The Internal Market 2.0 9781509939039, 9781509939060, 9781509939053

Citation preview

THE INTERNAL MARKET 2.0 This edited volume brings together leading authors and actors in EU internal market law and policy, revisiting the classic themes in a contemporary context and considering (re-)directions for the future. The EU would not be where and what it is today without its internal market. It is the cradle of the EU’s most important legal doctrines and the source of the most significant amount of European integration. And, as Brexit has underlined, it remains the primary political reason for EU membership. Considering the well-established and fundamental nature of internal market law, it is striking to find many crucial doctrinal questions still unanswered today, as explored by this book. Furthermore, these questions now find a new legal, social and political context: one that is acutely aware of the contested nature of the EU and its policies and the need to embed the internal market project in a broader setting of constitutional norms and values. This need is made all the more pressing by the rapidly changing and often disruptive technological context. This book contributes to finding a new direction for continued European integration in changing times, by rethinking, and where necessary reinventing, the role and purpose of this area that remains the EU’s beating heart. Volume 102 in the Series Modern Studies in European Law

Modern Studies in European Law Recent titles in this series: Administrative Regulation Beyond the Non-Delegation Doctrine: A Study on EU Agencies Marta Simoncini The Interface Between EU and International Law: Contemporary Reflections Edited by Inge Govaere and Sacha Garben The Rise and Decline of Fundamental Rights in EU Citizenship Adrienne Yong The Court of Justice and European Criminal Law: Leading Cases in a Contextual Analysis Edited by Valsamis Mitsilegas, Alberto di Martino and Leandro Mancano The EU as a Global Regulator for Environmental Protection: A Legitimacy Perspective Ioanna Hadjiyianni Citizenship, Crime and Community in the European Union Stephen Coutts Critical Reflections on Constitutional Democracy in the European Union Edited by Sacha Garben and Inge Govaere Constitutional Law of the EU’s Common Foreign and Security Policy: Competence and Institutions in External Relations Graham Butler The Juridification of Individual Sanctions and the Politics of EU Law Eva Nanopoulos Sixty Years of European Integration and Global Power Shifts: Perceptions, Interactions and Lessons Edited by Julien Chaisse Fundamental Rights and Mutual Recognition in the Area of Freedom, Security and Justice: A Role for Proportionality? Ermioni Xanthopoulou Law and Judicial Dialogue on the Return of Irregular Migrants from the European Union Edited by Madalina Moraru, Galina Cornelisse and Philippe De Bruycker Framing Convergence with the Global Legal Order: The EU and the World Edited by Elaine Fahey EU Citizenship at the Edges of Freedom of Movement Katarina Hyltén-Cavallius For the complete list of titles in this series, see ‘Modern Studies in European Law’ link at www.bloomsburyprofessional.com/ uk/series/modern-studies-in-european-law

The Internal Market 2.0 Edited by

Sacha Garben and

Inge Govaere

HART PUBLISHING Bloomsbury Publishing Plc Kemp House, Chawley Park, Cumnor Hill, Oxford, OX2 9PH, UK 1385 Broadway, New York, NY 10018, USA HART PUBLISHING, the Hart/Stag logo, BLOOMSBURY and the Diana logo are trademarks of Bloomsbury Publishing Plc First published in Great Britain 2020 Copyright © The editors and contributors severally 2020 The editors and contributors have asserted their right under the Copyright, Designs and Patents Act 1988 to be identified as Authors of this work. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior permission in writing from the publishers. While every care has been taken to ensure the accuracy of this work, no responsibility for loss or damage occasioned to any person acting or refraining from action as a result of any statement in it can be accepted by the authors, editors or publishers. All UK Government legislation and other public sector information used in the work is Crown Copyright ©. All House of Lords and House of Commons information used in the work is Parliamentary Copyright ©. This information is reused under the terms of the Open Government Licence v3.0 (http://www.nationalarchives.gov.uk/doc/ open-government-licence/version/3) except where otherwise stated. All Eur-lex material used in the work is © European Union, http://eur-lex.europa.eu/, 1998–2020. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication data Names: Garben, Sacha, 1982- editor.  |  Govaere, Inge, editor. Title: The Internal Market 2.0 / edited by Sacha Garben and Inge Govaere. Other titles: Internal Market two Description: Oxford, UK ; New York, NY : Hart Publishing, an imprint of Bloomsbury Publishing, 2020.  |  Series: Modern studies in European law ; volume 102  |  Includes bibliographical references and index. Identifiers: LCCN 2020035649 (print)  |  LCCN 2020035650 (ebook)  |  ISBN 9781509939039 (hardback)  |  ISBN 9781509942640 (paperback)  |  ISBN 9781509939053 (epdf)  |  ISBN 9781509939046 (Epub) Subjects: LCSH: Trade regulation—European Union countries.  |  Freedom of movement— European Union countries.  |  Free trade—European Union countries.  |  European Union countries— Economic integration.  |  Law—European Union countries—International unification. Classification: LCC KJE6417 .I589 2020 (print)  |  LCC KJE6417 (ebook)  |  DDC 343.2408—dc23 LC record available at https://lccn.loc.gov/2020035649 LC ebook record available at https://lccn.loc.gov/2020035650 ISBN: HB: 978-1-50993-903-9 ePDF: 978-1-50993-905-3 ePub: 978-1-50993-904-6 Typeset by Compuscript Ltd, Shannon

To find out more about our authors and books visit www.hartpublishing.co.uk. Here you will find extracts, author information, details of forthcoming events and the option to sign up for our newsletters.

TABLE OF CONTENTS Notes on Contributors������������������������������������������������������������������������������������������������ vii 1. Introduction: The ‘Internal Market 2.0’����������������������������������������������������������������1 Sacha Garben and Inge Govaere PART I THE INTERNAL MARKET AND ITS DEVELOPMENT OVER TIME 2. The Development of the Free Movement Principles Over Time�������������������������25 Stefan Enchelmaier 3. In Search of the Limits of Article 30 of the EEC Treaty Revisited����������������������65 Eric White 4. Internal Market Dynamics: On Moving Targets, Shifting Contextual Factors and the Untapped Potential of Article 3(3) TEU�����������������������������������75 Inge Govaere PART II THE FOUR FREEDOMS 5. The Classic Freedom? The Free Movement of Goods: Old Doctrines, New Cases and Contemporary Reflections���������������������������������������������������������95 Eleanor Spaventa 6. The ‘Social Freedom’? The Free Movement of Persons in EU27�����������������������111 Niamh Nic Shuibhne 7. The Freedom to Provide Services: The Controversial Freedom?�����������������������137 Bruno de Witte 8. Free Movement of Capital and Protection of Social Objectives in the EU: Critical Reflections on the Case Law Regarding Golden Shares and Privatisations���������������������������������������������������������������������161 Ilektra Antonaki

vi  Table of Contents PART III THE INTERNAL MARKET IN DIGITAL TIMES 9. Single Market 2.0: The European Union as a Platform������������������������������������187 Andrea Renda 10. The Internal Market and the Online Platform Economy���������������������������������213 Vassilis Hatzopoulos 11. ‘Tinkering or Fundamental Overhaul?’ The Past, the Present and the Future of the Digital Single Market�����������������������������������������������������233 Claire Bury and Irene Roche Laguna PART IV CRITICAL REFLECTIONS ON THE INTERNAL MARKET AND ITS FUTURE 12. The Fundamental Question of Minimum or Maximum Harmonisation��������261 Stephen Weatherill 13. From Supranationality to Managing Diversity: A (Re-)New(ed) Paradigm for the Establishment of the Internal Market?���������������������������������285 Kai P Purnhagen 14. The Internal Market in its Historical Context. Has the ECJ ‘Over-Constitutionalised’ the Internal Market?������������������������������������������������313 Peter Behrens 15. Originalism at the European Court of Justice���������������������������������������������������323 Gareth Davies 16. The ‘Fundamental Freedoms’ and (Other) Fundamental Rights: Towards an Integrated Democratic Interpretation Framework�����������������������335 Sacha Garben Index��������������������������������������������������������������������������������������������������������������������������371

NOTES ON CONTRIBUTORS Ilektra ANTONAKI is a guest researcher at the Europa Institute of Leiden Law School and legal assistant at the CJEU Peter BEHRENS is Professor Emeritus at the University of Hamburg Claire BURY is Deputy Director General at the Directorate General Communications Networks, Content and Technology at the European Commission and Visiting Professor at the College of Europe Gareth DAVIES is Professor of European Law at VU Amsterdam Bruno DE WITTE is Professor of European Union Law at Maastricht University and the European University Institute Stefan ENCHELMAIER is Professor of European and Comparative Law at the University of Oxford, and Fellow of Lincoln College, Oxford Sacha GARBEN is Permanent Professor of European Union Law at the European Legal Studies Department of the College of Europe Inge GOVAERE is Professor of European Law at Ghent University and Director of the Ghent European Law Institute (GELI) as well as Director of the European Legal Studies Department at the College of Europe Vassilis HATZOPOULOS is Professor of EU Law and Policies at Panteion University, Athens and Visiting Professor at the College of Europe Niamh NIC SHUIBHNE is Professor of European Union Law at the School of Law, University of Edinburgh and Visiting Professor at the College of Europe Kai PURNHAGEN is Chaired Professor of Food Law at the University of Bayreuth Andrea RENDA is Senior Research Fellow at CEPS and Professor of Digital Innovation, College of Europe. Member of the EU High Level Expert Group on AI Irene ROCHE LAGUNA is Deputy Head of the Unit dealing with the Implementation of the Regulatory Framework at the European Commission and Visiting Professor at the College of Europe Eleanor SPAVENTA is Professor of European Law at Bocconi University Stephen WEATHERILL is the Jacques Delors Professor of European Law at Somerville College and Law Faculty, University of Oxford Eric WHITE is a Consultant at Herbert Smith Freehills

viii

1 Introduction: The ‘Internal Market 2.0’ SACHA GARBEN AND INGE GOVAERE

I.  The EU, the Internal Market and Yet Another ‘Crisis’ By establishing an internal market, the EU has gradually become a closely interwoven and interconnected market, economy and society – precisely as the Member States themselves have consciously and deliberately decided it should be. Merging the markets into one and creating economic interdependence has been instrumental to maintain peace and stability among the EU Member States for over 70 years. Quite paradoxically, the Brexit saga is testimony of the success of European integration for exposing the difficulty of disentangling the Gordian knot of how to secure the advantages of participation in (some aspects of) the internal market without the concurrent obligations of EU membership. For one thing, this means a common understanding that EU-wide solidarity and concertation is needed not least in times of crisis. As disasters and disaster-decisions in one Member State affect the situation in other Member States, the problem of one is the problem of all. At the time that this book goes to press, the EU is yet again in the throes of a ‘crisis’.1 The COVID-19 virus and ensuing global pandemic has sadly resulted in many untimely deaths and

1 It should be noted that the use of the word ‘crisis’ in relation to the EU has suffered from a certain degree of inflation over the past years, especially to the extent that the EU itself has almost continuously been described to be in a state of crisis, in relation the economy and the euro, terrorism, migration, Brexit, the Rule of Law, the environment, and throughout it all – and because of it all – a general identity/legitimacy crisis. However, without in any way seeking to downplay the personal, humanitarian and political tragedies implied by these critical developments or the importance of the challenges faced by the EU in these respects, it deserves to be noted that is not immediately clear that in all these cases the European Union as such and in its entirety has really been in crisis. Of course, to the extent that media and public discourse tend to conflate critical circumstances that the EU as one of the most important global political actors has to deal with and may not be able to fully resolve, with a failure and crisis of the EU itself, the crisis exists in any event because it exists in people’s minds. For a political order, crisis is, to some extent, in the constituent’s eye. Yet as academics we have a responsibility to disentangle those situations that really, truly, pose an existential, life-or-death danger to the European integration project as a whole, and those that are either crises that are not specific to the EU or those that pose some problems for the EU specifically but in an ultimately surmountable and manageable way. The very fact that circumstances are unstable, that difficult decisions have to be adopted and prove difficult to adopt, and even that fundamental changes are made to the organisation, membership

2  Sacha Garben and Inge Govaere has caused governments across the world to take unprecedented measures restricting personal and economic activity and movement. The political, societal and economic consequences of these developments have yet to fully materialise. As is becoming commonplace when difficult circumstances appear on the international scene, the fact that the EU institutions or even the EU Member States individually and/or collectively cannot swiftly and decisively solve the issue, is portrayed as an illustration of the failure of the European Union as such. This is often more than a little unfair: the issue may not actually be within the EU’s power, as the Member States may not have provided the EU with the necessary competences to act in a way that can resolve or successfully manage the issue,2 and it may not even be appropriate for the EU to impose the uniform, harmonised approach that implicitly seems to be the expectation of those who attack the Union for a lack of action. At the same time, while the cause of the particular difficulties may lie wholly outside the EU’s control, and while it may simply lack the competences to take full and effective control over them, it is often also true that the EU’s policies that are in place will very likely in some way significantly interact with both the problems and the solutions in any given transnational challenge. This means that the EU inevitably gets tied into the issue. In particular as these common institutions were created, whenever we are faced with a common problem that furthermore has specific interactions with common policies, it is not entirely unreasonable for people to expect a strong common response. All the more so because, in light of its still more limited input legitimacy, the EU is extra reliant on its output legitimacy.3 Add to this the temptation of national governments under pressure to deflate as much tension and blame to the EU level as they can, while having little incentive to act in the common good rather than in their own (constituencies’) self-interest, and we can see how the Union quickly becomes a focal point of criticism. And thus in the context of the ‘Corona-crisis’, we have seen Member States unilaterally closing borders with other Member States, restricting or banning export of personal protective equipment (such as protective glasses, face masks, gloves, surgical overalls and gowns), medicines and medical devices (in one case extending to 1,324 products)4 and entering into corrosive rows over the financial measures to be taken, with the Union’s response in turn branded as unsatisfactory. and material scope of the project itself, is in itself nothing out of the ordinary or threatening for the EU: dynamic development and political difficulty is what it lives and breathes. The responsibility to distinguish crisis from challenges is deontological but also societal: a realistic, non-alarmist analysis is necessary to try and prevent crisis-narratives from unnecessarily becoming self-fulfilling prophecies. 2 See generally S Garben and I Govaere (eds), The Division of Competences between the EU and the Member States: Reflections on the Past, the Present and the Future (Oxford, Hart Publishing, 2017). 3 On these notions, their importance, and interaction, see F Scharpf, ‘De-constitutionalisation of European Law: The Re-empowerment of Democratic Political Choice’ in S Garben and I Govaere (eds), The Division of Competences between the EU and the Member States: Reflections on the Past, the Present and the Future (Oxford, Hart Publishing, 2017). 4 European Commission, Communication, ‘Coordinated economic response to the COVID-19 Outbreak’ COM(2020) 112 final, Annex 2, 4.

Introduction: The ‘Internal Market 2.0’  3 Furthermore, in these particular circumstances, the scientifically mandated response of social (or rather physical) distancing just so happens to hit the EU precisely where it hurts: the free movement that – despite some political difficulties over the past years – remains its hallmark and crown achievement,5 and one of its core competences. This inevitably draws the EU deeply into the issue and its management. Moreover, the positive side of that interconnectedness is the potential that lies in the infinitely larger and richer set of resources and solutions when pooled together, bolstering the problem-solving capacity of individual States because of the collective: ie, stronger together. Lives can be saved by sharing ICU-capacity and other facilities, cooperation can accelerate vaccine development, financial support can mitigate the recession. Balancing the case for unity and the case for diversity, the ‘subsidiaritycalculus’6 thus indicates that national decisions should be taken in accordance with a common framework even if they can (and perhaps should) diverge on substance, and that there has to be a certain measure of solidarity in the sharing of risk and resources. The general approach should thus be, in EU law generally as in the case of responding to disasters such as the COVID-19 pandemic, to ‘manage diversity’.7 While perhaps it could be said that the legal and political framework for financial solidarity, as well as arguably the one for civil protection, has proven to be inadequate in the face of this task in the Corona-crisis,8 the internal market framework has instead actually proven to function relatively well. Many of the containment measures taken by the Member States of course restrict the free movement of goods, services and people, but they fall squarely in the regime for legitimate justification in Article 36 of the Treaty on the Functioning of the European Union (TFEU), Article 45(3) TFEU and Article 52(1) TFEU.9 The Member States have a margin of discretion in this regard, as ‘account must be taken of the fact that the health and life of humans rank foremost among the assets and interests protected by the Treaty and that it is for the Member States to determine the level of protection which they wish to afford to public health and the way in which that level is to be achieved’, which ‘may vary from one Member State to another’.10 They are mandated to take a ‘better safe than sorry’ approach on the basis of the precautionary principle.11 The 5 Niamh Nic Shuibhne in ch 6 discusses the political difficulties surrounding the free movement of persons in the context of Brexit, however noting that 82% of Union citizens support the freedom to live, work, study and do business in other Member States and that ‘far too often that policymaking is driven disproportionately by the 18 per cent’. 6 K Gutman, The Constitutional Foundations of European Contract Law: A Comparative Analysis (Oxford, Oxford University Press 2014) 296. 7 See ch 13 by Kai Purnhagen. 8 C Beaucillon, ‘International and European Emergency Assistance to EU Member States in the COVID-19 Crisis: Why European Solidarity Is Not Dead and What We Need to Make It both Happen and Last’ (25 April 2020) European Papers, European Forum, Insight 1. 9 As regards non-economically active citizens, the reference is Art 29 of Directive 2004/38/EC. 10 Case C-198/14 Visnapuu EU:C:2015:751, para 118 and case law cited. 11 Although the Court of Justice of the European Union (CJEU) tends to apply this more leniently towards the Union legislator than measures imposed by the Member States, see Cases C-192/01 Commission v Denmark EU:C:2003:492; C-333/08 Commission v France EU:C:2010:44; C-343/09 Afton Chemical EU:C:2010:419.

