalized economy in the Centurial Change poses a huge challenge for China. As a developing country with a socialist market economy, China is now facing many policy difficulties. From theory to practice, China needs to rationally look at the innovation of international taxation rules. While accepting and restoring the basic and legitimate taxation rights of market countries with the economic concept of a multilateralist community, China must strive to break the hegemonic global economy to realize fairness of the tax base in the global circular system of both the digitalized economy and the industrial economy. Although reforms can be implemented gradually, only the construction of a comprehensive and substantive international taxation system is the correct path for the healthy development of the national and global economy in this digital era.

45.4.1 Background: Recognition and Positioning of China’s Strategic Role While the Centurial Change is the basic historical background of international taxation reform, the new stage of digitalized economy and international taxation reform is a leading force in it; only by thoroughly combing the basis of the strategic pattern of the Centurial Change can we correctly understand the historical logic for international taxation reform. There are two important manifestations of change in the world pattern. One is the decline in the overall strength and hegemony of the Northern developed countries in the sense of productivity level and the dynamic rise in the strength of emerging economies in the South and developing countries. The latter’s contribution to world economic growth has reached 80%, and the process of reforming the world order has begun. Secondly, the Western road, which is based on the Western concept of the capitalist system aiming at singularization, is being balanced by the Eastern road, by taking the socialist value system with Chinese characteristics as the cornerstone, and recognition of multilateralism as the goal at the current stage. The progress of the Eastern forces, associated with different political choices, is the most positive factor in the ‘Centurial Change’; the rebalance requirement of the contrast of strength is obvious, and the impact is extremely far-reaching.14 The new trend in the ‘Eastern– Western’ and ‘Southern– Northern’ framework of analysis of the world economy calls for adjustment of the existing international

14 L. Deng, ‘Research on China’s International Taxation under the Great Changes Unseen in a Century’, International Taxation 2 ( 2020), 1–11.

The Chinese Perspective on International Tax Law 837 taxation order. The reality of multi-polarization and the difficulties of globalization are intertwined, giving international tax competition new connotations in addition to the pursuit of tax benefits. As the tax competition in developed countries often embodies more protectionism, and the corresponding international tax cooperation is more focused on the new problems of political multi-polarization and counter-economic globalization, the world needs to reorganize the historical roles and functions of the participants to reshape the international tax order. The OECD BEPS framework and the changing trend of the ‘post-BEPS era’ will better promote the system design in a deep reshaping of international tax rules. If the ‘BEPS era’ is a result of dealing with tax base erosion and profit shifting of multinational corporations under economic globalization, then the ‘post-BEPS era’ is moving forwards in step with the reality of world multi-polarization and the difficulties of economic globalization. In this context, today’s research on international taxation must not only continue to examine the impact of economic globalization and de-globalization within the OECD BEPS framework under the changing circumstances, but will also have to look outside such framework and shift its focus to the profound impact of the world’s multi-polarization. When studying international tax reform in the ‘post- BEPS era’, it is necessary to actively study the participation, feedback and, influence of developing countries, including China, on the BEPS Project and pay more attention to the demands of developing countries. Furthermore, today’s international taxation research will, of course, focus partly on China’s contribution to ‘Belt and Road’ tax cooperation, and prevent the tendency to place Belt and Road tax cooperation inside the OECD BEPS framework. Special efforts should be made to increase China’s institutional supply for multilateral governance of international taxation. In this regard, China’s Belt and Road Initiative15 and US trade protectionism contrast the rationality of today’s reconstruction of the international economic order. China is contributing its market, capital, and technology to the growth of the world economy and should further consolidate it place as a cornerstone of world multilateralism governance. The continuous development of the Belt and Road Initiative calls for the inclusive growth and sustainable development of the global economy and reflects the historic responsibility of creating a community with a shared future for mankind.

45.4.2 Tax Base Exploration and Design of China’s Policy System Based on the preceding discussion, as a developing digital country with a socialist market economy, China should establish a comprehensive and active international tax policy for the digitalized economy from the three levels of tax base security, fairness,

15 The

initiative is a strategy initiated by China that seeks to connect Asia with Africa and Europe via land and maritime networks with the aim of improving regional integration, increasing trade, and stimulating economic growth.

838 Bristar Mingxing Cao and development. The first is to actively participate in the BEPS mechanism of global cooperation in anti-tax avoidance which is led by the G20 and OECD, and to strengthen information transparency and administration. The second is to promote the reform plan to expand the formal integrity of fair rules and to design fairer tax jurisdictions across the entire industrial chain from the industrial economy to the digitalized economy, so as to fundamentally achieve a balance between the digitalized economy and the real economy. Finally, based on the concept of the Belt and Road Initiative, we advocate a joint tax base construction and profit sharing by rationally using tax incentives, in order to help developing countries strengthen their complete fiscal capacity building, and to promote economic growth in least developed regions and emerging industries. Returning to the basic principle of ‘profits should be taxed in the place where economic activity occurs and value is created’ put forward by the OECD and G20, in order to promote the fair development of tax base between market-source countries and resident countries, we propose a Chinese scheme of international taxation—Base Exploration and Profit Sharing, which would be the Neo-BEPS scheme. Neo-BEPS would be the embodiment of the Belt and Road Initiative principle of negotiation, construction, and sharing in the field of international taxation. Driven by the concept in the Belt and Road Initiative of inclusive growth, Neo-BEPS would creates a more proactive international order which would be growth-oriented and tax-sharing, and would adopt a simple, definite, efficient, and fair international tax system to promote the joint exploration of tax bases, profit sharing, economic development, and the realization of substantial social justice and order.16 The basic idea of Neo-BEPS would be to deepen the integration and adjustment of the two cornerstones of judgment standards that support the current international tax rules (the permanent establishment nexus and the arm’s-length principle) to become the Principle of Personal Existence. In response to the principle of independent transactions, under which states currently administer tax based more on residents and permanent establishments, under the new mechanism states would operate taxation through effective incentives for individuals by cross-tax crediting. In this way, the principle of independent transactions would be achieved through tax competition between the individual and the enterprise, the unity of individual interests, enterprise interests, and national interests realized under market conditions, with profits taxed where the economic activities occur and where the value is created, and the tax base would also be realized. The main points of the detailed operation of the mechanism would be that a multinational digital enterprise would itself calculate its payable global income tax according to the national tax law of its residence. The multinational enterprise would distribute income tax and pay it to the national treasury of the user’s location according to the contributions of its global users (including producers and consumers). The income 16 B. M. Cao, ‘Neo-BEPS: China’s Proposal for International Tax Reform from the Perspective of the Belt and Road Initiative’, in M. Lang and J. Owens, eds, Removing Tax Barriers to China’s Belt and Road Initiative (Alphen aan den Rijn: Wolters Kluwer, 2019), 3.

The Chinese Perspective on International Tax Law 839 tax allocated by the enterprise to the national tax authority would be applied by that authority to credit or deduct personal income tax or to enjoy more and better public services. In that process, multinational companies would have incentives to distribute income tax to users, because it could help them to win more users and expand the size of their market. The country of residence would be happy to surrender some tax for resident business enlargement in market countries, and the market country would also obtain tax benefits which it would not without such scheme.17 It seems that we will not arrive at a global consensus in the short term, either at the UN level or within the OECD framework, but we still have to base reform on the sound principles of ‘neutrality, fairness, and flexibility’. To achieve a neutral principle of the tax system, any mechanism should allow various enterprises (not only digital enterprises, but also traditional enterprises) to determine the price of various production-element owners, such as users and capital in the market. ‘Fairness’ means that the mechanism should arrive at fair results under market conditions, and achieve fair results all for enterprises, users, market countries, and resident countries. ‘Flexibility’ means leaving behind the narrow aim of tax base competition for the realization of the vitality of economic development, the creation of the driving force of economic development, and promotion of the prosperity of the global community.

45.5 Conclusion In the era of a digitalized economy, international taxation-governance mechanisms need to be reshaped. With the increasingly deepening recognition in the multi- dimensional framework of cross-border transactions in the digitalized economy, in the multi-party competition pattern of residence, source, market, and tax havens, as well as in the urgent need to deal with the problems of tax base erosion and tax base mismatch, we will come up with a Chinese viewpoint and Chinese plan for the global governance of international taxation in the digitalized economy. This requires us to base the reform on China’s characteristic digitalized economy, by coordinating the policy goals of technological progress, fairness of the rules, and growth sharing in carefully determining the scope of application, scientifically establishing tax nexus, reasonably setting income thresholds, comprehensively recognizing the role of governments, reasonably re-evaluating tax incentives, and actively promoting the sharing of growth in order to gradually establish a coupled tax system for digital production and consumption, a matching tax system for digital and real economies, and a coordinated tax system for tax revenue and economic growth.

17 W. Wang and B. M. Cao, ‘Systematic Response to the Challenges of the Digital Economy: An International Taxation Governance System of Global Market Unity and Competition (GMUC)’, Working Paper. Peking University Law School, 2021.

840 Bristar Mingxing Cao Perhaps our ideal is to step up the anti-tax avoidance mechanism of the OECD BEPS framework to a new order with the Neo-BEPS framework in the Belt and Road Initiative. But there is a prerequisite that needs serious attention; that is, there is no supranational sovereignty in the world and stages of development are not synchronized; under these conditions, in order to achieve fairness that both conforms to the law and achieves substantive fairness, the design of rules has a long way to go. However, in this special historical period of the Centurial Change, the world has to look into the fog of bilateral, multilateral, regional, and global ‘rule changes’ that developed countries are accelerating, brewing, and dominating, and effectively counteract the hegemonism, protectionism, and singularity present under the banner of ‘rules’, and promote the establishment of a multilateral system based on fair and reasonable rules. This requires us to participate intensively in the great changes in the global governance system and to actively propose Chinese solutions for the better protection of the rights and interests of the developing countries, including those of China.

Chapter 46

T he Indian Perspe c t i v e on In ternationa l Tax L aw Kuntal Dave

46.1 Introduction: Indian Economy and International Trade [India] will embark on rationalization and simplification of the tax regime to make it non-adversarial and conducive to investment, enterprise and growth. (Statement by the President of India before a joint sitting of Parliament on 9 June 2014 to describe the current government’s approach to tax policy) a unified approach to the nexus and profit allocation challenges is a promising one that merits serious attention . . . A solution that is simple to implement, simple to administer and simple to comply with is needed. (Indian Finance Minister at the G20 Finance Ministers and Central Bank Governors meeting in October 2019)

India collaborates on various initiatives on matters relating to the OECD’s tax policy formulation, and is an important contributor to a range of OECD standard- setting activities. India, as member of the Inclusive Framework, is also committed to implementing the Base Erosion and Profit Shifting (BEPS) Project and has introduced several measures as recommended in the BEPS Action Plans in its domestic tax law. India plays a significant role in the world’s economy, trade, and commerce. According to the International Monetary Fund (IMF), India’s GDP of $2.6 trillion is 3.01% of global GDP. During the financial year 2019/20, India attracted a foreign direct investment inflow of $49.98 billion. With increased globalization, changes in business models, and the rise of the digital economy, it is difficult to ring-fence businesses by the jurisdictions involved. India seeks to implement tax policies which are unambiguous,

842 Kuntal Dave transparent, and predictable and to provide rules that contain effective dispute- resolution mechanisms to improve investor confidence. However, triggering to the concern of the global investor community, the Indian tax authorities have, on various occasions, acted in a way that have given rise to unprecedented tax controversies, thus raising questions about its commitment to investors as well as to its treaty obligations. There have been cases1 where the tax authorities have applied the provisions of domestic tax law in an attempt to override a double taxation avoidance agreement (DTAA). On various occasions, including in the decision in Volkswagen Finance,2 the tax authorities have also sought to tax income earned by foreign companies outside India, deeming such income as having its source in India. To meet that end, the provisions of domestic law have been given a very wide meaning. Clearly, the controversy surrounding the Vodafone case3 and the retrospective amendment of the tax law led to fear in the minds of the investor community, with such action being perceived as ‘tax terrorism’. The worldwide Covid pandemic gave rise to a very delicate situation, with a contracting economy on the one hand, and the expectation of support and relief from governments on the other hand. Even during such difficult times, India introduced relaxation measures and extensions to various compliance due dates to help the economy. In addition, in the budget presented in February 2021, the government did not increase taxes in order to support taxpayers through the difficult period. The recently introduced faceless assessment and appeal schemes, will bring about a radical change in the tax assessment and litigation process and will reduce bias and other issues hindering impartiality. The following sections will provide an overview of income tax law in India as well as its interplay with its treaty obligations, as prevailing on 1st May, 2021

46.2 Overall Scheme of the Income-Tax Act, 1961 Section 5 of the Income-tax Act of 1961 determines the scope of total income for residents and non-residents. Taxation depends on residential status of the taxpayer and on place of the accrual and receipt of income. A person’s residential status is determined for the previous year.4 A person (which includes individuals, partnerships, and companies) 1

PILCOM 116 taxmann.com 394 (SC); AAR-Bid Services Division (Mauritius), Ltd [2020] (2) TMI 1183. 2 115 taxmann.com 386. 3 [2012] 1 SCR 573. 4 The previous year is defined to mean the financial year (i.e. from April to March) immediately preceding the assessment year (also from April to March). Income earned in the previous year is assessed in the assessment year (section 3 of the Act).

The Indian Perspective on International Tax Law 843 Table 46.1 Determination of residential status Residential status

Type of taxpayer

Taxable income

Resident

All

Global income is taxable

Not Ordinarily Resident

Only individuals and HUFs*

Foreign income is taxable only if derived from a business controlled in or a profession set up in India. Income earned abroad, e.g. salary, dividend, interest, capital gains, royalties, etc., are not taxable unless received in India.

Non-Resident

All

Taxable income includes income received or deemed to be received in India, or accrues or arises or is deemed to accrue or arise in India.

* Hindu Undivided Family: a traditional joint family considered as a separate taxable entity.

can be either resident or non-resident (NR); however, resident individuals are further classified as ‘resident’ and ‘not ordinarily resident’. Different tests are prescribed for determining residential status. See Table 46.1. Section 9 of the Act provides elaborate source rules for determining the income of NRs that can be taxed in India which it defines as ‘all income accruing or arising, whether directly or indirectly, through or from any business connection in India, or through or from any property in India or through or from any asset or source of income in India or through the transfer of a capital asset situate in India’. Thus, the scope is sufficiently wide to include all income which has source in or a nexus with India. Continuity, a real and intimate connection, attribution of income, and common control are the fundamental ingredients of a business connection.5 The amended provisions of section 9 now cover: (1) enlarged rules for concluding ‘agency Permanent Establishment (PE)’ which is prevalent under the UN Model Convention (MC), OECD MC, and the BEPS Action Plans;6 (2) transactions of ‘indirect transfer’;7 and (3) nexus through a significant economic presence (SEP) in India.8 5 See R. D. Agarwal and Co. 56 ITR 20 SC; Hira Mills Ltd 14 ITR 417; and Qualcomm International Inc. 93 taxmann 80. 6 Action 7 of the BEPS Project deals with the artificial avoidance of PE status. AP 7 seeks to update the definition of a PE to prevent artificial avoidance of PE status especially through commissionaire agreements, specific-activity exclusions, splitting contract, fragmentation of activities, etc. These changes were brought about through the MLI to modify art. 5 of tax treaties. 7 When the shares of a foreign company are transferred and such shares directly or indirectly derive significant (more than 50%) value from the shares of an Indian company and assets located in India, this results in the ‘indirect’ transfer of the shares of the Indian company and is chargeable to capital gains tax in India. 8 SEP is a nexus concept introduced in the Act in line with BEPS Action Plan 1. It establishes a nexus between the income of non-residents and source in India in instances where there may not be a physical presence in India, but there is a significant economic presence, especially in highly digitalized business models. Having an SEP in India is similar to having a PE in India, resulting in all income earned through

844 Kuntal Dave Section 90(2) of the Act offers a unique choice whereby a NR can choose to be governed by the provisions of the Act or the applicable DTAA, whichever is more beneficial. Consequently, despite a source rule in a DTAA conferring taxing right to India, a NR can opt for domestic law to apply if that is more beneficial. An example of this is fees for technical services, where the rate under a DTAA is often higher than the 10% rate under the Act. Tax credits on foreign-sourced income to which a DTAA applies are covered in section 90. Besides providing for ordinary tax credit, some DTAAs provide other methods for avoiding or relieving double taxation: (1) full exemption;9 (2) exemption with progression;10 and (3) full credit.11 The majority of Indian DTAAs contain a ‘tax sparing clause’ which acts as an added incentive for cross-border trade and business, as it results in intended double non-taxation. However, with the implementation of the Multilateral Instrument (MLI), such unfair and unequitable benefits are likely to undergo change.

46.3 Rates of Tax for Residents and Non-R esidents Rates of corporate tax, including surcharge and health and education cess, are based on the residence of the company. The effective tax rate for domestic companies (including foreign-owned subsidiaries) and other legal entities are in the range of 15–35%, depending on specified conditions being fulfilled. A NR company is taxed at around 42% on net business income. Other reduced tax rates (ranging from 5% to 20%) apply depending on the nature of the income and fulfilment of applicable conditions.

46.4 Interplay Between Domestic Law and a Treaty (Including the MLI) Under the Income-tax Act, a taxpayer resident in another jurisdiction can choose to be covered by the provisions of the Act or a DTAA, whichever are more beneficial. the SEP becoming taxable in India. SEP has been defined as a ‘transaction in respect of any goods, services or property carried out by a non-resident in India, including the provision of download of data or software in India, subject to payment threshold to be prescribed; or systematic and continuous soliciting of business activities or engaging in interaction with such number of users as may be prescribed, in India through digital means’. In May 2021, the CBDT issued the thresholds, as transactions amounting to INR 20 million or more, and having 300,000 users or more, which would be effective from financial year 2022/23. 9

DTAAs with Austria (for Austrian residents only), Greece, and Hungary. DTAAs with Australia, Cyprus, Germany (for German residents only), Japan (for Indian residents only), UK, etc. 11 DTAA with Germany in respect of Indian tax paid on dividend/interest/royalty/FTS/capital gains on shares/director’s fees/income of artistes and sports persons. 10

The Indian Perspective on International Tax Law 845 Conditions are prescribed in the Act for a taxpayer to secure treaty entitlement, such as furnishing of a tax residency certificate (TRC) from the relevant jurisdiction and other necessary information/documents. Additionally, treaty benefits can be denied if general anti-avoidance rules (GAARs), are involved (see Section 46.9.1). A circular12 issued by the Central Board of Direct Taxes (CBDT) to clarify the eligibility of NRs to claim treaty benefits, as upheld in a case before the Supreme Court,13 has still not dissuaded tax authorities from denying such treaty benefits. At the core of the controversy are the method and manner of interpreting the term ‘liable to tax’. Since the Act did not define the term, it led to several controversies, especially with regard to specified NR entities14 and residents of specified jurisdictions.15 The Finance Act, 2021 inserted a definition of the term16 and appears to be an attempt by the lawmakers to consolidate, declutter, and demystify the issues surrounding treaty entitlement.

46.5 Interplay Between a Treaty and the Multilateral Instrument India was among the sixty-seven signatories to the MLI in Paris on 7 June 2017. The government ratified it on 13 June, 2019, depositing the instrument of ratification on 25 June 2019, along with the final position in respect of covered tax agreements (CTAs). The MLI entered into force for India from 1 October 2019 and on various dates for other state-relevant CTAs. On and from the financial years 2020/21 or 2021/22, as the case may be, the provisions of the MLI have effect for the respective DTAAs with jurisdictions which have notified their DTAAs with India as a CTA (thirty-eight as at 14 August 2020). India has:

(1) (2) (3) (4)

expressed no reservations with regard to articles 4–15 of the MLI; opted for article 3 and 17 not t0 apply; selected option C of article 5 and option A of article 13; and expressed a reservation on article 16, paragraph 1.1.

In relation to the articles on which India has expressed reservations, having regard to the position of the other jurisdiction, the relevant treaty benefits may be subjected to

12

Number 789 of 2000. Azadi Bachao Andolan [2002] 125 Taxman 826 (SC). 14 e.g. partnership firms (see Linklaters LLP 40 SOT 51; A. P. Moller TS-555-ITAT-2013; Linklaters & Paines 65 SOT 266; P & O Nedlloyd Ltd 369 ITR 282; Maersk Line UK Ltd 68 taxmann.com 173. 15 e.g. Mauritius and the UAE. 16 Section 2(29A): ‘liable to tax’ in relation to a person and with reference to a country, means that there is an income-tax liability on such person under the law of that country for the time being in force and shall include a person who has subsequently been exempted from such liability under the law of that country.’ 13

846 Kuntal Dave further conditions (e.g. relief through CAs, application of the principal purpose test (PPT), application of the simplified limitation of benefits (SLOB), application of a minimum holding period, the rule of splitting up contracts, anti-fragmentation rules, etc.) or may impact the DTAA’s interpretation. Broadly speaking, articles 6 and 7 are likely to apply to all Indian CTAs; the anti-abuse rules of the relevant articles are likely to apply to a few Indian CTAs; and article 12, 13, 14, and 15 (amending art. 5 of the MC) will impact a large number of CTAs.

46.6 India’s Position on Select Articles of the MLI With regard to India’s overall approach to addressing BEPS, the MLI will have a great impact. Instances of treaty abuse will be addressed and undesirable interpretations of DTAAs will be challenged. India has reserved the right for the following articles not to apply to its CTAs in their entirety: • article 3 on transparent entities; • article 5 on methods of eliminating double taxation; • Part VI of the MLI (arts 18–26) on mandatory and binding arbitration. For article 7 of the MLI (PPT and SLOB provisions), India has chosen to opt for applying SLOB provision to its CTAs and, therefore, where the treaty partners make a similar election, SLOB will apply. For article 8 (minimum holding period for taxation of dividends), India has opted to apply the article except for CTAs which already require a longer holding period. In respect of capital gains from alienation of shares or interests of entities deriving their value principally from immovable property, India has opted to apply article 9 to require that such gains are taxable if more than 50% of their value was so derived at any point during the 365 days preceding alienation. Articles 12, 13, and 14 of the MLI (avoidance of PE status) are covered in Section 46.10.

46.7 Interpretation of Tax Treaties It is recognized that rules for the interpretation of DTAAs should differ from those applicable to tax laws. Unlike tax laws, which contain highly technical legislative language relevant to a specific jurisdiction, DTAAs are based on a mutual understanding between two or more contracting states. DTAAs specify general tax principles whereas domestic

The Indian Perspective on International Tax Law 847 tax laws seek to impose tax under specific circumstances. In applying DTAAs, most countries invariably use the meaning of the term defined in their domestic law. Indian courts have recognized the concept of uniform interpretation and held that judgments pronounced by foreign courts are also relevant for interpreting DTAAs.17 Rules laid down by the judiciary on the interpretation of tax treaties can be summarized as: • The office of a judge is not for legislating but to express the intention of the legislature. The whole statute should be read together and construed together.18 • The approach should be purposive and integrative, seeking to give effect to the intention of the legislation.19 • The rules of interpretation in respect of international treaties are different from those applicable in respect of domestic law.20 • A tax treaty is an agreement. It is not a statute of taxation and on that basis the agreement should be construed as ut res magis valeat quam pereat (to make it workable rather than redundant), as should all agreements.21 • When an expression has not been defined in a DTAA, it will have to be interpreted in harmony with the treaty read as a whole.22 • If two views are possible, the view which is favourable to tax payer should be adopted.23 • The principle of literal interpretation does not apply when interpreting a DTAA.24 • Support can be provided by any applicable examples provided in the treaty.25 • Identical or similar expressions used in different treaties should be given the same interpretation.26 The OECD Commentary or a report of the Technical Advisory Committee27 is to be taken into account in the interpretation of a DTAA. • the treaty is to be read in whole and not in isolation28 and must be interpreted in good faith.29 • A retrospective amendment in the Act cannot be read into the treaty.30

17

Visakhapatnam Port Trust 144 ITR 146. Stock v. Frank Jones (Tipton) Ltd. 19 In co Europe Ltd v. First Choice Distribution. 20 Azadi Bachao Andolan 263 ITR 706 (SC). 21 Union Texas Petroleum Corp v. Critchley (1988) STC 69. 22 Ensco Maritime Ltd 91 ITD 459. 23 Vegetable Products Ltd 88 ITR 192 (SC); Reliance Communications Ltd 52 CCH 292. 24 Azadi Bachao Andolan 263 ITR 706 (SC), Hindalco Industries 94 ITD 242. 25 SNC Lavalin International Inc. 332 ITR 314. 26 Patni Computer System Ltd 84 CCH 116; Asia Satellite Telecommunication Co. Ltd 332 ITR 340. 27 Asia Satellite Telecommunication Co. Ltd (332 ITR 340). 28 FRS Hotel Group Lux Sarl 404 ITR 676 AAR. 29 Sanofi Pasteur Holding SA 354 ITR 316. 30 New Skies Satellite NV 382 ITR 114; Taj TV Ltd 161 ITD 339. 18

848 Kuntal Dave • Recourse can be made to extrinsic material31 (e.g. past history relevant to the treaty, preparatory work, interpretative protocols,32 resolutions, committee reports, subsequent agreements, the conduct of the parties, other treaties in pari materia) to eliminate interpretation leading to ambiguity, absurdity,33 being unworkable,34 or being unreasonable. The judgment of the Bombay High Court in the case of Aberdeen Institutional Commingled Funds LLC35 is notable as it reaffirmed the concept of private international law, wherein the laws of a foreign country are recognized and can be relied upon for interpretation when the issue pertains to a person to whom such law applies. In the case at hand, the status of a foreign trust converted into a company under the laws of its jurisdiction of incorporation was accepted instead of determining its status under Indian law. Such foreign laws can only be ignored or disregarded if they are contrary to the public policy of India and must relate to the principles of morality and justice.36 The MLI, which is similar to one of the subsequent agreements entered into, will form an important basis for the interpretation of DTAAs and will be applied alongside existing DTAAs, modifying their application, unlike a protocol which directly amends the DTAA’s text. The MLI will either supplement, complement, supersede, or modify the application of existing DTAAs to bring them in line with the BEPS Action Plans recommendations.

46.8 Resolution of Tax Disputes Elaborate provisions are incorporated in the Income-tax Act as well as DTAAs to provide mechanisms for addressing tax disputes. On the one hand, there are provisions in the Act to seek advance rulings37 and a dispute-resolution provisions in the Act38 as well as in DTAAs (i.e. the mutual agreement procedure (MAP)). It is reported that India has improved its administration, thereby encouraging many multinational enterprises (MNEs) to seek advance rulings and enter into advance pricing agreements (APAs). It has also settled many MAP cases within the average prescribed time period of twenty- four months. On the grounds of sovereignty, India does not favour using arbitration

31

Rajeev Sureshbhai Gajwani 129 ITD 145. Steria India Ltd 386 ITR 390. 33 Norasia Lines Malta Ltd 279 ITR 268. 34 NHAI OMP 633/2012 dt. 8 July 2014; Set Satellite Singapore Ltd 106 ITD 175; Ashapura Minichem Ltd 131 TTJ 291. 35 [2019] 104 taxmann.com 63 (Bombay). 36 Technip SA v. SMS Holding (P.) Ltd [2005] 60 SCL 249 (SC). 37 Ch. XIX B, advance rulings; as well as s. 92CC (advance pricing agreements) and s. 92CB (power of board to make safe harbour rules). 38 S. 144C (reference to dispute-resolution panel); and Ch. XX, appeals and revision. 32

The Indian Perspective on International Tax Law 849 to settle tax disputes; it has challenged, unsuccessfully, the right of taxpayers to invoke arbitration under the provisions of bilateral investment protection agreements and has amended its model in that regard and has also, as expected, not agreed to the MLI proposal for tax arbitration for tax dispute resolution.

46.9 Recent Developments in Indian Domestic Law Impacting International Tax Law 46.9.1 Introduction of General Anti-Avoidance Rules The introduction of general anti- avoidance Rules (GAAR) were introduced in Chapter X-A of the Income-tax Act to counteract abusive tax avoidance and protect the country’s tax base from erosion. GAAR came into effect from April 2017 and transactions concluded prior to that date have been grandfathered. There are certain monetary thresholds and exemptions from applicability of GAAR and it is to be invoked only where there is an ‘impermissible avoidance agreement’ (IAA), the main purpose of which is to obtain a tax benefit. A procedure has also been laid down for tax officers to have pre-apply to the designated authority in order to invoke GAAR. When invoked, a transaction, or part or step thereof, may be disregarded, combined, or recharacterized to tax the transaction. Taxpayers are entitled to appeal against GAAR orders either before the appellate tribunal or by a writ before the jurisdictional High Court. The advance- ruling mechanism helps to ensure that there is tax certainty before any transactions are structured. The CBDT has clarified39 certain issues in this regard: (1) GAAR will not infringe a taxpayer’s right to select a method for implementing a transaction. (2) GAAR will not be invoked merely on the grounds that an entity is located in a tax-efficient jurisdiction. (3) There is no application of GAAR if the tax avoidance is sufficiently addressed by an LOB clause in a treaty. (4) GAAR can coexist with specific anti-avoidance rules. (5) If a court or National Company Law Tribunal40 has ‘explicitly and adequately’ considered the tax implications while sanctioning an arrangement of business reorganization or restructuring, then GAAR will not apply.

39

See Circular No. 7 of 2017 of 27 January 2017. A National Company Law Tribunal (NCLT) is a special court that hears matters and applications pertaining to the Indian Companies Act, 2013, especially applications for business combinations, arrangements, demergers, etc. 40

850 Kuntal Dave As stated earlier, if the purpose of an IAA is to obtain a tax benefit, the DTAA benefit can be denied by invoking GAAR. As the MLI contains the PPT in order to deny DTAA benefits, it will be interesting to see the interpretation and administration by the tax authorities on the interplay between GAAR and the MLI. It is also important to note that to invoke GAAR a detailed procedure has been laid down even though to apply PPT under the MLI no such procedures have been set down. There is no grandfathering under the MLI whereas the same has been provided under GAAR.

46.9.2 Interplay of GAAR and LOB The intention of LOB is to ensure that DTAA benefits are limited to genuine residents of either treaty state. GAAR, on the other hand, could be invoked to disregard a structure in order to determine residence. The Shome Committee41 on GAAR stated that a LOB clause, being a specific law, should take precedence over GAAR. This is also well accepted practice in other jurisdictions that have GAAR. In that regard, it is pertinent to examine the relevant provisions of the Act. Section 90 (2A) contains a non-obstante clause which reads: Notwithstanding anything contained in sub-section (2), the provisions of Chapter X-A of the Act shall apply to the assessee even if such provisions are not beneficial to him . . .

As a result, section 90(2), which allows for DTAA to be applied instead of the Act to the extent that it is more beneficial, can be overridden and, if an arrangement is held to be impermissible by invocation of GAAR, DTAA benefits can be denied. There is a lack of clarity about whether LOB under a DTAA has dominance over GAAR under the Act. Thus, at present, GAAR may override even those DTAAs that have a LOB clause. This leads to a potential situation where an arrangement might be hit by GAAR in spite of passing the LOB test. If an arrangement is primarily for commercial purposes, that would be permissible and any incidental tax benefits would be immaterial. On the other hand, even if part of an arrangement was designed to obtain a tax benefit, it could be hit by GAAR. Not satisfying the LOB clause would only mean that the taxpayer concerned would not be entitled to DTAA benefits. On the other hand, if an arrangement is deemed to be an IAA under GAAR, it could be disregarded, recharacterized, or considered as never having been entered into; the repercussions would extend beyond mere tax liability. This has serious implications for international transactions which may be compliant with DTAA LOB provisions. 41

The Shome Committee was established by Dr Manmohan Singh and was responsible for guiding the Indian government on the application of GAAR. The Committee issued a report for consideration by the government highlighting the challenges and issues in the implementation of GAAR.

The Indian Perspective on International Tax Law 851

46.9.3 Equalization Levy Several measures were discussed as part of BEPS Action Plan 1 and effort to tackle taxation of the digital economy. As multilateral solutions have stalled, several countries have introduced unilateral measures, especially in the form of digital services taxes (DSTs). In 2016, India introduced an Equalization Levy (EL) of 6% as a separate levy to income tax that applied to certain transactions involving non-residents. ‘Specified services’ include online advertisement, provision of digital advertising space, or any other facility or service for the purpose of online advertising and it includes any other service as may be notified by the government, provided by a NR to a resident, or to an Indian PE of a NR. The scope of the EL was expanded in 2020 and a 2% EL was applied to certain e- commerce transactions, such as the facilitation of goods or services by ‘e-commerce operators’ to: (1) Indian residents; (2) NRs in respect of sale of advertisements targeted at residents or using Indian IP addresses; (3) a person who buys goods or services using an Indian IP address The 2% EL functions more as an indirect tax to be discharged by the e-commerce operators, unlike the 6% EL which works like a withholding tax to be deducted by the payer. Since EL is a separate levy from the Income-tax Act, DTAA provisions do not apply and nor would tax credit be available to the NR, potentially resulting in double taxation. The definition of the terms ‘e-commerce operator’ and ‘e-commerce supply or services’ are fairly wide in scope and may cover various digital transactions and services. Along with SEP, these changes to the Act may have overlapping coverage on certain digital transactions, potentially resulting in conflicting claims on characterization. Accordingly, the interplay and coordination between these provisions need to be evaluated in detail. However, the Act provides that any amount covered under EL will be exempt from tax under the Act. Further, the wide scope of EL is likely to have an impact on a number of MNEs, given the increasingly pervasive nature of digitalization. The procedures relating to collection and recovery of EL are expected to increase the compliance obligation for NRs. Overall, the implementation of such unilateral measures to tax digital transactions indicates India’s inclination to address the broader tax challenges posed by digitalization by seeking additional taxing rights for the user or market jurisdiction. The Indian government has indicated that the measure will be withdrawn once a multilateral, consensus-based solution is achieved.

852 Kuntal Dave

46.9.4 Master File and Country-by-Country Reporting As a part of BEPS Action Plan 13, India has introduced rules for maintaining and furnishing the master file (MF) and country-by-country reporting (CbCR). An MF is to be filed by every constituent entity (CE) of an international group whose consolidated revenue is INR 5 billion and the value of international transactions with an associated enterprise (AE) exceeds INR 500 million, or its purchase, sale, transfer, lease, or use of intangible property exceeds INR 100 million. The MF’s contents are as per Action 13 with certain additional information. Stakeholders have raised concerns about the confidentiality and security of information, and in the case of India the responsibility for the security of such information lies with the income tax authorities. The timelines for filing an MF work well for an Indian outbound group but may not be consistent for an inbound group if the parent entity of an inbound CE has a different accounting year and filing deadline. Also, from a practical perspective, inbound Indian CEs may not have access to the MF unless filed or made available by the parent entity. CbCR is to be filed by every CE of an international group whose consolidated revenue is INR 55 billion where: • the parent entity is a resident of a country with which India has a CbCR exchange mechanism; or • an exchange framework exists but there has been a systemic failure in exchanging information.

46.9.5 APAs An APA is an agreement between the CBDT and the taxpayer determining in advance the arm’s-length price or specifying the manner of the determination of an arm’s-length price (or both), in relation to an international transaction. APAs remain in force for maximum of five years and are binding on both taxpayers and tax authorities. Currently, India has entered into 300 APAs. Covid-19 resulted in an extraordinary economic crisis and resulted in huge differences in the functional profiles of the parties involved in APAs and with changes in critical assumptions having a huge impact on profitability. Accordingly, MNEs can assess the impact and opt to renegotiate their APAs with the tax authorities, as provided under the Act.

46.9.6 Faceless Assessment Scheme The newly implemented faceless assessment scheme (FAS) provides taxpayers with a seamless, painless, and faceless form of assessments and appeals with better efficiency,

The Indian Perspective on International Tax Law 853 transparency, and accountability, eliminating the physical interface between the taxpayer and tax officers. FAS covers all assessments including transfer pricing, except cases involving serious fraud, major tax evasion, sensitive cases and search matters, cases involving undisclosed foreign income and assets, and Benami property (i.e. property whose beneficial owner is a person other than the recorded and purported owner). Some of the key features of the FAS are : • selection of the assessment unit/jurisdiction to be automated and randomized using data analytics and artificial intelligence; • electronic issuing of notices and elimination of tax office visiting; • draft orders, review, and finalization by different officers; • system-driven appeals allotment; • online replies to notices, document submissions, and hearings; • digital trail of every action of the Revenue and the taxpayer with recall value and search functions, cross linkages, digital tracking, etc.

46.9.7 Foreign Income of Residents 46.9.7.1 Black Money Act Residents are taxed on their global income and the tax legislation has been strengthened by introducing several provisions to tax foreign-source income. The Black Money (Undisclosed Foreign Income and Assets) and Imposition of Tax Act, 2015 was enacted to tax previously undisclosed foreign income and assets. The Finance Act 2012 expanded the scope of ‘income escaping assessment’ to include undisclosed assets located outside India. The time limit to reopen earlier tax assessments was also increased from four/six years to sixteen years, reduced to ten years from 2021 onwards. The Black Money Act contains far more severe implications, penalties, and opportunities for prosecution than were previously prescribed.

46.9.7.2 Place of Effective Management Prior to the financial year 2016/17, foreign companies were regarded as a tax resident of India if they were ‘wholly controlled and managed’ in India. In 2015, the residence test was amended to prescribe that a foreign company would be considered a tax resident in India if its place of effective management (POEM) was in India. The challenges faced by a company deemed by virtue of POEM to be resident in India (e.g. calculation of depreciation and valuation of assets, set-off and carry- forward losses, foreign tax credit) are sought to be addressed by the insertion of a new section 115JH. CBDT Circular No. 06/2017 of January 2017 lays down guiding principles for determination of the POEM of a company.

854 Kuntal Dave The Act uses POEM to determine the residency of a foreign company whereas, from an international law perspective, it is used as a tiebreaker rule. The principles espoused by Circular No. 06 while not decisive but guiding in nature, lack clarity with regard to a number of terms. The lack of any definite established legal factors for the determination of POEM could lead to several disputes and litigation.

46.10 Change in the Concept of Permanent Establishment With regard to the various judicial precedents and frequent amendments to the Indian laws on the subject of PE, it should be noted that ‘there is nothing permanent even in the concept of PE’. Table 46.2 summarises some of the recent court rules in India on the subject of PE. A careful reading of the decisions in the table will provide some important points for consideration which are discussed in the following sections.

46.10.1 ‘Fixed Place of Business’ and Use of Examples Even though court rulings against a taxpayer are due to typical fact patterns, the threshold for the existence of a PE appears to have been lowered. The authorities have referred to various examples cited in the OECD and other international literature. The majority of these examples had a fact pattern of the customer’s location being in the source state, however, controversy may arise when customers are not situated in the source state. It is quite likely that many other developing countries will examine these rulings closely. The question that arises is whether principles can be established through use of these examples.

46.10.2 Disposal Test In later judgments (e-funds),42 the Indian Supreme Court has affirmed the High Court’s analysis and conclusions. Consequently, it is incorrect to rely on factors such as a close association between the entities, the results of review and application of financial accounting and reporting studies, dependency on the earning of an Indian entity, the allocation of direct/indirect cost, the method and manner of remuneration to Indian entity and payment of royalties, and reports filed before regulatory authority outside India by the group. 42

Appeal no. 6082/2015 dated 24th October, 2017.

The Indian Perspective on International Tax Law 855 Table 46.2 Decisions by Indian courts on PE Decision delivered in

PE Type

Authority

Decision in Favour of

Determinative Factors/Criteria

Basic PE/Fixed Place PE* Formula Onea

April 2017

E-Funds

October 2017 Supreme Court Taxpayer

Dominos Indiab May 2018 c

Supreme Court Revenue ITAT, Delhi

Taxpayer

MasterCard

June 2018

AAR

Revenue

Nokiad

June 2018

ITAT, Delhi

Taxpayer**

• Initial onus on revenue. • 3 tests: Permanency. Fixed base & disposal. • Exclusion for preparatory & auxiliary services. • Sufficiency of fixed place at the disposal through the time required for business • Legal right to Own/Use • Automatic equipment can create a PE.

Service PE* E-funds

October 2017 Supreme Court Tax-player

MasterCard

June 2018

AAR

Revenue

• Service must be furnished at India. • Customer’s location & receipt of services. • Stewardship activities.

Agency PE* E-Funds

October 2017 Supreme Court Taxpayer

Dominos India May 2018

ITAT, Delhi

Taxpayer

MasterCard

Jun 2018

AAR

Revenue

Nokia

June 2018

ITAT, Delhi

Taxpayer**

• Authority or exercise of authority to conclude contract • Securing orders exclusively for Foreign enterprise. • ALP remuneration.

Subsidiary PE E-Funds

October 2017 Supreme Court Taxpayer

MasterCard

June 2018

AAR

Revenue

Nokia

June 2018

ITAT, Delhi

Taxpayer**

• Independent to fixed place PE. • Virtual projection –alter ego. • ALP remuneration.

■ Associated/Direct PE ■ Unassociated/Indirect PE

*Including through Indian subsidiary, as applicable. **Dissenting Member, vide note attached to the decision, opined that PE existed at India. Source: Slide by Kuntal Dave at IFA Congress 2018, Seoul, Seminar on Recent Developments in International Taxation. a Formula One : 394 ITR 80 (SC). bDominos India : 94 Taxmann 296. cMasterCard : 94 Taxmann 195. dNokia : I.T.As. No. 1963&1964/DEL/2001.

856 Kuntal Dave In contrast, in the F1 decision, the Supreme Court considered it necessary to examine the manner in which commercial rights had been exploited. Whether the fixed place of business (in this instance, the circuit) was at the disposal of the enterprise was based on who had real and dominant control over the event. The overall exposure on Indian soil was examined and, since the taxable event took place in India, it was held that the disposal test was met. The decision not only lowered the threshold for PE but also established a precedent leading to a major shift in the manner of conducting the ‘disposal test’. Additionally, the dissenting judge in the Nokia case stated that in view of the F1 case, in instances of indirect PEs the disposal test could be satisfied by reference to the agent who acted as the proxy in conducting the business. This, however, applies only as long as transactions are not at arm’s-length price.

46.10.3 Exploitation of Information on Indian Consumers While examining the relevance of the cost of equipment in the fixed place of business test in the MasterCard case, the tax authority acknowledged that information stored on a server outside India was valuable, particularly when such information is sold. Consequently, any commercial exploitation of information related to Indian customers/ users may attract scrutiny from the tax authorities. This has the potential to force a major shift in cross-border taxation of FAANG.43

46.10.4 Sanctity of the Article on Subsidiary PE Concerns have been raised about whether all arrangements are exposed to the PE risk and regarded as devices to artificially avoid PE. Will an Indian subsidiary undertaking marketing or administrative work at a fixed location in India imply the existence of a PE in all cases? This may result in questioning the very sanctity of article 5(7). A study of the court rulings suggests that where transactions are undertaken at arm’s- length price, the Indian subsidiary cannot be treated as a PE. The importance of undertaking transactions at arm’s-length price was highlighted in the Nokia case. Remunerating an Indian entity at an arm’s-length price is of key importance in defending the non-existence of a PE. Similarly, in structures involving the issuance of guarantees on behalf of an Indian subsidiary and agreements for non-dilution of equity holdings, the appropriate remuneration at arm’s-length price will be relevant criterion in while examining existence of a PE. This is consistent with the decision of the Supreme Court in the Morgan Stanley case.44

46.10.5 Dependent Agent PE—Scope under Domestic Law The newly substituted clause for a dependent agent PE (DAPE), amongst other changes, explicitly covers the activity of a person who ‘habitually plays the principal role leading 43

44

Acronym for Facebook, Amazon, Apple, Netflix, and Google. DIT v. Morgan Stanley and Co. Inc. [2007] 292 ITR 416 (SC).

The Indian Perspective on International Tax Law 857 to conclusion of contracts’. This may lead to ambiguity on the determination of the scope of the erstwhile DAPE rules for the relevant period vis-à-vis the enforceability of other clause specifically dealing with the role of ‘securing orders’ for NRs. In the case of MasterCard, agreements routed through the Indian subsidiary were considered to have fulfilled the test of ‘habitually securing orders’.

46.10.6 Business Operations versus Preparatory Auxiliary Activities Performance of services complementing a core business can lead to the conclusion that the Indian company was acting as a proxy. There appears to be very thin line distinguishing these activities, especially in context of digital transactions. It is also difficult to pinpoint key operations and where they occur. Issues may further arise if significant activities occur at multiple locations. Which activity or jurisdiction should take priority: (1) the location of credit card transaction; (2) the location of bank approving the transaction; or (3) the location where the data analysis took place? At the heart of these analyses is ascertaining whether the business is carried out in the source state or with the source state. It is pertinent to note that the OECD MC has undergone a change from reference to the place in which business is carried out to business carried on through a fixed location. In recognition of the ineffectiveness of applying the present tax rules for establishing nexus to the source state solely through a physical presence, particularly in the digital world, new definitions and rules are being explored. The perceptions of the adjudicating authority will play a determinative role and these rules will be tested for cogency and persuasiveness so that they can be cited in other jurisdictions. The concept of a PE has been evolving due to the jurisprudence, the legislative amendments, amendments to the model tax conventions and commentaries, and rules deviating under different DTAAs and caused by the MLI.

46.10.7 Select Amendments in the Income Tax Law 46.10.7.1 Amendments relating to the scope of ‘business connection’ As explained earlier, the concept of a business connection has undergone a sea- change in India in the past few years. Its scope has expanded from time to time due to developments in tax policy, domestically as well as internationally, and the emergence of relevant jurisprudence. The recent changes expand the scope of the Act to cover DAPE, transactions of an indirect transfer of the shares of an Indian company, and SEP.45

45

See Section 46.2.

858 Kuntal Dave

46.10.7.2 Attribution of profits to business connection/PE According to the settled legal position, only income attributable to business activities carried out in India is taxable in India. The method and manner of attributing such income is prescribed under rule 10 of the Income-tax Rules, 1962. In April 2019, the CBDT published a ‘Report on Profit Attribution to Permanent Establishments’ in which draft revised rules for computing profit attributed to a business connection were released, and public comments from stakeholders were sought. Two approaches have been proposed for profit attribution; namely, the function-asset-risk approach (FAR) and the demand and supply approach (D&S). The draft rules reject FAR by underlining India’s position on the OECD MC. For D&S, the report outlines two methods—‘formulary apportionment’ and ‘fractional apportionment’. Final rules in this regard are awaited. The proposed inclusion of market factors in the exercise of income attribution requires international agreement and changes to India’s DTAAs. Non-adoption by treaty partners may result in double taxation.

46.10.8 India’s Reservations to Article 5 OECD MC India’s reservations to article 5 of the OECD MC 2017 are discussed here to amplify the evolving Indian tax policy and approach in respect of expanding the PE-based source rules. • DAPE: India has observed that non-inclusion of the words ‘routinely’ and ‘to which it is closely related’ in paragraph 6 will not eliminate instances of the existence of a PE in India. Consequently, person(s) acting exclusively for an enterprise cannot be considered independent agents and a PE would arise even in case of a low-risk distributor arrangement where goods are sold exclusively for an entrepreneur. Many Indian DTAAs already contain clauses similar to article 12 of the MLI and, therefore, India has chosen to apply that article to its CTAs to provide for the creation of a PE through commissionaire arrangements and where persons play a principal role in securing, negotiation, and the conclusion of a contract. • A website may constitute a PE: India has observed that an enterprise can be considered to have a PE through a website which simply collects data (which can be shared with any another enterprise). • Construction PE: any work undertaken on site shortly after completion of construction, including repairs under guarantee, may be considered in determination of a PE. India has not stated a reservation to article 14 of the MLI which contains the anti-abuse ‘splitting of contracts’ provisions and, therefore, subject to a matching position by treaty partners, such a measure will be included. • Preparatory or auxiliary activities: collection of data for the purpose of determination or quantification of risk by an insurance company is not of a preparatory or auxiliary character. India has opted to apply option A under article 13 of the MLI

The Indian Perspective on International Tax Law 859 such that if the treaty partner makes the same election, the exemption for specific activities under the CTA will only be available if the activities have the overall character of being preparatory or auxiliary in nature. Further, the ‘anti-fragmentation’ rule will be included in such modified CTAs.

46.10.9 Recent Trends of Indian Tax Treaties for Determination of PE Post-BEPS, India has signed or renegotiated several DTAAs, notably with Mauritius, Singapore, Cyprus, China, and Hong Kong. Except for article 5(3) of the India–China DTAA which addresses PE avoidance through splitting of contracts, none of the other measures of anti-fragmentation, preparatory and auxiliary character-exemption, SEP, etc. have been inserted. Therefore, although India prefers PE-related anti-avoidance measures, it has been unable to negotiate its inclusion in the DTAAs.

46.11 Changes to the Concept of Fees for Technical Services and Royalties India follows source-based taxation, on a gross basis, for fees for technical services (FTS) and royalties. Such payments by a resident are taxable in India unless they pertain to business or a profession carried on outside India. Royalties or FTS payable by an NR are also taxable for the purpose of Indian business. Royalties and FTS are subject to tax at a rate lower than applicable to income of a different nature. Most Indian DTAAs provide for source-state taxation of royalties and FTS. However, with respect to business income, it is necessary for PE income to be taxable in the source country. Therefore, in the absence of a PE, the characterization of Indian source income as either royalties, FTS, or as business income is critical for taxation in India, which is dependent on such a characterization. FTS is defined as any consideration (including any lump sum consideration) for the rendering of any managerial, technical or consultancy services. The definition in certain DTAAs46 is narrower due to the ‘make available’47 clause. Even under the broader 46

See the tax treaties with the USA, UK, Singapore, and Switzerland. A ‘make available’ clause generally means, by providing FTS, that the provider of the services has equipped the recipient of the services in such a manner that the recipient of the services is enabled to do provide services independently/apply such knowledge independently. On the other hand, the a most-favoured-nation clause generally occupies its position in the protocol to a DTAA under which one country agrees to accord to the other contracting state treatment that is no less favourable than that 47

860 Kuntal Dave definition, as in the Act, it is possible to argue that standardized digital services are not FTS. The ‘make available’ clause only further strengthens the case. Taxing payments for the use of software under FTS has long been under debate. Important cases in this regard are: • Kotak Securities:48 not all services are FTS; charges paid by a member to transact business on the stock exchange is not a technical service. • Madyanchal Vidyut Vitran Nigam Ltd:49 transmission charges collected only involving the use of technology do not qualify as FTS. • Sky Cell:50 charges paid as subscriptions for telephone services do not entail a technical service agreement between the telecom service provider and user and are not FTS. As explained earlier, the EL covers all NR ‘e- commerce operators’ providing ‘e- commerce supply or service.’ The scope of the EL was widened to cover all types of digital services, going far beyond the current species of DSTs in other jurisdictions. With regard to the overlapping provisions and prevailing ambiguity on the interpretation of the relevant provisions, NRs are bound to face avoidable tax controversies and instances of double taxation. The clarification brought by the Finance Act, 2021 is a welcome move as it provides clarity and also specifies the scope of the EL. In respect of royalties, it is irrelevant where they arise or where services are rendered or an intangible was developed, as long as they were utilized for business or a source of income in India. However, if the payment for digital services/products is outside the scope of a ‘royalty’ as defined, then the general test of a business connection determines taxability in India. which it accords to certain other third states. It is relevant to note that the protocol forms part of the DTAA unless provided otherwise. It is interesting to see the interplay of a ‘make available clause’ with an MFN clause. In the case of Steria (India) Ltd 386 ITR 390, it was held that:

The words ‘a rate lower or a scope more restricted’ occurring in clause 7 of the Protocol envisages that there could be a benefit on either score i.e., a lower rate or more restricted scope. One does not exclude the other. The other expression used is ‘if under any Convention, Agreement or Protocol signed after 1-9-1989 between India and a third State which is a member of the OECD’. This also indicates that the benefit could accrue in terms of lower rate or a more restrictive scope under more than one Convention which may be signed after 1-9-1989 between India and a State which is an OECD member. The purpose of clause 7 of the Protocol is to afford to a party to the Indo-France Convention the most beneficial of the provisions that may be available in another Convention between India and another OECD country. 48

67 taxmann.com 356(SC). 88 taxmann.com 664 (Allahabad). 50 119 Taxman 496 (Madras). 49

The Indian Perspective on International Tax Law 861 Payments made by a resident for the right to use software have been the subject of dispute for nearly two decades. Recently, the Supreme Court51 settled the issue in favour of the taxpayers, holding that amounts paid by resident Indian end users/distributors to NR computer software manufacturers/suppliers as consideration for resale/use of computer software through end user licence agreements/distribution agreements, were not royalties for the use of copyright in the computer software, and hence there is no income taxable in India and no consequent obligation to withhold taxes. This decision has had a far-reaching impact on the IT sector (one of the most important contributors to the Indian economy).

46.12 Taxation of Interest India follows source-based taxation for interest, and taxes interest payable by a resident to an NR, unless the interest pertains to business or a profession carried on outside India. Interest payable by one NR to another is also taxable if the debt was for the purpose of business in India. Interest is taxable at rates other than those for income of a different nature. Interest payments are subject to withholding tax (WHT) at the rates between 5% and 20%, based on the nature of the debt and the identity of the lender. Thin capitalization rules were introduced in section 94B, in line with BEPS Action Plan 4. The provisions restrict total interest deduction to a resident taxpayer to the extent of 30% of EBIDTA (Earnings before interest, taxes, depreciation and amortization), if interest is paid to a foreign AE or in respect of a debt guaranteed by a foreign AE. Siemens Gamesa Renewable Power Pvt. Ltd52 has filed a writ petition, admitted by the Chennai High Court, challenging the constitutional validity of section 94B’s restrictions on deduction of interest for debt issued by a non-AE but guaranteed by a NR AE.

46.13 Withholding Tax from Payment to Non-R esidents Section 195 of the Act requires that tax must be withheld at the ‘rates in force’ from any payment to a NR if the sum paid is chargeable to tax in India. This includes instances of payment from one NR to another if such amount is taxable in India, such as in the case of an indirect transfer of shares. WHT rates depend upon the specific nature of the payment. In case of specific interest payments on bonds and foreign currency debt, or royalties, FTS, etc., concessional 51

52

Engineering Analysis Centre of Excellence (P.) Ltd, 125 taxmann.com 42 (SC). High Court of Madras, WP No. 19436, WMP No. 22874 of 2018.

862 Kuntal Dave WHT rates are prescribed. In many cases where a specific WHT rate is prescribed, the WHT is a final tax and NRs are exempted from having to file tax returns. An exhaustive procedure is laid down where taxpayers must obtain certification from a chartered accountant or apply to the tax authorities for determination of WHT. The Act also requires that the taxpayer obtain a valid government-issued TRC and certain additional information from the NR, in the absence of which beneficial DTAA provisions cannot be applied while determining the WHT. Further, if an NR neither provides a TRC nor has Indian tax registration, higher WHT rates are prescribed, irrespective of the nature of the payment.

Chapter 47

The Brazi l ia n Perspecti v e on Internationa l Tax L aw Fernando Souza de Man

47.1 Introduction Brazil, as one of the BRIC countries,1 is at a crossroads regarding its international standing on tax matters. That is because even though Brazil is not a developed country,2 it is hard to treat Brazil as a traditional developing state as it was, in 2019, the ninth biggest economy in the world.3 Consequently, Brazil is a prime candidate for an analysis on how the growth of countries might lead them to adopt tax positions which are more in line with the views expressed in the OECD Model Tax Convention. This study becomes more interesting when it is considered that historically Brazil has adopted tax views which do not conform to the guidance prescribed in the OECD Model Tax Convention, such as its intensive focus on source taxation.4 In that sense, Brazil has always been somewhat of a rogue nation in the international tax arena, as its positions—as shown throughout this chapter—deviate from the OECD consensus on these matters and go further than the UN compromise. Nonetheless, considering 1 Brazil,

Russia, India, and China as coined in a Goldman Sachs report from the beginning of the 2000s: Goldman Sachs, ‘Building Better Global Economic BRICs’, Global Economics Paper 66 (2001), https://www.goldmansachs.com/insights/archive/archive-pdfs/build-better-brics.pdf. 2 Despite the difficulty of properly defining what it means to be a developed or developing country, it is safe to say that Brazil, with a Human Development Index of 0.765 (84th out of 189 countries), is not a developed country. UN, Human Development Index Report 2020, http://hdr.undp.org/en/content/lat est-human-development-index-ranking. 3 World Bank, Gross Domestic Product 2019, https://databank.worldbank.org/data/download/ GDP.pdf. 4 For more on this issue, see Section 47.2.1.

864 Fernando Souza de Man the most recent international tax developments and the OECD discussions since the Base Erosion and Profit Shifting (BEPS) Project for enhanced taxation at the source of income,5 one might argue that Brazil was simply ahead of its time, in the sense that a historically rogue nation can nowadays be a trend-setter, a savant on source taxation of international income. To ascertain the Brazilian perspective on international tax law, we will first present the Brazilian domestic system, determining who is taxed and on what, the Brazilian position on the permanent establishment (PE) concept, the existence of anti-abuse rules in the Brazilian domestic tax system, taxation of income earned by companies resident in tax havens or that benefit from a fiscal-privileged regime, and the interaction between domestic tax rules and the double tax conventions signed by Brazil. From that, we will build on the Brazilian policy regarding tax treaties, especially considering its main deviations from the guidance of the OECD and UN Model Tax Conventions as regards taxation of royalties and technical services and taxation of independent personal services. Later, the chapter will focus on the influence of BEPS actions and the Multilateral Instrument (MLI) on the Brazilian domestic tax rules as well as the influence of the most recent UN Model Tax Convention. This study will allow us to assert the differences between the Brazilian tax treaties and the main model tax conventions, as well checking whether the Brazilian perspective of broad source-taxing rights is becoming mainstream.

47.2 Brazilian Approach to Taxation of International Income 47.2.1 Domestic Rules on Income Taxation When thinking about the Brazilian perspective on international tax law, it is of utmost importance to first study the domestic rules focused on taxation of income: (1) earned abroad by a resident company; and (2) earned in Brazil by a non-resident company. Only after such study can we proceed with an analysis of the double tax conventions signed by Brazil to address Brazilian tax treaty policy.

5

OECD, ‘Addressing the Tax Challenges of the Digital Economy, Action 1—2015 Final Report’ (2015), https://www.oecd.org/tax/addressing-the-tax-challenges-of-the-digital-economy-action-1-2015-final-rep ort-9789264241046-en.htm; OECD, ‘Addressing the Tax Challenges of the Digitalization of the Economy’ (2019), https://www.oecd.org/tax/beps/public-consultation-document-addressing-the-tax-challenges-of- the-digitalisation-of-the-economy.pdf; OECD, ‘Tax Challenges Arising from Digitalisation—Report on Pillar One Blueprint’ (2020), https://www.oecd.org/tax/beps/tax-challenges-arising-from-digitalisation- report-on-pillar-one-blueprint-beba0634-en.htm.

The Brazilian Perspective on International Tax Law 865

47.2.1.1. Taxation of residents For the taxation of Brazilian-resident companies, up until 1995 Brazil adopted a territorial system, therefore there was no taxation of profits earned abroad. From 1996 on, those companies were subject to tax on their worldwide income.6 On that issue, income earned abroad is part of the Brazilian tax base, but losses cannot reduce the Brazilian tax base7 unless there is a consolidation of accounts and taxation according to current controlled foreign corporation (CFC) rules.8 Regarding CFC rules, which should be a subset of anti-avoidance rules (special anti-avoidance rules, SAAR), since the introduction of the system of worldwide taxation, Brazilian tax authorities have been trying to expand the Brazilian tax base to also include income earned abroad by foreign companies controlled or associated with the Brazilian company,9 and this has been mainly done by reference to CFC legislation. Nevertheless, as repeatedly stressed by Brazilian authors over the years, the Brazilian CFC rules are over-inclusive as they apply regardless of whether there is an abusive situation, of the level of taxation at source, and whether the taxpayer has passive or active income.10 Consequently, only if the term SAAR is used loosely can the Brazilian CFC regime a SAAR. Apart from having such a broad CFC regime that is difficult to classify as a SAAR, there is a dispute over whether Brazil has even introduced a general anti-avoidance rule (GAAR). The crux of the discussion is article 116 of the National Tax Code, which states that tax authorities might disregard abusive situations if they respect procedures established in an ordinary law. However, Brazilian legislators have never enacted such an ordinary law, so despite reference made by the tax authorities to that provision, taxpayers correctly point that the absence of an ordinary law establishing procedures for disregarding such dealings impedes the application of article 116. On that matter, the tax authorities make recourse to the business-purpose doctrine to disregard legal transactions which they consider abusive, but the Brazilian courts do not support that position.11

6

Art. 27 of Law-Decree 5.844/43 is the legal basis for taxation of resident companies, while art. 25 of Law 9.249/25 states that income earned abroad is part of the Brazilian tax basis. Moreover, for Brazilian domestic law, partnerships are companies for matters of taxation, being subject to corporate income tax (art. 7 Law Decree 2.303/1986). In its position on art. 1 of the 2014 OECD Model Tax Convention, Brazil stated that it reserved its right to extend coverage of the convention to partnerships since partnerships are legal entities under its domestic legislation. This position is not part of the 2017 OECD Model Tax Convention, even though there was no change in the domestic legislation regarding the legal status of partnerships. 7 Brazil, Law 9.249/1995, art. 25, para. 5. 8 Brazil, Law 12.973/2014, art. 78. Such consolidation was possible until 2022. 9 Schoueri and Galdino, ‘Controlled Foreign Company Legislation in Brazil’, in Kofler et al., eds, Controlled Foreign Company Legislation, vol. 17 (Paris: IBFD, 2020), s. 7.1. 10 Schoueri and Galdino, ibid.; Ferreira, Brazil: Corporate Taxation, Country Tax Guides (IBFD), s. 10; Bianco and Tomazela Santos, ‘Brazilian Anti-Avoidance Legislation: Recent Refinements and Major Deviations from International Practice’, Bulletin for International Taxation 71/6 (2016), s. 3.2. 11 Bianco and Tomazela Santos, ibid.

866 Fernando Souza de Man Ultimately, despite the lack of real CFC rules and GAAR, the tax authorities are not toothless in the fight against abusive situations as Brazil has been fighting against base erosion, by means of limiting the amount of deductions allowed, since the 1950s,12 and it has also adopted other SAARs, such as transfer pricing13 and thin-capitalization rules.14 Specifically with regard to the Brazilian transfer pricing rules, it is interesting to note that they deviated from the OECD guidance on the matter and, because of that, they were a point of discussion between the OECD and Brazil, but recently the conversion of the Brazilian Transfer Pricing Rules to the OECD standard has been approved by the Brazilian Congress, effective from January 202415 Despite not being a member of the OECD or having adhered to the MLI, it is undeniable that the work conducted at the OECD level regarding BEPS and the MLI has the support of the Brazilian government. On that matter, the Brazilian government presented a draft law establishing that taxpayers in Brazil would have to disclose operations which lead to a reduction or deferral of taxes when those operations have no relevant non-tax reasons, adopt an unusual form, or have clauses that deviate from the effect of typical contracts or are listed in an act specified by the Brazilian tax authorities.16 That part of the bill was not transposed into law as Congress rejected the establishment of the new obligation for taxpayers. This was a positive outcome as, despite referring to BEPS Action 12, the proposed bill did not actually respect the criteria set by BEPS Action 12 for the establishment of mandatory disclosure rules, having rules that were too broad and not focused on actual abusive situations.17

47.2.1.2 Taxation of non-residents In relation to the taxation of non-resident companies, the domestic legislation states that Brazil will tax income earned by non-residents in Brazil at the rate of 15% of gross

12

Such measures were already part of Ordinance 436/1958 and Law 4506/1964. Law 9.430/1996. 14 Law 12.249/2010. 15 The OECD worked with the Brazilian tax authorities to reconcile Brazilian rules with the OECD standards, as can be seen in OECD, Receita Federal do Brasil, Transfer Pricing In Brazil: Towards Convergence with the OECD Standard (Paris: OECD, 2019), www.oecd.org/tax/transfer-pricing/transfer- pricing-in-brazil-towards-convergence-with-the-oecd-standard.htm. The assimilation suggested by the OECD has been criticized by Brazilian scholars in Taveira Torres et al., ‘Brazil and OECD: building a future of certainty and equality. Public Statement regarding the OECD-Brazil transfer pricing project’ (July 2019), http://kluwertaxblog.com/2019/07/30/brazilian-tp-missed-opportunities-ahead/, but the Brazilian Government enacted a Provisional Measure in 2022 (PM 1152/22) outlining the conversion of Brazilian transfer pricing rules to the OECD standard, and this measure has recently been approved by the Brazilian Congress (March 2023), to be effective from January 2024. 16 Provisional Measure 685/2015, art. 7. 17 Souza de Man and Martini, ‘BEPS and Developing States: How Developing States are using BEPS to Justify Aggressive Amendments to Domestic Legislation’, in Arts, Jansen, and Korving, eds, De internationalisering van het belastingrecht—The Internationalization of Tax Law: Bundel ter Gelegenheid van het 25-Jarig bestaan van de Opleiding Fiscaal Recht aan de Universiteit Maastricht (Maastricht: Shaker Publishing, 2016). 13

The Brazilian Perspective on International Tax Law 867 income.18 This taxation is in line with the prevalent idea that Brazil should also collect tax proceeds of income earned in Brazil and paid by a resident (i.e. in which Brazil was the source of production and payment (or only source of payment when dealing with income from services income)). In thinking about payments made by resident companies to companies abroad, this approach is based on the idea of the avoidance of base erosion, as the Brazilian paying company would deduct the payments. This possibility of source taxation is independent of the existence of a fixed place of business, as prescribed in the double tax conventions and in the legislation of states that also include the PE concept in their domestic laws. For Brazil, the fact that a non- resident has earned income in Brazil and a local resident has paid such amount, already provides a sufficient link for taxation (save in the case of income from services which is subject solely to the latter condition). In fact, despite including the term in all its double tax conventions, Brazil does not even adopt the PE concept in its domestic legislation.19 This disregard of the PE concept is also a consequence of the fact that as the domestic legislation imposes such a lengthy and burdensome procedure for carrying on a business via a PE, non-resident enterprises favour establishing a subsidiary instead.20 Moreover, based on domestic legislation, branches and agencies of non-resident companies receive the same tax treatment as resident companies,21 therefore non-residents do not have a strong incentive to go through the bureaucracy in order to carry on business via a subsidiary, a resident company for tax matters. Naturally, this does not mean that Brazil does not tax the income earned by non-residents if a double tax treaty has been signed between Brazil and the residence country of the non-resident income earner, but that this income is taxed based on another provision of the treaty, most likely article 12 as it will be seen later. As a sidenote, it is noteworthy that this preference for source taxation, normally adopted by developing states, has met strong resistance over the years, but the recent BEPS work on the taxation of the digital economy is heading towards a similar compromise, establishing the existence of income-earning activity at source as sufficient for the allocation of taxing rights to the state.22 Another interesting trait of the Brazilian taxation of non-residents is that while non- residents are normally subject to a 15% tax on gross income, this rate increases to 25% if the beneficiary is resident in a state that does not tax the income or taxes at a rate lower

18 Law-Decree 5.844/1943, art. 97(a). Royalties and fees for technical services are subject to the same tax rate, as made clear by Provisional Measure 2.159-70/2001, art. 3. 19 Souza de Man, ‘Taxation of Services in Treaties between Developed and Developing Countries: A Proposal for New Guidelines’, IBFD Doctoral Series, 39 (Amsterdam: IBFD, 2017), ch. 6, s. 6.1.1.1; Schoueri, ‘Brazil’, in Brauner and Pistone, eds, Brics and the Emergence of International Tax Coordination (2018), s. 4.3.1; Rocha, ‘Brazil’s Treaty Policy’, Bulletin for International Taxation 71/6 (2017); Souza Carvalho and Andrade Costa, ‘Brazil—Permanent Establishments and the Taxable Presence of Non- Residents in Brazil’, Bulletin for International Taxation 71/6 (2017). 20 Souza de Man, ibid., s. 6.1.1.1. 21 Law 4.131/1962, art. 42. 22 OECD, ‘Addressing the Tax Challenges of the Digital Economy, Action 1—2015 Final Report’.

868 Fernando Souza de Man than 20%.23 Hence, Brazil has already addressed in its domestic legislation dealings with residents of tax havens or of states that provide for a fiscal-privileged regime, and has thus increased tax collection at source. This is unmistakably an anti-abuse rule which has been part of the Brazilian tax law system for more than twenty years. Considering that the Brazilian domestic tax rules prescribe the taxation of residents and non-residents alike, based on broad source rules, the latter will only escape taxation of their earnings in Brazil if there is a double tax convention applicable which restricts the Brazilian source-taxing rights. Hence, it would be reasonable to assume that Brazil would have a vast tax treaty network, but up until now this has not been the case. In fact, Brazil does not have a double tax convention with the USA or Germany, two of the biggest investors in Brazil.24 In the following section we will delve into the Brazilian tax treaty network and its more distinctive trends, such as taxation of technical services as royalties and the maintenance of an article on independent personal services. But before doing so it is important to set out how tax treaties and domestic laws interact in the Brazilian legislation. The Brazilian Tax Code states that treaties revoke or modify domestic tax rules and will be observed by supervening legislation.25 Although the wording is imprecise (a tax treaty does not revoke domestic legislation but applies in place of the legislation whenever both documents deal with the same issue), this provision clearly establishes that Brazilian tax law does not allow for treaty overrides. Notwithstanding this, until recently there were still discussions in Brazil on whether supervening legislation could alter tax treaties. The confusion arose due to a decision of the Supreme Court that recognized, in a non-tax case, that such an override was possible. However, since that decision, the courts have recognized that, based on the Brazilian domestic tax system, no treaty override is possible. Such outcome, supported by court decisions and part of doctrine, is based on the idea that tax treaties are lex specialis in relation to domestic rules, as prescribed in the National Tax Code and in the Brazilian Constitution. The parties acknowledge that it is not a matter of hierarchical superiority of treaties when compared to domestic rules, but is a matter of specialty.26

23

Law 9.779/1999, Art. 8. has yet to sign a double tax convention with the USA, although it signed a double tax convention with Germany in 1976, but Germany terminated the agreement in 2005. 25 National Tax Code, art. 98. 26 Souza de Man, ‘Taxation of Services in Treaties between Developed and Developing Countries: A Proposal for New Guidelines’, s. 6.1.; Xavier, Direito Tributário Internacional do Brasil, 6th ed. (Rio de Janeiro: Forense, 2007), 117. Supporting the idea that tax treaties cannot be altered by domestic legislation, but not based on lex specialis, rather, on the idea that treaties provide for a limitation on domestic rules and these can only be applicable within the limited framework prescribed by the tax treaty, see Schoueri, Direito Tributário (São Paulo: Saraiva, 2011), 94; Schoueri and Barbosa, ‘Brazil’, in Lang et al., eds, Tax Rules in Non-Tax Agreements (IBFD: Amsterdam, 2012), 151. 24 Brazil

The Brazilian Perspective on International Tax Law 869

47.2.2 Brazilian Tax Treaty Network As of May 2023, Brazil had signed forty-three double tax conventions,27 of which thirty-six are in force and effective.28 The double tax treaties signed by Brazil so far have been divided into three distinct stages when considering the treaty partners and the changes undergone in the Brazilian domestic legislation.29 In the first stage, comprised of treaties signed in the 1960s and 1970s, Brazil focused on signing treaties with developed states. At the time, there were no Brazilian multinationals investing abroad, so Brazil did not stress the need to guarantee residence taxing rights for itself, focusing instead on having tax-sparing clauses so that tax benefits granted to foreign investors would not be nullified by the investor´s state of residence.30 Furthermore, Brazil negotiated broad source-taxing rights over income from technical services and technical assistance by including them under the royalty definition and guaranteed itself source-taxing rights also as regards ‘other income’. The second stage of Brazilian tax treaty policy encompasses treaties signed in the 1980s and 1990s. In these treaties, signed with other developing countries,31 Brazil did not push for the inclusion of tax-sparing clauses but, when they did exist, they were reciprocal.32 This approach confirms that Brazil did, indeed, view tax-sparing clauses as a necessary means of guaranteeing that investors from developed states would be able to enjoy the tax breaks available in Brazil, but that the Brazilian negotiators did not think such clause would be important in tax treaties with fellow developing states, in which the flow of income between states would be more balanced. In that sense, tax-sparing clauses were part of Brazilian tax treaty policy as weapons to enable Brazil to be more attractive to foreign investment. In fact, considering that Brazil had a territorial system of taxation until 1996, during the first thirty years during which Brazil signed double tax treaties, it did so with the objective of attracting foreign direct investment.33 27 Brazil has signed double tax conventions with: Japan (1967), Portugal (1971), France (1971), Finland (1972), Belgium (1972), Denmark (1974), Spain (1974), Sweden (1975),Austria (1975), Germany (1975), Italy (1978), Luxembourg (1978), Argentina (1980), Norway (1980), Ecuador (1983), Philippines (1983), Canada (1984), Hungary (1986), Slovakia (1986), Czech Republic (1986), India (1988), South Korea (1989), Netherlands (1990), China (1991), Finland (1996), Portugal (2000), Paraguay (2000), Chile (2001), Ukraine (2002), Israel (2002), Mexico (2003), South Africa (2003), Russia (2004), Venezuela (2005), Peru (2006), Trinidad & Tobago (2008), Turkey (2010), Switzerland (2018), Singapore (2018), United Arab Emirates (2018), Uruguay (2019), Norway (2022) and United Kingdom (2022) 28 The treaties with Portugal (1971), Finland (1972), and Germany (1975) have been terminated (only the treaty with Germany was not renegotiated), the treaties signed with Norway (2022) will replace the treaty signed in 1980, and it is unknown at the time when the treaties with Paraguay (2000), Uruguay (2019) and United Kingdom (2022) will become effective. 29 Schoueri, ‘Brazil’, s. 4.2.1. 30 Ibid. 31 The 1980s mark the point at which the Brazilian tax treaty policy shifted from signing treaties mainly with developed states to signing treaties with developing non-OECD member states. Ibid., s. 6.1. 32 Ibid. 33 Ibid.; Rocha, ‘Brazil’s Treaty Policy’. It could be argued that signing tax treaties is beneficial for attracting foreign direct investment, and such argument would explain why Brazil, a territorial country, signed tax treaties between 1967 and 1995. Nonetheless, the effect tax treaties have on attracting foreign

870 Fernando Souza de Man The third stage of the Brazilian tax treaty policy includes treaties signed from 2000 until 2010. Once again, tax-sparing clauses are not present, which one might argue was a sign of a shifting of Brazilian tax treaty policy as Brazil also became a capital exporter, or a simply a recognition that in those treaties investment would occur both ways in a balanced manner, therefore a tax-sparing clause was not necessary34. This was also the first stage in which Brazilian tax treaties included limitation-on-benefits (LOB) clauses.35 After a hiatus of eight years without signing a tax treaty, Brazil entered, in the author’s opinion, a new stage of its international tax policy. This stage, which is the fourth stage, started with the signing of a double tax treaty with Switzerland.36 In this new stage, which started after Brazil had filed a request to become an OECD member state,37 the influence of the OECD BEPS Actions is clear: (1) the adoption of article 4(3) of the 2017 OECD Model Tax Convention; (2) the inclusion in all double tax conventions of a provision dealing specifically with entitlement to benefits (all measures recommended in BEPS Action 6); (3) the inclusion of article 1(2) to deal with transparent entities (as prescribed in BEPS Action 2); or (4) the new wording of article 5 (which stems from BEPS Action 7). Up until that time, the double tax conventions signed did not include tax-sparing provisions either. Considering that recently this has been the trend in Brazilian double tax conventions, one might wonder whether the adoption of tax-sparing provisions is still a trademark of Brazilian tax treaty policy. Approximation to the OECD guidance does not mean that Brazil is abandoning its pro-source-taxation stance. All these conventions prescribe source taxation of royalty income and adopt a provision on the taxation of technical services in line with article 12A of the 2017 UN Model Tax Convention,38 apart from maintaining a provision on the taxation of independent personal services. Of all the stages of Brazilian tax treaty policy, this is the one in which the influence of the OECD and UN Model Tax Conventions is clearer. Despite the amendments modelled after the OECD BEPS work, which consider the varied possibilities for source taxation of income adopted in the Brazilian double tax treaties, the Brazilian double tax treaties are still more influenced by the UN Model Tax Convention than by the OECD Model.39 Nonetheless, one can no longer affirm, as was direct investment is controversial, with certain authors stating that they do, indeed, attract foreign direct investment while others state that there is no empirical evidence of such link. On this issue, see Neumayer, ‘Do Double Taxation Treaties Increase Foreign Direct Investment to Developing Countries?’, Journal of Development Studies 43/8 (2007); Chisik and Davies, ‘Gradualism in Tax Treaties with Irreversible Foreign Direct Investment’, International Economic Review 45/1 (2004); Davies, ‘Tax Treaties, Renegotiations, and Foreign Direct Investment’, Economic Analysis & Policy 33/2 (2003); Barthel et al., ‘The Relation between Double Taxation Treaties and Foreign Direct Investment’, in Lang et al., eds, Tax Treaties: Building Bridges between Law and Economics (Amsterdam: IBFD, 2010). 34

Schoueri, ‘Brazil’ e.g. DTC Brazil–Israel (2002). 36 DTC Brazil–Switzerland (2018) 37 Brazil officially filed a request to become an OECD member in May 2017. 38 In their protocols, these treaties include payments for technical assistance under the scope of the royalty article. 39 Rocha, ‘Brazil’s Treaty Policy’, s. 4. 35

The Brazilian Perspective on International Tax Law 871 possible before this new batch of double tax conventions, that Brazil is not converging, at least to a certain extent, towards the international tax regime40 and that the OECD has had a limited influence on Brazilian tax treaty policy.41 Thus, albeit that Brazil has never changed its stance on the need for the allocation of more tax treaty rights for source states, and has successfully negotiated to maintain those rights, which exist in its domestic tax legislation, in its tax treaties at every stage of the Brazilian tax treaty policy has seen Brazil increasing it use of provisions focused on guaranteeing the rights of residence states to tax international income. Concomitantly, Brazil is also adopting new provisions drafted by the UN to expand the taxing rights of source states when compared with provisions existing in the OECD and UN Models. In this sense, it can be said that with time Brazil is slowly moving from a position of reflecting its domestic legislation in its tax treaties to one of adopting the provisions negotiated at the international level (i.e. Brazil is adopting fewer provisions created with broader source-taxing rights than prescribed in the model conventions and following more the guidance of the OECD and UN Model Tax Conventions).

47.3 Considerable Deviations of Brazilian Double Tax Conventions from the OECD and UN Models As demonstrated in the previous section, in its domestic tax rules Brazil prescribes source taxation of income earned by non-residents irrespective of the existence of a PE, taxing the non-resident whenever they have their income sourced in Brazil. As a rule, taxation will only occur if the source of production (the place where income is earned) and the source of payment (the place where the payer is located) are in Brazil; however, when dealing with services income, Brazil requires only that the source of payment is in Brazil. Regardless of such broad domestic source-taxing rules, if the double tax treaties signed by Brazil adopt the OECD or even the UN position on the matter (the difference being the existence of a service PE in art. 5(3)(b) of the UN Model Tax Convention), Brazil will not be able to exercise the taxing rights prescribed in its domestic law unless the income is attributed to a PE in Brazil (a concept, as seen in Section 47.2.1, which is alien to Brazilian domestic tax law). That is why it is crucial for Brazil to negotiate tax treaties that reflect its domestic position on the taxation of non-residents. Up until now, Brazil has been quite successful in mirroring its domestic rules in the double tax conventions it has signed. The taxation of technical services and independent personal services are primary examples of how Brazil negotiates treaties which allow for

40 Ibid. 41

Schoueri, ‘Brazil’, s. 4.2.2.1.

872 Fernando Souza de Man the exercise of broad source-taxing rights prescribed in its domestic legislation and maintains more taxing rights than those allocated to source states by both the OECD and UN Model Tax Conventions.

47.3.1 Taxation of Fees for Technical Services The term ‘fees for technical services’ is not including in the wording of any provision of the OECD Model Tax Convention, being under the scope of article 7, as is made clear in the Commentaries on the OECD Model Tax Convention.42 The same applied to the UN Model Tax Convention prior to the introduction of article 12A. Thus, irrespective of the model adopted, if Brazil had followed the guidance of the model tax conventions prior to the 2017 UN Model, it would only have been able to tax income from technical services earned by non-residents in the case of attribution of that income to a non-resident PE in Brazil. That approach would have been a restriction on the taxing rights prescribed by Brazilian domestic tax law, therefore Brazil made it clear in its position on article 12 and its commentaries that it reserved the right to include technical services and assistance under the scope of the article on royalties.43 Since its first double tax convention, signed in the 1960s, Brazil has achieved considerable success in this quest, having prior to 2017 only five double tax conventions in which technical services are not treated in the same manner as royalties.44 Article 12 of Brazilian double tax treaties is based on article 12 of the 1980 UN Model Tax Convention,45 establishing hared competence over the taxation of income from royalties, but it goes further than that model by including fees for technical services and assistance under the scope of article 1 instead of the scope of article 7. Thus, by stating that fees for technical services are under the scope of article 12, Brazil is saying that it can tax such income regardless of the existence of a PE, it being sufficient that a Brazilian resident makes the payment. Despite the inclusion of technical services within the scope of article 12, there is still discussion on whether all technical services come under article 12 or only technical services ancillary to royalty payments, with the Brazilian tax authorities stating since 2014 that all technical services, even those not linked to a transfer of technology, come under article 12, and thus providing for a broad definition of technical services.46 There are Brazilian scholars who point out that this approach is incorrect as activities which

42

OECD Commentary on art. 12. OECD, Non-OECD Economies’ Positions on the OECD Model Tax Convention, Positions on art. 12 (royalties) and its Commentaries, para. 7. 44 DTCs with Austria, Finland, France, Japan, and Sweden do not include such provision in their protocols. 45 Souza de Man, ‘Taxation of Services in Treaties between Developed and Developing Countries: A Proposal for New Guidelines’, s. 6.1.1.2.1. 46 Normative Declaratory Act 5/2014. 43

The Brazilian Perspective on International Tax Law 873 do not lead to imparting knowledge should not be included in the scope of article 12.47 The author agrees with these scholars on a de lege ferenda basis but, considering that in the double tax conventions signed by Brazil no distinction is made between technical services that involve the transfer of technology and ones that do not, the author considers that the wording of the treaty should prevail, and the provision should be applicable to all technical services.48 Consequently, article 12 of the Brazilian double tax conventions grants Brazil more taxing rights as the source state than would the OECD and UN Model Tax Conventions.