4  Sacha Garben and Inge Govaere normative precedence accorded to the protection of health and the empowerment of Member States in upholding that interest is entirely appropriate, ethically as well as constitutionally, even if it fragments the internal market and limits free movement.12 At the same time, it is also appropriate to ensure that the measures do not do unnecessary harm, and therefore EU law correctly rejects an ‘anything goes’ approach. It is settled law that the national measures need to comply with the principle of proportionality, and thus have to be appropriate, necessary and proportionate stricto sensu. The European Commission needs to be able to oversee these measures, through notification, to limit the impact of the national measures to what is necessary and to ensure they do not in the bigger scheme of things create a counterproductive effect, by placing them in the European context of supply and demand and production chains. The Commission has indicated that draconian export bans are unlikely to meet that test.13 The Commission has furthermore demanded that Member States keep their borders open in relation to the circulation of essential goods and services,14 to frontier workers15 and to allow the entry of EU citizens residing on their territory.16 Most Member States seem to comply with this, and if they do not, they can be referred to the Court of Justice of the European Union (the Court). The internal market framework can be perfectly well enforced against Member States’ kneejerk (over-)reactions, and they know it. And thus, for instance, after the Commissioner for the Internal Market Thierry Breton intervened, Germany and France authorised the export of masks, gowns and other protection to Italy, despite their initial national requisition measures.17 This shows how the internal market as an area constructed largely through the Community Method over the years provides distinct advantages, also in times of crisis,18 over those EU policies dependent on the intergovernmental or ‘EU Method’19 of executive dominance.20 12 See also S Coutts, ‘Citizenship, Coronavirus and Questions of Competence’ (25 April 2020) European Papers, European Forum Insight 1. 13 COM(2020) 112 final. 14 European Commission, ‘Guidelines for border management measures to protect health and ensure the availability of goods and essential services’, C(2020) 1753 final. 15 European Commission, ‘Guidelines concerning the exercise of the free movement of workers during COVID-19 outbreak’, C(2020) 2051 final. 16 C(2020) 1753 final. 17 Beaucillon, above n 8. 18 For the argument contra, see C Timmermans, ‘The Competence Divide of the Lisbon Treaty Six Years After’ in S Garben and I Govaere (eds), The Division of Competences between the EU and the Member States: Reflections on the Past, the Present and the Future (Oxford, Hart Publishing, 2017). 19 Angela Merkel, Speech given on the occasion of the opening ceremony of the 61st academic year of the College of Europe in Bruges, 2 November 2010. The method consists in ‘coordinated action in a spirit of solidarity, each of us in the area for which we are responsible (that is to say, the institutions and Member States) but all working towards the same goal’, and is a compromise between the ‘Community Method’ and the pure intergovernmental method. For discussion, see D Dero-Bugny, ‘The Dilution of the Community Method and Diversification of Intergovernmental Practices’ (2014) 134 Revue de l’OFCE 65. 20 For a defence of the Community Method and the case to expand legislative competence to counter the use of other forms of harmonisation (by stealth): S Garben, ‘Restating the Problem of Competence Creep, Tackling Harmonisation by Stealth and Reinstating the Legislator’ in S Garben and I Govaere (eds), The Division of Competences between the EU and the Member States: Reflections on the Past, the Present and the Future (Oxford, Hart Publishing, 2017).

Introduction: The ‘Internal Market 2.0’  5 This is not to say that the internal market framework is perfect, and that it should not be further improved also in light of the COVID-19 experience. There certainly is scope to fine-tune the ‘managed diversity’ of the internal market, in health and in other areas, as Kai Purnhagen argues in chapter thirteen. His contribution traces a development in EU internal market law from supranationality to managed diversity and proposes as a steering principle (alongside the traditional hallmarks of subsidiarity, proportionality and conferral) the principle of solidarity. Applying this framework specifically to COVID-19,21 the case is made to adopt ‘a different approach to the Union’s competence to control national measures on the basis of proportionality’, instead based on ‘the principle of solidarity between the Member States, the Union’s obligation to protect public health, and the right to health care within Union law’.22 On this approach, ‘scarce products needed for combatting COVID-19 need to be channeled to those who need them most, and this means keeping internal Union borders open for the supply of essential products, especially food, as well as medical equipment, devices and medicines’.23 It is argued that ‘the internal market is at the service of the health of Europe’s population’, assuming a concept of Union public health and Union solidarity, ‘rather than the usual interpretation where public health is a policy area in which domestic solidarity within Member States’ is the principle that is ‘traded-off against market integration’ – ostensibly a novel interpretation of internal market law ‘not based on case law or any other legal instrument’.24 Novel as it may be, the Commission seems in full agreement: it too has emphasised the importance of solidarity in the scrutiny of national COVID-19 restrictions, stating: The Single Market is at the heart of the European Union. In times of crisis it is the solidarity instrument to ensure that essential goods necessary to mitigate health risks outbreak can reach all those in need. By making sure those goods are available across the EU, the Single Market contributes to the protection of our health. Unilateral national restrictions to the free movement of essential supplies to the healthcare systems create significant barriers and affect dramatically Member States’ capacity to manage the COVID-19 outbreak.25

II.  Uniformity Versus Diversity The concern for the appropriate balance between unity and diversity in the internal market is one of the main red threads that emerges in this edited volume.

21 K Purnhagen et al, ‘More Competences than You Knew? The Web of Health Competence for European Union Action in Response to the COVID-19 Outbreak’ (2020) 11(2) European Journal of Risk Regulation, 297–306. 22 ibid, 8. 23 ibid, 9. 24 ibid. 25 COM(2020) 112 final, 3.

6  Sacha Garben and Inge Govaere As discussed above, the COVID-19 situation has showcased the inbuilt balance between common standards and national regulatory autonomy in the framework of public health justification to national restrictive measures. It would seem that in relation to the pandemic, EU internal market law is actually ‘doing its job well’ in accordance with the criteria developed by Gareth Davies:26 it is functional, provides concepts and a framework which encourages substantively good results, it provides limits to the EU’s powers and provides the basis for a legitimating discourse.27 Also in other areas of the internal market, some of our authors find a shift towards a ‘deferential’ approach that accommodates national concerns. Bruno de Witte in chapter seven notes in relation to the freedom to provide services, that generally speaking, the Court seems to leave considerable deference to choices made by national governments and to the fact finding by national courts. There are roughly equal numbers of judgments in which the Court finds national measures compatible or incompatible with Art 56. So, all in all, the view that the Court of Justice still functions as an ‘engine of liberalisation’ seems less persuasive than before, at least in the field of services.28

Yet in other parts of internal market law, there may be some further room for improvement in how diversity is accommodated. Illektra Antonaki in chapter eight questions whether sufficient space is left for national regulatory choices, in the specific context of the free movement of capital. She argues that the very broad interpretation of the notion of ‘capital restrictions’ under Article 63 TFEU and the treatment of golden shares and forms of public ownership as inherently incompatible with the internal market ‘raises significant concerns, firstly, regarding the division of competences between the EU and the Member States in the fields of corporate governance and property ownership systems, and secondly, regarding the protection of public interest objectives’. She considers that the definition of capital restrictions should be refined and that a ‘procedural proportionality test’ should be applied in relation to golden shares, as this would ‘guarantee an objective judicial review that respects national regulatory diversity in a sensitive field that is not harmonised at the EU level, whilst at the same time ensuring compliance with procedural requirements that create a transparent legal framework protecting the legitimate expectations of foreign investors and providing effective judicial remedies against arbitrary decisions of public authorities’ and that more weight should be given to Article 345 TFEU29 as a ground for justifying restrictions on capital movements. 26 G Davies, ‘The Competence to Create an Internal Market: Conceptual Poverty and Unbalanced Interests’ in S Garben and I Govaere (eds), The Division of Competences between the EU and the Member States: Reflections on the Past, the Present and the Future (Oxford, Hart Publishing, 2017). 27 ibid. 28 Referring to J Zglinski, ‘The Rise of Deference: The Margin of Appreciation and Decentralized Judicial Review in EU Free Movement Law’ (2018) 55 Common Market Law Review 1341; and for the term ‘engine of liberalisation’ to N Dunne, ‘Liberalisation and the Pursuit of the Internal Market’ (2018) 43 European Law Review 803, 809–10. 29 ‘The Treaties shall in no way prejudice the rules in Member States governing the system of property ownership’.

Introduction: The ‘Internal Market 2.0’  7 Eleanor Spaventa in chapter five discusses the Scotch Whisky case,30 in which the Court considered the justification of the Scottish government’s minimum price per alcohol unit, the aim of which was to reduce alcohol-related deaths and health problems caused by excessive alcohol consumption. The Court found that taxation would be, in principle, a less restrictive way to achieve the same aim since it would not have an impact on the formation of prices, while leaving the final assessment to the national court, which ultimately found in favour of the Scottish government thus meaning that the rules were maintained. As Spaventa notes, what is particularly interesting from the perspective of national regulatory autonomy is the specific context of devolution in the case. All parties agreed that, for the purposes of the case, and in line with established case law, Scotland was to be treated as if it were the UK, but while consistent with international law practice, … the same approach is more debatable when transferred point blank to the realm of the proportionality assessment, especially when the latter includes a detailed examination of regulatory alternatives. In particular, in Scotch Whisky the issue, described by the national court as the elephant in the room, was that whereas the Scottish Government has devolved jurisdiction in the field of health, it does not have competence in the field of taxation. The assessment of regulatory alternatives therefore became an exercise in legal fiction – the measures examined by the Court of Justice as being a suitable regulatory alternatives were measures that the regulator could not in fact adopt.

Considering the fact that Article 4(2) TEU obliges the EU to ‘respect the equality of Member States before the Treaties as well as their national identities, inherent in their fundamental structures, political and constitutional, inclusive of regional and local self-government’, it indeed may be necessary for the Court to take better account of national competence division in the application of the proportionality assessment, as it already sometimes does in relation to the interpretation of secondary legislation.31 Gareth Davies notes in chapter fifteen that in more general terms EU internal market law still errs on the side of uniformity rather than diversity: Free movement is understood as imperative, as urgent, and the burden of adaption to its requirements is placed on the Member States, either directly, through negative harmonisation, or through positive harmonisation. There is an absence in the case law of any suggestion that there should be limits to movement – except in de minimis situations, which, by definition, barely matter – or that we should accept that the existence of diverse institutions and regimes is a good in itself which entails certain constraints on ease of movement. Sometimes movement gives way to national institutions in the law,

30 Case C-333/14 Scotch Whisky Association and Others v The Lord Advocate and The Advocate General for Scotland EU:C:2015:845. 31 Case C-213/07 Michaniki AE v Ethniko Symvoulio Radiotileorasis and Ypourgos Epikrateias EU:C:2008:73; see discussion by S Garben, ‘Collective Identity as a Legal Limit to European Integration in Areas of Core State Power’ (2020) 58 Journal of Common Market Studies 41–55, 51.

8  Sacha Garben and Inge Govaere but this is contingent and temporary, not a desirable end point. The vision of the Court is one in which states redesign themselves, or are redesigned, to fit the requirement of complete inter-State openness.

Some of the cases discussed by Sacha Garben in chapter sixteen, notably AlemoHerron32 and AGET,33 would confirm the Court’s reluctance to relinquish an absolute free movement paradigm in favour of national competence to set regulatory standards. What is specifically problematic from this perspective in both these cases, is that they concerned minimum harmonisation measures which explicitly and deliberately should leave Member States the freedom to set higher standards of protection, a fact of constitutional relevance that was ignored by the Court. In Alemo-Herron, a judgment that Stephen Weatherill has found so deplorable that it should be ‘consigned to the bottom of an icy lake’,34 the Court – controversially35 relying on Article 16 of the EU Charter on the freedom to conduct a business – interpreted a minimum harmonisation directive on safeguarding of employees’ rights in the event of transfers of undertakings to exclude a more protective UK clause on collective agreements without even establishing an infringement of the internal market provisions. In AGET, the Court goes very far in that same direction in relation to a minimum harmonisation directive on collective redundancies, although ultimately condemning the (indistinctly applicable) Greek rules on prior authorisation of collective redundancies on the basis of Article 49 TFEU instead. Such an unaccommodating approach to minimum harmonisation is constitutionally problematic, not in the least since, as Stephen Weatherill explores in-depth in chapter twelve, [m]inimum harmonisation represents a realistic attempt to accommodate diverse national tradition and to promote scope for dynamic regulatory innovation within the process of integration. In the development of the regulatory framework that underpins the internal market there are constitutional commitments that favour minimum rule-making over maximum measures contained in EU primary law, most prominently in Articles 12, 114(3), 168(1) and 169(1) TFEU and Articles 35, 37 and 38 of the Charter. These several provisions assert the EU’s role as a proper source of regulatory protection, but it does not and should not follow that it should be the exclusive source.

32 Case C-426/11 Mark Alemo-Herron and Others v Parkwood Leisure Ltd EU:C:2013:521. 33 Case C-201/15 Anonymi Geniki Etairia Tsimenton Iraklis (AGET Iraklis) v Ypourgos Ergasias, Koinonikis Asfalisis kai Koinonikis Allilengyis EU:C:2016:972. 34 S Weatherill, ‘Use and Abuse of the EU’s Charter of Fundamental Rights: on the improper veneration of “freedom of contract”’ (2014) 10 European Review of Contract Law 167. 35 X Groussot, GT Pétursson and J Pierce, ‘Weak Right, Strong Court – The Freedom to Conduct Business and the EU Charter of Fundamental Rights’ (2014) Lund University Legal Research Paper 01/201; J Prassl, ‘Freedom of contract as a general principle of EU law? Transfers of undertakings and the protection of employer rights in EU labour law’ (2013) 42 Industrial Law Journal 434; P Syrpis and T Novitz, ‘The EU Internal Market and Domestic Labour Law: Looking Beyond Autonomy’ in A Bogg et al (eds), The Autonomy of Labour Law (Oxford, Hart Publishing, 2015); M Bartl and C Leone, ‘Minimum Harmonisation after Alemo-Herron: The Janus Face of EU Fundamental Rights Review’ (2015) 11 European Constitutional Law Review 140.

Introduction: The ‘Internal Market 2.0’  9 A model built not only on EU rule-making but also on national level contributions above an EU minimum offers a compelling alternative. In fact the stakes are high enough to justify reliance on Article 4(2) TEU as a provision that favours minimum over maximum harmonisation, for only by foreclosing the pre-emptive effect of EU law is space left for the expression of national identity in its many diverse forms across the Member States. So there is a strong case to be made for regarding maximum harmonisation as applicable only exceptionally, where its use has been carefully justified in prevailing sector-specific conditions.

On this basis it would seem that especially in cases where the EU legislator itself has specifically decided for minimum harmonisation, the Court should refrain from turning the measure into a maximum harmonisation standard instead.

III.  Contextualising and Embedding the Internal Market The foregoing citation from Weatherill’s chapter touches on a second main red thread that reveals itself in the various contributions to this book: the need to interpret the market in the broader constitutional context in which it finds itself, especially post-Lisbon. As Inge Govaere highlights in chapter four, the Lisbon Treaty radically departs from the earlier Treaties ‘by opting for a strong proactive and purposeful wording of the internal market objective’. The former Article 2 TEC is now repealed and replaced, in substance, by Article 3 TEU, the third paragraph of which reformulates specifically the internal market objective.36 She points out that the wording used … expresses a strong assertion that the internal market shall be established and thus ‘proactively’ pursued. Moreover, Article 3(3) TEU dictates that such a purposeful internal market shall also proactively pursue other higher and horizontal Treaty objectives. The Lisbon Treaty thus provides the necessary, but seemingly also a compulsory, legal framework for the establishment of a more proactive internal market policy which positively intertwines with other EU policies. It cannot go unnoticed that this echoes the provisions of general application as listed in Articles 7–13 TFEU.

In particular, this provides legal ground to revisit the CJEU’s case law on the internal market harmonisation legal basis of Article 114 TFEU in Tobacco Advertisement,37 36 ‘The Union shall establish an internal market. It shall work for the sustainable development of Europe based on balanced economic growth and price stability, a highly competitive social market economy, aiming at full employment and social progress, and a high level of protection and improvement of the quality of the environment. It shall promote scientific and technological advance. It shall combat social exclusion and discrimination, and shall promote social justice and protection, equality between women and men, solidarity between generations and protection of the rights of the child. It shall promote economic, social and territorial cohesion, and solidarity among Member States. It shall respect its rich cultural and linguistic diversity and shall ensure that Europe’s cultural heritage is safeguarded and enhanced’. 37 Case C-376/98 Federal Republic of Germany v European Parliament and Council of the European Union EU:C:2000:544.

10  Sacha Garben and Inge Govaere where it held that a mere finding of disparities between national rules is not sufficient to rely on Article 114 TFEU, but that there also has to be an ‘obstruction of the fundamental freedoms’ or a need ‘to prevent the emergence of future obstacles to trade’.38 As Govaere argues, with the repeal of former Article 3 TEC and the express conferral of shared competence for the internal market in Article 4 TFEU, it seems difficult to anchor these restrictive conditions in the Treaties post-Lisbon. The continuation of the ‘obstacles’ test furthermore does not seem to follow from the wording of Article 114 TFEU itself, as it more generally provides for the possibility to adopt EU ‘measures for the approximation of the provisions laid down by law, regulation or administrative action in Member States which have as their object the establishment and functioning of the internal market’.39 Such a broadened interpretation of the internal market legal basis would perhaps be perceived as problematic from the perspective of ‘competence creep’40 by some, but it should be kept in mind that of all the various forms of ‘harmonisation by stealth’,41 measures regulating the internal market in the public interest are from a legitimacy point of view the least of our concern.42 As Peter Behrens states in chapter fourteen: [L]egislative acts of the Union are, as any legislation, always the result of political compromises within the Union’s institutions and among Member States’ governments all of which are democratically elected. Irrespective of whether such powers are used for the harmonisation of national laws in order to facilitate the use of the economic freedoms or for the implementation of the Treaty’s competition rules, no directive or regulation will be enacted without an appropriate balancing of conflicting political preferences and interests … The decision-making process within the institutions of the EU requires and allows nevertheless careful attention to political preferences developing in the Member States.

Legislation on the basis of Article 114 TFEU is adopted through the most democratic form of international decision-making that is available, benefiting from high levels of input and ‘throughput legitimacy’43 especially when compared with all other forms of competence creep such as economic policy coordination, or for that matter, negative integration of sensitive national areas of competence on the basis of the directly effective free movement provisions.44 As Garben argues in chapter sixteen, 38 Case C-547/14 Philip Morris EU:C:2016:325, see in particular paras 58–60. 39 See also S Garben, ‘The Principle of Legality and the EU’s Legitimacy as a Constitutional Democracy: A Research Agenda’ in S Garben, I Govaere and P Nemitz (eds), Critical Reflections on Constitutional Democracy in the European Union (Oxford, Hart Publishing, 2019). 40 S Weatherill, ‘Competence Creep and Competence Control’ (2004) 23 Yearbook of European Law 1, S Garben, ‘Competence Creep Revisited’ (2019) 57 Journal of Common Market Studies 205. 41 S Garben, EU Higher Education Law: The Bologna Process and Harmonization by Stealth (Alphen aan de Rijn, Kluwer Law International, 2011). For this term, I am indebted to Bruno de Witte. 42 Garben, ‘Competence Creep Revisited’, above n 40. 43 V Schmidt, ‘Is There a Deficit of Throughput Legitimacy in the EU?’ in S Garben, I Govaere and P Nemitz (eds), Critical Reflections on Constitutional Democracy in the European Union (Oxford, Hart Publishing, 2019). 44 Garben, ‘Competence Creep Revisited’ above n 40.

Introduction: The ‘Internal Market 2.0’  11 one of the traditional normative justifications for the constitutionalisation of the free movement provisions is the argument of transnational democracy,45 also referred to as the argument from ‘transnational effects’.46 Market integration is, to the extent that it de-nationalises decision-making to ensure that affected interests from other Member States are taken into account, ‘desirable on account of democracy itself ’.47 This transnational democracy argument is indeed very strong in relation to positive market integration through the adoption of EU-level rules regulating the internal market, and this serves as an important counterbalance to the often-heard argument that the EU legislative process is not as democratically legitimate as national decision-making, and mitigates the concerns of competence creep in relation to the use of Article 114 TFEU. However, the transnational democracy argument has instead been most widely used in the context of negative integration,48 while there its validity is much less obvious than in relation to positive market integration, as examined in detail in Garben’s chapter. Since the negative application of the free movement provisions is, from a democratic perspective, the less refined, less effective and more contentious method of market integration, the democratic argument would require that precedence is given to positive market integration, for which we need a broad interpretation of Article 114 TFEU. Allowing the EU to use Article 114 TFEU to regulate the market in a more general sense, which as Govaere argues in chapter four is supported by both a textual and contextual interpretation of the Treaties post-Lisbon, would furthermore allow us to overcome what Davies has called the internal market’s ‘conceptual poverty’.49 Davies has argued that ‘despite all the ways in which national competence is squeezed while EU competences are interpreted broadly, there is another way in which internal market legislative competences can be seen as too narrow: in the goals that may be pursued’.50 Limiting the possible uses of Article 114 TFEU to only two, the Court seems to be imposing a substantive interpretation of what a market is and requires. Suppose the legislature does take one of the mainstream views that establishing a functioning market entails eg shared values or redistribution. It would seem that the Court will not allow them to pursue this path with law.51

He continues that there is nothing neutral about the proposition that establishing a market requires merely movement and competition. This is a particular political standpoint, and a very

45 C Joerges, ‘European Law as Conflict of Laws’ in C Joerges and J Neyer, ‘“Deliberative Supranationalism” Revisited’ (2006) 20 EUI Working Paper Law 22. 46 A Somek, ‘The Argument from Transnational Effects I: Representing Outsiders through Freedom of Movement’ (2010) 16 European Law Journal 315. 47 ibid. 48 ibid. 49 Davies, above n 26. 50 ibid. 51 ibid.

12  Sacha Garben and Inge Govaere contentious one. It is close to what is commonly referred to in critique as ‘neo-liberalism’ and however vague and misused that term may be, it is hard to fully rebut claims that the EU is a neo-liberal project as long as its core competence is not permitted to be used in any other than a neo-liberal way: as long as establishing a market and making it function cannot include ensuring it is embedded in wider society.52

The argument that a free market interpretation of the internal market should be avoided is further supported by Article 3(3) TEU which labels and commits the EU to be a ‘social market economy’. It is by reference to that normative underpinning that Antonaki in chapter eight argues for a different approach to the clash between economic and social objectives in the field of free movement of capital and more specifically in the case law concerning golden shares and privatisations. She considers that this can make a crucial difference in the perennially contested balancing exercise between economic freedoms and social values. As Weatherill notes in chapter twelve, ‘(e)conomic integration is and always has been at the forefront of EU activity, and it has been compellingly argued that the single market “remains the EU’s core business”, but there is much more to the EU’s mission than trade, as a glance at Articles 2 and 3 TEU reminds’. Indeed, other important embedding values in the EU’s constitutional framework include notably those contained in the EU Charter of Fundamental Rights, post-Lisbon having the ‘same legal value as the Treaties’.53 As Niamh Nic Shuibhne states, this entails ‘legal equivalence for Treaty freedoms and fundamental rights’.54 She furthermore points out that in accordance with Article 2 TEU, ‘[t]he Union is founded on the values of respect for human dignity, freedom, democracy, equality, the rule of law and respect for human rights’, which could provide an argument for priority of fundamental rights over free movement provisions.55 Garben further explores in chapter sixteen how the ‘fundamental freedoms’ of the internal market and fundamental rights can be woven into a coherent, integrated adjudication framework. Considering in more general terms that the EU is in need of a more coherent theory on the basis of which it could define and improve its legitimacy, the chapter makes the case to root any such theory in the two equal principles of democracy and constitutionalism56 – the latter consisting in the Rule of Law and fundamental rights57 – and argues that it is imperative that this frame ensures an

52 Referring to J Caporaso and S Tarrow ‘Polanyi in Brussels: Supranational Institutions and the Transnational Embedding of Markets’ (2009) 63 International Organization 593. 53 Art 6(1) TEU. 54 N Nic Shuibhne, ‘Fundamental Rights and the Framework of Internal Market Adjudication: Is the Charter Making a Difference?’ in P Koutrakos and J Snell (eds), Research Handbook on the Law of the EU’s Internal Market, Research Handbooks in European Law (Cheltenham, Edward Elgar Publishing, 2017). 55 ibid. 56 See J Tully, ‘The Unfreedom of the Moderns in Comparison to Their Ideals of Constitutional Democracy’ (2002) 65 Modern Law Review 204. 57 See also Garben, ‘The Principle of Legality and the EU’s Legitimacy as a Constitutional Democracy’, above n 39.

Introduction: The ‘Internal Market 2.0’  13 appropriate embedding of the internal market.58 This ‘thick constitutional context’ should provide the context and telos for the interpretation of, in particular, the directly effective ‘negative’ provisions on goods, services, workers, establishment and capital, both as regards their individual substantive meaning and in terms of their relationship to other rights. As becomes clear from the analysis, this does mean that there is a legitimate place for the free movement provisions as part of EU fundamental rights and constitutionalism, but only to the extent that they serve human dignity and foster a robust and well-functioning transnational democracy. This agues for a rather more limited interpretation of what is a prima facie restriction, keeping a firm focus on a discriminatory element. This provides something of a middle way between the thought-provoking yet controversial argument to de-constitutionalise the internal market out of a concern for democracy on the one hand,59 and the approach of internal market-exceptionalism that seems to consider ‘the fundamental economic freedoms’ to be the normative apex and constitutional foundation of the European integration process. This ‘embedding process’ should be multidirectional, as Niamh Nic Shuibhne argues in chapter six. She takes issue with the ‘prevailing register of disconnect’ between the internal market and in particular EU social policy, sharing the concern for more meaningful integration of internal market and social goals60 but approaching the issue from the other way around, asking, in essence, can EU social policy really be effective if it overlooks the effects and implications of free movement of the EU-bespoke dimension of the internal market around which social policy should place a critical frame of protection? The example of retaining the status of worker in a host Member State is used in section IV to illustrate more concretely why disconnecting rather than integrating economic and social policy threads is sub-optimal. That discussion confirms that long-standing systemic vulnerabilities of EU social policy – notably, asymmetries of competence and a reluctance to assign responsibility for action – were not reconceived through the adoption of the Pillar, therefore most likely prolonging disconnect between the social and the mobility dimensions of the internal market. Echoing the view of the Commission, the chapter therefore concludes that transformative progress – exhibiting more meaningful integration of the internal market’s social and economic dimensions – ultimately ‘requires more political courage and determination than 25 years ago, and greater efforts than ever to close the gap between rhetoric and delivery’.

58 In this regard, it concurs with the view of Schiek, who has proposed to ‘constitutionally condition’ the internal market. See D Schiek, ‘Towards More Resilience for a Social EU – the Constitutionally Conditioned Internal Market’ (2017) 13 European Constitutional Law Review 611. 59 D Grimm, The Constitution of European Democracy (Oxford, Oxford University Press, 2017); D Grimm, ‘The Democratic Costs of Constitutionalisation: The European Case’ (2015) 21 European Law Journal 460; Scharpf, above n 3. See discussion by Peter Behrens in ch 14. 60 Schiek, above n 58; see also ch 16 by Sacha Garben.

14  Sacha Garben and Inge Govaere

IV.  Change, Clarity and Continuity in EU Internal Market Law Yet despite a degree of consensus among our authors that the internal market should be (re)interpreted in light of its constitutional context, paying due regard to the rewording of Article 3(3) TFEU by the Treaty of Lisbon and in order to do justice to the various overriding values contained in the EU Charter of Fundamental Rights and in the Treaties, the case law has arguably remained rather static. Davies writes in chapter fifteen that, most of (internal market) law, and most of its impact, comes from repeated application of old-established principles. Where there is controversy, it is usually because the Court is sticking to a rule, not because it is going in a new direction. Dassonville expressed the idea that it is not form, but effects which matter, and that effects should be understood in a commonsensical way: is there, or might there be, an actor whose movement is made more difficult? Laval, Viking, the free movement of medical patients, and the cases on Free Movement of Capital are just applications of this simple approach. Cassis expressed the idea that each actor in the market should be regulated by their home state. The law on services, the Services Directive and Centros are applications of this. In all of these controversial situations arguments were made that the situation was different – trade unions were special, medical care was special, free choice of company location threatened models of incorporation, services deregulation would undermine society, removing all obstacles to investment would amount to dis-embedding capitalism. Whatever one might think about the merits of these arguments, or about the Court’s final decisions, the characteristics of the judgments are that faced with a new context, and a plausible argument for a new approach, the Court instead stuck to its old one. Free movement law is really just footnotes to Dassonville.

The controversial cases cited by Davies generally pre-date Lisbon. But Govaere exposes in chapter four that the fundamental changes in the EU’s constitutional framework introduced by the Lisbon Treaty have also not (yet) led the Court to reconsider the internal market and its own pre-Lisbon interpretation thereof. As she says, it is striking to find that the fundamentally changed internal market objective, from reactive and instrumental to proactive and purposeful, is not at all addressed by the CJEU in relation to the establishment of the internal market as such. In fact, it would appear that the very first sentence of Article 3(3) TEU has so far never been considered by the CJEU in post-Lisbon case law. Not even in relation to the interpretation of specific internal market provisions, such as Article 114 TFEU. Instead, the CJEU simply points back to pre-Lisbon rulings as the basis for a consistent line of case law which still stands good post-Lisbon. It nonetheless seems highly problematic for the CJEU to bluntly do so without explaining the underlying rationale. How does this square with the reformulation of the internal market objective, but also with the above-mentioned repeal of former Article 3 TEC which was very much the anchorage of the pre-Lisbon case law?

Such a non-changing approach might have been justifiable if one was to consider that the established framework is so clear and functions so well that there is no

Introduction: The ‘Internal Market 2.0’  15 compelling reason to fundamentally revisit it (even if it would still remain difficult to square with the explicit stipulations in the EU’s constitutional texts). But can we really say that there is a satisfactory degree of legal certainty as regards the fundamental concepts and their functioning in the EU internal market, and that they work well? Views on these issues differ, including among our collected authors. Many identify a certain ‘stabilisation’ trend in their chapters, with accumulating amounts of case law and, importantly, secondary legislation colouring in the grey spaces, while still identifying some fundamental uncertainties. De Witte in a comprehensive survey of the law on services in chapter seven finds that secondary legislation occupies most of the field, with gambling as the most significant ‘gap’ where litigation in relation to Article 56 TFEU remains central. Eric White states in chapter three, revisiting his seminal article on ‘the Limits of Article 30 of the EEC Treaty’61 by reference to the past 30 years of case law in the area of free movement of goods that ‘more clarity has been achieved’, although noting that ‘all cases are fact specific and the case law relating to a vague yet fundamental treaty provision will inevitably have many conflicting strands’. Spaventa agrees in chapter five that ‘the sheer number of cases have provided national courts with a reasonable idea of the boundaries of Article 34 TFEU’ and that in the area of free movement of goods ‘the case law has pretty much settled’, but also points out that there nevertheless still is ‘a certain lack of clarity in the dicta of the Court’, and identifies as new frontiers of controversy the extensive duties to act and of good administration imposed on law makers. Stefan Enchelmaier in chapter two considers that the lack of clear definitions by the Court has led to ‘unnecessary confusion’ in relation to Keck,62 and that even if the uncertainty was ‘avoidable’, the resulting lively discussions by commentators encouraged the Court to try a new general formula: ‘the elevation of “market access” to be the decisive criterion, culminating in the 2009 judgment in Moped trailers’.63 He considers, however, that this ‘is little more than wordplay: “market access” is merely another way of saying, “free movement between Member States”. We already knew that that must not be impeded. The question is not whether, but how’. According to Enchelmaier, we do know how: ‘the old tools – the discrimination test and the prohibition of universal bans (“internal frontiers”) – are still in perfect working order’. Yet could it be that the tools, while generally known, are in themselves inherently unclear? Spaventa in chapter five considers that the case law on goods is best understood as a ‘system of rebuttable and non-rebuttable presumptions’.64 61 EL White, ‘In Search of the Limits to Article 30 of the EEC Treaty’ (1989) 26 Common Market Law Review 235. 62 Joined Cases C-267/91 and C-268/91 Criminal proceedings against Bernard Keck and Daniel Mithouard EU:C:1993:905. 63 Case C-110/05 Commission of the European Communities v Italian Republic EU:C:2009:66. 64 The presumptions being as follows: (1) Directly discriminatory rules are always caught by Arts 34 (and 35) TFEU and justifiable (usually) only pursuant to Art 36 TFEU. (2) Product requirements (all rules concerning the physical qualities of a product) are caught by Art 34 TFEU as there is an absolute

16  Sacha Garben and Inge Govaere In that system, it actually transpires, it is safest to assume that national measures somehow affecting goods are in fact caught as prima facie restrictions that need justification, unless they are very obviously non-discriminatory selling arrangements. The main ‘clarity’ then comes from the near all-encompassing nature of the notion of a ‘restriction’, and from knowing that the real game is played in terms of justification. It would seem that the same can roughly be said for the other free movement provisions. And from there, we go back again: there may be reasonable certainty as to what are in principle accepted public interest reasons to justify a restriction, but the level of scrutiny by the Court as regards the necessity of the measures compared with other (theoretical) policy options and the extent to which it will leave the ultimate decision on compatibility to the national court is largely unpredictable. The question of whether internal market law, its concepts and methodology, its substance and application, ‘works well’, is even more contentious. Views differ on what Davies in chapter fifteen calls the ‘currently fashionable critique of the state of EU law’, namely that it is over-constitutionalised, ie, ‘that too many policy choices are fixed in EU law, because they are embedded in Treaties which are very hard to change. Matters which ought to be the subject of political debate are instead constitutionalised, and then no longer meaningfully contestable’. Davies considers that ‘(t)o the extent that this is true, it is because the Court has chosen to interpret the Treaties in a certain way’. Behrens provides a passionate defence of this status quo of constitutionalisation and the Court’s role in internal market law in chapter fourteen. The over-constitutionalisation critique is according to Behrens, at least in the version put forward by Dieter Grimm: [B]ased on a fundamental lack of understanding of the indispensable legal underpinnings of an internal market within the framework of the EU. Far from a ‘self-empowerment’ of the Court at the expense of the political institutions of the EU (including its Member States), its jurisprudence has merely unfolded what may have initially been hidden in text of the Treaty by following generally accepted rules of interpretation. The twin concepts of direct effect and supremacy of the economic freedoms and the competition rules admittedly but deliberately reduce the scope of political intervention by Member States’ governments or parliaments with the free flow of goods, services, capital and persons across national borders or with undertakings’ freedom to compete for presumption that they affect intra-Community trade, and must be justified pursuant to mandatory requirements of public interest and/or Art 36 TFEU. This means that it is not open to a Member State to rebut the presumption of effect on inter-community trade. (3) Certain selling arrangements on the other hand benefit from a rebuttable presumption of compatibility with Art 34 TFEU, unless they are deemed directly or indirectly discriminatory. Crucially no ‘hard evidence’ is required, leaving some interpretative leeway to the judiciary in assessing the restrictive effect of the rule before deciding whether to declare those indirectly discriminatory and in need of justification. Furthermore, what is relevant in those cases is not so much discrimination on grounds of place of production of the good in question, but rather market access discrimination. (4) Rules banning advertising and rules regulating long-distance sales are presumed to have an effect on intra-EU trade and must be justified. (5) Rules prohibiting the use of a product must be justified because they affect market access; rules regulating use must be satisfied if they have an effect on market access. (6) All other rules that hinder directly or indirectly, actually or potentially intra-EU trade, or rules that affect market access, need to be justified.