47.3.2 Taxation of Independent Personal Services Similar to what occurs with the taxation of technical services, Brazilian double tax conventions deviate from the OECD Model with regard to the taxation of independent personal services. While the OECD Model prescribed exclusive residence taxation unless income is attributed to the fixed base,49 Brazil, apart from adopting that threshold, also prescribes in its treaties that income will be taxed in Brazil if payments are made by a Brazilian resident, or a PE located in Brazil.50 The reference to the PE concept does not diminish the possibilities for source taxation, as the provision is worded in a manner in which it is not required that the activities are carried on through the PE, but rather that the payments are borne by the PE (i.e. the focus is on the source of payment, as in the domestic tax rules).51 This broad source rule has not been adopted in more recent treaties, which follow the UN Model Tax Convention by prescribing that taxation at source income is attributed to a fixed base or if the person remains in the source state for more than 183 days.52 Also, diverging from the OECD guidance which deleted article 14 from its Model Convention, and following the UN perspective on the matter, Brazil has decided to maintain this provision in the tax treaties signed since 2000. Despite of the prominence of this provision in Brazilian tax treaty policy, it remains to be seen whether it is nowadays superfluous, considering the addition in most recent treaties of article 12A, which might overlap with article 14 and is applied regardless of article 14 (i.e. independent personal services which are also technical services will be taxed at source irrespective of the existence of a fixed base or the physical presence of

47 Xavier, Direito Tributário Internacional do Brasil; Schoueri, ‘Brazil’, s. 4.3.2.1. 48

Souza de Man, ‘Taxation of Services in Treaties between Developed and Developing Countries: A Proposal for New Guidelines’, s. 6.1.1.2.1. 49 OECD Model Tax Convention, art. 14. 50 e.g. DTCs with France (1972), Austria (1975), Norway (1980), India (1988), Finland (1996), and Israel (2002) 51 de Man, ‘Tributação de Serviços na Fonte e as Convenções Modelo: Renascimento dos Ideais do Artigo 14 OCDE’, in de Oliveira et al., eds, Direito Tributário Atual, vol. 26 (São Paulo: Dialética, 2012), 49–50. 52 e.g. DTC with Switzerland (2018), Singapore (2018), United Arab Emirates (2018), and Uruguay (2019).

874 Fernando Souza de Man the service provider in the source state for more than 183 days).53 Note, however, that despite prescribing limited tax rates at source, taxes based on article 12A are on the gross amount earned, while article 14 allows for taxation on the net amount, so the adoption of article 12A can, indeed, make article 14 redundant and lead to over-taxation of the non-resident income earner and an increase in the compliance burden.54 It should be borne in mind that the basis of the OECD work on Pillar One is on taxation in the source state irrespective of the existence of a fixed place of business in the source state, so that it might occur that by still resorting to a physical presence threshold, article 14 will be more restrictive on source taxation than article 12A and the OECD guidance on Pillar One. Taking the recent UN and OECD work on detaching taxation at source from the criterion of physical presence, it is doubtful whether article 14 will continue to have a prominent place in the tax treaty policy of developing states, including Brazil, who as mentioned earlier is adopting article 12A of the UN Model Tax Convention.

47.4 Influence of BEPS Work and MLI on Brazilian Tax Treaty Policy Brazil, apart from being a member of the G20, is also a member of the BEPS Inclusive Framework. Even though it did not sign the Multilateral Instrument (MLI), Brazil is committed to the implementation of the recommendations issued in the BEPS Actions. The BEPS outcomes resonated within the Brazilian government which, as seen earlier, intended to establish a mandatory disclosure regime. Despite the regime never coming to fruition, and going further than the OECD proposal,55 it is undeniable that the attempt by the Brazilian government shows that it supports at least part of the BEPS outcomes. This assumption is corroborated by an attempt to limit the deductibility of interest on net equity, also not approved by Congress, the introduction of guidance on the rulings system, and the exchange of information with other tax authorities,56 which is in line with BEPS Action 5 and with signing the Convention on Mutual Administrative Assistance on Tax Matters, the Multilateral Competent Authority Agreement on the Exchange of Country-by-Country Reports, and the Multilateral Competent Authority Agreement on the Automatic Exchange of Financial Account Information, as well as establishing a country-by-country reporting system (measures in line with BEPS 53 Souza

de Man, ‘The Past, Present and Future of the Taxation of Independent Personal Services in Double Tax Conventions: Is There Still a Place for Article 14 on Model Tax Conventions?’, in Báez Moreno and Castro Arango, eds, Problemas de Tributación Internacional en Iberoamérica: una vision desde los diez años del OITI (Bogotá: Universidad Externado de Colombia, 2019), s. 47.3.1. 54 Ibid., s. 47.3.1. 55 See Section 47.1.2. 56 Normative Rulings 1.658/2016 and 1.689/2017.

The Brazilian Perspective on International Tax Law 875 Action 13) and drafting an instruction regulating the dispute-resolution mechanism on Brazilian double tax conventions (BEPS art. 14). The inclusion of BEPS Actions outcomes in its domestic rules and most recent double tax conventions, mean that it can be inferred that despite not signing the MLI, Brazil does indeed see its drafting, in which Brazil participated, as a positive outcome. The issue is then that Brazil prefers to deal with the matters contained in the MLI on a case- by-case basis, with renegotiation of specific double tax conventions,57 as recently carried out in its renegotiation of the double tax convention signed with Argentina. As a result, that convention now has a new title and preamble in line with the BEPS outcomes, as well as a provision on a mutual agreement procedure, exchange of information, and limitation of benefits58 (i.e. Brazil has adopted the Recommendations of BEPS Action 6 in the treaty). In that sense, as a member of the BEPS Inclusive Framework, Brazil has duly incorporated the BEPS minimum standards in its domestic rules and some of its tax treaties (the one signed with Argentina and the more recent double tax conventions signed since 2018).

47.5 Influence of the 2017 UN Model on Brazilian Tax Treaty Policy The most significant innovation of the 2017 UN Model Tax Convention has been the introduction of article 12A to deal specifically with taxation of technical services. This article, which prescribes taxing rules similar to those in article 12 (royalty income), has been part of all double tax conventions signed by Brazil since 2018, as mentioned earlier.59 Thus, at first it seems that by adding such provision to its double taxation conventions, Brazil is increasing the potential for source taxation of income from technical services as, following the UN Model Tax Convention, such income would be under the scope of article 7 and would only be taxed at source if attributable to a PE in the source state. Nonetheless, as highlighted previously,60 in its tax treaties Brazil has successfully negotiated the inclusion of technical services under the scope of article 12, adopting a version of article 12 in its tax treaties that prescribes broader source-taxing rights than the UN Model Tax Convention, so the introduction of article 12A should not

57 According to Schoueri and Galendi Júnior, this is the outcome that most respects the independent Brazilian tax treaty policy, as the adoption of the MLI is likely to affect the balance of the allocation of taxing rights achieved by the treaty-negotiating parties and lead to an outcome which would not be accepted in bilateral negotiations: Schoueri and Galendi Júnior, ‘Interpretative and Policy Challenges Following the OECD Multilateral Instrument (2016) from a Brazilian Perspective’, Bulletin for International Taxation 71/6 (2017), se. 3.1. 58 Decree 9.482/2019. 59 See Section 47.2.2. 60 See Section 47.3.1.

876 Fernando Souza de Man lead to an increase of taxing rights for the source state. The benefit of having article 12A instead of continuing to list technical services under the royalty article is that now there is no discussion of whether all technical services should be taxed at source or only those linked to royalty payments, a debate which has been ongoing in Brazil for the past ten years, as mentioned previously.61 Furthermore, the adoption of article 12A shows Brazil´s alignment to the criteria of source taxation championed by the UN Model and balances the influence the recent OECD Model has had on the latest Brazilian double tax conventions.

47.6 Conclusion The study of the Brazilian domestic rules on international taxation and the double tax treaties signed by Brazil shows that the country has long favoured broad source-taxing rights in its domestic rules and that, in general, it was able to replicate those in its double tax conventions. Thus, despite engaging with double tax conventions, in its limited tax treaty network Brazil has taken advantage of every situation to increase its source-taxing rights, going further than the OECD and UN Model Tax Conventions. More recently, despite its strong focus on source taxation, Brazil has adopted the provisions of the OECD in a more consistent manner in its double tax conventions. One could argue that this approximation to the model tax conventions was already expected as Brazil is converging towards the international tax regime, based on the OECD Model Tax Convention.62 But as mentioned in the chapter, Brazil is also converging with the provisions of the UN Model that prescribe broader source taxation, such as article 12A. Thus, one might think that Brazil is leaving behind the behaviour of a rogue agent in favour of a state conforming to international standards of the international tax rules. While there is some truth to that statement, as Brazil is indeed more active in the international tax arena and is embracing the tax treaty provisions recently negotiated at the OECD and UN level, it cannot be overlooked that such rules are in line with Brazilian domestic tax rules; for example, focusing on anti-avoidance rules in double tax treaties and avoiding double non-taxation, bodes well, for instance, for the Brazilian CFC rules, which, as explained earlier, are too broad to be viewed as a GAAR and should be seen more as anti-deferral rules. In the same sense, the adoption of article 12A of the UN Model Tax Convention and a willingness to participate in the discussions on Pillar One are in accordance with the domestic tax rules of favouring taxation at source irrespective of the existence of a physical presence therein. Therefore, it seems that the issue is not that Brazil is, in general, conforming to the international rules, but rather that the international rules being proposed are 61

Section 47.2.2. ‘The International Tax Regime and the BRIC World: Elements for a Theory’, Oxford Journal of Legal Studies 33/4 (2013). 62 Baistrocchi,

The Brazilian Perspective on International Tax Law 877 in accordance with established Brazilian tax treaty policy and Brazilian domestic tax law, which contains rules that favour taxation at the place where income is earned and not where formal requirements, such as the existence of a fixed place of business, are fulfilled. Ultimately, considering the latest developments at the OECD and UN level, it is valid to consider whether Brazil was a trend-setter and not a rogue agent, favouring taxation based on real economic criteria over formal considerations such as the existence of a fixed place of business.

Chapter 48

T he Germ an Perspe c t i v e on In ternationa l Tax L aw Gerhard Kraft

48.1 Taxation in Germany—A General Overview 48.1.1 Preliminaries The current German international tax system encompassing all rules dealing with cross-border taxation is considered extremely complex. There is little doubt that this complexity is partly due to the incremental development of the complexity of economic systems all over the world now in place to deal with cross-border trade activity. In addition, Germany presides over an unprecedented system of anti-abuse rules with a particular focus on international taxation.

48.1.2 Notion of Tax For international as well as for domestic tax purposes, the notion of tax under German definitory standards refers to a mandatory financial charge or some other type of levy typically imposed upon an individual or other legal entity, the so-called taxpayer, by a governmental organization in order to fund various public expenditures. So, as in most jurisdictions, a tax under the German legal perception is not directly linked to a specific governmental service. Taxes are said to be levied in order to finance public goods; this is why taxes need to be compulsory, otherwise one would face a severe free-rider problem. Economic and legal definitions of taxes sometimes marginally differ depending on the perspective. As in most countries, failure to pay taxes in Germany, along with evasion of or resistance to taxation, bears severe criminal law repercussions and is punishable.

880 Gerhard Kraft It has been argued that German prosecutors are particularly eager to fight international tax evasion, and there appears to be strong empirical evidence for that line of reasoning.

48.1.3 History of German International Taxation Knowledge of international tax history is an important foundation for understanding current international tax practice and examining the possibilities for future international tax reform. As holds true for any history, it is difficult to decide what to include and what to leave out in terms of the history of international taxation in Germany. It is similarly difficult to decide where to start, and where to stop. Like most international tax systems of comparable countries, the German international tax system has largely developed on an incremental basis. Clearly, one short chapter cannot expect to fully address all relevant historical international tax issues. Nonetheless, it appears to be important to provide the reader with the foundations necessary for understanding other aspects of German international taxation. Probably a reasonable starting point for looking at German international tax history might be the decline of the Holy Roman Empire (Latin: Sacrum Imperium Romanum; German: Heiliges Römisches Reich). The Holy Roman Empire as a multi-ethnic complex of territories in Western and Central Europe developed during the early Middle Ages. There is little doubt that international tax and customs matters were of relevance even before the final dissolution of the Holy Roman Empire which continued to exist until 1806 and formerly came to an end as a consequence of the Napoleonic Wars. Those wars, following the course of the French Revolution, involved drastic change with regard to the political order of the old Reich. It is worth noting that the most important and final break in terms of international, constitutional and canon law was the Reichsdeputationshauptschluss (formally the Hauptschluß der außerordentlichen Reichsdeputation—Principal Conclusion of the Extraordinary Imperial Delegation) passed on 25 February 1803 by the Reichstag (‘Imperial Diet’). Of purely anecdotical interest, the tithe, the most important medieval tax, was abolished by the French Revolution in 1789 whereas it took until 1848 for the German territories to rescind it as a result of the (failed) 1848 revolution. The former Holy Roman Empire was replaced by the German Confederation (Deutscher Bund). The German Confederation was an association of thirty-nine predominantly German-speaking sovereign states in Central Europe, created by the Congress of Vienna in 1815. It was dominated by the State of Prussia with a heavily agrarian society where land ownership played a central role. Germany’s nobles, especially those in the east called Junkers, dominated public life and paid no land tax until 1861. The establishment of the Prussian-led Zollverein (the customs union) in 1834 formed a cornerstone of German international tax history and needs to be viewed in the light of 1869, a year in which Prussia concluded a treaty with Saxony ‘for the Elimination of Double Taxation of State Nationals of both Sides’. The 1869 tax treaty concluded by

The German Perspective on International Tax Law 881 Prussia and Saxony demonstrates a remarkable resemblance to modern bilateral tax treaties. In 1902, the ‘bubbly tax’ (Schaumweinsteuer) was introduced in order to finance the imperial navy. After the First World War, the tax and fiscal reforms headed by German Finance Minister Matthias Erzberger in 1919 and 1920 fundamentally reshaped German public finances. The Erzberger reform was the first, and probably the only real, tax reform that enabled a German government to almost fully reorganize, modernize, and centralize taxes and public finances. Relevant features, especially with regard to international taxation, still today form the backbone of the current international tax regime. After the First World War, some remarkable steps towards an international tax treaty network can be identified. The first double tax treaty concluded by the German Reich was entered into in 1925 with the Italian kingdom. An important development with regard to international tax issues was the introduction of the Reich Flight Tax (Reichsfluchtsteuer) in 1931. The Reich Flight Tax was a capital control law implemented in order to stem capital flight from the Weimar Republic. The legal basis for such tax was created by a decree on 8 December 1931 by Reichspräsident Paul von Hindenburg. The tax was assessed on departure from an individual’s German domicile, provided that an individual disposed of assets exceeding 200,000 Reichsmark, whereby an alternative yearly income of 20,000 Reichsmark applied. The tax rate amounted to 25%. In 1931, the Reichsmark was fixed at an exchange rate of 4.2 Reichsmark per US$; 200,000 Reichsmark was equivalent to US$47,600 (equivalent to US$800,000 in 2019). In 1934, during the Nazi regime in the so-called Third Reich, the use of the Reich Flight Tax shifted away from dissuading wealthy citizens from moving overseas and was instead used as a form of ‘legalized theft’ to confiscate Jewish assets. This was accomplished by substantially changing the existing decree with the ‘Law Concerning Revision of the Specifications of the Reich Flight Tax’ (Gesetz über Änderung der Vorschriften über die Reichsfluchtsteuer), issued on 18 May 1934.1 The Reich Flight Tax in Nazi Germany was extended six times during the Third Reich and was amended on 9 December 19422 to remain in force indefinitely. After the collapse of the Third Reich, the international tax systems in the former ‘German Democratic Republic’ and West Germany developed dramatically differently. Whereas the German Democratic Republic subscribed to a socialist legal and economic order with all the consequences extending to international tax matters, the Federal Republic of Germany became deeply integrated in the Western-type democracies and international relations. This was not without impact on the international tax situation in the Federal Republic of Germany. Thus, even in 1972, in Germany the so-called Foreign Tax Act was enacted containing rules on transfer pricing, exit taxation, controlled foreign corporations, and abusive family-foundation schemes. The main part of this tax, the rules concerning controlled foreign corporation (CFC) taxation or, in the US analogue

1

2

RGBl. 1934 I, 392–393. RGBl. I, 682.

882 Gerhard Kraft ‘Subpart-F-legislation’, were patterned after the US prototype. After German unification in 1990, the Western-oriented international tax system was formerly adopted by the newly created Germany. Today, Germany has agreed on close to 100 double taxation treaties. Literally all features of modern international tax system—including rules concerning transfer pricing, exit taxation, CFCs, treaty and directive shopping— are implemented in Germany.

48.1.4 Constitutional Foundations of International Taxation in Germany 48.1.4.1 Constitutional sources of international tax law Germany, as with most constitutional states, has numerous sources of tax law which can be divided between legislative, administrative, and judicial sources. First and foremost, the German Constitution (Grundgesetz (GG) or Basic Law) should be highlighted. The German Constitution forms the legal basis for taxation insofar as that body of law establishes the framework in which the basic principles governing tax law are laid out. There are no specific rules regarding international taxation in the Constitution, nevertheless it is of the utmost importance for the design of the domestic and the international tax system. In concrete terms, the German Constitution lays down the principles governing domestic and the international taxation, inter alia but not limited to, such as the ability-to-pay principle, equality in taxation, the lawfulness of taxation, and the welfare state principle. Empirically, the two most prominent and focal features of fundamental constitutional rights impacting on international taxation in Germany are the role of equality under the law and the rule- of- law concept. Most sub- principles of relevance in German international tax doctrine go back to the clause in the German Constitution guaranteeing equality under the law. Among those are the consistency principle (Folgerichtigkeitsgebot) and the ability-to-pay-principle (Leistungsfähigkeitsprinzip). Apart from this, the rule of law (Rechtsstaatsprinzip) plays the most prominent role with respect to international tax matters.

48.1.4.2 Legislative sources of international tax law As far as legal hierarchy is concerned, underneath the Grundgesetz, as in most democratic countries, Germany experiences a group of statutes that are published according to legislative procedural formalities and are often referred to—in the English translation—as the tax codes or tax laws. Accordingly, the German tax system spreads over several different tax Acts. Whereas most of them are not particularly designed to specifically govern international tax matters, some (e.g. the Foreign Tax Act) are particularly designed to explicitly deal with international taxation issues. Statutes in Germany are the primary source of international tax law; they are the supreme law of the land outside the Constitution and tax treaties. Because of the importance of statutory law,

The German Perspective on International Tax Law 883 the legislative history underlying the enactment of these statues is also an important source of federal tax law. In this respect, German is no different from other democratic jurisdictions where legislative history usually consists of judicial committee reports or even documented debates in the parliamentary bodies. Tax treaties, or double tax arrangements (DTAs), are another important legislative source of law. As mentioned earlier, German double tax treaties number close to 100 and are obviously relevant when taxpayers have ties with two different countries, usually either another country of residence or another source country in which German resident individuals, corporations, or other entities generate income.

48.1.4.3 Administrative sources of international tax law There are numerous administrative sources of international tax law in Germany. Terminologically, such administrative sources of tax law in their English translation are usually referred to as ‘regulations’ (Richtlinien), ‘guidelines’ (Erlasse), ‘revenue ruling’ (BMF-Schreiben), or in short used under the umbrella term of ‘rulings’. Similar notions are not uncommon and are the most frequently encountered. In Germany, such administrative sources are in most cases not binding on the courts but are binding on the administrative bodies. That is, even if a public (tax) agent considers the wording and/or spirit of a tax regulation as unlawful, they are obliged to apply it. Essentially, regulations and similar instruments in Germany with regard to international taxation may be viewed as the officially promulgated view of the executive arm or the government. As neither a taxpayer nor a court are obliged to follow a regulation, such ordinances can be overruled by the courts when a taxpayer successfully challenges the before the courts. Regulations and other administrative sources of tax law often provide very detailed examples with immense practical impact.

48.1.4.4 Judicial sources of international tax law Judicial opinions form a third source of international tax law in Germany. Local courts that hear international tax cases, publish their judicial opinions that explain, supplement, and, in some cases, create new law. The authority of these opinions varies based on which court wrote the opinion and when it was written. Sometimes the courts themselves impose restrictions on the precedential authority of their judicial opinions. As a general rule, the courts responsible for international tax matters not only construe the statutes enacted by the legislative bodies, but they may also decide the law and sometimes write legal history with respect to international tax principles.

48.1.5 Germany as an EU Member Country 48.1.5.1 Primary EU law Despite Germany’s far-reaching tax autonomy as an EU member state, German tax law, as with the tax law of all other EU member states, has to conform to European law. Whereas it is true that no express provision concerning the interrelationship between

884 Gerhard Kraft EU law and the national laws of the member states exists, it is worth noting in this context that the primacy of EU law established by the European Court of Justice plays a considerable role as far as the perception of international tax law in Germany is concerned. Respecting the spirit and wording of the seminal court case usually referred to as Costa/ ENEL3 appears to be a well-accepted and consistently applied principle of international law in Germany that EU law takes precedence over national law. Thus, in terms of international taxation in Germany, the primacy of EU law applies both in relation to the founding Treaties and the EU directives. As a logical consequence of the aforementioned case law and as a corollary of Germany’s EU membership, Germany needs to design international tax law within the confines set by EU primary law. As European law only has a mandate for the harmonization of indirect taxes, especially value-added tax,4 Germany, like all other EU member states, today presides over a direct international tax system that has developed along the lines predetermined by primary and secondary European law. Clearly, the Treaty on the Functioning of the European Union (TFEU) defines the internal market as an area without internal borders in which the free movements of goods, persons, services, and capital is ensured.5 It is well accepted in the German tax community that, in line with European Court of Justice (ECJ) case law, primary European law is supranational and ranks above national law. Its practical transformation by the German tax authorities, however, lags behind the application of this noble principle. Thus, the relevant provisions of the fundamental economic freedoms, such as the free movement of goods,6 free movement of persons,7 freedom of establishment,8 freedom to provide services,9 and free movement of capital and payments,10 are often violated on account of their day-to-day application by the German legislator and the administrative arm. It appears that the German tax courts particularly enjoy the infamy of a profound eagerness to submit controversial cases to the ECJ. For many years, Germany enjoyed a proverbial ‘model student reputation’ as far as the transformation of European secondary law into German domestic tax law was concerned. That reputation dramatically deteriorated with the transposition of the Anti-Tax Avoidance Directive (ATAD)11 into German law. Council Directive (EU) 2016/ 1164 of 12 July 2016 laid down rules against tax avoidance practices that directly affect the functioning of the internal market and required member states to apply those measures

3 Case 6/64 Flaminio Costa v ENEL, reference for a preliminary ruling: Giudice conciliatore di Milano, Italy. 4 Art. 113 TFEU. 5 Art. 26(2) TFEU. 6 Arts 28–32 TFEU. 7 Arts 45–48 TFEU. 8 Art 49–55 TFEU. 9 Arts 56–62 TFEU. 10 Arts 63–66 TFEU. 11 Council Directive (EU) 2016/1164 of 12 July 2016 (ATAD I), followed by Council Directive (EU) 2017/952 of 29 May 2017 (ATAD II).

The German Perspective on International Tax Law 885 as from 1 January 2019. Only in 2021 did Germany initiate the transposition process which gave rise to questions regarding Germany’s ability and preparedness to comply with EU requirements. Finally, state aid issues have played an important role in German domestic and international tax thinking over recent years. However, after a series of clarifying ECJ rulings, German enthusiasm to suspect state aid problems behind, quite literally, every tax matter has significantly waned.

48.1.5.2 Secondary EU law Germany used to be a reliable proponent for the timely implementation of EU tax directives; including international business tax frontrunners such as the Parent– Subsidiary Directive, the Merger Directive, and the Interest and Royalties Directive. Most subsequent EU tax directives were implemented into German tax law in due course. Therefore, it should come as no surprise that the seminal Francovich12 judgment on state liability did not play a crucial role in German international tax law. Although perhaps that observation will have to be viewed from a different perspective once the deficiencies of the implementation the ATAD finally reach the courts. As mentioned earlier, Germany’s reputation as a model student was frivolously discarded by its non-compliance with the deadline of 31 December 2018 for transposition of the ATAD. Only on 24 March 2021, did the German government take notice of the draft of the German ‘Law implementing the EU Anti-Tax Avoidance-Directive’ (‘Draft Law’; ATAD I and II). On that date, without any discussion, the German government adopted the draft bill for an Act on the implementation of the ATAD. The draft law is intended to implement the ATAD provisions on exit taxation, on CFC rules, and on hybrid mismatches as set out in the ATAD. The draft bill has been heavily criticized in the German technical literature for a variety of reasons.

48.1.6 Germany’s Tax Treaty Policy 48.1.6.1 Preliminaries As noted previously, Germany currently has close to 100 double taxation treaties. With some exceptions, the vast majority imply the exemption of profits received from subsidiaries abroad often combined with an activity clause under which profits received will only be exempted if and insofar as they stem from a so-called active investment. Some of the German double tax conventions are extremely complex, such as the German–US and the German–Swiss treaties which are generally viewed as among the most complicated tax treaties in the world.

12

Joined Cases C-6/90 and C-9/90 Andrea Francovich and Danila Bonifaci & Others v. Italian Republic (Francovich) [1991] ECR I-5357.

886 Gerhard Kraft

48.1.6.2 Relevance of MLI On 6 November 2020, the German upper house of parliament approved the ratification of the Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting (MLI). As Germany was among the sixty-eight initial jurisdictions that signed the MLI on 7 June 2017 in Paris, it will come as a surprise that only fourteen German double tax conventions qualify as ‘covered tax agreements’ (CTAs). Clearly, the MLI aims to combat base erosion and profit shifting (BEPS) by permitting jurisdictions to implement tax treaty-related measures developed through the OECD/G20 BEPS Project in their existing bilateral tax treaties in a coordinated and efficient manner. Currently, it is unclear whether the German implementation will be viewed as a non-starter or a complete success; often good intentions do not meet reality.

48.1.7 The Role of Courts in International Tax Matters in Germany 48.1.7.1 The role of Germany’s Supreme Court in international tax matters Germany’s Supreme Court does not act as a tax court as its primary function. Nevertheless, some noteworthy cases have been decided by the court on tax matters. Recently, the field of international taxation has gained considerable relevance whereas over the previous two decades its tax exposure primarily lay with inheritance matters. Thus, in a very prominent judgment released on 17 December 2014, the First Senate of the Federal Constitutional Court declared sections 13a, 13b, and 19 section 1 of the Inheritance and Gift Tax Act (Erbschaftsteuer-und Schenkungsteuergesetz—ErbStG) as unconstitutional based on violations of the Constitution which resulted in the incompatibility of the submitted provisions with its equal treatment clause.13 Two comparatively recent decisions specifically concerning international tax matters have received broad attention from the international tax community. In a holding dated 15 December 2015, the Supreme Court ordered that so-called treaty overrides in national statutory law were permissible under the Constitution. The decision caused unprecedented controversy in the scholarly German international tax world. Whereas the Supreme Tax Court’s First Senate in line with the prevailing view in the technical literature argued in favour of a violation of constitutional principles, the Supreme Court held that the federal legislature should not be barred from enacting statutes even if they contravene international treaties within the meaning of article 59 section 2 sentence 1 of the Constitution. Such was the decision of the Second Senate in an order dealing with the double taxation treaty concluded between Germany and Turkey in 1985. The Supreme Court took the position that article 59 section 2 sentence 1 of the Constitution

13

Art. 3 sec. 1.

The German Perspective on International Tax Law 887 provides that, in the national context, international treaties have the same rank as statutory federal law unless they fall within the scope of a more specific constitutional provision dealing with the relationship between domestic and international law and according a different rank to such international treaties (referred to as ‘opening clause’— Öffnungsklausel). In order to substantiate its position, the Supreme Court referred to articles 1, 2, 23, and 24 of the German Constitution. The principle of democracy requires that within the boundaries set by the Constitution, later legislatures are able to revoke the legal acts of previous legislatures. Neither the rule of law nor the principle of the Constitution’s openness to international law (Grundsatz der Völkerrechtsfreundlichkeit des Grundgesetzes) yield a different result. Although the latter principle is also of constitutional rank, it does not entail an absolute constitutional duty to obey all rules of international law. In a further order concerning international tax matters published by the Supreme Court on 31 March 2021, the Third Chamber of the Second Senate of the Federal Constitutional Court granted relief in constitutional complaint proceedings in which a corporation had challenged a judgment of the Federal Finance Court (Bundesfinanzhof). In that judgment, the Federal Finance Court had declared lawful an income correction made on the basis of section 1(1) of the Foreign Transaction Tax Act (Außensteuergesetz—AStG) in respect of the partial write-down of an unsecured intra-group loan in default. The judgment violated the complainant’s right to be lawfully judged under article 101(1), second sentence of the Constitution because the Federal Finance Court did not refer the case to the Court of Justice of the European Union (CJEU) for a preliminary ruling pursuant to article 267(3) TFEU. The Supreme Court’s main lines of reasoning were based on two constitutional grounds: namely, a violation of the general guarantee of the right to equality (art. 3(1) GG) in its manifestation as a prohibition of arbitrary measures and its fundamental procedural right to be lawful judged (art. 101(2) second sentence GG). The third Chamber concluded that the constitutional complaint was admissible and well founded. It held that there was no need to decide here whether the challenged judgment violated article 3(1) of the Constitution in its manifestation as a prohibition of arbitrary measures on the grounds that the Federal Finance Court, in applying the arm’s-length principle as required by section 1 of the German Foreign Tax Act (AStG), appeared to presume that fully secured loans are the arm’s-length standard, without providing any reasons regarding the usual amount of collateral for the specific offsetting agreement and a possible correlation between that amount and the agreed interest rate. In any event, the Supreme Tax Court’s judgment violated the complainant’s fundamental procedural right to be lawful judged under article 101(1), second sentence of the Constitution because of the way the Federal Finance Court had handled its duty of referral to the CJEU pursuant to article 267(3) TFEU. Under well-established legal principles, the CJEU is a lawful judge within the meaning of article 101(1), second sentence of the Constitution, and the third Chamber of the Constitutional Court reasoned that in the case at issue the Supreme Tax Court’s failure to comply with the duty of referral under article 267(3) TFEU violated the complainant’s well-established rights. This

888 Gerhard Kraft was correct because the question of EU law relevant for a court decision was yet to be addressed in the CJEU’s case law. The Supreme Court thus held that neither an acte clair nor an acte éclairé situation according to CJEU case law could be identified.

48.1.7.2 The role of Germany’s Supreme Tax Court in international tax matters International tax case law lies in the hands of the responsible special court, the Supreme Tax Court, or often referred to as Federal Fiscal Court (Bundesfinanzhof), domiciled in the city of Munich. The Federal Fiscal Court is one of five federal supreme courts in Germany, established according to article 95 of the German Constitution. In terms of responsibility, the Supreme Tax Court is the federal court of appeal for tax and customs matters in cases which have already been heard by the lower courts, essential the local Tax Courts (Finance Courts). Established in 1950, the German Supreme Tax Court succeeded the Supreme Finance Court of the German Reich—Reichsfinanzhof— established in 1918. By virtue of the court’s organizational chart, the Court’s First Senate is responsible for international direct tax issues, and therefore will handle the bulk of international direct tax questions. Other senates may be entrusted with questions arising from international tax issues in terms of international inheritance tax issues (Second Senate), or international VAT issues (Fifth and Eleventh Senates). On rare occasions, other senates will decide international tax issues if, among other things, an international tax question is of relevance on account of a particular trial. An example is the Fourth’s Senate responsibility for partnership taxation and questions arising from cross-border activities of German-domiciled partnerships involved in international business transactions. It is of particular note that in the past the First Senate, with considerable self- confidence, applied the CILFIT14 doctrine, a case law-generated ECJ doctrine which states that if a judgment or rule of law is sufficiently clear, then a member state court has no duty to refer a question for a preliminary ruling to the ECJ. The German Supreme Tax Court often viewed a situation as so evident as to leave no room for any reasonable doubt. The court’s future application of the CILFIT doctrine will most probably have to change considerably as the German Supreme Court has ruled that a lack of reflection on decisions of the ECJ by the Supreme Tax Court, and its presentation of cases to the ECJ, could deprive taxpayers of their lawful right.

48.1.7.3 The influence of ECJ case law on the German international tax system Germany, as a member state of the EU, is often confronted with the decisions of the ECJ. In order to address the relationships between ECJ case law and German international tax law it is necessary to analyse the place and the importance of Germany, as a member state’s obligations derive from the EU legal order. As ECJ case law is necessarily casuistic 14

Case 283/81 CILFIT v. Ministero della Sanità (CILFIT) [1982] ECR 3415.

The German Perspective on International Tax Law 889 and praetorian law and evolves constantly, it should not be surprising that there is no clear dogmatic line as far as the impact of the ECJ’s case law on German international taxation is concerned. Nevertheless, there are striking features resulting from ECJ case law that have impacted on German international tax law. First, the difference between resident and non-resident taxation has been applied to a wide extent, widely involving EU citizens or EU business undertakings. Secondly, the German court are comparatively eager to present cases to the ECJ in order to receive a preliminary ruling. Thirdly, there has been a tendency in the past for the German Ministry of Finance to strongly dislike ECJ decisions in favour of taxpayers and to sometimes delay implementation of the case law longer than can be reasonably justified. Lastly, as the CJEU is a lawful judge within the meaning of article 101(1), second sentence of the German Constitution, it will be very likely that in the future many cases will be presented to the CJEU. Therefore, ECJ case law will remain the most effective driving force in terms of international tax developments in Germany despite OECD efforts and similar initiatives.

48.1.7.4 Lower German tax courts Local tax courts, sometimes described as ‘Finance Courts’ (Finanzgericht) are special courts established by one of the sixteen states (Bundesländer). In Germany, local tax courts are the courts of first instance for legal recourse in financial disputes. It is the judges’ obligation to rule on legal disputes between tax citizens and the tax authorities. Depending on their internal organizational rules, the local tax court’s senates hear cases on international tax law. Clearly, the level of specialization at such senates is not comparable with the Supreme Tax Court’s First Senate level of specialization. The following enumeration lists all German local tax courts: three Fiscal Courts of North Rhine-Westphalia (Düsseldorf, Cologne, Münster), Fiscal Court of Baden- Wuerttemberg (Stuttgart with branch offices in Freiburg), Fiscal Courts of Bavaria (Munich; Augsburg, Nuremburg, Fiscal Court of Berlin—Brandenburg (Cottbus), Fiscal Court of Bremen, Fiscal Court of Hamburg, Fiscal Court of Hesse (Kassel), Fiscal Court of Lower Saxony (Hannover), Fiscal Court of Saxony-Anhalt (Dessau), Fiscal Court of Thuringia (Erfurt), Fiscal Court of Saxony (Leipzig), Fiscal Court of Schleswig- Holstein (Kiel), Fiscal Court of Rhineland-Palatinate (Neustadt), and Fiscal Court of Saarland (Saarbrücken).

47.1.8 German Anti-Abuse Tax Legislation concerning International Tax Matters 48.1.8.1 The German view on international tax avoidance Long before the international tax avoidance issue became en vogue at the OECD level, Germany was conspicuous in having general and specific anti-abuse clauses aiming at schemes particular with international ramifications. Germany was a front-runner as far

890 Gerhard Kraft as the application of general anti-avoidance rule (GAAR) provisions, such as section 42 of the General Tax Code, to international tax matters by the German tax administration. Additionally, the German legislator was among the first to introduce specific anti- avoidance rules concerning CFC legislation, anti-treaty-shopping clauses, exit taxation, and rules on foreign-family foundations.

48.1.8.1.1 General overview The German GAAR in section 42 of the General Tax Code (hereinafter GTC) consists of two paragraphs. The GTC is sometimes referred to as the ‘Federal Code of Tax Procedure’.15 The GAAR restricts the principle that every taxpayer may order their affairs so as to minimize their tax liability. Section 42 GTC thereby disallows the tax effects of any legal arrangement that constitutes an abuse of rights or—in German terminology ‘Rechtsmissbrauch’. Essentially, such legal norm intends that taxpayers may not arrange their affairs to manipulate or distort the economic reality of a transaction. Any respective manipulation or distortion is deemed to be tax avoidance and, consequently, is countered by the doctrine of Rechtsmissbrauch or, in the context of tax law, by the more specific concept of abuse of possible legal arrangements. In German tax terminology, the suitable term is ‘Gestaltungsmissbrauch’ which can be freely translated a\s ‘abuse of structuring options’. Thus, to summarize, section 42 GTC requires the taxpayer to be taxed according to the true economic substance of the transaction in question rather than according to its artificial form. The German Ministry of Finance provides a non-technical English translation,16 the heading of which reads ‘Section 42—Abuse of tax planning schemes’. Section 42 paragraph 1 GTC in its English translation reads that: it shall not be possible to circumvent tax legislation by abusing legal options for tax planning schemes. Where the element of an individual tax law’s provision to prevent circumventions of tax has been fulfilled, the legal consequences shall be determined pursuant to that provision. Where this is not the case, the tax claim shall in the event of an abuse within the meaning of subsection 2 arise in the same manner as it arises through the use of legal options appropriate to the economic transactions concerned.

Section 42 paragraph 2 GTC tries to define the notion ‘abuse’. It is translated by the German Ministry of Finance as follows: An abuse shall be deemed to exist where an inappropriate legal option is selected which, in comparison with an appropriate option, leads to tax advantages unintended by law for the taxpayer or a third party. This shall not apply where the

15

Prebble and Opacic, Bulletin for International Taxation (Apr. 2008), 151. for another translated version, KPMG, Deutsches Außensteuerrecht, German International Taxation, 1st ed. (2010), 1.1.1 Tax Procedure Law (selected provisions); another translation of sec. 42 GTC can be found in Prebble and Opacic, ibid. 16 Cf.

The German Perspective on International Tax Law 891 taxpayer provides evidence of non-tax reasons for the selected option which are relevant when viewed from an overall perspective.

With respect to the application of the German GAAR, section 42 GTC, one of the main problems is caused by the fact that such legal provision does not specify exactly what constitutes an abuse of possible, or feasible, legal arrangements. It should therefore not be surprising that the lower tax courts as well as the German Federal Tax Court have shaped the German GAAR’s interpretation and scope. Apart from the GAAR located in section 42 GTC, Germany has a long history of specific anti-avoidance rules (SAARs). It is noteworthy that Germany was the second country after the USA to introduce CFC rules in 1972. Furthermore, rules on transfer pricing, exit taxation, treaty shopping, thin capitalization (interest barrier rules), licence barriers, hybrid mismatches, and double-dip-arrangements today form, inter alia, integral parts of the German anti-tax-avoidance arena. The relationship between the German GAARs and SAARs has been one of the most controversially discussed methodological topics in German scientific tax literature over recent years. With regard to double tax treaties, Germany subscribes to a dichotomy type of approach. The first part of the anti-abuse-strategy focuses on agreeing to explicitly reserve the right to apply domestic GAARs and SAARs. This strategy has, for instance, been applied in the recently revised or concluded treaties with Australia, China, Costa Rica, and Finland. As such clauses directly point to the German GAAR, they could be interpreted as ‘treaty GAARs’. Nevertheless, Germany does not apply an approach to explicitly incorporate GAARs in double tax treaties. Rather, as the second bedrock of the its anti-abuse-strategy, Germany relies on limitation-on-benefits clauses (LOB) in tax treaties with other countries. Thus, Germany has been eager to implement these clauses in treaties. The most prominent examples of extremely complex and complicated LOB clauses in recent treaties generally considered to be the treaties with Japan, Australia, and the USA.

48.1.8.1.2 The tax avoidance scheme, arrangement, or transaction According to the section 42 GTC, the definition of the first primary element of the German GAAR (i.e. the tax avoidance scheme, arrangement, or transaction) requires that two abuse or avoidance elements are fulfilled: an ‘inappropriate legal arrangement’ and a thereby generated ‘legally unintended tax benefit’. Conversely speaking, the tax authority is obliged to demonstrate that an inappropriate legal structure has caused an unintended tax advantage or tax benefit. The taxpayer may then rebut that finding by arguing that non-tax reasons can be identified to countermand the accusation of avoidance. Therefore, the legal design criteria of a tax avoidance scheme, arrangement, or transaction gravitate around the notion of an ‘inappropriate legal arrangement’ and a ‘legally unintended tax benefit’. Consequently, the first core question focuses on what constitutes an appropriate arrangement. There should be little doubt that ‘appropriateness’ or

892 Gerhard Kraft ‘inappropriateness’ requires a judgmental assessment. Older case law17 and a respective decree issued by the German Ministry of Finance18 circumscribe ‘inappropriate’ with notions such as ‘unusual’, ‘not economically justifiable’, ‘complicated’, ‘cumbersome’, ‘intricately’, ‘unnatural’, ‘artificial’, ‘ineffective’, ‘paradoxical’. The decree contains examples indicating the inappropriateness of an arrangement: • the arrangement would not have been chosen by rational third parties considering the underlying facts and the economic purpose without the intended tax benefit; • the interposition or other use of relatives or related parties or companies was exclusively motivated by tax considerations; • the shifting or other transfer of income or assets to another taxpayer was motivated exclusively on tax grounds. According to the prevailing view in the literature,19 a mere intent to save taxes does not render an arrangement abusive.

48.1.8.1.3 The tax benefit, gain, or advantage GAAR approach The second core question is what should be understood by a ‘legally not intended tax benefit’. In order to evaluate whether an arrangement entails tax benefits, the German Ministry of Finance20 takes the position that a comparison needs to be made between the tax effects of the factual (i.e. realized) arrangement and the tax effects of a hypothetical appropriate arrangement. The effects of tax benefits are not limited to the taxpayer in question but may extent to third parties with a ‘certain proximity’ to the taxpayer in question. This group of persons encompasses related parties and/or relatives. Treaty GAARs As pointed out earlier, Germany has not agreed to treaty GAARs; rather, in several treaties it has reserved the right to apply the domestic GAAR. It should be evident, therefore, that no material differences in the application of a domestic GAAR should arise if applied to a treaty or to a non-treaty situation. Domestic SAARs Domestic SAAR appear to be much more specific in comparison to the German GAAR. This not only applies in respect of the conditions for their application but also in terms of the legal consequences outlined by the SAARs. 17 Cf. e.g., German Federal Tax Court, 19 August 1999, I R 77– 96; German Federal Tax Court, 1 February 2001, IV R 3/00. 18 Decree of the Federal Ministry of Finance, 31 January 2014, Federal Tax Gazette 2014 I, 290, explanation of sec. 42, at 2.2. 19 Hübschmann, Hepp, and Spitaler, § 42 AO m.no. 71; Tipke and Kruse, § 42 AO m.no. 19; Spindler, StbJb. 08/09, 55; Spindler, Stbg. 10, 52; German Federal Tax Court, 17 February 2004, I R 9/03; German Federal Tax Court, 16 September 2004, IV R 11/03. 20 Decree of the Federal Ministry of Finance, 31 January 2014, Federal Tax Gazette 2014 I, 290, explanation of sec. 42, at 2.2.

The German Perspective on International Tax Law 893 In terms of conditions, the application of SAARs can be illustrated with the examples of the CFC legislation and the domestic anti-treaty-shopping provision. All other SAARs tend to act similarly (i.e. in a very detailed and specific way compared to the GAAR). German CFC rules21 very precisely specify all typical features of CFC legislation. First and foremost, the rules for defining a CFC under the German perception are outlined in a detailed manner accompanied by quantified threshold requirements. Further, the definition of CFC income, rules for computing and attributing income, and provisions to prevent or eliminate double taxation set out the mechanics of the German CFC legislation with much more precision than could be done in a GAAR.22 The German domestic anti-treaty-shopping rule, section 50d, paragraph 3 of the Income Tax Act, outlines the conditions when treaty and/or EU benefits will be denied under domestic law. The legal consequences take a very particular approach to quantifying treaty benefits according to the shareholding structure of a foreign treaty country-domiciled corporation. Reduction is granted according to the fictitious entitlement to treaty reduction of direct and indirect shareholders of such company.

48.1.8.2 CFC legislation As noted previously, Germany was the second country after the US to implement CFC rules in 1972. The Foreign Tax Act describes a CFC as a foreign company whose capital or voting rights are either directly or indirectly majority-owned by German residents at the end of its fiscal year. The rules even apply when there are partnerships in the participation chain. Section 8 of the German Foreign Tax Act (Außensteuergesetz—AStG) defines an explicit catalogue of income that is accepted as active income. Income from portfolio investments and from holding patents and licences are not included. The relevant low- tax threshold amounts to 25%. Thus, a German-domiciled investor in a controlled foreign corporation will be liable for taxes on all the passive investment income of the investment corporation. Foreign taxes paid by the subsidiary on that passive income can be credited against German taxes on application.

48.1.8.3 Special areas 48.1.8.3.1 Related party transfer pricing For many years, German has introduced extensive related-party transfer pricing rules. Empirically, transfer pricing issues appear to constitute one of the most relevant components of any field tax audits of German subsidiaries of multinational groups. To a large extent, German transfer pricing rules are in accordance with the OECD reports and recommendation. They are, however, more detailed and in some cases 21

Cf. ss. 7–14 Foreign Tax Act. an example, the German GAAR contains 117 words whereas the German CFC provisions flanked by administrative pronouncements contain 11,688 words. Cf. Decree of the Federal Ministry of Finance, 14 May 2004, BMF IV B 4-S 1340-11/04, Federal Tax Gazette 2004 I, supplement, 3. 22 As

894 Gerhard Kraft more specific. Based on the premise that related parties should trade at arm’s length, transfer pricing has been the subject of widespread legislative, administrative, and certainly judicial guidance. The basic principle under which each unit involved in a group should be looked on as an independent entity has been widely modified and subjected to formal qualifications. Added to this is that in most cases the burden of proof is on the taxpayer that all their arrangements are in line with fictitious third-party transactions. Data-base analyses play a crucial role and often the data bases are not accessible to taxpayers. There are formal requirements for the acceptance of transfer pricing arrangements by the tax administration which should be substantiated by written agreements concluded in advance.

48.1.8.3.2 Exit taxation When a shareholder moves abroad, section 6 of the Foreign Tax Act subjects built-in gain in an investment of at least 1% in a corporation’s capital (sec. 17 German Income Tax Act—EStG) to German taxation. Thus, the previously untaxed hidden reserves will be taxed even without cash inflow deeming as a fictitious sale of the share package. Through this rule, Germany secures the right to tax the hidden reserves in privately held shares generated by corporations if the taxpayer is resident in Germany at the time they move abroad. The reason for such a rule lies in the mechanics of double taxation treaties. Under most double taxation treaties, the state of residence enjoys the right to tax capital gains from the sale of privately held shares in a corporation, similar to the general rule set out in article 13 paragraph 5 of the OECD Model Tax Convention. Therefore, Germany, in principle, would lose the right to tax silent reserves (built-in gain) in the shares if a taxpayer moves abroad. In order to prevent the tax-free export of such a built- in gain, German exit tax legislation has long invented countermanding instruments with a tendency for excessive taxation. Currently, the future developments concerning the German exit tax system are heavily and controversially discussed. It has been argued, that after the ECJ ruling in the Wächtler23 case dated 26 February 2019, governmental plans are not in line with primary EU law.

48.1.8.3.3 Thin capitalization Germany has a long track record with respect to thin-capitalization issues. The interest limitation rules currently in place restrict the maximum amount of net interest expense a business entity can deduct in arriving at the tax base of 30% of the relevant profit after adding back the interest expense and depreciation and amortization and after deducting interest income. This system is generally referred to as EBITDA and is available for

23

Case 58-581/17 Martin Wächtler v. Finanzamt Konstanz, ECLI:EU:C:2019:138.

The German Perspective on International Tax Law 895 set-off; this encompasses the interest limitation rules of not only inter-company financing schemes but also third-party financing arrangements. This holds true even if there is no recourse to an affiliated company or related person. Some exceptions apply to the application of the thin-capitalization rules: • the amount of interest expense is less than €3 million; • the business in question does not belong, or only partially belong, to a tax group; • the business belongs to a tax group and its equity ratio as of the preceding balance sheet date exceeds or is at least equal to that of the (worldwide) group. The equity ratio is measured by the proportion of the shareholder’s equity relative to the balance sheet total but failure to reach the equity test of the group by two or less percentage points is be disregarded. The group clause and the escape clause, or equity test, for corporations will only be applicable if it can be demonstrated that no detrimental shareholder debt financing as defined by law has been implemented.

48.1.8.3.4 Family foundations Germany has an in-depth regime dealing with foreign-family foundations. Under section 15 of the Foreign Tax Act, certain attributions similar to CFC rules apply to specific foreign trusts and family foundations. In principle, the family-foundation attribution regime comes into play in cases in which either the settlor or a beneficiary is resident in Germany. Further, the family foundation’s undistributed income accumulated by the trust or family foundation may be attributed to the German resident in proportion to the that person’s ‘share’ in the trust or foundation’s assets. Under the applicable rules, the trust or foundation’s different types of income will be attributed directly to the German resident as if that individual had received the income themselves. This means that the flat tax rate applicable for capital income of individuals (26.375%) will apply as far as the trust or foundation has received such capital income. The individual’s progressive tax rate will come into play far as other income is attributed to the German settlor or beneficiary. For trusts or foundations tax resident in a member state of the EU or the EEA, an escape clause might be relevant in cases in which it can be demonstrated that neither the settlor nor the beneficiaries have any control over the trust or foundation’s assets. Furthermore, the tax rules concerning foreign-family foundations contain additional provisions with regard to the attribution of undistributed income of CFC to the trust or foundation. This applies to cases in which foreign companies are controlled by the trust or foundation which itself receives ‘passive income’ being subject only to low taxation. There is, however, an escape clause for companies with residence in a member state of the EU or the EEA where a specific substance test is met. Consequently, the income of a CFC attributed to the trust or foundation in a first step may then be attributed to the German settlor or beneficiary in a second step.

896 Gerhard Kraft Cases in which the trust or foundation is itself a beneficiary of another foreign trust or family foundation follow similar rules. The other trust’s or foundation’s undistributed income may be attributed to the ‘first-tier’ trust or foundation, the favoured trust or foundation. Again as expected due to EU primary law provisions, an escape clause for trusts or foundations resident in a member state of the EU or the EEA will apply in cases in which it can be demonstrated that the favoured trust or foundation has no control over the other trust or foundation’s assets.

48.1.8.3.5 Hybrid mismatches A new foreseeable tax regime will significantly change the German international tax landscape in the near future. A new section 4K of the Income Tax Act will be implemented in the German income tax system containing so-called anti-hybrid rules. The envisaged rules will deal with deduction/non-inclusion (D/NI) situations and reverse hybrid situations, and with double deduction situations and tax residency mismatches. Further, specific provisions will be designed to counter imported mismatch arrangements.

48.2 Concluding Remarks 48.2.1 International Tax Education in Germany Traditionally, tax education in general academia is shared between law and business/ economic university departments. At the inception, probably in the 1960s, only a handful of scholars were dedicated to international tax topics. At present, international tax education forms an integral and relevant part of a business student’s education with a concentration on tax, accounting, and finance. Although the study of international tax law is not mandatory for trainee lawyers, there is outstanding scholarly penetration as far as international tax issues are concerned. Some of the most prominent international tax researchers hold chairs in law faculties with an unprecedented international reputation. Nevertheless, it should be noted that international tax education broadly relies on the efforts of business professors specializing in taxation with a particular focus on international tax. In recent decades, more practically oriented so-called universities of applied sciences as well as government institutions of higher tax education have also assumed a crucial role in international tax formation. Some of those institutions’ teachers and researchers have an outstanding teaching and research output. Thus, it is fair to say that there are excellent chances for young academics interested in specializing in international taxation to start a promising professional career after having experienced a sound theoretical and academic basic foundation.

The German Perspective on International Tax Law 897

48.2.8 Some Reflections Concerning the Future International Tax Environment Germany, as a highly internationally integrated and interdependently entwined economy, will most likely be among the prominent trendsetters in respect of international tax developments in the future. This can be expected as taxation is one of the most important variables in economic decision making, particular internationally. As many commercial decisions have long-term implications, they are likely to be affected by the way the German international tax system develops in the future.

Chapter 49

T he Perspect i v e of t h e EAC on Internat i ona l Tax L aw Afton Titus

49.1 Introduction The effective taxation of the digital economy is a priority for all countries. For developing countries, however, the digital economy has exacerbated the challenges they face when implementing the current international tax law rules. At present, these rules do not fairly and effectively attribute profits to where the income is being generated.1 Instead, a disconnect is created between the place where the profits are generated and the place where the profits are taxed.2 This is a long-standing problem. At the very early stages of mass digitalization, calls were made for the design of fair rules for the taxation of the digital economy.3 While the term ‘fairness’ is often used, it commonly goes undefined despite having several meanings.4 The aim of this chapter is to determine what ‘fair’ international tax law rules would look like for African countries. In this respect, fairness is understood in terms of its legal meaning as encompassing equality, certainty, and legitimacy.5

1 M. Ndajiwo, ‘The Taxation of the Digitalized Economy: An African Study’, International Centre for Tax and Development Centre (ICTD) Working Paper 107, Institute of Development Studies (2020), 1, 10. 2 G-24 Working Group on Tax Policy and International Co-operation, ‘Proposal for Addressing Tax Challenges Arising from Digitalization’ (2019). 3 J. VanderWolk, ‘Direct Taxation in the Internet Age: A Fundamentalist Approach’, Bulletin for International Taxation 54 (2000), 173. 4 I. J. J. Burgers and I. J. Mosquera Valderrama, ‘Fairness: A Dire International Tax Standard With No Meaning?’, Intertax 45 (2017), 767, 768. 5 Ibid., 771.

900 Afton Titus Fairness cannot be evaluated in a vacuum and the East African Community (EAC) has been selected as a suitable context for this purpose. The EAC has been selected because the regional bloc has the objective of harmonizing direct tax laws across the region;6 the ambition to form a federation;7 and exhibited promising growth rates—even during the global pandemic.8 Moreover, this regional bloc has been ranked as the most integrated regional economic bloc in Africa.9 Furthermore, the taxation of the digital economy is particularly relevant for the EAC. It plans to create a ‘Single Digital Market’ across East Africa which would result in it becoming the world’s ninth largest digital economy.10 This is bolstered by the fact that some of the EAC partner states have some of the highest rates of internet penetration across Africa.11 Furthermore, the impact of increased digitalization across the continent has meant that mobile e-commerce has dominated online retail in Africa.12 From an EAC perspective, this is also significant because 50% of Kenya’s total e-commerce, for instance, was generated by mobile online shopping.13 This means that it is essential that the EAC consider how to effectively tax the digital economy. This chapter argues that the fairest and most effective way for the EAC to do this is through the adoption of a corporate income tax system based on unitary taxation and formulary apportionment. Given the large number of multinational enterprises (MNEs) housed in the region and the way in which they often act as monopolies, such a corporate income tax system would be effective in taxing all MNEs, not only those prominent in the digital economy. It is further argued that as an interim measure, the EAC should consider the viability of adopting a digital services tax (DST), similar to that recently enacted in Kenya. Ancillary to this, the EAC partner states should consider strengthening their taxing rights in their double taxation agreements (DTAs) through the adoption of article 12A and the draft article 12B of the UN Model Treaty.14 Finally, this chapter argues that the OECD’s two-pillar approach in its current form is not suited for an African-country context. These proposals should undergo significant amendments 6

East African Community, Protocol on the Establishment of the East African Community Monetary Union (30 Nov. 2013), art. 8(1). 7 East African Community, Treaty for the Establishment of the East African Community (30 Nov. 1999), preamble and art. 11(3). 8 A. Mpoke-Bigg, ‘East Africa Holds Its Ground as Africa’s Fastest Growing Region, Despite COVID- 19 Disruption’, African Development Bank (10 July 2020), available at https://www.afdb.org/en/news-and- events/press-releases/east-africa-holds-its-ground-africas-fastest-growing-region-despite-covid-19-dis ruption-36809. 9 Africa Regional Integration Index, Report 2019, available at https://www.integrate-africa.org/filead min/uploads/afdb/Documents/ARII-Report2019-FIN-R40-11jun20.pdf. 10 World Bank, A Single Digital Market for the East African Community (2018), 1. 11 Statista, Internet Penetration Rate in Africa as of January 2021, By Region (2021), available at https:// www.statista.com/statistics/1176668/internet-penetration-rate-in-africa-by-region/. 12 Statista, Share of Mobile E-commerce in Selected African Countries in 2020 (2021), available at https:// www.statista.com/statistics/1175996/share-of-mobile-e-commerce-in-africa/. 13 Ibid. 14 United Nations, United Nations Model Double Taxation Convention between Developed and Developing Countries (2017).

The Perspective of the EAC on International Tax Law 901 before they are should be considered a viable option for the EAC. The chapter will briefly mention some of these necessary changes. In setting out these arguments, the chapter will evolve as follows: Section 49.2 will analyse the unitary taxation and formulary apportionment system and why it would be suited for the EAC. Section 49.3 will consider the viability of adopting a DST in the EAC before Section 49.4 details the DTA provisions which would assist in protecting that domestic taxing right. Finally, Section 49.5 will briefly analyse the OECD’s two-pillar approach before the chapter concludes. In making these recommendations, the author is aware of the great number of calls made for countries not to adopt tax measures ‘unilaterally’ and the need to seek ‘international consensus’. Unilateralism is another term which is frequently used without a clear understanding of its meaning. It has been argued that often the term in international tax means to ‘act against the United States’.15 Notwithstanding the uncertainty, the author will assume that what is meant is that one country acts on its own without regard for the implications of its actions for other countries which are connected to it through a network of global trade and treaties.16 The chapter therefore argues that in adopting a regional approach the EAC cannot be said to be acting unilaterally (in the strictest meaning of the term). More importantly, the negative connotation attached to unilateralism cannot be applied here because all the broader implications of the proposed recommendations are carefully analysed with a view to facilitating more fairness in international tax rules.

49.2 Unitary Taxation and Formulary Apportionment Unitary taxation treats a group of companies linked to each other by a sufficient degree of control as a single economically integrated unit.17 The profits or losses of each of the group companies are consolidated to form the taxable profit of the group as a whole.18 Once such consolidated taxable profit is determined, the profits are apportioned in terms of a formula to the countries within which the group companies operate.19 This formula is usually based upon an equal weighting of the labour, physical assets, and sales

15

W. Cui, ‘What is Unilateralism in International Taxation?’, American Journal of International Law, Unbound 114 (2020), 260. 16 Based on the definition of ‘unilateralism’ in the Cambridge English Dictionary: ‘the process or fact of deciding a policy or action without involving another group or country’. 17 W. Hellerstein, ‘Income Allocation in the 21st Century: The End of Transfer Pricing’, International Transfer Pricing Journal 12 (2005), 103, 105. 18 Ibid. 19 Ibid.

902 Afton Titus factors.20 While unitary taxation and formulary apportionment are complementary, they are capable of being applied separately.21 Unitary taxation is particularly effective in taxing the profits arising from the digital economy because a great deal of the complexity is a consequence of the sheer size of the groups involved.22 Commentators have noted that digitalization has not changed the core business activities which generate profits.23 As such, if the EAC were to adopt a system which effectively taxes MNEs, it would also have a system which effectively taxes the profits of the digital economy. This system would be appropriate for the EAC for the following reasons. First, it has long been established that the prevailing standard used to account for group company transactions—the arm’s-length principle—is problematic,24 especially for developing countries.25 This is true for the EAC as well. Despite having the longest standing transfer pricing department of all the EAC partner states, along with an increase in departmental staff complement, the Kenyan revenue authority takes approximately four to five years to finalize transfer pricing cases.26 This statistic is particularly poignant given that a 2021 report estimates that Kenya lost US$885 million from transfer pricing in 2017.27 Similar transfer pricing losses have been reported in Uganda.28 Therefore, if the EAC were to adopt a unitary taxation system with formulary apportionment, transfer pricing would no longer be a concern as the intra-group transactions would be ignored. The EAC partner states could instead pool their resources to create a large unit to focus on implementing the new unitary taxation system and circumventing potential manipulations of formulary apportionment.

20 S. Picciotto, ‘Taxing Multinational Enterprises as Unitary Firms’, ICTD Working Paper 53, Institute of Development Studies (2016), 1, 27. 21 Hellerstein, ‘Income Allocation in the 21st Century’. 22 C. Lau and A. Halkyard, ‘From E-Commerce to E-Business Taxation’, Asia-Pacific Tax Bulletin 9 (2003), 2, 7. 23 G- 24, ‘Proposal for Addressing Tax Challenges Arising from Digitalization’; United Nations Committee of Experts on International Co- operation in Tax Matters, Tax Issues Related to the Digitalization of the Economy: Report (2019), 1, 3. 24 R. S. Avi-Yonah, ‘The Rise and Fall of Arm’s Length: A Study in the Evolution of US International Taxation’, University of Michigan Law School, John M. Olin Center for Law & Economics Working Paper No. 07-017 (2007); K. Sadiq, ‘The Fundamental Failing of the Traditional Transfer Pricing Regime— Applying the Arm’s Length Standard to Multinational Banks Based on a Comparability Analysis’, Bulletin for International Taxation 58 (2004), 67. 25 A. Turrino, ‘Back to Grass Roots: The Arm’s Length Standard, Comparability and Transparency— Some Perspectives from the Emerging World’, World Tax Journal 10 (2018), 303; United Nations, Practical Manual on Transfer Pricing for Developing Countries (2017), para 1.4.7. 26 A. Waris, ‘How Kenya Has Implemented and Adjusted to the Changes in International Transfer Pricing Regulations: 1920–2016’, ICTD Working Paper 69, Institute of Development Studies (2017). 27 Global Financial Integrity, Illicit Financial Flows in Kenya (2021), 23. 28 Global Financial Integrity, Illicit Financial Flows in Uganda (2021), 15.

The Perspective of the EAC on International Tax Law 903 Secondly, a large number of MNEs operate in East Africa with many of them headquartered there.29 Waris notes that in this respect Kenya is unique in that it hosts both foreign and local MNEs.30 This means that the EAC partner states will not only be home to the MNE’s factors of production, but would also offer a sizable market. Unitary taxation would therefore allow the EAC to properly account for all the roles its partner states play in the MNE’s profit generation. Thirdly, the implementation of unitary taxation with formulary apportionment would allow for the legal elements that make up the understanding of ‘fairness’ to be met. Equality would be fostered insofar as MNEs would be taxed based on their actual business decisions and the effect they have on its profits rather than on the performance of comparable companies under a transfer pricing analysis.31 In this way, local and foreign MNEs would be taxed in the same way as unitary taxation allows for less opportunities for value to be attributed to low-tax jurisdictions. Moreover, inter-nation equity is achieved when profits are apportioned between countries based on actual economic activities such that countries receive revenue commensurate to the use of that country’s infrastructure and services in profit generation.32 It has been recognized that formulary apportionment more readily facilitates inter-nation equity.33 Certainty would be provided by the fact that a system of unitary taxation and formulary apportionment would have to be legislated.34 Finally, legitimacy would be afforded to such a system because it would have to be adopted after a careful and reasoned governmental decision-making process whereby that particular country’s context would be analysed along with the implications for that country’s participation in the international community. That is not to say that the system proposed here is perfect. The disadvantages of this system have been discussed at length and are identified as: defining the tax base;35 difficulty in determining the consolidation rules and at which stage companies are to be considered sufficiently controlled so as to constitute a group;36 the arbitrariness of the selected factors;37 the fact that even the factors may be subject to

29 Waris, ‘How Kenya Has Implemented and Adjusted to the Changes in International Transfer Pricing Regulations: 1920–2016’, 13; African Business, Africa’s Top 250 Companies in 2022, available at https://african.business/dossiers/africas-top-companies/#ranking>. 30 Kenya High Commission, Why Invest in Kenya, available at https://www.kenyahighcom.org.uk/ why-invest-in-kenya. 31 L. M. Kauder, ‘Intercompany Pricing and Section 482: A Proposal to Shift from Uncontrolled Comparables to Formulary Apportionment Now’, Tax Notes (1993), 485. 32 K. Sadiq, ‘Unitary Taxation— The Case for Global Formulary Apportionment’, Bulletin for International Taxation 55 (2001), 275; Picciotto, ‘Taxing Multinational Enterprises as Unitary Firms’, 27. 33 K. Brooks, ‘Inter- Nation Equity: The Development of an Important but Underappreciated International Tax Value’, in J. G. Head and R. Krever, eds, Tax Reform in the 21st Century: A Volume in Memory of Richard Musgrave (2009), 23. 34 S. I. Langbein, ‘The Unitary Method and the Myth of Arm’s Length’, Tax Notes 30 (1986), 625. 35 G-24, ‘Proposal for Addressing Tax Challenges Arising from Digitalization’, 7. 36 N. Hertzig, M. Teschke, and C. Joisten, ‘Between Extremes: Merging the Advantages of Separate Accounting and Unitary Taxation’, Intertax 38 (2010), 334. 37 Ibid.

904 Afton Titus manipulation;38 and that the factors do not properly capture the unique aspects of the digital economy.39 However, many of these concerns can be addressed. In terms of defining the tax base and the rules for consolidation, it has been recommended elsewhere that the EAC adapt the European Union’s proposals for a common consolidated corporate tax base (CCCTB) to define the tax base and the rules for consolidation.40 In terms of determining at which stage companies should be linked together, it is recommended that the EAC look to the domestic laws of the EAC partner states to set thresholds for control when determining this at a regional level. For instance, the Tanzanian Income Tax Act41 defines companies as being ‘associates’ once an entity, either alone or together with other associates and either directly or indirectly, ‘controls or may benefit from fifty percent or may of the rights to income or capital or voting power’ of the other entity.42 This definition is mirrored in the Ugandan Income Tax Act.43 When considering the choice of factors by which to apportion income, this is ultimately a political decision.44 There will therefore always be some degree of arbitrariness. The author is of the view that this is not fatal to the idea of a unitary taxation system with formulary apportionment. Neither is the fact that the formulary apportionment factors, once decided upon, may be subject to manipulation. Picciotto has commented that even if developing countries were to adopt a single factor of sales in implementing formulary apportionment, they would still probably earn more income from this lone factor than they currently do under the arm’s-length principle.45 Developing countries like the EAC partner states therefore have a lot more to gain than to lose from adopting unitary taxation with formulary apportionment. The peculiarities of the digital economy may, however, require some form of adjustment to the formula’s factors. It has been argued that a fourth factor, a data or user 38 M. F. De Wilde, ‘Tax Competition Within the European Union— Is the CCCTB Directive a Solution?’, Erasmus Law Review 7 (2014), 24; M. P. Devereux et al., Corporate Tax Reform in the EU: Weighing the Pros and Cons, Oxford Centre for Business Taxation (20 Mar. 2011), available at https:// voxeu.org/article/eu-corporate-tax-reform-weighing-pros-and-cons. 39 M. Vascega and S. van Thiel, ‘The CCCTB Proposal: The Next Step Towards Corporate Tax Harmonization in the European Union’, European Taxation 51 (2011), 379; N. Haskic, ‘The Arm’s Length Principle and the CCCTB: Solutions to Transfer Pricing Issues for Individual Countries and the European Union?’, Revenue Law Journal 19 (2009), 11. 40 A. Titus, ‘How Can the East African Community Guard Against Base Erosion and Profit Shifting While Working Towards Deeper Integration?’, World Tax Journal 9 (2017), 565; A. Titus, ‘The Design of a Corporate Income Tax System for the East African Federation and How to Protect It’, PhD thesis, UvA- DARE (2020), 147, available at https://dare.uva.nl/search?identifi er=60392d65-0b46-480d-a99c-6fca 5567b16; E. D. Siu et al., ‘Unitary Taxation in Federal and Regional Integrated Markets, ICTD Research Report 3 (2014), 65. 41 Tanzania: Income Tax Act, ch. 332, rev. ed. 2019. 42 Ibid., s. 1. 43 Uganda: Income Tax Act, 1997, ch. 340, s. 3(2)(h). 44 Hertzig et al., ‘Between Extremes: Merging the Advantages of Separate Accounting and Unitary Taxation’. 45 Picciotto, ‘Taxing Multinational Enterprises as Unitary Firms’, 28.

The Perspective of the EAC on International Tax Law 905 participation factor, should be added to account for the value added by consumer- generated data.46 Moreover, Vėgėlytė suggests how such user participation may be graded to reflect its varying importance in the generation of a particular MNE’s profits.47 It is therefore recommended that the EAC consider incorporating this fourth factor into the formula. Such incorporation would allow it to account for the burgeoning e- commerce within the region. While some may argue that unitary taxation and formulary apportionment may be too drastic and especially considering the double taxation implications, Picciotto counters that the OECD itself in its Base Erosion and Profit Shifting (BEPS) Project and transfer pricing documentation incorporates some elements of unitary taxation.48 Its use is therefore not a foreign concept in international tax law. Moreover, the G24 and also Avi-Yonah and Tinhage note that formulary apportionment is compatible with DTAs.49 It has also been argued that the EAC should consider implementing unitary taxation with formulary apportionment as a minimum alternative tax.50 In sum, while unitary taxation with formulary apportionment as a system is imperfect, it offers the EAC promising opportunities to effectively tax the profits arising from the digital economy. Because these profits are typically embodied in large MNEs, of which there are many in the EAC, such a system is particularly suited to the EAC context.

49.3 Digital Services Tax As an interim measure, the EAC may wish to consider imposing a DST. A DST is a tax on the revenue generated from the provision of digital services over online platforms. This may also include the revenues gained from the gathering of consumer data. While some have argued that a DST is more in the nature of an indirect tax,51 Lamensch persuasively

46 G-24,

‘Proposal for Addressing Tax Challenges Arising from Digitalization’, 7; Titus, ‘How Can the East African Community Guard Against Base Erosion and Profit Shifting While Working Towards Deeper Integration?’, 595. 47 E. Vėgėlytė, ‘Deconstructing User Participation: Why in the Digital Era Advertising Income is Different from Other Business Income’, International Transfer Pricing Journal 27 (2020), 180. 48 Picciotto, ‘Taxing Multinational Enterprises as Unitary Firms’, 7. 49 G- 24, ‘Proposal for Addressing Tax Challenges Arising from Digitalization’; R. S. Avi-Yonah and Z .P. Tinhaga, ‘Formulary Apportionment and International Tax Rules’, in S. Picciotto, ed., Taxing Multinational Enterprises as Unitary Firms (2017), 67. 50 A. Titus, ‘The Promise of Non- Arm’s Length Practices—Is the Destination-Based Cash Flow Tax or Unitary Taxation the Panacea of which Developing Countries are in Search?’, in I. J. Mosquera Valderrama, D. Lesage, and W. Lips, eds, Taxation, International Cooperation and the 2030 Sustainable Development Agenda (2021). 51 Ndajiwo, ‘The Taxation of the Digitalized Economy’, 13; M. D. Astuti, ‘Three Approaches to Taxing Income from the Digital Economy—Which is Best for Developing Countries?’, Bulletin for International Taxation 74 (2020), 721, 727.

906 Afton Titus argues that DSTs are more like a direct tax.52 This is because with DSTs the person paying the tax suffers a loss of income, whereas with indirect taxes the final burden is placed on someone else.53 Moreover, a DST is designed to be levied on an annual basis which is similar to direct taxes.54 There are several benefits to the introduction of a DST. DSTs are easily introduced through domestic legislation; they would cause no treaty conflicts because the tax would fall outside the purview of DTAs; and it has the potential to raise large amounts of revenue because it is levied on gross turnover.55 Conversely, the disadvantages of DSTs are that a high rate may result in the tax being more than the service provider’s profit margin; double taxation may come about because it is not covered by DTAs; and issues of non- discrimination may arise if residents and non-residents are taxed differently.56 Another drawback is that DSTs readily allow for the tax to be passed on to the consumer.57 It is recommended that the EAC carefully weigh up the advantages and disadvantages of a DST before introducing one. Having said that, Kenya already has a DST. A 1.5% tax is levied on the gross transactional value of a taxpayer whose income derives from a ‘digital market place’ in Kenya.58 A ‘digital market place’ is defined as ‘a platform that enables the direct interaction between buyers and sellers of goods and services through electronic means’.59 The tax is levied against both residents and non-residents. The next question is whether the other EAC partner states should implement a DST as well. The current circumstances in the EAC partner states make it clear that a coordinated approach to the taxation of the digital economy is necessary. Kenya also imposes an excise tax on mobile-based transactions,60 while Uganda has an excise tax on mobile money transactions along with a daily levy on the use of social media platforms.61 Tanzania has imposed a registration and licensing fee for media websites, online TV, radio channels, and podcasters.62 There are some advantages for the EAC in implementing a DST. First, governments would be perceived by citizens to be actively doing something to tax the digital economy. Secondly, the introduction of domestic legislation introducing a DST is relatively easy. Thirdly, the EAC partner states would be joining a growing list of countries which already have DSTs.

52 M. Lamensch, ‘DSTs: A Critical Analysis and Comparison with the VAT System’, in P. Pistone and D. Weber, eds, Taxing the Digital Economy: The EU Proposals and Other Insights (2019). 53 Ibid. 54 Ibid. 55 Astuti, ‘Three Approaches to Taxing Income from the Digital Economy’, 727. 56 Ibid.; Ndajiwo, ‘The Taxation of the Digitalized Economy’, 13. 57 Ibid. 58 Kenya: Income Tax Act, ch. 470, 1973 (as amended), s. 12E read with para. 12 of the Third Sch. 59 Ibid., s. 3(ba). 60 Kenya: Finance Act 10 of 2018, s. 32(b)(i). 61 Uganda: Excise Duty (Amendment) Act 2 of 2018, s. 4(5) read with Sch. 2. 62 Tanzania: Electronic and Postal Communications (Online) Content Regulations 2018.

The Perspective of the EAC on International Tax Law 907 A DST, however, would come with some disadvantages. First, there is a great deal of negative coverage about countries undertaking ‘unilateral action’ to implement DSTs. The concern appears to be that many of these measures are largely uncoordinated. In an African context, however, some attempt at coordination is being attempted. The Africa Tax Administration Forum (ATAF) recommends that African countries impose a DST of between 1% and 3% of gross turnover.63 It further recommends that such a tax is to cover digital services revenue derived either directly or indirectly from a country.64 Finally, ATAF suggests a de minimis threshold so as to exclude smaller service providers.65 Should the EAC follow the ATAF’s recommendations, this would be part of a coordinated approach. Secondly, concern has been expressed that a DST would be ineffective in that the taxpayer may ultimately pass the tax burden on to customers.66 Arguments like this are familiar and have been raised against the introduction of source taxes in general.67 It is argued that, in such event, customers have a choice whether to accept the charges or contract with other suppliers who did not impose such charges. The imposition of additional charges should be treated as a contractual concern.68 However, in the case of monopolies where customers have little choice, governments may step in with laws to protect its market—similar to the measures taken by Australia.69 The next question is whether a DST may be considered fair. The tax would achieve equality if it is imposed on both residents and non-residents operating within the digital arena. In terms of the broader discussion about whether DSTs could be used to facilitate inter-nation equity, it is argued that it may well be a movement in that direction. Such a tax would allow source countries to gain greater taxing rights in an international tax system which is currently heavily skewed towards residence countries. Any double taxation arising from such source taxes could be addressed in the same way as other taxes—through the use of DTAs as is further discussed later. A DST would also allow for certainty in that the tax would have to be legislated. Finally, a DST which has been imposed after careful analysis by governments who have also considered the broader effect of such a tax with respect to its interactions with other countries, would necessarily be legitimate.

63

Africa Tax Administration Forum (ATAF), Suggested Approach to Drafting DSTs Legislation (2020).

64 Ibid. 65 Ibid. 66

Ndajiwo, ‘The Taxation of the Digitalized Economy’, 13. e.g., E. C. C. M. Kemmeren, ‘Legal and Economic Principles Support an Origin and Import Neutrality-Based over a Residence and Export-Neutrality-Based Tax Treaty Policy’, in M. Lang et al., eds, Tax Treaties: Building Bridges Between Law and Economics (2010). 68 Titus, ‘How Can the East African Community Guard Against Base Erosion and Profit Shifting While Working Towards Deeper Integration?’, 183. 69 J. Porter, ‘Australia Passes Law Requiring Facebook and Google to Pay for News Content’, The Verge (24 Feb. 2021), available at https://www.theverge.com/2021/2/24/22283777/australia-new-media-bargain ing-code-facebook-google-paying-news. 67 See,

908 Afton Titus Factors which may persuade the EAC to legislate a DST include that such a tax may be considered fair; Kenya already has such a tax; the ATAF provides some level of coordination; and governments would be perceived as doing something to tax the digital economy.

49.4 Treaty Adaptions Should the EAC adopt the above domestic measures, the taxing rights so created would have to be preserved in DTAs. This could be done through the incorporation of provisions such as articles 12A and 12B of the UN Model Treaty. In 2019, the UN produced a report in which it pointed out that the decision to include only the supply side of profit generation as a basis for the allocation of global taxing rights was a policy decision.70 As such, further policy decisions could be made that would incorporate provisions such as article 12B, which would allow for source countries to levy a withholding tax on payments made for automated digital services either on a gross or net basis.71 The article has no threshold requirements and thus a country could impose the withholding tax without considering the amount of turnover generated within its borders or even the size of the MNE.72 The call for the incorporation of article 12B into DTAs has been echoed by the UN Financial Accountability, Transparency and Integrity Panel.73 The article comes not long after the introduction of article 12A into the UN Model Treaty. Article 12A enables source countries to levy a withholding tax on payments for the receipt of technical services and made to non-resident service providers. Like article 12B, there is no requirement that the non-resident service provider has a physical presence in the source country. It is envisaged that article 12B will not apply in instances where article 12A applies. Both articles have been proclaimed as potential solutions to the challenges of taxing the digital economy.74 As the newest instalment, article 12B does represent a more direct attempt to capture the profits arising from the digital economy. The advantages these articles present to developing countries are that they constitute a simple mechanism by which to grant all market jurisdictions the opportunity to tax

70 United Nations, 21st Session of the Committee of Experts on International Cooperation in Tax Matters (2020), available at https://www.un.org/development/desa/financing/events/21stsession- committee-experts-international-cooperation-tax-matters. 71 Ibid., 14. 72 Astuti, ‘Three Approaches to Taxing Income from the Digital Economy’, 726. 73 United Nations, Financial Integrity for Sustainable Development: Report of the High Level Panel on International Financial Accountability, Transparency and Integrity for Achieving the 2030 Agenda (2021), 24. 74 See M. T. Malan, ‘New Article 12A of the UN Model Regarding Fees for Technical Services Ahead of Its Time Or A Step Too Far?’, Bulletin for International Taxation 73 (2019), 58, 68; Astuti, ‘Three Approaches to Taxing Income from the Digital Economy’, 726.

The Perspective of the EAC on International Tax Law 909 the income from automated digital services;75 they add no administrative costs for revenue authorities while also not increasing the compliance burden of taxpayers;76 they cause the least amount of disruption to established practices;77 and finally they allow countries to maintain tax sovereignty.78 However, these articles have been met with some criticism. Sixdorf and Leitsch, for instance, argue that article 12A seems to draw an arbitrary distinction between technical and other services in that the authors find it difficult to think of a service that does not require some degree of specialized knowledge or skill to enable its delivery.79 Moreover, they argue that article 12A distorts international trade in that it only encourages service providers to pass on the tax to the consumer.80 Finally, they argue that the application of article 12A represents a violation of the principle of non-discrimination between local and foreign service providers.81 Malan further expresses concern that article 12A does not have any thresholds.82 The issue here is that smaller service providers or those who have only an insignificant interaction with the source country may be subject to the withholding tax.83 This could mean that the withholding tax could exceed the service provider’s profit margin in that country.84 Similar arguments have been raised in respect of article 12B. Astuti, for instance, notes that the application of article 12B to small and medium enterprises without the use of some sort of threshold may place such smaller enterprises at risk.85 Moreover, the gross basis of taxation may result in the over-taxation of some MNEs.86 Finally, the bilateral nature of a possible inclusion of the proposed provision in a DTA may be costly for developing countries.87 Perhaps the most compelling argument against the incorporation of article 12B is Moreno’s point that article 12B is unnecessary given that article 12A would already adequately cover automated digital services.88 One way for developing countries to ensure that article 12A does in fact do so, Moreno recommends, is for developing countries 75

Astuti, ibid.

76 Ibid.

77 S. Greil and T. Eisgruber, ‘Taxing the Digital Economy: A Case Study on the Unified Approach’, Intertax 49 (2021), 53, 66. 78 Ibid. 79 F. Sixdorf and S. Leitsch, ‘Taxation of Technical Services under the New Article 12A of the UN Model—Improved Taxation Or A Step in the Wrong Direction?’, Bulletin for International Taxation 57 (2017), 234, 240. 80 Ibid. 81 Ibid. 82 Malan, ‘New Article 12A of the UN Model Regarding Fees for Technical Services Ahead of Its Time Or A Step Too Far?’, 62. 83 Ibid. 84 Ibid. 85 Astuti, ‘Three Approaches to Taxing Income from the Digital Economy’, 727. 86 Ibid. 87 Ibid. 88 A. B. Moreno, ‘Because Not Always B Comes After A: Critical Reflections on the New Article 12B of the UN Model on Automated Digital Services’, World Tax Journal, 13 (2021), 501.

910 Afton Titus to clarify in their domestic legislation that technical services include automated digital services.89 This would then do away with the needs to renegotiate DTAs so as to include an article 12B-like provision.90 On an evaluation of these point, it is noted that some of the EAC partner states have already successfully included article 12A-like provisions in their concluded DTAs. For instance, In 2013, Rwanda included an article 12A-like provision in its renegotiated DTA with Mauritius.91 Therefore, it is recommended that the other EAC partner states follow its lead. Moreover, given the incorporation of article 12A-like provisions in the DTAs of some EAC partner states, the most logical course would be for the EAC to follow Moreno’s recommendation to clarify in their domestic legislation that technical fees include automated digital services. It may also serve the EAC well to further follow Moreno’s recommendations on a multilateral level by forming a minority position in the Commentary on article 12B by stating that article 12B is unnecessary given the scope of article 12A.92 In the author’s view, this course of action presents to the EAC the easiest and perhaps the most effective means of protecting their source right of taxation.

49.5 OECD Proposals As a last resort, the EAC may wish to consider the OECD’s proposals. The drive to tax the profits of the digital economy has occupied the OECD since the introduction of Action 1 of the OECD/G20 BEPS Project in 2015.93 The latest instalment of the OECD’s efforts is its Pillar One and Pillar Two proposals.94 On the face of it, Pillar One seeks to extend the ability of market jurisdictions to tax MNEs with a global turnover over €20 billion and with profitability above 10%.95 Moreover, the market jurisdictions within which such MNEs operate would only be entitled to tax the MNE’s ‘residual profits’,96 known as ‘Amount A’, once that MNE earns at least €1 million in revenue from that market jurisdiction, with some allowances made for smaller jurisdictions.97 This is done primarily through the use of a revenue-based 89

Ibid., 531.

90 Ibid.

91 Agreement Between the Government of the Republic of Mauritius and the Government of the Republic of Rwanda for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income (20 Apr. 2013). 92 Ibid., 531. 93 OECD/G20, Addressing the Tax Challenges of the Digital Economy— Action 1: 2015 Final Report (2015). 94 OECD/G20, Statement on a Two- Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy (8 Oct. 2021). 95 Ibid., 1. 96 This is defined as: ‘[P]‌rofit in excess of 10% of revenue’ in OECD/G20, Statement on a Two-Pillar Solution, 2. 97 Ibid., 1.