Introduction: The ‘Internal Market 2.0’  17 the benefit of consumers. But: if the High Contracting Parties really meant what they laid down in the Rome Treaty, i.e. the establishment of a common (internal) market, they must have also meant to accept the indispensable legal underpinnings thereof as unfolded by the ECJ’s interpretation of the Treaty.

Yet, a critical enquiry into the normative justifications for treating the free movement provisions akin to fundamental rights (or even, in some cases, as superior thereto) in terms of their capacity to circumscribe democratic power and choice on both the national and the EU level, reveals that the critique of over-constitutionalisation can to a certain extent be accepted. As Garben argues in chapter sixteen, this does not support the complete ‘de-constitutionalisation’ of internal market law as a whole, but it does argue for a limitation of the negative reach of the internal market provisions. And we must furthermore reiterate the findings in section II above that in a number of areas of internal market law, there is too much emphasis on uniformity and not on the ‘managed’ diversity that Purnhagen argues for in chapter thirteen. Spaventa in chapter five notes in relation to goods that the current approach is not without its conceptual and practical problems: conceptually because it steers decision-making towards technocraticism and away from the reality of political compromise, contributing to the narrative, dear to the Court and to a certain extent characterising the European integration project at many levels, that the choice between different regulatory frameworks is technical rather than inherently political. Practically because it overestimates the possibility to carry out extensive impact assessments, and it curtails policy experimentation. Furthermore, and as noted by Weatherill in the context of justifications, the Court holds an optimistic view of the overall ability of the legislature to ensure the coherence of regulatory frameworks, a view which fails to take into account the reality of democratic systems which are dynamic in nature, with changing constituencies and priorities.

As to the internal market’s main normative controversy, namely the ‘tension between market liberalisation and the protection of non-market values’,65 it is impossible to provide a single, generally applicable verdict. As de Witte concludes in chapter seven, ‘the balance between deregulation and re-regulation is situated differently in the different service domains. As a consequence, the normative evaluation of the action of the Court of Justice and of the EU’s political institutions is necessarily bound to be context-specific’, and this nuanced insight would seem to apply to all the areas of internal market law. Still, a main sore spot continues to crop up in this context, namely that of the position of labour law, collective rights and social interests in the Court’s vision of the EU’s ‘economic freedoms’, as becomes clear from the analyses especially in the chapters of Antonaki, Nic Shuibhne and Garben. It would seem that here above anything, a reinterpretation of the internal market by reference to Article 3(3) TFEU and the various overriding values contained in the EU Charter of Fundamental Rights and in the Treaties is due.66

65 As

stated by Bruno de Witte in ch 7. above n 58.

66 Schiek

18  Sacha Garben and Inge Govaere Ultimately, it does seem fair to conclude that despite Herculean efforts of experts from academia and practice (as also on display in various contributions to this book) to systematise and rationalise the Treaty provisions, legislation and case law governing the internal market, the degree of technical, theoretical and normative uncertainty about the meaning of free movement of goods, services, workers, capital and the freedom of establishment remains somewhat startling, especially if one considers their foundational importance in the European integration process. And that may be because (even) after all this time, we (still)67 don’t know what these provisions, in the bigger scheme of things, are – and should be – for. This translates into endless discussion about, for instance, to what extent ‘rules whose effect is to limit commercial freedom even where such rules are not aimed at products from other Member States’68 should be caught as prima facie restrictions,69 what type of justification test should be applied to ‘restrictive’ public interest rules of all sorts,70 how the free movement provisions should be interpreted in relation to each other,71 as well as fundamental human/social rights,72 and whether their ‘negative reach’ should correspond to the EU’s ‘positive’ competence to adopt harmonising legislation – opening in turn another can of worms about the limits of Article 114 TFEU and the place of non-market interests in internal market law, and the choice between minimum and maximum harmonisation.73 All this could even be said, only half-jokingly, to translate into uncertainty as to how to actually even call these provisions. Are they ‘free movement provisions’ – which would suggest an emphasis on equal treatment rather than general economic freedom; ‘internal market provisions’ – which places them alongside the legal bases for positive integration and suggests an emphasis on their joint contribution to the same

67 It could be argued, in line with Gareth Davies in ch 15 in this volume, that there once was – in the early days of European integration – a more pronounced ‘vision’ on what EU law and the provisions on goods, establishment, services, workers and capital were to mean, and that the current uncertainty comes from the fact that this vision no longer corresponds to current social and political thinking. ‘It is based on a vision of the Treaties which was bon ton in right-thinking EU circles sixty years ago – when these were the circles that mattered – but resonates rather weakly with the general public now’. Yet integration coûte que coûte is not much of a vision and cannot carry the label ‘constitutional’. For that latter point, see M Cahill, ‘European Integration and European Constitutionalism: Consonances and Dissonances’ in D Augenstein, ‘Integration Through Law’ Revisited: The Making of the European Polity (Surrey, Ashgate, 2012). 68 Joined Cases C-267/91 and C-268/91 Keck and Mithouard, above n 62, para 14. 69 See on the issue of discrimination, market access and selling arrangements, ch 2 by Stefan Enchelmaier in this volume. For an authoritative case to ‘restrict restrictions’, see C Barnard, ‘Restricting Restrictions: Lessons for the EU from the US?’ (2009) 68 Cambridge Law Journal 575. 70 See ch 8 by Ilektra Antonaki in this volume in the context of the free movement of capital. For a case to adapt proportionality to the nature of the restriction, especially in ethically contentious questions, see F de Witte, ‘Sex, Drugs & EU Law: The Recognition of Moral and Ethical Diversity in EU Law’ (2013) 50 Common Market Law Review 1545, 1566. 71 On the convergence between the free movement provisions, see ch 2 by Stefan Enchelmaier. 72 On this issue see ch 16 by Sacha Garben. See also S de Vries et al, Balancing Fundamental Rights with the EU Treaty Freedoms: The European Court of Justice as ‘Tightrope’ Walker (The Hague, Eleven International Publishing, 2012). 73 See ch 12 by Stephen Weatherill in this volume.

Introduction: The ‘Internal Market 2.0’  19 project and thus an argument for convergence; ‘four freedoms’ – which suggests an emphasis on economic freedom rather than equal treatment; or even ‘fundamental freedoms’ – which suggests that they are fundamental rights and deserve to be interpreted as such? For each of the above doctrinal questions, there are judgments and AG opinions pointing in different directions. Certainly, to a significant degree, competing interpretations of core concepts and doctrines are, for anyone except a die-hard positivist, inherent in the indeterminate nature of the law and part of its very method.74 Yet, a minimum degree of coherence, knowability and predictability of the law, and its limits,75 are cornerstones of the Rule of Law. And the legitimacy of EU law, and of the judiciary’s work within it, is dependent on sound legal methods of interpretation, embedded in a sound legal methodology – which has to comprise a substantive vision of the constitutional order.76 Here we must note that Weiler’s turn-of-the century finding about the vacuity of European constitutionalism still holds true and goes to the core of the problem.77 He considered that European integration has not produced a European legal order of constitutionalism without a formal constitution, but the opposite: ‘a constitutional legal order the constitutional theory of which has not been worked out, its long-term, transcendent values not sufficiently elaborated, its ontological elements misunderstood, its social rootedness and legitimacy highly contingent’.78 More concretely still, Cahill points out the ‘conceptual dissonance’ between European integration and constitutionalism, namely that the latter ‘has a long-established and well developed internal connection to orienting ideals of democracy, self-determination, representation, constituent power, separation of powers and fundamental rights, both as legitimating foundations and as outcomes to be achieved’ while the former is not sufficiently oriented to these ideals (but instead mostly to integration itself).79 The case of the provisions on goods, services, workers, capital and establishment are emblematic thereof. It is common ground that they have been central in the CJEU’s constitution building, yet they operate in a normative and theoretical vacuum – one could even say they have created at least part of that vacuum due to their blunt force combined with, ironically, their effectiveness to integrate – but their inadequacy to create.80 This is regrettable, all the more because 74 D Kennedy, ‘A Left Phenomenological Alternative to the Hart/Kelsen Theory of Legal Interpretation’ (2007) 40 Kritische Justiz 296. 75 In a federal order such as the EU, we should add the constitutional concern of respecting the competence and autonomy of the States, enshrined in EU law through the principles of conferral, subsidiarity and national constitutional identity. 76 See also G Conway, The Limits of Legal Reasoning and the European Court of Justice (Cambridge, Cambridge University Press, 2012). 77 JHH Weiler, The Constitution of Europe: Do the Clothes have an Emperor and Other Essays on European Integration (Cambridge, Cambridge University Press, 1999). 78 ibid, 8. 79 See Cahill, above n 67, 28. 80 Davies considers that ‘establishing’ an internal market entails something more than ‘integrating’ it: ‘For what does it mean to establish a functioning internal market? What is required? One can

20  Sacha Garben and Inge Govaere it is unnecessary, seeing as EU primary law now – thanks to the dynamic evolution of the integration process – does provide a rich constitutional context in which to anchor such a maturing development. This seems at once an opportunity and an obligation, to come to an ‘Internal Market 2.0’.

V.  Embedding the Digital Dimension The need for a thick and comprehensive constitutional understanding of, and approach to, EU law is all the more necessary in light of the fundamental changes that technology and digitalisation brings to our societies and markets, to our constitutional rights and the functioning of our democracies. Our next book will delve specifically into the question of constitutional democracy in the digital age,81 but this current edited volume already highlights the mounting importance of this issue in its third part, devoted to the Digital Single Market. Andrea Renda in chapter nine, considers that the digital transformation is changing the traditional, textbook economics of market integration, based on tenets such as economies of scale and the four freedoms. The rise of the digital economy requires a radical change in the policies for the Single Market, as well as in the trade policies that underpin the whole market integration process. Trends such as the virtualisation, servitisation and platformisation of the economy …, coupled with the rise of the Internet of Things and Artificial Intelligence, make market integration at once more appealing and increasingly challenging for EU policymakers, projecting the Single Market into a complete new dimension, in which the ‘Fifth Freedom’ (the free circulation of non- personal data) is intertwined with new concerns with the need to protect fundamental rights, and at the same time secure Europe’s technological sovereignty.

Vassilis Hatzopoulos in chapter ten offers a detailed analysis of the various ways in which specifically the ‘online platform economy’ is disrupting the internal market and offers some proposals for regulatory solutions at EU level in that regard. He points out that: The platform economy questions established legal categories both under national and under EU law. EU internal market law is being affected in many ways, from core market argue that it requires trust between communities, effective communication, perhaps a certain shared understanding and expectations. Cultural and educational similarities can make trade easier, as can a shared language. One can also certainly argue, and indeed it is fiercely argued by many in these days of permanent economic crisis, that a market without redistribution cannot truly work – that it will lead to stresses and inequalities that will ultimately destroy it or the fabric of the societies in which it is embedded. It may be that an internal market requires a high degree of political and economic integration. To ‘establish’, it may be noted, implies creating a thing in a way that is solid and rooted. An established market is not a disembodied one, but one that is solidly connected with and rooted in its society’. Davies, above n 26, 565. 81 S Garben, I Govaere and P Nemitz (eds), Democracy and Fundamental Rights in the Digital Age (Oxford, Hart Publishing, planned for 2021).

Introduction: The ‘Internal Market 2.0’  21 access conditions and workers’ rights to consumer and data protection. In all these areas the platform economy either stretches the scope and content of existing rules or, more radically, operates in a legal vacuum. The regulatory indeterminacy which has benefited platforms and has allowed them to thrive in the early phases of their development, is now reaching its limits as the size of the platform economy grows by the day. Organised opposition by the incumbents, unhappy user experiences, unexpected negative externalities and even scandals, touching upon life, integrity, dignity and other fundamental rights of users on both sides of the two-sided market, have backfired against platforms and have prompted divergent regulations and/or judicial reactions in different states. Hence, the risk of fragmentation in the internal market along national, and even local, lines is now as real as ever.

This demonstrates clearly how important it is to deal with all these issues in an integrated and holistic fashion, and to have a well-developed normative framework that pays due regard to democracy and fundamental rights embedding concerns for market integration, innovation and competitiveness. As Claire Bury and Irene Roche-Laguna consider in chapter eleven: ‘the Commission should strive to set out a third way of doing policy which is human-centric, ethical and founded on respect for fundamental rights, distinct from both a laissez-faire approach and a top-down controlled model’. They continue: ‘Europe will not be made all at once, or according to a single plan. It will be built through concrete achievements which first create a de facto solidarity’. Thus spoke the Schuman declaration: its purpose was to create an incipient single market for steel and coal under one common High Authority that confronted Germany and France with a de facto solidarity. Alfonso Mattera, one of the ‘fathers’ of the single market, also understood the single market as an instrument of peace. The same applies for the Digital Single Market, which is a translation of these de facto solidarities into the online world. More than ever in the online world, where information and data know no borders and can spread in a millisecond, Member States must continue working together to face common challenges. In so doing they should keep in mind that economic growth and technological development are not objectives per se, but must serve the wellbeing of EU citizens and the pursuit of their common values.

VI.  A Final Word of Dedication This edited volume has tasked itself with imagining an ‘Internal Market 2.0’. This has necessitated, first, an assessment of the lex lata, taking account of the developments from the initial days of European integration to the EU’s current constitutional order, as meticulously carried out by our authors in parts I and II of this book. It has furthermore brought us to shine a spotlight on one of the most influential economic and societal developments of current times, namely widespread digitalisation and technological change, and what it implies for the internal market and the EU legal and political order more generally, as expertly analysed by our authors in part III. Finally, this has led the authors of part IV to

22  Sacha Garben and Inge Govaere courageously confront some important outstanding controversies and questions that the internal market as conceived at present faces, and to reflect constructively on the necessary solutions and (r)evolutions for the lex ferenda. The diversity of the views of our collected authors is matched by the unity of their dedication to think in the best interests of the European integration project that we all hold dear. It is in that spirit that we have dedicated this edited volume to the memory of our dear friend and colleague Alfonso Mattera, a true advocate of the merits of European integration in particular through the establishment of the internal market. In his seminal book Le marché unique européen: ses règles, son fonctionnement, he succinctly pictured the 1992 internal market objective in the following terms: ‘c’est un dessein imaginatif, ambitieux, porteur d’avenir et de progrès’.82 Nearly 30 years later we see it as our task, as he undoubtedly would, to constantly revisit and if necessary, to reinvent the internal market to fit the nature of the ambition. As a final word it is befitting to express our heartfelt thanks, as always, to Mrs Valérie Hauspie without whom this publication would simply not have been possible, for her invaluable editorial and organisational assistance.



82 A Mattera, Le marché unique européen: ses règles, son fonctionnement, 2nd edn (Paris, Jupiter, 1990) 1.

part i The Internal Market and its Development Over Time

24

2 The Development of the Free Movement Principles Over Time STEFAN ENCHELMAIER

I. Introduction The Articles of the Treaty on the Functioning of the European Union (TFEU) that will be discussed in the following have not changed since the Treaty Establishing the European Economic Community. The Court of Justice of the European Union has, however, over the decades refined its interpretation of these provisions. The jurisprudence on each freedom did not develop in lockstep with that on the other freedoms. The case law of the Court of Justice of the European Union (the Court) on the free movement of workers, so far as it concerns the questions addressed below, set in towards the end of the 1960s and continued steadily thereafter. This precociousness is due not least to the steer the Court found in the secondary law adopted in the first decade of the EEC to facilitate the free movement of workers. The jurisprudence on the free movement of goods and on the freedom to provide services began in the first half of the 1970s with the judgments in Dassonville1 and van Binsbergen.2 Its development reached another milestone later that decade in Cassis de Dijon.3 This judgment, however, had precursors roughly two years, and only a few weeks earlier in, respectively, Thieffry4 and van Wesemael.5 The case law on the free movement of goods since Cassis has been characterised as the ‘pacemaker’ among the four freedoms.6 By the late 1980s, however, following the ‘Sunday trading’ line of cases,7 the need for a revision was widely perceived.8 1 Case 8/74 Dassonville EU:C:1974:82. 2 Case 33/74 van Binsbergen EU:C:1974:131. 3 Case 120/78 Rewe-Zentral AG v Bundesmonopolverwaltung für Branntwein EU:C:1979:42 (Cassis de Dijon). 4 Case 71/76 Jean Thieffry v Conseil De L’Ordre Des Avocats À La Cour De Paris EU:C:1977:65. 5 Joined Cases 110/78 and 111/78 van Wesemael EU:C:1979:8. 6 E Steindorff, ‘Gemeinsamer Markt als Binnenmarkt’ (1986) 150 Zeitschrift für Handelsrecht 687, 692. 7 One of several examples is Case 145/88 Torfaen Borough Council v B&Q plc EU:C:1989:593. 8 See, for many, EL White, ‘In Search of the Limits to Article 30 of the EEC Treaty’ (1989) 26 Common Market Law Review 235.

26  Stefan Enchelmaier The judgment in Keck9 was supposed to achieve this, but it also sparked new controversies. These arose mainly around the ill-defined concept of ‘(rules relating to certain) selling arrangements’. The uncertainties that surfaced then are still with us. An early question was whether ‘(rules relating to certain) selling arrangements’ had any relevance for the other freedoms, notably the freedom to provide services and the free movement of workers. Regarding the free movement of goods, restrictions on use had been dealt with occasionally over the years,10 but became the central issue in a judgment of the late 2000s, namely Moped trailers.11 One of the questions that received a good deal of attention during the proceedings (if not in the Court’s judgment itself) was whether and how the idea of ‘(rules relating to certain) selling arrangements’ could be applied to restrictions of use. In that judgment, and in a number on the other freedoms, ‘market access’ gained prominence among the Court’s formulae for assessing free movement cases. Another recent attempt by the Court at restructuring the assessment of free movement cases reflects the recently acquired centrality of human rights in the EU legal order. This piece will consist of a second section that takes stock of the relevant provisions of the Treaty. Also within view comes some secondary law that has its origin in legislation of the 1960s on free movement of workers. It will be seen that the prohibition of discrimination is central to both. For this reason, the third section will trace the development of the concept of discrimination in the Court’s case law, primarily of the 1970s. The Sunday trading cases of the second half of the 1980s, and the line of cases beginning with Keck in the first half of the 1990s are the subject matter of the fourth section. In that section, the implications of the definition of the internal market as ‘an area without internal frontiers’ are understood as the complement to the earlier discrimination test. The fifth section on the convergence of the freedoms analyses judgments in the wake of Keck in which the concepts developed in the previous sections can be seen at work, and judgments in which the Court openly reflected on the transfer of the Keck approach to the cases before it. The same section considers the suitability of ‘market access’ as a universal criterion for assessing free movement cases. It also analyses the Court’s Carpenter judgment, in which human rights assumed an apparently new role in the assessment of free movement cases. The sixth section concludes, and is followed by an annex containing schematic overviews of central themes of the piece, namely of the regulatory scheme of the internal market, the concept of discrimination, and a systematisation of the Court’s case law on the free movement of goods. 9 Joined Cases C-267/91 and C-268/91 Criminal Proceedings against Bernard Keck and Daniel Mithouard EU:C:1993:905. 10 See, for instance, Case C-473/98 Kemikalieinspektionen v Toolex Alpha AB EU:C:2000:379. 11 Case C-110/05 Commission v Italy (Moped trailers) EU:C:2009:66; see also Case C-142/05 Åklagaren v Percy Mickelsson and Joakim Roos EU:C:2009:336.

Development of the Free Movement Principles  27

II.  The Wording of the Treaty and of Early Secondary Law The wording of the Treaty holds a number of cues for the subsequent development of the case law. Before the first substantive amendment by the Single European Act (1986), the fundamental provision concerning free movement was what is now Article 18 TFEU. It prohibits within the scope of application of the Treaty ‘any discrimination on grounds of nationality’, and refers to ‘special provisions’ further on in the Treaty. One such provision is Article 45, according to which free movement of workers entails the ‘abolition’ of any such discrimination between workers of the Member States. In a similar vein, the second sentence of Article 36 prohibits ‘arbitrary discrimination’. A comparison of the wording of Article 45(3) with the functionally identical Article 52 reveals that ‘discrimination’ is unequal (‘special’) treatment, also referred to as ‘apply[ing] dissimilar conditions to equivalent transactions’ in the words of Articles 101(1)(d), 102(c), and in yet another provision, ‘charging different rates and imposing different conditions for the carriage of the same goods’ in Article 95(1). Such unequal treatment amounts to discrimination if there is no justification on one of the grounds listed in Articles 36, 45(3), (4), 51, 52, 62, or 65. It is in this sense that Article 7(4) of Regulation 492/2011 (with the same number and wording in Regulation 1612/68) declares any clause governing conditions of work ‘null and void insofar as it lays down or authorises discriminatory conditions in respect of workers who are nationals of the other Member States’. A finding of discrimination is, hence, the result of the assessment, not its starting point. For unequal treatment that is not (yet) so classified, the Treaty and secondary law use neutral phrases such as ‘special treatment for foreign nationals’, Article 52(1); ‘[without] distinction on grounds of nationality’, Article 61; or ‘conditions not applicable in respect of [a Member State’s] own nationals’, 3(1)(a) of Regulation 492/2011 (identical to Article 3(1), first indent, of Regulation 1612/68). These provisions, especially Article 61, point to another subdivision, viz that between distinctly applicable restrictive measures (‘special treatment’), and measures that do not make a distinction between nationals of (or goods, services, and investments originating in) different Member States, correspondingly referred to as ‘indistinctly applicable measures’. The Treaty leaves the classification open (ie, encompasses both types) when it speaks of ‘obstacles’ in Articles 46(b) and 50(2)(c), and of ‘restrictions’ in Articles 34, 35, 49, 56, 61 and 63. This is also what is meant by ‘less favourable in their direct or indirect effect’, Article 92, and ‘favouring certain undertakings’, Article 107(1). ‘Effects’ are, furthermore, what counts under Articles 30, 34 and 35. At other times, the Treaty envisages indistinctly applicable measures when it uses expressions such as ‘under the conditions laid down for its own nationals’, Article 49, second paragraph; ‘accord nationals of the other Member States the same treatment as their

28  Stefan Enchelmaier own nationals’, Article 55; ‘same conditions as imposed by [a Member State] on its own nationals’, Article 57, third subparagraph; and ‘applicable irrespective of nationality’, Article 3(1)(b) of Regulation 492/2011 (originally Article 3(1), second indent, of Regulation 1612/68). The text of Article 92 in particular (‘less favourable in their direct or indirect effect’) yields one more insight. ‘Special treatment’ (Article 52) will entail less favourable effects ‘directly’: that is the very purpose of singling out goods, persons, services and capital from other Member States. It will also invariably be its factual outcome. By contrast, restrictions that do not draw a ‘distinction on grounds of nationality’ (Article 61) may still be felt more keenly by producers, providers, workers, or investors from other Member States. Such restrictions will not on their face, but only in combination with other legal or factual circumstances and hence ‘indirectly’, be less favourable towards these economic operators. This clashes with the aim of the Treaty and of secondary law to ensure, in the words of Recital 6 to Regulation 492/2011 (Recital 5 to Regulation 1612/68), equality of treatment ‘in fact and in law’. In the absence of a justification for the unequal treatment in either case, the resulting discrimination can be called ‘direct’ and ‘indirect’ discrimination, respectively. The concept of discrimination is at the centre of recurring ambiguity in the Court’s case law. In its most recent as much as in earlier judgments, the Court makes no stringent use of the term ‘discriminatory’. Sometimes it uses the word in a neutral, purely descriptive sense, to mean distinctly applicable measures. In this usage, the term does not denote illegality. It merely designates unequal treatment in law, before any justification is contemplated. The wording of Article 18 and a number of the other provisions discussed above would, by contrast, indicate that ‘discrimination’ implies the illegality of unequal treatment. Some judgments are, therefore, phrased more carefully. In Weigel,12 for instance, ‘indistinctly applicable’ is correctly translated as ‘independent of nationality’ rather than as ‘non-discriminatory’, which is the laxer usage in, for instance, Anomar.13

12 Case C-387/01 Weigel EU:C:2004:256, para 53. 13 Case C-6/01 Anomar EU:C:2003:446, para 69: ‘legislation such as that in issue in the main proceedings which, although it does not discriminate on grounds of nationality, restricts the freedom to provide services’, but see the mixed formulation in para 68: ‘the Portuguese legislation does not discriminate between the nationals of the various Member States. That legislation must therefore be regarded as applying without distinction’. See, more recently, the formulation, ‘to ensure that the restriction, first, does not discriminate on grounds of nationality, next, is justified by an overriding reason relating to the public interest’ in Cases C-473/17 Repsol Butano EU:C:2019:308, para 45; C-293/14 Hiebler EU:C:2015:843, para 55; and C-539/11 Ottica New Line di Accardi Vincenzo EU:C:2013:591, para 39: ‘[i]t is not disputed, first of all, that Regional Law No 12/2004 does not discriminate on grounds of nationality’, but then at para 51, ‘[s]uch legislation risks bringing about unequal access to the establishment of opticians’ shops in the various areas of the region concerned’. See also Case C-64/08 Engelmann EU:C:2010:506, para 34: ‘It is apparent from the Court’s case-law that, to the extent that a restriction, such as that which has been found to exist in the present case, is discriminatory, it is compatible with European Union law only if it is covered by an express derogating provision, such as Article 46 EC, namely public policy, public security or public health’.

Development of the Free Movement Principles  29

III.  The Concept of Discrimination Unfolded A.  ‘Direct’ and ‘Indirect’ Discrimination and the Early Case Law on Workers Developing the concept of discrimination was at the heart of the Court’s jurisprudence on the four freedoms in the 1970s. On the eve of that decade, the Court in Ugliola held that Article 45 prohibited the Member States from ‘indirectly introducing discrimination in favour of their own nationals alone’.14 In that case, German law extended the protection of employment relationships during absence for national service only to men who served in the German armed forces, not to those nationals of other Member States living in Germany who were called up in their home country. ‘National service’ is not the same criterion as ‘nationality’ for the enjoyment of the protection. The disadvantage for foreign nationals of not being able to benefit from employment protection resulted from the fact that they are not subject to the draft into the German army. This liability, in turn, was not laid down in said provisions of employment law, but in a separate statute. The disadvantage in this way arose only from the combined effect of both pieces of legislation, and hence from the law in issue only indirectly, not from that law of its own, and thus not directly. As a consequence of this judgment, it will not be enough to find that a provision of national law applies in equal measure to all workers of whatever nationality. This will only lead to the next question, namely whether anyone regardless of nationality finds it equally easy to comply with the legislation, or to qualify for its benefits. In other words, a finding of indistinct applicability (‘national treatment’) does not mark the end of the assessment. It merely shifts the focus of the assessment to the factual repercussions of a given national rule. Shortly after Ugliola, in Sotgiu, the Court summarised the position to the effect that the rules regarding equality of treatment, both in the Treaty and in Article 7 of Regulation No 1612/68, forbid not only overt discrimination by reason of nationality but also all covert forms of discrimination which, by the application of other criteria of differentiation, lead in fact to the same result.15

It can readily be seen that the Court here uses ‘overt’ synonymously with ‘direct’ (discrimination), and that ‘covert’ corresponds with ‘indirectly’.

B.  ‘Direct’ and ‘Indirect’ Hindrances in Dassonville Around the same time, in Dassonville the Court first broached the interpretation of the phrase ‘measures having equivalent effect to quantitative restrictions’ in Article 34.

14 Case 15 Case

15/69 Ugliola EU:C:1969:46, para 6. 152/73 Sotgiu EU:C:1974:13, para 11.

30  Stefan Enchelmaier Its formula – ‘all trading rules enacted by Member States which are capable of hindering, directly or indirectly, actually or potentially, intra-Community trade are to be considered as measures having equivalent effect to quantitative restrictions’16 – is not entirely new. The Court had already used ‘direct or indirect, actual or potential’ (hindrance) for deciding whether an agreement between undertakings was such as to ‘affect trade between Member States’ in the sense of Article 85(1) EEC (now Article 101(1) TFEU).17 The meaning of the prohibition addressed at private parties was in that judgment derived from what the Member States were not allowed to do.18 In Dassonville, the Court simply inverted this to give contours to the corresponding prohibition addressed to the Member States, as AG Trabucchi had recommended.19 A few months later, the Court transferred this idea to the case law on Articles 45 and 56.20 Although the ‘direct or indirect’ formula had been familiar since the 1960s from the legislation and jurisprudence on the free movement of workers, the Court in Dassonville did not make any connection with these precursors. This may have been because the case did not necessitate the Court’s elaborately establishing unequal treatment between Belgian products and imports. The Belgian rule in issue – the requirement for drinks bearing a foreign protected designation of origin to be accompanied by a certificate of authenticity issued by the authorities of the country under whose laws the designation was protected – did not apply to Belgian products. The rule was therefore distinctly applicable: it provided for ‘special treatment’ (to borrow the language of Article 52) for foreign products. The national court could with some conviction come to the conclusion that, going by the Court’s guidance, the requirement of a certificate of authenticity did qualify as a measure having equivalent effect to a quantitative restriction. The

16 Dassonville, above n 1, para 5. 17 Joined Cases 56/64 and 58/64 Consten and Grundig v Commission EU:C:1966:41 [1966] ECR 299 341, last two paras: ‘It is only to the extent to which the agreement may affect trade between Member States that the deterioration in competition caused by the agreement falls under the prohibition of Community law contained in Art 85; otherwise it escapes the prohibition. In this connexion, what is particularly important is whether the agreement is capable of constituting a threat, either direct or indirect, actual or potential, to freedom of trade between Member States in a manner which might harm the attainment of the objectives of a single market between states’ (emphasis added). 18 Consten and Grundig, above n 17, 340, second para: ‘an agreement between producer and distributor which might tend to restore the national divisions in trade between Member States might be such as to frustrate the most fundamental objectives of the Community. The Treaty, whose preamble and content aim at abolishing the barriers between states, and which in several provisions gives evidence of a stern attitude with regard to their reappearance, could not allow undertakings to reconstruct such barriers. Art 85(1) is designed to pursue this aim, even in the case of agreements between undertakings placed at different levels in the economic process’. 19 Dassonville, above n 1, [1974] ECR 857, right-hand column. 20 Case 36/74 Walrave and Koch EU:C:1974:140, para 18: ‘The abolition as between Member States of obstacles to freedom of movement for persons and to freedom to provide services … would be compromised if the abolition of barriers of national origin could be neutralized by obstacles resulting from the exercise of their legal autonomy by associations or organizations which do not come under public law’.

Development of the Free Movement Principles  31 passage in the judgment on the permissibility (justification) of such rules, by contrast, was presumably less helpful. It boiled down to the requirement that the restriction be ‘reasonable’, without any elaboration.21

C.  Van Binsbergen, van Wesemael and the Existence of Grounds of Justification Beyond the Wording of the Treaty Only few months later, in the services case van Binsbergen, the Court developed the point more fully. It explained that depending on the nature of the services in issue, specific requirements imposed on the service provider were compatible with the Treaty where they have as their purpose the application of professional rules justified by the general good – in particular rules relating to organisation, qualifications, professional ethics, supervision and liability – which are binding upon any person established in the state in which the service is provided.

This would apply where otherwise, ‘the person providing the service would escape from the ambit of those rules being established in another Member State’.22 When the Court speaks of ‘compatibility with the Treaty’ despite the restrictive nature of the measures in question, it is addressing the question of justification. In this respect, we learn that the rules must be ‘binding upon any person established in the state in which the service is provided’. That is, they must be indistinctly applicable. Examples are rules addressing questions of ‘organisation, qualifications, professional ethics, supervision and liability’. The rules must, furthermore, be ‘justified by the general good’. The Court does not spell out what the ‘general good’ encompasses beyond the examples mentioned. Nevertheless, it is readily apparent that this concept is wider than the grounds of justification that Article 52 makes available for ‘special treatment for foreign nationals’, ie, for distinctly applicable measures. These written grounds are, as it were, a selection of (particularly important) aspects, or emanations, of the general good. The selective nature of the lists in the Treaty articles, the character of the grounds as exceptions from the fundamental principle of free movement, translates into the necessarily restrictive interpretation given to the grounds enumerated.23 These catalogues of grounds of justification in the Treaty cannot be added to, the grounds must in themselves be interpreted narrowly, and they

21 Dassonville, above n 1, para 6: ‘In the absence of a Community system guaranteeing for consumers the authenticity of a product’s designation of origin, if a Member State takes measures to prevent unfair practices in this connexion, it is however subject to the condition that these measures should be reasonable’. 22 Van Binsbergen, above n 2, paras 10–12; likewise in van Wesemael, above n 5, para 28. 23 Case 67/74 Bonsignore EU:C:1975:34, paras 5–6.

32  Stefan Enchelmaier cannot be applied by analogy to other chapters of the Treaty.24 Van Binsbergen, by contrast, allows Member States to select any aspect of the general good to justify indistinctly applicable restrictions. This leaves the second condition in van Binsbergen, namely that the service provider would otherwise ‘escape from the ambit of those rules being established in another Member State’. The Court elucidated this in a subsequent case, van Wesemael. The judgment reiterated what was already implicit in van Binsbergen: a Member State must not, in principle, impose on service providers operating out of other Member States the same rules that apply to providers established within its jurisdiction, precisely because these foreign providers do not become established on its territory. As already seen in van Binsbergen, this rule suffers an exception for restrictions imposed in pursuit of the general good, such as the requirement to obtain a licence or to submit to the supervision of public authorities.25 Crucially, however, these restrictions are not objectively justified when the provider comes under the public administration of a Member State or when the person providing the service is established in another Member State and in that state holds a licence issued under conditions comparable to those required by the state in which the service is provided and his activities are subject in the first state to proper supervision covering all [the provider’s] activity whatever may be the Member State in which the service is provided.26

In short, unjustified is the duplication of requirements that the service provider has already met in the Member State of his or her establishment.27 Member States can only insist on enforcing their own rules if the interests they pursue are not protected to the same standard by the rules in the other Member State to which the producer or provider is subject.28 For that, and to that extent, Member States have to give operators from other Member States regulatory credit, so to say: they have to recognise such guarantees as the foreign providers can offer that they will observe the ‘general good’ of the host Member State, howsoever defined by that State.