The Perspective of the EAC on International Tax Law 911 allocation key informed by the location of the users or consumers of the MNE’s goods or services.98 This is accompanied by an ‘Amount B’ which is calculated according to the arm’s-length principle and is to apply to in-country baseline marketing and distribution activities.99 The OECD states that Amount B is to be simplified in the near future.100 Collectively, this approach is called the ‘Unified Approach’. Meanwhile Pillar Two seeks to introduce a global minimum tax through two main rules—an income inclusion rule (IIR) and a tax on base eroding payments—as mechanisms to combat BEPS. In the short time that these proposals have been in circulation, they have generated a large amount of comment. In respect of Pillar One, Dourado has noted that the recognition of the right of market jurisdictions to be included in the allocation of taxing rights is a step in the right direction.101 Moreover, Guedes comments that the OECD’s push to reach a consensus-based international agreement should be commended.102 Elliffe also sees the OECD’s proposals in a positive light and argues that it is an attempt to fix a broken international tax system through the use of multilateral processes to ensure more effective source taxation and the reduction of tax competition globally.103 Overwhelmingly, however, the pillars have been subject to criticism. Brauner comments that the pillars share the same lack of focus as the entire BEPS Project.104 Brauner further comments that the proposal lacks justification for its introduction of a new nexus.105 Such justification has had to come from other quarters.106 These concerns are echoed by Dourado who comments that a fundamental flaw of Pillar One is that it arranges the profits into a hierarchical order based on the role of the country involved rather than recognizing a global profit that is to be distributed amongst the countries involved in its creation.107 And for all the complexity of the Unified Approach, market countries are allocated only a small portion of the related profits.108 In fact, Dourado surmises that the Unified Approach operates more as a means by which to defend supply-side countries from the claims of market countries.109 This is echoed in Shay’s

98

Ibid., 2. Ibid., 3. 100 Ibid. 101 A. P. Dourado, ‘The OECD Report on Pillar I Blueprint and Article 12B in the UN Report’, Intertax 49 (2021), 3. 102 E. F. D. L. Guedes, ‘Tax Challenges of the Digital Economy: An Evaluation of the New OECD Nexus Rule Based on Revenue Thresholds’, Bulletin for International Taxation 76 (2022), 197. 103 C. Elliffe, ‘The Brave (and Uncertain) New World of International Taxation Under the 2020s Compromise’, World Tax Journal 14 (2022), 237, 250. 104 Y. Brauner, ‘Editorial: Lost in Construction: What is the Direction of the Work on the Taxation of the Digital Economy?’, Intertax 48 (2020), 270, 272. 105 Ibid., 271. 106 See X. Li, ‘A Potential Legal Rationale for Taxing Rights of Market Jurisdictions’, World Tax Journal 13 (2021), (published online) who argues that in terms of a revised benefit principle, market jurisdiction could be justified by deeming users as a conduit through which market jurisdictions provide goods. 107 Ibid., 7. 108 Ibid., 6. 109 Ibid., 8. 99

912 Afton Titus calculations where he notes that Amount A would have only 3.75% of corporate profits allocated to market jurisdictions.110 Similar concerns have been raised regarding Pillar Two. Brauner comments that Pillar Two is nothing more than a minimum tax levied in favour of residence countries.111 As such, Schreiber notes that this represents a major shift in the OECD’s long-held position that tax rates fall exclusively within the purview of sovereign states.112 Some of these concerns have been allayed through the release of the OECD’s Pillar Two model rules. Most notably, these rules would force residence countries to recognize the source country’s qualified minimum domestic minimum top-up tax (QDMTT) in that residence countries would not be able to use the IIR when QDMTTs are present.113 Despite this relatively positive development for source countries, the criticisms of Pillar Two nonetheless remain. For instance, some commentators have questioned the value of Pillar Two. Hey, for instance, argues that Pillar Two is nothing more than a more technical controlled foreign company (CFC) rule.114 Similarly, Arnold argues that CFC rules would in fact be more effective at combating profit shifting from high-to low-tax jurisdictions.115 Others disagree with Pillar Two’s premise that all tax competition is bad and should be eliminated. Dagan, for instance, argues that the drive to completely eliminate tax competition would have adverse effects on developing countries which may yet derive benefits from using tax incentives.116 On an overall evaluation of the effectiveness of the proposed pillars, commentators are sceptical about its usefulness. The pillars have been described as distinctly lacking in principles and driven by political and pragmatic factors;117 severely limiting countries from designing a tax base that is different from that chosen in the pillars;118 and incoherent in its tax policy formulation.119

110

S. E. Shay, ‘The Deceptive Allure of Taxing “Residual Profits”’, Bulletin for International Taxation 75 (2021), 527, 532. 111 Brauner, ‘Editorial: Lost in Construction, 271. 112 U. Schreiber, ‘Remarks on the Future Prospects of the OECD/G20 Programme of Work—Profit Allocation (Pillar One) and Minimum Taxation (Pillar Two)’, Bulletin for International Taxation 74 (2020), 338. 113 M. Devereux, J. Vella, and H. Wardell- Burrus, ‘Pillar 2: Rule Order, Incentives, and Tax Competition’, Oxford University Centre for Business Taxation, Policy Brief 3 (2022). 114 J. Hey, ‘The 2020 Pillar Two Blueprint: What can the GLoBE Income Inclusion Rule Do that CFC Legislation Can’t Do?’, Intertax 49 (2021), 7, 13. 115 B. Arnold, ‘The Evolution of Controlled Foreign Company Rules and Beyond’, Bulletin for International Taxation 73 (2019), 631. 116 T. Dagan, International Tax Policy Between Competition and Cooperation (2017), 4. 117 Brauner, ‘Editorial: Lost in Construction’, 271; A. J. Martín Jimenez, ‘Value Creation: A Guiding Light for the Interpretation of Tax Treaties’, Bulletin for International Taxation 74 (2020), 197, 214. 118 M. P. Devereux et al., ‘The OECD Global Anti-Base Erosion Proposal’, Oxford University Centre for Business Taxation (2020), 20. 119 Martín Jimenez, ‘Value Creation: A Guiding Light for the Interpretation of Tax Treaties’, 206.

The Perspective of the EAC on International Tax Law 913 Despite this, the EAC partner states are expected to provide their cooperation in support of the OECD’s proposals. This chapter argues that providing such cooperation at this stage is not in the best interest of the EAC partner states. First, all of the EAC partner states would be classified as ‘user’ or ‘market’ countries with their large collective consumer market. This means that for most of the EAC partner states, aside from Kenya, they would be relying mainly on Amount A to garner revenue from the MNEs operating within their jurisdictions. Being solely reliant on Amount A when the MNEs may have no physical presence in a particular jurisdiction is a risky prospect considering that there is no guarantee that the calculations would even produce a residual profit at all. It has been noted that the actual profits that end up being allocated to market jurisdictions is likely to be miniscule.120 This likelihood is further confirmed by the OECD itself which indicated that only a ‘modest’ gain would be achieved in the raising of global corporate tax revenue.121 Secondly, for countries like Kenya who have many MNEs operating within their jurisdiction, the adoption of Pillar One would mean that an additional layer of calculations and administration is to be added on top of its already complex transfer pricing processes. The prospect of making further significant demands on revenue authorities in the EAC to administer the proposed Unified Approach seems unreasonable and pointless. This is especially true considering that Kenya would have to give up its DST in order to adopt Pillar One. Thirdly, all of the EAC partner states have special economic zones in terms of which tax incentives are offered to encourage foreign direct investment.122 If the EAC partner states implement Pillar Two, the only way for their tax incentives to be preserved is if they were to legislate QDMTTs and, in so doing, protect their tax sovereignty by preventing the IIR from being used at their expense.123 The author has argued elsewhere that it would be best for the EAC to adopt a regional approach should they take this step.124 Ultimately, the OECD’s introduction of excessively complex rules only serves to mask the fact that decisions regarding the allocation of taxing rights is ultimately a question of fairness.125 It is doubtful whether the OECD’s approach is equitable when it preserves 120 Astuti,

‘Three Approaches to Taxing Income from the Digital Economy’, 725; J. Li, ‘The Legal Challenges of Creating a Global Tax Regime with the OECD Pillar One Blueprint’, Bulletin for International Taxation 75 (2021), (online). 121 OECD, Tax Challenges Arising from Digitalisation— Report on Pillar One Blueprint (2020), para. 154. 122 Burundi: Establishment Law No. 1 of 2017 in respect of Special Economic Zones; Kenya: Special Economic Zone Act 16 of 2015; Rwanda: Law No. 5/2011 of 21/3/2011, Law Regulating Special Economic Zones in Rwanda; Tanzania: Special Economic Zone Act 2006 (as amended); Uganda: Free Zones Act, 2014 and The Free Zones (General) Regulations, 2016, South Sudan: Export Processing Zones are under development (see EAC, ‘EAC Investment Guide’, available at https://investment-guide.eac.int/index. php/operating-environment/export-processing-zones. 123 Devereaux et al., ‘Pillar 2: Rule Order, Incentives, and Tax Competition’. 124 A. Titus, ‘Pillar Two and African Countries: What Should Their Response Be? The Case for a Regional One’, Intertax 50 (2022), 711. 125 Schreiber, ‘Remarks on the Future Prospects of the OECD/G20 Programme of Work’, 339.

914 Afton Titus the status quo of portioning the lion’s share of global corporate profits to residence countries while leaving source countries with very little unless source countries implement QDMTTs and as the OECD directs. This becomes even more questionable when one considers that the legitimacy of the OECD as an institution to make proposals on global tax policy has long been at issue.126 Consequently, it is recommended that the EAC partner states not support the OECD proposals in their current form. The OECD proposals are not designed with African developing countries in mind. Until they are, it is recommended that the EAC turn to other mechanisms by which to effectively tax the digital economy, such as a DST as discussed earlier.

49.6 Conclusion The world has changed. It has become apparent that big data companies and the country that supports them are more vulnerable than they were before. And far less capable of acting in retaliation than they were in the past. Cui argues that these vulnerabilities have translated into a shift in global power dynamics in favour of source countries.127 This is because of the nature of digital service businesses and a consequence of globalization. The defining characteristic of many digital businesses is the low cost involved in their operations. This means that the majority of the value is related to market power— something which is purely source-based.128 This fact alone should justify stronger source rules.129 The nature of digital businesses also means that it needs markets. Moreover, the countries that provide those markets are no longer isolated voices easy to ignore. Globalization, increased channels of communication, and the interconnectivity of the world now mean that if a group of countries act in concert to demand greater taxing rights, it is unlikely that the digital service provider will be able to recoup the loss of market access by going elsewhere.130 The world is only so big. This is illustrated by the Google/Australia dispute where Google eventually withdrew its threat that it would take its business elsewhere.131 This means that there is more scope for regional blocs like the EAC to make better, more compelling policy decisions. 126

I. J. Mosquera Valderrama, ‘Legitimacy and the Making of International Tax Law: The Challenges of Multilateralism’, World Tax Journal 7 (2015), (published online); P. Pistone, ‘Coordinating the Action of Regional and Global Players during the Shift from Bilateralism to Multilateralism in International Tax Law’, World Tax Journal 6 (2014), 1; J. Bossuyt and L. Brosens, ‘Legitimacy in International Tax Law- Making: Can the OECD Remain the Guardian of Open Tax Norms?’, World Tax Journal 12 (2020), 313. 127 Cui, ‘What is Unilateralism in International Taxation?’, 264. 128 Ibid. 129 Ibid. 130 Ibid. 131 Porter, ‘Australia Passes Law Requiring Facebook and Google to Pay for News Content’.

The Perspective of the EAC on International Tax Law 915 In considering which approach to adopt in the drive to tax the digital economy, it has been noted that both the OECD/G20 and the UN are moving in the same direction in recognizing that greater taxing rights should be allocated to market jurisdictions.132 The same may be said of DSTs and unitary taxation with formulary apportionment. However, the author argues that while many of the approaches may be considered fair from a legal perspective, the OECD’s proposals require some serious adjustments before the same can be said of them. Consequently, this chapter recommends that the EAC does not follow the OECD’s proposed approach. This is because the proposals are overly complex, demand a great deal of administrative resources, and would require the EAC partner states to surrender some tax sovereignty. The trade-off is simply not worth it. Instead, this chapter recommends that the EAC consider adopting a unitary taxation with a formulary apportionment corporate income tax system and, in the interim, a DST in line with the ATAF’s recommendations. Also, in order to protect its domestic taxing rights, it is recommended that the EAC push for the incorporation of an article 12A-like provision in their DTAs. This is especially because some of the EAC partner states have already been successful in including such provisions in their DTAs. While this would be a coordinated regional approach, it is still not a global consensus. Notwithstanding this, it may well be the tipping point needed to bring about more tangible fairness in international tax relations.

132

Dourado, ‘The OECD Report on Pillar I Blueprint and Article 12B in the UN Report’.

Chapter 50

T he Japanese Pe rspe c t i v e on In ternationa l Tax L aw Masao Yoshimura

50.1 Introduction The design of an international taxation system, especially a tax treaty policy, depends on a country’s environment, including its economy, the size of its imports and exports, and the relationship it has with its trading partners. After the Second World War, Japan experienced rapid economic growth and became a mature capital-exporting country in the 1980s. The change in Japan’s economic environment resulted in the evolution of its international tax law policy. This chapter will focus on three issues related to Japan’s international taxation rules and will determine the environmental factors that influenced the transformations in Japan’s policy on international taxation. These significant environmental changes, which occurred in the twenty-first century, will be examined against the experiences of other countries (including the globalizing value chain and international profit shifting) and the policy changes that Japan has made in this regard will be identified. What changed Japan’s policy? (1) How did Japan respond to offshore outsourcing by Japanese multinational enterprises (MNEs)? (2) How did Japanese companies and the government react to the Base Erosion and Profit Shifting (BEPS) Project? In responding to these questions, this chapter is organized as follows. First, a brief description of the characteristics of Japan’s international tax rules will be made. Secondly, an analysis will be conducted on how Japan moved to a territorial system in 2009 and changed its treaty policy to tighten restrictions on withholding tax in source countries in response to corporate offshoring and globalized value chains. In the 2000s, the Japanese government improved its administrative enforcement and strengthened counter rules

918 Masao Yoshimura to tackle cross-border profit shifting by multinational corporations (MNCs). However, given conservative tax strategies adopted by Japanese firms, the government introduced some exceptions to mitigate their compliance costs. Thus, the next section of the chapter will examine these exceptions. Fourthly, a review of relevant research on the perceived absence of aggressive tax planning practices by Japanese firms will be conducted. Finally, the conclusions will be presented.

50.2 Characteristics of Japan’s International Taxation System 50.2.1 Strong Influence of the US Rules In 1951, Japan signed the Treaty of Peace, which legally ended the state of war. The treaty re-ushered Japan into the international community and prepared Japanese companies for their entry into overseas markets and the acceptance of foreign direct investment. Following the Second World War, Japan developed a legal framework for economic activities, including taxation, with references to US law. It, anticipated a rapid increase in outbound and inbound transactions and introduced taxation rules for international transactions drawing from US law.1 There are two methods for eliminating double taxation on the activities of foreign business—the exemption method and the credit method. To eliminate double taxation, Japan introduced a foreign tax credit system in 1953 instead of the exemption system adopted by many European countries. Japan also signed its first tax treaty with the USA in 1954,2 and agreed on provisions that would prevent double taxation and limit the source tax rate. Japan’s treaty policy was modelled after this first tax treaty until 1971 when Japan and the USA wholly revised the treaty.

50.2.2 Close Coordination with the OECD In 1964, Japan joined the OECD and participated in harmonizing international taxation rules. Since then, Japan’s treaty policy has been strongly influenced by the OECD Model Tax Convention.

1 Y.

Masui, ‘International Taxation in Japan: A Historical Overview’, Tax Notes International 21 (18 Dec. 2000), 2813. 2 Japan was the sixteenth country to conclude a treaty on the avoidance of double taxation and the prevention of fiscal evasion with respect to taxes on income with the USA. M. B. Carroll, ‘Tax Conventions with Germany and Japan’, Taxes 32 (1954) 897, 901.

The Japanese Perspective on International Tax Law 919 Learning from the US experience and the OECD discussions, Japan developed rules to tackle international tax avoidance in the 1970s.3 As the USA began to apply transfer pricing rules to Japanese companies aggressively, Japan responded by enacting its transfer pricing tax system in 1986. Japan also established thin capitalization rules in 1992 based on the OECD report4 and reviewed the foreign tax credit system which had supported the overseas activities of Japanese companies and included provisions limiting tax credit from 1988.5 Recently, the Japanese government has actively promoted the BEPS Project that was initiated in 2013.6 Japanese tax authorities and taxpayers considered the minimization of Japanese tax liabilities by US IT companies as unfair. Japanese multinationals also believed that they were at a competitive disadvantage as they did not engage in aggressive tax planning and thus incurred higher tax rates than MNCs headquartered in other countries. For these reasons, Japan quickly implemented most recommendations contained in the BEPS Final Reports. Japan also expected to lead other countries in the implementation of the BEPS proposals, thereby curbing aggressive tax planning, increasing tax revenues, and levelling the playing field.

50.2.3 Economic Stagnation and Tax Policy Changes After the period of rapid economic growth, Japan suffered prolonged recession and stagnation around 2000. On the one hand, it relaxed the rules on controlling foreign direct investment to reform the financial market, hoping that foreign MNCs would increase their direct investment in the country and contribute to the growth of innovative industries. On the other hand, the expansion of employment and consumption in Asia, including China, led Japanese MNCs to develop offshoring production capacity to those countries.

3 According to Professor Nakazato, the Japanese government shifted from its previous export- stimulus policy and adopted a policy of strengthening taxpayer compliance to curb tax avoidance. M. Nakazato, ‘Internationalization of Japan’s International Tax Law’, Japanese Annual of International Law 35 (1992), 78, 88. E.g. Japan introduced CFC rules in 1978, following the US Subpart F rules. For more information on the history of the CFC rules up to the 2009 reform, see A. Asatsuma, ‘The Controlled Foreign Corporation Legislation’, Asia-Pacific Tax Bulletin 11/3 (2005), 363. 4 OECD, Issues in International Taxation No. 2— Thin Capitalisation and Taxation of Entertainers, Artistes and Sportsmen (Paris: OECD Publishing, 1987). 5 See, e.g., S. Sultan, ‘Japan Blocks Tax Credit Loopholes’, International Tax Review 5 (1994), 40 . The basic structure of Japan’s international taxation rules was similar to that of the USA. However, based on a comparison of tax burdens on FDI, Slemrod estimated that Japan was more generous in taxing foreign-source income. J. Slemrod, ‘Competitive Advantage and the Optimal Tax Treatment of the Foreign-Source Income of Multinationals: The Case of the US and Japan’, American Journal of Tax Policy 9 (1991), 113. 6 At the time, Mr Masatsugu Asakawa, who held a key position in the Ministry of Finance, was the chairman of the OECD Finance Committee. The first meeting of the Inclusive Framework on BEPS was held in Kyoto, Japan, in June 2016 under his chairmanship.

920 Masao Yoshimura Considering Japan’s aging population, the change significantly impacted the overall tax system, including corporate taxes. Japan expanded tax revenue from its consumption tax (value-added tax), introduced in 1989 to finance social security benefits, and gradually raised the tax rate, which is now at 10%. However, the enormous budget deficit resulted in a corporate tax rate which remains higher than the OECD average, which cannot compete with its Asian trading partners such as China and Korea. Tax reform was necessary for maintaining tax revenues and strengthening the competitiveness of Japanese companies in foreign markets. Therefore, Japan dramatically overhauled its international tax rules and treaty policies in the twenty-first century. First, it moved from a worldwide tax system to a territorial system and introduced the 2009 Act providing an exemption for dividends paid by foreign subsidiaries. The government intended that the repatriated dividends would be utilized for research and development in Japan. Secondly, in 2003, Japan made significant revisions to its tax treaty with the USA, representing a landmark change in treaty policy. Previously, Japan stuck to its policy of reserving the taxing right of the source country where it collected taxes from royalties paid to licensors in the other country, especially in the USA. However, as Japanese companies had been offshoring and receiving more patent fees and dividends from their US subsidiaries, Japan decided to encourage Japanese companies to receive royalty payments from their US subsidiaries by agreeing to waive source tax on royalties. Instead, Japan negotiated and won with the US limitations on withholding tax imposed on investment income such as interest and dividends in order to increase the cash flow received by Japanese headquarters.

50.3 Promoting Offshoring and Value Chain Globalizing 50.3.1 2003 Tax Treaty Revision The tax treaty between Japan and the USA is of exceptional significance to the Japanese government.7 This is not only due to Japan’s economic relationship with the USA, its principal trading and direct investment partner, but also because it will lead to a comprehensive review of Japan’s general treaty policy.

7 M. Asakawa, ‘The New Japan–US Tax Treaty’, Asia-Pacific Tax Bulletin 10(4) (2004), 205–209; G. M. Thomas, ‘Opportunities and Challenges in New International Tax Era in Japan’, Tax Notes International 34 (14 June 2004), 1161–1173; Y. Masui, ‘Overhaul of the Japan–US Tax Treaty in 2003’, The Japanese Annual of International Law 49 (2006), 55–70; P. A. Fuller, ‘The Japan–US Income Tax Treaty: Signaling New Norms, Inspiring Reforms, or Just Tweaking Anachronisms in International Tax Policy?’, The International Lawyer 40 (2006), 773–908.

The Japanese Perspective on International Tax Law 921 In 1954, the first tax treaty was signed after Japan had just rejoined the international community and needed inbound investment to develop its economy. Through negotiations with the USA, the Japanese government recognized the need to clarify source rules and the permanent establishment (PE) concept in treaty provisions and domestic law.8 This first treaty served as a model for the series of tax treaties that followed.9 In 1971, following its rapid economic growth and membership of the OECD, Japan revised its tax treaty with the USA with reference to the OECD Model Tax Convention.10 For example, Japan, which had previously adopted a force-of-attraction rule, abandoned the rule and accepted a rule conforming to the OECD Model Tax Convention, where only profits directly attributable to a PE were subject to tax as business income. However, since Japanese companies imported a significant amount of technology, the Japanese Ministry of Finance refused to exempt withholding tax on royalties from intangible assets and negotiated maintenance of the 10% tax rate on royalties paid to US companies. Although the economic relationship between Japan and the USA has been further strengthened, the tax treaty has remained untouched for many years until 2003. The USA has nursed the intention of exempting royalties from withholding taxes, and both Japanese and US industries have urged their respective governments to revise the treaty.11 However, the Japanese government was reluctant to change the treaty as that would have resulted in a loss of tax revenue. Japanese companies’ total royalties to US companies were larger than those received by Japanese companies from US companies. This imbalance of royalty payments also existed globally. However, the situation has changed. First, under the alternative minimum tax (AMT) implemented in 1986, the alternative minimum foreign tax credit was generally limited in accordance with US law to 90% of AMT liability. Many US MNCs could not fully benefit from foreign tax credits and were subjected to double taxation due to Japanese withholding tax and US AMT. See Table 50.1 for withholding tax rates on investment income. Secondly, Japan needed to redesign its treaty policies to consider the dividends and royalties received by Japanese parent companies from profitable foreign subsidiaries, as Japanese companies had been offshoring more of their manufacturing capacity since the 1990s. In addition, the Japanese government assumed that Japan’s future savings rate would decline due to its aging population. It believed that Japan would need more foreign investment and that it would benefit in the long term by reducing or exempting withholding taxes on investment income.12 8

Nakazato, ‘Internationalization of Japan’s International Tax Law’, Japanese Annual of International Law 35, 81. 9 Y. Masui, ‘The Influence of the 1954 Japan–United States Income Tax Treaty on the Development of Japan’s International Tax Policy’, Bulletin for International Taxation 66 (2012), 243–251. 10 Y. Komatsu, ‘The 1971 Income Tax Convention between Japan and the United States of America’, Law Japan 7 (1974), 51. 11 See, e.g., Japan– US Business Council, Joint Statement Joint Statement 37th Annual Japan—US Business Conference (11 July 2000). 12 Asakawa, ‘The New Japan–US Tax Treaty’, IBFD (2004), 206; Masui, ‘Overhaul of the Japan–US Tax Treaty in 2003’, The Japanese Annual of International Law 49 (2006), 55–70.

922 Masao Yoshimura Table 50.1 Withholding tax rates on investment income 1971 Treaty

2003 Treaty

General

15%

10%

parent–subsidiary

10%

Exempt or 5%

Interest

10%

General: 10% Financial institutions: Exempt

Royalties

10%

Exempt

Dividends

Source: Reproduced from Yoshihiro Masui, ‘Overhaul of the Japan–US Tax Treaty in 2003’, The Japanese Annual of International Law (2006) 49, 60 © 2006, HeinOnline.

The following two items are particularly important in the 2003 treaty. First, Japan strengthened the restrictions on source-country taxation of investment income, including royalties. As previously noted, Japan made concessions on royalty exemptions and instead won tax exemptions for interest received by financial institutions and dividends paid between parent and subsidiary companies.13 The amount of investment income that Japan received was larger than that received by the USA, even at that time. Thirdly, because the 2003 treaty allowed greater scope for tax exemption for investment income, it introduced the limitation-on-benefits (LOB) provision to prevent tax avoidance (treaty shopping) through conduit companies. Japan accepted the LOB provision developed by the USA; the rule was stipulated in subsequent treaties.14 As expected by the Japanese government,15 the amount of royalties received by Japanese MNCs continued to increase, with royalty receipts exceeding payments worldwide in 2003 and receipts from the USA even exceeding payments in 2013. This fact shows that as Japanese companies increased their production capacity abroad (e.g. car- manufacturing plants), their foreign subsidiaries increased their royalty payments to their parent companies according to their production capacity. The facts presented in the previous paragraph confirm that parent companies in Japan own and control significant intangible assets and receive large amounts of royalties from their subsidiaries without aggressive tax planning. It is said that Japanese MNCs tend to avoid international tax avoidance (see Section 50.5) and the forementioned may be the reason behind that trend.

13

Ibid., 66. the time, the Japanese tax authorities began to intensify investigations to prevent treaty shopping. See Thomas, ‘Opportunities and Challenges in New International Tax Era in Japan’, Tax Notes. Int'l 34 (2004), 1161. 15 Ibid., 1163. 14 At

The Japanese Perspective on International Tax Law 923

50.3.2 Shift to Territorial System Japan, like the USA, has adopted a worldwide system and has considered the global income in MNCs to be subject to its tax jurisdiction. Foreign subsidiaries were subject to taxation in the host country, and dividends repatriated from subsidiaries to their parent companies in Japan were subject to Japanese corporate taxation. Japanese corporate tax law allowed an indirect tax credit for the host country tax imposed on their profits, from which the subsidiaries paid dividends to their parent companies to avoid double taxation. However, after the 2009 tax reform, Japan exempted 95% of dividends received from foreign subsidiaries from taxation where a domestic corporation owned 25% or more of the shares.16 Since dividends are exempted from Japanese taxation under the territorial system, the local withholding tax on such dividends is excluded from the (direct) foreign tax credit. Thus, it is crucially important for Japanese MNCs that the taxation in a source country is restricted under the tax treaty with the host country. Due to changes in Japanese companies’ economic environment and global trends, Japan has adopted a territorial system. Interestingly, the UK also implemented reforms to adopt a territorial system in 2009.17 In 2008, a study group from the Ministry of Economy, Trade, and Industry (METI) identified problems with the existing worldwide system.18 It analysed the expected effects if Japan were to move to a territorial system.19 The report concluded that Japanese companies operating overseas should return the profits of their foreign subsidiaries to Japan whenever needed, irrespective of the amount and regardless of the tax burden in Japan.20 16 T. Miyatake, ‘Japan’s Foreign Subsidiary Dividends Exclusion’, Proceedings paper at University of California at Berkeley School of Law, Institute for Legal Research & Robbins Religious and Civil Law Collection, Sho Sato Conference (9–10 Mar. 2009), 1–16; Y. Masui, ‘Taxation of Foreign Subsidiaries: Japan’s Tax Reform 2009/10’, Bulletin for International Taxation 64 (2010), 242–248; T. Morotomi, ‘Japan’s Shift to Territoriality in 2009 and the Recent Corporate Tax Reform: A Japan–United States Comparison of Taxing Income from Multinationals’, Pittsburg Tax Review 14 (2017), 173–217. 17 For the impact of the UK debate on the revision of international taxation rules, see M. Herzfeld and M. Honda, ‘Moving to Territorial: Lessons from Japan’, Tax Notes International 89 (8 Jan. 2018), 119–123. 18 Subcommittee on International Taxation, ‘Report on the Repatriation of Japanese Companies’ Overseas Profits’ (‘Waga Kuni Kigyo no Kaigai Rieki no Shikin Kanryu ni tsuite’) (2008) (in Japanese). 19 The study group suggested that Japan consider the following two facts as an essential basis for tax reform: (1) Japanese manufacturers with overseas operations were increasing their share of overseas production, which rose from 24.2% in 2000 to 31.4% in 2007. Similarly, the amount of profit made by foreign subsidiaries of Japanese companies quadrupled between 2001 and 2006. Since foreign markets would probably grow more in strength than the mature and aging Japanese market, Japan needed to ensure that the tax rules did not hinder the competitiveness of Japanese companies in foreign markets. (2) The study group believed that most of the profits earned by foreign subsidiaries were retained by those subsidiaries and not repatriated to Japan. In 2001, the annual amount of retained profits was 137.8 billion yen, but the subsidiaries kept about 2 trillion yen every year from then onwards. The METI estimated that the resulting balance retained and reinvested overseas was 17 trillion yen in 2006. 20 Considering corporate decision- making on dividend repatriation, the fundamental problem is that Japan’s corporate tax rate is higher than that of other countries. The Japanese government, which

924 Masao Yoshimura The study group believed that Japanese companies should use repatriated dividends to expand their R&D functions and strengthen their intangible asset management and financial management,21 which would help improve long-term domestic employment. A questionnaire by the METI showed that CEOs expected that the funds returned to the headquarters would be used mainly for positive domestic investment, such as capital investment and R&D. After the tax reform, some economists conducted empirical studies to measure the effects of Japan’s policy changes. According to Tajika et al., an increase in repatriated dividends was observed between 2008 and 2009 when there was strong demand for funds at the Japanese headquarters (e.g. capital expenditures, repayment of loans, and dividends to shareholders).22 They argue that this amendment paved the way for the parent companies to effectively use the retained earnings of foreign subsidiaries which the previous tax rules had blocked. Hasegawa & Kiyota found that dividends from subsidiaries with high retained earnings increased according to the intended purpose of the amendment.23 Moreover, subsidiaries became more sensitive to the withholding tax rate at the source to determine the receipt of dividends. However, contrary to prior expectations, they did not find an increase in dividends from subsidiaries in low-tax locations where the additional burden on repatriated dividends would have been more significant under the previous system. Additionally, Feld et al. analysed the impact of taxes on dividend repatriations on foreign mergers and acquisitions using information from 17,907 foreign mergers and acquisitions that occurred among OECD countries between 2004 and 2013.24 They estimated that the transition to the territorial system increased the number of overseas mergers and acquisitions by Japanese firms by 16.1% ($3.1 billion per year in value terms). This study may be evidence that tax reform boosted the competitiveness of Japanese MNCs in overseas markets.

was suffering from a budget deficit, could not provide a solution for lowering the corporate tax rate and decided to maintain Japan’s high corporate tax rate by shifting to a territorial tax system. Masui, ‘Taxation of Foreign Subsidiaries: Japan’s Tax Reform 2009/10’, 242–248. 21 Subcommittee

on International Taxation, ‘Report on the Repatriation of Japanese Companies’ Overseas Profits’, 9. 22 E. Tajika, M. Hotei, and K. Shibata, ‘The Effects of Dividend-Exemption Method on Repatriation of Income from Abroad: The Case of 2009 Japanese Tax Reform (Zeisei to Kaigai Kogaisha no Rieki Sokin: Honsha Shikin Juyo Karamita 2009 Nendo Kaisei no Bunseki)’, Economic Analysis (Keizai Bun- seki) 188 (2014), 68–92 (in Japanese). 23 M. Hasegawa and K. Kiyota, ‘The Effect of Moving to a Territorial Tax System on Profit Repatriation: Evidence from Japan’, Journal of Public Economics 153 (2017), 92–110. 24 L. P. Feld et al., ‘Repatriation Taxes and Outbound M&As’, Journal of Public Economics 139 (2016), 13–27.

The Japanese Perspective on International Tax Law 925

50.4 Tackling Cross-B order Profit Shifting Japan changed its tax policy to support overseas operations, however, the government was also concerned about income shifts to overseas subsidiaries and implemented several reforms to curtail the risk of such transfers. For example, after transitioning to a territorial system, Japan tightened its controlled foreign corporation (CFC) rules and interest-deduction limitation rules. Japanese tax authorities began to intensively investigate whether Japanese companies had any adjustable profit transfers under transfer pricing rules. Even recently, the Japanese government, especially the Ministry of Finance, has been an ardent proponent of the BEPS Project and has faithfully implemented its recommendations.25 In areas where the government needed to revise the rules, such as the CFC rules, profit-stripping provisions, and transfer pricing regulations, Japan reviewed the rules based on the BEPS recommendations and implemented amendments, taking into account the compliance burden on industry associations.

50.4.1 Transfer Pricing Challenges In the 1980s when trade friction with the USA intensified, Japanese companies were targeted by the Internal Revenue Service using transfer pricing rules. In response, Japan also introduced transfer pricing provisions in 1986 and gained the option to apply them to US companies.26 In the 1990s, Japan’s National Tax Agency (NTA) applied the rules to foreign MNC affiliates operating in Japan. Since 2001, Japan has been working on enacting legislation based on the OECD Transfer Pricing Guidelines and it has adopted a list of transfer pricing methods, including the profit method and the transactional net margin method, in addition to the traditional comparative methods.27 The NTA accepted the OECD Transfer Pricing Guidelines (1995 version) and provided administrative guidelines in 2001. By 2000, Japanese companies had

25

T. Miyatake, ‘Recent Japanese Tax Developments’, Tax Notes International 84 (21 Nov. 2016), 783– 788; H. Takahashi, ‘Japan Branch Report on Assessing BEPS: Origins, Standards and Eesponses’, Cahiers de Droit Fiscal International vol. 102A (2017), 451–462; M. Ohno, ‘The Adoption of BEPS in Japan’, in K. Sadiq et al., eds, Tax Design and Administration in a Post-BEPS Era: A Study of Key Reform Measures in 18 Jurisdictions (Birmingham: Fiscal Publications, 2019), 145–159. 26 M. M. Yoshimura, ‘The Tax War between the United States and Japan under Internal Revenue Code 482: Is There a Solution’, Wisconsin International Law Journal 12 (1993), 401; J. H. Guttentag and T. Miyatake, ‘Transfer Pricing: US and Japanese Views’, Tax Notes International 9 (7 Feb. 1994), 375. 27 Since 2011, the tax authorities have chosen transfer pricing methods based on the ‘most appropriate method’ rule following the OECD Transfer Pricing Guidelines (2010 version).

926 Masao Yoshimura rapidly relocated their manufacturing capacity from Japan to other Asian countries and increased imports of end products assembled by their foreign subsidiaries. From 2003, the NTA began aggressive transfer pricing inspections of these offshored manufacturing entities and issued tax assessments for several Japanese multinational corporations in 2004 and 2005. As with other developed countries, it focused on royalties from foreign subsidiaries to Japanese parent companies. A residual profit method was applied to determine the transfer prices of taxpayers, emphasizing the parent companies’ leading role or intangible assets in the manufacturing process or research and development activities. These aggressive investigations by the tax authorities soon faced a counterattack from business groups. First, the METI study group, which included representatives from the industry group, published a transfer pricing report in 2007 and requested the NTA to issue clear and detailed rules on the application of transfer pricing. It also requested the NTA to enhance the advance pricing arrangement (APA) process to ensure tax certainty. Responding to the requests, the NTA rewrote its administrative guidelines on transfer pricing rules, providing more detailed information, and published the Reference Case Studies on Application of Transfer Pricing after conducting a public consultation. The NTA also significantly increased the number of its workforce involved in the APA process, which significantly increased APA processing in Japan. Secondly, the National Tax Tribunal revoked some of those assessments.28 The tribunal required the NTA to specifically investigate the value contributions of each of the relevant parties when applying the profit methods. After all, the NTA had attempted to regain tax revenue by strengthening transfer pricing enforcement against the relocated subsidiaries of Japanese multinationals but had failed to achieve satisfactory results. The NTA overestimated the functions of the head office and could not win the tribunal’s support because its argument lacked concrete case analysis. Business groups and the METI strongly criticized it as damaging tax certainty. Although some of the NTA’s challenges to large companies have failed, its transfer pricing investigations have been rising.29 The NTA conducts inspections on small and medium-sized enterprises (SMEs) as well as large businesses since even SMEs expand their business overseas as their business partners relocate. As the NTA broadens the scope of its transfer pricing investigations, its detailed guidelines will improve taxpayer compliance and reduce disputes between the tax authorities and taxpayers. It will also guarantee tax certainty for taxpayers.

28 Decision of the National Tax Tribunal of 27 January 2010 (TDK Co. case) and Decision of the National Tax Tribunal of 18 March 2013 (Takeda Pharmaceutical Co. case). 29 See, e.g., T. Miyatake, ‘Transfer Pricing Disputes in Japan’, in E. Baistrocchi and I. Roxan, edd, Resolving Transfer Pricing Disputes—A Global Analysis (2012), 415–438; N.i Oka, ‘International Taxation in Japan: Revised Transfer Pricing Guidelines by the National Tax Agency’, International Transfer Pricing Journal 21 (21 Mar. 2014), 99–106.

The Japanese Perspective on International Tax Law 927 In 2019, Japan revised its transfer pricing rules based on the BEPS Final Report30 and gave the tax authorities a new set of weapons. First, Japan defined ‘intangibles’ for the purpose of transfer pricing regulations and added the discounted cash flow method to the list of transfer pricing methods as outlined in the report. Secondly, the revised rules enable tax authorities to assess and determine the transfer price of ‘specified intangibles’ (hard-to-value intangibles, HTVI) based on the best method considering ex post outcomes derived from the transferred HTVI. It is not yet certain how these provisions will change the relationship between the taxing authority and the taxpayer.

50.4.2 2017 CFC Reforms In 2017, Japan revised the CFC rules based on the BEPS Action 3 Final Report.31 Japan had already moved to a territorial system in 2009, deducting 95% of dividends received from foreign subsidiaries from taxable income. Thus, Japanese multinational corporations can transfer profits to foreign subsidiaries and return the profits to their parent companies without paying additional Japanese taxes. This amendment increased the need to strengthen CFC rules to curb profit shifting.32 At the time, Japan applied the CFC rules to foreign subsidiaries if the effective tax rate borne by the subsidiaries was 25% or less while excluding subsidiaries that performed certain levels of substantial activities and management functions. Since Japan had adopted the entity approach, all income of the relevant subsidiaries was consolidated with that of the parent companies and taxed at the Japanese corporate tax rate.33 As tax competition intensified, Japan lowered the threshold (effective tax rate) for triggering the application of the CFC rules from 25% to 20% in 2010 and ‘less than 20%’ in 2015 to exclude major trading partners such as the Netherlands and the UK. In 2010, Japan adopted an element of the income approach.34 The ‘asset income’ of foreign subsidiaries trapped by the triggering rate criteria was taxed at the Japanese corporate tax rate, even if it satisfied the requirement of substantial activities. In other words, lawmakers stipulated provisions to consolidate asset income (some categories of passive income) of foreign subsidiaries with taxable income of their parent companies

30 OECD, Aligning Transfer Pricing Outcomes with Value Creation, Actions 8– 10— 2015 Final Reports: OECD/G20, Base Erosion and Profit Shifting Project (Paris: OECD Publishing, 2015). 31 OECD, Designing Effective Controlled Foreign Company Rules, Action 3—2015 Final Report: OECD/ G20, Base Erosion and Profit Shifting Project (Paris: OECD Publishing, 2015). 32 L. Lokken and Y. Kitamura, ‘Credit vs. Exemption: A Comparative Study of Double Tax Relief in the United States and Japan’, Northwestern Journal of International Law & Business 30 (2010), 621, 646. 33 The entity approach adopted by Japan directly impacted the location decisions of Japanese companies when establishing subsidiaries, considering the corporate income tax rate. S. Clifford, ‘Taxing Multinationals Beyond Borders: Financial and Locational Responses to CFC Rules’, Journal of Public Economics 173 (2019), 44, 49. 34 R. Ichitaka, ‘Key Practical Issues to Eliminate Double Taxation of Business Income’, Asia-Pacific Tax Bulletin 17 (2011), 448, 457.

928 Masao Yoshimura if they bore low effective tax rates. As Japanese lawmakers lowered the rate threshold, they believed that Japanese MNCs could now transfer passive income to their active subsidiaries and enjoy the tax benefits. The BEPS Final Report recommended strengthening the CFC rules based on the income approach, including a proposal for tackling so-called ‘cash box’ subsidiaries. As mentioned earlier, the Japanese government reviewed its hybrid rules35 and amended its tax laws in 2017.36 It defined ‘a shell company’ as a foreign subsidiary without a substantial office, and ‘a cash box company’ as a foreign subsidiary whose passive income exceeded 30% of its gross income or whose passive assets (including intangible assets) exceeded 50% of its total assets. Where a subsidiary falls into one of these categories, it must subject its total income to Japanese taxation unless its effective tax rate is 30% or more. As for the provisions based on the entity approach, Japan previously regarded active activities and management ability as a carve-out checklist but has now reversed the order. The CFC rules are generally applicable to companies that do not meet the checklist. In addition, for provisions based on the income approach, the scope of asset income has been expanded to ‘passive income’ including income from derivatives trades or foreign exchange and extraordinary income calculated by the formula. This income approach applies to all foreign subsidiaries even if they meet all the requirements in the checklists. These amendments dramatically expanded the scope of application of the CFC rules. However, during the review process at the Ministry of Finance, the METI and industry groups strongly argued that the proposal did not balance the compliance burden on companies with the risk of profit shifting. Therefore, the Japanese government finally proposed to exclud the application of both the entity-based rules and the income-based rules due to the minimal risk of profit shifting for the applicable foreign subsidiaries with an effective tax rate of 20% or more. Thus, the revised framework is identical to the previous one, although the reform overturned the application flow diagram. This was the result of the government’s efforts to balance the compliance burden for Japanese companies and the BEPS risk caused by those risks. Similarly, in the revision of the rules regarding interest deductions, the government considered compliance costs for companies in designing the structure of the rules.

35 Regarding

Japan’s hybrid approach, the OECD Final Report explicitly stated that ‘Because this approach ultimately considers different streams of income rather than just attributing all the income of an entity, it is essentially a version of a transactional approach’ (OECD, Designing Effective Controlled Foreign Company Rules, Action 3—2015, 55 fn. 21.) 36 For details, see Y. Nishikori and H. Sato, ‘Japan Upgrades Its CFC Regime’, Tax Notes International 89 (19 Feb. 2018), 705); S. Urushi and M. Suzuki, ‘Controlled Foreign Company Legislation in Japan’, in G. Kofler et al., eds, Controlled Foreign Company Legislation (Amsterdam: IBFD Publications, 2020), 375–404.

The Japanese Perspective on International Tax Law 929

50.4.3 Limitation Rules on Interest Deduction In response to the BEPS Final Reports,37 Japan’s 2019 tax reform strengthened the existing interest-deduction limitation rules.38 Japan already had thin-capitalization and earnings-stripping rules; however, both regulations applied to intra-group loans. Under Japan’s existing thin-capitalization rules, both the total and internal debt-to- equity ratio must not exceed the threshold of 3:1. In addition, lawmakers introduced the earning-stripping rules as a backstop for the thin-cap rules in 2012 before the launch of the BEPS Project. The previous threshold set for the earnings-stripping provisions was 50% of adjusted taxable income. Disallowed interest could be carried forward for seven years, a somewhat shorter period than Japan’s net operating loss carry-forward period of ten years. The BEPS Final Report showed that the 2019 tax reform expanded the scope of the earnings-stripping rules to all loans, including third-party loans, and lowered the earnings-stripping threshold from 50% to 20%. If net applicable interest payments exceed 20% of adjusted income in a relevant taxable year, the taxpayer cannot deduct the excess amount of interest costs from its taxable income. The disallowed interest can be carried forward for seven years as stipulated by the previous rules. Business associations expressed their opposition to extending the earning-stripping limitation to third-party loans, emphasizing the risk of double taxation.39 Considering this opposition, the 2019 tax reform provided generous exclusions from the definition of ‘applicable interest payments’ under the new earnings-stripping rules. The BEPS Final Report recommended minimizing exceptions to the regulations, however, Japan focused on balancing the cost of compliance for companies with the risk of causing profit shifting. As the BEPS 4 Recommendation was merely a common approach, the Japanese government decided to identify and address situations where profit shifting is likely to occur. As long as the recipient of the interest is subject to Japanese tax, the Japanese government determines that there was no need to capture it in the earning- stripping regime.40 Under the current earnings-stripping rules, the law defines excludable interest as follows.

37 OECD, Limiting Base Erosion Involving Interest Deductions and Other Financial Payments, Action 4–2015 Final Report: OECD/G20 Base Erosion and Profit Shifting Project (Paris: OECD Publishing, 2015). 38 K. Ogata, ‘Japan Branch Report on Interest deductibility: the implementation of BEPS Action 4’, Cahiers de Droit Fiscal International vol. 104A (2019), 415–433. 39 In response to the discussion draft on the fixed-ratio rule recommending that it be applied to a group’s net interest expense, Keidanren (the Japan Business Federation) argued that it would ‘hamper ordinary business operations, undermining incentives for investment’. It contended that much of the interest income taxed on the receipt side could suffer from economic double taxation. See Keidanren, ‘Comments on the Public Discussion Draft on BEPS Action 4—Elements of the Design and Operation of the Group Ratio Rule’ (12 Aug. 2016). Article can be reached at https://www.keidanren.or.jp/en/policy/ 2016/066.html. 40 Ogata, ‘Japan Branch Report on Interest deductibility: the implementation of BEPS Action 4’, 428.

930 Masao Yoshimura (1) For corporate bonds issued by the taxpayer and held by dispersed third-party investors: (a) if the payment is subject to withholding tax or included in Japan’s taxable income in the recipient’s hand, the total amount of the interest paid on the bond will be excluded from the interest limitation rules; (b) in other cases, if the bonds are issued in Japan, 95% of the amount paid will be excluded from the limitation; if the bonds are issued overseas, 25% of the payment will be excluded. (2) Other types of interest payments are allowed to be excludable from the applicable interest payments only if the receipts are subject to Japan’s income tax in the hands of the recipients. Interest expenses subject to Japanese taxation will be excluded from the new earnings- stripping rules, allowing Japanese companies to ignore the interest-deduction limitation when borrowing from ordinary domestic corporations. Despite the ambitious proposals in the BEPS Final Report, Japan has implemented moderate amendments considering the burden on corporations. Because the increase of an MNC’s tax burden reduces its competitiveness, it is politically impossible to design a tax system without considering the country’s level of tax compliance.

50.5 The BEPS Project and Japanese Corporate Behaviour 50.5.1 Tax Compliance Behaviour of Japanese Multinational Corporations When the BEPS Project launched, Japanese industry groups argued that Japanese companies did not practise aggressive tax planning and that implementing the BEPS recommendations would only increase compliance costs for ‘good’ Japanese companies.41 From the examples provided in this chapter, it can be seen that the government has carefully considered this argument as well as the regulations for combating tax avoidance, sometimes making exceptions to the BEPS recommendations. Several empirical studies agree that there is a behavioural culture among Japanese MNCs to not to involve themselves in tax avoidance. For example, according to Markle and Shackelford, Japanese firms have been facing higher effective tax rates than firms in other countries42 41 For

a case study of the involvement of the Keidanren (the Japan Business Federation) and BEPS Action 13, see K. Tsuda, ‘Keidanren’s role on the BEPS negotiation: a professional enabler under the Japanese chairmanship’, Conference Paper for TJN19 (2019), 1–27. 42 K. S. Markle and D. A. Shackelford, ‘Cross- Country Comparisons of Corporate Income Taxes’, National Tax Journal 65 (2012), 493–528; K. S. Markle and D. A. Shackelford, ‘The Impact of Headquarter

The Japanese Perspective on International Tax Law 931 and Hasegawa shows that Japanese firms are less responsive to changes in tax rules than US firms.43 The question is why Japanese MNCs tend to hesitate to engage in tax planning that lightens their tax burden and benefits their shareholders, and whether lawmakers would regard that behavioural culture as an unchanged prerequisite in the future. Izawa opines that the following three factors have shaped the behavioural culture of Japanese MNCs:44 • lack of experience; • the Japanese government’s coordinated tax regime learned from foreign precedents; and • a preference for nationality. First, Japanese companies do not have sufficient experience in managing global tax risks. Except for a few companies, Japanese companies operated within Japan (i.e. the territories of the Japanese Empire) before the war, therefore they were subject to harmonized tax rules.45 In addition, the Japanese government did not feel the need to conclude tax treaties with other countries. Unlike continental Europe and the USA, it had not negotiated comprehensive tax treaties with other independent countries prior to the Second World War.46 After Japan signed the Treaty of Peace in 1951, Japanese companies expanded overseas and became aware of local taxation, or double taxation. They adapted to international taxation rules for the first time, but focused on expanding the scale of their businesses rather than adjusting their tax burden during a period of rapid economic growth. The Japanese government also supported the overseas expansion of Japanese companies by providing a generous foreign tax credit system.47 Even today, many Japanese MNCs do not consider tax departments as an essential function of their businesses and do not provide them with many resources.48 Fewer Japanese companies have a formally documented tax strategy than companies in and Subsidiary Locations on Multinationals’ Effective Tax Rates’, Tax Policy and the Economy 28 (2013), 33–62. See also, J. H. Collins and D. A. Shackelford, ‘Corporate Domicile and Average Effective Tax Rates: The Cases of Canada, Japan, the United Kingdom, and the United States’, International Tax and Public Finance 2 (1995), 55–84. 43 M. Hasegawa, ‘Territorial Tax Reform and Profit Shifting by US and Japanese Multinationals’, KIER Discussion Paper No. 1016 (2019), 1–39. 44 R. Izawa, ‘Who, me? Tax Planning and Japanese Multinational Enterprises, 1887–2019’, Faculty of Economics, Shiga University, Working Paper Series No. 291 (2019), 1–21. 45 Ibid, 7. 46 Ibid., 4–5. 47 H. Kaneko, ‘Basic Structure of the Foreign Tax Credit System of Japan’, Bulletin for International Fiscal Documentation 40 (1986), 148; Nakazato, ‘Internationalization of Japan’s International Tax Law’, 84–86; Masui, ‘International Taxation in Japan: A Historical Overview’, 2815. 48 Izawa, ‘Who, me? Tax Planning and Japanese Multinational Enterprises, 1887–2019’, 10. Article can be reached at https://www.econ.shiga-u.ac.jp/ebr/No291.pdf.

932 Masao Yoshimura other countries, and they do not have well-developed internal control procedures for managing tax risks. Thus, Japanese companies are not accustomed to managing tax risks and lack a decision-making process to determine the acceptability of aggressive tax planning.49 Secondly, Japan emphasized cooperation between the public and private sectors in the process of policy implementation. The government focuses on issues already known abroad in advance by consulting with the private sector before the problems arise in Japan. For example, Japan introduced comprehensive CFC rules and thin-capitalization rules based on the US experience and OECD discussions before companies generally started to engage in aggressive tax planning. Thus, many Japanese MNCs view international tax management as being compliant with Japan’s international tax rules. Thirdly, most of the board members and key employees in Japanese companies are Japanese. They have been willing to perform essential head office functions in Japan, including managing intangible assets. Izawa argues that their explicit or implicit preference for nationality has limited the tax avoidance options available to Japanese companies.50

50.6 Concluding Remarks Based on the foregoing discussion, the metamorphosis of Japan’s international taxation rules can be summarized as follows. First, Japan has changed its treaty policy primarily based on its relationship with the USA. Immediately after the war, as a capital-importing country, it emphasized the retention of source-country taxation rights. However, as Japanese multinationals expanded overseas and increasingly received more revenue from their foreign subsidiaries, Japan changed its position in the 2003 Japan–US tax treaty negotiations towards lower source taxation or exemption. Furthermore, as a capital-exporting country, it was natural for the Japanese government to take a cooperative stance on several OECD projects. Next, in the twenty-first century, partly due to the stagnant Japanese economy, the government encouraged manufacturers to expand their offshore production and to spend the foreign profits on headquarter activities. Japan revised its tax treaties to conform to the interests of capital-exporting countries and moved to a territorial system, exempting dividends that were repatriated from overseas subsidiaries. Along with these incentives, Japan strengthened its CFC rules and interest-deduction limitation rules to prevent the diversion of income derived from the activities of MNCs headquartered abroad. However, to balance the compliance costs for Japanese

49 T. Kozu, ‘Recent Trends in Tax Management in Japan: Planning and Compliance— From a Governance Perspective’, Zeitschrift für Japanisches Recht 47 (2019), 19, 23. 50 Izawa, ‘Who, me? Tax Planning and Japanese Multinational Enterprises, 1887–2019’, 13.

The Japanese Perspective on International Tax Law 933 companies and the BEPS risks, the Japanese government introduced exceptions not included in the BEPS recommendations. Thirdly, the Japanese government significantly contributed to the success of the BEPS Project and revised international tax rules based on the 2015 BEPS final recommendations. However, Japanese companies that were sceptical about tax planning perceived the BEPS Project as a distraction that would only increase their compliance costs. Japan has worked with the private sector to develop international taxation rules; the revision of regulations in the 2000s is a good example. Some believe the relationship contributed to the high compliance culture among Japanese companies. However, this policymaking process in Japan tends to over-represent the voices of mature industries in decision making. For example, while many European countries decided to introduce a digital services tax, no such request was found in Japan. This silence may indicate the weakness of the political influence of start-ups offering digital services. Moreover, Japanese manufacturers, as with US platform businesses, opposed the spread of unilateral measures introduced by some market countries.51 Will the honeymoon period between the Japanese government and Japanese MNCs continue? Izawa insists that Japanese corporate culture will probably change,52 and thinks that the outlook appears to be correct. As noted earlier, the profits earned by the subsidiaries of Japanese MNCs are increasing so rapidly that they have placed more significant functions on regional headquarters, including finance and tax functions. Regional centres may employ several foreign experts and may be free from the conservative culture of the Japanese headquarters. Moreover, an enormous budget deficit will make it difficult for the Japanese government to continue to provide generous regulations for Japanese industries. Recently, Japan has been strengthening its taxation rules concerning MNCs in conformity with the BEPS Project. Times are changing—in an aging society, the Japanese government may project Japan as a market country, rather than an exporting country.53

51 For

an analysis of the comments on the OECD/G20 project by Japanese industry groups, see K. Aoyama, ‘Challenges of the Digital Economy on International Taxation Rules from the Perspective of Global Business Society’, Policy Research Institute, Ministry of Finance, Japan, Public Policy Review 17/ 1 (2021), 1–27. 52 Izawa, ‘Who, me? Tax Planning and Japanese Multinational Enterprises, 1887–2019’, 13. 53 e.g. the Ministry of Finance is not explicitly opposed to the ongoing Pillar One, which calls for the reallocation of tax revenues from Japanese manufacturers (e.g. Toyota) to market jurisdictions.

Pa rt V I I I

E M E RG I N G I S SU E S AND THE FUTURE OF I N T E R NAT IONA L TAX L AW

Chapter 51

T he Em erging C onse nsu s on Valu e Cre at i on Theory and Practice Allison Christians

51.1 Introduction The idea that income should be taxed where value is created has been called axiomatic, a fundamental principle, and even a paradigm for international taxation, but the precise meaning of the term is elusive. This is so because the nature of value and its origins are themselves elusive. But being elusive does not render the concept useless. Instead, its flexibility gives the term a certain familiar role in policymaking in that, like many other fundamental principles of international taxation, it can be used to create common ground among people that have distinct and even polarized ideas about what the tax system should look like. This chapter reviews some of the key ambiguities in the concept of value creation and argues that the term is positioned to become a common shorthand to assure various targeted audiences that their disparate goals are being met by whatever policy is being promoted. This will draw value creation into the revered canon of tax jargon: oft repeated, rarely explained, and susceptible to all manner of misuse and mischief, yet potentially useful to consensus-building. The discussion begins with a brief review of the emergence of value creation as part of the lexicon of international taxation, examining its quick take-up and causes for intrigue. It then examines three features that simultaneously make value creation an intuitively compelling concept yet ultimately render it an incomplete tax principle, namely: that in tax terms, value is a legislated construct rather than a science; that value is inconsistently identifiable and quantifiable; and that value and its causes of origin often defy geo-location, even though this is the express purpose of using the term in the first place. These features all point to interesting problems of tax cooperation that will almost certainly remain unresolved in the current iteration of global tax policymaking.

938 Allison Christians The chapter therefore concludes that the concept of value creation is likely to continue to intrigue and frustrate scholars while serving as a useful rhetorical flourish for advocates of various policy preferences across the worldwide tax community.

51.1.1 The Emergence and Intrigue of Value Creation Perhaps unusual in tax history, we can point to the precise moment when value creation received its place in international tax discourse; it happened on 6 September 2013 when the G20 issued a Leader’s Declaration ‘welcom[ing] the establishment of the G20/ OECD BEPS project’ with the apparent organizing mandate that ‘[p]‌rofits should be taxed where economic activities deriving the profits are performed and where value is created’.1 It seems likely that in floating the term, the finance ministers of the G20 were not engaged in a studied attempt to upend some prior agreed consensus but were most likely seeking to reflect what they already believed to be a consensus view in international tax circles. Later statements by the OECD2 seem to confirm this view. Even so, value creation was in fact a new term in the international discourse.3 Using it opened a door to a detailed look at what, exactly, we mean by value in the context of an international system whose raison d’être is to coordinate national tax regimes to prevent, as nearly as possible, the duplication of taxes on cross-border capital flows while still affording each nation as much autonomy as possible over the design of their national tax rules and regulations. The fact that the G20 statement was issued in the context of an expansion of the OECD’s mandate to also prevent ‘double non-taxation’ of such capital flows does not seem to have figured in the decision to adopt the value-creation vocabulary. Yet the rise of the idea of value creation alongside the expansion of efforts to prevent double non- taxation is significant. If double non-taxation is to be prevented through multilateral action, a number of assumptions about what this term means, how the phenomenon comes about, who is responsible for stopping it, and how to go about doing so all come into play.

1 G20

Leaders’ Declaration, St Petersburg Summit, ‘Addressing Base Erosion and Profit Shifting, Tackling Tax Avoidance, and Promoting Tax Transparency and Automatic Exchange of Information’ (6 Sept. 2013), 50. 2 The OECD is an intergovernmental organization comprised of thirty- seven members, most of which are the world’s richest countries. See OECD, ‘About us’, https://www.oecd.org/about/ (accessed 26 July 2022). Although it has long been characterized as an exclusive rich countries’ club, the OECD has played a dominant role in coordinating tax policy across its member states and, through their collective dominance, around the world. For a discussion about how and why the OECD plays this role, see A. Christians, ‘How Nations Share’, Indiana Law Journal 87 (2012), 1407. 3 The idea that profit is produced as a causal matter where ‘economic’ activities are ‘performed’ is apparently equally deserving of scrutiny. The enquiry seems to parallel that of value creation, but if so, why would both terms be needed in explaining the G20’s mandate for the OECD? This puzzle is worthy of further analysis.

The Emerging Consensus on Value Creation: Theory and Practice 939 As such, using the language of value creation invited an enquiry into the nature and origins of value at the very time that policymakers were engaging in a theoretical enquiry into the ‘rights’ that jurisdictions might or might not have to say that a given stream of income ‘belonged’ to them in the sense that they could exclude others from making a similar claim. The exercise of claiming of income is important because it effectively creates a hierarchical list of primary and secondary taxing rights operationalized by national and treaty-based double tax coordination regimes. If the ‘wrong’ jurisdiction is allowed to claim value, either double taxation or double non-taxation could result.4 Deciding the circumstances under which a given jurisdiction is in some mutually accepted sense ‘entitled’ to make a tax-relevant claim over a given income item thus became a necessary step to fulfil the goals of the coordination regime.5 It is for this reason that the language of value creation immediately sparked the interest of many tax scholars, including myself, perhaps well beyond what the G20 finance ministers or OECD policymakers anticipated. The term offered a new opportunity to examine one of the most fundamental normative and practical dimensions of international tax coordination efforts, namely, that attributing income to a given origin, also known as a source, is ultimately how the tax system allocates wealth across nations.6 4 See,

e.g., R. Altshuler and H. Grubert, ‘The Three Parties in the Race to the Bottom: Host Governments, Home Governments and Multinational Companies’, 7 Florida Tax Review 137 (2005); see also E. Cederwell, ‘Making Sense of Profit Shifting: Pascal Saint-Amans’ (22 May 2015), at http://taxfou ndation.org/blog/making-sense-profit-shifting-pascal-saint-amans. Quoting Mr Saint-Amans for the statement that: The international common principles drawn from national experiences to share tax jurisdiction have not kept pace with the changing business environment. . . . The current national tax rules are not coordinated, so there are gaps and frictions among different countries’ tax systems which are not dealt with by bilateral treaties or in the design of national tax rules. Sophisticated multinational enterprises (MNEs) can take advantage of these technically legal tax asymmetries. They are often aided by some governments willing to engage in harmful tax practices, which can be addressed with greater transparency and economic substance requirements. 5 The

idea of states having rights or being entitled to do things, including exercise the power of taxation, is a subject of great debate. For a discussion see, e.g., M. Levi, Of Rule and Revenue (Berkeley, CA: University of California Press, 1988); S. A. Dean, ‘More Cooperation, Less Uniformity: Tax Deharmonization and the Future of the International Tax Regime’, Tulane Law Review 84 (2009), 125, 144; A. Christians, ‘BEPS and the Power to Tax’, in A. Christians and S. Rocha, eds, Tax Sovereignty in the BEPS Era (Alphen aan den Rijn: Kluwer, 2017). 6 This question has been asked in various ways since the dawn of multilateral talks on income tax coordination. E.g. in examining how to prevent double taxation of multilateral income for a League of Nations report that ultimately served as a template for decades to follow, a committee of ‘technical experts’ suggested that each state should ‘tax the portion of the income produced in its territory’. League of Nations, Report presented by the Committee of Experts on Double Taxation and Tax Evasion (Geneva: League of Nations, 1927). As I have noted in prior work, it was also clear to these same technical experts that assigning an exclusive origin to any one part of a given income stream would be scientifically impossible. See A. Christians, ‘Taxing According to Value Creation’, Tax Notes International 90 (18 June 2018), 1379. Nevertheless, the expert’s logic followed the jurisdictional tie of ‘economic allegiance’ that developed in an earlier report. G. W. J. Bruins et al., Report on Double Taxation Submitted to the Financial Committee Economic and Financial Commission, League of Nations Doc. E.F.S.73.F.19 (Geneva: League of Nations, 1923). For this reason, Li, Bao, and Li argue that value creation is conceptually linked to

940 Allison Christians This brings value creation into the realm of foundational tax principles, implicating well-worn concepts that continue to provoke analysis precisely because they are never neatly encapsulated artefacts of scientific precision but are the product of years, if not decades, of political, social, and cultural history. The G20 declaration on 6 September 2013 not only initiated the use of the term value creation in income taxation, but it did so in particular to explain how profits generated by multinationals ought to be allocated across jurisdictions for the purposes of sharing the possible tax revenues to be collected thereupon. This is not a definitional exercise, nor a heuristic one. It is a political one. Value creation was a simple yet powerful way to signal the work to be done. Accordingly, it was quickly picked up by the OECD as well as a range of people representing non-governmental and civil society organizations, think tanks, and multinational corporations. None of these early adopters attempted to define value, let alone value creation. Instead, a conventional wisdom about the familiarity, even banality, of the term seemed to develop overnight. Tax scholars soon noted and took issue with the proposed terminology, most characterizing its ambiguity as a fatal flaw. But ambiguity can be a feature rather than a bug when it comes to developing multilateral consensus. After all, the entire international tax system rests on a number of equally ambiguous concepts forged in the early days of multilateral tax cooperation efforts. Among these are the notoriously slippery ‘nexus’-creating terms source and residence, each of which has attracted volumes of policy and scholarly analysis, as well as nearly constant iterations of legal reform.7 These terms can be fairly summarized as little more than highly loaded words with multiple meanings and vexing nuances that cannot be explained away by any objective standard. Added to these are any number of ideas strewn through the literature and the discourse with abandon, such as the infamously incoherent notion of ‘tax sovereignty’ and the abundantly popular ‘fair share’.8 (and reflects an ongoing commitment of the international community to) the concept of economic allegiance. Jinyan Li, N. J. Bao, and H. Li, ‘Value Creation: A Constant Principle in a Changing World of International Taxation’, Canadian Tax Journal 67/4 (2019), 1107–1134. For a comprehensive analysis of the origins of today’s international tax system, see S. Jogarajan, ‘The Conclusion and Termination of the “First” Double Taxation Treaty’, British Tax Review 3 (2012), 283–306; S. Jogarajan, ‘Prelude to the International Tax Treaty Network: 1845–1914 Early Tax Treaties and the Conditions for Action’, Oxford Journal Legal Studies 31/4 (2011), 679–708; S. Jogarajan, ‘Stamp, Seligman and the Drafting of the 1923 Experts’ Report on Double Taxation’, World Tax Journal 5/3 (2013), 368–392; and S. Jogarajan, Double Taxation and the League of Nations (Cambridge: Cambridge University Press, 2018). 7

See, e.g., L. Lokken, ‘What Is This Thing Called Source’, International Tax Journal (May–June 2011), 21, 53; T. Edgar and D. Holland, ‘Source Taxation and the OECD Project on Attribution of Profits to Permanent Establishments’, Tax Notes International 37/6 (2005), 525–539; A. Christians, ‘Drawing the Boundaries of Tax Justice’, in K. Brooks, ed., Quest for Tax Reform: The Carter Commission 50 Years Later (Toronto: Carswell, 2013), 53; S. Gadžo, ‘The Principle of “Nexus” or “Genuine Link” as a Keystone of International Income Tax Law: A Reappraisal’, Intertax 46/3 (2018), 194–209. 8 See, e.g., D. Ring, ‘Democracy, Sovereignty and Tax Competition: The Role of Tax Sovereignty in Shaping Tax Cooperation’, Florida Tax Review 9 (2009), 555; R. Eccleston and A. Elbra, eds, Business, Civil Society and the New Politics of Corporate Tax Justice: Paying a Fair Share (Cheltenham: Edward Elgar, 2018).

The Emerging Consensus on Value Creation: Theory and Practice 941 Each of these terms is susceptible to equal parts use and misuse, each attracting ebbs and flows of scholarly attention, and each has served to sustain generation after generation of multilateral tax consensus building. In joining this discourse, value creation is a new term but it occupies a familiar role. Given this role, it may be considered folly to try to interpret or explain value creation. It would be easy to simply dismiss it as a politically expedient tool and not a normative principle worthy of analysis. This may seem especially appropriate given that those who first embraced the term seem now willing to abandon it. But it is worth questioning whether the move away from the term is in any part owing to the immediate and intense scrutiny and attention it attracted. Value creation struck a nerve for many people interested in international tax. It may have done so precisely because the concept embodied a whole host of assumptions that concern those who work and study in this area. As such, it is at least worth attempting to understand what questions and challenges arise in the statement that income should be taxed where value is created.

51.1.2 Value is a Legislated Construct Valuation is really important in taxation systems. Regardless of the type of tax, a transfer of value from one person to another typically triggers an assessment if not an automatic payment of tax due. Value triggers taxation in an ad valorum tax system (literally: in proportion to the value, as in a sales or value-added tax) and on the calculation of net income associated with the transfer in an income tax system. As such, every tax system has to deal with the idea of value. Yet value is not a single concept with general applicability. In his influential 1937 treatise on the valuation of property, James C. Bonbright stated that ‘[W]‌hen one reads the conventional value definitions critically, one finds, in the first place, that they themselves contain serious ambiguities, and in the second place, that they invoke concepts of value acceptable only for certain purposes and quite unacceptable for other purposes.’9 He concluded that defining value is as much a matter of identifying (and justifying) the methods to be used in crafting such a definition as it is interpreting the definition itself.10 Further, Bonbright could identify no single definition of value with general applicability for tax purposes, nor even for a given type of taxation such as that on income. He concluded that ‘the subject must be divided and sub-divided by reference to the particular uses of the valuation and to the specific types of property’.11 9

J. C. Bonbright, Valuation of Property (New York: McGraw-Hill, 1937), 11. Bonbright, ibid., stating that ‘[T]‌he problem of defining value, for the many practical purposes for which the term is used, is an exceedingly difficult one, deserving quite as much attention as does the technique in proof.’ 11 Bonbright, ibid., 451, stating that in many property law contexts, the reason for an inquiry into value was to recompense an aggrieved party in a lawsuit, but this is not the case in taxation, where the object is to calculate a contribution to the treasury, and noting that ‘it would be vain to seek any general concept of “value for tax purposes,” applicable alike to the general property tax and the inheritance tax, or to a 10

942 Allison Christians Since value is a regular fixture in tax systems, it is perhaps not surprising that, most of the time, the simplest and best solution for many applications of value in taxation has been to work on the baseline assumption that market prices reflect value.12 Accordingly, tax systems customarily (with myriad exceptions, as Bonbright and many others have noted) assume that value implies market value, and market value implies fair market value. Using market prices to identify value is necessary but insufficient in the construction of a comprehensive tax system. It is necessary because if tax authorities were unable to use market prices as a proxy for value, they would be forced to undertake valuation of every tax-relevant transaction by some other means, namely through independent appraisal. Influencing and shaping the appraisal process would quickly become a cottage industry. Taxpayers would certainly object to unfavourable valuation assessments and present their own valuations in turn; review and appeals systems would have to be established for even the most routine and everyday value transfers, and the administration of the tax system would be hijacked by both real and petty disagreements about the intrinsic value of everything. Think, for example, of the kinds of arguments employees bring to seek a downward adjustment of income attributed to them by virtue of receiving expenses-paid trips and tickets to sporting events as employee performance awards. Similarly, think of the valuation issues involved in dealing with discounted goods and services provided to employees. If valuation was always to be administratively determined and then subject to constant contestation, tax authorities would quickly be overrun. The only solution is to work with the market price wherever possible.13 Even so, market prices are simply not an adequate account of value in a perhaps surprising number of situations. To see why this is so, we need only examine the way that lawmakers have defined value in practice. A common definition found across legal subject areas where appraisal of value is relevant—including taxation, banking, labour law, competition, bankruptcy, expropriation and eminent domain, and consumer protection—is that an accurate estimation of value can be perceived from the market price (only) when it derives from unrelated parties who willingly transact under conditions in which neither is compelled to trade and both have reasonable knowledge of the material facts.14 This definition, of course, requires a cascading set of sub-definitions, capital-value tax and an income tax. Indeed, there cannot even be any such thing as “value for income- tax purposes,” or “value for general-property-tax purposes,” valid under all circumstances’. 12 See, e.g., B. Marks, ‘Valuation Principles in the Income Tax Assessment Act’, Bond Law Review 8 (1996), 114, 117 (‘[C]‌ourts and administrators have generally treated the meaning of the terms ‘value’, ‘market value’, ‘fair market value’ and ‘open market value’ as both equivalent and interchangeable whether they applied to income tax, capital transfer, death and estate duties or stamp duties’). 13 See, e.g., K. Brooks, ‘Delimiting the Concept of Income: The Taxation of In-Kind Benefits’, McGill Law Journal 49 (2004), 255(examining the application of fair market valuation to employee non-cash benefits in Canadian tax law). 14 In Canadian legislation, see, e.g., Competition Act, RSC 1985, c. C-34 s. 52.1(3)(c) and (d) (using the concept of fair market value in restricting deceptive telemarketing practices); Bankruptcy and Insolvency Act, RSC 1985, c. B-3 s. 2 (defining ‘transfer at undervalue’ as ‘a disposition of property or provision of services for which no consideration is received by the debtor or for which the consideration

The Emerging Consensus on Value Creation: Theory and Practice 943 each of which raises a range of questions including which kinds of relationships should alter the various assumptions, what it means to be ‘willing’ to transact, and what to do with market-wide distortions (e.g. monopolies and quasi-monopolies, as well as manias, panics, and crashes).15 It also sidesteps the immensely difficult problem of value transfers that do not take place in a standard market context. If it is not obvious that given these difficult questions it is all but impossible to articulate a single, precise definition of value, consider the sources that lawmakers and courts go to for interpretive guidance. Black’s Law Dictionary is often cited by courts and scholars for the general legal proposition laid out earlier, namely that value is fair market value and that fair market value requires unrelated parties, consent, and knowledge.16 The same general definition is repeated in a wide range of scholarship on the role of valuation.17 Leaving aside valid critiques of the practice of deferring to dictionary writers for legal interpretation, Black’s Law Dictionary provides a clue as to just how difficult it is to go from a conceptual idea of what defines value to an administrable rule set. For instance, consider that Black’s definition of value explains that a price obtained from a sale that is ‘forced by the necessities of the owner’ is not a market price, that is to say, does not reflect its fair value.18 This merely opens up multiple new lines of enquiry: what is it to be forced? What are the necessities? Which necessities? Consider workers who accept substandard working conditions and wages in order to avoid even worse conditions elsewhere. These workers are obviously ‘forced’ by ‘necessities’ to trade their services for remuneration at less than they would were they not completely overpowered in the marketplace of wage-setting and wage-taking. As such, it is clearly

received by the debtor is conspicuously less than the fair market value of the consideration given by the debtor’). For Canadian jurisprudence interpreting the definition of fair market value, see, e.g., Lake Erie & N. Ry Co. v. Brantford Golf & Country Club [1917] 32 DLR 219, 229; Woods v. The King [1951] SCR 504; Pastoral Fin. Ass’n [1914] AC 1088. In US legislation, see, e.g., 29 USC § 1002(26) (labour) (defining the term ‘current value’ as ‘fair market value where available and otherwise the fair value as determined in good faith by a trustee or a named fiduciary’); 12 CFR § 703.11 (banks and banking) (using the term fair value in regulations for valuing securities); 18 CFR §§ 4.10ff (conservation of power and water resources) (using the term fair value for purposes of appraising specified constructed projects); 19 CFR §§ 351ff (customs duties) (explaining calculations of export price, fair value, and normal value). 15 The constant expansion of law to cover all the possible variables is the predictable, if sometimes lamented, result. See, e.g., B. Manning, ‘Hyperlexis and the Law of Conservation of Ambiguity: Thoughts on Section 385’, The Tax Lawyer 36 (1982), 9. 16 ‘Market Value’, Black’s Law Dictionary, 6th ed. (St Paul, MN: West Publishing, 1990) and 2nd ed. (1910). The definition further notes that a market price is one that ‘would be fixed by negotiation and mutual agreement, after ample time to find a purchaser, as between a vendor who is willing (but not compelled) to sell and a purchaser who desires to buy but is not compelled to take the particular article or piece of property’. 17 See, e.g., Howard L. Braitman, ‘The Eye of the Beholder: A Fresh Look at Fair Market Value’, Taxes 52 (1974), 269 (echoing the definition and examining its use in US tax statutes and regulations); Marks, ‘Valuation Principles in the Income Tax Assessment Act’ (reviewing the concept of valuation in Australia’s tax law by reference to the same elements as those laid out in Black’s Law Dictionary and analysing each of the elements of the definition). 18 ‘Market Value’, Black’s (1990); ‘Market Value’, Black’s (1910).

944 Allison Christians not the case that substandard wages are fair market prices. Yet market wages are routinely accepted without question as the definitive value of labour for tax purposes. As explained in more detail below, allowing this error to stand also allows the uncounted value to be attributed to some other factor, which is precisely the problem that led to the articulation of value creation as an ideal. To state the principle of market value is thus to highlight a range of troublesome assumptions that must carry a huge conceptual load when applied to value in a tax system.19 In practice, the application has led to volumes of rules, regulations, interpretations, and guidance of all forms, which has in turn spawned a gargantuan, world-spanning tax-compliance industry, whose deft navigation of legal ambiguities has in turn spawned further rules, regulations, interpretations, and guidance. None of these recursive cycles of lawmaking has ultimately clarified the concept of value, however.

51.1.3 Value is Inconsistently Identifiable and Quantifiable A second major feature of value creation that is both fascinating and troubling for application in a tax system is that not all value has market value, and we really have no idea what to do about that. Even if, contrary to the argument in the previous section, it was possible to come to a clear, objective standard of value, it is obvious that lawmakers are either unwilling or unable (or both) to account for the kinds of value that are not represented in simple market price terms. Two main categories of uncounted value make the point clear, namely, that of underpaid and unpaid labour and that of negative externalities. The failure to account for unpaid or underpaid labour and negative externalities in national accounts was the main thesis of Marilyn Waring’s groundbreaking work beginning in the late 1980s. In If Women Counted,20 Waring demonstrated that the collected wisdom around what things create valued includes some plainly incoherent ideas. In addition to demonstrating the incoherence of excluding the value of uncompensated labour in national accounts, Waring described environmental destruction as a paradox of value creation. How could it be, she asked, that keeping the environment free of toxic substances is assigned no value in economic terms, while expending resources to clean up after a toxic waste accident is valuable because it involves investment of capital and job creation? Waring’s work revealed the pernicious social and cultural effects of failing to recognize or correct these valuation assumptions and mistakes. Less direct in her 19 Some

of these have been discussed at length by scholars, policymakers, and judges and have been accounted for in specific tax rules. For some comprehensive analyses, see, e.g., M. G. Kemp, ‘The Meaning of Market in Fair Market Value’, UBC Legal Notes 2 (1958), 593, 602; D. S. Goldberg, ‘Fair Market Value in the Tax Law: Replacement Value or Liquidation Value’, Texas Law Review 60 (1982), 833; E. L. Gordon, ‘What Is Fair Market Value?’, Tax Law Review 8 (1952), 35; Marks, ‘Valuation Principles in the Income Tax Assessment Act’; H. Wurzel, ‘The Tax Basis for Assorted Bargain Purchases or: The Inordinate Cost of “Ersatz” Legislation’, Tax Law Review 20 (1964), 165. 20 M. Waring, If Women Counted (New York: Harper & Row, 1988).

The Emerging Consensus on Value Creation: Theory and Practice 945 work but every bit as obvious, these assumptions and mistakes have also had pernicious effects for tax systems. The problem of failing to account for some sources of value in economic terms is that the tax system requires valuation, as we saw earlier, not only to determine the existence of some transaction or activity that is meant to trigger a tax, but also to ensure that the correct taxpayer is assigned the tax base so created (whether it be income or some other tax base). This is problematic in the domestic context, and the problems only intensify when the task is to distribute a given tax base among jurisdictions. When some value sources are ignored, the tax system necessarily assigns their value to other things that are counted. Think of the idea of goodwill in the corporate context. When the market capitalization of a company exceeds the estimated market value of all of its assets, the extra value is swept up in the generalized category of goodwill. Similarly, when a company is able to underpay its workers or export negative costs to the public—by, for example, dumping waste into local waterways—we do not see a line in the books showing the profits that were made possible by (or in tax terms, attributed to) those things. But we do see a line for the company’s unique intellectual property items, such as trademarks, licences, and so on. Since the former is invisible and the latter visible, the legally recognized intangibles act as a sponge for the value that was actually created by underpaying workers or creating negative externalities. The point was driven home by renewable energy advocate Assaad W. Razzouk in a call for environmental impacts to be reflected in public company income statements, ‘because otherwise [the stated] earnings are wrong’.21 Razzouk argued that ‘all companies in all sectors hold assets that are at risk in terms of climate change’, and that failure to account for them in financial statements constitutes a material omission: ‘They are invisible and their earnings, in other words, are inflated.’ Were these costs to be properly reflected, Razzouk argued, some consumer products would be eliminated as simply unprofitable. The observation demonstrates that inconsistently accounting for different types of value starts as a problem outside the tax system. This probably explains why it is so difficult for the tax system to offer a ready fix. Indeed, there are no doubt many simple and rational reasons why taking these things into account must be impossible in a tax system, but the fact is that not taking them into account is a choice in many cases. As shown in prior and forthcoming scholarship, both the underpayment of labour and the costs associated with externalizing many environmental risks can be quantified, at least in some respects.22 This scholarship points to research and methodologies developed by experts in the field of living-wage analysis and 21 A.

W. Razzouk, Podcast (2 Apr. 2021), https://theangrycleanenergyguy.com/podcast/episode-49/ . He added that, ‘What we need . . . is a provision that the company running that risk takes against its earnings in case it messes up, and we then need to decrease that company’s earnings by that amount and change its valuation, unless it never discharges into waterways and dirty our drinking water.’ 22 A. Christians and L. van Apeldoorn, ‘Taxing Income Where Value is Created’, Florida Tax Review 22/1 (2019); A. Christians, ‘Designing a More Sustainable International Tax System’, Dalhousie Law Journal 44/1 (2021); A. Christians and L. van Apeldoorn, Tax Cooperation in an Unjust World (Oxford: Oxford University Press, 2021); A. Christians, ‘Carbon Pricing and the Income Tax’, Canadian

946 Allison Christians life-cycle analysis as two ready sources of quantitative information about certain types of uncounted value. It is admittedly the case that incorporating new sources of quantification will serve to heap complexity on top of complexity. However, there is no clear justification for accepting complexity to one point and not the next, if the different levels of acceptance have distributional consequences (as this prior scholarship demonstrates). In any case, making a continued choice to ignore some kinds of heuristic and methodological tools while readily using other (arguably, equally flawed) ones means that we simply do not consistently account for all the things that produce value. The consequence of inconsistently recognizing and accounting for value is that we continuously risk assigning value to the wrong factors as a matter of course. This has distributional consequences given that the motivating purpose of the value-creation concept is to assign (primary) taxing rights and tax revenues to different jurisdictions, according to some kind of coordinated consensus. This brings us to the third and related feature of value and its origins that make its use in tax contexts both fascinating and troublesome.

51.1.4 Value and its Origins Defy Geo-Location This third feature of value creation, closely related to the previous two, is that we really cannot isolate the distinct variables that causally produce value, whether as an economic matter or as a matter of building an administrable tax system. This is to say that any given dollar to be taxed is the product of multiple inputs and variables, each of which is inextricably combined in the production of that dollar. Yet if the tax system is to assign the tax base among different taxpayers, some method of dividing the dollar is necessary. The ability to do this is especially important in the international context since the act of dividing can also be the act of assigning a primary right to tax to one jurisdiction instead of another. Yet, the task is impossible. First, it is effectively impossible as an economic matter in the same way that multinational profit itself is said to be an indivisible product of synergies gained through the act of owning instead of transacting the various stages and processes of production.23 Market prices that would arise in a counterfactual world of autonomous independent enterprises transacting with each other at arm’s length can only be guessed at (more on this later). In an interconnected world with international financial flows, an overwhelming amount of value is generated by combining multiple factors through multiple processes in multiple jurisdictions.24 These factors include traditional value-generating inputs like capital, labour, natural resources, and other

Tax Journal 20/1 (2022); A. Christians and T. D. Magalhães, ‘The Case for Taxing Away Unsustainable Profits’, George Washington Law Review 91/1 (2023). 23

This was reflected in the original multilateral income tax coordination efforts, as noted earlier. Christians and van Apeldoorn, ‘Taxing Income Where Value is Created’; L. van Apeldoorn, ‘Exploitation, International Taxation, and Global Justice’, Review of Social Economics 77/2 (2019), 163. 24 See

The Emerging Consensus on Value Creation: Theory and Practice 947 tangible things, as well as well-functioning marketing and distribution channels.25 The latter include legal and financial institutions that, effectively, make it possible for corporations to exist, transact, and produce value as well as extracting value in the form of rent—another major challenge in applying the value-creation idea. There is simply no way to accurately fragment a given tax base to assign causal credit to one or another of the various inputs that contributed to its existence. Relatedly, even if the economic problem could be overcome, the administrative barriers posed by implementing any given fragmentation approach are probably insurmountable. For example, even if it were possible to solve the inconsistent accounting problem discussed, for example by implementing living-wage and life-cycle assessments to account for the value that is created (or more accurately, extracted) from imposing negative externalities, it is probably the case that all but a few governments would be utterly overwhelmed in their attempts to implement such a system and monitor compliance. In traditional assessment, audit, and appeal processes, including treaty-based dispute-resolution processes, the administrative burdens would begin with the inevitable conflicts between taxpayers and governments, and would compound as governments then seek to resolve disputes by mutual agreement. If the various available heuristic and methodological tools are asymmetrically adopted or accepted, the interpretive exercise quickly expands to an intricate web of potentially unresolvable inconsistencies and differences of opinion. A good illustration is found within the arcane area of transfer pricing, which is currently the main tool used to compute taxes on multinational profits but is also the main tool that distributes taxing rights and tax revenues across jurisdictions.26 Transfer pricing rules spring forth from the notion that parties must be unrelated in order for a price to be said to reflect market value. Relationships matter for valuation because the freely transacted prices of parties who share an interest in their collective tax consequences are almost certain to reflect non-economic decisions.27 When parties are related, we say that they do not deal with each other at arm’s length. The metaphor evokes the idea that market participants are presumed not to be altruistic but to seek advantage for themselves wherever possible, including by trading on knowledge and power asymmetries. When they do not seek these advantages, they are 25 K.

G. Dau-Schmidt, ‘Dividing the Surplus: Will Globalization Give Women a Larger or Smaller Share of the Benefits of Cooperative Production?’, Indiana Journal of Global Legal Studies 4/1 (1996), 51, 53(‘The economy, whether at a local, national, or global level, can be thought of as a problem of cooperative production. At any of these levels, there are many resources that can be productively employed in joint production to people’s benefit: natural resources, capital, and labor’). See also C. H. Koch Jr, ‘Cooperative Surplus: The Efficiency Justification for Active Government’, William & Mary Law Review 31/2 (1990), 431 (using the theory of cooperative surplus at the national level as an efficiency- based argument derived from economics against constitutional libertarianism). 26 See, e.g., J. Monsenegro, ‘Value Creation and Transfer Pricing’, in W. Haslehner and M. Lamensch, eds, Taxation and Value Creation (Amsterdam: IBFD, 2021). 27 See, e.g., Swiss Bank Corp. v. Minister of Nat’l Rev., [1974] SCR 1144, 1152 (Can.) (when parties are not dealing at arm’s length, nothing can ensure that the transaction ‘will reflect ordinary commercial dealing between parties acting in their separate interests’).

948 Allison Christians not like other participants in the market so their negotiated prices are suspect. This does not mean the negotiated exchange prices of related parties are morally bad or scientifically wrong. It simply means that the tax system has a valuation problem in that it can no longer simply trust the prices claimed to represent the value transferred among the related taxpayers. It must therefore turn to something other than the parties’ own representations. It is a matter of utter fascination that when the parties are related, the typical solution in an income tax is for the taxpayers to support their claimed exchange prices by conducting a search for market transactions that are sufficiently like their own, but that occur between unrelated parties.28 The business of deciding what is sufficiently alike is a heuristic and methodological quagmire on its own. But the fact that transfer pricing creates distributional consequences (by effectively assigning income to one jurisdiction or another, which then ideally allows one or the other but not both jurisdictions to tax such income) makes it that much more important in any discussion of where value might be created. It is a matter of (cross-border) legal recognition of the corporate entity, together with that of various forms of assets and ownership claims, that multinational companies can be said to be made up of multiple entities, each of which can have legal rights, and therefore each can be seen to do things like pay the other for the right to use intellectual property or charge for the use of borrowed money. These inter-company transactions are wholly fictional in that in a consolidated income statement they would all cancel each other out. The bottom-line profit of the multinational as a whole is the relevant outcome for investors, after all. But each and every one of those inter-company transactions is in effect a claim that a value-creating asset or activity exists somewhere within the company, and the profit earned by the multinational should be accordingly assigned. It then falls to each of the potentially affected governments to inspect these claims to see if the taxpayer is being entirely candid. In doing so, the tax authorities generally accept the many legal fictions involved—the organizational documents that declare the existence of the separate entities, as well as the contracts of property ownership and obligation held by the related parties, their terms, their contexts, and conditions, and so on. It is not (generally) the aim of transfer pricing adjustment enquiries to question or invalidate the existence of an entity or the claim of ownership of a legal asset. Instead, the aim of the enquiry is to interrogate whether the claim of the amount of value created, as indicated by a given exchange price involving those entities or assets, can be said to be comparable to that which other taxpayers, who are not in similarly controlled situations vis-à-vis their exchanging parties, would have likely claimed. The transfer pricing adjustment exercise therefore involves not only the taxpayer and the tax authorities separately identifying ‘comparable’ transactions but also: making any

28

OECD Transfer Pricing Guidelines, 43 (‘‘comparability analysis’ is at the heart of the application of the arm’s length principle’).

The Emerging Consensus on Value Creation: Theory and Practice 949 number of adjustments to the comparables to account for the unique circumstances of the parties and the transaction in question; eliminating ‘outliers’ from the possible pool of comparables based on any number of subjective factors; writing copious tomes explaining all of these decisions; and then writing copious tomes explaining why the other party’s copious tomes ought not be accepted. The lengths of artifice that can be reached in this exercise, by both taxpayers and tax authorities, are nothing short of astounding to everyone involved in the process. They include imagining wholly counterfactual markets, market participants, and transactions, and justifying all manner of profit-motivated data sets and econometric methodologies, most of which is well beyond the interest or capacity of regular tax lawyers to understand, let alone the general public or the lawmakers tasked with designing workable tax systems. It is no wonder that worldwide there are thousands of unresolved transfer pricing disputes awaiting resolution by various processes and bodies. Most of these disputes will be resolved through opaque administrative processes that are designed not to illuminate or refine anyone’s conception of value but rather to deliver a bespoke (and typically unexplained) solution solely to the specific taxpayer involved.29 Transfer pricing is but an example of the complex regulatory work that goes into identifying value in a tax system, especially when the object of doing so is to decide where value is created. Approaches of varying complexity are similarly found in the assignment of income rules, deemed distribution rules involving controlled corporate enterprises, deemed sale and income rules involving partners in their dealings with partnerships, and various constructive receipt-and other income-deeming rules of all kinds, which exist across tax systems the world over. The unifying theme of all of these regimes is that value is a concept that defies generalization. As such, it is subject to infinite interpretation and manipulation.

51.2 Conclusion Value is not a rule or a measure but a dynamic social construct whose creation is a puzzle for the ages. Yet value creation is an intriguing idea that captures a central question for taxation, namely: what should nations view as their share of the economic gains that arise from cross-border activities? This chapter has demonstrated that the question cannot be answered by reference to any economic or legal definition of value or any empirical examination of the various interrelated origins of value. Yet its definitional ambiguity does not render the concept useless. On the contrary, it seems likely that since the

29 A. Christians, ‘How Nations Share’, Indiana Law Journal 87 (2012), 1407(explaining how taxpayers and tax authorities resolve cross-border disputes, including transfer pricing, through a combination of internal domestic appeals and treaty- based mutual agreement procedures which conclude in confidential resolutions, typically without giving reasons or rationales).

950 Allison Christians question it evokes is enduring, the idea of value creation will continue to be drawn into international tax discourse going forward. If so, the term is likely to be as instrumentally useful to some as it is a source of frustration and irritation to others. This chapter has not resolved this tension, but it has illuminated some of its interesting and troubling aspects. Doing so reveals the potential of value creation to settle into international tax discourse as an enduring rhetorical tool alongside similarly manipulable concepts like nexus, sovereignty, or fairness. It is the very mixture of intuitive familiarity and political malleability of these terms that makes them useful, perhaps even more so than a clear and unambiguous alternative might. In its more recent documents related to the base erosion and profit shifting (BEPS) initiative, the OECD pointedly discontinued its use of the term. This has already led to some scholars predicting that far from embedding in the collective consciousness, value creation will rapidly disappear from the international tax vernacular altogether. Yet it is also clear that the current cycle of global tax-norm development and policymaking, as every cycle before it, is destined to deliver an incomplete consensus. As a significant point of contention in this cycle, value creation presented an opportunity to think about the objectives of international tax cooperation and reflect upon the interconnecting nature of ideas, rules, norms, laws, and practical realities. Since complete global tax harmonization is impossible, these themes are sure to arise again. Value creation is as well positioned as any other term to serve as the touchstone.

Chapter 52

The All o cation of Tax i ng Righ ts under Pi l l a r One of the OECD Prop o s a l Aitor Navarro

52.1 Introduction Pillar One is a proposal by the OECD Secretariat to address changes in the allocation of taxing rights of business profits at an international level. It comes as a reaction to the tension generated by the inadequacies of the international tax regime that are significantly exacerbated by the digitalization of the economy.1 The proposal aims at achieving a long-standing, consensus-based solution on the matter. Pillar One comprises two components. Amount A assigns taxing rights to market jurisdictions on profits obtained by multinational enterprises (MNEs) operating above certain thresholds through a residual profit split that relies on formula-based calculations. Amount B aims to standardize the remuneration of related party distributors that perform baseline marketing and distribution activities. Notably, the rationale of Pillar One is in opposition to long-standing standards on the allocation of taxing rights of business profits such as the requirement of physical presence to allow source taxation or the arm’s-length standard as a profit-allocation mechanism. This chapter first addresses the context prior to the Pillar One proposal. It explicitly explains how a consensus was constructed around the concept that the international tax regime is outdated, and the formation of the Unified Approach advocated by the OECD in which Pillar One is integrated as a possible solution (Section 52.2). The specific components of the proposal are briefly described (Section 52.3), followed by a

1 A,

Báez Moreno and Y. Brauner, ‘Taxing the Digital Economy Post BEPS ... Seriously’, Columbia Journal of Transnational Law 58 (2019), 121, 157–161.

952 Aitor Navarro normative evaluation of Pillar One and its compromise-based rationale (Section 52.4). The chapter concludes with remarks on the feasibility of Pillar One and the future of the international tax regime (Section 52.5).

52.2 The Context Prior to the Pillar One Proposal The rationale of Pillar One can only be understood by examining the reasons justifying a review of the fundamental features of the international tax framework. Concisely, the central issue refers to the high level of electivity on the key components instigating the levying of taxes on business profits. The taxation of business profits rests exclusively on the supply side, specifically on the location of production factors, while the demand side—location of the market—is utterly ignored.2 Currently, the main value drivers that allow for the extraction of economic rents revolve around the relevance of intangible assets and human capital in a knowledge-based economy.3 These factors are extremely mobile and can be allocated in a tax-efficient manner at the will of MNEs. Moreover, the physical presence of production factors plays a vital role for determining the allocation of taxing rights on business profits with the paradigm being the permanent establishment (PE) threshold contained in almost all tax treaties in force. Yet, currently, several businesses can be heavily involved in the economic life of different jurisdictions without meaningful physical involvement.4 Undeniably, the digitalization of the economy has exacerbated the stated issues.5 Additionally, the definition of a taxable person is primarily contingent upon the individual components of MNEs, leading to the taxation of each separate entity’s profits. Residence, which is a relevant nexus in this respect, is often defined with criteria that are highly malleable and equally prone to electiveness.6 Moreover, the arm’s-length 2 See

M. F. de Wilde, Sharing the Pie: Taxing Multinationals in a Global Market (2017), 488ff. See also V. Chand, ‘Allocation of Taxing Rights in the Digitalized Economy: Assessment of Potential Policy Solutions and Recommendation for a Simplified Residual Profit Split Method’, Intertax 47 (2019), 1023, 1024–1206 and literature cited therein. 3 See OECD, Tax Challenges Arising from Digitalisation—Interim Report (2018), 34–42. See also R. J. S. Tavares, ‘Multinational Firm Theory and International Tax Law: Seeking Coherence’, World Tax Journal 8 (2016), 243, 246. S. Beer et al., ‘Exploring Residual Profit Allocation’, IMF Working Paper (2020), 20–21. 4 See OECD, Tax Challenges Arising from Digitalisation—Interim Report (2018), 51. In the EU, see European Commission, ‘A Fair and Efficient Tax System in the European Union for the Digital Single Market’, COM(2017) 547 final (2017), 7. In the UK, see HM Treasury, ‘Corporate tax and the digital economy: position paper’ (2017), 8–9. 5 See M. P. Devereux and J. Vella, ‘Implications of Digitalization for International Corporate Tax Reform’, in S. Gupta, et al., eds, Digital Revolutions in Public Finance (International Monetary Fund, 2017). 6 See a detailed account in E. Escribano, Jurisdiction to Tax Corporate Income Pursuant to the Presumptive Benefit Principle (Kluwer Law, 2019), 63ff.

Allocation of Taxing Rights 953 standard, which is the dominant profit-allocation proxy between related enterprises, serves as a catalyser for a tax-driven allocation of production factors among the said components of MNEs. For instance, risk assumption, which highly influences the generation of returns, may be easily reallocated through contracts.7 As a result, several businesses have been able to benefit from the described status quo and access markets with a relatively low tax cost, through commissionaire structures—addressed in the Base Erosion and Profit Shifting (BEPS) Project8—and limited risk distributors.9 The combination of these features led to an outcome in which market countries were only able to minimally tax the revenues generated by MNEs while their profits were not taxed (or were low taxed) elsewhere.10 Furthermore, electivity fosters competition between jurisdictions to attract investment through tax-related incentives resulting in a ‘beggar-thy-neighbour’ approach to attract investment at the expense of tax collection.11 Competition not only generates overall revenue losses but also introduces distortions to an efficient allocation of entrepreneurial resources.12 A normative evaluation of the described framework based on proxies such as revenue raising, fairness, efficiency, and simplicity of administration indicates its defectiveness.13 Yet, the unanimous claim for a revision of the allocation of taxing rights on business profits can only be understood as a result of political momentum without which Pillar One would have been inconceivable. Neither the 1998 Ottawa Taxation Framework14 nor the 2005 OECD report on electronic commerce concluded that the international tax framework had to be redesigned.15 Not even the BEPS Project was ‘directly aimed at changing the existing international standards on the allocation of taxing rights on cross- border income’.16 At a different level, claims to address the fact that the existing regime has significantly favoured the interests of developed countries were never thoroughly

7 See R. J. Vann, ‘Taxing International Business Income: Hard-Boiled Wonderland and the End of the World’, World Tax Journal 2 (2010), 291, 333. See also W. Schön, ‘International Taxation of Risk’, Bulletin for International Taxation 68 (2014), 280. 8 See OECD, Preventing the Artificial Avoidance of Permanent Establishment Status. Action 7—2015 Final Report (2015). 9 See A. Musselli and A. Musselli, ‘Stripping the Functions of Affiliated Distributors’, International Transfer Pricing Journal 15 (2008), 262. 10 W. Schön, ‘Ten Questions about Why and how to Tax the Digitalized Economy’, Bulletin for International Taxation 72 (2018), 278, 279. See also A. Christians and T. D. Magalhães, ‘A New Global Tax Deal for the Digital Age’, Canadian Tax Journal 67 (2019), 1153, 1162. 11 See R. S. Avi-Yonah, ‘Globalization, Tax Competition, and the Fiscal Crisis of the Welfare State’, Harvard Law Review 113 (2000), 1573. R. Altshuler and H. Grubert, ‘The Three Parties in the Race to the Bottom: Host Governments, Home Governments and Multinational Companies’, Florida Tax Review 7 (2005), 153; T. Dagan, ‘International Tax and Global Justice’, Theoretical Inquiries in Law 18 (2017), 1, 13ff. 12 M. P. Devereux et al., Taxing Profit in a Global Economy (Oxford University Press, 2021), 44. 13 Ibid., 113ff. 14 OECD, ‘Electronic commerce: taxation framework conditions’ (1998). 15 OECD, ‘E-Commerce: Transfer Pricing and Business Profits Taxation’ (2005). 16 OECD, ‘Action Plan on Base Erosion and Profit Shifting’ (2013), 11.

954 Aitor Navarro addressed,17 even when sound policy reasons justified otherwise. In the international tax arena, political impulse is thus critical. Specifically, political momentum may be traced back to the economic downturn of 2007–2008.18 The adopted fiscal measures left several states with the need to raise revenue. Additionally, a series of scandals involving the tax-planning schemes of multinational and high net worth individuals created political breeding grounds to instigate a reform of the international tax regime.19 The BEPS Project, fostered by the OECD and the G20, emerged as a reaction to the ‘perception that the domestic and international rules on the taxation of cross-border profits are now broken and that taxes are only paid by the naïve’.20 Meaningful results were achieved,21 however the fundamental issues described earlier remained unaddressed. Indeed, the final report on the tax challenges generated by the digital economy (Action 1) was of a descriptive nature and did not contain specific proposals on the subject.22 Meanwhile, political pressure on the issue of MNE taxation derived from noteworthy reactions such as the opening of state aid infringement procedures in the EU,23 the expansion of source-taxing rights on DTCs,24 or the enactment of unilateral measures to tax digitalized business models,25 among others. These outcomes combined with a strong response by the Trump administration against measures that allegedly discriminate US MNEs26 placed pressure on the OECD to produce a consensus-based solution.27 After a disordered examination of proposals that never were thoroughly 17 A. Martín Jiménez, ‘BEPS, the Digital(ized) Economy and the Taxation of Services and Royalties’, Intertax 46 (2018), 620, 623; T. Dagan, ‘Tax Treaties as a Network Product’, Brooklyn Journal of International Law 41 (2016), 1081, 1101–1104. Cf. Y. Brauner, ‘What the BEPS’, Florida Tax Review 16 (2014), 55, 63–64. 18 See Devereux et al., Taxing Profit in a Global Economy, 4ff. 19 See an examination of these scandals and the resulting ‘leak-driven lawmaking’ in S. Oei and D. Ring, ‘Leak-Driven Law’, UCLA Law Review 65 (2018), 532. 20 OECD, ‘Addressing Base Erosion and Profit Shifting’ (2013), 13. 21 For an account, see R. Mason, ‘The Transformation of International Tax’, American Journal of International Law 114 (2020), 353. 22 OECD, Addressing the Tax Challenges of the Digital Economy, Action 1—2015 Final Report (2015). 23 See the proceedings of the European Commission in Apple (SA.38373), Amazon (SA.38944), Fiat (SA.38375), McDonald’s (SA.38945), Starbucks (SA.38374), and Ikea (SA.46470). See an analysis in S. Buriak and I. Lazarov, ‘Between State Aid and the Fundamental Freedoms: The Arm’s Length Principle and EU Law’, Common Market Law Review 56 (2019), 905. 24 See, e.g., UN, Model Double Taxation Convention between Developed and Developing Countries (2017), art. 12A, advocating for source-taxing rights on fees for technical services. 25 These measures range from the equalization levy adopted in India to the resulting differing versions of digital services taxes by a handful of European countries, among others. See Báez Moreno and Brauner, ‘Taxing the Digital Economy Post BEPS ... Seriously’, 162ff. See also A. Christians and T. D. Magalhães, ‘The Rise of Cooperative Surplus Taxation’ (2020), http://dx.doi.org/10.2139/ssrn.3687011 (30 March 2021). 26 See R. Mason and L. Parada, ‘The Legality of Digital Taxes in Europe’, Virginia Tax Review 40 (2020), 175. 27 OECD, Programme of Work to Develop a Consensus Solution to the Tax Challenges Arising from the Digitalisation of the Economy (2019), 7, on unilateral measures taxing digital services: ‘This dissatisfaction has created a political imperative to act.’

Allocation of Taxing Rights 955 considered,28 the OECD focused its resources on embarking on a unified solution drafted by its Secretariat.29 This so-called ‘Unified Approach’ comprises two pillars. Pillar One is the proposal analysed in the present chapter, and it focuses on revising existing profit allocation and nexus rules.30 Pillar Two consists of achieving minimum taxation for MNEs.31 Henceforth, a description of the Pillar One proposal and policy considerations related to its adoption are provided.

52.3 Elements of the Proposal This section describes the allocation of taxing rights proposed in Pillar One which comprises two components.32 Amount A consists of a sales-based formulary residual profit split to allow market jurisdictions to tax a small portion of in-scope MNEs’ profits.33 It addresses the need to revisit taxing rules in response to a changed economy in which ‘businesses can, with or without the benefit of local physical operations, participate in an active and sustained manner in the economic life of a market jurisdiction, through engagement extending beyond the mere conclusion of sales, in order to increase the value of their products, their sales and thus their profits’.34 Yet, this taxing right would be implemented on a stand-alone basis as an overlay of the existing international taxation rules.35 Amount B, on the other hand, aims to standardize the remuneration of related party distributors that perform baseline marketing and distribution activities to enhance certainty and reduce both controversy as well as compliance and administration costs.36 At the time this chapter was written,37 more than 136 countries had reached a political agreement on the adoption of Pillar One that aims at finishing all implementation stages 28 See OECD, Tax Challenges Arising from Digitalisation—Interim Report (2018); OECD, Addressing the Tax Challenges of the Digitalization of the Economy—Public Consultation Document (2019). For an account, see L. V. Faulhaber, ‘Taxing Tech: The Future of Digital Taxation’, Virginia Tax Review 39 (2019), 145. 29 Pascal Saint-Amans presented Richard Collier, Associate Fellow at Oxford’s Saïd Business School and senior OECD advisor, as the main drafter of the proposal. See OECD, ‘OECD Tax Talks 13’ (2019), https://www.oecd.org/tax/beps/tax-talks-webcasts.htm (accessed 9 November 2021). 30 OECD, Tax Challenges Arising from Digitalisation—Report on Pillar One Blueprint (2020) (the 2020 Blueprint). 31 OECD, Tax Challenges Arising from Digitalisation—Report on Pillar Two Blueprint (2020). 32 Measures to enhance certainty and implementation issues remain outside the scope of this chapter. 33 See an example—although built around the 2020 Blueprint proposal—in S. Greil and T. Eisgruber, ‘Taxing the Digital Economy: A Case Study on the Unified Approach for a Case Example’, Intertax 49 (2021), 53. 34 2020 Blueprint, 19. 35 Ibid., 66. 36 Ibid., 160. See I. Dykes and L. Keegan, ‘The OECD Pillar 1 Blueprint: Why Amount B Matters’, Bulletin for International Taxation 75 (2021), 124. See also H. Higinbotham et al., ‘Amount B: Facts and Circumstances Matter Even for Routine Distributors’, Tax Management International Journal 50 (2021), 1. 37 The last substantial review of the chapter took place on 11 May 2023.

956 Aitor Navarro by 2023.38 The components of Amounts A and B described hereafter are a combination of those reflected in the October 2020 Blueprint and the October 2021 Statement on a Two-Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy encapsulating the said agreement,39 as well as the multiple draft public consultation documents published in 2022. The final drafting of the specific rules may be subject to further changes.

52.3.1 Amount A—Taxing Right for Market Jurisdictions 52.3.1.1 Scope In-scope companies are MNEs with a global turnover above €20 billion and profitability (profit before tax/revenue) above 10%40. Extractives and regulated financial services are excluded.41 It is estimated that less than eighty MNEs will fall within the scope of Amount A.42

52.3.1.2 Nexus and revenue-sourcing rules The nexus rule will allow the allocation of Amount A to a market jurisdiction when the in-scope MNE derives at least €1 million in revenue from that jurisdiction. For smaller jurisdictions with a GDP lower than €40 billion, the nexus will be set at €250,000.43 Revenue is sourced to the end market jurisdictions where goods or services are used or consumed.44 Although the detailed rules are yet to be determined, the 2022 consultation document already provides detailed provisions on this matter. A sourcing rule is determined for each type of in-scope revenue accompanied by a list of the acceptable— and hierarchized—indicators that an MNE would use to apply the principle and identify the jurisdiction of source. The indicator that appears first in the hierarchy was to be used unless information on it was unavailable or unreliable. For instance, the sourcing of online advertising services based on the real-time location of the viewer is linked to the jurisdiction of the real-time location of the viewer of the advertisement, identified primarily through the user profile information of the viewer.45 An in-scope MNE should prepare documentation: (1) describing the functioning of its internal control framework

38 OECD, Secretary-General Tax Report to G20 Finance Ministers and Central Bank Governors. Italy October (2021), 4–5 (2021 Agreement). 39 Ibid., 10–16. 40 See OECD, Pillar One –Amount A: Draft Model Rules for Domestic Legislation on Scope (2022). 41 See OECD, Pillar One –Amount A: Extractives Exclusion (2022), and OECD, Pillar One –Amount A: Regulated Financial Services Exclusion (2022). 42 M. Devereux and M. Simmler, ‘Who Will Pay Amount A?’, EconPol Policy Brief 36 (2021). 43 2021 Agreement, 10. 44 Ibid., 11. See also OECD, Pillar One –Amount A: Draft Model Rules for Nexus and Revenue Sourcing (2022) (2022 Nexus and Revenue Sourcing). 45 2020 Blueprint, 74. 2022 Nexus and Revenue Sourcing, 16–17.

Allocation of Taxing Rights 957 related to revenue sourcing; (2) containing aggregate and periodic information on the results of applying the indicators for each type of revenue and in each jurisdiction; and (3) explaining the indicator used and, if relevant, why a secondary indicator was applied instead.46

52.3.1.3 Tax base calculation The relevant measure of profit or loss of the in-scope MNE will be determined by reference to financial accounting income, with a small number of adjustments47. Losses will be carried forward. The taxable base is calculated using the consolidated group financial accounts of the group in accordance with generally accepted accounting principles (GAAP) that produces equivalent outcomes to those of the international financial reporting standards (IFRS). Other GAAPs may be eligible on a case-by-case basis. No book-to-book harmonization adjustments are considered necessary, and only limited book-to-tax adjustments would apply for determining an adjusted profit before taxes (PBT) that will be used as a reference.48 Segmentation will occur only in exceptional circumstances where, based on the segments disclosed in the financial accounts, a segment meets the scope rules.49

52.3.1.4 Tax base allocation For in-scope MNEs, 25% of residual profit defined as profit in excess of 10% of revenue will be allocated to market jurisdictions with nexus using a revenue-based allocation key.50 The distribution of the allocable tax base among eligible market jurisdictions will probably be calculated by reference to the locally sourced revenue.51

52.3.1.5 Elimination of double taxation Different mechanisms are proposed to prevent or reduce double taxation.52 The main one would apply at the level of the paying entity that would be identified following a four-step process: (1) ascertain the entities that perform activities leading to a sustained contribution to the ability of the group to generate residual profits; (2) apply a profitability test to ensure the identified entities have the capacity to bear Amount A’s tax liability; (3) allocate, in order of priority, the Amount A tax liability to the entities that have a connection with the market(s) where Amount A is allocated; and (4) allocate, on a pro-rata basis, where no sufficiently strong connection(s) is (or are) found or where

46

2020 Blueprint, 96–98. See overall OECD, Pillar One –Amount A: Draft Model Rules for Tax Base Determinations (2022). 48 2020 Blueprint, 105–109. 49 2021 Agreement, 11. 50 Ibid. 51 2020 Blueprint, 126–129. 52 Ibid., 132–159. 47

958 Aitor Navarro those with a market connection lack the necessary amount of profit. Either exemption or credit could be applied to eliminate double taxation.

52.3.2 Amount B—Fixed Remuneration for Baseline Marketing and Distribution Activities 52.3.2.1 Scope Defined in accordance with a positive list of typical functions performed, assets owned, and risks assumed at arm’s length by routine distributors and a negative list that reflects those that are intended to not be covered.53 For instance, an entity purchasing goods for resale within the market that owns or leases warehousing facilities and holds limited market and inventory risks would be in-scope for Amount B. On the contrary, an entity performing activities related to the development, enhancement, maintenance, or protection of marketing intangibles would be out of scope. Additionally, quantitative indicators would be used to ascertain whether the entity has the profile of a routine distributor or if it performs additional activities and assumes additional risks that may render it out of scope. This is primarily based on the assumption of costs such as marketing, advertising and R&D expenses, inventory write-downs, and others.54

52.3.2.2 Quantum A fixed return on sales to remunerate baseline marketing and distribution activities under Amount B is intended to deliver an outcome that approximates results determined in accordance with the arm’s-length standard.55 Such a return would vary corresponding to the specific features of the region and industry where in-scope entities operate as well as the functional intensity that is displayed. Further work is required to define these aspects, the benchmarking process, and the update frequency of the adopted references.56

53.3.2.3 Implementation The proposal advocates for a narrow scope of Amount B that would prevent conflicts with existing domestic law and treaties.57 In fact, the implementation of the proposal could occur at the level of domestic regulations.

53

Ibid., 162–169. See also OECD, Pillar One –Amount B (2022). 2020 Blueprint, 166–168. 55 Ibid., 169–170. 56 Ibid., 170–171. 57 Ibid., 172. 54

Allocation of Taxing Rights 959

52.4 A normative Evaluation of Pillar One: Compromise as the Main Policy Concern The term that best describes the rationale of Pillar One is ‘compromise’. The consensus- oriented approach adopted by the OECD—leaning on the Inclusive Framework as a catalyser—aimed at satisfying all relevant parties.58 A clear indicator refers to the fact that the design of Amount A was built partially following the rationale of three previous proposals:59 (1) The user-participation proposal, championed by the UK, was premised on the concept that soliciting the sustained engagement and active participation of users is a critical component of value creation for certain highly digitalized businesses—specifically, social media platforms, search engines, and online marketplaces. It would allocate part of the residual profit to the jurisdictions where those businesses’ active and participatory user bases are located. (2) The significant economic presence proposal, raised by the G24, seeks to expand the nexus beyond physical presence. Taxing rights would then be reallocated on a fractional apportionment basis with user participation added to the traditional apportionment factors, explicitly sales, assets, and employees.60 (3) The marketing intangibles proposal, supported by the USA, would require marketing intangibles and risks associated with such intangibles to be allocated to market jurisdictions. These would be entitled to tax some or all of the non- routine income properly associated with such intangibles and their attendant risks while all other income would be allocated among members of the group based on existing transfer pricing principles. As shown in the previous section, Amount A replicates different features of these proposals. It allows taxation in the market jurisdiction beyond physical presence and beyond digitalized business models by means of a residual profit split that relies on 58 See A. P. Dourado, ‘The OECD Unified Approach and the New International Tax System: A Half-Way Solution’, Intertax 48 (2020), 3, 7. See also S. Buriak, ‘A New Taxing Right for the Market Jurisdiction: Where are the Limits?’, Intertax 48 (2020), 301, 302. 59 The proposals were depicted in OECD, Addressing the Tax Challenges of the Digitalization of the Economy—Public Consultation Document (2019), 9–23. See Chand, ‘Allocation of Taxing Rights in the Digitalized Economy’. See also C. Elliffe, ‘International Tax Frameworks: Assessing the 2020s Compromise from the Perspective of Taxing the Digital Economy in the Great Lockdown’, Bulletin for International Taxation 74 (2020), 532, 541. 60 See G24 Working Group on Tax Policy and International Tax Cooperation, ‘Proposal for Addressing Tax Challenges Arising from Digitalization’ (2019), https://www.g24.org/wp-content/uplo ads/2019/03/G-24_proposal_for_Taxation_of_Digital_Economy_Jan17_Special_Session_2.pdf.

960 Aitor Navarro formula-based calculations. In fact, it extends the taxing rights of market jurisdictions further than a physical nexus and an arm’s-length allocation of taxing rights as projected in the three proposals. Hence, the goals of the proponents are satisfied to a certain degree. Moreover, the fact that Amount A would be additionally affixed to the existing international tax superstructure, combined with the rather modest reallocation of taxing rights it entails, should reduce the reluctancy of production countries, mainly the USA. Similarly, MNEs will still be able to profit from the tax-efficient allocation of production factors that the ‘old regime’ permits while having to bear a relatively minimal increase of the total tax burden. The design of the proposal also reveals the need to forge a product that the main players— the USA and the European countries— consider tolerable.61 European countries were willing to tax US digitalized MNEs more than the current rules allow. To tip the balance, the proposal incorporated several elements that are clearly pointing towards increasing the options for including the USA in the agreement. In fact, a rejection of the agreement by the North American country would most probably imply its failure.62 First, as stated, it maintains the prevalence of the ‘old regime’ that has benefited US MNEs for decades, including the central role that the arm’s-length principle plays as a profit-allocation mechanism.63 In that respect, Pillar One may be described as a vaccine inoculating a small portion of market taxation to induce immunity and achieve the preservation of the old rules. Secondly, Amount B will significantly reduce the risk for tax assessments directed at putting limited risk-distribution schemes into question that are frequently used by US MNEs to access markets while escaping taxation of economic rents therein. Thirdly, the scope of Amount A does not only cover high tech MNEs—mainly US groups—but any MNE above the threshold pinpointed in the previous section in order to avoid claims of discrimination. Fourthly, the agreement on Pillar One would entail the removal of those unilateral measures taxing the digitalized economy contested by the USA.64 Fifthly, the proposal would entail a significant boost for mandatory dispute-resolution mechanisms,65 the primary promoter of which has

61

See A. Christians, ‘A Unified Approach to International Tax Consensus’, Tax Notes International 96 (2019), 497, 499. 62 Estimates show that 63.8% of Amount A profits belong to companies headquartered in the USA. See Devereux and Simmler, ‘Who Will Pay Amount A?’, 5. 63 See J. C. Fleming Jr, R. J. Peroni, and S. E. Shay, ‘Worse than Exemption’, Emory Law Journal 59 (2009), 79, 119ff. 64 2020 Blueprint, 211; See OECD, Pillar One –Amount A: Draft Multilateral Convention Provisions on Digital Services Taxes and other Relevant Similar Measures (2022). On 21 October 2021, Austria, France, Italy, Spain, and the UK issued a joint statement with the USA by which they agreed to withdraw their digital service taxes at the time Pillar One takes full effect. The Joint Statement can be accessed at http://www.lamoncloa.gob.es/serviciosdeprensa/notasprensa/hacienda/Documents/2021/211021-acue rdo_impuestos_digitales.pdf. 65 See 2021 Agreement, 11, where it is stated that in-scope MNEs will benefit from dispute-prevention and resolution mechanisms in a mandatory and binding manner. The only carve-out allowed in the agreement refers to developing economies that are eligible for deferral of their BEPS Action 14 peer review 4 and have no or low levels of mutual agreement procedure disputes.

Allocation of Taxing Rights 961 been the USA in previous years.66 In addition to these points, it must be highlighted that all the proposals to modify the 2020 Blueprint version of Amount A suggested by the USA were adopted in the October 2021 agreement. The most notable changes consisted in the significant elevation of the revenue threshold from €750 million to €20 billion and the elimination of the distinction between automated digital services (ADS) and consumer-facing businesses (CFB) that was critical in defining the scope and nexus of the 2020 Blueprint proposal.67 The significant effort to integrate different views must be praised as it is not an easy task to reconcile the interests of the involved countries to achieve a consensus solution. However, defining what consensus entails in the context of the Pillar One proposal is relevant. As Christians highlights, ‘priority is on time to consensus, and not process to consensus, certainly not on normative assessment of the substance of consensus’.68 Definitely, the fact that consensus is critical for the feasibility of the proposal should neither camouflage legitimacy concerns69 nor the difficulties of fitting the Pillar One proposal within an appropriate set of normative standards.70 In fact, hereafter, a brief contrast of Pillar One with long-standing normative canons on taxation matters (i.e. fairness, efficiency, and administrability) will be made.71 The examination will begin with Amount A, which encompasses the most significant rupture from current tax- allocation standards. (1) From a fairness perspective, two dimensions must be distinguished. First, the appropriateness of allocating taxing rights on business income to market states from an inter-nation equity perspective. The well-known paradigms of economic allegiance, the benefit principle, and the concept of value creation—as overlapping as they are— would support such an allocation. Specifically, economic allegiance entails the distribution of taxing rights in accordance with the level of the taxpayer’s economic interests

66 See Y. Brauner, ‘International Tax Treaty Dispute Resolution, the Mutual Agreement Procedure, and the Promise of Mandatory Arbitration for Developing Countries’, in Y. Brauner, ed., Research Handbook on International Taxation (Edward Elgar Publishing, 2020), 205. 67 Cf. 2020 Blueprint, 19–64 and the slides prepared by US representatives in the meeting with the steering group of the Inclusive Framework, available at https://mnetax.com/wp-content/uploads/2021/ 04/US-slides-for-Inclusive-Framework-meeting-of-4-8-21-2.pdf. The US proposal followed certain suggestions posed in M. J. Graetz, ‘A Major Simplification of the OECD’s Pillar 1 Proposal’, Tax Notes International 101 (2021), 199. 68 See Christians, ‘A Unified Approach to International Tax Consensus’ , 500. 69 See I. Ozai, ‘Institutional and Structural Legitimacy Deficits in the International Tax Regime’, World Tax Journal 12 (2020), 53. 70 In OECD, Programme of Work to Develop a Consensus Solution to the Tax Challenges Arising from the Digitalisation of the Economy (2019), 7, it is stated that the solution to be adopted ‘must be underpinned by sound economic principles and conceptual basis’, yet the Pillar One documentation produced by the OECD lacks a normative evaluation on the soundness of the proposal. 71 For an account of the impact of these standards on international taxation matters, see S. Gadžo, Nexus Requirements for Taxation of Non-Residents’ Business Income (IBFD, 2018). For an evaluation of the current international tax regime in accordance with these normative standards, see Devereux et al., Taxing Profit in a Global Economy, 113ff.

962 Aitor Navarro in each sovereign state competing for the tax base.72 The benefit principle requires that each taxpayer’s contribution is in line with the benefits that they receive from public services.73 The value-creation notion refers to the need that income should be taxed where economic activities deriving the profits are performed and where value is created.74 Undoubtedly, MNEs have strong economic interests in market countries, benefit from the public sphere therein, and their products or services carry no value unless there is customer demand.75 That being said, the use of these fairness parameters to justify the taxation of business profits by market jurisdictions is rather unsatisfactory. These concepts justify taxation in market countries as much as they may be used to support taxation in production countries because of the exact same noted reasons.76 Furthermore, the mentioned fairness parameters would justify the maintenance of the current status quo as market jurisdictions are already able to tax added value at market by means of destination-based consumption taxes.77 On top of everything, once it is decided that market jurisdictions should have the possibility to tax business income, the mentioned fairness parameters are useless for determining the allocation of the tax base.78 If inter-nation equity is approached from the perspective of revenue raising and redistribution among states due to the implementation of Amount A, the implications are rather modest. Amount A would entail slight changes on taxing rights among a few key states79 and a minimal increase of worldwide corporate tax revenues, between US$13 and US$36 billion in absolute numbers.80 The OECD decided not to publish per- country estimations to avoid disturbing the ongoing negotiations. 72 See N. H. Kaufman, ‘Fairness and the Taxation of International Income’, Law and Policy in International Business 29 (1998), 145. 73 R. A. Musgrave and P. B. Musgrave, Public Finance in Theory and Practice (1989), 219. 74 The concept as defined was first mentioned in G20, G20 Leaders’ Declaration, St Petersburg (2013), 12–13, (http://www.g20.utoronto.ca/2013/Saint_Petersburg_Declaration_ENG.pdf). See an in- depth analysis of the value-creation notion in S. I. Langbein, and M. R. Fuss, ‘The OECD/G20-BEPS-Project and the Value Creation Paradigm: Economic Reality Disemboguing into the Interpretation of the “Arm’s Length’ Standard” ’, The International Lawyer 51 (2018), 259. Notwithstanding that its relevance in Pillar One must be relativized as no specific references to the concept from a policy perspective are made in the 2020 Blueprint. 75 Schön, ‘Ten Questions about Why and how to Tax the Digitalized Economy’, 285. 76 See Martín Jiménez, ‘BEPS, the Digital(ized) Economy and the Taxation of Services and Royalties’, 626. 77 On the VAT destination principle, see OECD, International VAT/GST Guidelines (2017), 15–17. Cr. J. Becker and J. Englisch, ‘Taxing Where Value is Created: What’s “User Involvement” Got to Do With It?’, Intertax 47 (2019), 161, 164–165 and literature cited therein. 78 See J. Hey, ‘ “Taxation Where Value is Created” and the OECD/ G20 Base Erosion and Profit Shifting Initiative’, Bulletin for International Taxation 72 (2018), 203, 205. 79 See Christians and Magalhães, ‘A New Global Tax Deal for the Digital Age’, 1157. See also L. Parada, ‘The Unified Approach Under Pillar One: An Early Analysis’, Tax Notes International 96 (2019), 983, 985. 80 The estimates were updated in 2023, although no report has been issued yet in this regard. Previous estimates elaborated in 2020 showed worldwide revenue gains ranging between US$5 and US$12 billion. See OECD, Tax Challenges Arising from Digitalisation—Economic Impact Assessment (2020), 15. For a critical review, see L. Eden, ‘Winners and Losers: The OECD’s Economic Impact Assessment of Pillar One’, Tax Management International Journal 49 (2020), 597.

Allocation of Taxing Rights 963 The second perspective on fairness that is worth mentioning refers to the ability- to-pay principle as a corollary of inter-taxpayer equity.81 As stated earlier, the point of departure to calculate the taxable base of Amount A refers to the financial accounting standard used in the ultimate parent jurisdiction for preparing consolidated financial statements. Eligible GAAPs are, in principle, those producing equivalent or comparable outcomes to the IFRS. Yet, other GAAPs are also to be allowed on a case-by-case basis.82 Thus, remarkably, the measurement of profit will vary depending on the location of the ultimate parent entity. Hence, two identical MNEs with different ultimate parent entity locations may display diverging profits. On the other hand, from the perspective of each market jurisdiction, Amount A would be determined by means of accounting standards that are not necessarily the same as those adopted in the domestic accounting regulations. Hence, from the perspective of each market jurisdiction, two identical MNEs that equally derive profits from users/consumers located in the same market may end up displaying differing figures regarding the allocable profit in that jurisdiction. Overall, for the sake of simplicity, neither a unified set of rules to determine the consolidated profit nor book-to-book adjustments are contemplated.83 Accordingly, the appropriateness of the said disparities should be measured as a trade-off between inter- taxpayer equity as projected through the ability-to-pay principle and administrability. Lastly, it must be stressed that ability to pay may be useful for determining the tax base but not for allocating it among different jurisdictions.84 Hence, overall, a benchmark cannot be derived to allocate taxing rights from a fairness perspective. (2) Efficiency demands minimizing the distortions that taxes create over economic choices and behaviour. As stated previously, the current international framework on business profits taxation is based on the location of production factors, and the main value drivers generating economic rents are currently intangible assets and human capital. Both may be easily relocated at will and for tax-driven reasons, hence the present status quo leads to inefficient outcomes. On the contrary, taxation based on immobile factors—such as the location of the market—is efficient, as MNEs have no power to decide where users and consumers are located.85 In this respect, the rationale of the source rules envisaged in Amount A accords with this logic, and it thus seems to meet efficiency ends.86 That notwithstanding, it must not be forgotten that Amount A is conceived as a stand-alone rule that applies alongside the current international tax regime. Hence, overall—and considering the relative impact of Amount A in the overall tax burden of 81

See Gadžo, Nexus Requirements for Taxation of Non-Residents’ Business Income, 202–205. 2020 Blueprint, 105. 83 Ibid. 84 W. Schön, ‘International Tax Coordination for a Second-Best World (Part I)’, World Tax Journal 1 (2009), 67, 73. 85 W. Schön, ‘One Answer to Why and how to Tax the Digitalized Economy’, Intertax 47 (2019), 1003, 1007; Devereux et al., Taxing Profit in a Global Economy, 168ff. 86 Yet, Eden considers that there is a margin of freedom for tax authorities and MNEs to manipulate Amount A components in a way that would affect the size of the gain or loss. See L. Eden, ‘Pillar One Tax Games’, Tax Management International Journal 50 (2021), 1, 8. 82

964 Aitor Navarro MNEs—the efficiency gains that Amount A may introduce are negligible vis-à-vis the international tax framework as a whole. However, efficiency also comprises the notion of neutrality over business decisions. In this respect, focusing on the design of Amount A, the main concern refers to the undesirability of ring-fencing. The OECD itself stated that the digitalization of the economy is a comprehensive phenomenon, and it would thus be difficult, if not impossible, to ring-fence the digital economy from the rest of the economy.87 Yet, Amount A presents multiple ring-fencing layers that have been already described in Section 52.3. Additionally, Amount A leads to a well-known phenomenon that may pose distortions contrary to the efficiency standard, namely the breach of the continuity requirement (i.e. going from full to zero taxation as a result of the decrease of one threshold unit).88 This would happen if an MNE with a turnover of exactly €20 billion decreases its revenue by €1. That said, the truth is that the negative consequences of ring-fencing should be weighted with the aim of each of the mentioned measures delimiting the scope and allocation of taxing rights to determine their appropriateness.89 For instance, it was claimed in the 2020 Blueprint that monetary thresholds are justifiable due to practicality reasons. It was considered that the compliance and administrative burden placed for both taxpayers and tax administrations should be commensurate with the benefits in terms of the available reallocation for market jurisdictions. The original €750 million threshold was adopted from the recommendation to adopt country-by-country regulations, was chosen because it ‘is estimated to exclude 85–90% of MNE groups while still covering MNE groups responsible for approximately 90% of global corporate revenues’.90 The final €20 billion threshold will obviously cover a significantly lower number of MNEs. (3) Administrability requires a tax system to be certain, convenient— easy to administer—and efficient from an enforcement-cost perspective. Yet, complexity is one of the most salient features of corporate taxation. Amount A is certainly not an exception, especially when compared to other proposals to address the taxation of the digitalized economy, such as source withholding taxes.91 Often, complexity is the irremediable consequence of a better constructed set of tax regulations.92 However, that is hardly the case for Amount A. In fact, as shown throughout this chapter, Amount

87 OECD, Addressing the Tax Challenges of the Digital Economy, Action 1—2015 Final Report (2015), 54.

88 Schön, ‘One Answer to Why and how to Tax the Digitalized Economy’, 99–101. See also Gadžo, Nexus Requirements for Taxation of Non-Residents’ Business Income, 218–219; Báez Moreno and Brauner, ‘Taxing the Digital Economy Post BEPS ... Seriously’, 170. 89 Schön, ‘Ten Questions about Why and how to Tax the Digitalized Economy’, 281. 90 OECD, Transfer Pricing Documentation and Country‑by‑Country Reporting, Action 13—2015 Final Report (2015), 21. No reference to how the OECD assessed these numbers is given. 91 See Báez Moreno and Brauner, ‘Taxing the Digital Economy Post BEPS ... Seriously’. 92 See S. D. Pollack, ‘Tax Complexity, Reform, and the Illusions of Tax Simplification’, George Mason Independent Law Review 2 (1994), 319, 321.

Allocation of Taxing Rights 965 A combines an odd mix of complexity and simplification.93 On the one hand, the definition and enforcement of the revenue-sourcing rules, as well as the determination of the paying entity and the resulting elimination of double taxation rules are rather intricate and account for a significant level of complexity. These aspects are probably the most complex in the proposal. On the other hand, a crucial aspect of Amount A, specifically the determination of the residual profits to be allocated to market jurisdictions, simply consists of a fixed percentage. Hence, the proposal renounces finetuned calculations such as the accurate delineation of routine profits or the determination of the specific value generated by users/consumers, among others, that were formed as part of certain proposals noted earlier.94 The OECD does not elaborate on why administrability justifies the simplification of specific components of the proposal while leaving others complex. Yet, over and above that, from the perspective of the international tax framework as a whole, Amount A undoubtedly increases complexity. The conception of Amount A as an extra layer of regulation on top of existing rules will generate additional compliance costs for taxpayers and tax authorities alike.95 Compared to Amount A, the aim of Amount B is rather modest, and its normative assessment much simpler. Amount B entails the use of a formula that attempts to replicate arm’s-length remuneration of related party distributors, performing baseline marketing and distribution activities. The only meaningful gain from a normative perspective vis-à-vis the old regime refers to administrability. By relying on fixed, predetermined returns, Amount B would enhance certainty and reduce controversy as well as decrease compliance and administration costs generally associated with transfer pricing regulations.

52.5 Final Remarks on the Feasibility of Pillar One and the Future of the International Tax Regime The success of the Pillar One proposal is rather uncertain. Aside from the fact that certain relevant components are still to be defined,96 success very much depends on the consensus of those countries where in-scope MNEs are located and the largest market countries, which are overlapping in a great deal of cases. Specifically, if the USA does 93 For

a quantitative measurement of complexity in Pillar One—based on the 2020 Blueprint proposal, see J. Colliard, L. Eden, and C. Georg, ‘Tax Complexity and Transfer Pricing Blueprints, Guidelines, and Manuals’, Tax Management International Journal 50 (2021), 1. 94 As stated by Schön, ‘the final outcome represents more a partial move towards formulary apportionment applying a “sales factor” than anything specifically related to the “digital economy” ’. See W. Schön, ‘Is There Finally an International Tax System?’, World Tax Journal 13 (2021), s. 3.1.2. 95 See Devereux et al., Taxing Profit in a Global Economy, 128. 96 Cf. the sections labelled ‘next steps’ in the 2020 Blueprint and the 2021 Agreement.

966 Aitor Navarro not adopt Pillar One, the project will seemingly collapse. Considering the pace of the negotiations and the fact that the implementation of Amount A requires agreements at a multilateral level accompanied by changes at the domestic law level,97 their foreseeable opposition means that the chances for success are outright minimal. Additionally, the last hard law multilateral attempt by the OECD has left a bittersweet taste. The Multilateral Convention (MLC) has been adopted by ninety-five jurisdictions and has already updated a significant number of tax treaties in force. Yet, some very relevant jurisdictions—such as the USA and Brazil—have not signed it, and wide adoption has come at the expense of multiple carve-outs.98 The MLC was directed at updating the existing double taxation convention network and not at reaching truly multilateral obligations. Pillar One, on the other hand, requires the achievement of a genuinely multilateral instrument that would fundamentally change the allocation of taxing rights among states, breaking with several long-standing principles of the international tax regime. The ambitiousness of the proposal is directly proportional to the scepticism that it raises. On a different level, it is still unclear whether countries that have enacted unilateral measures to address the taxation of the digitalized economy will agree to their removal, as the 2021 Agreement establishes. As Dourado states, the recently debated concept of value creation in its reference to users has not been taken into account in the proposal; an OECD policy decision to keep this intentional gap could lead to the maintenance or further enactment of unilateral taxes that are justified on user value creation.99 Additionally, Pillar One leaves the primary concern of developing countries, specifically the expansion of source taxation, unaddressed. This fact may compromise the stability of the new consensus if it is implemented as proposed. The Unified Approach truly constitutes a missed opportunity for the international community.100 It helps very little in mobilizing revenue for developing countries to meet the UN sustainable development goals.101 Concisely, there is a significant imbalance between the relevance of fundamental changes that Pillar One makes to the existing rationale of the allocation of taxing rights and the revenue collection derived from it. The fact that Pillar One applies alongside the existing ‘old regime’ implies that none of the flaws of the international tax regime are significantly dealt with. This fact leads to peculiar outcomes such as the de-legitimization of the arm’s-length principle as an adequate allocation standard, on the one hand, and its preservation, on the other hand. In fact, the so-called ‘2020 Compromise’ is more about maintaining the old ‘1920 Compromise’ than about fundamentally reforming it, at least 97 OECD, Tax Challenges Arising from Digitalisation—Report on Pillar One Blueprint (2020), 205ff. 98

See R. García Antón, ‘Multilateral Dynamics in Bilateral Settings: Back to Realpolitik’, British Tax Review 4 (2019), 462. 99 A. P. Dourado, ‘The OECD Unified Approach and the New International Tax System: A Half-Way Solution’, Intertax 48 (2020), 3. 100 Christians and Magalhães, ‘A New Global Tax Deal for the Digital Age’, 1157. 101 Cf. IMF, OECD, UN, World Bank, ‘Taxation & SDGs’ (2018), https://www.oecd.org/ctp/tax-global/ first-global-conference-of-the-platform-for-collaboration-on-tax-february-2018.pdf.

Allocation of Taxing Rights 967 regarding the true implications of Pillar One. Yet, an optimistic observer could reply that, once the fundamentals of old regime have been breached, the implementation of further changes to achieve a truly fair, efficient, and administrable set of international allocation of taxing rights standards may be more easily achieved. The implications of the proposal are yet to be seen.

Chapter 53

Effective Mi ni mum Tax ation unde r Pi l l a r T wo of th e OE C D Prop osal (‘ G l oBE ’ ) Joachim Englisch

53.1 Background and Evolution of the GloBE Initiative With the publication of the 15 Action Item Reports on Base Erosion and Profit Shifting (BEPS) in 2015,1 a comprehensive reform of international tax law seemed to have succeeded thanks to concerted efforts by the OECD and the G20. At the time, it was widely anticipated that international collaboration in tax matters would henceforth focus on the implementation of the numerous BEPS recommendations and standards. To this effect, the Inclusive Framework on BEPS was established, and emerging and developing countries that were not represented in the OECD/G20 were invited to join the process. However, the BEPS negotiations had only led to an agreement on measures to curb ‘aggressive’ international tax planning and tax avoidance. The conflict over a possible realignment of the traditional allocation of taxing rights in reaction to the digitalization of the economy was not resolved due to a veto by the USA. At the time, the OECD merely raised the prospect of new report on the subject in 2020. However, unilateral tax measures after 2015 and especially the US tax reform enacted in 2017 generated an unexpected momentum for further reform. The Inclusive Framework therefore decided to aim for an internationally consensus-based solution to meet the challenges of taxing the digitalized economy by 2020. Initially, the only item 1 OECD, Explanatory Statement, Base Erosion and Profit Shifting Project, 2015 Final Reports (2015), 13ff.

970 Joachim Englisch on the agenda of this work stream was a discussion on the reallocation of taxing rights in favour of market and user countries. In 2018, a German–French initiative introduced the idea of an international effective minimum tax of low-taxed foreign profits.2 While earlier such initiatives had not garnered much support within the EU or the OECD, the previous introduction of a form of international minimum taxation by the USA—the GILTI3 regime—had opened what their proponents considered to be a window of opportunity. The Inclusive Framework quickly decided to analyse this option. However, it did so not alternatively but additionally, in parallel to its deliberations on new nexus criteria and profit-allocation rules in favour of market jurisdictions. This established the duality of ‘Pillar One’ and ‘Pillar Two’. The second pillar on minimum taxation was initially envisaged as a complementary anti-BEPS instrument, in contrast to the first pillar on a genuine realignment of the international tax system.4 This was reflected in the name for the new minimum tax concept: ‘GloBE’, as a short form for the ‘Global anti-Base Erosion’ proposal.5 This notwithstanding, an additional and more far-reaching objective of international minimum taxation was soon also mentioned. It could not only make BEPS less attractive, but also establish a floor for international tax competition for foreign direct investment.6 Work on the GloBE concept then progressed swiftly. Agreement on key design features was reached relatively quickly, as evidenced by the detailed October 2020 Blueprint for Pillar Two.7 While certain technical aspects as well as the crucial issue of the minimum tax rate still needed to be resolved at this time, political agreement was largely secured by then. However, it did not yet translate into an official endorsement of the GloBE proposal, reportedly because some countries refused to have an agreement on Pillar Two without simultaneous agreement on the more controversial Pillar One. A package compromise on the conceptual cornerstones of both pillars was then reached by 130 member countries of the Inclusive Framework on 1 July 2021. Notably, it was agreed that unlike the Pillar One solution, GloBE will not constitute a minimum 2

See Franco–German joint declaration on the taxation of digital companies and minimum taxation, https://w ww.consilium.europa.eu/media/37276/fr-de-joint-declaration-on-the-taxation-of-digital- companies-final.pdf (accessed 22 March 2021); OECD, Report on the GLOBE BEPS 2.0 Initiative (2018) (on file with the author). 3 Global intangible low-taxed income regime, see Internal Revenue Code, s. 951A. 4 See OECD/ G20 Base Erosion and Profit Shifting Project, ‘Addressing the Tax Challenges of the Digitalisation of the Economy—Policy Note’ (2019), 2, available at https://www.oecd.org/tax/beps/pol icy-note-beps-inclusive-framework-addressing-tax-challenges-digitalisation.pdf. 5 OECD/G20 Inclusive Framework on BEPS, ‘Programme of Work to Develop a Consensus Solution to the Tax Challenges Arising from the Digitalisation of the Economy’ (2019), 25, available at https:// www.oecd.org/tax/beps/programme-of-work-to-develop-a-consensus-solution-to-the-tax-challenges- arising-from-the-digitalisation-of-the-economy.pdf. 6 Ibid., 25ff; OECD, ‘Global Anti-Base Erosion Proposal (‘GloBE’)—Pillar Two, Public Consultation Document’ (2019), 6, available at https://www.oecd.org/tax/beps/public-consultation-document-global- anti-base-erosion-proposal-pillar-two.pdf.pdf. 7 OECD/G20 Base Erosion and Profit Shifting Project, ‘Tax Challenges Arising from Digitalisation— Report on Pillar Two Blueprint: Inclusive Framework on BEPS’ (2020), available at https://doi.org/ 10.1787/0e3cc2d4-en.

Effective Minimum Taxation 971 standard to be implemented by all signatory states. Instead, it is conceived as a ‘common approach’ aimed at the international coordination of what quintessentially remain domestic tax measures: member countries of the Inclusive Framework that opt for the implementation of an international effective minimum tax regime are politically committed to adhere to the GloBE concept, model rules, and guidance; member countries that do not implement GloBE merely need to ‘accept’ its application by other jurisdictions. This notwithstanding, on 1 July 2021nine member countries of the Inclusive Framework still abstained from supporting this approach, including three EU member states. After additional rounds of negotiation, the minimum rate was fixed at 15% and a more generous transition carve-out for substance-based activities was provided during a prolonged transition phase. On this basis, a refined political agreement was reached on 8 October 2021, among 136 member jurisdictions including all former EU holdout countries.8 In December 2021, the Inclusive Framework adopted the detailed ‘GloBE Model Rules’ as model legislation for the implementation of the minimum tax in domestic legal systems.9 These Model Rules are largely based on the concepts and rules already outlined in the Blueprint from 2020, but they also featured a few innovations. The Model Rules are further fleshed out by a Commentary that was published in March 2022.10 Moreover, it was envisaged that an ‘implementation framework’ would be developed by the end of 2022, dealing with administrative procedures including, importantly, certain simplification measures and safe harbours. It is also under consideration whether the coordinated implementation of GloBE should be further enhanced by a multilateral convention, which unlike the 2021 model rules and guidance would create ‘hard’ international public law obligations for its signatory states. Finally, certain necessary tax treaty amendments that will be discussed more in detail later will be facilitated by model treaty provisions and a multilateral instrument. Just two days after the publication of the GloBE Model Rules, the EU Commission tabled a proposal for a directive on the harmonized implementation of the international minimum tax in all EU member states.11 In general, the proposal is closely aligned with 8

See OECD/G20, ‘Inclusive Framework on BEPS, Statement of 136 member jurisdictions on a Two- Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy’ (8 Oct. 2021). Meanwhile, 137 of the 140 members joined the October 2021 Statement, see https://www.oecd.org/ tax/beps/oecd-g20-inclusive-framework-members-joining-statement-on-two-pillar-solution-to-addr ess-tax-challenges-arising-from-digitalisation-october-2021.pdf (accessed 9 November 2021). 9 OECD/ G20 Base Erosion and Profit Shifting Project, ‘Tax Challenges Arising from the Digitalisation of the Economy—Global Anti-Base Erosion Model Rules (Pillar Two)’ (2021), available at https://www.oecd.org/tax/beps/tax-challenges-arising-from-the-digitalisation-of-the-economy-global- anti-base-erosion-model-rules-pillar-two-commentary.pdf. 10 Base Erosion and Profit Shifting Project, ‘Tax Challenges Arising from the Digitalisation of the Economy—Commentary to the Global Anti-Base Erosion Model Rules (Pillar Two)’ (2022) available at https://www.oecd.org/tax/beps/tax-challenges-arising-from-the-digitalisation-of-the-economy-global- anti-base-erosion-model-rules-pillar-two-commentary.pdf. 11 Commission, Proposal for a Council Directive on ensuring a global minimum level of taxation for multinational groups in the Union, COM(2021) 823 final (22 Dec. 2021).

972 Joachim Englisch the international model. However, it also provides for a mandatory extension of the minimum top-up tax for certain domestic group entities: when a parent company of the respective entity—not necessarily the ultimate parent of the multinational group— is located in a member state, and in the case of purely domestic groups with annual sales above the GloBE threshold of €750 million. This addition to the international agreement seeks to ensure the compatibility of the minimum tax with the freedom of establishment in article 49 Treaty on the Functioning of the European Union (TFEU). Initially, the required unanimity in the Council had not been reached; the adoption of the proposed directive failed first because of a Polish veto and then due to a Hungarian veto. However, this blockade was eventually resolved in the second half of 2022, and the proposed Directive has been adopted. In the USA, Congress (narrowly) adopted a corporate alternative minimum tax (CAMT)12 that would complement GILTI so as to bring the overall regime closer to the agreement on the GloBE model. But the CAMT, too, still differs from the GloBE concept, especially because it allows worldwide blending of profits and taxes instead of the country-by-country application of the minimum tax as envisaged by the GloBE rule set. It therefore remains to be seen whether, at least transitionally, CAMT will be accepted as a GloBE equivalent for the purpose of coordinating the collection of the minimum tax. From a principled perspective, it should not. In the rest of the world, only the UK and South Korea had drafted legislation to implement GloBE by July 2022. Against this background, it is not realistic to expect a broad application of GloBE- style minimum taxes before 2024, if at all, contrary to the ambitious 2023 timeline set in the October 2021 Statement of the Inclusive Framework. Meanwhile, an alternative minimum tax approach has been proposed in the literature,13 but its chances of being adopted instead of GloBE are now practically zero on the grounds of both political dynamics and its administrative complexity.

53.2 General Analysis 53.2.1 Concept The GloBE concept calls for a ‘top-up tax’ on foreign-sourced profits of large MNEs whose effective tax rate (ETR) is below the politically agreed minimum rate of 15%. For the purpose of calculating the ETR, the profits of all group entities domiciled or established in a particular tax jurisdiction will be determined as an aggregate amount by reference to financial accounting income, with certain adjustments. They would then be assessed in relation to the total burden of income taxes and similar ‘covered taxes’ 12

See Internal Revenue Code, ss. 55ff. S. Picciotto et al., ‘For a Better Globe— METR: A Minimum Effective Tax Rate for Multinationals’ (2021), available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3796030. 13 See

Effective Minimum Taxation 973 attributable to those profits. If the resulting ETR is below the minimum tax rate, the difference will determine the potentially applicable top-up tax rate. However, top-up tax will be collected only on ‘excess profits’, defined as the difference between the GloBE income calculated for a particular jurisdiction and the so-called ‘substance-based income exclusion’ (SBIE). The SBIE refers to a formulary carve-out of an amount of profit deemed to represent a routine return of investments in certain human resources and tangible assets. Moreover, the ensuing GloBE top-up tax burden on excess profits would be reduced by any eventual top-up tax collected by way of a qualified domestic minimum top-up tax (QDMTT). The latter is defined as a replica of GloBE applied by a source country to any undertaxed excess profits of its own domestic MNEs. Effectively, this means that a domestic collection of a GloBE-style minimum tax takes precedence over the top-up tax collection elsewhere.14 To the extent that the international top-up tax is not soaked up by a QDMTT, two basic mechanisms are foreseen for its collection: an income inclusion rule (IIR) if a parent company is liable to pay top-up tax, or an undertaxed payment rule (UTPR) if the tax is collected from another constituent entity of the group. Unlike in the earlier stages of discussion,15 the international agreement no longer conceives the IIR and the UTPR as separate minimum tax instruments with different approaches regarding scope and ETR calculation. Rather, they merely represent different forms of collecting the top-up tax, which is calculated uniformly as outlined in the preceding paragraph.16 In order to avoid double (minimum) taxation and facilitate administration and compliance, priority is given to the IIR with a top-down approach in the case of several tiers of parent companies. The UTPR serves as a backstop to the IIR in case that the parent jurisdiction(s) did not implement GloBE, or if they are themselves low-tax jurisdictions. It will not apply to MNEs ‘in the initial phase of their international activity’.17 This also has certain implications for international GloBE revenue allocation. The chosen rule order tends to favour the developed industrialized countries and classic holding locations, while developing countries will hardly ever be able to generate revenue from the prioritized IIR. However, as stated earlier, every country can at least theoretically avoid this treasury-transfer effect by implementing its own prioritized QDMTT. In practice, some developing countries might still find this difficult, for example due to a lack of the administrative capacity needed to enforce the required and complex quasi-GloBE rules, or because it is legally constrained by its investment agreements.18 In any event, the Inclusive Framework considered it necessary to add 14 See B. Arnold, ‘The Ordering of Residence and Source Country Taxes and the OECD Pillar Two Global Minimum Tax’, Bulletin for International Taxation 76 (2022), 218, 224–225. 15 See, e.g., OECD/G20 Base Erosion and Profit Shifting Project, ‘Addressing the Tax Challenges of the Digitalisation of the Economy, Public Consultation Document’ (2019), paras 92ff, available at https:// www.oecd.org/tax/beps/public-consultation-document-addressing-the-tax-challenges-of-the-digital isation-of-the-economy.pdf; OECD, ‘Programme of Work’, 25ff. 16 See also the Model Commentary on art. 2, paras 68–69. 17 As defined in the ‘Statement of 136 member jurisdictions’, 4. 18 On investment agreements, see UNCTAD, World Investment Report 2022, 151.

974 Joachim Englisch another element to the international agreement in order to reach a more geographically balanced outcome: a minimum withholding tax on certain categories of outgoing payments to related companies that are particularly susceptible to BEPS. If the income resulting from the payment is taxed below a minimum rate of 9% in the jurisdiction where the recipient is established, a treaty-based ‘subject to tax rule’ (STTR) would allow developing countries to levy a withholding top-up tax. The ensuing tax burden would be included in the calculation of the ETR for the purposes of IIR and UTPR, which is why the minimum withholding tax would effectively take precedence over the actual GloBE minimum taxation. Unlike the core GloBE concept, however, this kind of minimum taxation would be prompted by a too low nominal tax rate in the jurisdiction of the recipient of the payment, adjusted only for eventual beneficial regimes that apply specifically to the relevant category of income. Moreover, the withholding tax would be levied on gross income rather than net income. These simplifications cause the minimum withholding tax to be less accurate, but they take into account the limited administrative capacities of many developing countries that would apply the STTR.

53.2.2 Policy Rationale As mentioned earlier, the original German–French GloBE initiative primarily intended to make BEPS less attractive for MNEs, thereby complementing the more specific measures recommended in the 2015 BEPS Action Item Reports. Unlike the latter, it was not conceived to directly address abusive or ‘aggressive’ arrangements, but instead to reduce the economic incentives for them.19 Arrangements that would push the tax burden below the minimum tax level by taking advantage of low-tax regimes or low-tax locations would no longer produce any financial gain for the MNE, irrespective of the structures or transactions used to benefit from the low effective tax rate. A second possible objective that emerged in the course of the deliberations within the Inclusive Framework and in public consultations in 2019 was to establish a floor for international tax competition in general. Depending on its design, a minimum tax could generally limit international tax arbitrage and thus the influence of different national ETRs on discrete investment decisions, beyond BEPS considerations. While a low level of business taxation would remain in the toolbox of countries to attract foreign direct investment, the use of this instrument would be moderated by the threshold of the minimum tax level. This second effect was considered desirable to reduce the adverse fiscal effects of ‘excessive’ international tax competition and the associated shift of tax burdens to labour or consumption.20

19 See specifically on transfer pricing, C. Elliffe, ‘OECD/ International— International Tax Frameworks: Assessing the 2020s Compromise from the Perspective of Taxing the Digital Economy in the Great Lockdown’, Bulletin for International Taxation 74 (2020), 532, 547. 20 See OECD, ‘Public consultation document’, para. 90.

Effective Minimum Taxation 975 This more far-reaching objective would imply a fundamental paradigm shift in the assessment of international tax competition by the G20 and OECD.21 Before and during the BEPS Project, the consensus was that only ‘harmful’ international tax competition should be eliminated, because and to the extent that it encourages profit shifting without also shifting ‘substance’, enables non-transparent arrangements, or ring-fences tax privileges to certain cross-border transactions or investments.22 By contrast, no need was seen to act on ‘fair’ tax competition for investment projects through generally applicable low tax rates or broadly defined preferential regimes. This approach became particularly evident in the agreement on the modified nexus approach for so-called intellectual property (IP) boxes under BEPS Action 5.23 It was therefore not surprising that this second potential objective of the minimum tax concept was much more controversial than the first, both in the public consultations held by the OECD and among numerous members of the Inclusive Framework. The most prominent objection was that a general limit on international tax competition would undermine national tax sovereignty. Reportedly in response to this criticism and in order to reach a compromise between supporters and opponents of the minimum tax initiative, the aforementioned substance-based ‘carve-out’ for routine returns on investments in tangible fixed assets and jobs was included in the concept. Moreover, the language in the official documentation of the negotiation process has also been toned down; the 2020 Blueprint mentions only the aim of using GloBE to address the remaining BEPS challenges, and the 2021 Statements do not mention any objective at all. Prima facie, GloBE has thus been scaled back to a broad anti-BEPS measure.24 On a closer look, however, certain elements that would curb international tax competition for investment would remain according to the 2020 Blueprint and to the October 2021 political agreement. The envisaged ‘carve-out’ would only shield presumed routine returns on real investments and wage costs from minimum taxation.25 In contrast, ‘excess’ profits generated from investment in, and exploitation of, intangible assets such

21

See M. Devereux et al., ‘The OECD Global Anti-Base Erosion Proposal’, Oxford University Centre for Business Taxation Report (Jan. 2020), Executive Summary, 1–2, available at https://oxfordtax.sbs. ox.ac.uk/files/oecdglobeproposalreportpdf; C. Cipollini, ‘International/OECD/G20—GloBE Proposal and Possible Carve-Outs: Is There a Future for Preferential Tax Regimes?’, World Tax Journal 12 (2020), 217, 222; L. Eden, ‘Taxing Multinationals: The GloBE Proposal for a Global Minimum Tax’, Tax Management International Journal 49/1 (2020), 7. 22 See OECD/ G20, ‘BEPS Action Item 5, Final Report 2015’, paras 155ff; OECD, ‘Explanatory Statement, Base Erosion and Profit Shifting Project, 2015 Final Reports’, para. 2; OECD, ‘Designing Effective Controlled Foreign Company Rules, Action 3, 2015 Final Report’, paras 6ff. 23 See OECD/G20, ‘BEPS Action Item 5, Final Report 2015’, 24ff. 24 See J. Hey, ‘Global Minimum Taxation (GloBE): What Is It About and What Could be a European Answer?’, in G. Kofler, R. Mason, and A. Rust, eds, Thinker, Teacher, Traveler—Reimagining International Tax: Essays in Honor of H. David Rosenbloom (2021), 254; Devereux et al., ‘The OECD Global Anti-Base Erosion Proposal’, 4. 25 See in more detail Section 53.3.3.

976 Joachim Englisch as IP would continue to be subject to the minimum tax. This applies even if they are backed up by some ‘substance’ in the respective low-tax location, within the meaning of the development, enhancement, maintenance, protection, and exploitation (‘DEMPE’) approach of the OECD Transfer Pricing Guidelines.26 In particular, low- tax regimes in the form of IP boxes would fall within the scope of GloBE regardless of whether they observe the requirements of the modified nexus approach of BEPS Action 5. Also, a carve-out in this respect was ultimately rejected by the Inclusive Framework, because it would undermine the ‘regulatory objectives and effectiveness’ of GloBE.27 But if GloBE were designed as a pure anti-BEPS instrument, this claim would make no sense, because compliance with the modified nexus approach should already ensure the absence of BEPS.28 In reality, the GloBE minimum tax still has the potential to establish a floor for international tax competition if implemented sufficiently broadly. But this floor would then apply only to taxation of excess profits, and indeed none is established regarding the traditional corporation tax on overall business profits. Moreover, GloBE can also be expected to mitigate international business tax competition above this floor; however, this effect is mostly limited to competition for highly mobile and highly profitable business activities of in-scope MNEs that involve relatively little ‘substance’.29 Since mobile IP and other intangible assets are the most important value drivers in today’s digitalized and globalized economy, more so than tangible assets and labour used for routine activities, the GloBE initiative thus still retains, to a certain degree, its character as a measure against ‘excessive’ business tax competition for investment.30 The ‘mute presence’ of this second objective of the minimum tax regime is also confirmed by its mention in an OECD brochure accompanying the October 2021 agreement.31 However, it should be acknowledged that it is pursued only half-heartedly in the actual GloBE Model Rules, because apparently not all Inclusive Framework member countries warmed to it.

26 OECD, ‘Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations’ (2017), ch. VI on intangible assets, esp. at B. 27 See also ‘Draft Report to Pillar Two’, CTPA/CFA/WP11/NOE(2020)21 (Summer 2020), para. 285 (on file with the author). 28 Cf. Cipollini, ‘International/ OECD/G20—GloBE Proposal and Possible Carve-Outs’, 235; B. da Silva, ‘Taxing Digital Economy: A Critical View Around the GloBE (Pillar Two)’, Frontiers of Law in China 15 (2020), 111, 124. 29 For an in-depth analysis, see M. Devereux, J. Vella, and H. Wardell-Burrus, ‘Pillar 2: Rule Order, Incentives and Tax Competition’, Oxford University Centre for Business Taxation Policy Brief (2022), available at https://oxfordtax.sbs.ox.ac.uk/pillar-2-rule-order-incentives-and-tax-competition; J. Englisch, ‘GloBE Rules and Tax Competition’, Intertax 50 (2022). 30 For a different opinion see Hey, ‘Global Minimum Taxation (GloBE)’. 31 See OECD, ‘Brochure: Two- Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy’ (Oct. 2021), 4.

Effective Minimum Taxation 977

53.2.3 Pro and Contra GloBE The author of this chapter has already extensively discussed the advantages and drawbacks of GloBE in earlier publications.32 This section is therefore limited to some essential aspects.

53.2.3.1 Effects on BEPS To begin with, the dual policy objective of the GloBE minimum tax has some merit, in particular from an economic point of view. First, regarding its anti-BEPS impetus, the shifting of book profits to a low-tax jurisdiction is unproductive from an economic perspective because it is not accompanied by a more efficient use of resources in value creation. Associated tax planning costs are therefore wasteful and welfare-reducing. Moreover, the corresponding revenue loss for the treasury on the losing side can lead to an undersupply of public goods and services, especially in developing and emerging countries. And while developed countries whose revenues are negatively affected by BEPS can often resort to other sources of taxation,33 that may have undesirable distributional effects in social policy terms. In addition, BEPS can give MNEs a competitive advantage in relation to smaller, especially local, suppliers that do not have the same international tax-planning options. Finally, the resulting concentration of taxing rights in low-tax countries that offer tax-sheltering opportunities is increasingly seen as contradicting the principles of international distributive justice. Admittedly, it was precisely the aim of the original BEPS initiative to address these issues. It has therefore frequently been a criticism that the G20/OECD should have awaited the full implementation of the 2015 BEPS recommendations, which might render additional anti-BEPS measures (e.g. minimum tax) superfluous.34 However, an abstract analysis of the 2015 BEPS package of measures has already strongly suggested that profit shifting, while becoming somewhat more costly, will not effectively be stopped, especially regarding schemes that rely on the allocation of intangible assets and risk within a corporate group.35 Initial empirical studies confirm this impression.36

32 See

J. Englisch and J. Becker, ‘International—International Effective Minimum Taxation—The GLOBE Proposal’, World Tax Journal 11 (2019), 484, 488ff; J. Becker and J. Englisch, ‘Implementing an International Effective Minimum Tax in the EU’ (2021), https://papers.ssrn.com/sol3/papers.cfm?abst ract_id=3892160. 33 See Devereux et al., ‘The OECD Global Anti-Base Erosion Proposal’, 8ff. 34 See, e.g., da Silva, ‘Taxing Digital Economy’, 119; further references can be found in Englisch and Becker, ‘International—International Effective Minimum Taxation—The GLOBE Proposal’, 491. 35 See, e.g., IMF Policy Paper No. 19/007, ‘Corporate Taxation in the Global Economy’ (10 Mar. 2019), 54ff; V. Chand and G. Lembo, ‘OECD/International—Intangible-Related Profit Allocation within MNEs based on Key DEMPE Functions: Selected Issues and Interaction with Pillar One and Pillar Two of the Digital Debate’, International Tax Studies 3 (2020), 35. 36 See K. Clausing, ‘5 Lessons on Profit Shifting From US Country- by-Country Data’, Tax Notes International (2020), 759ff.

978 Joachim Englisch Against this background, GloBE would be useful to enhance and complement the measures conceived in 2015.

53.2.3.2 Effects on tax competition and tax sovereignty The second objective of GloBE (i.e. reining in international tax competition for investment) also has its merits, even if the effect is essentially limited to entrepreneurial high-margin business activities and for IP holding locations.37 At least in respect to such activities, GloBE reduces potential distortions of investment decisions caused by tax incentives or low-tax regimes. Other factors that better reflect the value-creation capacity of the economy thereby gain relevance, which ensures a more efficient global allocation of resources and thus welfare gains.38 Moreover, countries that are being driven by—rather than driving—international tax competition would regain some leeway for raising business taxes on profits. This would allow especially developing countries to provide a higher level of public goods and services; developed countries, in particular, would have more freedom to implement their own policy choices on their desired distributional balance of their tax systems and thus on their society’s prevailing views on tax justice. Countries would no longer be pressured into foregoing taxes to an extent that is no longer appropriate even by the standards of public choice theory39 or the benefit principle. Moreover, large multinational companies naturally benefit more from international tax competition than their local small and medium-sized competitors. The resulting differences in the cost of capital or distortions of competition are neither desirable from an economic nor from a tax fairness point of view. As a mere minimum tax and due to political compromises, GloBE would only curb and not eliminate BEPS, and it would only moderately limit international tax competition, as explained earlier.40 But this limited impact is to some degree also a concession to the insight that international tax competition also has its benefits, and it addresses the objection of an inappropriate interference with national tax sovereignty. It is often assumed that international tax competition works as a necessary corrective to the tendency of politicians to spend excessively and pursue clientele politics.41 Moreover, it can be one of only a few instruments of small and peripheral economies to attract foreign direct investment42 and thereby generate positive spillovers for their

37 See also Elliffe, ‘OECD/ International— International Tax Frameworks: Assessing the 2020s Compromise from the Perspective of Taxing the Digital Economy in the Great Lockdown’, 548. 38 See Devereux et al., ‘The OECD Global Anti-Base Erosion Proposal’, 37. 39 Based on J. Buchanan, Public Finance in Democratic Process: Fiscal Institutions and Individual Choice (Chapel Hill, NC: University of North Carolina Press, 1967). See also J. Edwards and M. Keen, ‘Tax Competition and Leviathan’, European Economic Review 40 (1996), 113ff, with further references. 40 For a critical comment see B. Arnold, ‘The Evolution of Controlled Foreign Corporation Rules and Beyond’, Bulletin for International Taxation 73 (2019), 631, 645. 41 Cf. the overview provided by Edwards and Keen, ‘Tax Competition and Leviathan’, 115. 42 See e.g., K. Andersson, ‘Should We Use Value Creation or Destination as a Basis for Taxing Digital Businesses? Krister Andersson’s Comments on the 2018 Klaus Vogel Lecture Given by Professor Michael Devereux’, Bulletin for International Taxation 72 (2018), 684, 688.

Effective Minimum Taxation 979 economy. From this perspective, any attempt to neutralize international tax competition is met with scepticism. A related concern is that a minimum tax runs counter to national sovereignty in tax matters. If a sufficiently large number of countries were to levy a minimum tax, a country’s sovereign right to pursue an investment-friendly tax policy would be rendered partially ineffective.43 The GloBE minimum tax, however, has been neither conceived nor designed to eliminate international tax competition. It would merely establish a—low—floor to it, and it would moderately curb competition for certain high-margin projects above that floor. In a similar vein, while GloBE indirectly constrains the fiscal and economic sovereignty of low-tax states to attract direct investment through their fiscal policies, it also partially restores the tax sovereignty of countries with a preference for higher business taxes that are not sustainable in a situation of unfettered tax competition. While an imperfect compromise, GloBE can thus conceptually still be defended as seeking an adequate balance between the positive and negative aspects of business tax competition, and the respective national interests of countries that are in different economic positions. That notwithstanding, it should be acknowledged that the internal procedures of the Inclusive Framework combined with the ambitious time frame of negotiations often made it difficult for developing countries to fully identify and assert their national interests in the process. The consensus reached in 2021 was therefore less inclusive than it could have been, which calls for future institutional reform.44 The main alternative channels are: first, taxes that are (at least partially) borne by businesses but not covered by the GloBE minimum tax, such as payroll taxes or, possibly, the personal income taxes of highly skilled personnel; secondly, direct subsidies; and, thirdly, regulatory competition in fields such as environmental or labour law standards. It is highly unlikely that any single alternative could wholly undo the effects of GloBE. However, where such a shift to other means of competition leads to imbalanced or undesirable outcomes, they should be addressed through separate measures that are adapted to the respective context.

53.2.3.4 Other aspects It has been criticized in various submissions during public consultations on GloBE that the latter would be liable to distort competition in local markets by undermining capital-import neutrality. Multinational firms that sell products or services in a jurisdiction where the ETR is below the minimum tax rate would incur higher tax burdens than local businesses if they had to pay top-up tax on the profits generated in that jurisdiction. The potential magnitude of this problem under a ‘pure’ minimum tax concept would crucially depend on the chosen minimum rate. But this aspect has now become much less of a concern in any event due to the SBIE carve-out for a 43 See

Arnold, ‘The Evolution of Controlled Foreign Corporation Rules and Beyond’, 647; Devereux et al., ‘The OECD Global Anti-Base Erosion Proposal’, 5, 16 and 19; da Silva, ‘Taxing Digital Economy’, 121ff. 44 See also A. Christians and L. van Apeldoorn, ‘The OECD Inclusive Framework’, Bulletin for International Taxation 72 (2018), 226.