24 Case 46/76 Bauhuis EU:C:1977:6, paras 12–14. 25 Van Wesemael, above n 5, para 29. 26 ibid, para 30. 27 See, for a contemporary example, Case 16/78 Choquet EU:C:1978:210, para 8, condemning a Member State’s ‘[i]nsistence on a driving test which clearly duplicates a test taken in another Member State for the classes of vehicle which the person concerned wishes to drive’. 28 Case 279/80 Webb EU:C:1981:314, paras 17, 20: ‘The freedom to provide services is one of the fundamental principles of the Treaty and may be restricted only by provisions which are justified by the general good and which are imposed on all persons or undertakings operating in the said state in so far as that interest is not safeguarded by the provisions to which the provider of the service is subject in the member state of his establishment … Such a measure would be excessive in relation to the aim pursued, however, if the requirements to which the issue of a licence is subject coincided with the proofs and guarantees required in the state of establishment’ (emphasis added).

Development of the Free Movement Principles  33

D.  Cassis de Dijon and the Justification of Indistinctly Applicable Restrictions Up to this point, the question was open which (if any) grounds of justification Member States could invoke beyond those expressly mentioned in the Treaty for restrictions to free movement. Van Binsbergen had made a start by allowing restrictions imposed on all providers in the interest of the ‘general good’ as exemplified by the concerns of ‘organisation, qualification, professional ethics’ etc. That judgment had not, however, given the national court more detailed orientation. The opportunity to clarify this came with Cassis de Dijon,29 the next case on free movement of goods. We have seen above the subdivision of restrictions into distinctly and indistinctly applicable measures. Distinctly applicable measures, singling out nationals or goods from other Member States, are diametrically opposed to the idea of creating one, continent-wide market to which everyone everywhere has, in principle, free and equal access. Member States’ ability to adopt such measures must therefore be tightly controlled. The Treaty does so by allowing only a limited number of permissible aims that Member States may pursue by means of distinctly applicable measures. As already seen, to the grounds listed in Article 36 and the equivalent provisions on the other freedoms, none may be added, and those that are enumerated must be interpreted narrowly. This changes with the other type of national provisions, indistinctly applicable measures: the same rules apply to all goods coming from or going to other Member States, as well as to those made and sold within the borders of that Member State alone. With such rules, there is no ‘special treatment’ of goods etc from other Member States in law. In fact, however, it may turn out that foreign goods find it specifically more difficult than those made in that Member State to comply with the rules. Products made in accordance with the rules in force in other Member States may not conform to the rules applicable in the importing Member State. The producer and/or its distributors will then, like Rewe in Cassis de Dijon, be left with the choice either to adapt the product at potentially prohibitive expense, or not to market it in that Member State at all. It is this factual inequality that makes indistinctly applicable national measures problematic. Some inequality, however, is unavoidable with all legal rules. As soon as the legislature stipulates any conditions for a particular legal consequence to arise, some people will meet the conditions, while others will not: those whose beverage contains less than 25 per cent alcohol must not call it ‘fruit liqueur’, but those may whose drink contains that much or more; those who fail to rescind a contract within the statutory cooling-off period are stuck with the contract and must pay for something they no longer want, while those are free who rescinded



29 Above

n 3.

34  Stefan Enchelmaier in time. Some are ‘in’: they benefit from the rule; some are ‘out’, they are denied the benefit, or they have to shoulder a (heavier) burden. With distinctly applicable measures, the inequality is intentionally created. With indistinctly applicable measures, by contrast, the inequality is the inescapable consequence of any and all legislation. That has consequences for the Member States’ ability to justify each type of measure. With the first, Member States are limited to the grounds of justification that the Treaty expressly allows. Nevertheless, if on these grounds distinctly applicable measures can be justified, then they must also be available for the prima facie less harmful indistinctly applicable measures. What is more, in the case of indistinctly applicable measures Member States are not limited to the grounds written in the Treaty. The rescission period in the example above will have been enacted by the Member State to give consumers a window during which they can change their mind. Afterwards, however, the reliability of contractual relationships (pacta sunt servanda) must prevail. In the interest of that reliability, the legislator accepts that those who miss the deadline cannot get away from the contract, while those who keep it can. In other words, the policy motivation that underlies the legislation doubles as the ground of justification for the inequality that the rule inevitably entails. The Court in Cassis de Dijon coined the phrase ‘mandatory requirements’ for such grounds of justification beyond Article 36, and defined them as, ‘a purpose which is in the general interest and such as to take precedence over the requirements of the free movement of goods’.30 These policy motivations are potentially endless in number: anything is permissible that is not incompatible with the Treaty or any binding secondary Union law adopted on the basis of the Treaty. As long as Member States observe the principle of proportionality (to which the Court referred obliquely by stipulating that the national measure be ‘necessary [in order to satisfy mandatory requirements]’), they are free to seek to achieve anything that they have identified as the ‘general good’ to which the Court referred in van Binsbergen. Member States enjoy legislative autonomy. The idea of ‘regulatory credit’ introduced above in the discussion of van Wesemael is since Cassis de Dijon commonly referred to as ‘mutual recognition’.31 This term has its origin in the Treaty, namely Article 57(1) EEC (now Article 53 TFEU), which since 1957 empowered the Community/Union to ‘issue Directives for the mutual recognition of diplomas, certificates and other evidence of formal qualifications’. Even in the absence of such Directives, in Thieffry the Court had held that, there is an unjustified restriction on [the freedom of establishment] where, in a Member State, admission to a particular profession is refused to a person covered by the Treaty who holds a diploma which has been recognized as an equivalent qualification by the 30 Cassis de Dijon, above n 3, para 14. 31 This usage has over the years extended from the Court’s jurisprudence to the Union’s legislation, see for a recent example Regulation (EU) 2019/515 of the European Parliament and of the Council of 19 March 2019 on the mutual recognition of goods lawfully marketed in another Member State and repealing Regulation (EC) No 764/2008 [2019] OJ L91, 1–18.

Development of the Free Movement Principles  35 competent authority of the country of establishment and who furthermore has fulfilled the specific conditions regarding professional training in force in that country, solely by reason of the fact that the person concerned does not possess the national diploma corresponding to the diploma which he holds and which has been recognized as an equivalent qualification.

In this context, ‘in each Member State, the recognition of evidence of a professional qualification for the purposes of establishment may be accepted to the full extent compatible with the observance of the professional requirements mentioned above’.32 In other words, qualifications obtained in other Member States must be taken on their merits, and assessed in light of the public interest the Member State seeks to protect. In Cassis, the Court substituted the qualities that goods have received on account of their compliance with the rules in force in the Member State of their production, for the qualifications that individuals have obtained in the course of their training in another Member State. In each case, the other Member State must give credit for such guarantees to its own public interest that result from conformity with the exporting Member State’s legislation. The familiar phrase, ‘mutual recognition’, must not be taken literally, though. The recognition is not ‘mutual’ because it is irrelevant for the outcome in Cassis what treatment German-made fruit liqueur experienced in France, or in Thieffry, whether French-trained lawyers would have their degrees recognised as equivalent in Belgium. As usual, one Member State is not absolved from observing its obligations under EU law by other Member States failing to abide by theirs.33 The recognition is also not automatic, but merely conditional: it will not take place if the 32 Thieffry, above n 4, paras 19, 23. 33 See, eg, Joined Cases 90/63 and 91/63 Commission v Belgium and Luxembourg (Milk import licences) EU:C:1964:80, para 631: ‘In [the governments’] view, since international law allows a party, injured by the failure of another party to perform its obligations, to withhold performance of its own, the Commission has lost the right to plead infringement of the Treaty. However this relationship between the obligations of parties cannot be recognized under Community law. In fact the Treaty is not limited to creating reciprocal obligations between the different natural and legal persons to whom it is applicable, but establishes a new legal order which governs the powers, rights and obligations of the said persons, as well as the necessary procedures for taking cognizance of and penalizing any breach of it. Therefore, except where otherwise expressly provided, the basic concept of the Treaty requires that the Member States shall not take the law into their own hands. Therefore the fact that the Council failed to carry out its obligations cannot relieve the defendants from carrying out theirs’; more recently, Case C-5/94 Hedley Lomas EU:C:1996:205, para 20: ‘A Member State may not unilaterally adopt, on its own authority, corrective or protective measures designed to obviate any breach by another Member State of rules of Community law’; Case C-266/03, Commission v Luxembourg (Inland waterways agreement) EU:C:2005:341, para 35: ‘under the system laid down by Art 258 TFEU, … a Member State cannot, in any event, plead the principle of reciprocity and rely on a possible infringement of the Treaty by another Member State in order to justify its own default’; Case C-45/07 Commission v Greece (International Maritime Organisation) EU:C:2009:81, para 26; most recently in Case C-98/16 Comission v Greece (Donations mortis causa) EU:C:2017:346, para 51 (available only in French): ‘en adoptant et en maintenant en vigueur une législation qui prévoit un taux préférentiel des droits de succession pour les legs effectués en faveur d’organismes sans but lucratif qui sont établis dans d’autres États membres de l’Union ou de l’EEE [EEA] sous réserve de réciprocité, la République hellénique a manqué aux obligations qui lui incombent en vertu de l’article 63 TFUE et de l’article 40 de l’accord EEE’.

36  Stefan Enchelmaier Member State succeeds in justifying its rules on one of the grounds in Article 36, or a mandatory requirement. All the same, if the host State fails to extend such credit where it is due, it commits indirect discrimination, because the rules in question are ex hypothesi indistinctly applicable. If, on the other hand, the provider has no guarantees to offer because it is not subject to any meaningful supervision by the Member State that authorised its activities, it will be subject to the whole gamut of the host Member State’s legislation (provided the rules are also proportional).34 Following Cassis de Dijon, doubts arose as to which sorts of national measures could benefit from ‘mandatory requirements’. Barely a year after its seminal decision, the Court therefore clarified that [i]t is only where national rules, which apply without discrimination[35] to both domestic and imported products, may be justified as being necessary in order to satisfy imperative requirements … that they may constitute an exception to the requirements arising under Article [34].36

Such rules require a justification if, although they apply to national and imported products alike, their effects are protective in nature.37 The Court was equally emphatic that distinctly applicable measures, by contrast, cannot be justified on any grounds other than those written in the Treaty.38 34 See, for an example, Case C-46/08 Carmen Media Group Ltd EU:C:2010:505, where the eponymous company was issued by the authorities of Gibraltar with an ‘offshore bookmaking licence’ against nothing more than payment of the applicable fees. This licence allowed Carmen Media to offer its (betting) services anywhere except in Gibraltar. Nevertheless, the Court emphasised (at para 50) that, ‘the question of the applicability of Art 56 TFEU is distinct from that of whether a Member State may take measures to prevent a provider of services established in another Member State from circumventing its internal legislation’. 35 ‘Discrimination’ here means ‘distinction’, see above by and in n 13. 36 Case 788/79 Gilli and Andres EU:C:1980:171, para 6. The Court found a more succinct formula without the imprecise use of the term ‘discriminatory’ in Case C-21/88 Du Pont de Nemours Italiana EU:C:1990:121, para 14: ‘[Imperative requirements] recognized by the Court in its case law … may be taken into consideration only in relation to measures which are applicable to domestic products and to imported products without distinction’. Similarly, the Court held in Case 274/87 Commission v Germany (Beer purity) EU:C:1989:51, para 5: ‘[T]he contested rules prohibit the marketing of the products concerned … regardless of whether they are national or foreign products. Such measures which are applicable without distinction are not affected by the prohibition of measures having equivalent effect, provided that they are necessary in order to meet certain imperative requirements’; and in Case C-434/85 Allen and Hanburys v Generics EU:C:1988:109, para 35: ‘[I]t is only where national rules apply without distinction to both domestic and imported products that they do not fall under the prohibition laid down by Art [34] of the Treaty if they are necessary in order to satisfy imperative requirements’. 37 Case C-193/80 Commission v Italy (Vinegar) EU:C:1981:298, para 20. A more elaborate version of this can be found in Case C-16/83 Prantl EU:C:1984:101, para 21: ‘[E]ven national legislation on the marketing of a product which applies to national and imported products alike falls under the prohibition laid down in Art 34 TFEU if in practice it produces protective effects by favouring typical national products and, by the same token, operating to the detriment of certain types of products from other Member States’. 38 Case 113/80 Commission v Ireland (Souvenirs) EU:C:1981:139, para 11: ‘The orders concerned in the present case are not measures which are applicable to domestic products and to imported products without distinction but rather a set of rules which apply only to imported products and are therefore discriminatory in nature, with the result that the measures in issue are not covered by the decisions

Development of the Free Movement Principles  37

E.  Walloon Waste and the Concept of ‘(In)distinctly Applicable’ Restrictions All the same, the matter remains controversial. The protection of the environment, in particular, was suggested as a ground that should apply across the board. The Walloon waste judgment39 is invariably cited as proof,40 most recently by Advocate General Bot.41 In Walloon waste, a rule enacted by the Belgian region of Wallonia was in issue that prohibited (among others) the disposal in Wallonia of waste originating in other Member States. More specifically, the Walloon legislation defined ‘waste not produced in Wallonia’ as ‘waste from a foreign state or a region of Belgium other than Wallonia’.42 Separate but substantially identical rules spelled out that both types of waste must not be brought into Wallonia. There might have been differences in the exceptions; the legislation was not clear on that point. Regardless, the Court only discussed the question whether the Walloon rules treated differently Walloon waste on the one hand, and foreign as well as non-Walloon Belgian waste on the other.43 We had learned earlier that Wallonia’s limited tipping capacities were being overwhelmed by an inflow of waste from other regions (of Belgium, that is).44

[in Cassis de Dijon and others] which relate exclusively to provisions that regulate in a uniform manner the marketing of domestic products and imported products’; similarly in Case C-177/83 Kohl v Ringelhan EU:C:1984:334, paras 14 and 19, and in Joined Cases C-321/94 to C-324/94 Pistre and others EU:C:1997:229, para 52. 39 Case C-2/90 Commission v Belgium (Walloon waste) EU:C:1992:310. 40 E Spaventa, ‘On discrimination and the theory of mandatory requirements’ (2001) 3 Cambridge Yearbook of European Legal Studies 457, 462 f; C Barnard, ‘Fitting the remaining pieces into the goods and persons jigsaw?’ (2001) 26 European Law Review 35, 54: the ECJ ignores the distinction between Art 36 and mandatory requirements, as and when it fits it; V Hatzopoulos, ‘Exigences essentielles, impératives ou impérieuses: une théorie, des théories ou pas de théorie du tout?’ (1998) 34 Revue trimestrielle de droit européen 191, 198 f: ‘example de voluntarisme juridique extrême’; J Snell, Goods and Services in EC Law. A Study of the Relationship Between the Freedoms (Oxford, Oxford University Press, 2002) 179 f; N Nic Shuibhne, The Coherence of EU Free Movement Law. Constitutional Responsibility and the Court of Justice (Oxford, Oxford University Press, 2013) 203. 41 Joined Cases C-204/12 to C-208/12 Essent Belgium (no 1) EU:C:2013:294 opinion, paras 87–98, culminating in his calling (at para 96) for an ‘[e]xpress elevation of environmental protection to the level of an imperative requirement relating to public interest that may be invoked in order to justify measures that restrict the freedoms of movement even where such measures are discriminatory’ – a merry mixture of ‘mandatory requirements’ with the list in Art 36, and of ‘discriminatory’ with ‘distinctly applicable’, supposedly foremost ‘dictated by a concern for legal certainty’, para 93. AG Bot was, however, more guarded in Case C-492/14 Essent Belgium (no 2) EU:C:2016:257 opinion, para 65: ‘A phenomenon whereby the discrimination is concealed emerges clearly from the case-law, although the process that is at play in reaching that result is shrouded in a certain degree of mystery’. AG Trstenjak also openly called for allowing the protection of the environment as a ground of justification for distinctly applicable measures in Case C-28/09 Commission v Austria (Inntalautobahn No 2) EU:C:2011:854, paras 78–91, as had AG Jacobs in paras 32–41 of his opinion in Case C-136/00 Danner EU:C:2002:558. 42 See the national rules reproduced on page I-4433 of the printed European Court Reports for 1992. 43 Walloon waste, above n 39, paras 33–36. 44 ibid, paras 30, 31.