980 Joachim Englisch deemed routine return on ‘substance-based’ activities.45 Most local sales and distribution activities of multinationals in a given market jurisdiction tend to involve routine rather than entrepreneurial functions and little or no intangible assets, so that the ensuing profits would not be subject to the minimum tax regime.46 Any eventual remaining distortions would arguably be outweighed by the efficiency gains from levelling the playing field in jurisdictions where multinationals currently enjoy competitive advantages over local firms that do not have access to tax havens and low-tax regimes. It has furthermore been claimed that levying a top-up tax primarily in the (ultimate) parent jurisdiction is not in line with the principle of taxation at the place of value creation, as promoted in the 2015 BEPS Action Item Report.47 But this is not a particularly relevant objection. First, it is not necessarily correct because profits generated in low-tax IP holding jurisdictions often have their root cause in R&D and innovation that occurred in the head office jurisdiction before the nascent IP was shifted.48 Secondly, and more importantly, the GloBE minimum tax operates as a subsidiary surtax. It therefore need not adhere to international standards of fair allocation of taxing rights; rather, its place of collection should be based on considerations of practicability and efficiency. And, finally, the decision to credit a QDMTT liability against the international top-up tax on undertaxed excess profits provides source countries with the option of a first tax grab even for minimum tax revenue. An effective international minimum tax would reduce the possibilities for international tax arbitrage and thus increase the effective marginal tax burden on investment and the cost of capital. This could have a dampening effect on global investment volumes.49 According to the OECD, however, the expected adverse effects would only be very moderate;50 this is corroborated by a more recent study carried out for UNCTAD.51 The concrete design of the GloBE minimum tax does, indeed, minimize the risk of negative effects on the scale of investment to a great degree. In particular, the increase of the marginal tax burden on business profits that an effective international minimum tax entails is primarily relevant for the scale of investment projects that do not generate ‘excess profits’ or economic rents, but only routine profits. However, such projects are typically not affected by the minimum

45

For a more detailed discussion, see Section 53.3.3. For similar conclusions but different reasoning, see J. Hey, ‘The 2020 Pillar Two Blueprint: What Can the GloBE Income Inclusion Rule Do That CFC Legislation Can’t Do?’, Intertax 49 (2021), 7, 9. 47 Cf. Devereux et al., ‘The OECD Global Anti- Base Erosion Proposal’, 3 and 5; Cipollini, ‘International/OECD/G20—GloBE Proposal and Possible Carve-Outs’, 247. 48 See Englisch and Becker, ‘International—International Effective Minimum Taxation—The GLOBE Proposal’, 494. 49 See Devereux et al., ‘The OECD Global Anti-Base Erosion Proposal’, 4–5 and 10. 50 See OECD, ‘Tax Challenges Arising from Digitalisation—Economic Impact Assessment: Inclusive Framework on BEPS’, OECD/G20 Base Erosion and Profit Shifting Project (2020), 153ff. 51 See UNCTAD, ‘World Investment Report 2022’ (2022), ch. 3. 46

Effective Minimum Taxation 981 tax at all due to the SBIE exemption for substance-based activities. In addition, the GloBE Model Rules generally treat deferred taxes as equivalent to covered taxes that have actually been paid for the purposes of ETR calculation. This ensures that accelerated expensing or deferred recognition of income for tax purposes, and even full expensing of capital assets or cash-flow taxation and thus marginal tax burdens of zero, remain fully effective as tax incentives for real investments. In any event, any remaining negative effects of GloBE on investment are inherent in any effective taxation of capital income and they are therefore a necessary consequence of corresponding distributional preferences. The risk of a somewhat reduced scale of investment is therefore a legitimate point of criticism only from the point of view of those countries which, even without international tax competition, would opt for an effective tax on corporate profits of less than 15%. If the minimum tax system were adopted by a large number of relevant economies, it moreover cannot simply be assumed that the top-up tax burden would be shifted by affected MNEs to less mobile taxpayers (i.e. consumers and employees) rather than borne by the owners of the firm.52 The existing empirical findings on the incidence of corporation tax in open economies do not reflect the situation of an internationally coordinated and comprehensive introduction of an additional tax burden. Any international minimum tax, and certainly also the GloBE, would add a layer of complexity to international taxation, even if only for a limited number of large multinational groups. This causes additional compliance costs and risks for businesses and administrative costs for tax administrations. The decision to determine the ETR for each jurisdiction (‘jurisdictional blending’),53 in particular increases the administrative complexity and compliance burden of the GloBE rules. This has been accepted because the alternative of ‘global blending’ would considerably dilute the effectiveness of the minimum tax.54 Another significant driver of complexity are the numerous adjustments to financial accounting income required for the purpose of calculating the GloBE ETR, including adjustments to the use of deferred taxes in the numerator of the formula.55 There is therefore a need for pragmatic and effective simplification measures, allowing initial ETR estimates based on data that is typically already available, in order to substantially reduce the number of country-specific full ETR calculations that an in-scope MNE has to carry out.

52

For a different view, see Devereux et al., ‘The OECD Global Anti-Base Erosion Proposal’, 8ff. For a more detailed discussion, see Section 53.3.1. 54 See also S. Shay, J. Fleming, and R. Peroni, ‘Designing a 21st Century Corporate Tax—An Advance US Minimum Tax on Foreign Income and Other Measures to Protect the Base’, Florida Tax Review 17 (2015), 669, 706; Chand and Lembo, ‘OECD/International—Intangible-Related Profit Allocation within MNEs’, 36; A. Dourado, ‘The Global Anti-Base Erosion Proposal (GloBE) in Pillar II’, Intertax 48 (2020), 152, 156. 55 For details, see ‘GloBE Model Rules’, art. 4.4. 53

982 Joachim Englisch

53.3 Critical Design Elements The GloBE Model Rules and Model Commentary together feature almost 300 pages of technical details, and additional elements are still being deliberated by the relevant OECD Working Parties and the Inclusive Framework. This section will analyse only the most critical design features involving important trade-offs and political judgments.

53.3.1 Jurisdictional Blending One of the most decisive policy choices of the GloBE design is the endorsement of ‘jurisdictional blending’, whereby the ETR—and potentially top-up tax—is calculated for each jurisdiction. The main alternative would consist in calculating it only once for all foreign profits of a multinational group (‘global blending’), as under the US GILTI rules. Jurisdictional blending requires essentially a four-stage approach.56 First, the GloBE base for each constituent entity—including PEs—must be calculated. As discussed later, the starting point for this exercise would be local accounting information used for the preparation of the group consolidated financial accounts. Secondly, the relevant entity income must be assigned to a particular jurisdiction. In line with the country-by-country reporting (CbCR) rules, this would essentially be carried out on the basis of tax residence, or location in the case of PEs, with special rules for tax-transparent entities, reverse-hybrid entities, etc. Thirdly, it must be determined which covered taxes can be allocated to the respective entity income. This step is guided by the principle of ‘taxes follow profits’. It does not matter which jurisdiction levies the tax, but rather which taxes are levied on the profits assigned to a particular jurisdiction, including, for example, foreign withholding taxes or controlled foreign corporation (CFC) taxes.57 Moreover, deferred taxes will, in principle, be treated as equivalent to actually paid taxes for this purpose. Finally, the jurisdictional ETR is computed as the relation between the sum of GloBE income that has been assigned to a specific jurisdiction and the entirety of corresponding covered taxes. ‘Jurisdictional blending’ is preferable to ‘global blending’ even though it increases complexity and enforcement costs, because it is significantly more efficient with respect to the dual objective of GloBE. Under the ‘global blending’ approach, profit shifting to low-tax jurisdictions or other profit sheltering arrangements would continue to be attractive as long as the ETR on all foreign group profits did not fall below the minimum tax level.58 The incentive for certain jurisdictions to offer a generally low ETR or 56

See ‘Pillar Two Blueprint’, paras 248ff. For a detailed analysis of the interaction with CFC taxes, see Arnold, ‘The Ordering of Residence and Source Country Taxes and the OECD Pillar Two Global Minimum Tax’, 270. 58 See H. Grubert and R. Altshuler, ‘Fixing the System: An Analysis of Alternative Proposals for the Reform of International Tax’, National Tax Journal 66 (2013), 671 697ff; C. Garbarino, ‘A Critical Evaluation of the International Impact of the Global Minimum Tax’, International Tax Studies 6 (2022), 8. 57

Effective Minimum Taxation 983 preferential low-tax regime in order to attract book profits or IP would remain largely unchanged, because MNEs could be expected to still exploit those opportunities for international tax arbitrage. Since they cannot fully avoid a taxable presence in medium- and high-tax jurisdictions, they would usually have some profits taxed at a higher-than- minimum effective rate to potentially set off against their low-taxed income.

53.3.2 ETR Calculation Based on Consolidated Financial Accounts The GloBE base and thus the ETR calculation do not rely on the tax accounting rules of the state that would be entitled to collect eventual top-up tax (i.e. by default, those of the ultimate parent jurisdiction). In this respect, the GloBE concept departs from the usual approach in the context of unilateral CFC rules. Instead, the GloBE base is tied to the accounting rules used to compute the financial consolidated accounts of the MNE group. This means that the GloBE base is, in principle, determined by the accounting standard that is applicable to the ultimate parent entity.59 As a caveat, the respective standard would have to be either the international financial reporting standards (IFRS) or a— generously defined— equivalent and therefore ‘acceptable’ standard; other standards will be reviewed and material divergences to IFRS would have to be adjusted for.60 There are various reasons for this approach. First, it would allow for a relatively uniform calculation of the ETR in all participating countries, which in turn is important for the international coordination of the collection of top-up tax.61 Countries will only be willing to refrain from applying their own minimum tax regime when they can be sure that a sufficient amount of minimum tax is already levied in a prioritized jurisdiction. Secondly, this approach can build on information that will often already be available for non-tax purposes, reducing compliance costs. Moreover, and in a similar vein, it avoids an MNE group having to make multiple calculations for the same profits where the top- up tax is collected by more than one country, as could occur where the UTPR is applied. Finally, there would be synergies with the similar approach envisaged for the purposes of allocating taxation rights to market states (‘Amount A’) in the context of Pillar One.62 In order to keep up with the underlying ambition of simplification, the determination of entity-level income to be assigned to a particular jurisdiction would, in principle, start with the information that has been gathered at the level of the individual constituent entity for the purpose of preparing the consolidated group balance sheet.63 There would 59

See arts 3.1.2 and 10.1 (definition of ‘Consolidated Financial Statements’) of the GloBE Model Rules and the Model Commentary on art. 3, paras 5–6. 60 See the relevant definitions in art. 10.1 of the GloBE Model Rules and the Model Commentary on art. 10, para. 4. 61 See Dourado, ‘The Global Anti-Base Erosion Proposal (GloBE) in Pillar II’, 155. 62 See OECD, ‘Pillar Two Blueprint’, paras 417ff. 63 See the Model Commentary on art. 3, para. 3.

984 Joachim Englisch thus be no need to break down the consolidated accounts into country-specific data. For similar reasons, there would only be a limited number of book-to-tax adjustments. In general, adjustments are limited to deviations from IFRS that are internationally common, material, and moreover give rise to permanent rather than merely temporary differences (e.g. exemption of dividends and gains from significant shareholdings; non- deductibility of bribes and penalties).64 As a caveat, the Model Rules also require certain adjustments regarding deferred tax positions that will be taken into account in the numerator of the ETR formula in order to address merely temporary differences65; in practice, this will cause significant compliance costs. The principle of precedence of financial accounting can lead to some apparent inconsistencies regarding the effects of tax-expenditure schemes on the ETR calculation. In particular, investment tax credits that are a common R&D incentive in national income tax laws, or other forms of incentivizing tax credits, can be treated as either (tax- free) government grants under IAS 20 or as a reduction in income taxes under IAS 12. Their exact financial accounting treatment depends essentially on whether, and if so to what extent, they are refundable in the case of an excess credit. The GloBE rules have been aligned with that approach, even though they operate with a general four-year limit for increased legal certainty.66 Small variations in the design of a tax credit can therefore have significant implications for the ETR calculation, because a decrease in taxes reduces the numerator in the ETR fraction and therefore has a disproportionally larger impact on the ETR than a mere broadening of the GloBE base in the denominator by the same amount. However, this inconsistency is inherent in the minimum tax concept because it only targets investment incentives delivered via the tax system and not direct or quasi-direct subsidies.67 Moreover, in reality this may well be of lesser concern than in theory, because the two forms of subsidies are not easily interchangeable given the greater budgetary effects and transparency of direct subsidies. There are some more serious concerns, however, that certain tax- planning opportunities are inherent in the use of financial accounting as a starting point for the GloBE base. This could allow MNEs to deflate the denominator of the ETR fraction and thus ‘artificially’ increase the ETR to avoid, or mitigate, top-up taxation. One possible strategy that has been pointed out68 is the use of hybrid financial instruments for intra- group financing, where payments are treated as interest under IFRS but as dividends for local taxation purposes. Other possible schemes could rely on accounting discretion to 64

See OECD, ‘Pillar Two Blueprint’, paras 175ff; Model Commentary on art. 3, para. 21. See art. 4.4 of the GloBE Model Rules. 66 See arts 3.2.4 and 4.1.2(d) of the GloBE Model Rules in combination with the definition of ‘Qualified Refundable Tax Credit’ in art. 10.1. 67 See Arnold, ‘The Evolution of Controlled Foreign Corporation Rules and Beyond’, 644 fn. 100; R. Pinkernell and X. Ditz, ‘Säule 2 des Arbeitsprogramms des Inclusive Framework on BEPS der OECD— kritische Anmerkungen zum GloBE-Proposal‘, Internationale Steuer-Rundschau (2020), 1, 9. 68 See M. de Wilde, ‘Is There a Leak in the OECD’s Global Minimum Tax Proposals (GLOBE, Pillar Two)?’, Kluwer International Tax Blog (1 Mar. 2021), available at http://kluwertaxblog.com/2021/03/01/is- there-a-leak-in-the-oecds-global-minimum-tax-proposals-globe-pillar-two/ (accessed 22 March 2021). 65

Effective Minimum Taxation 985 reduce the GloBE base or to stay below the personal scope threshold of annual group turnover. While this would also reduce reported earnings and thus have undesirable repercussions for investors and creditors of the MNE, the benefit of tax savings might well outweigh those opposing forces. Moreover, such strategies would not only permit (minimum) tax avoidance, but they would eventually also undermine the effectiveness of financial accounting information with respect to its primary functions.69 It is therefore likely that, over time, the number of mandatory book-to-tax adjustments will have to increase to reduce these detrimental effects.

53.3.3 Exception for Substance-Based Profits As mentioned previously, the levy of GloBE top-up tax is limited to ‘excess profits’, whereas no minimum tax liability arises with respect to the routine return on ‘substantive activities’ even if the local ETR is below 15%. To this effect, the deemed routine profit is determined under a formulaic carve-out, the so-called SBIE70 which takes into account two factors of value creation as indicators of local substantive activities that are less susceptible to BEPS risks and therefore not covered by the GloBE minimum tax: eligible employees and certain tangible fixed assets. A fixed percentage of 5% applies to payroll and to the carrying value of qualifying tangible assets, and up to the resulting amount the profits assigned to the respective jurisdiction will not be subject to any top- up tax. During a transition period of ten years, higher percentages of 10% of payroll and 8% for tangible assets will initially apply as part of the October 2021 compromise agreement, to be gradually reduced each year.71 Arguably, those percentages mark the upper limit of a reasonable estimate of mere routine profits that can be generated without the use of significant intangible assets; they should not be increased any further. This carve-out approach is clearly inspired by the so-called qualified business asset investment (QBAI) exemption under the US GILTI rules,72 with an added payroll component. This is somewhat ironic considering that in the USA, the Biden administration originally sought to eliminate this very exemption. However, it was apparently not possible to secure broad international endorsement of GloBE without this carve-out. Preferably, a carve-out should instead have shielded IP regimes that are compliant with BEPS Action 5 from the application of the minimum tax. Without such grandfathering, GloBE disproportionately interferes with the principle of protection of legitimate expectations.

69

See, e.g., the Open Letter of Concern from 264 US Accounting Academics (5 Nov. 2021), available at https://tax.unc.edu/index.php/news-media/open-letter-of-concern-from-264-accounting-academics- regarding-including-financial-accounting-income-in-the-tax-base/ (accessed 9 November 2021). 70 See art. 5.2.2 of the GloBE Model Rules. 71 For details, see ‘Statement of 136 member jurisdictions’, 4. 72 For a detailed analysis of QBAI, see C. Pérez Gautrin, ‘US Tax Cuts and Jobs Act: Part 1—Global Intangible Low-Taxed Income (GILTI)’, Bulletin for International Taxation 73 (2019), 36, 45.

986 Joachim Englisch In the 2020 Blueprint, the interaction of the carve-out with the formula for calculating a jurisdiction’s ETR was still marked as an open issue. In that regard, the Blueprint contemplated different design options,73 which quintessentially represented two alternative approaches. Under the first approach, the carve-out would also have been taken into account for the purpose of calculating the local ETR, which would then have represented the effective tax rate only for the excess profits, under the assumption that all covered should be attributed exclusively to those excess profits. Under the alternative approach, which was eventually endorsed in the GloBE Model Rules, the SBIE is ignored for the purposes of calculating the jurisdictional ETR, so that the latter represents the effective tax rate for the totality of MNE profits attributable to a particular jurisdiction. The SBIE only becomes relevant to determine whether an ETR below 15% results in a top-up tax liability due to the existence of excess profits, and if so how much top-up tax arises.74 It has been shown that GloBE would have had the potential to establish a floor for international tax competition with respect to traditional corporation tax on overall business profits only under the first, hypothetical approach.75 On the other hand, determining the GloBE ETR without regard to the SBIE potentially mitigates tax competition until the threshold of 15% effective taxation on overall profits is reached, which the other approach would not have achieved.76 It has been argued that the ‘substance-based’ carve-out assimilates the rationale for GloBE to the one underlying traditional CFC regime, in line with the BEPS Action 3 Report.77 It has therefore been suggested that GloBE should merge with existing CFC regimes, eliminating the high group-revenue threshold and levying top-up tax at the regular rate applicable to the (ultimate) parent entity, while possibly maintaining the calculation of the ETR based on IFRS and jurisdictional blending.78 However, while the finding of a certain convergence of the objectives pursued by the GloBE and CFC regimes is convincing,79 the aforementioned conclusions are not fully so. Having a common base and a common (minimum) top-up tax in all jurisdictions provides GloBE with some distinct advantages, facilitating its international coordination and reducing the risk of international head office competition and tax-driven inversions. Both of those elements, in turn, imply phasing in GloBE with relatively high thresholds, to ensure proportionate compliance and administrative costs.80 Against this backdrop, it should be expected that traditional CFC rules will continue to operate alongside GloBE,

73

See ‘Pillar Two Blueprint’, para. 335. See arts 5.1 and 5.2.2 of the GloBE Model Rules. 75 See Devereux et al., ‘What Is the Substance-Based Carve-Out under Pillar 2? And How Will It Affect Tax Competition?’, EconPol Policy Brief 5 (2021), 39. 76 See Englisch, ‘GloBE Rules and Tax Competition’. 77 OECD/G20, ‘BEPS Action Item 3, Final Report 2015’. 78 See Hey, ‘The 2020 Pillar Two Blueprint’, 12–13. 79 See also F. De Lillo, ‘Introducing Pillar Two: Towards a Global Minimum Effective Tax Rate’, in A. Perdelwitz and A. Turina, eds, Global Minimum Taxation? (Amsterdam: IBFD, 2021), 3, 14. 80 See OECD, ‘Pillar Two Blueprint’, paras 114ff. 74

Effective Minimum Taxation 987 at least for a considerable transition period. This is also the solution of the US GILTI regime.

53.3.4 Mechanisms to Collect Top-Up Tax As mentioned earlier, top-up tax is to be levied primarily by way of an IIR in the ultimate parent jurisdiction, resembling the mechanics of a CFC regime. If the former jurisdiction did not implement GloBE, a top-down approach in the participation chain would be pursued, with certain exceptions due to avoidance concerns.81 In the case of PE profits that are exempt under tax treaty law, the levying of the top-up tax would have to be supported by a switch-over clause. Only to the extent that top-up taxes cannot be attributed to a group entity upstream of the low-taxed constituent entity, would the UTPR apply in the source state(s). The UTPR therefore seeks to counter the relocation of corporate headquarters to countries without an IIR, and to ensure that profits in the (ultimate) parent jurisdiction are subject to GloBE, too. The 2020 Blueprint envisaged allocating taxing rights under the UTRP in a two-step process which was supposed to reflect intra-group (net) payment flows. For simplification reasons, that concept has, however, been abandoned in the GloBE Model Rules. Instead, the allocation of UTPR percentages to jurisdictions now follows the same ‘economic substance’ approach that also underlies the SBIE: half of it is determined by the ratio of local employees to overall employees in all UTPR jurisdictions, and the other half by the corresponding ratio with respect to the net book value of tangible assets.82 For EU and EEA member states, the IIR and the UTPR could be problematic regarding their consistency with the freedom of establishment enshrined in the European Treaties.83 But as stated previously, the EU has enacted Union legislation to harmonize the implementation of GloBE. The relevant Minimum Tax Directive extends the GloBE collection mechanisms to certain domestic constituent entities of in-scope MNEs in all member states in order to rule out any problematic discrimination. Under an alternative design modification that had been proposed by the author of this chapter in order to comply with the EU free movement rights, the foreign ‘undertaxed’ entities could themselves have become liable to pay top-up tax. The IIR and UTPR rules would

81

See ibid., paras 434ff, esp. para. 438; and the Model Commentary on art. 2, paras 3–10. See art. 2.6 of the GloBE Model Rules. 83 For a detailed analysis, see J. Nogueira, ‘GloBE and EU Law: Assessing the Compatibility of the OECD’s Pillar II Initiative on a Minimum Effective Tax Rate with EU Law and Implementing It within the Internal Market’, World Tax Journal 12 (2020), 465; P. Koerver Schmidt, ‘A General Income Inclusion Rule as a Tool for Improving the International Tax Regime—Challenges Arising from EU Primary Law’, Intertax 48 (2020), 983; L. de Broe and M. Massant, ‘Are the OECD/G20 Pillar Two GloBE- Rules Compliant with the Fundamental Freedoms?’, EC Tax Review 30 (2021), 83; J. F. P. Nogueira and A. Turina, ‘Pillar Two and EU Law’, in Perdelwitz and Turina, Global Minimum Taxation?, 283; Englisch, ‘Designing a Harmonized EU-GloBE in Compliance with Fundamental Freedoms’, EC Tax Review 30 (2021), 136. 82

988 Joachim Englisch then have been reduced to mere rules on the allocation of taxing rights (and could even also have been modified in that regard).84 Compliance with such a collection mechanism and its administration could have been further facilitated by standard procedural instruments of EU tax law, namely the so-called one-stop shop and joint audits.

53.4 Compatibility with Tax Treaty Law The 2020 Blueprint argues that only a few tax treaty changes are necessary to implement GloBE. This concerns the rather marginal instrument of a switch-over clause for PE profits, and the merely complementary STTR.85 In order to ensure an efficient and prompt amendment of the treaty network, the October 2021 Implementation Plan suggests creating a new multilateral instrument that would amend the existing network of bilateral treaties. In contrast, the IIR and the UTPR have been held to be compatible with standard tax treaty clauses inspired by the OECD Model Convention, especially its articles 7, 9, and 24.86 This is arguably correct,87 especially since the now proposed ‘substance-based’ carve-out moves GloBE closer in the direction of an anti-avoidance measure. However, it is not entirely clear, at least for tax treaties without a saving clause like article 1(3) of the 2017 OECD Model.88 It would therefore be advisable to include clarifications to that effect in a possible multilateral convention.

53.5 Implementation and Simplification Measures All forms of collecting the GloBE minimum tax require amendments to the national tax law of each state that seeks to implement it. The technical design of the respective instruments is left to each state, as long as the outcomes intended by the GloBE Model

84

For a detailed discussion of this and the following, see Englisch, ibid., 207. See OECD, ‘Pillar Two Blueprint’, para. 677. 86 See ibid., paras 679ff. 87 See Englisch and Becker, ‘International—International Effective Minimum Taxation—The GLOBE Proposal’, 517ff, with further references also to the opposing view; see also da Silva, ‘Taxing Digital Economy’, 128ff, with further references. 88 See Arnold, ‘The Evolution of Controlled Foreign Corporation Rules and Beyond’, 645; da Silva, ‘Taxing Digital Economy’, 132; B. Andrade Rodrígez and L. Nouel, ‘Interaction of Pillar Two with Tax Treaties’, in Perdelwitz and Turina, Global Minimum Taxation?, 242; V. Chand, A. Turina, and K. Romanovska, ‘Tax Treaty Obstacles in Implementing the Pillar Two Global Minimum Tax Rules and a Possible Solution for Eliminating the Various Challenges’, World Tax Journal 14 (2022), 3. 85

Effective Minimum Taxation 989 Rules are respected and the internationally agreed top-up tax is ultimately levied in the right amount at the right place. For example, the UTPR could be implemented by means of an independent top-up tax or by limiting the tax deductibility of payments, as acknowledged by the GloBE Model Rules themselves.89 There is probably not much added value in making the political commitment of the Inclusive Framework member countries to the GloBE Model Rules legally binding by means of a multilateral agreement90 which could, in any event, be unilaterally terminated at a later date. That notwithstanding, an internationally coordinated implementation will require an intensification of international administrative assistance, for which the current instruments of exchange of information and dispute settlement are not well equipped. It should therefore be considered to at least expand country-by-country reporting (CbCR) to include more GloBE-relevant information, and to strengthen the international compliance assurance programme91 and joint audits. To minimize compliance costs for MNEs within the personal scope of GloBE to the extent possible without compromising the effectiveness of GloBE, simplification measures that waive the requirement to calculate the ETR for certain ‘low-risk’ jurisdictions unless requested for auditing purposes should be part of the final package. Various options are still being explored by the Inclusive Framework, among them ‘safe harbour’ rules based on CbCR data and the development of ‘tax administrative guidance’. The latter would designate countries which are generally low risk or where MNEs need only check a very limited list of ‘red flags’ to decide whether a full GloBE declaration is required.92 Together with a colleague, the author of this chapter has developed and extensively vetted such an approach on behalf of the OECD Secretariat. We consider it feasible93 but it remains to be seen whether it will also be politically endorsed. Ensuring an efficient and impartial procedure for the necessary country-by-country assessments could prove to be a stumbling block.

53.6 Final Evaluation and Outlook The introduction of an effective GloBE minimum tax would pose considerable technical challenges for both the multinational corporations concerned, which number in the 89

See art. 2.4.1. of the GloBE Model Rules and the Model Commentary on art. 2, paras 43ff. was contemplated in OECD, ‘Pillar Two Blueprint’, paras 705ff, and in the Implementation Plan of the 8 October 2021 Statement. For a different opinion, see P. Pistone and A. Turina, ‘The Way Ahead: Policy Consistency and Sustainability of the GloBE Proposal’, in Perdelwitz and Turina, Global Minimum Taxation?, 432. 91 See OECD, ‘International Compliance Assurance Programme, Pilot Handbook 2.0’ (2019). 92 Also considered the most promising approach by Pistone and Turina in Perdelwitz and Turina, Global Minimum Taxation?, 423–425. 93 See C. Döllefeld et al., ‘A Simplification Safe Harbor for Pillar 2’, Tax Notes International 106 (2022), 1513. 90 This

990 Joachim Englisch thousands, and the national tax administrations. The drafters of GloBE have made significant efforts to make the new regime administratively feasible and reduce compliance costs. Nevertheless, it would add a considerable degree of complexity to international tax law in order to be sufficiently effective and internationally coordinated. If one does not reject the idea of a global minimum tax for fundamental reasons, as some do, the question arises as to whether the expected outcome outweighs the associated costs. According to the OECD’s calculations, additional tax revenue in the double-digit b illions could be expected globally.94 While this economic impact assessment is based on data before the outbreak of the Covid-19 pandemic, it has been confirmed by a simulation of members of the EU Tax Observatory. A recent UNCTAD report is also broadly in line with that estimate, as it predicts a significant growth in global revenues on FDI profits in a range between 15% and 20%.95 Furthermore, the minimum tax has the potential for certain (net) positive effects for global welfare despite an eventual dampening effect on corporate investment. It could provide countries with more liberty to design their tax mix and rates in line with their economic and distributional preferences. As a caveat, those potential benefits presuppose a critical mass of states willing to implement it, especially among large industrialized countries and classic holding locations with an industrial base.96 Assuming a more long-term perspective, the usefulness of an international minimum tax regime is intrinsically linked to the development of nexus criteria for the allocation of taxing rights for (excess) business profits: the less mobile the relevant connecting factors become, the less need to curb tax competition and BEPS via a minimum tax.

94

See OECD, ‘Economic Impact Assessment’, 15. See UNCTAD, ‘World Investment Report 2022’ (2020), 125–126. 96 See, in that regard, Devereux et al., ‘The OECD Global Anti-Base Erosion Proposal’, 11ff. 95

Chapter 54

Challenges of t h e E m erging Inte rnat i ona l Tax C onsensus for L ow - and M iddle -I nc ome C ountri e s Natalia Quiñones

54.1 Introduction Globalization has affected all areas of international regulation, given the global, borderless nature of the action by multinational enterprises (MNEs). Tax is not an exception, as MNEs have continued to expand their business to new markets, sometimes reaching more than 180 territories in the case of the biggest brands. However, the impact of globalization on taxation has not been the same across countries. A study published in 2019 indicates that globalization has had the effect of decreasing taxation overall, but in countries with a high capital–labour ratio, this trend is associated with increased taxation.1 In other words, globalization has indeed created opportunities for mobilizing profits and valuable assets out of the scope of labour-intensive countries, and most of that shift has benefited capital-intensive countries. Low-and middle-income countries (as defined by the World Bank2) are usually labour-intensive, while capital- intensive countries are usually associated with more industrialized economies or investment hubs. 1 P. Jha and G. Gozgor, ‘Globalization and Taxation: Theory and Evidence’, European Journal of Political Economy 59 (2019), 296–315. 2 World Bank Country Lending Groups, available at https://datahelpdesk.worldbank.org/knowle dgebase/articles/906519-world-bank-country-and-lending-groups (accessed 10 March 2021).

992 Natalia Quiñones Globalization and technological advancement have also deeply affected the way that companies do business. Nowadays, economic profits are obtained from an increasingly integrated, digitalized, and mobile operation across continents and, more recently, outside planetary borders.3 Meanwhile, domestic tax rules across the world continue to rely on a relatively stable physical presence of the taxpayer, especially in low-and middle- income countries given their historic reliance on natural resources. In all cases, however, the asymmetry between a borderless world for companies and a tax system based on those territorial frontiers has given rise to a scenario where countries try to attract investment (including through tax competition) while trying to prevent aggressive tax planning and the erosion of their tax base. Companies take advantage of the first type of initiatives, but are also overburdened with the costs of complexity and multiple taxation caused by the second type of rules. In this context, it seems impossible to create unilateral systems (except perhaps in the case of the USA, as argued by Professor Avi-Yonah4), that can effectively tax the profit without creating excessive burdens or multiple taxation. It is important to mention that at this level of business integration across regions, bilateral treaties are not enough to resolve the conflicts that arise when MNEs apply several different domestic tax legislations to a seamless multilateral operation. Furthermore, tax competition has made it easier for global taxpayers to achieve an overall reduction in the effective tax rate levied on their profits, which in turn has given rise to serious inequalities that are now recognized by citizens5 and academics around the world.6 Inequality in the distribution of taxing rights is also on the rise due to the rapid changes brought by technology. On the one hand, technology has achieved recent significant progress in reducing the time required to complete important tasks within the supply chain, and even in the time required to complete an entire business activity such as the construction of a building. A striking example of the latter can be found in the 3D-printing business, which now allows for the construction of a building, or even an entire housing complex, in a matter of days,7 when the permanence threshold for a construction permanent establishment on the OECD Model Tax Convention on Income and Capital (OECD MC) is twelve months, and in the United Nations Model Double Taxation Convention between Developed and Developing Countries (UN MC) is six 3 E.

Mazareanu (25 Aug. 2020), estimated the turnover of the global space economy was US$423.8 billion in 2019. More statistics on the space economy are available at https://www.statista.com/statistics/ 946341/space-economy-global-turnover/. 4 R. S. Avi- Yonah, ‘Constructive Unilateralism: US Leadership and International Taxation’, International Tax Journal 42 (2016), 17. 5 See World Inequality Data, a collaborative platform by academics and NGOs, available at https:// wid.world/. 6 See, e.g., A. Alstadsæter, N. Johannesen, and G. Zucman, ‘Tax Evasion and Inequality', American Economic Review 109/6 (2019), 2073–2103; see also M. Agranov and T. R. Palfrey, ‘The Effects of Income Mobility and Tax Persistence on Income Redistribution and Inequality’, European Economic Review 123 (2020), 103372. 7 See S. Lubell, ‘3D-Printing is Speeding Up the Automation of Construction’ (15 Feb. 2021), https:// www.metropolismag.com/technology/3d-printing-is-speeding-up-the-automation-of-construction/.

Challenges of the Emerging International Tax Consensus 993 months. On the other hand, technology has made it easier to conduct highly profitable business activities remotely, a trend that has been made evident to all in the Covid-19 pandemic and the global switch to digital and remote working. The failure of traditional international tax principles in addressing global mobility and technological changes has also been recognized by the OECD itself since the first base erosion and profit shifting (BEPS) reports were released in 2015.8 This failure has been more critical in the case of the developing and emerging economies, which have progressively lost taxing rights as MNEs have found new ways to obtain profits in developing and emerging country markets without triggering source-country taxation. It is undeniable that low-and middle-income countries are usually identified as source countries, and their international tax policies and treaties are based on the twentieth- century division of taxing rights between source and residence states. This traditional dichotomy recognized that source countries could obtain their ‘fair share’ in the international business income tax pie by taxing an MNE’s physical presence within their territorial borders. This consensus is still reflected in the treaty and domestic definitions of permanent establishment (PE) in most low-and middle-income countries, which rely on the presence of physical infrastructure or a personal agent in order to trigger PE taxation.9 As a result of the reduced requirements for physical presence and permanence in the significant economic activities that an MNE may perform in any given market, the scope of PE taxation in source countries has shrunk significantly. This erosion in the size of the slice of low-and middle-income countries can only grow with time, given the technological promise of reducing the time spent in industrial and commercial processes and of reducing the need for physical infrastructure for business activities. The reactions to this erosion have been varied in timing and scope. Some countries have chosen to make unilateral legislative changes, including the introduction of domestic service PE legislation, significant economic presence legislation,10 reducing thresholds, and levying withholding taxes on different types of payments in order to allow the deductibility for the local payer. Other countries have reacted through case law, widening the scope of source taxation through the interpretation of the standards by the judges.11 Another group of countries have chosen to implement new taxes for 8 OECD/G20 Base Erosion and Profit Shifting Project, Addressing the Tax Challenges of the Digital Economy, Action 1—2015 Final Report (Paris: OECD Publishing, (2015), https://doi.org/10.1787/978926 4241046-en. 9 For a comparative perspective on the PE concept in different jurisdictions, see A. A. Skaar, Permanent Establishment: Erosion of a Tax Treaty Principle (Alphen aan den Rijn: Kluwer Law International, 2020). A distinct mention of the irrelevance of physical location and permanence is made at the beginning of the book: ‘Today, permanence and physical location at a specific geographical spot are not a prerequisite for the performance of an economically significant business activity. The area of digital economy has proved that physical presence is not necessary at all’ (s. 2.4). 10 See, for example, Nigeria’s Section 13(2)(c) and (e) of the Companies Income Tax Act and the SEP Order issued in 2021. 11 See, e.g., the 2021 settlement between Netflix and the Italian revenue authorities, in which data centres belonging to third parties were deemed as a permanent establishment even in the absence of Netflix personnel on site. Milan public prosecutor statement available at xx.

994 Natalia Quiñones digital activities, mainly in the form of digital service taxes (DSTs) and equalization levies. In all cases, the only commonality has been unilateral action and, in most cases, a very low efficacy in recovering the revenue lost due to remote and quicker activities performed by MNEs when engaging with the domestic market. These unilateral actions have, in turn, increased the pressure on countries to find a new international consensus that will restore the equilibrium required to make the international tax system sustainable in the long term. In these globalized times, of course, an international tax consensus has a very different meaning than the meaning it had in the past, when OECD countries or even a smaller group of nations could impose the standards for the distribution of taxing rights. Nowadays, global economic output is distributed amongst different types of countries and emerging economies play a very important role. Furthermore, the increased mobility for economic activities now requires that low-and middle-income countries implement the standards at a fast enough pace so that tax does not become a distortive element in the way MNEs generate economic growth and scientific advancement. Given the situation described here, it is indeed in the interest of low-and middle-income countries to actively devise a solution to the tensions that are threatening to reduce their share in an already precarious consensus on the distribution of taxing rights.

54.2 Dominating Narratives Around Possible Solutions A number of possible solutions to the tensions described have occupied the dominating space in the narrative. The most dominant is indeed the work of the OECD Inclusive Framework, given the political push behind it due to the G20 and G7 endorsement. It is important to note at this point that the G20 now comprises eight middle-income countries: Argentina, Brazil, China, India, Indonesia, Mexico, South Africa, and Turkey. Of course, none of these are members of the Group of 7, which is the driving force behind the need to restore the stability of the international tax system, currently threatened by the tensions created by unilateral actions as described earlier. The reason for this is mainly that most MNEs suffering the consequences of diverging applications of the standards in each of the jurisdictions in which they do business are mainly based in the G7 countries, with the notable exception of China. In any case, the mandate for a consensus solution was made by the G20 to the OECD in 2018 with the approval of an interim report,12 and has gone through a number of approvals with a deadline that has now been extended until July 2021. 12 OECD/G20 Base Erosion and Profit Shifting Project, Tax Challenges Arising from Digitalisation— Interim Report 2018: Inclusive Framework on BEPS (Paris: OECD Publishing (2018), https://doi.org/ 10.1787/9789264293083-en.

Challenges of the Emerging International Tax Consensus 995 The OECD Inclusive Framework (IF) narrative for resolving the tensions brought by globalization and technology has focused on the approval of a two-pillar solution that draws from a policy note published in January, 2019.13 The policy note was the result of blending three competing proposals discussed in the IF Task Force for the Digital Economy: the USA, the UK, and the G24 proposal. The first two proposals mainly tried to apply the existing arm’s-length standards to grant a minimal portion of taxing rights to the countries where marketing intangibles were present, in the case of the US solution, and to the countries were users actively added value to the MNE, in the case of the UK solution. This ‘user participation’ solution was also restricted to highly digitalized companies; namely, social media platforms, online marketplaces, and search engines. The third solution, also known as the significant economic presence proposal, was brought by the G24, a group composed mostly of low-and middle-countries seeking to align their monetary and development policies. The introduction of this solution was disruptive in the sense of rejecting the arm’s-length standard and proposing formulary apportionment to deal with the distribution of taxing rights for the digitalizing economy. With the introduction of the G24 proposal, a new technical tension in the international tax system became apparent: a new group of countries supported formulary apportionment over the arm’s-length standard in the design of a new international tax system. This additional technical tension brings new complexity to the design of a stable solution, as the underlying technical tools to tackle the challenge of the digitalization of the economy must either strive to continue with the existing framework or must create an entirely new framework that will at least partially replace the arm’s-length paradigm. The OECD IF chose a ‘middle path’ in the 2019 policy note, admitting formulary apportionment but only for a minimal portion of the profits of a few companies that met a number of established thresholds.14 The rationale for narrowing down the application of formulary apportionment drew from the user participation and marketing intangibles proposals under Pillar One. As a result, the Pillar One document advocates for the continuation of the arm’s-length standard concurrently with formulary apportionment on the miniscule portion of the profits known as ‘Amount A’. This attempt to reconcile such diverging views has created yet a new tension on the narrative driven mainly by the concerns expressed by business on the probable disputes that will arise in the cases where an MNE is under an obligation to calculate Amount A while being subject to the arm’s-length standard over 100% of its profit. The tax certainty documents released for public consultation in May 2022 reveal that all the IF members which ratify the Pillar One multilateral convention (regardless of

13 OECD/G20 Inclusive Framework on BEPS, ‘Addressing the Tax Challenges of the Digitalisation of the Economy—Policy Note’ (2019), https://www.oecd.org/tax/beps/policy-note-beps-inclusive-framew ork-addressing-tax-challenges-digitalisation.pdf. 14 For an analysis of the estimated 108 taxpayers covered by Pillar One, see L. Eden, ‘Taxing The Top 100—Part 1: Who’s In, Who’s Out?’ (2021), available at https://news.bloombergtax.com/daily-tax-report/ taxing-the-top-100-part-1-whos-in-whos-out.

996 Natalia Quiñones the level of development or mutual agreement procedure (MAP experience) will participate in a dispute-prevention mechanism modelled on the International Compliance Assurance Program (ICAP). Under the proposed mechanism, the tax administration of the ultimate parent entity (UPE), also called the lead tax administration, will propose a solution to certainty requests by any resident company. This solution may be reviewed by a review panel composed of tax administrations randomly selected from volunteers listed by the taxpayer in jurisdictions where certainty is needed. Other interested jurisdictions may propose alternative solutions to be reviewed by a determination panel the composition of which is still being agreed, but whose decision will be binding for all tax administrations. Again, the stronger role assigned to high-income countries acting as lead tax administrations reveals a continued pressure on developing countries to conform to the standards set and applied in these more developed jurisdictions. Pillar Two, on the other hand, is inspired by the popular idea of creating a ‘floor’ to tax competition by implementing a minimum tax on the profits of the large MNEs in order to guarantee that the profits are taxed somewhere. This premise is universally attractive, especially after the public outcry produced by data leaks showing the opportunities available to MNEs to reduce taxation dramatically or, sometimes, entirely, without breaking any law. This narrative has dominated the press releases and reports available to the public, but little attention is being paid to the implications of the global anti-base erosion (GloBE) measure in the distribution of the residual taxing powers. The policy note was clear in its intention to preserve the traditional dichotomy of source and residence, again giving primacy to the residence state by assigning the rights to the ultimate parent residence jurisdiction in the application of the income inclusion rule (IIR). The only rule contained in the Blueprints for Pillar Two15 that could potentially bring further rights for low-and middle-income countries is the subject to tax rule (STTR), which could allow source countries to apply the full withholding tax rate to a payment protected by a treaty when there is evidence that the payment was taxed at a lower rate than the established minimum rate. The undertaxed payments rule (UTPR), initially designed to be applied by source countries, probably will not have any practical application as it can only be invoked if the residence country does not apply an IIR. However, the difficulty in establishing whether the payee was effectively taxed and at what rate has made it the target of criticism from business and some academics. The qualified minimum tax, on the other hand, allows source countries to levy the top-up tax if the MNE has an effective tax rate below 15% in their jurisdiction. Although this mitigates the idea of giving up taxing rights in favour of the UPE jurisdiction, it still limits the possibility for developing countries to use income tax benefits to attract investment or to correct domestic imbalances created by natural disasters or armed conflict. In general, the attempt to devise an all-inclusive solution has proven to be extremely difficult,

15 OECD,

Tax Challenges Arising from Digitalisation—Report on Pillar Two Blueprint: Inclusive Framework on BEPS (Paris: OECD Publishing, 2020), https://doi.org/10.1787/abb4c3d1-en.

Challenges of the Emerging International Tax Consensus 997 resulting in a very high degree of complexity in the technical writeup for both Pillar One and Pillar Two. The issue of complexity itself can be seen as a new challenge for low-and middle- income countries, which may feel alienated from the dominating narrative due to their lack of experience, for instance in applying profit splits domestically. In fact, complexity has probably been the most significant argument against the dominating narrative by the OECD IF. The public consultation processes have indeed revealed that the application of the new standards may prove too difficult even for large MNEs with the necessary resources and advice. Most public consultations held in 2021 and 2022 have asked for input on possible simplification measures for both Pillar One and Pillar Two,16 which shows that even the OECD Secretariat is concerned about the actual applicability of the solution on a worldwide scale. As discussed earlier, the dominating narrative of the OECD IF introduces new tensions and challenges for low-and middle-income countries, even when they are presented as participating ‘on equal footing’17 in all the discussions preceding the release of the public documents. The main challenge is related to the complexity and practical difficulties in the application of both Pillar One and Pillar Two. These challenges have been noted in other narratives, some of which have occupied a relevant space in the discussions. One of these counter-narratives was developed by the UN Committee of Experts on Tax Matters. The UN Committee opted for a simpler, less ambitious approach to tackle some of the challenges and tensions brought by the digitalization of the economy. The narrative adopted by the UN is limited in scope, as it can only be implemented in the framework of a bilateral treaty that has been signed or modified to include article 12B of the UN Model Convention. This implies a challenge of renegotiating existing treaties or agreeing to new treaties in a world where source countries are rethinking the advantages of having a treaty network at all.18 Furthermore, the article is limited to income arising from automated digital services (ADS), which only represents part of the businesses that can be conducted remotely and with less permanence in the market or source jurisdiction. The UN narrative champions simplicity in preserving the primacy of residence- country taxing rights while allowing for source-withholding. In fact, withholding taxes are the simplest and most effective tools for developing countries to collect revenue from foreign companies doing business in their territory. They also manage to 16 See

the public consultation documents and presentations available at https://www.oecd.org/tax/ planned-stakeholder-input-in-oecd-tax-matters.htm. 17 See, e.g., the description of the OECD IF in the statement published 31 January 2020, available at https://www.oecd.org/tax/beps/statement-by-the-oecd-g20-inclusive-framework-on-beps.htm#:~:text= The%20O E CD%2FG20%20In c lus i ve%20Fr amew ork%20on%20B E PS%20(Inclus i ve%20Fr amew ork),address%20the%20tax%20challenges%20of. 18 See, e.g., P. Janský and M. Šedivý, ‘Estimating the Revenue Costs of Tax Treaties in Developing Countries, The World Economy 42/6 (2019), 1828–1849, who concludes that tax treaties can cost up to 0.1% of GDP based on the experience of sub-Saharan Africa and some Asian countries such as the Philippines.

998 Natalia Quiñones address the concern for excess withholding by allowing the taxpayer to elect for net taxation, introducing certainty in the determination of said taxable profits. However, this simplified approach reveals yet another tension that has also been acknowledged by the literature:19 taxation of gross profits and the arm’s-length standard as currently defined in the OECD Transfer Pricing Guidelines are not fully compatible. Both dominating narratives have thus avoided centring the discussion on the adequacy of the arm’s-length standard going forward. The G24 proposal and some academics20 and even practitioners21 have claimed that the arm’s-length standard (ALS) is no longer fit for purpose, given the increased mobility of functions, risks, and assets across jurisdictions, and given the significant changes introduced by new technologies in the way that functions are performed. They also argue that the separate-entity approach is unrealistic given the increasing levels of integration within multinational groups. However, the resistance to changing the contents of the ALS or substituting it entirely is still widespread in the dominating narratives. The value-creation paradigm continues to justify the policy objective of preserving the ALS, even when the new circumstances demand a change in its content in order to reduce the space for the transfer of profits and losses by MNEs to more convenient jurisdictions. It is important to note, also, that substituting the ALS for formulary apportionment will not necessarily alter the current distribution of taxing rights, as the actual distribution will depend on the contents of the formula and the weight given to each of the factors. In other words, a formula that replicates the current panorama by functions, risks, and assets would not change the distribution of taxing rights. An incremental change in the distribution of taxing rights is therefore possible and highly achievable now that over 140 countries have agreed to the elements that create value. It is difficult to tell at this point if the deadlock in the consensus will give way to a more radical narrative that truly brings countries together in designing a global policy that will distribute taxing rights in a way that ensures stability in the long term. This would, of course, require addressing the tensions mentioned so far in this chapter, that is: (1) designing rules that do not depend on permanence or physical presence to assign taxing rights; (2) designing a simple system that can be applied and enforced equally by all jurisdictions, including by low-income countries with limited resources. Given the complexity involved in the application of the ALS and the separate entity approach, this system should either include net taxation based on global 19 A.

P. Dourado, ‘Editorial: The OECD Report on Pillar One Blueprint and Article 12B in the UN Report’, Intertax 49/1 (2021), 3–6. 20 See, e.g., R. S. Avi-Yonah and K. A. Clausing, ‘Toward a 21st-Century International Tax Regime’, Tax Notes International (2019), 839–849 or A. Ezenagu, ‘Faltering Blocks in the Arguments against Unitary Taxation and the Formulary Apportionment Approach to Income Allocation’ Asper Review of International Business & Trade Law 17 (2017), 131. 21 See A. Hartley, ‘Wide Support for Limited Formulary Apportionment at OECD Public Consultation’, International Tax Review (2019).

Challenges of the Emerging International Tax Consensus 999

formulary apportionment or allow for local gross taxation. Alternatively, the system could go a step further in cooperation and create a central International Tax Organization to administer a global tax for global companies. This seems plausible now that the tax base has been agreed in Pillar Two and the initial taxpayers have been identified in Pillar One; (3) excluding and replacing all incompatible systems in order to prevent tax disputes; (4) reducing the space for shifting tax losses and profits in spite of increasing mobility for assets, risks, and functions, even outside planetary borders; (5) ensuring the sustainability of the system by granting taxing rights to all jurisdictions involved in the generation of profits, or even to all jurisdictions; (6) creating a central body to develop the policy interpretations needed in the face of future changes in the business or technological landscape, as well as to decide any conflict; (7) ensuring that the approved measures do not interfere with technological advances and the digitalization in the economy, and that access to new services and technologies is still widely available across all jurisdictions.

Unfortunately, none of the current dominating narratives address all of the tensions required to create a stable international regime in the long term. Some academics have proposed other solutions that have not necessarily made it into the dominant narratives, but that could contribute valuable elements to the discussion. The challenge for low- and middle-income countries would be to review these alternatives and decide if the merit is justified for a long-term solution that can be brought to the main discussion either at the OECD IF or at the UN. Avi Yonah and Clausing,22 advocate for a sales-based formulary apportionment based on the ease of measuring the demand side of value creation. Although they favour a multilateral implementation, and applaud the significant economic presence proposal, they conclude that it is more feasible to start with unilateral implementation by the USA. Starting with a domestic US law seems sensible given the history of international taxation and the constant replication of US rules by the OECD and the EU.23 The proposed formula excludes assets and payroll, given the relative mobility of those factors as compared to the relative immobility of the location of customers or users. This may favour bigger markets over smaller markets, but the authors argue that the revenue for developing countries is still larger than the revenue they receive with the current application of the ALS with the tax planning and avoidance of physical presence and

22 R. S. Avi-Yonah and K. A. Clausing, “Toward a 21st-Century International Tax Regime’, Tax Notes International (26 Aug. 2019), 839–849, University of Michigan Public Law Research Paper No. 656, University of Michigan Law & Economics Research Paper No. 20-001, available at https://ssrn.com/abstr act=3488779. 23 On the Washington, Paris, Brussels trifecta, see M. Hearson, Imposing Standards: The North–South Dimension to Global Tax Politics (Ithaca, NY: Cornell University Press, 2021), doi:10.1353/book.83892.

1000 Natalia Quiñones permanence in the activities of source countries. The proposal also considers a sales threshold in order to replace physical presence or permanence requirements as a nexus for taxation. Báez and Brauner24 have proposed the implementation of a withholding tax on base- eroding payments made to non-residents, with a low standard rate and a higher rate for payments made to residents of non-cooperating jurisdictions. This solution would apply only when the payment is not otherwise taxed in the payer jurisdiction by virtue of domestic or treaty rules, thereby preserving the existing ALS but capturing some revenue for source jurisdictions with a mechanism that they are used to applying and have capacity to administrate. The withholding responsibility is transferred to financial institutions for the payments made by individuals. Christians and Magalhaes,25 on the other hand, is advocating for a global excess profits tax (GEP), based on the fact that after the Covid-19 pandemic the world economy is divided into a majority of firms experiencing financial contraction while a minority of large, highly digitalized, firms are experiencing extraordinary profits and windfalls due to the mobility restrictions brought by the pandemic. The general idea, thus, would be to capture those windfall gains and distribute them according to the Pillar One and Two rules, but creating a universal rate of 30% for the excess profits (Amount A under Pillar One) on top of the regular corporate income tax. This approach indeed increases the total tax paid by MNEs conducting business remotely and with new speedy technologies, and it would allocate the GEP to market countries in the same proportion as Pillar One allocates a residual portion of non-routine profits to the market jurisdiction. Devereux and others,26 in turn, published a robust report in 2021 recommending two proposals for reform: the Residual Profit Allocation by Income (RPAI) and the Destination Based Cash Flow Tax (DBCFT). The first proposal resembles the OECD IF Pillar One but allocates the entire residual profit to the destination or market jurisdiction, therefore widening the scope of application of formulary apportionment but maintaining the existing tensions involved in the application of the ALS to the routine profit. Furthermore, the distinction between routine and non-routine profits is to be made under the current ALS criteria, including the use of comparables which entails significant complexity and increases administrative costs. The DBCFT, in contrast, opts for allocating the entire profit to the country where the sales take place, thereby

24

A. Báez and Y. Brauner, ‘Taxing the Digital Economy Post BEPS ... Seriously?’, University of Florida Levin College of Law Research Paper No. 19-16 (1 Mar. 2019), available at https://ssrn.com/abstract=3347 503 or http://dx.doi.org/10.2139/ssrn.3347503. 25 See A. Christians and T. Diniz Magalhaes, ‘The Case for a Sustainable Excess Profits Tax’ (24 Mar. 2021), available at https://ssrn.com/abstract=3811709 or http://dx.doi.org/10.2139/ssrn.3811709 and T. Diniz Magalhaes and A. Christians, ‘Rethinking Tax for the Digital Economy After COVID-19’ (2020), available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3635907. 26 M. P. Devereux et al., Taxing Profit in a Global Economy (Oxford: Oxford University Press, 2021)..

Challenges of the Emerging International Tax Consensus 1001 eliminating the ALS and simplifying the enforcement. This proposal, however, has been opposed by a number of academics,27 especially if it were adopted unilaterally. Hongler and Pistone28 had an early suggestion for a virtual PE, granting taxing rights to source countries whenever a digital presence could be confirmed. This development operates in the framework of the ALS and existing treaties, but extends source-taxing rights in the face of digital activities involving any specific market. Reducing the shock associated with transitioning to a new regime is one of the virtues of this proposal, although it may require higher capacity-building investments in low-and middle-income countries in order to locally enforce the rules. Finally, Picciotto and others29 suggest that a minimum effective tax rate replaces the GloBE as set out in OECD IF Pillar Two. Under this system, MNEs would calculate the non-effectively taxed profits based on the effective tax rate (ETR) per country and then the aggregate result would be distributed among the countries that meet a substance rule based on the activities actually performed by the MNE in each jurisdiction. Although simpler than the combination of Pillar One and Pillar Two, the minimum effective tax rate still requires extensive administrative efforts and still provides incentives for countries to compete rather than cooperate with their domestic tax legislation.

54.3 What Can Low-and Middle-Income Countries Expect It is clear that the international tax system is facing a crisis due to the unresolved tensions and the competing dominating narratives which do not address the underlying issues that threaten the stability of the system. The question, therefore, is the role of low- and middle-income countries in the construction of a new solution, or whether it is in the interest of these jurisdictions to allow the current crisis to continue. At this point, it is important to recall the statistics mentioned in the first section of this chapter, which conclude that labour-intensive countries are presently losing revenue in the current scenario, even with unilateral measures that try to recapture lost revenue due to globalization and technological advances. It therefore seems logical to at least participate in the discussion, especially when the dominant narrative has opened a space for low-and middle-income countries regardless of their geopolitical status. 27 See,

e.g., J. Becker and J. Englisch, ‘Unilateral Introduction of Destination- Based Cash- Flow Taxation’, International Tax Public Finance 27 (2020), 495– 513, https://doi.org/10.1007/s10 797-019-09579-0. 28 P. Hongler and P. Pistone, ‘Blueprints for a New PE Nexus to Tax Business Income in the Era of the Digital Economy’, WU International Taxation Res. Paper Series No. 2015-15 (2015), https://ssrn.com/ abstract=2591829. 29 S. Picciotto et al., ‘For a Better GLOBE. METR: A Minimum Effective Tax Rate for Multinationals’ (2 Mar. 2021), available at https://ssrn.com/abstract=3796030.

1002 Natalia Quiñones It is indeed true that the Inclusive Framework has invited all countries to join the discussions prior to agreement to comply with the BEPS minimum standards as well as with the Global Forum on Transparency standards. This apparently low cost in joining the discussion comes at a high price for some low-and middle-income economies that have limited administrative capabilities, as complying with the exchange-ofinformation rules alone requires the implementation of software, capacity building, and dedicated personnel. However, if done properly, it can indeed increase the ability to collect taxes from high net worth individuals and some domestic corporations that commonly exploit the lack of access to information in low-and middle-income countries in order to shield wealth from local taxation. It also aids developing countries to halt illicit financial flows, enhancing the fight against corruption and financial crime. Joining the discussion, in any case, also comes with the challenge of sending professionals that have the capacity of understanding and contributing to the discussion in complex matters such as the profit-split method in transfer pricing. This has been made easier with the pandemic, as the costs of flying to Paris for every meeting may very well become impossible for the budget of tax administrations located in far geographies. Online participation, however, does not necessarily guarantee that the voices of a low-or middle-income country will be taken as relevant in the drafting process.30 As mentioned earlier, the G20 and G7 leadership are certainly the drivers for the OECD IF Secretariat’s technical work, and the Steering Group voices are the defining gauge for consensus before any document is released to the public. One of the opportunities that comes with joining the discussions on the dominant forum (the OECD IF) is precisely the requirement for consensus implied in the mandate. Although the definition of consensus in the context of the IF is not strictly approval by all but one of the participating jurisdictions, there is still a possibility of engaging with like-minded jurisdictions in order to express dissent in a way that is politically relevant, as will be discussed in the following section. The short deadlines initially proposed by the OECD Secretariat were accompanied by significant scepticism as to the probabilities of actually achieving a consensus solution that had enough political viability to become a reality. After the postponements due to the pandemic and the difficulties in reaching political agreements and commitments, the scepticism regarding the actual implementation of the two-pillar solution is even higher. Because of this situation, an increasing number of countries, including a few low-and middle-income countries, have opted for the adoption of unilateral measures (DSTs, equalization levies, and significant economic presence legislation) as a substitute for the multilateral agreement that may not happen. It is indeed logical to expect that the complexity will stop many countries from adopting a multilateral instrument containing rules that cannot be enforced with 30 On the decision- making powers of developing countries in the IF, see R. Corlin Christensen, M. Hearson, and T. Randriamanalina, ‘At the Table, Off the Menu? Assessing the Participation of Lower- Income Countries in Global Tax Negotiations’, Working Papers 15853, Institute of Development Studies, International Centre for Tax and Development (2020).

Challenges of the Emerging International Tax Consensus 1003 the current administrative capability of the local tax administration. As noted by the African Tax Administration Forum (ATAF),31 the challenges faced by most African countries are very different to the ones faced by their more industrialized counterparts in the IF. However, academics such as Rukundo32 continue to discourage the adoption of unilateral measures, especially if designed with unique elements. The rationale behind this critique concerns the difficulties that will most certainly arise if each jurisdiction chooses a different scope, rate, or general definitions in the application of their domestic legislation. As mentioned previously, for multinational companies operating in more than 150 jurisdictions, these differences represent an impossible scenario that may lead to a reduced supply of digital services in the jurisdictions where the rates are higher or the scope is broader. These unilateral actions must, therefore, be studied carefully, and hopefully implemented in agreement with regional competing jurisdictions as they may reduce the attraction of foreign investment for low-and middle-income countries acting alone. As seen in 2020, these actions may also trigger undesirable reactions from jurisdictions such as the USA, which has chosen to prosecute DSTs with trade sanctions that may prove costly for countries exporting products and services to the USA. In this context, designing measures that do not specifically target the largest, US-based multinationals may be helpful in avoiding retaliation for low-and middle-income countries. The timing of the measures is also key, given that a later adoption of a DST or equalization levy may bring a competitive advantage if implemented in isolation from other countries in the region. On the other hand, it may be politically costly to delay the design of said rules if there is no agreement from a critical mass of countries by July 2023. In that event, and given the short window of political opportunity for approving domestic tax reform, it is advisable for developing country parliaments to implement an early (maybe temporary) solution that will levy taxes from the companies that have profited the most from the pandemic, that is, highly digitalized companies. In the alternative scenario in which an agreement is reached by a critical mass of countries by July 2023, the political pressure on all IF countries to sign the multilateral instrument containing the agreement will be substantial, which is yet another incentive for developing countries to ensure that the simplified mechanisms adopted are satisfactory for their current capacities. The options for simplification can also be inspired in some of the non-dominant narratives discussed earlier, and in any case should leave room and enough flexibility to change and adapt the rules in the future. It is possible that Pillars One and Two will constitute a small step towards the change that the international tax system requires in order to resolve existing tensions. In that case, low-and middle-income countries can expect to reconvene and advance down that path in the

31 ATAF, ‘The Tax Challenges Arising in Africa from the Digitalisation of the Economy’, ATAF 1st Technical Note, CBT/TN/01/19 (2019). 32 S. Rukundo, ‘Addressing the Challenges of Taxation of the Digital Economy: Lessons for African Countries’, ICTD Working Paper 105, Brighton, IDS (2020).

1004 Natalia Quiñones near future, when fears associated with the transition shock are dissipated in the political and practical arenas. As for the proposed article 12B, there is no reason why low-and middle-income countries should refrain from including it in their treaty models and in any new negotiations, especially South- to- South treaties, as well as introducing it in any renegotiations that take place. However, only time will tell if countries are effectively including this treatment of ADS in their bilateral relationships. Low-and middle-income countries may consider the elements of non-dominating narratives to either support the discussions in the dominating forum or to construct unilateral solutions following the example of India33 or Nigeria.34 As to the administrative capabilities of low-and middle-income countries, the required investments will vary depending on the adopted solution. As many have noted,35 current capabilities are not necessarily aligned with the requirements of the complex solutions presently proposed under Pillar One and Pillar Two. In the case of an agreement at the level of the OECD IF, low-and middle-income countries should expect a need to expand their budgets for capacity-building programmes, software acquisition, and travel. A smaller budget may still be necessary in a world of unilateral DSTs and significant economic presence, as controlling the number of local users requires greater capacity in the tax administration, including the possibility of processing large amounts of data. Knowing that capacity-building needs will continue to grow, it seems wise to establish regional or global cooperative mechanisms where low-and middle-income countries can share the costs and take advantage of mutual learning curves. This type of cooperative mechanism, as shown in the next section, is certainly the most efficient way of dealing with the new tax challenges for low-and middle-income countries.

54.4 Towards a Strengthened Developing-C ountry Alliance Unilateralism has proven to be ineffective in protecting low-and middle-income countries’ tax base from the challenges and tensions brought by globalization and new technologies. Furthermore, unilateral measures like DSTs and equalization levies have

33 Indian

Department of Revenue, Central Board of Direct Taxes, ‘Public Consultation on the Proposal for Amendment of Rules for Profit Attribution to Permanent Establishment’, F. No. 500/33/ 2017-FTD.I (18 Apr. 2019). 34 Nigerian Ministry of Finance, ‘Companies Income Tax (Significant Economic Presence) Order’, FGP 65/52020/150 (10 Feb. 2020). 35 A. Sanchez-Castro, ‘Administrative Capability Analysis of OECD Proposals from the Perspective of Developing Countries’, Intertax 48/2 (2020).

Challenges of the Emerging International Tax Consensus 1005 added a layer of complexity, divergent interpretations, and an increased risk of excess taxation that can result in a shortage or costliness of digital services for the citizens of small markets and the developing world. It is thus suggested that a multilateral solution that truly seizes the opportunity to reshape the international tax standards is the best shot that low-and middle-income countries have in the long term. Multilateralism is also known for reducing costs and enhancing the learning curve for all jurisdictions involved, as is the case for the Global Forum on Transparency and the automatic exchange of information. This particular example shows that the narratives play a crucial role in achieving political consensus. In the case of the exchange of information, the dominating narrative posed the collaborative solution as a collective way of defeating tax evaders, rather than a fight between investment hubs and countries wanting the information to apply their domestic rules to a mobile resident taxpayer. In the case of the challenges posed by the digitalization of the economy, the collaborative or multilateral solution is still viewed as an antagonistic fight between residence and market countries in the dominating narratives. In contrast, narratives that bring global governments together are clearly more effective in achieving consensus and, in this case, it is true that all countries can come together to ensure that global mobility and new technologies will not only benefit the pockets of a few privileged shareholders, but rather benefit citizens of all markets where the MNEs make profits. In all cases, finding commonalities with other low-and middle-income countries will definitely increase the possibilities for including relevant simplification measures, feasible administration and collection, and, if enough countries join the effort, greater allocation of taxing rights in the long term. The most immediate proof of the power of joining voices towards a common goal is precisely the introduction of the G24 proposal and the admission of formulary apportionment as a viable alternative in a forum that had traditionally forcefully rejected this alternative. This example also shows that there are available forums for developing countries to come together in identifying common interests before discussing individually in the IF working groups, where it is already difficult to influence the discussion unless a relevant geopolitical power backs the country’s proposition. Also, developing country members of the IF can communicate directly to schedule preparatory or debriefing calls which are always useful in finding commonalities that can be brought forward as a consolidated developing-country position. Having a unified common voice is more likely to generate a change in the dominating narrative, especially in view of the consensus requirement in the OECD IF. Furthermore, broadening the scope of countries participating in the common view also has the advantage of preventing destructive patterns of aggressive tax competition between countries that could otherwise act as regional allies. Even in the case of a lack of agreement in the IF, having an agreement on the structure, content, and timing of the implementation of unilateral DSTs, equalization levies, significant economic presence, or changes to the domestic PE legislation will prove more efficient for low-and middle- income countries.

1006 Natalia Quiñones Ideally, working with a mindset based on commonalities rather than differences will also permeate the discussion environment in the OECD IF, so that in the medium term all member countries can finally address the underlying challenges in a way that is flexible enough to address new changes brought by new technologies and sustainable enough that countries will apply it with significant uniformity in the long term.

Chapter 55

Gl obal Tax Gov e rna nc e Irma Mosquera Valderrama

55.1 Introduction The term global tax governance has been used by political scientists and tax scholars to address the changes in international tax lawmaking and the interaction of countries (both developed and developing) with international (OECD, UN, IMF) and regional (EU) organizations. For instance, in political science, Thomas Rixen addressed the established institutional infrastructure to deal with double taxation (bilateral tax treaties, model treaties, and institutions, i.e. the OECD) and the need to find cooperative solutions to deal with the issue of double non-taxation which at that time (2008) was only addressed through unilateral (domestic) rules.1 In his work, Rixen discussed, in addition to double taxation, other concepts of international taxation such as source vs residence, fairness, and international tax neutrality from an international political economy perspective.2 Later, in 2016, Rixen, along with Peter Dietsch, provided a definition of global tax governance stating that it: consists of the set of institutions governing issues of taxation that involve cross- border transactions or have other international implications. This definition implies that global tax governance need not, but could, involve a full or partial shift of the power to tax, that is, the right to impose taxes on citizens, to the international level. Currently, the right to tax is firmly tied to the nation-state. While global tax governance circumscribes and shapes a nation’s power to tax in various ways, it exclusively consists of institutions governing the interaction among national tax systems.3 1

T. Rixen, The Political Economy of International Tax Governance (Springer, 2008), 1and 81. Ibid., 57–85. 3 P. Dietsch, and T. Rixen, T., ‘Global Tax Governance: What It Is and Why It Matters’, in P. Diestsch and T. Rixen, eds, Global Tax Governance: What is Wrong With It and How To Fix It (Rowman & Littlefield/ECPR Press, 2016), 3. 2

1008 Irma Mosquera Valderrama Tax scholars have also referred to the need to address fairness, neutrality, sovereignty, and the role of international organizations in dealing with tax cooperation and tax competition. For instance, in 2009, Ring discussed the question of ‘how sovereignty shapes arguments over the merits of tax competition and how sovereignty influences the design of responses to tax competition’.4 The design of the international tax regime has also been addressed by Brauner who explores the benefits of a truly global approach for efficiently resolving the challenges in international taxation. He proposes a full set of international tax rules in the form of a multilateral treaty. For this author: key elements of the success of such a treaty would be: (1) the participation of a significant part of the major economic and trade forces in the world; (2) acceptance of non-negotiable rules as domestic legislation, such as the interpretation and arbitration clauses, which also would apply to trade with nontreaty partners; (3) a flexible but binding interpretation system similar to the OECD model commentaries; and (4) an easily accessible conflict resolution system, which would be open to the individual residents of the treaty partners.5

In addition, in her article on hard law vs soft law in international taxation in 2007, Christians6 addressed the use of these terms to explain the degree of global adherence by countries to various tax practices. Later, Dourado also discussed the validity of global standards in tax law, in this case exchange of information.7 In these contributions, tax scholars, as has also been done by political scientists, have referred to the interaction between international organizations and countries when designing international tax rules. However, unlike political scientists, tax scholars at that time did not make reference to global tax governance but to sovereignty, tax competition, and tax cooperation.8 Therefore, there was a disparity in the interaction between tax and political scientist scholars when addressing issues of global tax governance. This has been resolved in recent research projects9 and recent articles by tax scholars.10 4

D. Ring, ‘Who is Making International Tax Policy: International Organizations as Power Players in a High Stakes World’, Fordham International Law Journal 33 (2009), 649. 5 Y. Brauner, ‘An International Tax Regime in Crystallization’, Tax Law Review 56 (2002), 259, 263. 6 A. Christians, ‘Hard Law, Soft Law, and International Taxation’, Wisconsin International Law Journal 25 (2007), 325. 7 A. P. Dourado, ‘Exchange of information and validity of global standards in tax law: Abstractionism and expressionism or where the truth lies’ (2013), https://cadmus.eui.eu/bitstream/handle/1814/26059/ RSCAS_2013_11.pdf?sequence=1&isAllowed=y. 8 M. Keen and A. Simone, ‘Is Tax Competition Harming Developing Countries More Than Developed?’, Tax Notes International 34/13 (2014), 1317–1326; Brauner, ‘An International Tax Regime in Crystallization’. 9 ERC Project GLOBTAXGOV, see Section 55.8. 10 S. A. Rocha and A. Christians, Tax Sovereignty in the BEPS Era (Kluwer Law International, 2016), 1137–1196; R. Azam, ‘Ruling the World: Generating International Tax Norms in the Era of Globalization and BEPS’, Suffolk University Law Review 50 (2017), 517; T. D. Magalhães, ‘What Is Really Wrong with Global Tax Governance and How to Properly Fix It’, World Tax Journal 10/4 (2018), 499–536; S. Kingma,

Global Tax Governance 1009 Notwithstanding the above, the boundaries between international tax cooperation and global tax governance are still indistinguishable. Some scholars discuss international tax cooperation as part of global tax governance and state that if all countries cooperate, then global tax governance will be achieved.11 However, other scholars may question the use of the terminology of global tax governance since the use of global governance may involve ‘imposing outcomes on people, to the benefit of some and at the expense of others’.12 For Hurd, ‘Global governance scholars who adopt the cooperation premise should be alert to its biases and costs. By assuming that global governance follows from mutual interests pursed through cooperation, they naturally interpret international institutions as devices aimed at the common good.’13 Therefore, further research needs to be conducted on what is meant by international tax cooperation and how it may or may not contribute to global tax governance. By assuming that the term global tax governance is used to impose outcomes on people, the question that should be asked is whether this is true and countries (mainly lower income/developing countries) still follow these (imposed) outcomes by the OECD, under what conditions can the model of global tax governance be feasible and legitimate for both developed and developing countries. This question is currently addressed in the framework of the GLOBTAXGOV ERC-funded project.14 This chapter aims to contribute to the discussion of global tax governance by proposing a framework with some of the elements that can be taken into account by tax scholars when studying this issue.15 For this purpose, the chapter is divided into Inclusive Global Tax Governance in the Post-BEPS Era (IBFD Publications, 2019); I. Ozai, ‘Institutional and Structural Legitimacy Deficits in the International Tax Regime’, World Tax Journal 12/1 (2019), 53–78; R. Mason, ‘The Transformation of International Tax’, American Journal of International Law 114/3 (2020), 353–402; L. V. Faulhaber, ‘Diverse Interests and International Legitimation: Public Choice Theory and the Politics of International Tax’, AJIL Unbound 114 (2020), 265–269. 11 Kingma, Inclusive Global Tax Governance in the Post- BEPS Era; I. Ozai, ‘Two Accounts of International Tax Justice’, Canadian Journal of Law & Jurisprudence 33/2 (2020), 317–339. 12 I. Hurd, ‘The Case Against International Cooperation’, International Theory 14/2 (2020), 1 and 20. In his article, Hurd challenges the view that international law, institutions, and organizations can be classified unproblematically as examples of international cooperation. He suggests that instead of using the terms international tax cooperation and governance interchangeably there should in their place be the cooperation theory, ‘an overtly political methodology that assumes that governance—global or otherwise—necessarily favors some interests over others’. 13 Ibid., 1, 20. 14 See Section 55.8. 15 For a political science approach, see, e.g., R. Corlin Christensen and M. Hearson, ‘The New Politics of Global Tax Governance: Taking Stock a Decade After the Financial Crisis’, Review of International Political Economy 26/ 5 (2019), 1068– 1088; W. Lips, ‘Great Powers in Global Tax Governance: A Comparison of the US Role in the CRS and BEPS’, Globalizations 16/1 (2018), 104–119; T. Büttner and M. Thiemann, ‘Breaking Regime Stability? The Politicization of Expertise in the OECD/G20 Process on BEPS and the Potential Transformation of International Taxation’, Accounting, Economics, and Law: A Convivium 7/1 (2017), https://doi.org/10.1515/ael-2016-0069; R. Eccleston and H. Smith, ‘The G20, BEPS and the Future of International Tax Governance’, in Dietsch and Rixen, Global Tax Governance: What Is Wrong With It and How To Fix It, 175–198).

1010 Irma Mosquera Valderrama seven sections. Section 55.2 will provide an introduction to the recent international tax standards, mainly exchange of information and base evasion and profit shifting (BEPS) as developed by the OECD with the political mandate of the G20. Section 55.3 will address the use of soft law vs hard law to introduce international tax standards. Section 55.4 will address the role of developing countries in the BEPS Inclusive Framework and the peer review of the BEPS minimum standards. Section 55.5 will address the validity of the outcome of these international tax standards and Section 55.6 will address the role of the actors in global tax governance (countries, international and regional organizations, business, and civil society, among others). The final section (Section 55.7) will provide some concluding remarks.

55.2 The Development of International Tax Standards and the Role of the OECD 55.2.1 Development of International Tax Standards In the past, international organizations such as the OECD and the UN provided recommendations, guidelines, and standard models that have been used by developed and developing countries unilaterally or in bilateral negotiations. Before the BEPS Project, the role of the OECD was to develop standards, recommendations, and models that were followed by OECD countries and in some cases (e.g. the OECD Tax Treaty Model) by non-OECD countries.16 Despite the role of the UN (with 193 countries) as a representative of developing countries, the OECD has taken an important role in the development of tax rules, for instance regarding the use of the OECD Tax Treaty Model17 and the OECD guidelines (e.g. Transfer Pricing Guidelines) by OECD and non-OECD countries.18 In the light of the current developments in taxation, it can safely be argued that, despite the call from

16 See for an earlier analysis of the role of the OECD, A. J. Cockfield, ‘The Rise of the OECD as Informal World Tax Organization Through National Responses to E-Commerce Tax Challenges’, Yale Journal of Law and Technology 8 (2005), 136–187; H. J. Ault, ‘Some Reflections on the OECD and the Sources of International Tax Principles’, Tax Notes International 70/12 (2013), 1195–1201. 17 Some of these reasons could be seeking accession to the OECD (political motivation) or the result of the, bargain negotiation power between OECD and non-OECD countries. See I. J. Mosquera Valderrama, ‘Legitimacy and the Making of International Tax Law: The Challenges of Multilateralism’, World Tax Journal 7/3 (2015), 355–356. See for a study of the use of the UN Model W. Wijnen and J. de Goede, The UN Model in Practice 1997–2013 (IBFD, 2013). 18 For an overview of the application of the arm’s-length principle and the Transfer Pricing Guidelines by non-OECD countries, see the country profiles available at http://www.oecd.org/ctp/transfer-pricing/ transferpricingcountryprofiles.htm.

Global Tax Governance 1011 the UN and civil society19 to give the UN a more representative role in international tax negotiations, the OECD continues to occupy this predominant role. However, this may change due to the discussions taking place since November 2022 at the United Nations level. These discussions have been initiated by Nigeria with the support of some African countries. These countries submitted a proposal for a draft resolution at the UN Second Committee on the Promotion of Inclusive and Effective International Tax Cooperation at the United Nations by means of developing a framework or instrument. This resolution has been approved by the General Assembly in December 2022. This resolution gives the United Nations, a more important role in the setting of international tax standards. At this time, input has been sought (i) on the content of this framework or instrument and (ii) on the role of the United Nations to achieve inclusive and effective international tax cooperation. From a tax and international relations perspective, Christians and Keohane20 identified the OECD as an organization with a networking role that has facilitated the spread of best practices. Keohane further stated that this network role takes place by means of ‘linking national officials and quasi-public bodies with their foreign counterparts for the purpose of joint decision making, coordination or information sharing’.21 Later the OECD introduced multilateral projects in taxation with the political mandate of the G20, for which the main focus was the introduction in 2009 of a global standard on exchange of information and, more recently (2014), a global standard on the automatic exchange of information. Unlike these projects that provide for assistance and exchange of information between countries, in the BEPS Project the OECD focuses on substantive issues that will change the international tax architecture of developed and developing countries when they are implemented. The BEPS Project has fifteen Actions, four of which (5, 6, 13, and 14) are minimum standards that countries should have in their tax systems. These Actions are soft law thus not legally binding; however, there is an expectation that they will be implemented by countries that are participating in the BEPS Inclusive Framework.22 The contents of the BEPS Project have been decided and approved by the BEPS 44 Group that includes the OECD, OECD accession countries, and G20 countries. At the 19 One example is the 2015 proposal by civil society, developing countries, and a few developed countries to upgrade the UN Tax Committee to an intergovernmental body which was rejected by other developed countries at the Addis Ababa Conference arguing for the leading role of the OECD in all tax issues. Mosquera Valderrama, ‘Legitimacy and the Making of International Tax Law’, 375. One recent example is the work carried out by the FACTI panel which, in its recent report (Feb. 2021) has provided fourteen recommendations including recommendation 14B that proposed creating an inclusive intergovernmental body on tax matters under the auspices of the UN, https://www.factipa nel.org/explore-the-report. The recommendation has been acknowledged and welcomed by civil society organizations such as the Tax Justice Network, https://www.taxjustice.net/press/heads-of-state-launch- historic-un-plan-to-stamp-out-global-tax-abuse/. 20 Christians, ‘Hard Law, Soft Law, and International Taxation’, 1–31. 21 R. O. Keohane, S. Macedo, and A. Moravcsik, ‘Democracy- Enhancing Multilateralism’. International Organization 63/1 (2009), 1, 20. 22 See Section 55.2.

1012 Irma Mosquera Valderrama meeting in Kyoto, Japan from 29 June to 1 July 2016, countries (including developing countries) committed to participate as BEPS Associates in the BEPS Inclusive Framework. As of December 2022, 142 jurisdictions (the BEPS 44 Group and ninety- eight additional jurisdictions) committed to the BEPS Inclusive Framework and to the implementation of the BEPS minimum standards. This description shows the leading role of the OECD in international tax matters and, therefore, the time is right to investigate the role of the OECD in the current model of global tax governance in respect of developed and developing countries.

55.2.2 The Role of the OECD vis-à-vis Developing Countries The introduction of the international tax standard of exchange of information, and later on, the BEPS Project by the OECD with the political mandate of the G20, has resulted in a change of paradigm in the role of the OECD in international tax lawmaking. The OECD consists of developed countries and the only way to become a member is by invitation from the OECD. In order to receive this, countries will need to initiate an accession process and comply with the OECD standards and conditions. Therefore, developing countries can only become members if the OECD decides to allow them to do so. This has been the case with Latvia, Lithuania, Colombia, and Costa Rica that formally became full members of the OECD in 2016, 2018, 2020, and 2021 respectively.23 In tax discussions, the OECD has stated that it does adhere to the political mandate of the G20.24 However, as has been argued in the past, the G20 does not represent all countries. The G20 is a political forum of governments with countries from Asia, North America, the Middle East, and Europe including the BRICS countries (Brazil, Russia, India, China, and South Africa) and non-OECD countries such as Argentina, India, South Korea, and Saudi Arabia. In addition, even if a developing country wishes to become a member, that is not possible. Countries cannot become members of the G20 since membership is a political decision made by the G20 countries.25 Based on this, in the past it was concluded elsewhere that the BEPS Project lacks input legitimacy in terms of participation and representation of developing countries in the agenda setting and decision making which is currently carried out by the OECD with the political mandate of the G20.26 Another tax scholar, Brauner, has also addressed the

23

http://www.oecd.org/about/members-and-partners/. the OECD documents on the BEPS Project and interviews/presentations by Pascal Saint Amans at https://www.oecd.org/tax/beps/about/#history. In addition, the OECD also submits OECD Secretary General Tax Reports at G20 meetings (e.g. under the presidencies of Germany, Argentina, Japan, and Saudi Arabia), http://www.oecd.org/g20/topics/international-taxation/publicationsdo cuments/. 25 Mosquera Valderrama, ‘Legitimacy and the Making of International Tax Law’, 353–354. 26 Ibid. 24 See

Global Tax Governance 1013 role of the OECD in the BEPS Project stating that the inadequacies of the project are also due to its political origins.27 The approach to legitimacy and accountability of the OECD multilateral initiatives has also been dealt with Lesage from a political science perspective. He justly argued that ‘both initiatives, which are attempts at universal governance, suffer from legitimacy issues because the G20 and OECD exclude most developing countries. Moreover, the policy outputs are not necessarily adjusted to the needs and interests of developing countries. In recent years, both the G20 and the OECD have attempted to address this issue through institutional fixes, extensive consultations with developing countries and modifications at the level of content’.28 Therefore, Lesage called for more legitimacy and accountability of international organizations. There are three organizations/political forums in which developing countries participate and which are also relevant for international taxation. The first is the United Nations Tax Committee. The second and third are political forums, that is the Group of 77 (G77) at the United Nations which is a coalition of 134 developing countries and the Group of 24. The Group of 24 consists of twenty-eight member countries established in 1971 as a chapter of the Group of 77 in order to help coordinate the positions of developing countries on international monetary and development finance issues. The Group of 24 also ensures that their interests are adequately represented in negotiations on international monetary matters. However, the role of these three organizations/political forums in international tax lawmaking is limited. The reasons for the limited role of the UN in taxation has been explained earlier. The roles of the G77 or G24 in taxation were limited, for instance, in the recent discussion of BEPS Pillar One. There was a proposal from the G24 countries presented in January 2019 to tax highly digitalized business.29 However. that proposal was disregarded by the OECD Secretariat proposal for a ‘Unified Approach’ presented in October 2019 which is currently under discussion by the members of the BEPS Inclusive Framework.30 The G77 has also been active, for instance in the development of the United Nations Handbook on Selected Issues in Protecting the Tax Base of Developing Countries.31 The handbook has been praised by the Tax Justice Network which states that ‘while technical 27 Y.

Brauner, ‘BEPS. An Interim Evaluation’, World Tax Journal 6/1 (2014), 10, 37–38. See also R. S. Avi-Yonah and H. Xu, ‘Evaluating BEPS’, Erasmus Law Review 10/1 (2017) 3. 28 D. Lesage, ‘The Current G20 Taxation Agenda: Compliance, Accountability and Legitimacy’, International Organisations Research Journal 9/4 (2014), 32, 32–41. 29 G24, ‘Proposal for Addressing Tax Challenges Arising from Digitalisation’, available at https://www. g24.org/wp-content/uploads/2019/03/G-24_proposal_for_Taxation_of_Digital_Economy_Jan17_Specia l_Session_2.pdf. 30 The removal of the G24 proposal from the OECD Secretariat’s Unified Approach has been questioned by civil society, see e.g. https://www.globaltaxjustice.org/en/latest/time-developing-countr ies-go-beyond-oecd-led-tax-reform. 31 United Nations, Handbook on Selected Issues in Protecting the Tax Base of Developing Countries (2017), available at https://www.un.org/esa/ffd/wp-content/uploads/2017/08/handbook-tax-base-sec ond-edition.pdf.

1014 Irma Mosquera Valderrama in style and cautious in approach, the UN tax handbook identifies a range of issues in which the OECD’s Base Erosion and Profit Shifting (BEPS) process has failed to deliver for lower-income countries’32 However, no further attempts have been made by the G77 countries to play a role in the development of international tax standards and the current international tax negotiations. The participation of developing countries in the BEPS Project will be discussed in Section 55.4. Since the introduction of the standard of exchange of information and the BEPS Project by the OECD with the political support of the G20, developed and developing countries are adhering to these standards. In respect of developing countries, tax and political science scholars have questioned their validity vis-à-vis developing countries. These concerns will be addressed in Section 55.4. The following section will address the use of hard law vs soft law instruments in international taxation.

55.3 The Use of Soft Law and Hard Law in International Taxation As mentioned earlier, in international taxation the rules have been developed by the OECD and the UN in the form of treaty models, guidelines, and recommendations. However, the introduction of exchange of information and the BEPS Project have created a new international tax framework with a mix between standards/best practices (soft law) and treaties (hard law). In respect of exchange of information, countries have signed tax information- exchange agreements in order to facilitate the exchange of information on request in addition to the already existing article 26 of the bilateral tax treaties (UN or OECD Tax Treaty Model). Furthermore, more than 140 countries have ratified the Multilateral Convention on Mutual Administrative Assistance in Tax Matters as amended by the 2010 Protocol.33 Regarding automatic exchange of financial accounting information, countries have also endorsed the Common Reporting Standard CRS,34 and more than 100 have signed the CRS Multilateral Competent Authority Agreement.35 These

32 A. Cobham, ‘New UN Tax Handbook: Lower-Income Countries vs OECD BEPS’ (2017), available at https://www.taxjustice.net/2017/09/11/new-un-tax-handbook-sets-lower-income-countries-oecd- beps/. 33 OECD, ‘Jurisdictions participating in the convention on mutual administrative assistance in tax matters’ (2020), available at https://www.oecd.org/tax/exchange-of-tax-information/Status_of_convent ion.pdf. 34 OECD, ‘International framework for the CRS’ (2021), available at https://www.oecd.org/tax/ automatic-exchange/international-framework-for-the-crs/. 35 OECD, ‘Signatories of the multilateral competent authority agreement on automatic exchange of financial account information and intended first information exchange date’ (2020), available at https:// www.oecd.org/tax/automatic-exchange/about-automatic-exchange/crs-mcaa-signatories.pdf.

Global Tax Governance 1015 instruments have made it possible to activate automatic exchange of information on request among OECD and non-OECD countries. In addition, a new forum has been created, the Global Transparency Forum with 162 countries participating on an equal footing on the implementation of the standards of exchange of information (on request and automatic). Within the framework of this forum, countries are being monitored and peer-reviewed on the implementation of the standards on exchange of information.36 The BEPS Project provides for four minimum standards, ten best practices, and one multilateral instrument to modify bilateral tax treaties. Countries that are not participants in the BEPS 44 Group, have committed to participate in the implementation of the minimum standards by becoming BEPS Associates in the BEPS Inclusive Framework. In addition, countries have signed (over ninety) and ratified (over thirty) the multilateral instrument that contains two of the four minimum standards (Actions 6 and 14). However, unlike the Multilateral Administrative Convention addressed previously, this multilateral instrument provides for a menu of options (opt-in, opt-out, reservations) that result in mini-bilateral negotiations among countries. In addition, in the BEPS Inclusive Framework, it has been established that countries participate on equal footing on the implementation of the four BEPS minimum standards and that there will be a process of monitoring and peer review of the implementation. The question is to what extent this new mix of instruments and the peer review in the framework of the new institutions (i.e. Global Transparency Forum and BEPS Inclusive Framework) result in countries regarding these standards as binding. In principle, for exchange of information there is a requirement to exchange information and, in the event that a country does not comply, it may be possible for the country to be listed as a non-cooperative (and/or harmful) tax jurisdiction. However, there are other additional factors in order for such countries to be added to the list.37 Regarding BEPS, the situation is more complex since, in principle, these are minimum standards (soft law). However, unless these minimum standards have been opted into (and ratified) in the multilateral instrument (hard law),38 there is no formal obligation to adopt the standards. However, there is an expectation that these standards will be implemented by countries that are participating in the BEPS Inclusive Framework.39 As rightly argued 36

http://www.oecd.org/tax/transparency/who-we-are/. on the content of the OECD and EU assessment, I. J. Mosquera Valderrama, ‘Regulatory Framework for Tax Incentives in Developing Countries After BEPS Action 5’, Intertax 48/4 (2020), 446– 459, https://kluwerlawonline.com/journalarticle/Intertax/48.4/TAXI2020039. 38 Assuming that the country does not provide for a tax treaty override. One example is the USA that does provide for one. This could also be one of the reasons why the USA participates in the BEPS Inclusive Framework but has not signed the multilateral instrument. 39 According to the OECD: 37 See,

The final BEPS Reports are consensus documents, reflecting the agreement of all OECD and G20 governments. As soft law instruments, the expectation is that the conclusions of the reports will be implemented in consistent manners, as has been the case with similar multilateral initiatives

1016 Irma Mosquera Valderrama by Christians, ‘becoming a BEPS Associate entails adopting the initiative’s minimum standards and joining a new coordination architecture—the “Inclusive Framework”— organized for the purpose of measuring and monitoring BEPS compliance across countries’.40 Therefore, the adoption and peer review of these standards has raised questions of the exact consequences of failure to comply. In order to monitor the implementation of these standards, the OECD has created a system of peer review based on its experiences in other areas (corruption, trade). According to the OECD: peer review can be described as the systematic examination and assessment of the performance of a State by other States, with the ultimate goal of helping the reviewed State improve its policy making, adopt best practices, and comply with established standards and principles. The examination is conducted on a non-adversarial basis, and it relies heavily on mutual trust among the States involved in the review, as well as their shared confidence in the process.41

In this system, countries functioning as peers with the assistance of the OECD Secretariat review the implementation of the four BEPS minimum standards by the countries participating in the BEPS Inclusive Framework. There are several actors involved in the peer review; that is, the collective body within which the review is undertaken, the Organization Secretariat (i.e. OECD), the reviewed country, and the examiner countries. In respect of exchange of information, the collective body is the Global Transparency Forum and for BEPS is the BEPS Inclusive Framework. In both cases, the participation of countries is on an equal footing. However, peer reviews will be adopted subject to a ‘consensus minus one’ rule. According to the OECD, this rule aims ‘at ensuring that no one jurisdiction, whether the jurisdiction under review or another jurisdiction with an isolated position, can block consensus on the adoption or publication of a report’.42 In the author’s view, further research should be conducted on the use of a peer review to evaluate compliance (or not) with the international tax standards including the BEPS minimum standards.43 The following section will address the participation and representation of developed and developing countries in the BEPS Inclusive Framework and the peer-review process. co-ordinated by the OECD, such as on exchange of information for tax purposes. A monitoring mechanism will be put in place to assess implementation as well as the impact of OECD recommendations over time with all interested countries participating on an equal footing. (OECD, ‘Myths and Facts about BEPS’, https://www.oecd.org/ctp/myths-and-facts-about-beps.pdf.) 40

A. Christians, ‘BEPS and the New International Tax Order’, BYU Law Review 6/4 (2016), 1603, 1606. (2003). Peer Review: A Tool For Co-Operation and Change. An Analysis of an OECD Working Method by F. Pagani. https://www.oecd.org/dac/peer-reviews/1955285.pdf. 42 OECD, ‘Inclusive Framework on BPES. Progress Report July 2016–June 2017’ (2017), http://www. oecd.org/tax/BEPS/inclusive-framework-on-BEPS-progress-report-july-2016-june-2017.pdf, 22. 43 This research is being carried out in the framework of the GLOBTAXGOV Project, see Section 55.8. 41 OECD

Global Tax Governance 1017

55.4 Participation and Representation The inclusiveness of the BEPS Inclusive Framework vis-à-vis developing countries has been addressed by several scholars and NGOs.44 These scholars have examined: (1) the legitimacy deficits of the BEPS Inclusive Framework;45 (2) the lack of true participation by developing countries in the decision-making process;46 (3) the limited contribution of developing countries to international tax reform;47 and (4) the limited capacity of developing countries to allocate resources (technical, personnel) efficiently to address the implementation of the BEPS minimum standards, the ratification of the multilateral instrument, and the discussions of Pillar One and Pillar Two.48 Some of these concerns have also been addressed by representatives of regional tax organizations (e.g. the African Tax Administration Forum),49 OECD representatives,50 IMF representatives,51 and representatives of tax administrations at several conferences.52

44 For NGOs see, e.g., Action Aid, ‘Levelling up. Ensuring a fairer share of corporate tax for developing countries’ (18 Aug. 2016), http://www.globaltaxjustice.org/sites/default/files/levelling_up_fi nal.pdf;BEPS; BEPS Monitoring Group, ‘Overall evaluation of the G20/OECD Base Erosion and Profit Shifting (BEPS) project’ (18 Aug. 2016), https://bepsmonitoringgroup.files.wordpress.com/2015/10/ general-evaluation.pdf; and Tax Justice Network, ‘Briefing on Base Erosion and Profit Shifting (BEPS) Implications for developing countries’ (2014), https://www.taxjustice.net/wp-content/uploads/2013/04/ TJN-Briefi ng-BEPS-for-Developing-Countries-Feb-2014-v2.pdf. 45 Mosquera Valderrama, ‘Regulatory Framework for Tax Incentives in Developing Countries After BEPS Action 5’; Ozai, ‘Two Accounts of International Tax Justice’. 46 S. Fung, ‘The Questionable Legitimacy of the OECD/ G20 BEPS Project’, Erasmus Law Review 10 (2017), 76; A. Christians and L. Van Apeldoorn, ‘The OECD Inclusive Framework’, Bulletin for International Taxation 72/4–5 (Apr./May 2018); Ozai, ‘Two Accounts of International Tax Justice’. 47 I. Burgers and I. Mosquera, ‘Corporate Taxation and BEPS: A Fair Slice for Developing Countries’, Erasmus Law Review 10 (2017), 29. See also blogpost, S. Picciotto, ‘Developing countries’ contributions to international tax reform’, https://www.ictd.ac/blog/developing-countries-contributions-international- tax-reform-oecd/. 48 R. Collier and N. Riedel, ‘The OECD/ G20 Base Erosion and Profit Shifting Initiative and Developing Countries’, Bulletin for International Taxation 72/12 (2018), 704. 49 See the presentation by Logan Wort (ATAF), ‘AT FACTI Panel Improving Cooperation in Tax Matters (Virtual Consultation), High- Level Panel on International Financial Accountability, Transparency and Integrity for Achieving the 2030 Agenda’ (FACTI Panel) (5 May 2020), Videorecording available at https://www.factipanel.org/events/virtual-consultation-session-improving- cooperation-in-tax-matters. 50 See the presentation Ben Dickinson, Head of the Global Relations and Development Division, OECD, OECD BEPS IF Online Conference, 27–28 January, and Tax Development Day 29 January 2021. See also Tax Sunday Conference organized by the World Bank and IMF, 18 October 2020, ‘International Tax Avoidance—What Has BEPS Accomplished for Developing Countries?’, Panel 2 from 20:22 at https://www.worldbank.org/en/events/2020/10/08/tax-Sunday-conference-2020#2. 51 E. Crivelli, R. A. De Mooij, and M. M. Keen, Base Erosion, Profit Shifting and Developing Countries (International Monetary Fund, 2015). 52 See, e.g., OECD BEPS IF Online Conference, 27– 28 January, and Tax Development Day 29 January 2021. See also Tax Sunday Conference organized by the World Bank and IMF, 18 October 2020, ‘International Tax Avoidance—What Has BEPS Accomplished for Developing Countries?’, Panel 2

1018 Irma Mosquera Valderrama In the past, it has been argued by this author53 that there is a lack of input legitimacy (participation and representation) of developing countries in the agenda setting and decision making of international tax standards. By participating in the BEPS Inclusive Framework, developing countries were required to adopt these standards although without being able to address the content of these standards. This lack of input legitimacy is also shown in the current discussions of BEPS Pillar One and Pillar Two that have resulted in the G24 being excluded from the ‘unified approach’ presented by the OECD Secretariat. In a 2020 study by Christensen et al.,54 they concluded that ‘structural obstacles to participation, including limitations on expert capacity, lack of political support and the absence of effective coalitions, are exacerbated by OECD/IF specific ways of working, e.g. “brutal” or choreographed meeting environments, and high participation costs’. Therefore, attention should be paid to the manner in which countries participate in the decision-making process of international tax standards. In addition, the participation and representation of developing countries in the peer- review process of the BEPS minimum standards should be carefully observed. Even though OECD and non-OECD countries participate in the BEPS Steering Group,55 government tax officials have called for a more inclusive role for developing countries in decision making and in the adoption of the peer-review report. For instance, in the meeting on 27 January 2021 of the BEPS Inclusive Framework, the tax official representative of Jamaica56 addressed the need for a more inclusive BEPS Steering Group.57 Therefore, more research should be performed on the participation of developing countries and the role of the OECD Secretariat, the BEPS Steering Group, and the peer- review process.58

Future directions presentations by representatives of Jamaica and Vietnam, at https://www.worldbank. org/en/events/2020/10/08/tax-Sunday-conference-2020#2. 53

Mosquera Valderrama, ‘Legitimacy and the Making of International Tax Law’. R. C. Christensen, M. Hearson, and T. Randriamanalina, ‘At the Table, Off the Menu? Assessing the Participation of Lower-Income Countries in Global Tax Negotiations’. December 2020, ICTD working paper 115. https://opendocs.ids.ac.uk/opendocs/bitstream/handle/20.500.12413/15853/ICTD_WP115.pdf. 55 According to the OECD FAQ, a steering group leads the Inclusive Framework. As its name suggests, the role of the Steering Group is to steer the Inclusive Framework and provide advice to help inform the Inclusive Framework’s decisions. The Steering Group is comprised of members of the Bureau of the Committee on Fiscal Affairs (i.e. from OECD countries) and members from BEPS Associate countries (i.e. from non- OECD countries), https://www.oecd.org/tax/beps/bitesize-beps/. See the composition of the BEPS Steering Group at https://www.oecd.org/tax/beps/steering-group-of-the-inclus ive-framework-on-beps.pdf. 56 Marlene Nembhard- Parker (Chief Tax Counsel, Legislation, Treaties, and International Tax Matters—Ministry of Finance, Jamaica). 57 See presentation Day 1, 27 Item 3a, ‘Are we on track to finance the SDGs: What role for Taxation’, programme available at https://www.oecd.org/tax/beps/agenda-oecd-g20-inclusive-framework-on- beps-meeting-january-2021.pdf, webcast available from 3:02 at https://www.oecd.org/tax/beps/oecd- g20-inclusive-framework-on-beps-meeting-january-2021.htm/ https://youtu.be/hEJqRHkLQa8. 58 This research is currently being carried out in the framework of the ERC GLOBTAXGOV Project, see Section 55.8. 54

Global Tax Governance 1019

55.5 Outcome International organizations and scholars have addressed the relationship between BEPS and sustainable development, as well as the contribution of the international tax standards to achieving the Sustainable Development Goals (SDGs).59 The challenges that are addressed, for instance, are the need for capacity building and regional cooperation. However, the discussion is fragmented on whether the role should be that of: (1) the United Nations (UN Tax Committee and the recently created Financial Accountability, Transparency and Integrity (FACTI) Panel); (2) the OECD alone or in cooperation with other international organizations (World Bank and IMF); and/ or (3) the countries themselves in a regional setting (Africa, Latin America, Asia). In addition, international organizations within the framework of the Platform for Collaboration on Tax have addressed the need to tackle other issues that are not BEPS- related (e.g. wasteful tax incentives and indirect transfer of assets). This has resulted in additional reports/toolkits that countries can also take into account when designing their domestic and international tax rules. The fast pace of these international tax developments makes it difficult for countries (including developing countries) and also scholars to follow all of them. Developing countries are still attempting to discern how to implement the BEPS minimum standards in their country and/or how to evaluate the suitability of the BEPS minimum standard (e.g. Botswana opting out of BEPS Action 13)60 and, in some cases, requesting technical assistance to do so (e.g. Cameroon for implementation of BEPS Action 13).61 Some of these challenges have also been addressed, for instance, by OECD representatives at the meetings of international organizations (World Bank/IMF).62 In the author’s view, it is clear that by introducing BEPS minimum standards, signing and ratifying the BEPS multilateral instrument, and participating in the current discussions on BEPS Pillar One and Pillar Two, developing countries will need a significant number of personnel as well as substantial technical knowledge and political will to ensure that participation in these initiatives is effective for achieving their own SDGs.

59 e.g.

PCT, IMF/WB meeting addressing tax and SDGs, and the 27 January 2021 OECD meeting addressing SDGS. 60 OECD/ G20 Base Erosion and Profit Shifting Project, ‘Country- by- Country Reporting— Compilation of Peer Review Reports (Phase 1): Inclusive Framework on BEPS: Action 13’ (2018), http:// dx.doi.org/10.1787/9789264300057-en, 12 and 13. 61 OECD/ G20 Base Erosion and Profit Shifting Project, ‘Country- by- Country Reporting— Compilation of Peer Review Reports (Phase 1): Inclusive Framework on BEPS: Action 13’ (2018), http:// dx.doi.org/10.1787/9789264300057-enat124. 62 Presentation by Ben Dickinson, Head of the Global Relations and Development Division, OECD, OECD BEPS IF Online Conference, 27–28 January, and Tax Development Day, 29 January 2021. See also Tax Sunday Conference organized by the World Bank and IMF, 18 October 2020, ‘International Tax Avoidance—What Has BEPS Accomplished for Developing Countries?’, Panel 2 from 16:53, https://www. worldbank.org/en/events/2020/10/08/tax-Sunday-conference-2020#2.

1020 Irma Mosquera Valderrama The challenges that developing countries face in the implementation of BEPS minimum standards and in achieving the SDGs have been addressed elsewhere.63 Some of these challenges are the use of transfer pricing rules, the use of medium-term revenue strategies, the need to introduce a regulatory (administrative) framework for tax incentives, and the need to introduce tax expenditure reporting, among others.64 In order to address these challenges and possible solutions, scholars from the fields of tax, political science, economics, and international relations need to cooperate so that comprehensive solutions can be presented.65 In addition, the solutions proposed by scholars and policymakers (ministries of finance, tax administrations, and/or international organizations) will require political will and, therefore, a dialogue should be established between tax officials, politicians, civil society, businesses, and business associations. That dialogue should occur at domestic, regional, and international levels. At this stage that dialogue takes place at the international level, however at regional and domestic levels it is limited. Tax officials at the ministry of finance and/or tax administrations may decide to introduce tax reforms following the international standards agreed at the OECD and/or BEPS Inclusive Framework level. These will thereafter need to be explained and discussed with politicians, businesses, and judges that may not be aware of international tax developments and the need to change the rules to achieve the SDGs. One interesting example in developing countries is the use of tax incentives for political reasons and the lack of transparency in the granting and regulation of tax incentives. Mosquera Valderrama and Balharová66 have recently argued that ‘there should be no room for administrative discretion on the granting of tax incentives, one person/body should be in charge of granting tax incentives and the incentive should be transparent (publicly available on the website of the tax administration or administrative agency)’. This has not been the case in developing countries and in the instance of the Philippines, the Covid-19 crisis resulted in flexibility for the president to grant fiscal and non-fiscal

63

See W. Lips and I. J. Mosquera Valderrama, ‘Global Sustainable Tax Governance in the OECD-G20 Transparency and BEPS Initiatives’, in C. Brokelind and S. Van Thiel, eds, Tax Sustainability in an EU and International Context (IBFD, 2020), 235–258 and I. J. Mosquera Valderrama, D. Lesage, and W. Lips, ‘Tax and Development: The Link between International Taxation, The Base Erosion Profit Shifting Project and the 2030 Sustainable Development Agenda, Working Paper Series no. W-2018/3, Bruges, UNU Institute on Comparative Regional Integration Studies (2018). 64 Mosquera, Lesage, Lips, ibid. 65 One example is I. J. Mosquera Valderrama, D. Lesage, and W. Lips, eds, Taxation, International Cooperation and the 2030 Sustainable Development Agenda, United Nations University Series on Regionalism vol. 19 (Springer, 2021). The book was edited in cooperation with two international political economist scholars and with contributions from economists, political scientists, tax officials, representatives of regional tax organizations, and tax scholars. 66 I. J. Mosquera Valderrama and M. Balharová, ‘Tax Incentives in Developing Countries: A Case Study: Singapore and Philippines in Taxation, International Cooperation and the 2030 SDG Agenda’, in Mosquera Valderrama, Lesage, and Lips), Taxation, International Cooperation and the 2030 Sustainable Development Agenda.

Global Tax Governance 1021 incentives which, according to the Department of Finance, ‘will be critical as the country competes internationally for high-value investments’.67 In its view, that Covid- 19 measure granting discretionary power to the president may affect the path taken by the Philippines to increase transparency and reduce the discretionary power to grant tax incentives. Therefore, even if it is not a BEPS issue, the use of tax incentives shows the different views/interests of actors in tax law decision making.68 The following section will address the role of actors in global tax governance.

55.6 Actors In global tax governance, there are different actors that can be mentioned, for instance at the international and regional levels: • international organizations (OECD, UN, G24, G77, WB, IMF), committees/panels (UN Tax Committee, FACTI Tax Panel), UN organizations (UNDESA,UNDP) and UN regional hubs (e.g. UN Economic and Social Commission for Asia and the Pacific UNESCAP, and UN Economic Commission for Africa UNECA); • advocacy groups (ICRIT, Tax Justice Network, BEPS Monitoring Group, Oxfam, Action Aid) and think tanks (CEP); • international and regional (tax administration) networks: Network of Tax Organizations NTO, Intra- European Organization of Tax Administrations IOATA, African Tax Administration Forum ATAF, Inter-American Centre of Tax Administrations CIAT, Study Group on Asian Tax Administration and Research SGATAR, International Tax Compact ITC.69 In addition, at a domestic level, it is also important to take into account the role of actors that will give shape to the content of the international tax standards and, in some cases, will influence or change the commitment made by countries’ representatives at

67 Website of theDepartment of Finance, https://taxreform.dof.gov.ph/tax-reform-packages/p2- corporate-recovery-and-tax-incentives-for-enterprises-act/ (accessed 8 August 2022). See Report to the Joint Congressional Oversight Committee (8 June 2020), https://www.officialgazette.gov.ph/downloads/ 2020/06jun/20200608-Report-to-the-Joint-Congressional-Oversight-Committee.pdf, 15 (accessed 8 August 2022). 68 These different views have been addressed in several (online) seminars organized in the framework of the GLOBTAXGOV Project. See summary discussions at blog GLOBTAXGOV, https://globtaxgov. weblog.leidenuniv.nl/reports-conference-papers/. 69 On the role of the EU, see in general C. Panayi, ‘Europeanization of Good Tax Governance’, Yearbook of European Law 36/1 (2017), 442. In respect of third (non-EU) countries including developing countries, see I. J. Mosquera Valderrama, ‘The EU Standard of Good Governance in Tax Matters for Third (Non-EU) Countries’, Intertax 47/5 (2019), 454–467.

1022 Irma Mosquera Valderrama the international level.70 These actors are tax administrations, ministries of finance, lawmakers (legislative), judges, advocacy networks, scholars, businesses, and business associations among others. The role of these actors in the implementation of BEPS has been addressed by Sadiq et al. in a comparative study of eighteen developed and developing countries.71 However, further empirical research needs to be conducted to ascertain the role (if any) of the actors in the implementation of BEPS and whether there are differences in its implementation and, if so, are those differences explained in the light of the differences in tax systems and tax culture. I have argued in the past that ‘the study of differences in culture provide the local tuning that makes that for a transplanted concept the rules are different in the recipient country than the ones in the donor’.72 In our view, the proposed approach to tax culture takes into account that ‘the differences in attitudes and beliefs in a country’s legal and tax system require the study of the way that the development of law takes place is “law in action” and not only the introduction of a concept in the formal law of the country “law in the books” ’.73 Therefore, when studying the transplant of BEPS minimum standards into countries belonging to the BEPS Inclusive Framework, I argue the need to address ‘the use of the concept of legal culture for identifying the role of the different parties (stakeholders) in the transplantation process and in the development of legal (tax) rules. These parties can be, for instance, the courts with tax competence, tax lawmakers, taxpayers, tax administrations, business associations, tax advisors, scholars, and civil society (NGOs). This study of legal culture will need not only desk research, but also empirical research (interviews and/or surveys) to the parties (stakeholders)’.74 This empirical research is currently conducted in the framework of the ERC GLOBTAXGOV Project.75

55.7 Concluding Remarks This chapter has highlighted some of the main elements that can be addressed when studying global tax governance and the current discussions from a tax scholarship 70

e.g. despite the Ministry of Finance agreeing to introduce the rule of commissionaire arrangements in the Multilateral Instrument, the Dutch Parliament when discussing the MLI decided not to agree until there was no specific provision for dealing with dispute resolution for commissionaire-arrangement structures. 71 K. Sadiq, A. Sawyer, and B. McCredie. Tax Design and Administration in a Post-BEPS Era: A Study of Key Reform Measures in 18 Jurisdictions. (Fiscal Publications, 2019). 72 Mosquera Valderrama, ‘Regulatory Framework for Tax Incentives in Developing Countries After BEPS Action 5’, 727. 73 Ibid., 728. 74 Ibid., 728. 75 In this project we carry out research in eight countries from different geographical regions, with different tax systems and different tax cultures (Mexico, Colombia, Senegal, Nigeria, India, the Netherlands, Spain, and Australia), see Section 55.8.

Global Tax Governance 1023 perspective. The elements that are addressed are the distinction between hard law vs soft law, the participation and representation of the OECD, G20, and non-OECD/non- G20 participants including developing countries in setting the international tax agenda, the results (outcome) of the current international discussions, and the role of the actors (international and domestic) in the implementation of international tax standards. In the author’s view, the study of global tax governance from different perspectives (tax, political, economy) facilitates the cross-fertilization of research findings between different disciplines. This approach affords the opportunity to provide solutions that are more comprehensive for achieving the goals highlighted by international organizations and countries (reducing BEPS practices and achieving the 2030 Sustainable Development Agenda). The topic of global tax governance is a relevant topic of discussion, however it does require certain caution when referenced by scholars, organizations, and policymakers in general. Not everything can be explained in terms of global tax governance, and the differences in approach by political scientists regarding global tax governance vs tax cooperation should also be taken into account. That notwithstanding, the current interests by tax scholars in research projects and literature on global tax governance allows them to facilitate the exchange of ideas in which not only tax technical issues are discussed, but attention is also given to policymaking and the role of actors in that process. This will allow an understanding of the differences in participation/representation in decision making, the differences in the commitment to international tax standards, and the different content of these standards on the transplantation of those standards into the recipient country’s tax system and tax culture.

55.8 Acknowledgements The writing and research conducted for this chapter is the result of the European Research Council research in the framework of the GLOBTAXGOV Project (2018– 2023) funded by the ERC under the European Union’s Seven Framework Programme (FP/2007–2013) (ERC Grant agreement no. 758671).

Chapter 56

The Fu tu re of L a b ou r Tax ation and t h e ‘ Ri se of the Rob ots ’ Georg Kofler

56.1 Introduction The discussion about international tax policy in the era of digitalization largely revolves around the taxation of profits of multinational enterprises, and this is indeed exemplified by the OECD’s project on ‘Base Erosion and Profit Shifting’ and the subsequent work on a new consensus of international profit-allocation and minimum taxation. That focus is understandable, but needs to be put in a broader perspective: in the EU in 2019, for example, taxes and net social security contributions amounted to about 41% of gross domestic product (GDP), and of that share less than 7% were corporate income taxes and more than 58% related to individual income taxes and social security contributions.1 The burden on labour income was and continues to be a major source of government revenue and the financing of social security systems. Correspondingly, the OECD average of the so-called ‘tax wedge’ for a single worker without children (i.e. a measure of the difference between labour costs to the employer and the corresponding net take-home pay of the employee) was 36% in 2019,2 implying a significant burden on labour and employment, while traditionally many countries have a lesser tax burden on capital. The relative tranquility in the legal framework for taxing capital and labour might, however, come into flux due to the increasing impact of digitalization, robotics, datafication, 1

See, e.g., the Eurostat Press Release, ‘Tax-to-GDP ratio at 41.1% in EU’, 160/2020 (29 Oct. 2020). See OECD, Taxing Wages 2020 (Paris: OECD Publishing, 2020). For 2020, the OECD average of the tax wedge decreased by 0.39 percentage points to 34.6%, which resulted predominantly from changes in tax policy settings and falling average wages due to the Covid-19 pandemic (see OECD, Taxing Wages 2021 (Paris: OECD Publishing, 2021), 14). 2

1026 Georg Kofler automation, and artificial intelligence (AI) on the economy and has resulted in intense debates not only about the future of work, but also the future of labour taxation and alternative sources of revenues in today’s globalized economy.3 Indeed, structural changes where wage income and tax revenues decrease and states’ social nets and public expenditures are strained might lead to significant budgetary challenges (‘double negative effect’4), increase income inequality, and impede the ability of states to implement redistributive policies. The ongoing ‘fourth industrial revolution’ or ‘automation revolution’ has created uncertainty and anxiety about the risks and chances of a possible transformation of the labour market, as there exists the perception that technological change is faster paced and broader based than in the past, making more jobs automatable than previously thought, with AI potentially being the harbinger of a jobless future, a picture elaborately painted in Martin Ford’s Rise of the Robots: Technology and the Threat of a Jobless Future.5 Conversely, it is argued, automation and AI also have the potential to complement and augment human capabilities, leading to higher productivity, greater demand for human labour, and improved job quality. Specifically for AI, however, little empirical evidence exists, and certainly AI as a broad technological platform is capable of both replacing work or complementing workers, ‘and it is thus partly a matter of societal and business choice how much job displacement AI will create’.6 It therefore comes as no surprise that, for example, also the OECD7 as well as the EU8 have put a focus on the impact of technological advancements, including AI, on the future of the labour market as one mega-trend, in addition to globalization, ageing, global population growth, migration, and ‘brain drain’ etc.9 This discussion is exacerbated by the foreseeable next generation of advances in AI, such as improvements in machine learning and new manufacturing technologies: while traditional automated machinery and robots typically compete with routine, low-skill work, specifically artificial intelligence has the potential to expand the range of tasks

3 Of course, there are multiple interconnections between technological and tax transformations apart from the impact on the taxation of labour. For an instructive overview see, e.g., V. P. Vishnevsky and V. D. Chekina, ‘Robot vs. Tax Inspector or How the Fourth Industrial Revolution will Change the Tax System: A Review of Problems and Solutions’, Journal of Tax Reform 4 (2018), 6–26. 4 B. McCredie, K. Sadiq, and L. Chapple, ‘Navigating the Fourth Industrial Revolution: Taxing Automation for Fiscal Sustainability’, Australian Journal of Management 44 (2019), 648, 648. 5 M. Ford, Rise of the Robots: Technology and the Threat of a Jobless Future (London: Oneworld Publications, 2015). 6 See D. Acemoglu et al., ‘AI and Jobs: Evidence from Online Vacancies’, NBER Working Paper 28257 (2021), 1. 7 See, e.g., OECD, OECD Employment Outlook 2019: The Future of Work (Paris: OECD Publishing, 2019) and M. Lane and A. Saint-Martin, ‘The impact of Artificial Intelligence on the labour market: What do we know so far?’, OECD Social, Employment and Migration Working Paper No. 256 (2021). 8 D. Spencer et al., ‘Digital automation and the future of work’, European Parliamentary Research Service, Scientific Foresight Unit (STOA), PE 656.311 (Jan. 2021). 9 For a detailed overview, see OECD, OECD Employment Outlook 2019: The Future of Work (Paris: OECD Publishing, 2019).

The Future of Labour Taxation 1027 that can be automated far beyond lower skilled jobs in labour-intensive sectors. As the OECD notes (after asking if we are ‘headed towards a jobless future’), ‘rapid progress in the ability of machines and artificial intelligence (AI) to automate an ever-widening number of job tasks performed by humans has the potential to accelerate the substitution of labour with capital and to induce significant productivity gains, requiring less labour input into the production process’.10 Some have even argued that this automation revolution and AI will make much of human labour, whether skilled or unskilled, routine or non-routine, superfluous (and might eventually create a so-called ‘singularity’ where machine intelligence exceeds human intelligence)11 or might at least result in rapid changes in the economy and painful transitions,12 which are hard to imagine in an environment of developed welfare states. It would, in the words of John Maynard Keynes, create technological unemployment ‘due to our discovery of means of economising the use of labour outrunning the pace at which we can find new uses for labour’.13 Chances and risks have also been expressed politically by the European Parliament, which dealt with civil law issues concerning ‘ever more sophisticated robots, bots, androids and other manifestations of artificial intelligence’ and highlighted, inter alia, that ‘despite the undeniable advantages afforded by robotics, its implementation may entail a transformation of the labour market and a need to reflect on the future of education, employment, and social policies accordingly’.14 Naturally, approaches such as labour market institutions, adult education, and social protection policies dominate the discussion about the future of work in the light of the rapid technological transformation.15 Indeed, already in 2016 the USA had issued a report on ‘Artificial Intelligence, Automation, and the Economy’, identifying three broad strategies; that is, to invest in and develop AI for its many benefits, to educate and train Americans for jobs of the future, and to aid workers in the transition and empower workers to ensure broadly shared growth.16 There is, however, also a deepening 10 OECD, OECD Employment Outlook 2019: The Future of Work (Paris: OECD Publishing, 2019), 44.

11 See, e.g., R. Kurzweil, The Singularity Is Near. When Humans Transcend Biology (New York: Viking, 2005). For an instructive analysis of the potential impact of such ‘singularity’ on taxation, see A. Korinek and J. E. Stieglitz, ‘Artificial Intelligence and Its Implications for Income Distribution and Unemployment’, NBER Working Paper 24174 (2017). 12 See for that discussion, e.g., E. Brynjolfsson and A. McAfee, The Second Machine Age: Work, Progress, and Prosperity in a Time of Brilliant Technologies (New York: W. W. Norton, 2014); M. Ford, Rise of the Robots: Technology and the Threat of a Jobless Future (London: Oneworld Publications, 2015); D. Susskind, A World Without Work: Technology, Automation and how We Should Respond (London: Penguin, 2020). For extensive context with regard to the labour market transitions during the previous three industrial revolutions, see C. B. Frey, The Technology Trap: Capital, Labor, and Power in the Age of Automation (Princeton, NJ: Princeton University Press, 2019). 13 J. M. Keynes, ‘Economic Possibilities for our Grandchildren’ (1930), in J. M. Keynes, ed., Essays in Persuasion (London: Palgrave Macmillan, 2010), 321, 325). 14 See Points E– K and specifically Point G of the European Parliament Resolution with recommendations to the Commission on Civil Law Rules on Robotics, P8_TA(2017)0051 (16 Feb. 2017). 15 See OECD, OECD Employment Outlook 2019: The Future of Work (Paris: OECD Publishing, 2019), 91–325. 16 Executive Office of the President, ‘Artificial Intelligence, Automation, and the Economy’ (20 Dec. 2016).

1028 Georg Kofler discussion about the sustainability of tax and social systems, including situations of larger scale structural worker replacement and technological unemployment. Much attention was therefore given to a motion for a resolution by the EU Parliament which demanded that, ‘for the preservation of social cohesion and prosperity, the likelihood of levying tax on the work performed by a robot or a fee for using and maintaining a robot should be examined in the context of funding the support and retraining of unemployed workers whose jobs have been reduced or eliminated’,17 an issue that was deliberately avoided in the final text approved by the European Parliament.18 However, the discussion of ‘taxing robots’ (or their use) or for changing existing tax policies in anticipation of substantial job losses due to automation substitution has since been ongoing and intensifying. This chapter aims at providing some context and an overview of the major reform proposals.

56.2 Automation and the Future of Work In 1920, Karel Čapek coined the term ‘robot’ for artificial, intelligent creatures in his play ‘R.U.R.—Rossum’s Universal Robots’.19 Now, over a century later, robotics, automation, and artificial intelligence are part of our lives, and it is generally expected that within a matter of years these rapidly developing technologies will have a significant impact on all sectors of the economy. Already now the observable decline in the labour share of national income across the OECD (and the corresponding decoupling of wages from productivity gains) has also been attributed in part to technological change.20 There are, of course, multiple dimensions of the economic effects of technological progress. For one, while workers are displaced by new technologies and a significant share of jobs is at risk of automation, there are also various channels through which technology may boost employment. Also, technological advancements may complement rather than displace labour. More generally, technological progress raises overall output, growing the size of the ‘economic pie’, but also generates a redistribution of the existing economic pie as it changes the market prices (e.g. wages) at which people transact in the economy.

17

Report with recommendations to the Commission on Civil Law Rules on Robotics, A8-0005/2017 (27 Jan. 2017). 18 European Parliament Resolution with recommendations to the Commission on Civil Law Rules on Robotics (2015/2103(INL), P8_TA(2017)0051 (16 Feb. 2017). 19 K. Čapek, Rossumovi Univerzální Roboti (Prague: Aventinum, 1920). 20 See also para. 45 in Lane and Saint-Martin, ‘The impact of Artificial Intelligence on the labour market: What do we know so far?’. Indeed, a number of OECD countries have seen a significant decline of the ‘labour share’, i.e. labour’s share of national income, over the past two decades (see, e.g., M. Pak and C. Schwellnus, ‘Labour Share Developments Over the Past Two Decades: The Role of Public Policies’, OECD Economics Department Working Papers No. 1541 (2019]).

The Future of Labour Taxation 1029 These price changes are, eventually, a zero-sum-game, benefiting some at the expense of others, creating (relative) winners and losers of innovation.21 It is primarily this ‘technological redistribution’ of income and its potential speed that also warrants attention from a policy (and also tax) perspective.22 While much of the observed and potential ‘technological redistribution’ resulting from traditional industrial robotics and automation is identified to the disadvantage of routine, low-skilled work, AI has affected higher paid, non-routine, cognitive work. Theoretically, displacements in these areas could mitigate the trend of increasing income inequality (but increase competition for lower skill work),23 but AI has also been found to be complementary to higher wage, higher education work, suggesting that it could increase income inequality.24 Many forces are in play, and multiple explanations exist why the visibility of productivity growth has been lagging despite substantial progress in AI and other technologies (‘productivity paradox’) and why mass unemployment has not materialized.25 Indeed, from a theoretical perspective, ‘the impact of AI on employment and wages is ambiguous, and it may depend strongly on the type of AI being developed and deployed, how it is developed and deployed, and on market conditions and policy’.26 Several strands of theoretical and empirical economic research feed into that discussion. An increasing body of research evaluates the impact of automation on jobs and wages and suggests that that automation will cause job displacement, and that recent technological progress, particularly in AI, is rapidly extending the range of tasks machines can perform. Traditionally, robotic automation technologies have been shown to disrupt labour markets and lead to reallocation dynamics, with negative changes well documented especially in the employment share of routine manual jobs. Indeed, the existing robotic automation technologies already have the ability to replace workers for certain tasks and hence lower labour demand (‘destruction’ or ‘displacement’ effect), but conversely those technologies raise productivity and create tasks that cannot (yet) be done by machines and hence increase labour demand (‘constructive’ or ‘reinstatement’ effect).27 Current economic studies on the socio-economic outcomes of automation clearly document the negative effects specifically of industrial robots on employment and wages and the positive impacts on productivity, but leave an unclear picture whether robot adoption is overall labour-replacing (and wage-reducing) or if displacement effects are cancelled out by reallocation effects. While some have found

21 See,

e.g., A. Korinek and J. E. Stieglitz, ‘Artificial Intelligence and Its Implications for Income Distribution and Unemployment’, NBER Working Paper 24174 (2017). 22 OECD, OECD Employment Outlook 2017 (Paris: OECD Publishing, 2017), 81ff. 23 See para. 17 in Lane and Saint- Martin, ‘The impact of Artificial Intelligence on the labour market: What do we know so far?’. 24 See ibid., para. 56. 25 See, with further references, ibid., paras 37ff. 26 See ibid., paras 43ff. 27 For a recent survey of the economic literature see, e.g., G. de Vries et al., ‘The Rise of Robots and the Fall of Routine Jobs’, ADB Economics Working Paper No. 619 (2020).

1030 Georg Kofler that industrial robots reduced the employment of low-skilled workers but not total employment,28 others have concluded that robots are overall labour-replacing (and wage- reducing).29 Limited empirical (i.e. backward-looking) work also exists on the impact of AI on employment and wages in the last decade, concluding that there was no overall decline in employment due to recent advances in AI (but possibly even a positive impact of AI on wage growth).30 It is, of course, ‘unclear to what extent future developments in AI will produce similar results’.31 A related field of economic research estimates which types of jobs are susceptible to automation on the basis of various technological projections regarding advancements of technology and the range of tasks machines can and will be able to perform (but do not take into account how other sectors and occupations will respond to these changes and hence do not address the number of jobs that technology generates). One set of studies focuses on the risk of automation for a subset of occupational titles, based on the tasks these occupations involve. Most influentially, in 2013 an Oxford study found that around 47% of total employment in the USA is in the ‘high risk category regarding job 28 G. Graetz and G. Michaels (‘Robots at Work’, Review of Economics and Statistics 100 (2018), 753– 768), e.g., focus on the variation in robot usage across industries in different countries and conclude that increased robot use contributed to annual labour productivity growth, while at the same time raising total factor productivity and lowering output prices; according to their findings, robots did not significantly reduce total employment, but reduced the employment of low-skilled workers. For Germany and its strong labour institutions, W. Dauth et al. (‘The Adjustment of Labor Markets to Robots’, Journal of the European Economic Association 19 (2021), 3104–3153) show that the adoption of industrial robots had no effect on total employment in local labour markets specializing in industries with high robot usage, as robot adoption led to job losses in manufacturing that were offset by gains in the business service sector; as for individual workers, robot adoption has not increased the risk of displacement for incumbent manufacturing workers, but rather that loss of manufacturing jobs is solely driven by fewer new jobs for young labour market entrants. Moreover, they found that in regions with higher exposure to automation, labour productivity increases while the labour share in total income declines. G. J. De Vries et al. (‘The Rise of Robots and the Fall of Routine Jobs’, ADB Economics Working Papers No. 619 (2020)) studied the relation between industrial robots and occupational shifts by task content and found that increased use of robots is associated with positive changes in the employment share of non-routine analytic jobs and negative changes in the share of routine manual jobs in many (especially high-income) economies. 29 D. Acemoglu and P. Restrepo (‘Robots and Jobs: Evidence from US Labor Markets’, Journal of Political Economy 128 (2020), 2188–2244) examine geographic variation in robot adoption across the USA and find that robots are overall labour-replacing. They conclude ‘that each additional robot per thousand workers reduces the local employment-to-population ratio by 0.39 percentage points and wages by about 0.77%. Because adopting robots creates benefits for other commuting zones via trade linkages, the implied aggregate effects are smaller—one additional robot per thousand workers reduces the aggregate employment-to-population ratio by 0.2 per-centage points and aggregate wages by 0.42%.’ 30 See, e.g., E. W. Felten, M. Raj, and R. Seamans, ‘The Occupational Impact of Artificial Intelligence: Labor, Skills, and Polarization’ (New York: NYU Stern School of Business, 2019); F. M. Fossen and A. Sorgner, ‘New Digital Technologies and Heterogeneous Employment and Wage Dynamics in the United States: Evidence from Individual-Level Data’, IZA Discussion Paper Series No. 12242 (2019); D. Acemoglu et al., ‘AI and Jobs: Evidence from Online Vacancies’, NBER Working Paper 28257 (2021). 31 See para. 57 in Lane and Saint- Martin, ‘The impact of Artificial Intelligence on the labour market: What do we know so far?’.

The Future of Labour Taxation 1031 automatability’ (i.e. jobs above a predicted automatability of 70% were defined as ‘high risk’) over a number of decades,32 and others have arrived at similarly high estimates.33 This occupation-focused approach has been criticized, arguing that occupations as a whole are unlikely to be automated as not all workers in the same occupation perform the same tasks and hence face the same risk of their jobs being automated. Focusing not on occupations but rather taking into account the variation in tasks within occupational groups (i.e. the task content of individual jobs), a second set of studies arrives at much lower estimates of jobs threatened by automation,34 again with their methodology and the relevance of the variation of tasks criticized.35 Based on this methodology, the OECD estimates that the share of jobs at high risk of automation (i.e. those with a probability of being automated of at least 70%) is around 14%, on average, across the OECD, with significant variations across countries (from 6% in Norway to 34% in the Slovak Republic) and across occupations (ranging from 1.1% for chief executives, senior officials, and legislators to 50.1% for food-preparation assistants).36 Moreover, ‘another 32% of jobs have a risk of between 50 and 70% pointing to the possibility of significant change in the way these jobs are carried out as a result of automation—i.e. a significant share of tasks, but not all, could be automated, changing the skill requirements for these jobs’. Overall, the analysis highlights that the risk of automation is higher among low-skilled workers rather than middle-skilled and high-skilled jobs, and that the occupations with the highest estimated automatability typically only require a basic to low level of education. While these studies consistently conclude that low-skilled occupations are highly exposed to automation technologies in a broader sense (and therefore at highest risk of automation), other studies suggest that primarily high-skilled occupations are exposed

32 C.

B. Frey and M. A. Osborne, ‘The Future of Employment: How Susceptible are Jobs to Computerization?’, Oxford Martin Working Paper (2013). This includes not only workers in transportation and logistics occupations, office and administrative support workers, and labour in production occupations, but also a substantial share of employment in service occupations. 33 See, e.g., M. Chui, J. Manyika, and M. Miremadi, ‘Where achines Could Replace Humans— And Where They Can’t (Yet)’, McKinsey Quarterly (July 2016) (45% in the USA); World Bank, World Development Report 2016: Digital Dividends (2016) (57% in the OECD). 34 See, e.g., M. Arntz, T. Gregory, and U. Zierahn, ‘The Risk of Automation for Jobs in OECD Countries’, OECD Social, Employment and Migration Working Papers No. 189 (2016), finding that only 9% of existing jobs are at risk of complete automation in the USA. 35 For methodological criticism see, e.g., S. W. Elliott, Computers and the Future of Skill Demand (Paris: OECD Publishing, 2017), 22 and M. A. Osborne and C. B. Frey. ‘Automation and the future of work –understanding the numbers’, Oxford Martin School Blog (13 Apr. 2018) (‘To take the example of the truck driver, our approach treats all truck drivers as equal: when autonomous vehicles arrive, all of them will become exposed to automation. The OECD study argues that their estimates are lower than ours because a large share of drivers will not find themselves exposed but without providing any data to show that this is the case. We would welcome the OECD’s publication of the distribution of workers that are exposed to automation by occupation’). 36 L. Nedelkoska and G. Quintin, ‘Automation, skills use and training’, OECD Social, Employment and Migration Working Papers No. 202 (2018) and OECD, OECD Employment Outlook 2019: The Future of Work (Paris: OECD Publishing, 2019), 44.

1032 Georg Kofler to AI;37 that is, ‘[w]‌hereas low-skill occupations are most exposed to robots, and middle- skill occupations are most exposed to software, it is high-skill occupations that are most exposed to artificial intelligence. Moreover, artificial intelligence is much more likely to affect highly-educated and older workers than these previous technologies.’38 It feeds into that line of research that a third approach focuses on human skills that are broadly relevant to work and relates those to the information technology (IT) capabilities described in the research literature to identify tasks and occupations where IT could substantially replace workers over the next few decades; this focus on worker skills rather than job tasks or activities suggests that 82% of current US employment is potentially automatable and vulnerable to displacement by IT in the near future.39 What, then, is the likely path for the future with regard to ‘traditional’ automation and evolving technologies, such as AI? In its 2019 report ‘The Future of Work’,40 the OECD notes that workers are indeed displaced by new technologies but takes a rather positive view on a range of countervailing forces41 through which technology creates and transforms jobs. It suggests that ‘a substantial contraction of employment is unlikely as a result of digitalisation and globalisation’ and that ‘despite the displacement effects of technological progress, employment in OECD countries has historically increased on average’.42 However, while these mechanisms may eventually lead to an overall increase in employment, ‘the importance of public policy to cushion the displacement effects of technology should not be downplayed, particularly because such risks are not distributed evenly across countries, regions, and socio-demographic groups’.43 Indeed, 37 E.

Brynjolfsson, T. Mitchell, and D. Rock, ‘What Can Machines Learn and What Does It Mean for Occupations and the Economy?’, AEA Papers and Proceedings 108 (2018), 43–47; E. W. Felten, M. Raj, and R. Seamans, The Occupational Impact of Artificial Intelligence: Labor, Skills, and Polarization (New York: NYU Stern School of Business, 2019); M. Webb, ‘The Impact of Artificial Intelligence on the Labor Market’, Stanford University, Department of Economics (2020). 38 Webb, ibid., 46. 39 S. W. Elliott, ‘Projecting the Impact of Information Technology on Work and Skills in the 2030s’, in J. Buchanan et al., eds, The Oxford Handbook of Skills and Training (Oxford: Oxford University Press, 2017), 557–575. 40 OECD, OECD Employment Outlook 2019: The Future of Work Paris: (OECD Publishing, 2019). See also, e.g., OECD, The Next Production Revolution: Implications for Governments and Business (Paris: OECD Publishing, 2017) and OECD, OECD Digital Economy Outlook 2017 (Paris: OECD Publishing, 2017). 41 Those countervailing factors include: (1) that technological progress can generate more jobs than it destroys within a given industry and that entirely new jobs may be created as a result of innovation, either to complement machine capabilities within existing occupational categories or in entirely new fields; (2) that by increasing productivity and reducing prices, certain technologies have a positive impact on employment in industries other than the ones when they are deployed; and (3) that automation can lower input costs for downstream industries, leading to output and employment growth in those industries. However, fowHor concerns that the digital economy is not creating the large number of jobs created by leading industries in the past, see OECD, ‘The Next Production Revolution: Implications for Governments and Business’ (2017), 31. Recent OECD work found that 40% of jobs created between 2005 and 2016 were in digitally intensive industries, see OECD, Going Digital: Shaping Policies, Improving Lives (Paris: OECD Publishing, 2019. 42 OECD, OECD Employment Outlook 2019: The Future of Work (OECD Publishing, Paris: 2019), 46. 43 Ibid., 47.

The Future of Labour Taxation 1033 as experts are not in agreement on the speed at which technology may be replacing work in the coming decades, ‘responsible policy making should aim to enhance the resilience of the labour market, effectively preparing for a range of potential futures’.44 Significant ‘risks of decreasing job quality and increasing disparities among workers loom large and should be the key focus of policy makers’, and ‘while the risk of an overall drop in employment is limited at the aggregate level, certain industries and regions may see net declines in the number of jobs available and policies are required to facilitate labour mobility and respond to regional disparities’.45 The OECD also notes that emerging economies generally face a higher risk of automation than more advanced ones but that despite technological feasibility automation may not yet be economically attractive in many emerging economies.46 However, the impact of automation on employment in emerging economies is multi-fold.47 For example, as the cost of industrial robots continues to decline and labour costs increase, the cost savings of using technology to replace labour are also starting to become significant in emerging economies. Also, there is empirical evidence that automation in advanced economies leads to a re-shoring of production and may contribute to job losses in emerging markets.48

56.3 Labour, Taxation, and the ‘Rise of the Robots’ While there is consensus that the technological advancements of the ‘fourth industrial revolution’ will bring at least short-term disruption of the labour market and a transition of the workforce in almost all sectors of production and services, the overall regional and sectoral long-term impact on employment, and correspondingly wage tax revenues and social security contributions, is less clear. This uncertainty and anxiety has resulted in an intense discussion about ‘taxing robots’ (or their use) and automation more generally or for changing existing income tax policies in anticipation of substantial job losses due to automation substitution as well as intense discussion of the justification and limitation of such ideas which is ongoing and intensifying.49 It is, moreover, connected with 44

Ibid., 44. Ibid., 44. 46 Ibid., 49–50. 47 See D. Alonso Soto, ‘Technology and the future of work in emerging economies: What is different?’, OECD Social, Employment and Migration Working Papers No. 236 (2020). 48 See, e.g., M. Faber, ‘Robots and Reshoring: Evidence from Mexican Labor Markets’, Journal of International Economics 127 (2020), art. 103384 (finding that US robot adoption lowers labour demand in Mexican export-producing sectors). 49 See, e.g., X. Oberson, ‘Taxing Robots? From the Emergence of an Electronic Ability to Pay to a Tax on Robots or the Use of Robots’, World Tax Journal 9 (2017), 247–261; M. M. Erdoğdu, and C. Karaca, ‘The Fourth Industrial Revolution and a Possible Robot Tax’, in I. Berksoy, K. Dane, and M. Popovic, eds, Institutions & Economic Policies: Effects on Social Justice, Employment, Environmental Protection & Growth (London: IJOPEC Publications, 2017), 103–122; R. Abbott and B. Bogenschneider, ‘Should 45

1034 Georg Kofler the perceived needs to reduce increased income inequality, to finance income-support policies and welfare programmes, education,50 etc. for displaced workers, or to consider a ‘universal basic income’ (UBI) to address long-term technological unemployment.51 Indeed, it has even been argued that ‘[a]‌moderate tax on robots, even a temporary tax that merely slows the adoption of disruptive technology, seems a natural component of a policy to address rising inequality’.52 The proposals for reform, therefore, also include very specific measures, such as imputing income to ‘robots’ or a ‘tax on the work performed by a robot’,53 or ‘automation taxes’ addressing labour-replacing technology. Specifically with regard to the income

Robots Pay Taxes? Tax Policy in the Age of Automation’, Harvard Law & Policy Review 12 (2018), 145–175; S. Dorigo, ‘Robots and Taxes: Turning an Apparent Threat Into an Opportunity’, Tax Notes International 92 (10 Dec. 2018), 1079–1085; T. Falcão, ‘Should My Dishwasher Pay a Robot Tax?’, Tax Notes International 90 (11 June 2018), 1273–1277; J. A. Soled and K. D. Thomas, ‘Automation and the Income Tax’, Columbia Journal of Tax Law 10 (2018), 1–48; U. Thuemmel, ‘Optimal Taxation of Robots’, CESifo Working Paper No. 7317 (2018); Vishnevsky and Chekina, ‘Robot vs. Tax Inspector’; S’ Ahmed, ‘Cryptocurrency & Robots: How to Tax and Pay Tax on Them’, South Carolina Law Review 69 (2019), 697–740; M. Barros, ‘Robots and Tax Reform: Context, Issues and Future Perspectives’, International Tax Studies 2 (2019); J. Daubanes and P.-Y Yanni, ‘The Optimal Taxation of Robots’, IEB Report 2/ 2019 (2019), 7–9; Englisch, ‘Digitalization and the Future of National Tax Systems: Taxing Robots?’, in W. Haslehner et al., eds, Tax and the Digital Economy: Challenges and Proposals for Reform (Alphen aan den Rijn: Kluwer, 2019), 261–281; R. Goulder, ‘Taxing Robots: Is Negative Depreciation in Your Future?’, Tax Notes International 95 (16 Sept. 2019), 1203–1206; M. Hammer, ‘Brave New World: Automation, Unemployment and Robot Taxes’, IBFD White Paper (2019); D. J. Hemel, ‘Does the Tax Code Favor Robots?’, University of Chicago Coase-Sandor Institute for Law & Economics Research Paper No. 900 (2019); R. F Mann, ‘I Robot: U Tax? Considering the Tax Policy Implications of Automation’, McGill Law Journal 64 (2019), 1–43; O. Mazur, ‘Taxing the Robots’, Pepperdine Law Review 46 (2019), 277–329; B. McCredie, K. Sadiq, and L. Chapple, ‘Navigating the fourth Industrial Revolution: Taxing Automation for Fiscal Sustainability’, Australian Journal of Management 44 (2019), 648–664; X. Oberson, Taxing Robots (Cheltenham: Edward Elgar, 2019); X. Oberson, ‘Taxing Robots?’, IEB Report 2/2019 (2019), 10– 11; U. Thuemmel, ‘A Case for Taxing Robots?’, IEB Report 2/2019 (2019), 12–13; V. Chand, S. Kostić, and A. Reis, ‘Taxing Artificial Intelligence and Robots: Critical Assessment of Potential Policy Solutions and Recommendation for Alternative Approaches’, World Tax Journal 12 (2020), 711–761; C. Dimitropoulou, ‘Scaling Back Tax Preferences on Artificial Intelligence-Driven Automation: Back to Neutral?’, World Tax Journal 12 (2020), 410–462; J. Englisch, ‘Taxation of Robots’, in Y. Brauner, ed., Research Handbook on International Taxation (Cheltenham: Edward Elgar, 2020), 369–384; F. Haase, ‘KI und Steuerrecht’, in M. Ebers et al., eds, Künstliche Intelligenz und Robotik (Munich: C. H. Beck, 2020), 636–653; K. Kisska- Schulze and K. Davis-Nozemack, ‘Humans vs. Robots: Rethinking Tax Policy For a More Sustainable Future’, Maryland Law Review 79 (2020), 1009–1056; R. Kovacev, ‘A Taxing Dilemma: Robot Taxes and the Challenges of Effective Taxation of AI, Automation and Robotics in the Fourth Industrial Revolution’, Ohio State Technology Law Journal 16 (2020), 182–217; R. de la Feria and A. Grau Ruiz, ‘Taxing Robots’, in A. Grau Ruiz, ed., Interactive Robotics: Legal, Ethical, Social and Economic Aspects (Berlin: Springer Nature, 2022), 557–575. 50 For the proposal of earmarked education taxes, see Chand, Kostić, and Reis, ‘Taxing Artificial Intelligence and Robots’, 745–756. 51 See, e.g., Mann, ‘I Robot: U Tax?’, 31–42. 52 R. Shiller, ‘Why robots should be taxed if they take people’s jobs’, The Guardian (22 Mar. 2017). 53 See Point K in the Report with recommendations to the Commission on Civil Law Rules on Robotics, A8-0005/2017 (27 Jan. 2017).

The Future of Labour Taxation 1035 tax system,54 the policy rationales of these ideas are manyfold and partly overlapping. Some concepts focus on replacing the tax revenue from the (routine) jobs lost to automation; that is, to drag the presumed disappearing labour tax base back into the tax system (which is not otherwise compensated for by higher profits due to lower costs following automation55), and to restore neutrality between human and automated work. Others aim at ‘corrective taxes’ to slow the pace of automation (or, conversely, incentivize employment of humans), while still others approach the debate from the perspective of the redistribution of income, specifically from non-routine workers to routine human workers,56 or from the perspective of an overall divergence of benefits not (only) between categories of workers but rather between labour and capital. Some of those narrower proposals should be highlighted briefly.57 One concept of ‘robot taxation’ focuses on the concrete displacement of workers and the disappearing labour tax base. It aims at imputing income in the amount of a hypothetical salary to labour-replacing ‘robots’ and subject it to income tax and social security contributions either at the level of the owner (i.e. a tax on the use of robots) or— rather futuristically and inspired by the discussion on the legal personality of certain sophisticated autonomous ‘robots’ (‘status of electronic persons’)58—at the level of the robots themselves (i.e. a tax on robots as taxpayers).59 This approach of taxing a deemed salary, it is argued, would correspond to the reality of automation replacing humans and 54 Also, some concepts focus on special automation taxes on specific fields of automation (e.g. automated cashiers in supermarkets, automated taxis) as well as object taxes on robots (similar to existing taxes on cars, motorbikes, etc.) or fees for using robots (to compensate for specific or general benefits such as infrastructure, control, supervision, etc. provided by the state) (see, e.g., Oberson, ‘Taxing Robots?’, 257–258; Oberson, Taxing Robots, 129–131). The latter idea was also highlighted in a motion for a resolution by the EU Parliament (see the Report with recommendations to the Commission on Civil Law Rules on Robotics, A8-0005/2017 (27 Jan. 2017), referring to a ‘a fee for using and maintaining a robot’). 55 See also the calculations in Oberson, Taxing Robots, 116–120, arguing that: (1) with a corporate tax rate that is typically lower than the individual income tax rate so that revenue gains from marginally higher corporate income do not offset revenue losses from marginally lower personal income; and (2) investments in robots are depreciable, often on an accelerated schedule. 56 See J. Guerreiro, S. Rebelo, and P. Teles, ‘Should Robots Be Taxed?’, Review of Economic Studies 88 (2021) (in print) (see also NBER Working Paper 23806). 57 For a comprehensive overview of the various proposals, see M. Barros, ‘Robots and Tax Reform: Context, Issues and Future Perspectives’, International Tax Studies 2 (2019). 58 See Point 59(f) of the European Parliament resolution of 16 February 2017 with recommendations to the Commission on Civil Law Rules on Robotics (2015/2103(INL), P8_TA(2017)0051 (16 Feb. 2017), and the detailed discussion of giving AI legal personality in J. J. Bryson, M. E. Diamantis, and T. D Grant, ‘Of, For, and By the People: The Legal Lacuna of Synthetic Persons’, Artificial Intelligence and Law 25 (2017), 273–291, and in J. Turner, Robot Rules (Basingstoke: Palgrave Macmillan, 2019), 173–205. See also the philosophical perspective by L. Floridi, ‘Robots, Jobs, Taxes, and Responsibilities’, Philosophy & Technology 30 (2017), 1–4, noting that ‘[i]‌f robots become one day as good as human agents . . . we may adapt rules as old as Roman law, according to which the owner of an enslaved person was responsible for any damage caused by that person (respondeat superior). As the Romans already knew, attributing some kind of legal personality to robots would deresponsabilise those who should control them.’ 59 See for a detailed proposal, Oberson, ‘Taxing Robots?’, 247–261; Oberson, Taxing Robots, 114–120 and 131–137; X. Oberson, ‘Taxing Robots?’, IEB Report 2/2019 (2019), 10–11.

1036 Georg Kofler is based on a comparability analysis that is also common, inter alia, in transfer pricing.60 On the other hand, these proposals (perhaps too confidently) assume that the technology used in automation can be linked to specific job losses and that there can be a clear differentiation between job-enhancing and job-replacing robots.61 While many technical issues of such taxes (e.g. economic double taxation, sourcing, deductibility of imputed income, etc.) might be solvable,62 decisive fundamental objections remain, among which is the decisive question of how to define a ‘robot’ in the light of the multitude of complex cyber-physical systems63 without leading to arbitrary distortions of a firm’s decisions. Also, it has been doubted if such a specific ‘robot’ tax, particularly one on robots as taxable persons, would be justifiable in the light of the ability-to-pay principle (or the insurance principle in social security systems)64 and be brought in line with considerations of neutrality, simplicity, certainty, and effectiveness.65 Other concepts take a broader perspective on labour-replacing technology and advance several options for an ‘automation tax’.66 These range from reducing incentives for investments in the automation of production,67 disallowing corporate tax deductions for automated workers68 to various ideas of general taxes that target the ratio of automation.69 The latter are aimed at substituting the wage taxes and social security 60 Oberson, Taxing Robots, 122. 61

See for that line of criticism, e.g., Mazur, ‘Taxing the Robots’, 302. See the detailed discussion by Oberson, Taxing Robots, 112ff. 63 Attempts at a workable definition in various fields of technology and law abound, but might not necessarily be suitable for taxation. The European Parliament in Point 1 of its Resolution of 16 February 2017, P8_TA(2017)0051, e.g., called on the ‘Commission to propose common Union definitions of cyber physical systems, autonomous systems, smart autonomous robots and their subcategories by taking into consideration the following characteristics of a smart robot’, and named: (1) the acquisition of autonomy through sensors and/or by exchanging data with its environment (inter-connectivity) and the trading and analysing of those data; (2) self-learning from experience and by interaction (optional criterion); (3) at least a minor physical support; (4) the adaptation of its behaviour and actions to the environment; and (5) absence of life in the biological sense. For further discussion and doubts regarding the feasibility of the EU Parliament’s definition for tax purposes, especially in that it excludes many types of labour- replacing robots (e.g. algorithms without physical form) but includes many forms of labour-enhancing robots, see, e.g., Falcão, ‘Should My Dishwasher Pay a Robot Tax?’ and Englisch, ‘Digitalization and the Future of National Tax Systems: Taxing Robots?’, 277–278; Mazur, ‘Taxing the Robots’, 298–299. 64 For detailed criticism, see Englisch, ibid., 264–272. 65 See, e.g., Chand, Kostić, and Reis, ‘Taxing Artificial Intelligence and Robots’, 733–745. 66 See, e.g., Abbott and Bogenschneider, ‘Should Robots Pay Taxes?’, 168–175. 67 For a detailed assessment, see C. Dimitropoulou, ‘Scaling Back Tax Preferences on Artificial Intelligence-Driven Automation: Back to Neutral?’, World Tax Journal 12 (2020), 410–462. For a brief discussion of the so-dubbed South Korean ‘robot tax’, see, e.g., Goulder, ‘Taxing Robots: Is Negative Depreciation in Your Future?’, 1206. More generally, the tax system may lead to automation even in cases where it not otherwise efficient; see Abbott and Bogenschneider, ‘Should Robots Pay Taxes?’, 163–168. 68 Abbott and Bogenschneider, ibid., 169–170; Oberson, Taxing Robots, 124–125. 69 See, e.g., Abbott and Bogenschneider, ibid., 171–172, referring to Meisel’s ‘payroll tax on computers’ (see W. Meisel, The Software Society: Cultural and Economic Impact (Bloomington, IN: Trafford Publishing, 2013), 220); Oberson, Taxing Robots, 120–122. Expanding that idea even further to a tradeable permits approach (similar to the Emissions Trading Scheme) where quotas of human employment can be bought and sold, F, D’Orlando, ‘Problems, solutions and new problems with the third wave of technological unemployment’, CreaM-Working Paper Series No. 2/2018 (2018). 62

The Future of Labour Taxation 1037 contributions that would have been paid by the worker and employer if the human worker had continued to perform the work that was replaced by automation. Such ‘automation tax’ would then be based on a ratio of, for example, a company’s revenues or profits to, say, the number of employees or gross employee compensation expense.70 The basic idea, however, is that such tax would increase as the revenue-per-employee ratio grows, making it less attractive to replace workers with automation. Similar proposals advocate an automation tax in the sense that enterprises pay additional amounts into an insurance plan or sovereign wealth fund if layoffs are due to automation.71 Also here, however, severe systematic concerns regarding such compensatory levies exist, primarily if one takes the perspective that (progressive) wage taxation is at least statutorily a burden on employees, whereas both costs for labour and capital and their tax treatment are neutral from the perspective of the employer, even if capital investments could be expensed immediately or depreciated on an accelerated schedule.72 Intuitively, of course, there are serious economic policy objections against specific ‘robot taxes’ or ‘automation taxes’ as they would potentially result in welfare losses. Such taxes could cause a drag on investment and innovation,73 higher prices for products and services, lower profits, decreased competitiveness of domestic production in a globalized economy, etc.; specifically the restraints imposed by international tax competition render the taxation of automation an unfeasible option, at least unless it is comprehensively coordinated on an international scale.74 This also seems to be the current political sentiment.75 Moreover, to date, very few economic studies on the optimal taxation of robots exist,76 and their value for practical conclusions is quite limited.77 Nevertheless, the concepts for specifically targeted ‘robot taxes’ and ‘automation taxes’ are valuable policy ideas to help focus the discussion. They also provide context for much broader debates on the tax system, as the potential impact of automation is to be viewed not only in the dimension of various types of labour, such as routine vs non- routine work, but also along the dimension of labour and companies and their owners, 70 For

the various considerations with regard to determining the ratio (revenues vs profits and number of employees vs employee compensation expense), see Abbott and Bogenschneider, ‘Should Robots Pay Taxes?’, 171–172. 71 See, e.g., ibid., 170–171. 72 See, e.g., Englisch, ‘Digitalization and the Future of National Tax Systems: Taxing Robots?’, 269–271. 73 See D. Spencer et al., ‘Digital automation and the future of work’, European Parliamentary Research Service, Scientific Foresight Unit (STOA), PE 656.311 (Jan. 2021), 46. 74 See Englisch, ‘Digitalization and the Future of National Tax Systems: Taxing Robots?’, 217–272. 75 For a brief overview of various political reactions see, e.g., Chand, Kostić, and Reis, ‘Taxing Artificial Intelligence and Robots’, 743–745. 76 Theoretical economic literature has addressed the ‘optimal taxation of robots’ in the light of the disruption of the labour market to the disadvantage of routine work and how redistributive policy should respond to automation. At the outset, these studies are not concerned with the taxation of robots to replace the tax revenue from the routine jobs lost to automation (as in these economic models automation increases output and overall tax revenue), but rather on the redistribution of income, specifically from non- routine workers to routine workers, through wage compression. See, e.g., Guerreiro, Rebelo, and Teles, ‘Should Robots Be Taxed?’; Thuemmel, ‘Optimal Taxation of Robots’. 77 See also Englisch, ‘Digitalization and the Future of National Tax Systems: Taxing Robots?’, 273–274.

1038 Georg Kofler who will probably be the biggest beneficiaries of these developments. This, in turn, has renewed the more general calls for shifting the burden of taxation from labour to capital, for example by taxing (unrealized) capital gains or by increasing corporate income tax, despite the traditional (and not undisputed78) optimal capital-taxation conclusion that neither capital income nor wealth should be taxed at all.79 Conversely, therefore, it has also been advocated to grant offsetting tax preferences for human workers to make automation relatively more expensive.80

56.4 Outlook It is hard to fathom a future where the ‘AI revolution’ has resulted in a world where for sustained periods everything from housing to education, food, clothing, etc. will become increasingly cheaper and where human labour has ceased to be the binding constraint to economic growth and its price has fallen towards zero. What will and should our legal, tax, and welfare systems look like? Would we only tax companies’ market capitalization and land? Would then everyone be entitled to an ‘AI dividend’ to share the wealth created by technology? This extreme picture is not, however, reflected in the current data and economic outlooks, which paint a rather optimistic picture of the overall effects of automation on employment and wages. It is nevertheless likely that major shifts in the labour market are and will be taking place, disrupting all parts of the economy and requiring a focus on those who ‘lose their jobs to robots’. It will need the right policy mix to steer and redistribute the benefits of automation, and taxation might play a key role in this endeavour. Indeed, there is not only an intense discussion about targeted ‘robot taxes’ and ‘automation taxes’ to address those concerns directly, but also an—arguably more fruitful and sustainable—‘ongoing debate about the unequal distribution of rents from processes of digital transformation and the unequal balance of taxation on capital and labour income, with a need to rebalance in favour of the latter’.81 Several concepts and ideas for dealing with the impact of this fourth industrial revolution from a tax perspective are on the table, all of them with their own strengths and weaknesses, and it is an important discussion that needs to be had. It is not, however, the only one. Technological advances might also intensify the debate

78 See,

e.g., C. W. Sanchirico, ‘Why the Optimal Long-Run Tax Rate on Capital is Zero . . . or Very High: The Missing Explanation’, ILE Resarch Paper No. 20-33 (2020). 79 See for that debate, e.g., Mazur, ‘Taxing the Robots’, 304– 328, favouring an elimination of ‘the distinction between labour and capital income for tax purposes and curtailing certain tax expenditures that significantly subsidize the creation of capital income’. 80 Abbott and Bogenschneider, ‘Should Robots Pay Taxes?’, 171. 81 D. Spencer et al., ‘Digital automation and the future of work’, 46.

The Future of Labour Taxation 1039 about nexus and profit allocation in international tax law, for example when autonomous robots replace ‘significant people functions’.82 Moreover, the rise of the gig and sharing economies, for example, leads to changes in the mix of the taxable status in the economy (from employee to self-employed or incorporated).83 ‘Gig workers’, ‘crowd working’, the ‘sharing economy’, ‘digital nomads’,84 and more generally a working environment that sometimes might not require physical presence at the employer’s place of business85 have not only challenged labour and social security laws around the world, but also domestic tax qualifications.86 The OECD has addressed some of these issues briefly in its 2018 Interim Report on BEPS Action 1,87 and continues that work. Those issues, however, are not isolated within jurisdictions but rather fundamentally challenge the traditional international notions of residence and employment. Being developed in times of relative immobility and traditional forms of employment, the allocation of taxing rights on cross-border active labour income in article 15 of the OECD and UN Models88 focuses the right to the state in which the employment is actually exercised (and has long abandoned the principle that the right to tax is that of the state in which the recipient of the income is resident), but only if there is sufficient nexus to the employment state, either because the engagement is not merely of a temporary nature or because the remuneration is deductible there: hence, the exclusive taxing right for the employment income remains with the taxpayer’s residence state unless the taxpayer either (actually) exercises the employment in the other state for more than 183 days in a twelve-month period or if the employer is a resident of that other state or has a permanent establishment there which bears the remuneration. A clear consequence of this system is ‘that a resident of a Contracting State who derived remuneration, in respect of an employment, from sources in the other State could not be taxed in that other State in respect of that remuneration merely because

82

Englisch, ‘Digitalization and the Future of National Tax Systems: Taxing Robots?’, 278–279. See also OECD, Policy Responses to New Forms of Work Paris: (OECD Publishing, 2019). 84 S. Kostić, ‘In Search of the Digital Nomad –Rethinking the Taxation of Employment Income under Tax Treaties’, World Tax Journal 11 (2019), 189–225. 85 For a tax treaty perspective see, e.g., OECD, ‘Updated guidance on tax treaties and the impact of the Covid-19 pandemic’ (21 Jan. 2021). 86 See, e.g., K. D. Thomas, ‘Taxing the Gig Economy’, University of Pennsylvania Law Review 166 (2018), 1415–1473; S. Kostić, ‘In Search of the Digital Nomad –Rethinking the Taxation of Employment Income under Tax Treaties’, World Tax Journal 11 (2019), 189–225; A. Milanez and B. Bratta, ‘Taxation and the Future of Work: How Tax Systems Influence Choice of Employment Form’, OECD Taxation Working Paper No. 41 (2019); C. Black, ‘The Future of Work: The Gig Economy and Pressures on the Tax System’, Canadian Tax Journal 68 (2020), 69–97; J. Li, A. Choi, and C. Smith, ‘Automation and Workers: Re-Imagining the Income Tax for the Digital Age’, Canadian Tax Journal 68 (2020), 99–124. 87 See paras 471–474 in OECD/G20 Base Erosion and Profit Shifting Project, Tax Challenges Arising from Digitalisation–Interim Report 2018: Inclusive Framework on BEPS (Paris: OECD Publishing, 2018). 88 Just as before art. VI of the London Model and art.VII of the Mexico Model. 83

1040 Georg Kofler the results of his work were exploited in that other State’.89 These rules are hence based on the physical presence of the employee in a certain state and have not significantly changed since the 1950s. New forms of labour and their exercise enabled by technological advancements might require a rethinking of the international allocation and enforcement of taxing rights in these situations.90

89 See 2017 OECD Commentary on art. 15 para. 1, which goes back (unchanged) to Annex F para. 3 on art. VII of the 1959 Report. 90 For a broad call for reform, see S. Kostić, ‘In Search of the Digital Nomad –Rethinking the Taxation of Employment Income under Tax Treaties’, World Tax Journal 11 (2019), 189–225.

Chapter 57

Digi taliz ation a nd t h e F u tu re of VAT i n t h e Eu ropean U ni on Michael Tumpel

57.1 Introduction According to Garnter’s glossary of information technology ‘digitalization is the use of digital technologies to change a business model and provide new revenue and value- producing opportunities; it is the process of moving to a digital business’.1 Developments in digital information technology over the past decades have moved beyond digitization to transform tax administrations and businesses alike. Whereas digitization is defined as the pure analogue-to-digital conversion of existing data2 and documents, digitalization leverages digital information and transforms the processes of doing business and collecting taxes. The development and use of digital technologies increase efficiencies and, ultimately, revenue. Digitalization leads to digital transformation of the business, the economy, and society. VAT is affected by digitalization in many ways. In this chapter, we will highlight two aspects of digitalization for VAT. First, digital transformation of businesses forces changes in VAT law in order to ensure the revenue stream. The OECD has addressed the challenges of digitalization in its Action 1 Final Report as part of its Base Erosion and Profit Shifting (BEPS) Project,3 which stated that as digitalization spreads across 1

Gartner Glossary Information Technology Glossary, ‘Digitalization’, available at https://www.gartner. com/en/information-technology/glossary/digitalization (accessed 30 March 2021). 2 Digitization vs. digitalization: Differences, definitions and examples— TruQC’, available at https://www.truqcapp.com/digitization-vs-digitalization-differences-definitions-and-examples/ (30 March 2021). 3 See OECD, Addressing the Tax Challenges of the Digital Economy, Action 1— 2015 Final Report, OECD/G20 Base Erosion and Profit Shifting Project (Paris: OECD Publishing, 2015).

1042 Michael Tumpel the economy, the digital economy is becoming the economy itself and ring-fencing it from the rest of the economy is no longer possible.4 Digitalization is breaking into all value chains, starting with travel, banking, telecommunications, automobiles, urban services, and healthcare.5 The digital economy is not limited to the GAFAM (Google, Apple, Facebook, Amazon, and Microsoft) and goes far beyond to include all sectors of the economy. It also cannot be isolated because that would require arbitrary lines to be drawn between what is digital and what is not.6 Digitalization has led to the development of new business models parallel or overlapping with traditional business or complementing those business (e.g. distance sales of goods)7 and also allows for conducting business on a substantially greater scale and over longer distances than was previously possible.8 These characteristics of the digital economy increase the already existing challenges of cross-border trade in goods and services for the VAT system. Over the last two decades, the EU has adjusted its VAT system to cope with those challenges and those changes to EU VAT law will be described and analysed in the following section. However, some adjustment and improvement of existing regulations are still needed in the future. A few of the new rules regarding distance sales and transfers of financial information came into force only recently. Whether those measures are effective or not will also be evaluated later in the chapter. Secondly, digitalization allows governments to process more information on the economic activities of taxpayers. Tax-enforcement technology will be improved through digitalization and can therefore contribute to efforts to lower tax evasion.9 Further extension of data collection and processing may possibly conflict with private data protection therefore the interests of tax authorities and citizens with regard of digitalization must be balanced to achieve an outcome which is in line with fundamental rights of individuals. Not only tax-enforcement technology by governments will be a factor in combating tax evasion; non-governmental peer-to-peer-networks based on blockchain technology could be applied to secure cross-border exchange of information about transactions within the scope of VAT.10 Tax authorities may encourage such initiatives as additional cost-efficient tools to avoid tax fraud. Peer-to-peer networks can also function as a marketplace where supplies of goods and services take place and if that evolves in economic activities which fall within the scope of VAT, tax authorities must register potential taxpayers and ensure compliance with all requirements of VAT law.

4

See ibid., paras 115 and 364. P. Collin and N. Colin, Task Force on Taxation of the Digital Economy (Paris: Ministry of the Economy and Finances and Ministry for Economic Regeneration, 2013), 3. 6 See para. 115 in OECD, Addressing the Tax Challenges of the Digital Economy, Action 1—2015 Final Report. 7 See ibid., para. 116. 8 See ibid., para. 116. 9 See B. Jacobs, ‘Digitalization and Taxation’, in S. Gupta et al., eds, Digital Revolutions in Public Finance (Washington, DC: IMF, 2018), 28. 10 See R. T. Ainsworth and A. Shact, ‘Blockchain (Distributed Ledger Technology) Solves VAT Fraud’, Boston University School of Law & Economics Working Paper No. 16-41 (2016), 15. 5 See

Digitalization and the Future of VAT in the European Union 1043 On the one hand, relevant information about the responsibility for VAT for potential taxpayers and easy access to fulfil those obligations must be provided. Digital solutions for these tasks already exist but others must be extended or newly implemented. On the other hand, tax authorities can use digital tools to monitor such transactions in peer- to-peer-networks. If there is no access to such networks or not enough relevant information can be obtained, tax authorities may rely on information provided by financial intermediaries. However, blockchain technology makes financial transactions possible without a specific financial intermediary using decentralized public ledgers to exchange information and transfer funds. Such schemes are difficult to monitor especially if the parties are disguised by encryption. However, the European Commission is working on ensuring the exchange of information regarding crypto-assets and e-money.11 Services without direct compensation in money are a feature not limited to but often key to digital business models. ‘Free’ services are often provided by digital businesses expecting user participation or user data in return. Whether those services fall within the scope of VAT is sometimes questionable since there is no direct link between the supply and a clearly defined consideration by the recipient of such services whose value can be expressed in monetary terms. We will analyse whether such transactions are taxable and what measures can be taken in order to ensure that services consumed are properly taxed if necessary.

57.2 History of VAT in the European Union 57.2.1 Key Features of EU VAT EU VAT is a general tax on consumption. However, tax is charged on every transaction which takes place in the production and distribution process including, but not exclusively, the retail trade stage. Businesses can deduct the amount of VAT borne by the various cost components and therefore avoid carrying the tax burden directly. Moreover, the burden of VAT should be shifted over to consumers spending their money on goods and services without being entitled to deduction of input VAT.

11 The

initiative will amend the Directive on Administrative Cooperation (DAC) to ensure that EU rules stay in line with the evolving economy and include other areas such as crypto-assets and e- money, see for further information European Commission, ‘Tax fraud & evasion—strengthening rules on administrative cooperation and expanding the exchange of information’, available at https://ec.eur opa.eu/info/law/better-regulation/have-your-say/initiatives/12632-Strengthening-existing-rules-and- expanding-exchange-of-information-framework-in-the-field-of-taxation-DAC8-/public-consultation (accessed 30 March 2021); see also Proposal for a COUNCIL DIRECTIVE amending Directive 2011/16/ EU on administrative cooperation in the field of taxation, COM/2022/707 final.

1044 Michael Tumpel From an economic perspective, the answer to the question who carries the burden of VAT does not depend on the legislator’s concept and perceptions but rather on the outcome of market forces. If demand is perfectly price elastic, every change of prices caused by a change of taxation (which increases the cost of businesses that are liable to tax) will change demand and therefore the profits of businesses. Businesses can either adjust prices and deal with the change of demand or absorb the tax and deal with the change of costs. In both cases, business profits would be impacted by VAT. Supplies of goods and services are subject to VAT in a specific EU member state if they take place within its jurisdiction. Whether or not VAT can be levied in a member states depends on the rules of place of supply. If comparable supplies of goods and services have the same place of supply, they are subject to VAT in the same member state, independent of analogue or digital supply. Digitalization would therefore have no direct consequence with regard to the ability of member states of the EU to levy VAT, if the rules regarding the place of supply were designed properly according to the destination principle and the assessment of VAT could be ensured. Therefore, all depends on how the place of supply is defined and whether the compliance of taxpayers can be expected. Digitalization gives tax authorities plenty of possibilities for encouraging compliance by taxpayers as reliable information about their obligations as well as digital tools to ease their realization can be provided. The VAT system was originally adopted as the common turnover tax system of the European Economic Community (EEC) in a purely analogue world. It proved to be neutral and robust. Fractioned tax payment as one of its key features prevented total loss of revenue if supplies to consumers remained untaxed and secured the robustness of the system. Neutrality of VAT guarantees the equal treatment of transactions irrespective of the type of supply and supplier. These key features of the VAT system will make the system no less effective for the digital economy if some rules are adapted due to the specific needs of digitalization.

57.3 VAT Directives 1967–2006 Preliminary studies showed that different systems of turnover taxes within the EEC led to distortion of competition. Modelled on the French TVA, the EEC implemented the VAT system as a common method of general consumption tax by adapting the member states’ existing tax systems in accordance with the First and Second Council Directives of 11 April 1967 on the harmonization of the laws of the member states relating to turnover taxes.12 Pursuant to article 2 of the Frist Directive:

12 See

First Council Directive 67/227/EEC of 11 April 1967 on the harmonisation of legislation of Member States concerning turnover taxes [1967] OJ 71/1301.

Digitalization and the Future of VAT in the European Union 1045 The principle of the common system of value added tax involves the application to goods and services of a general tax on consumption exactly proportional to the price of the goods and services, whatever the number of transactions which take place in the production and distribution process before the stage at which tax is charged. On each transaction, value added tax, calculated on the price of the goods or services at the rate applicable to such goods or services, shall be chargeable after deduction of the amount of value added tax borne directly by the various cost components. The common system of value added tax shall be applied up to and including the retail trade stage.