38  Stefan Enchelmaier In this situation, and given that the Community (as it then was) subscribed to the principles of self-sufficiency and of proximity, Wallonia was allowed to draw a distinction between consignments of waste produced in different ‘places’ (not ‘Member States’!): ‘having regard to the differences between waste produced in different places and to the connection of the waste with its place of production, the contested measures cannot be regarded as discriminatory’, the Court held.45 The Court’s analysis was prompted by the Commission’s argument at paragraph 33 that, ‘imperative requirements cannot be relied upon in the present case, given that the measures in question discriminate against waste originating in other Member States, which is no more harmful than waste produced in Wallonia’. The Court in its reply (paragraph 34) again uses ‘apply without distinction’ side by side with the formulation that it was, ‘assessing whether or not the barrier in question is discriminatory’.46 The Court returned to ‘discriminatory’ in its conclusion (paragraph 36, quoted above). Both the Commission’s argument and the Court’s answer are ambiguous. The Commission’s objection to Belgium’s attempt at justifying its measures can be read to mean that the protection of the environment in this specific case does not provide a justification (and hence, that there is discrimination), because waste from other Member States is no more harmful than waste originating in Belgium. Alternatively, it can mean that the protection of the environment cannot be invoked from the outset because the Belgian measure is distinctly applicable. The Court, in its turn, does not make it clear whether it is assessing the applicability of the Belgian measure (distinctly or indistinctly) or its justification, in the absence of which the measure would be discriminatory. The Court’s conclusion, ‘the contested measures cannot be regarded as discriminatory’, again allows both interpretations. When it speaks of the ‘differences between waste produced in different places’, it merely echoes (in the negative) the Commission’s ‘no more harmful than waste produced in Wallonia’. The Court presumably meant to say that the difference in the environmental impact of local (Walloon) waste and waste imported from wherever outside Wallonia implied that the rules were not distinctly applicable. This difference came from the principle that, it is for each region, municipality or other local authority to take appropriate steps to ensure that its own waste is collected, treated and disposed of; it must accordingly be disposed of as close as possible to the place where it is produced, in order to limit as far as possible the transport of waste.47

45 ibid, para 36. 46 ibid, para 34: ‘Imperative requirements can indeed be taken into account only in the case of measures which apply without distinction to both domestic and imported products … However, in assessing whether or not the barrier in question is discriminatory, account must be taken of the particular nature of waste’. 47 ibid.

Development of the Free Movement Principles  39 In other words, one place’s surplus waste is not another place’s concern. One need not allow the other to dump its problems on it. In legal terms the prohibition of discrimination, meaning unequal treatment without justification, is the negative flipside of the principle of equality. This principle can be formulated in four different ways, all of which mean essentially the same (but facilitate dealing with different situations): equals must be treated equally; unequals must be treated unequally; equals must not be treated unequally; and unequals must not be treated equally.48 The Court in Walloon waste is making the point that waste from one place is not the equal of waste from another place, so that it cannot demand to be treated equally. Where there is no equality of subject matter, unequal treatment cannot amount to discrimination. Instead, there are two sets of rules, each dealing with separate situations. ‘Place’ in the judgment translates into ‘region, municipality or other local authority’,49 the concern being a reduction as far as possible of waste shipments between these. Therein lies the problem with the Court’s reasoning: waste from Flanders and Brussels was no less of a burden on the strained Walloon capacity and on the environment there than was waste from neighbouring districts of France and Germany that had to travel the same distance to Walloon disposal sites. Nor was waste necessarily less burdensome that had to make its way to the tip from within the same region, Wallonia. The political categories, Member State – region – municipality, have no necessary correlation with the criterion that counts for the reduction in the use of transport, namely distance. To put it another way, ‘region’ is still too wide a circle to carry a differentiation according to distance travelled. All waste within the same radius from a disposal site is equal in this perspective.50 ‘Municipality’ or ‘local authority’ might have been sufficiently narrow, but these were not the territorial units that the Walloon legislation had chosen: the prohibition of waste imports applied to all of Wallonia without further subdivision. The question can also be approached from a different angle. The Commission brought its action under Article 258 TFEU against Belgium. This is because Wallonia is not a Member State and cannot, for that reason, be a ‘defendant’ in an action by the Commission for failure to comply with EU law. In Wallonia’s stead

48 The table in the Annex illustrates and explains this for all freedoms. 49 Walloon waste, above n 39, para 34. 50 Conversely, in Case C-388/01 Commission v Italy (Museum admissions) EU:C:2003:30 (local authorities granted free admission in favour of Italian nationals and persons aged over 60 or 65 years resident within their territory to municipal museums and other sites classified in Italian legislation as national monuments), the Court highlighted at para 24 that other persons resident in Italy were also subject to taxation that contributed to the running of the monuments. The exclusion of these persons from free admission was, therefore, not justified. This case shows that who must be included and who may be excluded, depends solely on whether in their person, the criterion stipulated in the legislation is realised. In Museum admissions, the criterion was a financial contribution to the upkeep of the monuments, and the exclusion was excessive; in Walloon waste, it was transportation of the waste over a distance, and the inclusion of all waste originating anywhere in Wallonia went too far.

40  Stefan Enchelmaier (vicariously, as it were), Belgium found itself summoned before the Court. Only the central government of a Member State is answerable for breaches of EU law by sub-central bodies vested with public authority.51 The procedural side is, of course, without prejudice to the substance of the litigation. The fact that Belgium was answerable rather than Wallonia did not extend the territorial application of the regional provision to the whole of Belgium. From the perspective of Belgium as a whole, the measure disadvantaged waste originating in Belgian regions other than Wallonia as much as it disadvantaged waste from other Member States. The import prohibition caught any waste of non-Walloon origin. Hence, although the goods benefiting from the rule were all Belgian (Walloon waste), some other Belgian waste (namely, from Flanders and Brussels) was in the less favoured category, together with all waste from other Member States. Such ‘mixed’ groups of the disadvantaged, including some national goods or persons among foreigners and imports, are characteristic of indistinctly applicable measures.52 Whichever way one looks at it, therefore, in Walloon waste the Court did not accept a justification of distinctly applicable measures on grounds of protection of the environment. Nor should the Court do so: Member States ought not to be given a freer hand than they currently enjoy to single out goods from other Member States. Should the Member States find the current exclusion of the environment from the grounds listed in Article 36 intolerable, it would be in their power to amend the Treaty.

F.  Lawyers’ Fees and the Question of ‘Unequal Treatment’ Although the discrimination test is well established in the jurisprudence, questions regarding even the fundamentals of the assessment still come up from time to time. Two actions brought by the Commission against Italy illustrate this – Third-party liability motor insurance53 and Lawyers’ fees.54 The problem that the Court tackled in both judgments is how to assess whether indistinctly applicable measures (in casu, regulations on services) in fact place providers from other Member States at a 51 Case 197/84 Steinhauser EU:C:1985:260, para 16; Cases C-145/97 Commission v Belgium (Furnished Accommodation) EU:C:1998:212, paras 1, 6, 14; C-383/00, Commission v Germany (Major-accidents hazards) EU:C:2002:289, paras 7–11, 18; C-423/00 Commission v Belgium (Major-accident hazards) EU:C:2002:32, para 16; C-388/01, above n 50, para 27; C-53/10 Mücksch EU:C:2011:585, para 30. 52 Museum admissions, above n 50, para 14; Du Pont de Nemours Italiana, above n 36 paras 12–13, for regional preferences in public procurement; more recently, Case C-20/12 Giersch EU:C:2013:411 para 45: ‘it is immaterial whether, in some circumstances, the contested measure affects, as well as nationals of other Member States, nationals of the Member State in question who are unable to meet such a criterion. In order for a measure to be treated as being indirectly discriminatory, it is not necessary for it to have the effect of placing all the nationals of the Member State in question at an advantage or of placing at a disadvantage only nationals of other Member States, but not nationals of the State in question’. 53 Case C-518/06 Commission v Italy (Third-party motor liability insurance) EU:C:2009:270. 54 Case C-565/08 Commission v Italy (Lawyers’ fees) EU:C:2011:188.

Development of the Free Movement Principles  41 disadvantage that their competitors established in the regulating Member State are spared. As a consequence, these measures amount to restrictions, and are hence in need of a justification. In the two judgments, the Court approached this question in two different and incompatible ways. In Motor insurance, the Court held that, [i]nasmuch as it obliges insurance undertakings which enter the Italian market to accept every potential customer, that obligation to contract is likely to lead, in terms of organisation and investment, to significant additional costs for such undertakings. If they wish to enter the Italian market under conditions which comply with Italian legislation, such undertakings will be required to re-think their business policy and strategy, inter alia, by considerably expanding the range of insurance services offered. Inasmuch as it involves changes and costs on such a scale for those undertakings, the obligation to contract renders access to the Italian market less attractive and, if they obtain access to that market, reduces the ability of the undertakings concerned to compete effectively, from the outset, against undertakings traditionally established in Italy.55

In Lawyers’ fees, by contrast, the Court found that, [r]ules of a Member State do not constitute a restriction within the meaning of the EC Treaty solely by virtue of the fact that other Member States apply less strict, or more commercially favourable, rules to providers of similar services established in their territory … The existence of a restriction within the meaning of the Treaty cannot therefore be inferred from the mere fact that lawyers established in Member States other than the Italian Republic must become accustomed to the rules applicable in that latter Member State for the calculation of their fees for services provided in Italy. By contrast, such a restriction exists, in particular, if those lawyers are deprived of the opportunity of gaining access to the market of the host Member State under conditions of normal and effective competition.56

Motor insurance is in the traditional mould: additional costs for foreign providers are the factual disadvantages that qualify an indistinctly applicable measure as a restriction. Such additional costs are the expenses necessitated by the adaptation to the regulations of the host Member State of the way an insurer does business in its home Member State or in any other Member State. By contrast Lawyers’ fees says, in essence, that a provider must take the market in another Member State as she finds it. The established providers at some point had to set up, or bring, their way of doing business in line with the applicable rules; a newcomer from another Member State cannot complain if she now has to do likewise. The exception that the Court concedes in Lawyers’ fees (viz that a restriction exists if lawyers are deprived of the opportunity of gaining access to the market of the host Member State under conditions of normal and effective competition)57

55 Third-party liability motor insurance, above n 53, paras 68–70, 64; also in Lawyers’ fees, above n 54, para 46. 56 Lawyers’ fees, above n 54, paras 49–51. 57 ibid, para 51.

42  Stefan Enchelmaier is unclear. If a foreign lawyer finds herself excluded from the Italian market altogether, she is certainly ‘deprived of the opportunity of gaining market access’. It is not enough for the Court, however, that she have market access at all. Such access must also be ‘under conditions of normal and effective competition’. The Court is silent as to what is ‘normal competition’ if not competition in accordance with the rules applicable in a given market; then again, the legality of those rules under EU law is the very object of the enquiry. Also, by which yardstick is the exclusion of some would-be competitors ‘ineffective’ and conversely, is untrammelled market access ‘effective’? The Court seems to assume that it knows the answers already, and thus turns in a circle regarding both ‘normal’ and ‘effective’ competition (or rather, drives the hapless national court into such a circle).58 On closer inspection, the homely truism with which the Court opens its assessment in Lawyers’ fees (‘[r]ules of a Member State do not constitute a restriction within the meaning of the EC Treaty solely by virtue of the fact that other Member States apply less strict, or more commercially favourable, rules’) resembles a formula that the Court traditionally uses. That formula is deployed, however, not in order to establish whether there is a restriction. Instead, it inoculates the Member State’s choice of the level of protection that it wants to afford one of the public interests objectives listed in Articles 36, 45(3), 52, etc, or to a mandatory requirement, against the objection that the national rule is surely excessive (not ‘necessary’) because other Member States make do with less demanding standards.59 Instead, every national rule must be assessed on its own merits. Only if compliance with the laws of another Member State undoubtedly achieves the same level of protection will it be excessive for the first Member State to insist on compliance with its own rules. This protects the legislative autonomy of the Member States that we encountered earlier: not only are they free to set national priorities, they are also in principle free to pursue them to an extent of their own choosing. This protection of national legislative autonomy is located at the justification stage of the assessment. It is without prejudice to the question whether there is a restriction in the first place. In that respect, there can be no reservation for the Member States. They have all subscribed to the undertaking of creating an internal market, with the concomitant abolition of all unequal treatment without

58 Equally critical of the Court’s case law on this point is J Snell, ‘The notion of market access: A concept or a slogan?’ (2010) 47 Common Market Law Review 437, 468 f: ‘Ultimately, the notion of market conceals rather than clarifies. The very ambiguity of the term may explain its use by and usefulness for the Court … Market access may simply provide a sophisticated-sounding garb that conceals decisions based on intuition’. See also G Davies, ‘Understanding market access: exploring the economic rationality of different conceptions of free movement law’ (2010) 11 German Law Journal 469. 59 See, for instance, Ottica New Line, above n 13, para 44: ‘it is for the Member States to decide on the degree of protection which they wish to afford to public health and on the way in which that protection is to be achieved. Since the level may vary from one Member State to another, Member States should be allowed a margin of discretion’.

Development of the Free Movement Principles  43 objective justification (ie, discrimination). They have to accept the embarrassment that this will occasionally cause them,60 that is, of having to justify their rules, even those that are indistinctly applicable, if economic operators from other Member States find it more difficult to comply with these rules than do their domestic competitors. In both Motor insurance and in Lawyers’ fees there can be no question (at least nobody challenged as incorrect) that foreign insurers as well as foreign lawyers, respectively, had to shoulder two sets of ‘sunk costs’,61 compared with the single set imposed on those who keep themselves to their home market. ‘Normal and effective competition’ must, as a minimum, mean competition that is not burdened by the dead weight of costs that do not protect a public interest but merely bolster the position of the domestic competitors, by lumbering market entrants from other Member States with ultimately wasteful expenditure.

G. Conclusion In sum, by the 1980s the discrimination test was firmly established in the Court’s jurisprudence: a national restriction of a freedom of the internal market requires a justification on one of the grounds written in the Treaty if that restriction is distinctly applicable. Absent a ground of justification, or if the restrictions are disproportionate, there will be direct or overt discrimination. Indistinctly applicable restrictions require a justification only if they are in fact more difficult to comply with (or to benefit from) for goods, persons, services, or investments from other Member States than they are for their counterparts in the Member State that imposes the restrictions. Such national measures can be justified on the grounds listed in the Treaty, and additionally by the pursuit of ‘mandatory requirements’ or ‘imperative reasons relating to the public interest’. In either case, the restrictions must also be proportionate, otherwise there is indirect or covert discrimination.62

60 This, too, is an idea that was in the Treaty from the outset, namely in what is now Art 351, 3rd para, TFEU: ‘In applying [agreements concluded before 1 January 1958 or, for acceding States, before the date of their accession, between one or more Member States on the one hand, and one or more third countries on the other], Member States shall take into account the fact that the advantages accorded under the Treaties by each Member State form an integral part of the establishment of the Union and are thereby inseparably linked with the creation of common institutions, the conferring of powers upon them and the granting of the same advantages by all the other Member States’ (emphasis added). 61 Sunk costs are investments in material (tangible) or immaterial (intangible) assets that are lost on exit from the market to whose regulations the services were tailored, because these assets cannot be put to any other use after exit. 62 For this and for the remaining pair, ‘traditional’ and reverse discrimination, see the table in the Annex.

44  Stefan Enchelmaier

IV.  ‘An Area without Internal Frontiers’: A Second Line of Defence Emerges A.  Sunday Trading and the Limits of the Concept of ‘Discrimination’ Private litigants in the Member States’ courts and their legal advisers quickly grasped the potential of Cassis de Dijon. In the years after Cassis de Dijon, there was a veritable explosion of ‘Euro defences’ like the one first raised by the Dassonvilles, viz that a national provision that thwarted a trader’s business plans had to be left out of application because it was incompatible with Article 34. Many of those defences were valid, and did a lot to speed up the integration of product markets within the EU. Some, however, were rather fanciful. Of those, some could be dealt with easily. One Monsieur André Gauchard, for instance, was fined for extending without planning permission the premises of the French supermarket of which he was the manager. In the French court, he alleged that French town planning legislation imposing the requirement of a planning permission was contrary to EU law. Mr Gauchard had not, however, specified what EU law exactly was violated. The Court, following a reference by the French court, assessed the question under the provisions on the freedom of establishment only. The Court considered the matter a purely internal situation to which the Treaty did not apply. Thus, the issue of the free movement of goods did not arise.63 The variation of the ‘Euro defence’ associated with what became known as the ‘Sunday trading’ line of case law, however, proved more troublesome. Ultimately, it led the Court to revise its case law on Article 34. The reader is taken to be familiar with the fact patterns of these cases: the traders’ argument was invariably that the enforced shop closure on Sundays reduced imports because fewer goods across the board were sold on the remaining days. The Court approached the question in its B&Q judgment in a manner vaguely reminiscent of Cassis de Dijon. It asked, first, whether rules such as those in issue pursued an aim which was justified with regard to Community law. It found that rules governing the opening hours of retail premises reflected certain political and economic choices. Their purpose was to ensure that working and non-working hours were so arranged as to accord with (the mandatory requirement of) ‘national or regional socio-cultural characteristics’. In the state of Community law at the time, this was a matter for the Member States. What was more, the rules were not even meant to govern the patterns of trade between Member States. Second,

63 Case 20/87 Ministère public v Gauchard EU:C:1987:532, para 9. AG Vilaça dealt with the question in more detail under II. of his opinion. On conflicts between the freedom of establishment and the free movement of goods, see below.