Member states should have enacted it as rapidly as possible but no later than 1 January 1970. The Second Directive determined the structure and procedures for application of the common system of VAT.13 The supply of goods and the provision of services within the territory of the country by a taxable person against payment, as well as the importation of goods, had been subject to VAT, whereas the tax rates and exemptions should have been fixed by the member states. This gave the member states a lot of latitude to adopt special measures in order to simplify the charging procedure in respect of the tax or to prevent certain frauds. In 1977, the Council adopted the Sixth Directive on the harmonization of the laws of the member states relating to turnover taxes. This would establish a common system of VAT and a uniform basis of assessment by 1 January 1978. It also clarified terms that had led to earlier difficulties, such as ‘taxable person’ and ‘taxable transaction’. Definitions of the place of supply of goods and services were implemented to avoid conflicts between the member states. The place where a supply of services is effected should, in principle, be defined as the place where the person supplying the services has their principal place of business or has a fixed establishment from which the service is supplied. However, there have been certain exceptions for specific services, such as those connected with immovable property, transportation, and other services. To avoid double taxation, non- taxation, or the distortion of competition, the member states may, with regard to specific supplies of services, consider the place of supply which would generally be situated within its territory, otherwise if it were situated outside the Community then the effective use and enjoyment of the services would take place outside the Community. Prior to the abolition of the fiscal frontiers for all transactions between member states in order to achieve the internal market as from 1 January 1993, it was necessary to limit the imposition of tax on imports and the remission of tax on exports to transactions with territories excluded from the scope of the common system of VAT. Hence, special

13 See Second Council Directive 67/228/EEC of 11 April 1967 on the harmonisation of legislation of Member States concerning turnover taxes—Structure and procedures for application of the common system of value added tax [1967] OJ 71/1303.

1046 Michael Tumpel rules for the taxation of intra-Community trade were implemented with the directive with a view to the abolition of fiscal frontiers.14 In a world before digitalization of the economy, it seemed reasonable that the location of the supply of services would be where the person supplying the services had their principal place of business or had a fixed establishment from which the service was supplied, because most services were rendered locally. Different tax rates and other disparities were important only for the economy close to the border. Taxation of supplies of goods, according to the destination principle as per the intra-Community transaction rules for businesses, precluded distortion of competition on the domestic market. However, consumers would benefit from the abolition of fiscal frontiers by taking home goods which they had acquired in another Member State without further taxation. To limit distortion of competition in the market for goods, however, distance sales above a threshold of 100,000 per year and supplies for the acquisition of new vehicles were also taxable in the member state of destination where those goods were supposed to be used by the final consumer. Small merchants whose revenues did not exceed the threshold could choose whether their supplies were liable in the member state of origin or destination. Importation from third countries outside the EU were subject to VAT in the member state within whose territory the goods were located when they entered the Community. The risk that suppliers would not pay the VAT to the tax authorities or customs was low as long as importers were only in contact with businesses in the member states of destination and the customs authorities were able to monitor importations in bulk. Beginning with the internet boom of the early 2000s, electronic services and online marketplaces boomed and brought new challenges not only for traditional domestic businesses but also for the VAT system and its administration. The rules applicable to VAT on radio and television broadcasting services and on electronically supplied services seemed to be inadequate for taxing such services consumed within the Community and for preventing distortions of competition in that area. This is why, effective 1 July 2003, all radio and television broadcasting services and electronically supplied services provided from third countries to persons established in the Community or from the Community to recipients established in third countries were taxed at the place of the recipient of the services.15 In order to facilitate compliance with fiscal obligations by operators providing electronically supplied services, who were neither established nor required to be identified for tax purposes within the Community, a one-stop shop regime was established. Operators supplying services by electronic means to non-taxable persons within the Community, could, if they were not otherwise

14 See Corrigendum to Council Directive 91/ 680/EEC of 16 December 1991 supplementing the common system of value added tax and amending Directive 77/388/EEC with a view to the abolition of fiscal frontiers [1992] OJ L272/72. 15 See Council Directive 2002/38/EC of 7 May 2002 amending and amending temporarily Directive 77/388/EEC as regards the value added tax arrangements applicable to radio and television broadcasting services and certain electronically supplied services [2002] OJ L128/41.

Digitalization and the Future of VAT in the European Union 1047 identified for tax purposes within the Community, opt for identification in a single member state and fulfil their VAT obligations for all member states electronically. This could be seen as a first step to address the challenges for VAT caused by digitalization.

57.4 VAT Directives 2006–2020 Because the Sixth Directive had been significantly amended on several occasions and was therefore difficult to read, the EU wanted to recast the VAT Directive in 200616 for reasons of clarity and rationalization. The recast directive would improve the structure and the wording of the directive but would not entail material changes to the existing legislation. In 2008, the Council of the EU adopted a directive amending the new VAT Directive (2006/112/EC) with regard to the place of supply of services.17 Changes in technology and globalization have created enormous changes in the volume and pattern of trade in services which are increasingly supplied at a distance. Only small steps had been taken to address that development until a large overhaul in 2010. Since that time, for supplies of services to taxable persons, the general rule with respect to the place of supply of services is based on the destination principle (i.e. the place where the recipient is established). Where services are supplied to non-taxable persons, however, the general rule is still that the place of supply of services is the place where the supplier has established their business. In many cases, exceptions from this general rule have been implemented which reflect the principle of taxation at the place of consumption. In particular, the place of supply of electronically supplied services to non-taxable persons who are established in a member state is deemed to be the place where the non-taxable person is established or where they have their permanent address or usually reside. This has been combined with a one-stop shop regime which was previously limited to operators established outside the EU but has now been extended to all operators established in a member state who carry out such transactions in other member states. Technological change and globalization have also resulted not only in an increase of cross-border service supplies but also in the explosive growth of electronic commerce and, therefore, of distance sales of goods, both supplied from one member state to another and from third territories or third countries to the Community. Hence, the special scheme for electronically supplied services supplied by taxable persons established within the Community but not in the member state of consumption would be extended to the intra-Community distance sales of goods and a similar special scheme would be introduced for distance sales of goods imported from third territories or third countries 16 Council Directive 2006/112/EC of 28 November 2006 on the common system of value added tax [2006] OJ L347/1. 17 See Council Directive 2008/8/EC of 12 February 2008 amending Directive 2006/112/EC as regards the place of supply of services [2008] OJ L44/11.

1048 Michael Tumpel by 1 July 2021.18 These schemes take the principle of taxation at destination into account. They also protect member states’ tax revenue by ensuring that those persons who facilitate distance sales of goods through the use of an electronic interface for the collection of VAT on those sales are the persons who are deemed to have made those sales when the seller resides outside the Community or the goods are imported and directly shipped to a consumer. For distance sales of goods imported from third countries to the Community, this is restricted to sales of goods which are dispatched or transported in consignments, with an intrinsic value not exceeding €150, for which a full customs declaration upon importation is required for customs purposes. At the same time, the exemption for the import of goods in small consignments with a negligible value of not more than €22 was removed. These measures create a level playing field for the businesses concerned and minimize their compliance burdens.

57.5 Improvements of the VAT System in the EU with Regard to Digital Transformation The EU VAT system provides for a neutral taxation of consumption within the EU if the principle of taxation at destination is consistently applied to cross-border supplies of goods and services. Major adaptations to secure taxation in the member state of destination have been implemented during the last decade as shown in the previous chapter. Whereas the place of supply of services to taxpayers is generally the place where the recipient is established, and is therefore in line with the principle of taxation at destination, for supplies to non-taxable persons the place of supply, as a general rule, continues to be the place where the service provider has established their business. However, importantly for the taxation of digital goods and services, the place of supply of electronically supplied services rendered to non-taxable persons is deemed to be at the place where the non-taxable person is established, or where they have their permanent address or usually reside. More and more services which were traditionally rendered locally are now offered globally via the internet. Therefore, EU law provides the tax authorities of the member states with the necessary tools to register suppliers and make them fulfil their obligations. What the Commission calls the ‘mini one-stop shop’ was originally

18 See

Council Directive (EU) 2017/2455 of 5 December 2017 amending Directive 2006/112/EC and Directive 2009/132/EC as regards certain value added tax obligations for supplies of services and distance sales of goods [2017] OJ L348/7; amended by Council Directive (EU) 2019/1995 of 21 November 2019 amending Directive 2006/112/EC as regards provisions relating to distance sales of goods and certain domestic supplies of goods [2019] OJ L310/1; postponed to 1 July 2021 by Council Decision (EU) 2020/ 1109 of 20 July 2020 amending Directives (EU) 2017/2455 and (EU) 2019/1995 as regards the dates of transposition and application in response to the COVID-19 pandemic [2020] OJ L244/3.

Digitalization and the Future of VAT in the European Union 1049 implemented as a special scheme for supplies of telecommunications, broadcasting, or electronic services but has since 1 July 2021 been extended to distance sales of goods making compliance easier for the operators. Supplies to non-taxable persons in the member states in which those suppliers are not registered can be declared via a web portal in the member state in which the supplier is identified. It is difficult for the tax authorities to administer a large number of suppliers and monitor their activities; however, a feasible solution has been found by involving the operators of the marketplaces which facilitate most of the online transactions and also handle the payments. According to article 9a of the VAT Regulation,19 where electronically supplied services are supplied through a telecommunications network, an interface, or a portal such as a marketplace for applications, a taxable person taking part in that supply will be presumed not to be acting in their own name but on behalf of the provider of those services, unless that provider is explicitly indicated as the supplier by that taxable person and that is reflected in the contractual arrangements between the parties. This ensures that the tax authorities have to deal with only a few operators of marketplaces rather than with many small businesses. Those marketplace operators generally have both the knowledge and the means to fulfil their obligations. Cross-border sales of goods directly to consumers have also increased through digitalization. Since the abolition of fiscal borders between the member states, distance sales to consumers pursuant to article 34 of the VAT Directive have been taxed in the member state in which the despatch or transport of the goods ends if the total value of such supplies to that member state exceeds €100,000 (or a lower threshold determined by the member states of not less than €35,000). From 1 July 2021, the place of supply of distance sales of goods is deemed to be the place where the goods are located at the time when dispatch or transport of the goods to the customer ends. The threshold has been lowered to a total of €10,000 per supplier according to article 59c of the VAT Directive as amended by Directive (EU) 2017/2455.20 Following the model of taxation of services pursuant to article 14a of the VAT Directive, where an operator of an electronic interface (e.g. a marketplace, platform, portal, or similar means) facilitates the supply of goods within the Community by a taxable person not established within the Community to a non-taxable person, the taxable person who facilitates the supply will be deemed to have received and supplied those goods themselves. In addition to this provision of the VAT Directive, effective from 1 July 2021, taxable persons who facilitate—through the use of an electronic interface such as a marketplace, platform, portal, or similar means—the supply of goods or services to a non-taxable person within the Community will be obliged to keep records of those

19

Council Implementing Regulation (EU) No. 282/2011 of 15 March 2011 laying down implementing measures for Directive 2006/112/EC on the common system of value added tax [2011] OJ L77/1. 20 Council Directive (EU) 2017/ 2455 of 5 December 2017 amending Directive 2006/112/EC and Directive 2009/132/EC as regards certain value added tax obligations for supplies of services and distance sales of goods [2017] OJ L348/7.

1050 Michael Tumpel supplies according to article 242a of the VAT Directive.21 Those records should be sufficiently detailed to enable the tax authorities of the member states where those supplies are taxable to verify that VAT has been accounted for correctly. The records must be made available electronically on request to the member states concerned and make it possible for them to monitor distance sales. Complementary to those regulations for monitoring the electronic commerce in order to keep track of transactions which are subject to VAT in a member state, starting in 2024 payment service providers will be required to keep sufficiently detailed records and to report certain cross-border payments determined as such by reason of the location of the payer and the location of the payee.22 This information will assist the member states to carry out controls effectively and to investigate suspected VAT fraud or to detect it. The European Commission has identified a remaining loophole for payments by crypto-assets and e-money. It opened a public consultation about a proposal for a Council Directive amending Directive 2011/16/EU23 for measures to strengthen the existing rules and expand the exchange-of-information framework in the field of taxation to include crypto-assets and e-money (DAC 8).24 It should be explored whether and to what extent the data collected could also be used for assessing the correct application of the VAT legislation. Especially in relation to transactions in anonymous, encrypted peer-to-peer networks, using crypto-currencies could circumvent all other measures of data collection and the obligations of taxpayers to combat VAT fraud. Over the last decade, the EU has implemented measures which consistently implement the principle of taxation at the member state of destination for electronically supplied services as well as supplies of goods purchased on the internet by consumers. At the same time, the EU has made it easier for suppliers to remit VAT and to fulfil their obligations in connection with VAT in the member states. Gathering information about transactions which are subject to VAT within the EU and monitoring such transactions has recently been a priority for the EU and must be implemented and evaluated in the next few years. Although there will always be loopholes for fraudsters to exploit the VAT system, it has been made gradually harder for them to succeed.

21 See Commission Implementing Regulation (EU) 2020/ 194 of 12 February 2020 laying down detailed rules for the application of Council Regulation (EU) No. 904/2010 as regards the special schemes for taxable persons supplying services to non-taxable persons, making distance sales of goods and certain domestic supplies of goods [2020] OJ L40/114. 22 See

Council Directive (EU) 2020/284 of 18 February 2020 amending Directive 2006/112/EC as regards introducing certain requirements for payment service providers [2020] OJ L62/7. 23 Council Directive 2011/ 16/EU of 15 February 2011 on administrative cooperation in the field of taxation and repealing Directive 77/799/EEC [2011] OJ L64/1. 24 See Proposal for a Council Directive amending Directive 2011/ 16/EU as regards measures to strengthen existing rules and expand the exchange of information framework in the field of taxation to include crypto-assets and e-money, Doc. Ares(2020)7030524 (2020).

Digitalization and the Future of VAT in the European Union 1051

57.6 VAT of Digital Services Without Consideration in Money Digital services are sometimes provided ‘free’ or without consideration in money but, obviously, those ‘free’ online services are monetized in other ways. By offering online services, suppliers harvest huge amounts of user data to extract and discover patterns that can be used for other applications. Those service suppliers often sell targeted advertisements to make money which, of course, can be subject to VAT in the EU if the recipient of those services is a business established in a member state. Those businesses are usually entitled to deduct input VAT so that the tax does not become a burden for the taxpayer and adds revenue to government coffers. From the perspective of a consumption tax, it seems reasonable to levy VAT on online services which are used by consumers (e.g. social media, search engines, web space). Although the scope of the VAT system in the EU is meant to be as broad as possible, it has some substantial limitations.25 Pursuant to article 2 of the VAT Directive, only the supply of goods or services for consideration within the territory of a member state by a taxable person acting as such will be subject to VAT. According to the settled case law of the Court of Justice of the European Union (CJEU): the possibility of classifying a supply of services as a ‘transaction for consideration’ requires only that there be a direct link between that supply and the consideration actually received by the taxable person. Such a direct link is established if there is a legal relationship between the provider of the service and the recipient pursuant to which there is reciprocal performance, the payment received by the provider of the service constituting the value actually given in return for the service supplied to the recipient.26

In many cases, the required legal relationship between the provider of the service and the recipient is established by accepting the terms of service during registration. However, it may not always be clear whether there is a legal relationship if the advertisement is presented on a website or as a search result. When there is a legal relationship and no payment in money is received for the service, it must be analysed whether there is payment in kind received by the provider of the service which constitutes the value actually given in return for the service supplied to the recipient. In many instances, users accept the usage of their personal data and relinquish their rights to the data by accepting the terms of service. However, there is

25 See M. Lamensch, ‘The Scope of the EU VAT System: Traditional & Digital Economy related Questions’, in M. Lang et al., eds, CJEU—Recent Developments in Value Added Tax 2017 (Vienna: Linde, 2018), 118. 26 CJEU: Case C‑432/15 Baštová, 10 November 2016, para. 28.

1052 Michael Tumpel usually no commitment to upload data (e.g. text or photos). Therefore, it seems questionable whether a direct link exists between the service supplied and the recipient’s data in return for the service. It can be deduced from the CJEU’s case law that the uncertain nature of the provision of any payment is such as to break the direct link between the service provided to the recipient and any payment which may be received.27 The consideration paid by the customer must neither be voluntary nor uncertain.28 Such uncertainty is typical for many of the online services which are provided without consideration in money because the provider does not know if and how much data will be received and whether its value can be expressed in monetary terms. It is clear from the settled case law that besides the direct link between the supply of services and the data collected, it must be possible to express the value of the latter in monetary terms.29 Moreover, the value of user data itself is very limited. Value creation happens through data mining, which requires several activities to transform the raw data into valuable knowledge.30 There is a controversy in the literature about whether online services which are supplied without consideration in money can be subject to VAT because personal and user-generated data are received by the supplier of those service.31 So far, the tax authorities seem to be reluctant to go down this route. However, it could be argued that consumption of such ‘free’ online services should be within the scope of VAT even if there is no direct link between the service received and a payment. The supply of services carried out free of charge by a taxable person can be treated as a supply of service for consideration under article 26 paragraph 1 lit. b of the VAT Directive if it is carried out for their private use or for that of their staff or, more generally, for purposes other than those of their business. Since the services are supplied for business purposes, the provision is not applicable. By contrast, the complementary rule of article 16 of the VAT Directive, which provides for the treatment of application by a taxable person of goods forming part of their business assets as a supply of goods for consideration, requires only their disposal free of charge. It is, in principle, immaterial whether or not the disposal was for business purposes.32 If article 26 paragraph 1 lit. b of the VAT Directive were changed to include a similar provision so that the supply of services free of charge were treated as a service for consideration, such online services would be subject to VAT.

27

See ibid., para. 29. See Case C-43/19 Vodafone Portugal, 11 June 2020, para. 44. 29 See Case C‑549/ 11 Orfey, 19 December 2012, para. 36; Case C‑283/ 12 Serebryannay vek, 26 September 2013, para. 38; Case C‑410/17 A Oy, 10 January 2019, para. 36. 30 See M. Olbert and C. Spengel, ‘Taxation in the Digital Economy: Recent Policy Developments and the Question of Value Creation’, ZEW Discussion Paper No. 19-010 (2019), 19. 31 See, e.g., S. Pfeiffer, ‘Comment on “Free” Internet services’, in Lang et al., CJEU—Recent Developments in Value Added Tax 2017, 134ff; G. Beretta, ‘VAT and the Sharing Economy’, World Tax Journal (2018), 417ff; M. Lamensch, ‘The Scope of the EU VAT System: Traditional & Digital Economy related Questions’, in ibid., 125ff; S. Pfeiffer, ‘VAT on “Free” Electronic Services?’, International VAT Monitor (2016), 158ff. 32 See Case C-48/97 Kuwait Petroleum, 27 April 1999, para. 22. 28

Digitalization and the Future of VAT in the European Union 1053 According to article 75 of the VAT Directive, the taxable amount of such services carried out free of charge could be the full cost to the taxable person of providing the services. This would guarantee an adequate taxation of consumption with regard to online services without payment. However, in those cases where money is paid but does not adequately reflect the value of the services provided to consumers, the taxable amount cannot be adjusted due to the CJEU’s case law.33 According to article 80 of the directive, an adjustment to ensure that the open-market value is the taxable amount to prevent tax evasion is only possible in respect of supplies involving family or other close personal ties, management, ownership, membership, or financial or legal ties as defined by the member state. In order to secure adequate taxation, the provisions of article 80 could be extended if a payment does not reflect the open-market value because the costs are covered by other activities related to the rendering of such services. Such a solution could, however, proof to be a slippery slope because it could possibly affect every supply of goods and services with a payment under the open-market value. For example, some shops offer rebates if customers use loyalty cards which makes it possible to track their shopping behaviour. Such schemes would probably also trigger a mark-up of the taxable amount. This can be seen as another example why ring-fencing digital services seems very difficult considering that digitalization (e.g. in the form of data-gathering from loyalty cards in regular shops) has percolated through the whole economy.

57.7 VAT and Consumption-Orientated Corporate Taxation of the Digital Economy Digitalization and globalization have not only posed a challenge for the VAT system but also challenged fundamental features of the international income tax system.34 The traditional allocation of taxing rights with respect to business profits ties it to a physical presence such as a permanent establishment. The digital economy does not rely on physical presence but rather offers its services globally via the internet. The international discussion drifted between finding a new definition for permanent establishments for digital businesses35 and the introduction of specific turnover taxes for digital 33 See Case C-412/03 Scandic,20 January 2005, paras 21 and 29 and the case law cited therein as well as Case C-40/09 Astra Zeneca, 29 July 2010, para. 28. 34 See para. 3 in OECD/ G20 Base Erosion and Profit Shifting Project, Tax Challenges Arising from Digitalisation—Report on Pillar One Blueprint: Inclusive Framework on BEPS, ( Paris: OECD Publishing, 2020). 35 See paras 3ff in OECD/ G20 Base Erosion and Profit Shifting Project, Preventing the Artificial Avoidance of Permanent Establishment Status, Action 7—2015 Final Report ( Paris: OECD Publishing, 2015); paras 267ff in OECD/G20 Base Erosion and Profit Shifting Project, Tax Challenges Arising from Digitalisation—Interim Report 2018: Inclusive Framework on BEPS ( Paris: OECD Publishing, 2018);

1054 Michael Tumpel services.36 At present, the OECD and the group that forms the Inclusive Framework is focused on new nexus and profit-allocation rules to ensure that the allocation of taxing rights with respect to business profits is no longer exclusively circumscribed by reference to physical presence as a Pillar One or a two-pillar global solution.37 It basically expands the taxing rights of market jurisdictions where the users are located38 and reserves to it a share of the residual profits of multinational enterprise groups if their revenues exceed certain thresholds. This requires rather complicated calculations as well as innovative dispute-prevention and dispute-resolution mechanisms between the tax authorities of the involved jurisdictions.39 The efforts of the Inclusive Framework to deliver a consensus-based solution for the reallocation of income tax revenue against the background of the digitalization of the economy seem moot. If implemented, it will set up a new profit-allocation system on top of the existing allocation of profits according to the transfer pricing system. This will shift income tax revenue to market countries and open the discussion for other changes to the international income tax system in favour of market countries. Until now, industrialized countries conjoined in the OECD were reluctant to grant taxing rights to developing countries even for royalties. New allocation rules which take into account market access must also address royalties but go far beyond them in scope. It must be kept in mind that, from an economic perspective, the answer to the question who has to carry the burden of taxation does not depend on a legal categorization as direct or indirect tax but rather on market forces. Consequently, collecting consumption taxes such as VAT by enforcing the destination principle may possibly be a much more sensible way to gain revenue from the digital economy than reforming the international income tax system.40 Jurisdictions in which digital businesses do not have a physical presence and therefore no nexus for income taxation are still able to levy consumption taxes. The EU has thus implemented provisions to ensure taxation in the member state where the consumer enjoys the goods and services as well as measures to enforce the destination principle. If existing loopholes, such as adequate taxation of digital services without payment or arm’s-length prices, can be closed, then fair compensation for the country which provides the consumer market will be achieved. Following

OECD, ‘OECD/G20 Inclusive Framework on BEPS: Progress Report July 2019–July 2020’ (2020), 20 and 49. 36 See paras 359ff in OECD/G20 Base Erosion and Profit Shifting Project, Tax Challenges Arising from Digitalisation—Interim Report 2018: Inclusive Framework on BEPS (Paris: OECD Publishing, 2018). 37 See para. 3 in OECD/ G20 Base Erosion and Profit Shifting Project, Tax Challenges Arising from Digitalisation—Report on Pillar One Blueprint: Inclusive Framework on BEPS (Paris: OECD Publishing, 2020). 38 See ibid., para. 6. 39 See paras 65ff in OECD/ G20 Inclusive Framework on BEPS, ‘Statement by the OECD/G20 Inclusive Framework on BEPS on the Two-Pillar Approach to Address the Tax Challenges Arising from the Digitalisation of the Economy’ (Jan. 2020). 40 See M. Olbert and C. Spengel, ‘Taxation in the Digital Economy: Recent Policy Developments and the Question of Value Creation’, ZEW Discussion Paper No. 19-010 (2019), 4.

Digitalization and the Future of VAT in the European Union 1055 Olberts and Spengel’s recommendation, researchers and policymakers should be encouraged to focus more on the role of consumption taxes, such as VAT in the EU and its role in the digital economy and its contribution to collecting a fair share of revenue in market countries.41

57.8 Can Digitalization Resolve VAT Fraud and Simplify VAT Collection? Digitalization and globalization increased the number and the volume of cross-border transactions. To monitor cross-border supplies of goods and services, tax authorities rely on close cooperation. Council Regulation (EU) No. 904/201042 provides a framework for administrative cooperation between the tax authorities of the member states in the field of VAT and key to this cooperation are electronic storage and transmission of certain data for VAT control purposes which allow for fast information exchange and automated access to information. In order to achieve these goals, the EU has established electronic databases on taxable persons and their intra-Community transactions. Every taxable person identified for VAT purposes must submit a recapitulative statement of intra-Community supplies of goods and services electronically pursuant to article 262 of the VAT Directive. The exemption for intra-Community supplies depends on compliance with the obligation to submit such statements according to article 138 paragraph 1a of the directive. During the last decade, the EU has established and expanded the one-stop shop scheme. Initially the one-stop shop scheme was designed to obtain a refund of VAT if the taxable person was not established in the member state of refund. Taxpayers use an electronic refund application and submit it to the member state in which they are established via the electronic portal set up by that member state. The scheme has been expanded to taxable persons supplying telecommunication services, television and radio broadcasting services, and electronically supplied services to non-taxable persons in member states in which they do not have an establishment in order account for the VAT due on those supplies via a web portal in the member state in which they are identified. As from 1 July 2021, it has also been used for distance sales. Therefore, digitalization makes it easier for taxpayers to comply with the obligations set forth out in the EU VAT system. It also helps to lower the administrative burden for businesses as well as tax authorities. Without the one-stop shop, suppliers would be required to register in each member state in which they supplied goods or services to their non-taxable customers. 41

See ibid.

42 Council

Regulation (EU) No. 904/2010 of 7 October 2010 on administrative cooperation and combating fraud in the field of value added tax [2020] OJ L268/1.

1056 Michael Tumpel Even though digitalization helps to collect, store, and exchange data about taxable persons and their transactions electronically, VAT fraud remains a problem in the EU. Especially in its fight against carousel or missing-trader fraud, the EU relies heavily on exchange of information as well as joint audits.43 An effective solution to promptly tackle the problem of missing-trader fraud in intra-Community trade would be to allow member states to provide that the person liable for payment of VAT on supplies listed in that article is the taxable person to whom the supply is made (the ‘reverse-charge mechanism’). However, the optional reverse-charge mechanism is limited to supplies of certain goods and services susceptible to fraud.44 Only member states that fulfil certain criteria regarding their level of fraud, in particular in relation to carousel fraud, are allowed to implement a temporary generalized reverse-charge mechanism (GRCM) with a certain threshold per transaction.45 Member states using the GRCM must introduce specific electronic reporting obligations on taxable persons to ensure the effective functioning and monitoring of the application of the GRCM. The implementation of an electronic reporting obligation to match transactions between businesses has been made possible by digitalization and would be a sensible way to reduce the risk of fraud. The European Commission takes a different approach by proposing a definitive VAT system for intra-EU trade based on the principle of taxation in the member state of destination of the supply of goods and services.46 Under this concept, the place of supply (even to other taxable persons) will be situated in the member state of arrival of the goods where the taxable person is liable for VAT. Where persons liable for VAT are not established in the member state where the tax is due, they will be able to settle a declaration, payment obligations, and deduction of input VAT via the one-stop shop system. The recapitulative statement and the exchange of information should be abolished. However, taxpayers would be forced to collect information about the VAT provisions, guidelines, and case law in twenty-seven jurisdictions if they wish to do business in all other member states. Multinational enterprises (MNEs) would have a definitive advantage because they already operate in many jurisdictions and have tax advisers in many member states. By contrast, it would be very difficult and costly for

43 See Council Regulation (EU) 2018/1541 of 2 October 2018 amending Regulations (EU) No. 904/ 2010 and (EU) 2017/2454 as regards measures to strengthen administrative cooperation in the field of value added tax [2018] OJ L259/1. 44 See Council Directive (EU) 2018/1695 of 6 November 2018 amending Directive 2006/112/EC on the common system of value added tax as regards the period of application of the optional reverse charge mechanism in relation to supplies of certain goods and services susceptible to fraud and of the Quick Reaction Mechanism against VAT fraud [2018] OJ L282/5. 45 See Council Directive (EU) 2018/ 2057 of 20 December 2018 amending Directive 2006/112/ EC on the common system of value added tax as regards the temporary application of a generalised reverse charge mechanism in relation to supplies of goods and services above a certain threshold [2018] OJ L329/3. 46 See Proposal for a Council Directive amending Directive 2006/ 112/EC as regards harmonising and simplifying certain rules in the value added tax system and introducing the definitive system for the taxation of trade between Member States, COM (2017) 569 final.

Digitalization and the Future of VAT in the European Union 1057 small and medium-sized enterprises to seek such information and representation in other member states. For them, it would not be a simplification.

57.9 Conclusion The future of VAT in the EU regarding the results of digitalization is already present. EU VAT law has already been adopted for the principle of taxation at destination for supplies of goods and services to non-taxable persons. This secures the principle of neutrality and fair competition even in a world where digitalization and globalization have led to an increase in cross-border transactions directly with consumers. However, a solution should be found for taxing ‘free’ online services to make sure that the digital businesses pay their fair share of tax. Digitalization has made it easier to electronically monitor cross-border transactions systematically. It also makes if more simple for businesses to fulfil their VAT obligations regarding cross-border trade.

Index

For the benefit of digital users, indexed terms that span two pages (e.g., 52–53) may, on occasion, appear on only one of those pages. Tables and figures are indicated by t and f following the page number ‘365 day’ period test 407–8, 409, 414–15 ability-to-pay principle 22–23, 24, 48, 278–79, 1035–36 abusive practices prevention agents 511–12 beneficial ownership 379, 380–81, 385–86 Court of Justice of the EU (CJEU) 602–4 EEA/EFTA 652 EU anti-tax-avoidance regimes 679–80 fundamental freedoms (EU) 625 national tax systems 103 principal purpose test (PPT) 407–8 accession agreements 596 accountability 1013 Achmea 721–23 acquiescence 59–60 acquis 596, 663–64, 666, 667, 672–73, 675, 676, 678 active business test 401n.56, 410 active income 26–27, 144–45, 149, 150, 155–56, 321 Adams, S. 10–11 adapting to the times 13–17 Addis Ababa Action Agenda 127, 129, 136 administrability 961, 963, 964–65 administrative agreements and arbitration 61–62 administrative capacity of tax authorities 671–74 administrative cooperation 180, 594, 694–95 administrative law 266 Administrative Procedure Code (APC) 663–64 administrative rules 48, 108

administrator 375–76 advance pricing agreements/arrangements (APAs) 548 alternative dispute resolution in EU 711 corresponding adjustments and transfer pricing 568 exchange of information and cooperation 53–54 state aid (EU) 638, 639, 640, 641, 643 transfer pricing and formulary apportionment 583–84 transfer pricing for MNEs 550–51 see also bilateral advance pricing agreements (BAPAs); multilateral advance pricing agreements (MAPAs); unilateral advance pricing agreements (UAPAs) Advance Pricing Agreements within the EU Guidelines 707 advance rulings 699, 773–74 advantage element 632 Advisory Body (AB) 217–18, 219–20, 226–27 Advisory Commission (AC) 708, 709–10, 714–19, 720, 722, 723–24, 725, 726–27 Africa 125, 127–28, 1019 regional double tax treaties 459–60, 461, 463, 465–66, 468–69 see also East African Community (EAC) as digitalized economy African Tax Administration Forum (ATAF) 97–98, 127–28, 129, 136–37, 457, 459–60, 488–89, 907, 908, 915, 1005, 1017

1060 Index agents in tax treaties 370, 373, 375, 497–514 dependent agents 497–98, 500–1, 510– 11, 514 independent agents 497–98, 500–1, 508 model tax treaties 498–503 OECD BEPS Report (2013) 505–10 commissionaire arrangements 503– 5, 507–10 profits attribution to permanent establishments 510–13 change implementation 511–12 digitalization of the economy 513 aggregate approach 436–37 aggressive tax planning schemes 53–54 Agreement on Subsidies and Countervailing Duties (SCM) 213 AI see labour taxation and automation/AI Albania 645, 663–64, 678 allocation conflicts 130–32, 291, 296, 440–41, 442–43, 445, 447 allocation of taxing rights 615 Alta Energy case 410–15 Alternative Dispute Resolution Commission (ADRC) 720, 723, 726–27 alternative dispute resolution in EU 697–727 advance pricing agreements (APAs) 711 complaint 712 cross-border disputes 698–705 bilateral investment treaties 700–1 double taxation conventions 702–5 evaluation and negotiation stage 711 formal application 711 prefiling stage 711 see also Arbitration Convention (AC); Tax Dispute Resolution Directive Amazon 154, 484 American Enterprise Institute 813–14 Amount A see Pillar One Amount B see Pillar One Andean Community 457 Andean Pact countries 485 anti-abuse provisions beneficial ownership 383, 386–89 charities 342 domestic law 48 double taxation 236 EU law 51–52

multilateralism 491 regional double tax treaties 462, 463, 469–70 transfer pricing 526, 537 see also base erosion and anti-abuse tax (BEAT) anti-avoidance provisions 31, 51–52, 117, 118, 527, 622–27, 834–35, 988 see also general anti-avoidance rule (GAAR); specific anti-avoidance rule (SAAR) Antibolus, P. 17–18 anti-competitive collusion 31 anti-fragmentation rules 131 anti-inversion rules 155 anti-money laundering information 355 Anti-Tax Avoidance Directive (ATAD) 452– 53, 686–95 BEPS 1.0 151 comparative tax law 280 controlled foreign corporation (CFC) rule 687–88 corporate law 200–1, 208 Court of Justice of the EU (CJEU) 594, 598–99, 602, 603–6 double taxation avoidance 324–25, 331–32 EEA/EFTA 656–57 EU alternative dispute resolution 707, 712 EU anti-tax-avoidance regimes 684, 685–86 exit tax rule 688–89 general anti-avoidance rule (GAAR) 689–90 implementation challenges 690–94 discretionary powers of member states 692–93 legislative amendments 691–92 third states 693–94 interest limitation 689 language factors 263 mismatch rules 690 partnerships 452 Anti-Tax Avoidance Directive (ATAD) I 51– 52, 581, 681–82, 686 Anti-Tax Avoidance Directive (ATAD) II 51– 52, 201, 201n.22, 452, 581, 681–82, 686 Anti-Tax Avoidance Directive (ATAD) III 681–82 Anti-Tax Avoidance Package (ATAP) 151, 152–53

Index 1061 anti-tax avoidance regimes see European Union anti-tax avoidance regimes AOL/TimeWarner 484 Apple 148, 154, 641–42, 804 apportionment criteria 617, 668–69 see also formulary apportionment; fractional apportionment arbitration clause 568 Arbitration Convention (AC) 707 exclusion 709 procedure 708 scope 708–9 taxpayers’ rights 709–10 arbitration proceedings 52, 523, 526, 562, 606, 712 see also mandatory arbitration Argentina 206, 226–27, 994 arm’s length principle (ALP) 824 1920s compromise and the 2020s compromise 21, 24–25, 30–31, 37, 38, 39–40, 41, 42, 43 agents 513 Anti-Tax Avoidance Directive (ATAD), 688–89 BEPS 2.0 153–54 consensus international tax law (CITL) 98–99 corresponding adjustments and transfer pricing 557–59, 560–61, 563, 567, 569–70 customary international law 80, 82–83 emerging tax consensus challenges 995, 997–1001 EU alternative dispute resolution 706, 708–9 EU anti-tax-avoidance regimes 684–85 EU/OECD VAT/GST comparison 796 international tax regime (ITR) 156 international trade law 223–24 low-and middle-income countries (LMICs) 126–27, 132–33, 141 OECD Transfer Pricing Guidelines 538–41 Pillar One 951, 952–53, 960–61 public international law 181, 182–83 state aid (EU) 637, 639, 640, 641, 643 transfer pricing and formulary apportionment 572–73, 574–76, 577, 578–80, 582, 583, 584–86, 587

transfer pricing for MNEs (OECD) 536–37, 538, 541, 543–44, 545, 546–49, 550–51 transfer pricing (OECD MC) 518, 519–20, 522, 523, 528, 529n.46, 530, 531–32, 533 VAT Directive and imports 756–57 arm’s length standard see arm’s length principle (ALP) Arnold, B. 912 artificial intelligence 550 see also labour taxation and automation/AI Artinger, K. 775, 777–78 Asakawa, M. 919n.6 Asia 125, 549, 552, 1019 asset-based ratio 200 associated enterprises 461–62, 520–27, 531 Association of South East Asian Nations (ASEAN) 477, 486 Astuti, M.D. 909 attribution of profits 510–14, 858 audits 579–80, 582, 583–84, 673–74 joint 363, 987–89, 1056 Augustus 8 Australia 100, 907 183-day test (bright-line test) 118–19 ceasing residency rule 119 charities 337 commencing residency rule 118–19 corporate law 207n.52 factor test 118–19 government official test 119 Offshore Banking Unit 488n.81 overseas employment rule 119 personal nexus 116n.38, 117n.45, 118, 119, 120n.57 principal purpose test (PPT) 395n.19, 406 transfer pricing 550–51 Austria 221, 387–88 Anti-Tax Avoidance Directive (ATAD) 691–92 Court of Justice of the EU (CJEU) 607 EEA/EFTA 645, 650 EU candidate countries 668 human rights law 419 multilateralism 493 partnerships 435 transfer pricing for MNEs (OECD) 552–53 VAT Directive and imports 761

1062 Index Authorized OECD Approach (AOA) 24–25 automated digital services (ADS) 38, 576–77, 960–61, 997–98, 1004 Automatic Exchange of Information (AEOI) regime 128–29, 140–41, 145–47, 1005 exchange of information 353, 354, 355, 363 global tax governance 1011, 1014–15 automation see labour taxation and automation/AI automation taxes 1034–35, 1036–37 autonomous concepts 171–72 autonomy of parties and choice of law 161–64 Avery Jones, J.F. 58, 183, 210, 288n.9, 288n.10, 294 Avi-Yonah, R.S. 73, 82, 93, 104, 181, 475–76, 581–82n.44, 817–18, 905, 992, 999–1000 Báez, A. 1000 Baistrocchi, E. 102–3 Baker, P. 191 balance of powers 262–63 balance-sheet and tax allocation rules 356–57 Balharová, M. 1020–21 bank secrecy 53, 140, 145–46, 369–70, 488, 491–92 Barbados 343, 410–11n.100, 505–6 Barker 275n.64 base erosion and anti-abuse tax (BEAT) 151–54 base erosion and profit shifting (BEPS) 15–16, 27, 45–46, 129–30 beggar-thy-neighbour approach 953 Belgium 116n.38, 668 Anti-Tax Avoidance Directive (ATAD) 691 coordination centres regime 637–38, 639– 40, 642 Court of Justice of the EU (CJEU) 604–5 exchange of information 368–69 fundamental freedoms (EU) 621–22 partnerships 435 state aid (EU) 636–38 tax ruling scheme 637–38 tax treaties 49 beneficial ownership 369–70, 402, 444 beneficial ownership limitation 394– 95, 404–5 beneficial ownership and tax treaties 373–89

anti-abuse mechanisms 386–88 domestic jurisprudence 381–86 EU and Court of Justice of the EU (CJEU) view 378–81 OECD MC 375–78 benefits 51 entitlement 467 sharing 827 see also benefits principle (BP); limitation on benefits (LOB) benefits principle (BP) 121n.58, 567 1920s compromise to 2020s compromise 24, 26–27 BEPS 2.0 154 Pillar One 961–62 single-tax principle 144–45 source taxation rights 238–39 unilateralism, bilateralism and multilateralism 476 benefit theory 111, 112 Benshalom, I. 581–82n.44 Bentley, D. 94–95 BEPS 1.0 148–51 and Tax Cuts and Jobs Act (TCJA) (USA) 151–53 BEPS 2.0 153–55 BEPS Action 2 Report 445–47 limitation and criticisms 446–47 BEPS Associates 1015–16 BEPS Project 584–86, 930–32, 977–78 BEPS Report (2013) 505–10 commissionaire arrangements 507–10 Berkholz 335–737, 743 Bermuda 505–6, 805 best interests 181 best practice 100, 124 Biden, J. 965–66, 985 bilateral adjustment of profits clauses 568 bilateral advance pricing agreements (BAPAs) 551, 552–53 bilateral agreements/bilateralism 17–18, 45– 46, 829–30 1920s compromise to 2020s compromise 23, 24–25, 26 agents 497–98, 501, 502–3, 511–12, 514 alternative dispute resolution in EU 699–701

Index 1063 BEPS 1.0 150 charities 336, 339–40, 343 consensus international tax law (CITL) 99–100 Court of Justice of the EU (CJEU) 592– 93, 594 customary international law 74, 78, 79, 80–81, 82, 86 double taxation treaties and language factors 251, 255 EEA/EFTA 656, 660 emerging tax consensus challenges 997– 98, 1004 EU ‘white supplies’, double taxation and VAT 768, 776 exchange of information 357 fundamental freedoms (EU) 609, 611, 616–17 global tax governance 1010, 1014–15 human rights law 417–18, 428–29 jurisdiction 65, 66, 68, 69–7 1 low-and middle-income countries (LMICs) 127–29, 140–41 national tax systems 87–88, 90, 92–93, 94, 103 personal nexus 109–10 public international law 180, 181, 192 regional double tax treaties 468, 471 source taxation rights 237 tax treaties interpretation 55–56 transfer pricing for MNEs (OECD) 553 triangular cases 301–2, 317 see also unilateralism, bilateralism and multilateralism bilateral tax rate 834 binding effect 595, 596–97 binding tariff information (BTI) decisions 780–81n.53, 784 blended profits 543 Bonbright, J.C. 941–42 border tax adjustments (BTAs) 222–23, 224, 228–29 Bosnia and Herzegovina 645, 663–64, 678 Botswana 1019 Braumann, C. 83, 93 Brauner, Y. 97, 98, 99–100, 490, 491, 911–12, 1000, 1008, 1012–13

Brazil 123, 127, 863–77, 994 anti-abuse rules 864, 867–68 anti-avoidance rules 876 anti-deferral rules 876 base erosion avoidance 866–67 BEPS Inclusive Framework and MLI 874–75 Constitution 868 controlled foreign corporation (CFC) rules 865–66, 876 corporate law 201n.22, 203, 206, 207n.51, 208, 209–10 country-by-country reporting system (CbC) 874–75 digital economy 867 dispute resolution mechanism 874–75 domestic law 864–69, 871–72, 875, 876–77 non-residents taxation 866–68 residents taxation 866 double non-taxation avoidance 876 double tax conventions 864, 867, 868, 869, 870–7 1, 874–76 and deviation from OECD and UN Models 871–74 exchange of information 875 fees for technical services (FTS) and royalties 864, 868–69, 870–7 1, 872– 73, 875–76 general anti-avoidance rules (GAAR) 865–66 independent personal services 864, 868, 871–72, 873–74 international trade law 224 lex specialis 868 limitation-on-benefits (LOB) clauses 870, 875 mandatory disclosure rules 866 minimum standards 875 mutual agreement procedure (MAP) 875 National Tax Code 865, 868 non-residents taxation 866–68, 871–72 permanent establishment (PE) 864, 871–72, 873, 875–76 Pillar One 874, 876, 966 resident-based taxation 866 specific anti-avoidance rules (SAAR) 865–66

1064 Index Brazil (cont.) tax havens 864, 867–68, 870 tax-sparing clauses 869, 870 tax treaty network 869–7 1 territorial system 865, 870 thin capitalization rules 866 transfer pricing 866 transparency 870 UN Model Double Taxation Convention 875–76 British Virgin Islands 505–6 broadcasting services 130–31, 792, 1046– 47, 1055 Bruins, Professor 22 Bulgaria 663–64, 672–78, 754–55 business agreement 792–93, 794–95 business connection 857 business operations versus preparatory auxiliary activities 857 business profits 302–3, 461 business purpose test 206 business restructuring 543 business-to-business (B2B) 482–83, 733–34, 737, 766, 768, 782, 783, 789, 791, 795, 796 business-to-consumer (B2C) 482–83, 733–34, 737, 768, 777, 782, 792, 793–94, 795, 796 Cameroon 1019 Canada 113n.27, 125 beneficial ownership 382 domestic law 47 international trade law 215n.7, 224 principal purpose test (PPT) 392–93, 395n.19, 403–4, 410–11, 412, 413–14 sojourning 116 transfer pricing and formulary apportionment 571, 573n.7, 574 transfer pricing for MNEs (OECD) 546–47, 550–51 value creation 942–43n.14 candidate countries see EU candidate countries capacity building 1019 Čapek, K. 1028–29 capital export neutrality (CEN) 241, 242, 321–22, 618

capital gains 302–3, 363 corporate law 205–8, 212 principal purpose test (PPT) 394–95, 399– 400, 411, 414–15 regional double tax treaties 464 capital import neutrality (CIN) 241, 242, 243, 321–22, 618 capital taxation 466 carbon tax 219–20, 228–29 Caribbean Community (CARICOM) 485 Carroll, M.B. 24–25, 154, 536 Carroll report 30–31, 536–37 carve-outs 192 cashboxes 149 cash pooling 543 Cayman Islands 146–47 certainty 789–90, 793–94, 795, 995–96, 1035–36 see also legal certainty chargeable event and chargeability of VAT 752–53 charges with equivalent effect 214 charities in tax conventions 335–52 OECD MC 339–44 donors 343 double taxation conventions (DTCs): residency 339–41 limitations on benefits (LOB) clause 342 no mutual recognition and non-discrimination 342–43 special tax regimes 341 OECD Model Convention on Taxes on Estates, Inheritances and Gifts (EIGMC) 348–51 commentary 348–49 gift and inheritance treaties (examples) 349–51 tax incentives to resident charities 337–39 UN MC 344 US Estate and Gift Model Convention (US EGMC) 351–52 US MC 345–48 donors 347–48 limitation on benefits (LOB) 346 non-discrimination and reciprocal exemption 346 residency 345

Index 1065 special tax regimes 345–46 welfare purposes 351 check-the-box mechanism (USA) 154, 209, 435 China 127, 153, 324, 544–45n.51, 550–51, 994 corporate law 203, 206, 208, 209 see also China as digitalized economy China as digitalized economy 821–40 anti-tax avoidance 837–38 arm’s length principle (ALP) 838 Base Exploration and Profit Sharing (Neo- BEPS scheme) 838, 840 Belt and Road Initiative 837–38, 840 Centurial Change 822–23, 836, 840 cooperation 821–22 fairness 821–22, 824, 839, 840 flexibility 839 gaming in establishment of market jurisdiction taxation rights 830–31 gaming in reform proposals 829–30 income tax 838–39 independent transactions principle 838 institutional challenges and theoretical exploration 822–29 institutional challenges and theoretical exploration, associated changes 822–24 market jurisdiction taxation rights 824–25 multilateralism 836, 837 multi-polarization 836–37 neutrality principle 839 objectivity 831 permanent establishment nexus 838 Pillar One: limited fairness of ring-fencing reform structure 821, 832–34 Pillar Two: formal fairness of design of ‘one size fits all’ rule 821, 834–35 preliminary basis for taxation reform 825–26 Principle of Personal Existence 838 recognition and positioning of strategic role 836–37 security 821–22 subjectivity 831 tax base erosion and profit-shifting 837, 839 tax base exploration and design of policy system 837–39

tax base mismatch 839 tax base security, fairness and development 837–38 tax competition 836–37, 838 tax havens 839 theoretical basis for market jurisdiction taxation rights 826–29 transparency 837–38 two pillar solution 830 Chrétien, M. 5–6 Christensen, J. 1018 Christians, A. 82, 84–85, 961, 1000, 1008, 1011, 1015–16 citizenship 111, 112–13 civil law/civil law countries agents 498, 500–1, 503–4 beneficial ownership 374–75, 379 comparative tax law 267–68n.17, 270–7 1, 272 corporate law 199, 202, 209 labour taxation and automation/ AI 1026–27 language factors 250 personal nexus 114–15, 121, 122 private international law 165, 167, 169–70 qualification conflicts 287 classification conflicts 436, 437–38, 440–41, 447, 450, 452 Clausing, K.A. 801, 803, 808, 810, 811, 813–14, 817–18, 999–1000 client-attorney privilege 368 Code of Conduct for Business Taxation (CoC) 633–35, 639, 667, 671–72, 682– 83, 694–95 Code of Conduct on Transfer Pricing Documentation (EU TPD) 549–50 Code of Conduct Working Group 633, 634, 643 coherence 621–22 collection of taxes, assistance in 467 collectives (or bulk) requests 358–59 collision-resolving regulations 376–77 Colombia 1012 comitology committee 779 comity 66, 83 commensurate with income standard 537– 38, 547–48

1066 Index commercial law 15–16, 199 commissionaire arrangements 131, 150, 498, 502–5, 507–12, 513–14 Committee on Dispute Resolution 726 Committee of Experts on International Cooperation in Tax Matters 502–3 Committee on Fiscal Affairs 101, 127–28, 194, 210, 293, 441, 480–81, 786 ‘Double Taxation Conventions and the Use of Conduit Companies’ 375–76 Committee of Technical Experts 22 commodity value formula 827–28 Common Consolidated Corporate Tax Base (CCCTB) 484–85, 571–73, 575, 576, 581, 584, 585, 598–99, 668–69 common core comparison 276–77 common law/common law countries 101 agents 498, 500–1, 503–4 beneficial ownership 373, 381, 382–83 comparative tax law 272 corporate law 199, 202 domestic law 58 language factors 263–64 national tax systems 104–5 personal nexus 114–16, 120, 121–22 qualification conflicts 287 Common Reporting Standard (CRS) 53–54, 128–29, 147 BEPS 2.0 155 exchange of information 354 global tax governance 1014–15 human rights law 426–27 international tax regime (ITR) 155 Multilateral Competent Authority Agreement 1014–15 multilateralism 486–87 single-tax principle 145 comparability analysis 539, 540t comparables 538, 539, 542 comparable uncontrolled price (CUP) method 537, 541–42, 546–47, 548– 49, 640 comparable uncontrolled transactions 539, 542 comparative tax law 12–13, 265–81 aim of: reasons for comparing 267 circulation of models 280

object of: what to compare 267–69 procedures 269–79 analysis 275–77 family taxation 277–79 foreign law, study of 274–75 method 270–72 micro and macro-comparisons 272–73 steps of comparison 273–74 compensation 827 stock-based 547–48 competences 593–94, 603, 664–65, 673– 74, 787 competent authorities 354, 361–62, 714–17 competition 661 fair 43, 51, 1057 rules 647 see also distortion of competition; tax competition complaints 713–15 compliance assurance programme 988–89 compulsory arbitration see mandatory arbitration conduit companies 373, 375–76, 380, 382–83, 386, 388n.93, 389, 402, 407–9 conferral of powers principle 592, 593–94 confidentiality 361–62, 367, 425–26 see also disclosure requirements; secrecy conflict clauses 192 conflict-of-laws rules 165–66, 592–93 conflict resolution see dispute resolution connecting factors 111, 119–20, 121, 161–62, 299–301, 302, 558–59, 857 see also genuine connection; sufficient connection consensus international tax law (CITL) 88, 94–104 consensus minus one rule 1016 consensus requirement 488, 961, 1005 constitutional law 265–66, 380–81 constructive or reinstatement effect 1029–30 consumer-facing businesses (CFB) 34, 960–61 see also business-to-consumer (B2C) consumption tax 828–29 destination-based 961–62 digitalization and VAT in EU 1044, 1051, 1054–55

Index 1067 EU ‘white supplies’, double taxation and VAT 765, 766 international tax regime (ITR) 143, 144 VAT Directive and imports 763 context 56, 522, 524, 525, 527 continuity requirement, breach of 964 control 522, 523–25, 526, 527 common 521 and dominion 382–83 financial 672–74 controlled foreign corporation/company (CFC) rules 1920s compromise to 2020s compromise 31, 37, 42 Anti-Tax Avoidance Directive (ATAD) 687–88, 692–94 BEPS 1.0 149, 151 domestic law 48 double taxation avoidance 324–25, 331 EU anti-tax-avoidance regimes 681–82 EU law 51–52 international tax regime (ITR) 155 jurisdiction 67 Pillar Two 982, 983, 986–87 private international law 167 Tax Cuts and Jobs Act (TCJA) (USA) and BEPS 1.0 151–53 transfer pricing and formulary apportionment 573 transfer pricing (OECD MC) 526 controlled transactions 523, 546–47 cooperation 45 1920s compromise to 2020s compromise 32 consensus international tax law (CITL) 94 digitalization and VAT in EU 1055 enhanced 668–69 EU candidate countries 671–74 global tax governance 1008–09 low-and middle-income countries (LMICs) 127–29, 141 mutual 125 national tax systems 88 public international law 186–87, 192 regional 475, 1019 Cooper, M. 801 Copenhagen criteria 596, 663–64, 672–74 corporate income tax (CIT)

1920s compromise to 2020s compromise 26–27 agents 505–6 BEPS 1.0 148 charities 335–36, 339, 341, 343–44 corresponding adjustments and transfer pricing 569 EEA/EFTA 654–55 EU ‘white supplies’, double taxation and VAT 780–81 international tax regime (ITR) 144, 155 labour taxation and automation/AI 1025– 26, 1037–38 partnerships 434, 435 single-tax principle 145 state aid (EU) 639 transfer pricing and formulary apportionment 585–86 corporate law 197–212, 601 beneficial ownership 385 dividends 202–4 entity types and liability for tax 209–11 equity and debt 198–201 restructurings 205–9 capital gains 205–8 exit taxes 208–9 corporate profits 487, 800f corporate tax 45 Anti-Tax Avoidance Directive (ATAD) 692 digitalization and VAT in EU 1053–55 double taxation avoidance 234, 329–31 EU candidate countries 667–69 EU/OECD VAT/GST comparison 793 language factors 263 Pillar One 964–65 Pillar Two 981, 986 regionalism 484–85 residence taxation 242 tax competition 246 transfer pricing and formulary apportionment 571, 575, 577–78 United States 802f see also Common Consolidated Corporate Tax Base (CCCTB); corporate income tax (CIT) corrective taxes 1034–35

1068 Index corresponding adjustments 555–70 EU Arbitration Convention 568 method 565–66 mutual agreement procedure (MAP) 566–68 regulation 556–65 assessment by other contracting state 560–62 legal basis 562–65 secondary correction 568–70 corruption 1002 Costa Rica 483, 1012 cost-benefit analysis 277 cost contribution arrangements (CCAs) 543, 545 cost-plus method (CPM) 537, 541, 542, 548– 49, 638 cost-sharing agreement 152, 537–38, 547 cost theory 112 country-by-country reporting (CbCR) agents 511–12 BEPS 1.0 150 exchange of information 355 low-and middle-income countries (LMICs) 137, 140 multilateralism 487, 488–89, 491–92 Pillar Two 982, 988–89 transfer pricing for MNEs (OECD) 536, 538, 544, 549, 550 country-of-origin principle 766–67, 770, 792 Court of Justice of the EU (CJEU) 47, 591–608 assessment 607–8 beneficial ownership and tax treaties 378– 81, 386, 387, 388 charities 349, 352 corporate law 208 EU law 50–51 exchange of information 371n.110 extra-EU: integration drive and disengagement 596–97 fundamental freedoms (EU) 611 jurisdiction extension and abuse of rights prevention 599–601 language factors 262 legal binding force 592–95 legal binding force, intra-EU: effective legal protection and integration 594–95 legitimation for interpretation of law 601–3

monopoly of interpretation 604–7 Directive law 604–6 double tax arrangements (DTAs) 606–7 norm-setting 603–4 partnerships 450 principal purpose test (PPT) 392–93 private international law 162–64, 174–75 public international law 187 regionalism 484 standardization 597–99 state aid (EU) 629, 630 see also Court of Justice of the EU (CJEU) and fixed establishment Court of Justice of the EU (CJEU) and fixed establishment 734–46 digitalization of society 744–46 head office (HO) and permanent establishment (PE) 735–37 new challenges 742–46 parent-subsidiary scenario 738–40 VAT group articulations 740–42 court, right to 422 Cover Statement (2020) 34–35 Covid-19 pandemic 11–12, 34–35, 39, 41, 156, 536, 633, 1000, 1020–21 Crawford, J. 74–75, 186n.44 credit absorption 109–10 full 321–22 maximum 321 ordinary 321–22 rating 543 triangular cases 304 see also credit method credit method 223–24, 321, 322–23, 329–30, 333–34, 377, 442–43, 447–49, 617–20 criminal activity see illicit/illegal behaviour criminal law 266, 364 criminal procedures 359 crypto-assets and crypto-currencies 745, 1043, 1050 cryptotypes 270, 274 Cui, W. 490, 899 cumulative marginal tax rate 233n.10 custom 179–81 customary international law (CIL) 26, 73–86, 181–83, 184, 189, 190, 192–93 constituent elements 74–78

Index 1069 objective element (general practice) 75, 76, 77, 80, 82 opinio juris 75, 77–78, 82–83, 84, 85–86 opinio necessitatis 75, 84 subjective or psychological element 75, 77–78, 85–86 corporate law 85–86, 197 evidence 78–85 tax treaty network 80–83 international organizations, role of 83–85 jurisdiction 65, 67–7 1 national tax systems 87–88, 91–94 qualification conflicts 291 tax treaties interpretation 55 customary rule 93–94 custom primacy thesis 74–75, 80–81 customs duties 214, 646, 673–74, 747 Customs Valuation Agreement 748, 763 Czech Republic 668, 676 Dagan, T. 479, 491, 912 Daniels, T. 447–49 Danske Bank 741, 743 Darwin, C. 249 data mining 1051–52 data protection 425–26, 1042 data sharing 488 d’Avout, L. 173–74 days-away method (New Zealand) 116 Deblauwe, R. 348 debt-equity ratio 199–200 deduction, right of 759–61 deduction without inclusion (D/NI) 452 deemed tax added back rule (DTABR) 834–35 defensive rules 452 deferred taxes 982 definition of international tax law 89–91 Deloitte 763 de minimis rule 30, 649–50, 689 democratic taxation 272–73 Denmark 11–12 beneficial ownership 379–80, 384, 388, 679–80, 685–86, 694–95 Court of Justice of the EU (CJEU) 600–1 EU alternative dispute resolution 723 EU and cross-border VAT 741–42 human rights law 420 partnerships 436

principal purpose test (PPT) 396–97 regionalism 484–85 depreciation or amortization rules 649 Déry, J.-M. 58 destination-based cash flow tax (DBCFT) 224, 228–29, 1000–01 destination-based consumption taxes 961–62 destination-based sales formula 577, 578, 586–87 destination principle 659, 732, 766–67, 777 digitalization and VAT in EU 1044, 1046, 1047, 1054–55 EU/OECD VAT/GST comparison 789, 790, 792, 793, 795, 796 destruction or displacement effect 1029–30 determination panel 995–96 developed countries 821, 833, 834, 836–37, 840, 869, 925–26 customary international law 85 global tax governance 1009, 1010, 1011, 1012, 1014, 1016, 1021–22 multilateralism 494–95 Pillar One 953–54 Pillar Two 973–74, 977, 978 regional double tax treaties 471 tax treaties 49 transfer pricing for MNEs (OECD) 541 developing countries 821–22, 833–34, 835, 836, 837–38, 840, 869, 874, 899, 902, 904, 908–10, 914 agents 502–3 BEPS 2.0 153 customary international law 85 digitalization and VAT in EU 1054 EU import VAT 747n.1 global tax governance 1009, 1010–12, 1014, 1016–18, 1019, 1020–23 international trade law 215–16 multilateralism 487, 494–95 OECD role 1012–14 Pillar One 962, 966 Pillar Two 969, 973–74, 977, 978, 979 regional double tax treaties 457, 460, 463, 468 tax treaties 49 transfer pricing for MNEs (OECD) 541 unilateralism, bilateralism and multilateralism 480

1070 Index Devereux, M.P. 244, 1000–01 DFDS 738 DG Competition Working Paper on State Aid and Tax Rulings 685 Dharmapala, D. 809 Dietsch, P. 1007 digitalization and VAT in EU 1041–57 corporate taxation 1053–55 digital services taxation without consideration of money 1051–53 improvements in VAT system for digital transformation 1048–50 key features 1043–44 VAT Directives (1967-2006) 1044–47 VAT Directives (2006-2020) 1047–48 VAT fraud and VAT collection simplification 1055–57 digitalized economy 15–16, 45–46, 992 1920s compromise to 2020s compromise 32–33, 35–36, 39 agents 505–6, 513 BEPS 1.0 149, 153 consensus international tax law (CITL) 97–98 Court of Justice of the EU (CJEU) and fixed establishment 593, 744–46 double taxation avoidance 319–20, 329 emerging tax consensus challenges 995, 997–98, 1000, 1003, 1005 global tax governance 1013 labour taxation and automation/ AI 1025–26 low-and middle-income countries (LMICs) 131–32 multilateralism 482–83 national tax systems 88, 103–4 Pillar Two 969–70, 976 public international law 182, 194–95 single-tax principle 145 see also China as digitalized economy; digitalization and VAT in EU; digital services tax (DST); East African Community (EAC) as digitalized economy digital services 355, 1002–03, 1004–05 digital services tax (DST) 12–13, 667–68, 829, 933

1920s compromise to 2020s compromise 33–34, 38 BEPS 2.0 153 emerging tax consensus challenges 993–94, 1002, 1003, 1004–05 jurisdiction 67–68 language factors 263 multilateralism 493–94 public international law 194–95 diplomatic immunities law 74–75, 190 diplomatic means 186 diplomatic missions and consular posts 189, 467 direct delivery approach 793 direct discrimination and trade in goods see under trade law direct effect (EU) 50–51, 667, 787 Directive law 604–6 directives for administrative cooperation (DACs) 682, 684 directors 465 direct taxation 54 1920s compromise to 2020s compromise 23 Court of Justice of the EU (CJEU) 591, 597–99, 602–3 domestic law 48 double taxation treaties and language factors 255 EEA/EFTA 646, 648, 650, 659, 660, 661 EU anti-tax-avoidance regimes 679, 681–82, 685, 694–95 EU candidate countries 664–65, 666–67, 668–69, 671–72, 677–78 EU and cross-border VAT 743, 744 EU law 50–51 EU ‘white supplies’, double taxation and VAT 768, 776, 777–78 fundamental freedoms (EU) 624 international trade law 213, 228–29 regionalism 483 direct use approach 793 disclosure requirements 53–54, 150, 425n.23, 694–95 discretion to provide information 364 discrimination 214, 607, 632, 655–56 clauses 54 direct 612, 613, 620 fundamental freedoms 612–14, 620–22, 626

Index 1071 horizontal 612, 616 indirect 612, 613 nationality 612, 613, 652 residency 50–51 reverse 634–35 vertical 612 see also non-discrimination disguised loan 569–70 disguised restriction 613 dispute resolution 19, 47 1920s compromise to 2020s compromise 36 agents 511–12, 513 BEPS 1.0 150 consensus international tax law (CITL) 99 corresponding adjustments and transfer pricing 566, 567 digitalization and VAT in EU 1053–54 emerging tax consensus challenges 995–96 EU law 52 EU ‘white supplies’, double taxation and VAT 778 horizontal 214–15 human rights law 422 international trade law 218–19 jurisdictional 277 mandatory 960–61 multilateralism 492 OECD Transfer Pricing Guidelines 550–53 Pillar Two 988–89 private international law 173–75 public international law 178–79, 186– 88, 191–92 transfer pricing for MNEs (OECD) 536 vertical 215 see also alternative dispute resolution in EU disputes 38, 40, 52, 546–49, 775–81 see also dispute resolution distance sales 1047–48, 1049–50, 1055 distortion of competition 598, 631, 643, 783, 1044, 1045, 1046–47 distributive justice 977 distributive rule 189–90, 394, 402, 404–5, 414–15, 444, 447–49 diverted profits tax (DPT) (UK) 93 dividends 1920s compromise to 2020s compromise 22–23, 24

beneficial ownership 373–74, 382, 385 charities 335–36, 343–44 corporate law 202–4, 212 corresponding adjustments and transfer pricing 569 double taxation avoidance 321, 324 EEA/EFTA 653–55 EU law 51 exchange of information 363 fundamental freedoms (EU) 618 human rights law 428–29 imputation credit system 619 inbound 654–55 outbound 653–54 principal purpose test (PPT) 406, 409 regional double tax treaties 462 triangular cases 302–3, 312, 313, 314–15, 316 domestic tax law 5–6, 45–46, 47–48, 55, 56–62 1920s compromise to 2020s compromise 26, 28–29 administrative agreements and arbitration 61–62 agents 505–6 Anti-Tax Avoidance Directive (ATAD) 691, 694 beneficial ownership 375–77, 380–86 charities 340–41, 345, 348, 349, 351 consensus international tax law (CITL) 94, 95, 99–100 corporate law 197–98, 199, 200–1, 202, 207, 210 corresponding adjustments and transfer pricing 555, 560–62, 563–64, 565, 567 Court of Justice of the EU (CJEU) 596–97, 604 customary international law 81, 83, 85, 86 double taxation treaties and language factors 253–54, 256–57 economic analysis of law 231–32 emerging tax consensus challenges 992, 999–1000, 1001, 1002–03 EU alternative dispute resolution 697, 703–4, 723 EU anti-tax-avoidance regimes 685–86 EU law 50–51 exchange of information 355, 361, 363, 364, 365, 366–67, 368, 369–7 1

1072 Index domestic tax law (cont.) fundamental freedoms (EU) 610, 623–24 global tax governance 1007, 1019, 1021–22 human rights law 423–24, 429, 430–31 jurisdiction 65, 66, 68, 69–7 1 labour taxation and automation/ AI 1038–40 language issue 61 low-and middle-income countries (LMICs) 127–28, 137 multilateralism 62, 488, 489–90, 492, 494 national tax systems 88, 93–94, 103, 104–5 OECD Model and its commentaries 58–61 partnerships 439–40, 442–43, 445, 450 personal nexus 108, 113–15, 121, 122 Pillar One 958, 965–66 Pillar Two 970–7 1 principal purpose test (PPT) 407–8, 412– 14, 416 public international law 178–79, 183– 84, 190 qualification conflicts 285, 287–88, 290, 291–93, 294, 295–97 regional double tax treaties 458, 459, 466, 471 tax treaties 48, 50 transfer pricing 520, 521, 522, 526, 527–28, 538, 546, 584–85 triangular cases 301–3, 305, 308, 313 unilateralism, bilateralism and multilateralism 475–76 see also under Brazil; India; private international law domicile 115–16 of choice 115 of dependence 115 of origin 115 Dong Yang 738, 739 double deduction (DD) 452 double incrimination principle 365 double negative effect 528, 1025–26 double non-taxation/double non-taxation prevention 47, 319, 323, 324, 326–27, 328, 333, 334, 825 BEPS 1.0 148, 149, 150, 151 BEPS 2.0 155 domestic law 57

EU and cross-border VAT 731–32, 743 international trade law 226–27 juridical 440–41, 447–49 multilateralism 489, 491–92 partnerships 436, 437–38, 440–41, 449– 50, 452–53 personal nexus 109, 120 qualification conflicts 285, 286–87, 290, 291 value creation 938–39 double non-transfer pricing 1007 double taxation arrangements (DTAs) 78–79, 98–99, 592–93, 604, 606–7 double taxation avoidance agreements (DTAAs) 124, 128–29 double taxation conventions (DTCs) alternative dispute resolution in EU 702–5 charities 336–37, 339–41, 342–44, 345, 346– 47, 352 customary international law 73, 74, 79, 80, 81–82, 83 fundamental freedoms (EU) 620–21 human rights 417–18, 424–27 Pillar One 954–55, 966 principal purpose test (PPT) 391, 392–93, 395–96, 397–98, 404–5, 406–7, 409–10, 412, 413–14, 415–16 public international law 189 transfer pricing 527, 528, 530, 572, 573, 574– 75, 584–85 Double Taxation Dialogue for VAT purposes 779–80 double taxation/double taxation prevention 4, 17–18, 45, 47, 319–34, 829 1920s compromise to 2020s compromise 22, 26, 27–28, 32, 38, 40, 42–43 agents 499–501, 502–3 Anti-Tax Avoidance Directive (ATAD) 693–94 basic effects of methods 320–22 beneficial ownership 33–34, 373, 375, 376– 77, 380, 385–86, 389 BEPS Reports (2015) 149, 150, 325–28 charities 335–36, 339 consensus international tax law (CITL) 98–100 corporate law 199, 203–4, 210, 212

Index 1073 corresponding adjustments and transfer pricing 567 Court of Justice of the EU (CJEU) 592–93 customary international law 74, 80–82, 84 design and applicability of method articles 322–23 digitalization and VAT in EU 1045 domestic law 48, 57–58, 233–35, 319–20, 321, 329–30, 334 economic analysis of law 231–36 EEA/EFTA 656 EU alternative dispute resolution 705–6, 707–9, 712, 713, 725, 726–27 EU anti-tax-avoidance regimes 695 EU and cross-border VAT 731–32 EU law 51, 52 EU/OECD VAT/GST comparison 787 fundamental freedoms 609, 615, 617–20, 622, 627 global tax governance 1007 international trade law 214, 223–24, 226 juridical 233–35, 319–20, 321, 323, 324, 325, 328–30, 334 jurisdiction 70–7 1 language factors 261, 262 legal 560 low-and middle-income countries (LMICs) 124, 125, 126–27, 131–32 multilateralism 481, 490 national tax systems 87–88, 90, 94 partnerships 436, 437–38, 440–41, 442–43, 446, 447–50, 452–53 personal nexus 108, 111, 112, 120 Pillar One 328–31, 957–58, 964–65 Pillar Two 331–33, 973 principal purpose test (PPT) 394–95, 398, 404–5, 413–14 private international law 165–66 public international law 181 qualification conflicts 285, 286–87, 290, 291, 294 regional double tax treaties 466 regionalism 483, 485 residence taxation 240–41 switchover rules (exemption vs credit) 323–26 tax treaties 48–50

transfer pricing 529n.46, 550–51, 552, 575 triangular cases 301–2, 304, 305, 313, 314 unilateralism, bilateralism and multilateralism 473, 474–75, 476 value creation 939 see also economic double taxation/ economic double taxation prevention; ‘white supplies’, double taxation and VAT in EU double taxation treaties (DTTs) 1920s compromise to 2020s compromise 24, 25, 26, 37–38, 41, 42, 43 corresponding adjustments and transfer pricing 559, 560–61, 562–64, 566, 568, 569 EU and cross-border VAT 744 exchange of information 353, 355, 362, 367 language factors 250, 263–64 public international law 180, 181, 184–85, 187, 190, 191, 192 see also regional double tax treaties; and under language Dourado, A.P. 479, 911–12, 966, 1008 dual criminality 145–46 dual residence cases 301, 309–11 R1-R2 tax treaty 309–10 see also reverse dual resident cases due diligence minimum standard 354 Due Process and Commerce Clauses 574n.14 earnings-stripping rules 200–1 East African Community (EAC) as digitalized economy 457, 459–60, 899–915 arm’s length principle 902 base eroding payments tax 910–11 certainty 899 common consolidated corporate tax base (CCCTB) 904 controlled foreign company (CFC) rule 912 corporate tax 900–1 de minimis threshold 907 digital services 908–10 digital services tax (DST) 900–1, 905–8, 914, 915 direct tax 900, 905–6 domestic law 906, 908, 910, 915

1074 Index East African Community (EAC) as digitalized economy (cont.) double tax arrangements (DTAs) 900–1, 905, 906, 907, 908, 909, 910, 915 e-commerce 900 equality 899, 903, 907 excise tax 906 fairness 899, 903, 907 formulary apportionment 900–5, 915 global minimum tax 910–11 income inclusion rule (IIR) 910–11, 912 indirect tax 905–6 legitimacy 899, 903 minimum alternative tax 905 non-discrimination 906, 909 OECD proposals 910–14 over-taxation 909 Pillar One and Pillar Two 910–13 qualified minimum domestic minimum top-up tax (QDMTT) 912, 913–14 residual profits 910–11 special economic zones 913 tax sovereignty 908–9, 913, 915 technical services 908, 909–10 transfer pricing 902, 903, 905, 913 treaty adaptations 908–10 Unified Approach 910–12, 913 unitary taxation 900–5, 915 withholding tax 908, 909 EBITDA (earnings before interest, tax depreciation and amortization) 200– 1, 689, 691 EC Communication on and External Strategy for Effective Taxation 682–83 Ecker, T. 777–78 economic advantage and selectivity 630–31 economic allegiance 22–23, 125, 961–62 economic analysis of law 231–47 double taxation 232–36 residence taxation and relief 240–43 source taxation rights 237–40 tax competition 244–47 economic criterion 596 economic double taxation/economic double taxation prevention 48, 200, 523, 1035–36 corresponding adjustments and transfer pricing 555–57, 558–59, 560, 562, 564– 65, 567–68

Economic and Financial Affairs Council (Ecofin) 682–83, 707 economic substance approach 987 Economists’ Report (1923) 22–23, 24, 26 Edward I 11–12 EEA/EFTA 611–12, 645–61 Anti-Tax Avoidance Directive (ATAD) 688–89 dividends 653–55 exit taxation 656–58 freedom of establishment 650–52 free movement of capital 652–53 future taxation issues 660–61 indirect taxes 658–59 information exchange, lack of and potential discrimination 655–56 insurance premium tax 660 non-discrimination principle 647–48 state aid provisions and taxes 648–50 effectiveness principle 394–95, 789–90, 793– 94, 795, 1035–36 effective remedy right 426–27, 725 effective tax rate (ETR) 374, 972–73, 974, 979– 81, 982–85, 986, 989, 1001 effet utile doctrine 258, 601 efficiency 28–29, 789–90, 793–94, 795, 953–54, 961, 963–64 EFTA Surveillance Authority (ESA) 648–49, 650 Einaudi, Professor 29 electronically supplied services 792, 1046–49, 1050, 1055 electronic reporting obligation 1056 Elliffe, C. 911 employment income 464–65 empowerment provisions 665–66 encryption 1043 equality of arms principle 425n.23 equalization levy (EL) 153, 829, 851, 993–94, 1002, 1003, 1004–05 equal treatment Anti-Tax Avoidance Directive (ATAD) 694 comparative tax law 278–79 Court of Justice of the EU (CJEU) 607 digitalization and VAT in EU 1044 EEA/EFTA 648 fundamental freedoms (EU) 612 private international law 161

Index 1075 public international law 189 state aid (EU) 630 see also non-discrimination equity 48, 198–201, 212 erga omnes binding rule 84, 86, 91, 595 Eritrea 113n.22 Erzberger, M. 881 escape clause 402–3, 404–5, 409 estoppel 59–60 ethical issues 245 Ethiopia 747n.1 Europe 125, 159–60, 221, 476, 548–49 see also European Union European Audit Office 673–74 European Commission of Human Rights (ECnHR) 419–21, 423, 430–31 European Convention on Human Rights (ECHR) 54, 190–92, 418–19, 420 see also under human rights law and tax treaties European Court of Human Rights (ECtHR) 54, 190–91, 418–19, 423–24, 428, 430–31 European Court of Justice (ECJ) see Court of Justice of the EU (CJEU) European Economic Community (EEC) 785–86 European Union 821–22, 829, 831–32, 833–34 1920s compromise to 2020s compromise 38 BEPS 1.0 148 BEPS 2.0 153 candidate countries see European Union candidate countries charities 338–39 Common Consolidated Corporate Tax Base (CCCTB) programme 824–25 comparative tax law 267, 268, 279 corporate law 203–4, 205 digitalization see digitalization and VAT in EU digital services tax 824–25, 830–31 dispute resolution see alternative dispute resolution in EU double taxation avoidance 333–34 double taxation treaties and language factors 255 exchange of information and cooperation 53–54, 355

fundamental freedoms see fundamental freedoms human rights law 429 import taxation see import taxation in EU international tax regime (ITR) 146, 155, 156 international trade law 214–15, 223–24, 228 labour taxation and automation/ AI 1025–26 language factors 250, 258–60, 262, 263–64 law (material law) 50–52, 62 partnerships 435–36 Pillar One 960–61 privacy and data protection 830–31 private international law 162, 164, 167 public international law 194 regionalism 483–85 state aid see state aid tax haven blacklist 682–83 transfer pricing 523n.24, 552–53 transfer pricing and formulary apportionment 571–72, 574–75, 581, 583, 584 unilateralism, bilateralism and multilateralism 477, 479 VAT see digitalization and VAT in EU; European Union VAT law and OECD VAT Guidelines see also Court of Justice of the EU (CJEU); EEA/EFTA; and under partnerships European Union anti-tax avoidance regimes 679–95 soft law instruments 682–84 state aid 684–85 tax avoidance as general principle of EU law 685–86 Tax Directives and anti-avoidance rules 679–82 see also Anti-Tax Avoidance Directive (ATAD) European Union candidate countries 663–78 administrative capacity of tax authorities, cooperation and mutual assistance 671–74 Bulgaria 674–78 own resources of budget of EU 667–7 1 European Union Commission Action Plan on VAT 785–86, 787 European Union Tax Observatory 990

1076 Index European Union VAT law and OECD VAT Guidelines 785–96 aim and legal status 787–88 core features of VAT 788–90 services and intangibles and place of taxation 790–95 comparison 794–95 EU law 790–92 OECD VAT/ Guidelines 792–94 exchange of information (EOI) 45 and cooperation of tax administrations 52–54 digitalization and VAT in EU 1043, 1050, 1056 EEA/EFTA 655–56, 661 emerging tax consensus challenges 1002 EU alternative dispute resolution 711 EU anti-tax-avoidance regimes 694–95 global tax governance 1014–15, 1016 human rights law 418, 425–27 low-and middle-income countries (LMICs) 127 multilateralism 488 Pillar Two 988–89 public international law 191–92 regional double tax treaties 467, 468 transfer pricing for MNEs (OECD) 536, 549–50 see also Automatic Exchange of Information (AEOI) regime; exchange of information (EOI) and tax treaties exchange of information (EOI) and tax treaties 353–7 1 competent authorities, confidentiality and specialty principle 361–62 industry-wide exchange of information 363–64 information upon request 362–63 international legal instruments 353 joint audit 363 limits 364–70 good faith and requests based on stolen data 364–66 OECD Model and limits clause 369–70 public policy 368–69 reciprocity principle 366–67

subsidiarity principle 366 trade or business secret 367–68 material scope 355–60 foreseeable relevance of request 355–57 identification taxpayer 357–59 third parties and transfer of information 359–60 personal scope 360 simultaneous examination 363 taxes covered 360 tax examination abroad 363 taxpayers’ rights 370–7 1 exchange theory of taxation 111–12 excise duties/taxes 51, 346, 646, 649, 747 EU candidate countries 665–66, 667, 675, 676, 677 exclusion rule 833 exclusivity requirement 350 exemption 304, 648–49 full 320, 650 on importation 757–59 partial 650 participation 48, 204, 324 with progression 320, 321–22 reciprocal 346 see also exemption method exemption method 223–24, 320–21, 322, 323, 324, 326, 328–29, 330, 332, 334, 442– 43, 617–20 exit taxation Anti-Tax Avoidance Directive (ATAD) 687, 688–89, 692–93, 694 corporate law 208–9 EEA/EFTA 656–58 EU anti-tax-avoidance regimes 681–82 EU law 51–52 human rights law 431 personal nexus 113 transfer pricing and formulary apportionment 573 exit value (resale market price) 542 Expert Committee Report to League of Nations Finance Committe 107–8 exports taxation 1045–46 see also capital export neutrality (CEN)

Index 1077 expropriatory taxation protection 192–93 Facebook C9P63, C45P28 fact gathering process 370–7 1 facts and circumstances tests 117, 120, 121–22 factual situation 168–70 faculty theory 113–14 fairness 1920s compromise to 2020s compromise 28–29, 42–43 domestic law 48 EU/OECD VAT/GST comparison 789–90, 793–94, 795 EU ‘white supplies’, double taxation and VAT 766 global tax governance 1007, 1008 Pillar One 953–54, 961–62, 963 principal purpose test (PPT) 392– 93, 413–14 private international law 162, 164 unilateralism, bilateralism and multilateralism 479 fair procedure, right to 699 fair trial right 52, 54, 191–92, 422–24, 699, 725 Falcão, T. 355n.6 family quotient systems 278–79 family taxation 277–79 Faroe Islands 659 Faulhaber, L. 153 FCE Bank 737, 740, 741 fees for technical services (FTS) Brazil 871–73 India 859–61 low-and middle-income countries (LMICs) 133–35 Feld, L.P. 924 Fiat Chrysler Finance Europe 640–41 fiduciary 370, 375–76 Financial Accountability, Transparency and Integrity (FACTI) panel 1011n.19, 1019 Finland 621n.73, 645, 651, 656 Fiscal Affairs Committee of League of Nations 24–25, 501, 519, 536 fiscal privileges and immunities 189–90 fiscal transparency 433–34, 435, 436, 438, 439– 42, 445–46, 447–50 ‘fishing expeditions’ 355–56

fixed establishment (FE) 732–34, 791–92, 796 see also Court of Justice of EU (CJEU) and fixed establishment fixed methods or approaches 435–36, 450 flexibility 789–90 Forbes ‘Global 2000’ lists 804, 816 force of attraction principle 132–33 Ford, M. 1026 foreign tax law 266, 267, 270 foreign tax law, see also under private international law foreseeable relevance of request 355–57, 362– 63, 366, 369–70 formulary apportionment 492 1920s compromise to 2020s compromise 30–31, 40, 42 emerging tax consensus challenges 995, 998–1001, 1005 transfer pricing for MNEs (OECD) 540, 541, 548 see also transfer pricing and formulary apportionment Forrest, L. 338–39 forum shopping 171–72 Forum on Tax Administration 488–89 Four Economists 476 fractional apportionment 25, 33, 131–32, 141, 519n.7, 536–37, 766, 959, 1044 France 9, 10–12, 830–31 agents 504–5, 511–12 Anti-Tax Avoidance Directive (ATAD) 692n.79 charities 337–38, 351 comparative tax law 269n.29, 277–79 corporate law 199, 209 digitalization and VAT in EU 1044 digital service tax 824–25 EEA/EFTA 655 EU candidate countries 668 exchange of information 356–57, 358–59, 365 human rights law 425n.23, 427 personal nexus 115 Pillar Two 969–70, 974 private international law 165, 166n.38, 168– 70, 172–73 taille 12–13 YouTube tax 824–25

1078 Index fraud beneficial ownership 379 carousel 1056 digitalization and VAT in EU 1042, 1045, 1055–57 EEA/EFTA 652 EU alternative dispute resolution 709, 718–19 EU anti-tax-avoidance regimes 679–80 EU candidate countries 669–70, 671–72, 676 EU ‘white supplies’, double taxation and VAT 770 fundamental freedoms (EU) 625 VAT 1050 Frederick II, Emperor 16–17 freedom of establishment 205, 605–6, 614, 650, 654, 656–57, 658 EEA/EFTA 650–52 EU alternative dispute resolution 705–6 EU candidate countries 666 Pillar Two 971–72, 987–88 freedom to provide and receive services 611, 647, 659, 660, 666, 731–32, 787 free movement of capital 50–51, 605–6, 611, 612, 613, 616, 617, 647, 650, 652–53, 655–56, 658, 666 free movement of workers 614, 647–48, 666 free-riding 31 ‘free’ services 1043, 1051 fringe benefits 269 full taxation norm 27 fundamental freedoms Anti-Tax Avoidance Directive (ATAD) 692–93 charities 338–39 Court of Justice of the EU (CJEU) 595, 599–600, 601, 602–3, 605–6 EEA/EFTA 647, 661 EU anti-tax-avoidance regimes 679–80, 685–86, 695 EU law 50–51 EU ‘white supplies’, double taxation and VAT 767 see also fundamental freedoms and tax treaties fundamental freedoms and tax treaties 609–27

allocation of taxing rights 615 anti-tax-avoidance provisions 622–27 discrimination and restrictions 612– 14, 620–22 double taxation elimination 617–20 most-favoured-nation (MFN) principle 616–17 objective factors 624–25 relevant provisions 611–12 subjective factors 624–25 fundamental rights 48, 54, 724, 725, 1042 Furman, J. 818 G20 45–46 1920s compromise to 2020s compromise 36 agents 505–6 BEPS 2.0 154–55 consensus international tax law (CITL) 98 exchange of information and cooperation 53 global tax governance 1011–13, 1014, 1022–23 international tax regime (ITR) 155–56 low-and middle-income countries (LMICs) 128–29 multilateralism 489–90 national tax systems 103 tax treaties 50 transfer pricing and formulary apportionment 571–72 United States 816 G24 131–32, 905, 959, 995, 1013 G77 1013–14 Gadzo, S. 89, 93–94 Gambaro, A. 272n.49 Gann, P.B. 112–13 Garcia-Bernardo, J. 809–10 general agreement (global settlement) 366 General Agreement on Tariffs and Trade (GATT) 54, 213, 214–17, 218–19, 220–24, 226–27, 228–29, 747–48, 749, 757, 763 General Agreement on Trade in Services (GATS) 213, 225–29 general anti-avoidance rule (GAAR) Anti-Tax Avoidance Directive (ATAD) 687, 689–90, 692, 694 beneficial ownership 388, 389

Index 1079 consensus international tax law (CITL) 99 double taxation 236 EU anti-tax-avoidance regimes 681–82, 684 national tax systems 105n.107 principal purpose test (PPT) 392–93, 397– 98, 402–3, 404, 407–8, 411, 412, 416 transfer pricing (OECD MC) 530 generalized reverse-charge mechanism (GRCM) 1056 generally accepted accounting principle (GAAP) 957, 963 general principles 179–81, 182–83 Geneva Model 322 genuine connection 66, 67, 237 Georgia 663–64 Germany 17–18, 116n.38, 879–97 ability-to-pay principle 882 activity clause 885 administrative sources 883 agents 505–6 anti-abuse tax legislation 879, 889–96 CFC legislation 893 domestic SAARs 892–93 exit taxation 894 family foundations 895–96 GAAR approach 892 hybrid mismatches 896 related party transfer pricing 893–94 tax avoidance 889–93 thin capitalization 894–95 treaty GAARs 892 Anti-Tax Avoidance Directive (ATAD) 691–92, 884–85 arm’s length principle 887, 894 base erosion and profit shifting avoidance 886 beneficial ownership 385–86, 387–88, 388n.93 bilateral agreements 886 charities 337–39 comparative tax law 267–68, 271, 278–79 consistency principle 882 constitutional sources 882 controlled foreign corporation (CFC) tax 881–82, 885, 889–90, 891, 893, 895 corporate law 200–1, 206, 209n.62 corresponding adjustments and transfer pricing 563–64, 569

court involvement 886–89 Court of Justice of the EU (CJEU) 599, 604–5, 607, 888–89 lower tax courts (Finance Courts) 889 Supreme Court 886–88 Supreme Tax Court (Federal Fiscal Court) 888, 891 covered tax agreements (CTAs) 886 deduction/non-inclusion (D/NI) situations 896 direct tax 888 domestic law 47, 879–80, 882, 884–85, 886– 87, 893 double deduction 896 double-dip-arrangements 891 double tax arrangements (DTAs) 882, 883, 885, 886–87, 891 double taxation/double taxation prevention 324, 325n.11, 893, 894 EBITDA 894–95 EEA/EFTA 650, 654–55 equal treatment clause 882, 886, 887 equity test 895 escape clause 895–96 as EU member country 883–85 primary EU law 883–85 secondary EU law 885 EU ‘white supplies’, double taxation and VAT 768, 769, 776–77, 780n.51 exchange of information 358 exit taxation 881–82, 885, 889–90, 891, 894 family-foundation schemes 881–82, 889– 90, 895–96 Foreign Tax Act 881–83, 887, 893, 894, 895 Foreign Transaction Tax Act (AStG) 887 fundamental freedoms 618n.56, 884 general anti-avoidance rule (GAAR) 889– 90, 891, 892 General Tax Code (GTC) 889–91 group clause 895 historical background of taxation 880–82 human rights law 419, 420 hybrid mismatches 885, 891, 896 Income Tax Act (EStG) 893, 894, 896 indirect taxes 884 Inheritance and Gift Act (ErbStG) 886 inheritance tax 888 interest limitation rules 894–95

1080 Index Germany (cont.) international tax education 896 judicial committee reports 882–83 judicial sources 883 labour taxation and automation/ AI 1030n.28 lawfulness in taxation 882 legislative sources 882–83 licence barriers 891 limitation-on-benefits (LOB) clause 891 Ministry of Finance 890, 891–92 notion of tax 879–80 opening clause 889 partnerships 435, 437, 447–49 personal nexus 115 Pillar Two 969–70, 974 primary law (EU) 883–85, 896 private international law 169n.54 qualification conflicts 287n.6, 293–94n.23 quantified threshold requirements 893 Reich Flight Tax 881 related party transfer pricing 893–94 reverse hybrid situations 896 rule-of-law 882 secondary law (EU) 884–85 specific anti-avoidance rules (SAARs) 891, 892–93 state aid issues 885 state liability 885 tax avoidance/tax avoidance prevention 884–85, 889–93 tax evasion 879–80 tax residency mismatches 896 tax treaties 49, 885–86 thin capitalization rules 891, 894–95 third-party financing 894–95 transfer pricing 548–49, 881–82, 891 treaty shopping 882, 889–90, 893 VAT 884, 888 VAT Directive and imports 753–54 welfare state principle 882 Gibraltar 632n.20 gifts 45, 335–36, 337, 339, 347, 348 gift treaties (examples) 349–51 Global anti-Base Erosion (GloBE) rules 37, 50, 194–95, 996–97 see also Pillar Two of OECD proposal and Global anti-Base Erosion (GloBE)

global blending 981, 982–83 global excess profits (GEP) 1000 global financial crisis (2008) 21, 29, 39, 41, 43, 95, 127, 145, 146 Global Forum 127, 140–41, 355, 470, 480, 488, 494–95 Global Forum on Transparency and Exchange of Information 128–29, 1002, 1005 Global Forum on VAT 483 global intangible low-taxed income (GILTI) (USA) 33n.69, 151–54, 155, 324–25, 808, 809, 811–12, 815, 969–70, 972, 982, 985, 986–87 global legislative extraordinary proceeding 174 global minimum tax 37, 38, 45–46, 182, 513, 834–35 global tax governance 1007–23 actors 1021–22 development of international tax standards 1010–12 OECD role and developing countries 1012–14 outcome 1019–21 participation and representation 1017–18 soft law and hard law 1014–16 Global Transparency Forum 1015, 1016 GLOBTAXGOV ERC-funded project 1009 Goldsmith, J.L. 77, 84 good administration principle 724, 725 good faith corresponding adjustments and transfer pricing 566 double taxation treaties and language factors 254 exchange of information (EOI) and tax treaties 364–66, 368–69, 370 jurisdiction 70–7 1 national tax systems 101 principal purpose test (PPT) 394–95, 396– 97, 398, 412–13 public international law 184–85, 186–87 tax treaties interpretation 56 transfer pricing for MNEs (OECD) 545 good practices 704 goods and services 733, 737, 746, 776 Goods and Services Tax (GST) 482, 768 see also VAT

Index 1081 goodwill 945 Google 154, 830–31 Google/Australia dispute 914 Gooijer, J. 340–41 government services 189–90, 465 grandfathering rule 689 grants and loans 648 Grazioxi, G.-R. 471 Greece 753–54 Greenland 659 group contribution schemes 651 group requests 53–54, 357–58, 359 guaranteed payments 437–38, 442–43, 447–49 Guedes, E.F.D.L. 911 Guinea 747n.1 Gurria, A. 148 Guttentag, J. 145–46 Guy de Lusignan 16–17 hacking 147 Hancock, J. 10–11 hard law 45–46, 83–84, 99, 103, 684, 1008, 1014–16, 1022–23 hard-to-value intangibles (HTVIs) 536, 545, 584–85 harmful tax competition 31–32, 40, 42, 97, 835 harmful tax practices 37, 39, 84–85, 149, 511–12 harmonization 88, 597–99, 787, 790, 795–96 double taxation prevention 781–83 indirect taxation 749, 771–75 Hartman, D.G. 802 Hasegawa, M. 924, 930–31 head office (HO) and permanent establishment (PE) 735–37 heard, right to be 699 Hearson, M. 96 Heckscher-Ohlin theory of trade 15–16 Hellier, C. 183 Henry III 16–17 Herdegen, M. 78–79, 81 Hey, J. 912 historical background 3–20 historic recurrences 9–13 mutation 13–20 permanence 6–13 Hongler, P, 1001 Horn, H. 218–19

Hudson, M. 18–19, 76 Huerta, R. 290n.16 Human Rights Committee 419n.6, 423 human rights law 190–92, 261 see also human rights law and tax treaties human rights law and tax treaties 417–32 cases before ECnHR and ECtHR 419–21 development 417–19 ECHR and ICCPR 421–31 fair trial right and mutual agreement proceedings (MAP) 422–24 leave a country, right to 431 non-discrimination 428–29 privacy right and double tax conventions 424–27 property right 429–31 Hungary 113n.23, 668, 769, 773–74, 971–72 Hurd, I. 1009 hybrid entities 440–47, 690, 694–95 domestic law 48 double taxation avoidance 327–28 partnerships 436, 437–38, 449–53 Pillar Two 984–85 private international law 170 regular 442, 443 reverse 442, 445, 452 see also BEPS Action 2 Report; OECD Partnership Report (1999) hybrid financial instruments 201 hybrid mismatch arrangements (HMAs) 445, 452 hybrid mismatches 149, 681–83, 686, 690, 896 IBFD 336, 349 Iceland 645, 656, 658, 659, 660–61, 663– 64, 678 identification taxpayer 357–59 collectives (or bulk) requests 358–59 group requests 357–58 IF Task Force for the Digital Economy 995 illicit/illegal behaviour 362, 365–66, 1002 banking solicitations 359 financial flows 140, 1002 see also abusive practices prevention; corruption fraud; money laundering; sham or simulated operation immediate or surrogate parent entity 549

1082 Index immovable property income 17–18, 306–9, 776, 1045 EU/OECD VAT/GST comparison 791, 794, 795 gains from alienation 306–8 income from use of 308–9 regional double tax treaties 460 impersonal taxes 107–8, 499–500 imports taxation 1045–46 see also capital import neutrality (CIN); import taxation in EU import taxation in EU 747–63 General Agreement on Tariffs and Trade (GATT) 747–48 harmonization for indirect taxation 749 TFEU principles 748–49 TFEU principles, non-discrimination of internal taxation 748–49 see also VAT Directive imputability and state resources 630 incidental questions 168–70 income allocation 377 apportionment, net vs gross 575–78 see also active income; income inclusion rule (IIR); income tax; passive income income inclusion rule (IIR) 331–32, 334 1920s compromise to 2020s compromise 37 BEPS 2.0 154–55 emerging tax consensus challenges 996–97 multilateralism 492 Pillar Two 973–74, 987–88 income tax 12–13, 45 comparative tax law 268–69, 272– 73, 277–78 Court of Justice of the EU (CJEU) 602–3 domestic law 48, 58 labour taxation and automation/ AI 1034–36 tax treaties 48–49 triangular cases 299–301 unilateralism, bilateralism and multilateralism 476 value creation 941, 948 see also corporate income tax (CIT); personal income tax; and under regional double tax treaties

independent personal services 464, 871– 72, 873–74 independent procedure 370–7 1 India 123, 127, 829, 841–62, 994, 1004 advance pricing agreements (APAs) 848– 49, 852 anti-abuse rules 846 anti-fragmentation rule 858–59 arm’s length price 852, 856 associated enterprise (AE) 852 bilateral investment protection agreements 848–49 business operations versus preparatory auxiliary activities 857 capital gains 846 CBDT Circular 853–54 CBDT Report on Profit Attribution to Permanent Establishments 858 Central Board of Direct Taxes (CBDT) 843–44n.8, 845, 849 confidentiality and security of information 852 constituent entity (CE) 852 corporate tax 844 country-by-country reporting (CbCR) 852 covered tax agreements (CTAs) 845, 846, 858–59 Covid-19 pandemic 842, 852 DBDT 852 demand and supply approach (D&S) 858 dependent agent permanent establishment (DAPE) 856–57, 858 digitalized economy 841–42, 851, 857, 859–60 digital services taxes (DSTs) 851, 860 disposal test 854–56 dispute resolution 848–49 dividends 846 domestic law 841, 842, 844–45, 847 domestic law impacting on international law 849–54 advance pricing agreements (APAs) 852 Black Money (Undisclosed Foreign Income and Assets) and Imposition of Tax Act 2015 853 equalization levy (EL) 851 faceless assessment scheme (FAS) 852–53 foreign income of residents 853–54

Index 1083 general anti-avoidance rules (GAAR) 849–50 general anti-avoidance rules (GAAR) and limitation on benefits (LOB) interplay 850 master file (MF) and country-by-country reporting (CbCR) 852 place of effective management (POEM) 853–54 domestic law and treaty law (including MLI) 844–45 double non-taxation 844 double taxation 858 double taxation avoidance agreements (DTAAs) 842, 844–47, 848–49, 850, 857, 858, 859–60, 862 EBIDTA 861 e-commerce operators 851, 860 efficiency, transparency and accountability 852–53 equalization levy (EL) 842–43, 851, 860 exchange of information 365–66 exemption with progression 844 exploitation of information on consumers 856 faceless assessment scheme (FAS) 842, 852–53 FDI 841–42 fees for technical services (FTS) and royalties 859–61 Finance Act 2012 853 Finance Act 2021 845, 860 fixed place of business 854 formulary apportionment 858 fractional apportionment 858 fraud 852–53 full credit 844 full exemption 844 function-asset-risk approach (FAR) 858 GDP 841–42 general anti-avoidance rules (GAARs) 844–45, 849–50 habitually securing orders test 856–57 immovable property 846 impermissible avoidance agreement (IAA) 849, 850 income escaping assessment 853 Income-Tax Act 1961 842–45, 848–49, 852

attribution of profits to business connection/PE 858 business connection 857 Income Tax Rules 1962 858 indirect tax 851 indirect transfer 843 interest taxation 861 interpretation of tax treaties 846–48 limitation on benefits (LOB) clause 849, 850 make available clause 859–60 master file (MF) 852 minimum holding period 845–46 MLI 846 most-favoured nation clause 859–60n.47 multilateralism 493, 851 mutual agreement procedure (MAP) 848–49 non-obstante clause 850 non-resident (NR) 842–43, 844, 845, 851, 856–57, 860, 861–62 OECD MC 858–59 permanent establishment (PE) 846, 851, 854–59 business operations versus preparatory auxiliary activities 857 decisions by Indian courts 855t dependent agent PE (DAPE) under domestic law 856–57, 858 determination 859 disposal test 854–56 exploitation of information on consumers 856 fixed place of business 854 subsidiary PE 856 place-of-effective management (POEM) 853–54 preparatory or auxiliary activities 858–59 principal purpose test (PPT) 407–8, 845–46, 850 residential status 842–43, 843t royalties 859–61 search matters 852–53 sensitive cases 852–53 shares or interest 846 significant economic presence (SEP) 824–25, 843, 851 simplified limitation of benefits (SLOB) 845–46

1084 Index India (cont.) splitting of contracts 858, 859 subsidiary PE 856 tax evasion 852–53 tax rates for residents and non-residents 844 tax residency certificate (TRC) 844–45, 862 tax sovereignty 848–49 tax sparing clause 844 tax terrorism 842 tax treaties interpretation 846–48 thin capitalization 861 transfer pricing 544–45n.51, 852–53 treaty law (including MLI) 844–46 unilateralism 851 website may constitute a permanent establishment (PE) 858 withholding tax 851, 861–62 indirect credit 321, 324 indirect taxation 54 Court of Justice of the EU (CJEU) 597– 98, 602–3 EEA/EFTA 646, 658–59, 660, 661 EU candidate countries 664–66, 667 EU import VAT 749 EU ‘white supplies’, double taxation and VAT 766, 777–78, 779 international trade law 213–14, 228–29 regionalism 484 see also EU VAT law and OECD VAT Guidelines; import taxation in EU; VAT in EU; ‘white supplies’, double taxation and VAT in EU Indonesia 123, 994 in dubio pro libertate concept 579–80 information asymmetries 31, 136 misuse of 425–26 rights 699 upon request 362–63 see also exchange of information (EOI) infringement procedure 773, 776 inheritance tax 17–18, 45, 165, 269, 335–36, 348 inheritance treaties (examples) 349–51 initial coin offering/initial token offering 745–46 institutions 96–98

instruments 98–101 insurance premium tax 660 intangibles 828 1920s compromise to 2020s compromise 33 DEMPE (development, enhancement, maintenance, protection and exploitation) standard 584–85 emerging tax consensus challenges 995 EU/OECD VAT/GST comparison 793– 94, 796 OECD Transfer Pricing Guidelines 544–45 Pillar One 963–64 Pillar Two 975–76, 977–78, 985 transfer pricing 537–38, 543 transfer pricing and formulary apportionment 576–77, 580 see also global intangible low-taxed income (GILTI) (USA); hard- to-value intangibles (HTVIs); marketing intangibles approach under United States intellectual property (IP) 1920s compromise to 2020s compromise 30 customary international law 85 international trade law 226–27 Pillar Two 975–76, 978, 980, 982–83, 985 principal purpose test (PPT) 401, 410 state aid (EU) 639 Tax Cuts and Jobs Act (TCJA) (USA) and BEPS 1.0 152 transfer pricing 548 intention of the parties 55–56, 58, 101 Inter-American Centre of Tax Administration (CIAT) 127–28 interest 1920s compromise to 2020s compromise 22–23, 24, 37 beneficial ownership 373–74, 378–80, 385 charities 335–36, 341, 343–44 domestic law 48 double taxation avoidance 332 EU candidate countries 677–78 EU law 51 exchange of information 363 regional double tax treaties 462–63 triangular cases 302–3, 312, 313, 314, 315, 316–17

Index 1085 interest deduction 149, 489–90 interest deduction limitation rules 687, 929–30 interest limitation rules 51–52, 681–82, 689, 691–93, 694 intergovernmental agreements (IGAs) 147, 354, 420–21n.13 interlinkage rule 447 intermediate approach 436–37 internal taxation 665–66, 748–49 internal theory 387–88 International Chamber of Commerce (ICC) 42 International Compliance Assurance Program (ICAP) 488–89, 995–96 International Court of Justice (ICJ) 70, 188 International Covenant on Civil and Political Rights (ICCPR) 418 see also under human rights law and tax treaties international decision harmony 257 International Economic Conference in Genoa (1922) 17–18 international economic law 78–79 International Financial Conference (1920) 22 International Financial Reporting Standards 583, 957, 963, 983– 85, 986–87 International Law Commission (ILC) 73, 75– 78, 80, 83, 91–92 International Monetary Fund (IMF) 124, 129, 140, 250, 841–42, 1017, 1019 International Tax Organization 998–99 international tax regime (ITR) 26, 27, 28, 30, 31, 41, 42, 43, 143–56 BEPS 1.0 148–51 BEPS 2.0 153–55 FATCA and AEOI regime 145–47 reasons for cross-border income taxation 143–44 single-tax principle 144–45 Tax Cuts and Jobs Act (TCJA) (USA) and BEPS 1.0 151–53 inter partes effect 83–84, 595 interpretation 101–3, 255–56, 257, 363, 396–97 autonomous 293 common 293–94

panel 264 qualification conflicts 288–90, 296, 297 and qualification distinction 287–88 rules 253–57 intervention criterion 742 intra-group contribution 651 intra-group services 543 investigation stage 370–7 1 investment inbound 90 law 192–93 outbound 89–90, 242–43 profits 23, 24 Iraq 340–41 Ireland 484–85, 504–5, 691–92 Ismer, R. 775, 777–78 Italy 116n.38, 221, 830–31 Anti-Tax Avoidance Directive (ATAD) 691 comparative tax law 267–68, 278–79 Court of Justice of the EU (CJEU) 599 EEA/EFTA 654 EU candidate countries 668 EU and cross-border VAT 737 EU ‘white supplies’, double taxation and VAT 768 multilateralism 493 partnerships 435–36 private international law 163–64, 169n.54 Izawa, R. 931, 932, 933 Jamaica 1018 Jansky, P. 809–10 Japan 917–33 advance pricing arrangement (APA) 926 agents 505–6 alternative minimum tax (AMT) 921 BEPS Project and corporate behaviour 930–32 cash box subsidiaries 928 charities 337 compliance behaviour of MNCs 930–32 conduit companies 922 consumption tax 920 controlled foreign company (CFC) rules 919n.3, 925, 927–28, 932–33 corporate tax 920, 923, 927–28 credit method 918

1086 Index Japan (cont.) debt-to-equity ratio 929 derivatives trades 928 discounted cash flow method 927 dividends 920, 921–22, 922t, 923, 924, 927, 932 double taxation/double taxation prevention 918, 921, 923, 929, 931 earnings-stripping rules 929 economic stagnation and tax policy changes 919–20 entity approach 927, 928 exchange of information 354 exemption method 918 FDI 918 fixed-ratio rule 929n.39 force-of-attraction rule 921 hard-to-value intangibles (HTVI) 927 hybrid rules 928 income approach 927–28 indirect tax 923 intangibles 921, 924, 925–26, 927 interest 920, 922, 922t interest-deduction limitation rules 925, 928–30, 932–33 international tax regime (ITR) 155 limitation-on-benefits (LOB) 922 mergers and acquisitions 924 Ministry of Economy, Trade and Industry (METI) 923–24, 926, 928 Ministry of Finance 921, 925, 928 MNEs 930–32 National Tax Agency (NTA) 925–26 National Tax Tribunal 926 OECD 918–19, 921 offshoring and value chain globalizing 919, 920–24, 925–26, 932 Japan-US Tax Treaty revision 2003 920–22 territorial system 923–24, 927 passive income 927–28 patent fees 920 permanent establishment (PE) 921 profit method 925–26 profit shifting 917, 925–30 CFC reforms 927–28

interest-deduction limitation rules 928–30 transfer pricing challenges 925–27 profit-stripping provisions 925 R&D 920, 924 royalties 920, 921–22, 922t, 925–26 shell companies 928 source-country taxation 922 source rules 921 tax avoidance 922 territorial taxation 920, 923–25, 927, 932 thin capitalization rules 919, 929, 932 third-party loans 929, 930 transactional net margin method 925 transfer pricing 550, 919, 925–27 Treaty of Peace 918, 931 treaty shopping 922 and USA relationship 918, 919, 920, 921, 932 value chain globalizing 917, 920–24 VAT 920 withholding tax 920, 921, 922t, 924 Jeffrey, R. 89 Jennings, R. 67 Jogarajan, S. 486 Joint International Task Force on Shared Intelligence and Collaboration (JITSIC) 488–89 Joint Statement of the US DOJ and Swiss Department of Finance (2013) 360 Joint Transfer Pricing Forum (JTPF) 706–7, 709, 710, 711 Jones, R.C. 24–25 Júnior, G. 875n.58 jurisdiction 65–7 1 basic principles 65–69 conflicts 160 extension and abuse of rights prevention 599–601 market-user 36, 37 personal nexus 108 to adjudicate 65–66 to enforce 65–66 to prescribe (prescriptive jurisdiction) 65–66, 67–69 to tax 80 unlimited 89 jurisdictional blending 981, 982–83, 986–87

Index 1087 jurisdictional principle 789 Kahn-Freund, O. 273–74n.58 Keen, M. 139, 244 Keith Committee 426n.27 Kenya 900–1, 902–3, 906, 908, 913 Keohane, R.O. 1011 Keynes, J.M. 1026–27 Kiyota, K. 924 Kleinbard, E.D. 804 Koele, I.A. 352 Konrad, K.A. 244 Korea 920 Kosovo 645, 663–64, 678 Kötz, H. 272n.49, 275n.64 Laazouzi, M. 701n.24 labour taxation and automation/AI 1025–40 future of work 1028–33 outlook 1038–40 ‘robot/automation tax’ 1036–38 Lamensch, M. 905–6 Lang, M. 172n.70, 289, 447–49 language 249–64 authentic version and translations 258–59 comparative tax law 274 domestic law 61 double taxation treaties 251–58 interpretation rules 253–57 OECD MC 253 restrictions 257–58 English 258n.36, 263 EU tax law 258–60 EU ‘white supplies’, double taxation and VAT 770–7 1 French 258n.36, 259, 262 German 263 multilingualism 258–59, 260–64 translation errors 259, 263–64 translations, non-binding 262–63 Lara Yaffar, A. 355n.6 last best offer method 720 last resort approach 668–69 Latin America 125, 127–28, 206, 468–69, 1019 Latvia 604–5, 668, 1012 law in action method 270–72 law on the book 271–72

League of Nations 9, 18–19, 22, 23, 25– 27, 30–31 agents 498–500, 501, 502–4 consensus international tax law (CITL) 96 customary international law 74, 78, 80– 81, 82 double taxation avoidance 322 EU alternative dispute resolution 702–3 multilateralism 494–95 national tax systems 90 personal nexus 108 source taxation rights 237–38 tax treaties 48–49 transfer pricing 518, 540 unilateralism, bilateralism and multilateralism 476 League of Nations Report (1923) 238n.31 leave a country, right to 431 legal certainty 261, 262–63, 392–93, 413–14, 416, 607, 664–65, 699 legal families 272, 275 legal formant 270–72, 274–75 legality principle 663–65 legal personality 434, 1035–36 legitimacy 479, 592, 603, 1012–13, 1018 Leitsch, S. 909 Lemaire, S. 701n.24 Lesage, D. 1013 letterbox companies 624, 693 lex fori 389 lex rei sitae (tax residence) 299–301, 302–3 lex specialis 291–92 lex specialis derogat generali 68 liberty, deprivation of 420 licence payments 373–74 Liechtenstein 645, 646, 655, 660–61 limitation on benefits (LOB) beneficial ownership 383, 387 BEPS 1.0 148, 149 charities 342, 346 fundamental freedoms (EU) 626 low-and middle-income countries (LMICs) 138 principal purpose test (PPT) 401n.56, 407–9, 410 limited liability 67 limits clause 369

1088 Index linking rules 445, 452–53, 690 Lisbon Strategy 668–69 Lithuania 651, 1012 location-specific rents 239–40 London Model (1946) 25–26, 48–49, 322, 480–81, 501, 519n.8, 703–4 look-through approach 451 Loretz, S. 244 loss transfers (group relief) 651 Louis XIV 12–13 lowest common denominator 367 low-and middle-income countries (LMICs) 123–41, 991–1006 allocation imbalances 130–32 cooperation 127–29 expectations 1001–04 international tax architecture 125–27 narratives and possible solutions 994–1001 offshore indirect transfers of assets 138–39 profits 125 profits attribution to PEs 132–33 relevant information, acquisition and processing of 140–41 royalties and fees for technical services (FTS) 133–35 strengthened alliance 1004–06 tax incentives, balanced use of 139–40 transfer pricing 135–37 treaty shopping 137–38 low taxation, undesired 324, 326–27, 328, 333, 334 loyalty principle 611, 787 lump-sum basis (forfait fiscal) 356 Luxembourg 201n.22 Anti-Tax Avoidance Directive (ATAD) 691, 692–94 Court of Justice of the EU (CJEU) 604–5 EEA/EFTA 655–56 exchange of information 353–57 human rights law 427 multilateralism 488–89 partnerships 435 principal purpose test (PPT) 399–400 regionalism 484–85 McLure, C. 143, 144 macro-comparisons 273, 275 Magalhaes, T.D. 1000

Malta 484–85 mandatory arbitration 422–23, 567, 607, 702, 704–5, 777–78 Manila Declaration on the Peaceful Settlement of International Disputes 186–87 Mansour, M. 139, 471 marginal tax burden 243 margin of appreciation 191, 419, 428, 430–31 margin of discretion 632, 642 market prices 942–43, 946–47 market-user jurisdictions 36, 37 Markle, K.S. 930–31 Marx, K. 827–28 Mason, R. 96, 236–37n.25, 491 material law 45, 50–52 Maullii, T. 17–18 Mavroidis, P.C. 218–19 Mexico 548, 994 Mexico Model (1943) 25, 48–49, 322, 480–81, 501, 519n.8, 703–4 micro-comparisons 273, 276 micro, meso and macro levels 827 Millennium Development Goals (MDGs) 124 minimum corporate tax anti-avoidance plans 97 minimum duration presence rule 114 minimum effective taxation 587, 1001 minimum standard 19 1920s compromise to 2020s compromise 37, 41 BEPS 1.0 148 consensus international tax law (CITL) 97, 99 corresponding adjustments and transfer pricing 566 Court of Justice of the EU (CJEU) 593 EU alternative dispute resolution 704–5 fundamental freedoms (EU) 622 global tax governance 1015–16, 1017, 1018, 1019–20, 1022 low-and middle-income countries (LMICs) 128, 131 multilateralism 490, 491 principal purpose test (PPT) 415–16 public international law 182 transfer pricing for MNEs 549–50, 551 minimum taxation 36, 40, 53, 103–4, 155, 304, 996–97 see also global minimum tax; Pillar Two

Index 1089 mismatch problems 822 mismatch rules 687, 690 Moldova 663–64 money laundering 355, 682–83 Montenegro 645, 663–64, 678 morality and justice principles 848 Moreno, A.B. 909–10 Mosquera Valderrama, I.J. 1020–21 most-favoured-nation (MFN) principle 16– 17, 477, 748 charities 343, 344 fundamental freedoms 616–17 international trade law 214–15, 225, 226– 27, 228–29 regional double tax treaties 462, 463, 469–70 motive test 604 movable property 17–18, 307–8 Muir Watt, H. 161 multilateral advance pricing agreements (MAPAs) 551 Multilateral Agreement on Administrative Cooperation in Tax Matters (MAATM) 147 multilateral agreements/multilateralism 19, 45–46, 824 1920s compromise to 2020s compromise 29, 32, 35, 40, 42 BEPS 1.0 150 charities 343 consensus international tax law (CITL) 97– 98, 99–100 Court of Justice of the EU (CJEU) 592–93 double taxation avoidance 330 EEA/EFTA 660 emerging tax consensus challenges 999– 1000, 1004–05 EU ‘white supplies’, double taxation and VAT 768 exchange of information 357 fundamental freedoms (EU) 609, 611 global tax governance 1008, 1011, 1013, 1015, 1017, 1019 international trade law 214 jurisdiction 66 low-and middle-income countries (LMICs) 127, 128–29, 140–41 national tax systems 87–88, 90, 103, 104 personal nexus 109–10

Pillar One 965–66 Pillar Two 988 public international law 181, 194–95 regional double tax treaties 468 soft 480–81 value creation 938 see also unilateralism, bilateralism and multilateralism Multilateral Competent Authority Agreement (MCAA) 53 multilateral tax rate 834 multilateral unified substantive rules 171–73 multilingualism 258–59, 260–64 multiple location entity (MLE) 793 multiple taxation 48, 712, 713 Musgrave, P. 240–41 mutation 13–20 adapting to the times 13–17 making progress 17–20 mutual agreement procedure (MAP) 47, 52, 56, 57, 98–99, 697 charities 340–41, 348–49 corresponding adjustments and transfer pricing 559, 561, 563, 565, 566–68 Court of Justice of the EU (CJEU) 593, 606–7 domestic law 61–62 emerging tax consensus challenges 995–96 EU alternative dispute resolution 702, 703– 6, 708, 710, 712, 714–15, 719, 723–24 EU ‘white supplies’, double taxation and VAT 777–78 human rights law 418, 420, 422–24 partnerships 447–49 public international law 187, 191–92 qualification conflicts 294, 295–96 regional double tax treaties 466–67, 470 Tax Dispute Resolution Directive 715 transfer pricing 520, 523, 551–53 triangular cases 310, 311 mutual assistance 468, 657, 661, 667 EU alternative dispute resolution 725 EU candidate countries 671–74 exchange of information and cooperation 52–53 multilateralism 491 private international law 167 regionalism 485–86

1090 Index mutual recognition 342–43, 346, 347, 351, 352, 436, 450 mutual respect 162–63 Myanmar 113n.23 Nakazato, M. 919n.3 national interest 352, 362 nationality 16–17, 112, 113 discrimination 612, 613, 652 non-discrimination 262 national neutrality (NN) 241–43 national tax law 47, 87–105 beneficial ownership 389 comparative tax law 267 consensus international tax law (CITL) 94–103 institutions 96–98 instruments 98–101 interpretation 101–3 corresponding adjustments and transfer pricing 562–64 Court of Justice of the EU (CJEU) 593–94, 600–1, 603–4 customary international law (CIL) 91–94 double taxation avoidance 326, 327, 328, 334 double taxation treaties and language factors 253, 256–57 EU/OECD VAT/GST comparison 787 fundamental freedoms (EU) 615 international tax law defined 89–91 language factors 250 partnerships 447–49 regionalism 484 tax treaties 50 transfer pricing and formulary apportionment 582 unilateralism, bilateralism and multilateralism 473, 474–75 national treatment principle 747–48 negative externalities 944–45, 947 neo-protectionism 586–87 Netherlands 54, 201n.22, 203, 204 agents 511–12 beneficial ownership 383–84 charities 340–41 EEA/EFTA 656 exchange of information 357–58, 368–69

human rights law 420 partnerships 435 private international law 170 regionalism 484–85 state aid (EU) 639 tax treaties 49 transfer pricing 546n.58, 548–49, 552–53 neutrality 243, 324, 733, 741, 829 competitive 321–22 digitalization and VAT in EU 1044, 1048– 49, 1057 EU/OECD VAT/GST comparison 789–90, 793–94, 795 EU ‘white supplies’, double taxation and VAT 777 global tax governance 1007, 1008 labour taxation and automation/ AI 1035–36 market 766 national 241–43 ownership 242, 243 Pillar One 964 private international law 161 VAT Directive and imports 761, 763 see also capital export neutrality (CEN); capital import neutrality (CIN) New Zealand 116, 121–22 nexus approach 81–82, 492, 513, 824, 956–57 1920s compromise to 2020s compromise 33, 36–37, 38, 40, 42 modified 401, 410, 975–76 territorial 93–94 see also personal nexus NGOs 42, 97–98, 482 Niboyet, J.P. 160 Nigeria 1004 1920s compromise and the 2020s compromise 17–18, 21–43, 103– 4, 966–67 2020s compromise 32–38, 104 content 36–38 critical assessment 40–41 issues for resolution and final agreement 35–36 milestones 33–35 alternatives to failure to achieve compromise 38

Index 1091 arm’s length principle (ALP) 24–25, 30–31 crisis of 1920s compromise and precedents of 2020s compromise 27–32 development and consolidation of 1920s compromise 25–27 harmful tax competition 31–32 residence, source and permanent establishment concepts 29–30 nominees or agents 370, 373, 375 non-discrimination 16–17, 54 charities 342–43, 344, 346, 348, 349 consensus international tax law (CITL) 98–99 Court of Justice of the EU (CJEU) 595 customary international law 80 EEA/EFTA 647–48, 661 EU candidate countries 666 EU and cross-border VAT 744 EU/OECD VAT/GST comparison 790 fundamental freedoms (EU) 616 human rights 419, 428–29, 431–32 internal taxation 748–49 international trade law 228 nationality 428 permanent establishments 428 public international law 181, 192–93 regional double tax treaties 466 stateless persons 428 transfer pricing 528 triangular cases 304–5 unilateralism, bilateralism and multilateralism 479 VAT Directive and imports 757 non-prosecution agreement request 360 non-taxation/non-taxation prevention 301–2, 447, 787, 1045 EU ‘white supplies’, double taxation and VAT 765, 767–7 1, 776, 777, 781, 783 see also double non-taxation/double non- taxation prevention non-union good entry into economic network as importation requirement 753–55 norm-setting 592, 603–4 Norr, M. 81–82 North Macedonia 645, 663–64, 678 Norway 358, 504–5, 668, 778, 1030–32

EEA/EFTA 645, 651–52, 653–54, 656, 657– 58, 660–61 notification rules 367 Obama, B. 146, 147, 805–6 objective/bright-line tests 117 OECD 924 1920s compromise to 2020s compromise 32, 39 beneficial ownership 386 charities 337, 338, 352 corporate law 205 Court of Justice of the EU (CJEU) 592–93 customary international law 78, 80–81 and developing countries 1012–14 East African Community (EAC) 910–14 emerging tax consensus challenges 994 EU/OECD VAT/GST comparison 787– 88, 796 exchange of information 367 FATCA and AEOI (USA) 147 global tax governance 1010, 1017, 1018–19, 1022–23 jurisdiction 69 labour taxation and automation/AI 1025– 27, 1028–29, 1030–32 language factors 250 low-and middle-income countries (LMICs) 124, 128–29 multilateralism 487, 488–89 partnerships 436 private international law 171–72, 174 transfer pricing and formulary apportionment 571–72, 574–75 unilateralism, bilateralism and multilateralism 478 OECD flow-through (or look-through) principle (General Principle) 442, 443, 444–47, 451 OECD General Report 446 OECD Model Convention on Taxes on Estates, Inheritances and Gifts (EIGMC) commentary 348–49 gift and inheritance treaties (examples) 349–51

1092 Index OECD Model Tax Convention beneficial ownership and tax treaties 375–78 changes of 60–61 domestic law 58–60 language and double taxation treaties 253 limits clause 369–70 qualification conflicts and tax treaties 286–87 R1-R2 tax treaty 310, 311 see also BEPS Action 2 Report; and under charities in tax conventions OECD Model Tax Convention and transfer pricing 517–33 associated enterprises (interpretation) 520–27 domestic law and treaty law 527–28 recharacterization 528–30 tax avoidance 531–33 treaty law 527–28 OECD Partnership Report (1999) 287, 438–44, 445–49 allocation conflicts 442–43 limitation and criticism 443–44 qualification conflicts 442–43 tax treaty entitlement 441–42 OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administration 535–53 arm’s length principle (ALP) 538–41 comparable uncontrolled price (CUP) method 541–42 cost-plus method (CPM) 542 development of transfer pricing 536–38 dispute resolution 550–53 disputes 546–49 functional analysis 539–40, 542, 544 intangibles 544–45 resale price (RP) method 542 transactional net margin (TNM) 543 transactional profit split (TPM) 543 OECD VAT Guidelines see EU VAT law and OECD VAT Guidelines OECD VAT Recommendation 483 off-balance sheet adjustment 569 offshore centres 27–28, 144, 396–97

offshore indirect transfers of assets 138–39 Olbert, M. 1054–55 one-stop shop 484, 758, 782, 987–88, 1046–47, 1048–49, 1055, 1056–57 on-the-spot supplies 788 opinio juris 69, 70, 75, 77–78, 82–83, 84, 85–86, 92, 93–94 opinio necessitatis 69, 75, 84 optimal tax theory 246 optimization 239–40 opt-in and opt-out 1015 optional system 435 Organisation for European Economic Co- operation (OEEC) 49, 477 origin principle 795 origin of wealth 237–38 Ottawa Taxation Framework (OTF) 482–83, 789, 790 ownership clauses 626 ownership interests 369–70 ownership neutrality (CON) 242, 243 own resources of budget of EU 667–7 1 pacta sunt servanda 70–7 1, 83, 412–13, 414, 527 Pan African Parliament (PAP) 469n.12 Panama 369–70 Panama Papers 145, 147 Papyrus, W. 7 Paradise Papers 145, 147 parent-subsidiary scenario 738–40 Pareto principle (80-20 rule) 39 participation exemption 48, 204, 324 partnerships 210–11, 433–53 domestic taxation 434–37 classification 434–35 fiscally transparent partnerships 436–37 foreign partnerships classification 435–36 guaranteed payments 437 EU 449–53 alternatives 452–53 Anti-Tax Avoidance Directive (ATAD) (EU) 452 Parent-Subsidiary Directive (P-S Directive) 451 primary law 449–50, 452 Royalties Directive 451

Index 1093 secondary law 450–53 symmetrical approach 436 tax treaties 437–49 guaranteed payments 447–49 treaty entitlement of partnerships 438–40 see also hybrid entities passive income 26–27 consensus international tax law (CITL) 98–99 double taxation avoidance 321, 324 FATCA and AEOI (USA) 146–47 international tax regime (ITR) 155 private international law 163–64 regional double tax treaties 469–70 single-tax principle 144–45 triangular cases 315 pass-through mechanism 210 patent boxes 149 payment arrangements 762–63 payroll taxes 979 peer review and monitoring 182, 488, 549–50, 551, 704, 1015–16, 1018 peer-to-peer networks and blockchain technology 1042–43, 1050 pensions 465 permanence 6–13 Permanent Court of International Justice 92 permanent establishment (PE) 12–13, 15–16, 825 1920s compromise to 2020s compromise 23, 24–25, 29–30, 40, 41, 42 agents 497–99, 500–4, 507–8, 509, 512, 514 artificial avoidance prevention 150 BEPS 2.0 153–54 construction 858 customary international law 80, 81–82 digital 38 digitalization and VAT in EU 1053–54 emerging tax consensus challenges 993– 94, 1005 EU ‘white supplies’, double taxation and VAT 769 and head office (HO) 735–37 international trade law 228 labour taxation and automation/ AI 1038–40

low-and middle-income countries (LMICs) 125, 130–31, 133–34, 138 multilateralism 488–89 Pillar One 952 Pillar Two 982, 987 profits attribution 510–13 public international law 181 transfer pricing 526–27, 535 virtual 1001 see also under India; reverse permanent establishment cases; and under triangular cases and tax treaties Permanent Establishment Report 538 personal data 1051–52 personal income tax (PIT) 23, 143–45, 155, 269, 335, 979 personality 767 personality, active and passive 66 personal nexus 107–22 national tax systems 93–94 objective criteria 117, 122 observations 122 subjective criteria 117 sufficient connection with individual 109–10 territorial taxation of individual 119–22 worldwide taxation of individual 110–19 personal property 17–18 Philippines 1020–21 physical presence 37, 42 Picciotto, S. 904, 905, 1001 Pickering, A. 468 Pillar One 11–12, 951–67, 969–7 1 1920s compromise to 2020s compromise 34–35, 36–38, 39, 40, 41 agents 513 Amount A 951, 955, 956–58, 959–61, 962– 66, 983, 995, 1000 double taxation elimination 957–58 nexus and revenue-scoring rules 956–57 scope 956 tax base allocation 957 tax base calculation 957 Amount B 951, 955, 958, 965 implementation 958 quantum 958

1094 Index Pillar One (cont.) scope 958 BEPS 2.0 153–54 compromise as main policy concern 959–65 context prior to proposal 952–55 digitalization and VAT in EU 1053–54 double taxation avoidance 319–20, 328– 31, 334 elements of proposal 955–58 emerging tax consensus challenges 995–97, 998–99, 1000, 1003–04 feasibility 965–67 global tax governance 1013, 1017–18, 1019 low-and middle-income countries (LMICs) 131–32, 141 multilateralism 492 private international law 171–72, 173 public international law 194–95 tax treaties 50 transfer pricing and formulary apportionment 540, 541, 571–72, 575, 576–77, 578, 581–84, 585 Unified Approach 951–52, 954–55, 965–66 Pillar Two 45–46, 954–55 1920s compromise to 2020s compromise 34–35, 36, 37–38, 39, 40, 41 agents 513 BEPS 2.0 153–54, 155 double taxation avoidance 319–20, 331–34 emerging tax consensus challenges 996–97, 998–99, 1000, 1001, 1003–04 global tax governance 1017–18, 1019 low-and middle-income countries (LMICs) 140 multilateralism 492 private international law 171–72 public international law 194–95 tax treaties 50 transfer pricing and formulary apportionment 585–86, 587 see also Pillar Two of OECD proposal and Global anti-Base Erosion (GloBE) Pillar Two of OECD proposal and Global anti- Base Erosion (GloBE) 969–90 background and evolution 969–72

BEPS, effects on 977–78 concept 972–74 critical design elements 982–88 ETR calculation based on consolidated financial accounts 983–85 jurisdictional blending 982–83 substance-based profits exception 985–87 top-up tax 987–88 evaluation and outlook 989–90 implementation and simplification measures 988–89 policy rationale 974–76 tax competition and tax sovereignty 978–79 tax treaty law, compatibility with 988 Pistone, P, 1001 Pitt, W. 12–13 place of importation of goods 751–52 place-of-effective management test as tiebreaker rule 309–11 place-of-supply (POS) rules 731–32, 733–34, 739–40, 746 Platform for Collaboration on Tax (PCT) 128, 137, 138, 1019 plurilateralism 475–76, 478, 488–89 Poland 668, 691, 738, 769, 773–74, 971–72 political criterion 596 political impulse 953–55 Pomerleau, K. 813–14 Portugal 338–39, 604–5, 699n.10, 723, 769 Posner, R. 77, 84, 231–32 Practice Manual on Transfer Pricing 137 pre-confirmation system (PCS) 550 preferable-destination principle 770 preferential tax rates 341 preliminary ruling procedure 594, 595, 596, 599 preparatory or auxiliary activities 858–59 primacy of EU law 610 primary law (EU) 50–51 alternative dispute resolution 722 Anti-Tax Avoidance Directive (ATAD) 692–93 anti-tax avoidance regimes 695 beneficial ownership 380–81, 388

Index 1095 Court of Justice of the EU (CJEU) 591–92, 600, 601, 602, 603–4 EU ‘white supplies’, double taxation and VAT 771 language factors 258 partnerships 449–50, 452–53 principal purpose test (PPT) 19 BEPS 1.0 148, 149, 151 charities 342 consensus international tax law (CITL) 99 double taxation 236 EU anti-tax-avoidance regimes 684 fundamental freedoms (EU) 626–27 guiding principle 392–93, 396–98 multilateralism 490 purposive interpretation 394–98 regional double tax treaties 467, 469–70 subjective element 399–403 see also principal purpose test (PPT) under tax treaty law principal purpose test (PPT) under tax treaty law 391–416 limits: Alta Energy case 410–15 objective element 392, 393, 400, 402, 403– 9, 413–14 object and purpose 406, 413, 415, 416 predictability 392–93, 414 purposive interpretation 393–98, 404–5, 410, 413–14, 416 subjective element 392, 393, 399–403, 405, 407–8, 413, 415 two-pronged test 392–93 prisoner’s dilemma 246 private companies 202 private and family life, respect for 191–92, 420, 424–27 private international law 159–75, 848 corporate law 198 domestic law and foreign law articulation 165–70 incidental questions 168–70 by judiciary 167–70 by legislature 167 questions/matters of fact 167–69 questions/matters of law 167–69 legal regulation of global economic phenomena 170–75

dispute resolution 173–75 multilateral unified substantive rules 171–73 state autonomy and connecting factors 162–63 taxpayer autonomy and connecting factors 163–64 private law 15–16, 265–66, 272, 273–74, 276– 77, 435 see also private international law private letter ruling (PLR) 550 procedural law 779 procedural rights 370–7 1 ‘Procedure 42’ 758 process and production methods (PPMs) 219–20 productivity paradox 1028–29 professors and teachers tax relief 468, 485–86 profit attribution 510–14, 858 blended 543 business 302–3, 461 distribution 826 routine and non-routine 37, 1000–01 see also residual profit profit allocation rules 33, 34, 36, 37, 38, 40, 42, 513, 563, 1053–54 profit before taxes (PBT) 957 profit shifting 39, 977–78 see also base erosion and profit shifting (BEPS); and under Japan profit splits 538, 1002 see also transactional profit split (TPM) profit transfer tax 651, 829 Programme of Work (PoW) 34 progression clause 619 prohibition provisions 665–66 property, protection of 54, 419 property rights 429–31 proportionality principle 359–60, 605–6, 620, 625, 633, 651, 657–58, 687, 761, 787 protection of legitimate expectations principle 985 protective concerns 66 proxies 793–94, 795, 953–54 public benefit rationale 342–43 public choice theory 246

1096 Index public companies 202 public interest considerations 173, 620, 651, 658 public international law 78–79, 177–95 consensus international tax law (CITL) 94 corporate law 197 dispute resolution 186–88 double taxation treaties and language factors 254 jurisdiction 65, 66 language factors 263–64 national tax systems 87–88 source taxation rights 237 substantive rules 188–93 fiscal privileges and immunities 189–90 human rights law 190–92 investment law and expropriatory taxation protection 192–93 treaties, custom and general principles 179–81 triangular cases 305–6 unilateralism, bilateralism and multilateralism 474 Vienna Convention on the Law of Treaties (VCLT) and double tax treaties 184–85 publicity of decisions 726 public law 198, 250, 271, 364, 423 see also public international law public policy 161, 364, 726, 1032–33 charities 337–38, 349, 352 exchange of information (EOI) and tax treaties 368–69 purpose limitation principle 361–62 Qatar 214–15 qualification conflicts 440–41, 442–43, 444, 447–49, 452 qualification conflicts, see also qualification conflicts and tax treaties qualification conflicts and tax treaties 285–97 interpretation 291–95 OECD MC and UN Model 286–87 tax literature 287–89 qualified business asset investment (QBAI) exemption 985 qualified domestic minimum top-up tax (QDMTT) 972–74, 980 qualified person 342, 344, 346, 407–8 quasi-rents 239–40

questionnaires 276–77 questions/matters of fact 167–69 questions/matters of law 167–69 quick fixes 744, 782 Qureshi, A. 178 Razzouk, A.W. 945 real estate income 302–3, 336 reasonableness test 88, 399 reasonable time requirement 420, 422 recharacterization 452, 528–30, 545 recharge approach 793, 796 reciprocal exemption 346 reciprocity principle 17–18 charities 342–43, 344 domestic law 58 EU ‘white supplies’, double taxation and VAT 777 exchange of information (EOI) and tax treaties 363, 366–67 principal purpose test (PPT) 412 public international law 181, 192 transfer pricing (OECD MC) 522 unilateralism, bilateralism and multilateralism 474–75 Recommendation on Tax Credits 682–83 ‘red flags’ 989 reductions 649 reference model (tertium comparationis) 277 refeudalization of tax systems 28–29 refrain or associate principle 173 regional double tax treaties 455–72 capital taxation 466 definitions 459–60 general definitions 459 permanent establishment 460 resident 459–60 undefined terms, treatment of 459 elimination of double taxation 466 emerging positions on regional basis 468–70 final provisions 468 entry into force 468 termination 468 identification 456–57 income taxation 460–66 associated enterprises 461–62 business profits 461 capital gains 464

Index 1097 directors 465 dividends 462 employment income 464–65 entertainers and sport persons 465 government service 465 immovable property income 460 independent personal services 464 interest 462–63 other income 466 pensions 465 royalties 463 shipping and air transport 461 students 465 technical services 463 professors and teachers 468 scope 458 persons covered 458 taxes covered 458 special provisions 466–68 benefits entitlement 467 collection of taxes, assistance in 467 diplomatic missions and consular posts 467 exchange of information 467 mutual agreement procedure (MAP) 466–67 non-discrimination 466 territorial extension 467 territorial extension 467 unique provisions 468 regionalism 483–86 reporting requirements 762 Report on Transfer Pricing and Multinational Enterprises 537 resale price (RP) method 541, 542, 546–47, 549 Resch, R.X. 289n.11, 290n.16 research and development (R&D) 401, 572–73 reservations against regional models 470 residence 572–73 residence-based taxation 13–14, 89–90, 149, 240–43, 866 residence jurisdiction 29–30, 295, 339–41, 345 residence principle 47, 502–3 resident test 116, 118, 119, 155 resident tiebreaker tests 114–15 residual profit 11–12, 37, 581–82, 959–60 Residual Profit Allocation by Income (RPAI) 1000–01 restricted direct use method 791–92

retail price method 537 revenue rule 166–67 revenue-scoring rule 956–57 reverse charge mechanism 482–83, 782, 788, 1056 see also generalized reverse-charge mechanism (GRCM) reverse dual resident cases 301, 315–17 R-S1 and R-S2 tax treaty 316–17 S1-S2 tax treaty 316 reverse permanent establishment (PE) cases 301, 312–15 R-SPE tax treaty 314 R-S tax treaty 313 S-SPE tax treaty 314–15 review panel 995–96 Revised Code of Conduct (2009) 705–6, 708– 9, 710, 718 Ring, D. 1008 ring-fencing 821, 832–34, 964, 975 risk assessment 488–89 Rixen, T. 147, 1007 robot taxation 1035–36 rollback clause 214–15, 633 Romania 338–39, 604–5, 674–75, 769 Rosenbloom, D. 476 round-tripping (circular arrangement) 398, 406 Roxan, I. 244 royalties 833 1920s compromise to 2020s compromise 37 beneficial ownership 373–74, 378–79, 383 charities 336, 341 digitalization and VAT in EU 1054 domestic law 48 double taxation avoidance 332 EU candidate countries 677–78 EU law 51 exchange of information 363 low-and middle-income countries (LMICs) 133–35 regional double tax treaties 463 state aid (EU) 639, 640 transfer pricing for MNEs (OECD) 546– 47, 548 triangular cases 302–3, 312, 313, 314, 315, 316–17 Ruding Report 484–85 Rukundo, S. 1002–03

1098 Index rule of law 48, 596–97, 663–64 rule of reason principle 620–21, 626 Russia 214–15, 488–89 Russian invasion of Ukraine 678 Rust Conference 222n.41 Rwanda 910 Sacco, R. 270–7 1, 272n.49 sacrifice theory 111 Sadiq, K. 95, 100, 1021–22 Saez, E. 813–14 safe harbour rules 200 1920s compromise to 2020s compromise 34, 35, 37 Anti-Tax Avoidance Directive (ATAD) 689, 694 low-and middle-income countries (LMICs) 127 Pillar Two 989 principal purpose test (PPT) 409–10 sales tax 577–78 Santayana, G. 6–7 saving clause 446–47 Schlesinger, R. 276–77 Schön, W. 224 Schoueri, P. 875n.58 Schreiber, U. 912 secondary law (EU) anti-tax-avoidance regimes 679–80, 694–95 beneficial ownership 380–81 language factors 258 partnerships 450–53 state aid 630 secondary response see linking rules secondary rule 690 secrecy 726 clause 361, 370–7 1 professional 370, 426–27 state 368 trade or business 364, 367–68 see also bank secrecy Seiter, G.M. 813–14 seizure of property 420 selective advantage 637, 639, 642, 644, 649 selectivity 631–32, 641–42, 650, 684–85 de facto 632 de jure 632

self-interest 77, 83, 84, 86, 474 Seligman, E.R.A. 9, 14, 22, 477 Sender, O. 74–75 separate accounting method 519n.7, 536–37 separate entity approach 98–99, 126–27, 436– 37, 537, 540, 541, 998–99 Serbia 645, 663–64, 678 serious misconduct 718–19 services 732, 793–94, 796, 1047 services and intangibles and place of taxation see under EU VAT law and OECD VAT Guidelines Shackelford, D.A. 930–31 shadow economy 145–46 sham or simulated operation 394n.8, 396– 97, 406 Shaviro, D.N. 243 Shay, S.E. 911–12 shell entities 146–47, 681–82, 687 shipping and air transport 461 Shome Committee 850 Sierra Leone 14 significant economic presence 30, 34, 131–32, 825–26, 959, 995, 1002, 1005 significant influence 525 similarity approach 435 simplicity 789–90, 793–94, 795, 953–54, 1035–36 simultaneous examination 363 Singapore 488–89 Singh, Dr M. 850n.41 single taxation 48, 319, 335–36, 339, 352 single-tax principle (STP) 144–45 1920s compromise to 2020s compromise 26, 27 BEPS 1.0 148, 149, 151 BEPS 2.0 153–55 customary international law 81–82 international tax regime (ITR) 155 TCJA and BEPS 1.0 152, 153 unilateralism, bilateralism and multilateralism 476 Sixdorf, F. 909 Skandia 740, 741–42, 743 Slemrod, J. 919n.5 Slovakia 668, 1030–32 Slovenia 668

Index 1099 ‘smart-contracts’ 745 Smith, A. 107, 111 social media platforms, search engines and online shopping malls 825–26, 959, 995, 1051 social security contributions 52, 780–81, 1025–26, 1035–37 soft law 45–46, 1014–16 comparative tax law 280 consensus international tax law (CITL) 99–101, 102 Court of Justice of the EU (CJEU) 592–93, 594 customary international law 83–84, 85 EU anti-tax-avoidance regimes 682–84 EU anti-tax-avoidance regimes 694–95 EU/OECD VAT/GST comparison 787–88 EU ‘white supplies’, double taxation and VAT 775, 777 global tax governance 1008, 1011, 1022–23 jurisdiction 70 multilateralism 488 national tax systems 103 private international law 162–63 unilateralism, bilateralism and multilateralism 477 SOLVIT 779–80, 783 source 29–30, 47, 502–3 source-based taxation 13–14, 81, 89–90, 149, 150, 237–40 source codes 745 source jurisdiction 295 source rules, statutory 121–22 sources of international law 79 sources of law and legal methods 45–63 domestic tax law 47–48 European law (material law) 50–52 exchange of information and cooperation of tax administrations 52–54 interpretation of treaties 55–62 interpretation of treaties, domestic law 56–62 tax treaties 48–50 source state principle 767 South Africa 459, 470, 994 South Asian Association for Regional Cooperation (SAARC) 485–86

Southern African Customs Union 477 South Korea 550–51, 972 Spain agents 511–12 charities 337–39 comparative tax law 267–68, 278–79 Court of Justice of the EU (CJEU) 599, 604–5 EU candidate countries 668 EU ‘white supplies’, double taxation and VAT 769 exchange of information 353–57 multilateralism 493 partnerships 436 special-purpose entities 505–6 special tax regimes 341, 344, 345–46 specialty principle 359, 360, 361–62 specific anti-avoidance rule (SAAR) 530, 694–95, 892–93 principal purpose test (PPT) 401–2, 405, 407–9 specific rules 792 Spengel, C. 1054–55 ‘splitting’ 278–79 Stamp, Sir J. 22 standing committee 778 standstill obligation 214–15, 633, 650 Starbucks 148, 639–41 state aid 629–44 advance pricing arrangements (APAs) 637–43 Anti-Tax Avoidance Directive (ATAD) 692–93 EEA/EFTA 647, 648–50, 661 EU anti-tax-avoidance regimes 684–85, 694–95 EU candidate countries 671–72 EU law 51 illegal 634, 642 infringement procedures 954–55 regionalism 484 tax competition 633–35 tax rulings 635–37 stateless income 489 state of permanent establishment (SPE) taxes 312, 313 statistical analysis 359

1100 Index status updates 356 Steichen, A. 244 stolen data 364–66, 368–69 students and tax relief 465, 485–86 sub-federal entities 573–74, 581, 584 subject to tax rule (STTR) 36, 37–38, 154–55, 332–33, 334, 373–74, 974, 988, 996–97 sub-Saharan Africa 139 subsequent agreements/subsequent practice 56 subsidiaries 48, 503–4 subsidiarity principle 366, 483, 593–94, 664– 65, 687, 787 subsidies 642, 643 substance-based income exclusion (SBIE) 972–73, 979–81, 985, 986, 987 substance-based profits exception 985–87 substance or motive test 605–6 substance-over-form doctrine 383, 384, 387 substantive rights 370–7 1 sufficient connection 109–11, 118, 119, 121, 122 supply chain restructuring 135–36 supremacy 667 Surmatz, H. 338–39 sustainable development 218–19 Sustainable Development Goals (SDGs) 127, 128, 129, 141, 966, 1019–20, 1023 Sweden 11–12 beneficial ownership 384–85 EEA/EFTA 645, 655 EU and cross-border VAT 740, 741–42 exchange of information 369–70 human rights law 420 partnerships 436 regionalism 484–85 unilateralism, bilateralism and multilateralism 478–79 switchover rules 323–26, 331–32, 333–34, 618n.56, 987, 988 Switzerland 15–16, 805 beneficial ownership 384 EEA/EFTA 645, 646, 655, 660–61 EU ‘white supplies’, double taxation and VAT 769 exchange of information 354, 355, 356–59, 360, 362, 364–66, 369–70 human rights law 420, 420–21n.13

private international law 169–70 transfer pricing and formulary apportionment 552–53, 573n.6 systemic integration 392–93 Tajika, E. 924 tangibles 576–77, 985, 987 Tanzania 904 tariffs 213, 214–16 taxable amount 755–57 taxable person 761–62 tax administration 528–30 tax arbitrage 489 tax avoidance/tax avoidance prevention 18– 19, 45–46, 821, 823, 834 1920s compromise to 2020s compromise 43 beneficial ownership 375, 378, 379, 380–81, 383, 386–87 BEPS 1.0 150 consensus international tax law (CITL) 97–98 corporate law 201 Court of Justice of the EU (CJEU) 594, 605–6 double taxation avoidance 324–25, 326, 328 EEA/EFTA 652 EU anti-tax-avoidance regimes 682–83 EU candidate countries 667, 671 fundamental freedoms (EU) 609, 615, 622, 627 international tax regime (ITR) 155 international trade law 228–29 low-and middle-income countries (LMICs) 129–30 multilateralism 489, 490 national tax systems 103, 105 OECD MC and transfer pricing 531–33 personal nexus 112 Pillar Two 969, 984–85 principal purpose test (PPT) 394–95, 396–97, 399–400, 404–5, 410, 413–14 public international law 185 qualification conflicts 286–87 regionalism 483 single-tax principle 145 TCJA and BEPS 1.0 151 transfer pricing 536, 544

Index 1101 transfer pricing and formulary apportionment 584–85 tax base allocation 957 tax base calculation 957 tax base erosion and profit shifting (BEPS) 822, 823–24 tax base reductions 341 tax burdens 245 tax competition 1920s compromise to 2020s compromise 27–28, 39, 40 asymmetric 246–47 economic analysis of law 244–47 global tax governance 1008 Pillar Two 978–79 private international law 162–63 state aid 633–35 see also harmful tax competition tax conventions see charities in tax conventions Tax Cuts and Jobs Act 2017 (TCJA) (USA) 151–53, 154, 806–11 evaluation 809–11 international incentives before and after Act 807t key provisions 806–9 Tax Directives and anti-avoidance rules 679–82 Tax Dispute Resolution Directive 712–26 Achmea 721–23 Advisory Commission (AC) 715–18 complaint 713–15 dispute resolution procedure 715–18 European Commission, role of 726 mutual agreement procedure (MAP) 715 publicity of decisions 726 scope 712–13 taxpayer role 718–21 Alternative Dispute Resolution Commission (ADRC) 720 decision-making 720 denying access to dispute resolution procedure 718–19 interaction with national proceedings 719 interpretation of directive 719–20

relationship with proceedings under other instruments 721 special rules for individual and smaller companies 721 taxpayers’ rights 723–25 taxes follow profits principle 982 tax evasion/tax evasion prevention 17– 19, 45–46 1920s compromise to 2020s compromise 22, 27–28 agents 499–501 beneficial ownership 375, 379, 383 consensus international tax law (CITL) 97–98 Court of Justice of the EU (CJEU) 605–6 digitalization and VAT in EU 1042, 1053 EU anti-tax-avoidance regimes 682–83 EU candidate countries 667, 671, 672 FATCA and AEOI (USA) 145–46, 147 international tax regime (ITR) 155 international trade law 228–29 language factors 250 multilateralism 490 principal purpose test (PPT) 394–96, 404–5, 413–14 public international law 181 qualification conflicts 286–87 single-tax principle 145 VAT Directive and imports 759 tax examination abroad 363 tax exemption through free zones 139 tax expenditure schemes 984 tax havens 822, 833 1920s compromise to 2020s compromise 27–28 consensus international tax law (CITL) 97 exchange of information 354 international tax regime (ITR) 155 international trade law 226–27 multilateralism 489 see also offshore centres tax holidays 139 tax incentives 139–40, 269, 1020–21 tax information exchange agreements (TIEAs) 109–10, 128–29, 147, 353, 354 Tax Inspectors Without Borders (TIWB) 136–37

1102 Index Tax Justice Network 1013–14 tax opacity 301, 434–36, 438, 439–42, 445–46, 449–50, 452 taxpayer role see under Tax Dispute Resolution Directive taxpayers’ rights 370–7 1, 709–10, 723–25 tax reduction 565 tax relief 650 double 444, 446–47 tax remission 565 tax residency mismatches 690 tax rulings 635–37 tax shaming 29 Tax and Social Security Procedure Code (TSSPC) 663–64 tax sovereignty 825, 835 1920s compromise to 2020s compromise 27–28, 32, 40–41, 42–43 consensus international tax law (CITL) 98–99 Court of Justice of the EU (CJEU) 592, 603 customary international law 81 double taxation treaties and language factors 255 EU candidate countries 664–65 EU/OECD VAT/GST comparison 787 global tax governance 1008 international trade law 229 multilateralism 495 national tax systems 88 personal nexus 109 Pillar Two 978–79 private international law 166–67, 173 state aid 643 unilateralism, bilateralism and multilateralism 474, 475 value creation 940–41 tax status of taxpayer or of procedure 356 tax treaties 15–16, 48–50, 55 abuse 99 agents see agents in tax treaties beneficial ownership see beneficial ownership and tax treaties entitlement 441–42 exchange of information see exchange of information (EOI) and tax treaties

fundamental freedoms see fundamental freedoms and tax treaties human rights law see human rights law and tax treaties network 80–83 objective method 55 policy 885–86 principal purpose test (PPT) see principal purpose test (PPT) under tax treaty law qualifications conflicts see qualification conflicts and tax treaties regional see regional double tax treaties triangular cases see triangular cases and tax treaties tax treaty arbitration (TTA) 552–53 tax uncertainty 38, 40 technical services 463 technological advancement 992–93, 995, 1001 telecommunication services 792, 1049, 1055 territoriality principle 66, 615 EU ‘white supplies’, double taxation and VAT 765, 766, 767, 770, 777–78 territorial taxation 110, 119–22, 323, 923– 24, 927 TFEU principles 748–49 thin capitalization rules 199–201, 519–20, 573, 689, 894–95 third countries Anti-Tax Avoidance Directive (ATAD) 690, 692–93 Court of Justice of the EU (CJEU) 591, 605–6 digitalization and VAT in EU 1046–48 EEA/EFTA 646, 655–57 EU anti-tax-avoidance regimes 682–83, 695 EU law 50–52 exchange of information 362 fundamental freedoms (EU) 617, 626–27 VAT Directive and imports 749, 752 third parties agents 503–4 beneficial ownership 383, 384 corresponding adjustments and transfer pricing 561, 567 EU alternative dispute resolution 709–10 fundamental freedoms (EU) 625

Index 1103 regional double tax treaties 471 transfer of information 359–60 transfer pricing and formulary apportionment 572–73 third states Anti-Tax Avoidance Directive (ATAD) 693–94 charities 343, 344 EU alternative dispute resolution 708–9 exchange of information 356, 360 principal purpose test (PPT) 394–95, 408 regional double tax treaties 469– 70, 471–72 Thomas, C. 82 threshold requirement 39 Thucydides 5–7 Thuronyi, V. 272 Tiebout, C.M. 246 Tiebout model 246 tiebreaker rule 356, 485–86 Tinhaga, Z.P. 905 tit-for-tat cases 214–15 top-up tax 972–73, 979–80, 981, 982, 983, 984– 85, 986–89, 996–97 see also qualified domestic minimum top- up tax (QDMTT) trade law 213–29, 477 direct discrimination and trade in goods 220–25 export subsidies 222–25 trade-related investment measures 221–22 direct discrimination and trade in services 225–28 indirect discrimination and border tax adjustments (BTA), excessive 216–20 tariffs and similar duties 214–16 trade-related investment measures 221–22 trade in services 213 training and tax relief 485–86 transactional net margin (TNM) 541, 543, 549, 577, 640–41 transactional profit split (TPM) 536, 540, 541, 543 transaction costs 31, 477–78 transaction value method 755–56 transfer pricing 45–46, 825–26

1920s compromise to 2020s compromise 30, 31, 37 agents 510–12 audits 579–80 BEPS 1.0 150, 153 consensus international tax law (CITL) 97, 99–100 corresponding adjustments see corresponding adjustments Court of Justice of the EU (CJEU) 606 digitalization and VAT in EU 1054 double taxation avoidance 319, 328–29 emerging tax consensus challenges 1002 EU alternative dispute resolution 704–5, 706, 709, 711 EU law 52 EU/OECD VAT/GST comparison 796 foreseeable relevance of request 356–57 global tax governance 1020 human rights law 423–24 jurisdiction 69–70 labour taxation and automation/ AI 1035–36 low-and middle-income countries (LMICs) 126–27, 135–37 multilateralism 488–90, 491 Pillar One 959, 965 principal purpose test (PPT) 401–2 private international law 166–67, 168 report 640 state aid (EU) 637–38, 639, 642–43 value creation 947, 948–49 see also OECD Model Tax Convention and transfer pricing; OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administration; transfer pricing and formulary apportionment transfer pricing and formulary apportionment 571–87 arm’s-length range 578–80 BEPS Project 584–86 income apportionment, net vs gross 575–78 institutional background 573–75 neo-protectionism 586–87 procedural issues 580–84

1104 Index Transfer Pricing and Multinational Enterprises Report (1979) 520 transparency 211 1920s compromise to 2020s compromise 31 agents 505–6 corporate law 210 EU alternative dispute resolution 710, 720 EU anti-tax-avoidance regimes 679 EU candidate countries 672 exchange of information 353, 369 fiscal 210–11, 212 global tax governance 1020–21 state aid (EU) 633–34 transfer pricing for MNEs (OECD) 549–50 see also fiscal transparency treaty shopping 386–87, 388, 394–95, 404–5, 406, 408, 413–14, 416 1920s compromise to 2020s compromise 28 agents 514 charities 342 EU alternative dispute resolution 700 fundamental freedoms (EU) 622, 626 low-and middle-income countries (LMICs) 137–38 triangular cases and tax treaties 299–318 definition 299–302 dual residence cases 309–11 R1 and R2 and S treaties 310–11 R1-R2 treaties 309–10 permanent establishment (PE) cases 301, 302–9, 317–18 immovable property income 306–9 R-PE treaties 304–6 R-S treaties 303–4 PE-S treaties 303, 305–6 R 312, 315 R1 and R2 and S treaties 310–11 OECD Model (2017) 311 place-of-effective management test as tiebreaker rule 310–11 R1 and R2 treaties 309–10 OECD Model (2017) 310 place-of-effective management test as tiebreaker rule 309–10 R1 and S1 treaties 315 R1 and S treaties 310, 311 reverse dual resident cases 315–17

reverse permanent establishment cases 312–15 R and PE treaties 303, 304–7, 308 R and S1 treaties 316–17 R and S2 treaties 315, 316–17 R and SPE treaties 313–14 R and S treaties 303–4, 310, 313, 314 R treaties 306 S1 treaties 315, 316 S2 treaties 315, 316 S and SPE tax treaties 312, 314–15 state of permanent residence (SPE) 312 state R 302–3 state R1 309 state R2 309 state S 302–4, 309, 312, 313 TRIMS Agreement 221, 222n.41, 228–29 true recipient 373–74 Trump, D. 146, 954–55 Trump Tariffs 214–15 trustees 370 trusts 373 Turkey 493, 645, 663–64, 668, 678, 754, 994 turnkey projects 130–31 turnover taxes 45, 519n.7, 577–78 Court of Justice of the EU (CJEU) 602 digitalization and VAT in EU 1044, 1045, 1053–54 EU and cross-border VAT 731–32 EU/OECD VAT/GST comparison 787 2020s compromise see 1920s compromise and the 2020s compromise two pillar solution 825 1920s compromise to 2020s compromise 34, 36 emerging tax consensus challenges 995, 1002 multilateralism 492–94 public international law 182, 195 see also Pillar One; Pillar Two UBS scandal 146–47 Uganda 902, 904, 906 Ukraine 214–15, 663–64 ultimate parent entity (UPE) 549, 995–97 unanimity principle 598–99

Index 1105 uncontrolled transactions 523, 546–47 UNCTAD 990 underpaid and unpaid labour 944–46 under-taxation avoidance 178–79 under-taxation and economic activity 631 undertaxed payments rule (UTPR) 1920s compromise to 2020s compromise 37–38 BEPS 2.0 154–55 emerging tax consensus challenges 996–97 multilateralism 492 Pillar Two 973, 974, 983, 987–89 Unified Approach (OECD) 34, 132, 1013 uniform application 262–63 unilateral advance pricing agreements (UAPAs) 551 unilateralism 824 1920s compromise to 2020s compromise 32, 38 agents 505–6 constructive 104 customary international law 80–81 global tax governance 1007, 1010 jurisdiction 66 Pillar One 954–55 tax treaties 48 see also unilateralism, bilateralism and multilateralism unilateralism, bilateralism and multilateralism 473–95 bilateralism 473–80, 484, 487, 490–91, 492, 495 multilateralism 473–83, 484, 486–95 model treaties 480–81 standards and tax policy transfer: VAT and e-commerce 482–83 regionalism 483–86 unilateralism 473–80, 487, 490, 491–92, 493–94, 495 unilateral taxation 69, 834, 969–70 unitary taxation 30–31, 109–10, 901–5 United Kingdom 11–12 1920s compromise to 2020s compromise 23n.5, 42 agents 504–5

automatic residence tests and automatic overseas tests 116, 118 beneficial ownership 373–74, 376–77 Brexit 146 charities 337–38 comparative tax law 271–72, 276, 277–79 corporate law 199, 200–1, 203, 206, 207n.52, 208, 209 Danegeld (land tax) 12–13 diverted profits tax (DPT) 93 double taxation avoidance 324 EEA/EFTA 651, 660 emerging tax consensus challenges 995 EU and cross-border VAT 737 Google tax 824–25 His Majesty’s Revenue and Customs (HMRC) 381–82 human rights law 426n.27 multilateralism 493 partnerships 435 PAYE 104–5 personal nexus 114, 114n.28, 115, 118, 119 Pillar One 959 Pillar Two 972 private international law 165 public international law 183 regionalism 484–85 Supreme Court 101 territorial system 923 transfer pricing for MNEs (OECD) 546–47 user contribution approach 131–32 withdrawal clause 596–97 United Nations 18–19, 48–49, 69, 78, 250 United Nations 12B plan 829–30 United Nations Ad-hoc Group of Tax Experts 126 United Nations Committee of Experts 129, 481, 512, 997–98 United Nations Financial Accountability, Transparency and Integrity Panel 908 United Nations General Assembly (UNGA) 78, 129 United Nations Handbook on Selected Issues in Protecting the Tax Base of Developing Countries 1013–14 United Nations Millennium Project 124

1106 Index United Nations Model Double Taxation Convention charities in tax conventions 344 qualification conflicts and tax treaties 286–87 United Nations Tax Committee 97–98, 134, 137, 139, 194, 1013 Practical Manual on Transfer Pricing (UN Manual) 127 United States 125, 132, 799–819, 901 35 percent rate 800 1920s compromise to 2020s compromise 23n.5, 32, 33, 34, 35, 42 Affordable Care Act 146 agents 511–12 anti-inversion provisions 815, 819 arm’s length principle (ALP) 817 base erosion and anti-abuse tax (BEAT) 808, 809, 812–13, 815 beneficial ownership 382–83 benefits 13–14 BEPS 1.0 148 BEPS 2.0 153, 155 border-tax adjustment 818 charities 337, 350 check-the-box mechanism 154, 209, 435, 803 comparative tax law 268, 276, 278–79 competition 13–14, 804 cooperation 815–16 corporate alternative minimum tax (CAMT) 972 corporate law/corporate tax 199, 200–1, 203, 206, 207n.52, 208, 209, 800–1, 800f, 802f, 806–7, 810–11, 812, 813 country-by-country administration 813, 814, 815 destination-based cash flow 817, 818 direct investment earnings 809–10 Double Irish Dutch Sandwich 146 double non-taxation/double non-taxation prevention 13–14, 818 double taxation/double taxation prevention 13–14, 818 effective tax rates 801 emerging tax consensus challenges 992, 995, 999–1000, 1003 EU/OECD VAT/GST comparison 786 export earnings subsidy 808–9 Foreign Account Tax Compliance Act (FATCA) 53, 145–47, 354, 420–21n.13, 487

foreign business income 799–800 foreign-derived intangible income (FDII) 808–9, 812, 814 formulary apportionment 817–18 generally accepted accounting principles (GAAP) 694 global intangible low-taxed income (GILTI) 33n.69, 151–54, 155, 324–25, 808, 809, 811–12, 815, 969–70, 972, 982, 985, 986–87 global tax governance 1015n.38 green card test 112 human rights law 420–21n.13 income (Subpart F income) 799–800 Inland Revenue Code 152–53 intangibles 817 intellectual property 814–15, 817 Internal Revenue Code 345, 537 Internal Revenue Service (IRS) 147, 354, 357, 517–18, 547–48, 552–53, 925 international tax regime (ITR) 155–56 international trade law 215n.7, 221, 223–24 Joint Committee on Taxation (JCT) 806, 809, 813–14 labour taxation and automation/AI 1027– 28, 1030–32 locations for MNC profit 805f management and control test 815 marketing intangibles approach 34, 131–32, 825–26, 959 Massachusetts formula 574, 576, 581 minimum tax 808, 812–14, 815–17, 819 multilateralism 486–87, 490, 493 national tax systems 104 neutrality 13–14 OECD countries 816 offshoring 801–3, 812, 813 one-time repatriation tax 806 partnerships 435 pass-through form 801 payroll, assets and sales (three-factor formula) 817–18 personal nexus 112–13, 121–22 Pillar One 817, 954–55, 959–61, 965–66 Pillar Two 969–70 portfolio interest exemption 154 potential future reforms to tax law 811–19 international coordination, importance of 815–16

Index 1107 long-term reforms 816–19 short-term reform plan 812–15 private international law 166–67, 168, 170 profit shifting 805–7, 808, 809, 815 purchasing power parity 804 R&D tax treatment 814–15 repatriation holiday 802 repatriation tax 802–3, 808, 809 Restatement (Fourth) of the Foreign Relations Law 65–66 ‘Section 301’ investigation 829, 830–31 single-tax principle 144–45 source-based taxation 13–14 special tax treatment 804 statutory tax rate 799–800, 801, 803 substantial presence test 116 taxation before Tax Act 2017 799–806 tax avoidance 803, 804, 817 tax base erosion due to profit shifting 803, 804, 811, 812 tax havens 805t, 809–10 Tax Policy Center 813–14 tax sanctuaries 801, 804–5, 810f, 811, 813, 814, 817–18 tax treaties 49 territorial system 806, 808 trade protectionism 837 transfer pricing 522, 537–38, 544n.46, 546, 547–48, 550–51, 553 transfer pricing and formulary apportionment 571, 573n.8, 573n.9, 574, 574n.14, 575, 581, 583, 586 Treasury 82, 813–14 under-taxation 812, 813 undertaxed profits rule (UTPR) 816 unilateralism, bilateralism and multilateralism 473, 479 value creation 942–43n.14, 943n.17 War Excess Profits Tax 517–18 Ways and Means Committee (Camp) 337, 805–6 ‘worldwide’ system 800, 803 WTO obligations 808–9, 814 see also Tax Cuts and Jobs Act 2017 (TCJA) (USA) United States Model Income Tax Convention, see under charities in tax conventions universal basic income (UBI) 1033–34

Universal Declaration of Human Rights 190–91 user participation 34, 825–26, 959, 995 value chains, value platforms and value stores 823–24 value consumption 826 value creation 826, 827, 937–50, 959, 961–62, 966 1920s compromise to 2020s compromise 33 BEPS 1.0 150 definitional ambiguity 940–41, 949–50 digitalization and VAT in EU 1051–52 emergence and intrigue 938–41 emerging tax consensus challenges 998, 999–1000 geo-location defiance 946–49 inconsistently identifiable and quantifiable 944–46 as legislated construct 941–44 Pillar Two 978, 980 principal purpose test (PPT) 415–16 source taxation rights 237–39 transfer pricing (OECD MC) 529–30 Van Hoorn, J. 338, 352 Vann, R. 498n.4 van Weeghel, S. 105n.107, 401n.56 VAT 45, 825 ad valorum tax system 941 chargeability and chargeable event 752–53 comparative tax law 268–69 Court of Justice of the EU (CJEU) 601 domestic tax law double 779 EEA/EFTA 646, 658 EU candidate countries 665–66, 667, 669, 670, 676 EU law 51 fraud 1050 fundamental freedoms (EU) 623 grouping regime and cost-sharing arrangements 743 identification number 782 international tax regime (ITR) 143 international trade law 215–16 multilateralism 482–83 value creation 941 see also digitalization and VAT in EU; ‘white supplies’, double taxation and VAT in EU; VAT in EU

1108 Index VAT Committee 779–80, 783, 784 VAT Cross Border Ruling (CBR) 781, 783 VAT Directive 749–63 chargeable event and chargeability of VAT 752–53 deduction, right of 759–61 entry of non-union good into economic network as importation requirement 753–55 exemptions on importation 757–59 payment arrangements 762–63 place of importation of goods 751–52 reporting requirements 762 subject matter and scope 749–50 taxable amount 755–57 taxable person 761–62 VAT Directives (1967-2006) 1044–47 VAT Directives (2006-2020) 1047–48 VAT in EU 731–41 cross-border context 731–32 fixed establishment 733–34 see also Court of Justice of the EU (CJEU) and fixed establishment Végélyté, E. 904–5 Vienna Convention on the Law of Treaties (VCLT) 55–56, 57, 59–60, 61, 184–85 Villiger, M. 77–78 Vogel, K. 102–3, 119–20, 131, 396n.23, 526 von Hindenburg, P. 881 Von Hippel, T. 338 von Schanz, G. 9 Ward, D.A. 58 Waring, M. 944–45 Waris, A. 903 war profits tax 11–12 Watts, A. 67 wealth tax 4 website may constitute a permanent establishment (PE) 858 Welmory 738 West African and Monetary Union 471 Western Balkans 678 whistle-blowing 147 ‘white income’ (non-taxed) 768 White Paper on levelling the playing field as regards foreign subsidies 642 ‘white supplies’, double taxation and VAT in EU 765–84

harmonization and double taxation prevention 781–83 harmonization in indirect taxation 771–75 legislative and administration cooperation VAT disputes 775–81 non-taxation or double taxation 768–7 1 wholly artificial arrangements 624, 625, 652 wilful default and gross negligence 709, 718 withholding tax 825 beneficial ownership 377, 378, 384 charities 335–36, 341, 343–44 corresponding adjustments and transfer pricing 569 EEA/EFTA 654–55 emerging tax consensus challenges 993–94, 996–98, 1000 EU law 51 FATCA and AEOI (USA) 147 human rights law 428–29 multilateralism 492 Pillar One 964–65 Pillar Two 973–74, 982 regional double tax treaties 462 tax treaties 49–50 Wittendorff, 528n.43 Wolfke, K. 76, 82–83 Wood, M. 74–75 World Bank 124, 129, 136–37, 140, 250, 1019 world income principle 559 world tax court 264 World Trade Organization (WTO) 54, 213, 214, 215, 218–20, 226–27, 228–29 dispute resolution 221–22, 478 EU import VAT 748 language factors 250 Panel 221 unilateralism, bilateralism and multilateralism 477, 482 worldwide taxation of individual 110–19 Xi Jinping 822–23 Zambia 136–37, 747n.1 Zamora, S. 78–79 Zucman, G. 809–10, 813–14 Zweigert, K. 272n.49, 275n.64