Development of the Free Movement Principles  45 it was necessary to ascertain whether the effects of such national rules exceeded what was necessary to achieve the aim in view. This was for the national court to determine.64 This is a rather unorthodox way of applying the test of proportionality. Proportionality (‘necessity’ in the formulation of Cassis de Dijon) is, put simply, about the relationship between ends and means. More precisely, it is about whether the ends (a ground of justification as in Article 36, or a ‘mandatory requirement’) can justify the restrictions entailed by the means (national provisions serving this ground or mandatory requirement). This will not be the case if the means are incapable of achieving the ends (they fall short), or if the means exceed what is necessary to achieve the ends (they go too far). In the B&Q judgment, this test is transformed into the question, first, whether the end is justified. The only criterion for deciding seems to be whether or not this is a matter for the Member States. This, however, begs the question: if the matter is for the Member States alone, why should they have to justify themselves? What is more, we have already seen that the choice of ground of justification and the decision on the level of protection to be afforded to this public interest is always for the Member State – what, then, have we learned from the Court’s explanation in B&Q? The second step, according to the Court, is to ask whether the effects of the rules (means), rather than the rules themselves, remain within the limits of what is necessary to achieve the aim. The Court gives no reason why it does not want to scrutinise the rules, but instead focuses on the effects on the patterns of trade. Those effects, however, were not even intended to begin with, the Court had said earlier. Proportionality, by contrast, implies an intentional employment of the means for the ends. Moreover, the effects do not have to be intended, anyway: it is enough if they occur, whatever solemn assurance the Member State gives as to whatever worthy outcome it really wanted to achieve. The key to understanding why the Court got into this muddle can be found in the Cassis de Dijon judgment itself. There, the Court said that ‘obstacles to movement … resulting from disparities between national laws’ are acceptable if certain conditions are fulfilled. Conversely, if the obstacles do not result from such disparities, the principles as set out in the judgment do not apply. With regard to Cassis de Dijon, the obstacle to trade resulted from the disparities between the German and French rules. A product lawfully produced in France had to be adapted to the requirements stipulated by the German rules, or else it would not be marketable in that Member State. The situation is different in the Sunday trading cases. It does not matter whether, say, garden gnomes from Germany may not be sold there on the same day as in England and Wales, or not on another day, or whether they may be sold in France on any day: they may still not be sold in England and Wales on a Sunday. In other words, it is not the



64 Torfaen

Borough Council, above n 7, paras 13–16.

46  Stefan Enchelmaier disparities between the national rules that create an obstacle to trade, it is the very existence of the rules in England and Wales that does.65 The legislation in issue in B&Q applied to all products, domestic and imported, and had the same factual repercussions on all. In other words, there was no discrimination, direct or indirect. Instead, equals were treated equally, which is in keeping with the principle of equality.66 The necessity to justify such rules under the Treaty articles on the free movement of goods cannot, therefore, arise as it had in Cassis de Dijon where one rule applied to all products, but not all products, especially not those from other Member States, found it equally easy to comply with the rule. As already seen, the concept of discrimination encompasses both direct discrimination as a consequence of distinctly applicable national measures without a justification, and indirect discrimination resulting from unjustified indistinctly applicable restrictions in Member States. Lastly, a national rule may apply to all goods, domestic and from other Member States, and it may also be equally difficult for all to comply with that rule, as in Sunday trading. This may also be the case because, for instance, the national rule is based on an international standard, and stipulates that goods must not be marketed if they do not conform to the standard, or that they may only be marketed after conformity is established. The rule might also impose a blanket prohibition on sales of the product in question. Rules of this type impose universal bans. Most such bans are conditional or temporary, as in the first two examples; few are absolute, as in the third. In this situation, there will be no unequal treatment in law or in fact. The question of discrimination will not arise. Article 26(2) TFEU, however, defines the internal market as an area ‘without internal frontiers’. According to the Court, the Treaty, by establishing a ‘common market’ (which can in this respect be considered identical with the ‘internal market’), and by progressively approximating the economic policies of the Member States, seeks to unite national markets in a single market having the characteristics of a domestic market.67 Any economically active person in the Union can decide to make the market in any Member State their geographic area of activity. The common market is ‘common’ in the sense that every national market is legally a market for everyone, regardless of where in the EU they are based. The commonality of this market does not mean the disappearance of the national markets as economically or geographically separate, nor the cessation of all

65 This argument was developed in AG Tesauro’s opinion in Case C-292/92 Hünermund v Landesapothekerkammer Baden-Württemberg EU:C:1993:932 [1993] ECR I-6787, 6800–15. 66 See again the table in the Annex. 67 Case 207/83 Commission v UK (Origin markings) EU:C:1985:161, para 17. In an earlier judgment, the Court had formulated more cautiously that the aim of creating the common market was ‘to merge the national markets into a single market bringing about conditions as close as possible to those of a genuine internal market’ (emphasis added), Case 15/81 Gaston Schul Douane Expediteur v Inspecteur der Invoerrechten en Accijnzen, Roosendaal EU:C:1982:135, para 33.

Development of the Free Movement Principles  47 regulatory activity on the part of the Member States. It merely describes the disappearance of legal obstacles to access to these markets. Within such a market, the Court held, economic interpenetration will result from a division of labour between Member States.68 To put it differently, the project of creating the internal market is about guaranteeing economic operators established anywhere in a Member State the legal possibility of access to the markets everywhere in the Union. A Member State that imposes a universal ban thereby erects a frontier between itself and those Member States in which the product may (still) lawfully be marketed. Universal bans therefore require a justification. Because such bans ex hypothesi apply indistinctly, grounds of justification therefore can be found in Article 36 or beyond, as ‘mandatory requirements’. If, on the other hand, an indistinctly applicable rule makes marketing equally more difficult for all goods irrespective of their origin, and if the rule imposes no universal ban, it will not require a justification at all. The rule merely reduces turnover across the board, but otherwise leaves the market equally open for everyone. Examples from the Court’s case law are shop opening hours, the prohibition of retail at a loss, and zoning legislation.

B.  Keck and the Mystery of ‘(Rules Relating to) Certain Selling Arrangements’ The Court developed this view in Keck and Mithouard.69 This judgment, too, can safely be taken to be familiar to the reader, so that the central paragraphs 15–17 need not be reproduced here. It is worth remembering, however, which question the Court was addressing: [National legislation imposing a general prohibition on resale at a loss] may, admittedly, restrict the volume of sales, and hence the volume of sales of products from other Member States, in so far as it deprives traders of a method of sales promotion. But the question remains whether such a possibility is sufficient to characterize the legislation in question as a measure having equivalent effect to a quantitative restriction on imports’.70

In the following paragraphs 15 and 16, the Court seems to juxtapose ‘product requirements’ and ‘(rules relating to certain) selling arrangements’. The Court in paragraph 15 gives some examples (‘such as’) of ‘product requirements’. The judgment is, however, silent as to what makes a rule relate to ‘certain selling arrangements’. Maybe the Court wanted to elaborate further on this new concept as more cases came before it. This would not, however, explain why it did not even make a start in Keck.



68 Origin

markings, above n 67, para 17. above n 9. 70 ibid, para 13 (emphasis added). 69 Keck,

48  Stefan Enchelmaier Another explanation is that ‘selling arrangements’ is but a convenient label for any rule that affects in the same manner, in law and in fact, the marketing of domestic and imported products.71 In other words, the rule must not stipulate or entail unequal treatment between the two groups of products. From paragraph 16 of the judgment, it is not apparent that the phrase ‘selling arrangements’ should have any content beyond this.72 Correspondingly, the expression ‘product requirements’ in paragraph 15 of the judgment is shorthand for measures which, if there is no justification, amount to material (indirect, covert) discrimination against imports. Discrimination and non-discrimination are mutually exclusive, every measure belongs either to the one or the other. With regard to ‘selling arrangements’ and ‘product requirements’, it can be less clear under which one a given provision falls.73 Some commentators have noticed that the Court speaks of ‘certain selling arrangements’, but if a category of ‘certain other selling arrangements’ exists,74 it remains entirely nondescript, so that the word ‘certain’ seems ultimately redundant – never mind that the distinction leaves us none the wiser as to what ‘(rules relating to certain) selling arrangements’ are in the first place. A simpler explanation would be that ‘selling arrangements’ are not part of the test at all. The test is contained in what follows each of the two ‘so long as’. It is the same test, familiar since Cassis de Dijon, as for ‘product requirements’ (paragraph 15). The differences between the two situations underlying the judgments are merely factual, not differences of principle (and concomitant tests): retail at a loss did have the same factual repercussions for all coffees, whereas the minimum alcohol content did not for all fruit liqueurs. The question can also be phrased the other way around: what are the questions (what is the test), according to paragraph 15, once we have qualified a national rule as one imposing ‘product requirements’? We ask whether it applies to all (‘rules [that] apply without distinction to all products’) and if so, whether all find it equally easy to comply with the rule (otherwise, ‘obstacles’ arise as a ‘consequence’ of these rules). In Cassis, the answer to the first question was, yes; to the second, no. 71 Similarly AG Fennelly in his opinion in Case C-190/98 Volker Graf v Filzmoser Maschinenbau EU:C:2000:49, para 19: ‘The formal distinction between product rules and selling arrangements is less important than the motivation which led to its adoption, which was to identify the circumstances in which different types of rules have the same undesired effect, that is, to affect access to the market’. 72 This is confirmed by the way the concept is introduced in later decisions. See, eg, Case C-254/98 Schutzverband gegen unlauteren Wettbewerb v TK-Heimdienst Sass EU:C:2000:12, paras 24–26, where the Court first states that the rule in question is a selling arrangement, and then proceeds to find that it affects the marketing of domestic products and imports differently. The prior categorisation is, therefore, not decisive. 73 See, eg, Case C-368/95 Familiapress v Heinrich Bauer Verlag EU:C:1997:325, paras 11, 12. The case concerned an Austrian ban on prize quizzes in magazines. Such competitions are a sales technique, their regulation can be said to concern ‘selling arrangements’. Nevertheless, the ban meant that publishers in Germany, where such competitions were legal and customary, had to change the content of their magazines, which is presumably characteristic of a ‘product requirement’. 74 C Barnard, The Substantive Law of the EU: The Four Freedoms, 5th edn (Oxford, Oxford University Press, 2016) 127.

Development of the Free Movement Principles  49 What do we ask in case we think we find ourselves in the presence of a rule relating to certain ‘selling arrangements’? We ask whether it applies to all (‘relevant traders’)75 and whether they all find it equally easy to comply with it (‘affect in the same manner … in fact’ – ‘in law’ is redundant as that was already the first question). In Keck, the answer to both questions was, yes. Both sets of questions are identical, and it is these questions (plus the one discussed below, concerning universal bans) that lead to the solution in Cassis as well as in Keck, and in all subsequent cases. The difference between ‘product requirements’ and ‘(rules relating to certain) selling arrangements’ is as nebulous as it is irrelevant. Only in passing, and because a number of commentators have argued otherwise,76 we note that there is not a word here of any presumption that, say, ‘product requirements’ do, and ‘(rules relating to certain) selling arrangements’ do not require a justification. This would be misconceived, anyway: the need for a justification comes from the inequality in law or in fact, not from what sort of rule caused it. Article 34 captures ‘[measures having] equivalent effect’, not ‘equivalent cause’ or whatever else. What is more, presumptions are alien to the procedure under Article 258 TFEU. The burden of proof for the presence of a restriction is on the Commission; for a justification for the restriction, it is on the Member State. It is likewise in any other litigation that involves Article 34, no matter whether ‘product requirement’ or ‘(rules relating to certain) selling arrangements’ or any other description of rules (see below on restriction of use) are in issue. Nor does it make a difference whether the question arises in proceedings in the Member States’ courts. The same distribution of the burden of proof is the starting point in national law: each party has to prove the facts that bolster their legal argument, whether they contend there is an obstacle to trade, or that the obstacle is justified. This distribution is the same for all comparable actions, involving purely national sales as much as imports and exports coming from or going to other Member States. The principle of equivalence is therefore observed. Nor do the rules on who has to prove what render the exercise of the right granted to individuals by Article 34 excessively difficult or downright impossible. The principle of effectiveness is hence also satisfied.77 As a consequence, there is no need to introduce any presumptions into national law, either.

75 For an example, see Case C-198/14 Visnapuu EU:C:2015:751, paras 106, 107: ‘the retail sale licence … is available only to manufacturers of alcoholic beverages established in Finland, thus excluding manufacturers established in other Member States. Consequently, the requirement to hold a retail sale licence in order to import alcoholic beverages with a view to their retail sale to Finnish consumers, at issue in the main proceedings, does not satisfy the first condition laid down in Keck and Mithouard’. 76 Nic Shuibhne, above n 40, 215, last para; Barnard, The Substantive Law of the EU, above n 74, 130, first and last paras; Barnard, ‘Fitting the remaining pieces into the goods and persons jigsaw?’, above n 40, 45; P Craig and G de Búrca, EU Law. Text, Cases, and Materials, 6th edn (Oxford, Oxford University Press, 2015) 690 sub (i). 77 As to these principles, see Case 33/76 Rewe v Landwirtschaftskammer für das Saarland EU:C:1976:188; Case 45/76 Comet BV v Produktschap voor Siergewassen EU:C:1976:191.

50  Stefan Enchelmaier The above two questions (does the rule apply to all regardless of origin or nationality? If so, do all find it equally easy to comply with the rule?) are simple tests for discrimination, direct (if the answer to the first question is, no) or indirect (if it is ‘yes’ to the first, ‘no’ to the second). We are, however, still looking for the answer to the Court’s initial question in Keck, posed in paragraph 13, whether a mere reduction in turnover, free from discrimination of any description (if the answer to both questions is, yes), is also caught by Article 34. This is where paragraph 17 of Keck comes in. We can leave to one side that the Court betrays some insecurity with the redundant, ‘impede access any more than [the rules impede] the access of domestic products’. Such inequality in law or in fact should already have been captured by a negative answer to one of the preceding questions. In that case, we need never have progressed beyond the discrimination test. One may also wonder what reason the Court might have had to make an apparently factual statement about the nature of such rules, especially given that the Court saw no need (or was unable) to give even an approximate description of what these ‘(rules relating to certain) selling arrangements’ actually were. Instead, paragraph 17 does not contain a factual statement, but a (normative) condition (a question, a test) additional to the two previous conditions. The condition is that the rules must not impose a universal ban. Such a ban is the only non-discriminatory limitation of imports that does not merely reduce their volume, and does so to the same extent as it reduces the turnover of domestic products. Such a ban erects an ‘internal frontier’, contrary to Article 26(2). Any neutral restriction short of a ban, any rule that leaves the market diminished but does not treat the remaining imports unequally in law or in fact, is permissible. This is the answer to the Court’s question in paragraph 13.78 Keck thus complements the discrimination test established in the earlier case law.79 Together they form a comprehensive system for dealing with the paradigms that can arise under Article 34 and beyond, in all free movement cases.80

78 The Court’s recent jurisprudence is anything but straightforward on this point. In Case C-456/10 ANETT EU:C:2012:241, para 20, the Court was asked, ‘whether Art 34 TFEU must be interpreted as precluding national legislation … which imposes prohibitions on tobacco retailers from importing tobacco products from other Member States’. It took the Court nine paragraphs to describe effects of this prohibition on imports, culminating in paras 42–43, where it explained: ‘All of these elements are capable of having a negative effect on the choice of products that the tobacco retailers include in their range of products and, ultimately, on the access of various products coming from other Member States to the Spanish market. In those circumstances, it must be held that the prohibition at issue hinders the access of these products to the market’. The prohibition was distinctly applicable, and for that reason if for no other required a justification. Inexplicably, the Court had found in passing at para 36 that, ‘nothing indicates that the national legislation at issue has the object or effect of treating tobacco coming from other Member States less favourably’, when the unequal treatment in law of imports was staring it in the face. 79 This is summarised and illustrated in the table in the Annex. 80 K Purnhagen, ‘Keck Is Dead, Long Live Keck? How the Court of Justice tries to avoid a Sunday trading saga 2.0’ in F Amtenbrink, G Davies, D Kochenov and J Lindeboom (eds), The Internal Market and the Future of European Integration, Essays in Honour of Laurence W Gormley (Cambridge, Cambridge University Press, 2019) 181, also identifies a general test applicable to all freedoms, derived from

Development of the Free Movement Principles  51

V.  The Convergence of the Freedoms A.  Glass Beads or Ball Bearings? In Search of the Limits to Keck ‘Convergence’ means the (affirmative answer to) the question whether there are principles common to all the freedoms of the internal market. More specifically, the question is whether there is a uniform methodology for assessing cases that arise under any of the freedoms. On the few occasions when the Court has reflected on the question, it seems to have assumed that there is. As the Court held in Gebhard, [i]t follows … from the Court’s case-law that national measures liable to hinder or make less attractive the exercise of fundamental freedoms guaranteed by the Treaty must fulfil four conditions: they must be applied in a non-discriminatory manner; they must be justified by imperative requirements in the general interest; they must be suitable for securing the attainment of the objective which they pursue; and they must not go beyond what is necessary in order to attain it.81

Similarly, the Court was adamant that across the board, no de minimis threshold applies to the freedoms: ‘according to the Court’s case-law, the articles of the Treaty relating to the free movement of goods, persons, services and capital are fundamental Community provisions and any restriction, even minor, of that freedom is prohibited’.82 In the same vein, the test developed above from the judgment in Keck – (prohibition of) discrimination and (of) universal bans – can be applied in the assessment of cases under any and all of the freedoms. Nevertheless, the matter is by no means clear-cut. The above quote from Gebhard addresses the question of the justification of restrictions in some detail. On the logically preceding question whether there is a restriction of a freedom, it only offers the formula ‘measures liable to hinder or make less attractive the exercise [of a freedom]’. That formula is gleaned from the case law on the free movement of persons and services. It does not occur in the jurisprudence on the free movement of goods. Likewise, the second quote does not attend to the question when there is a restriction. It merely holds that at any rate, it does not matter

Commission v Italy (Moped trailers), above n 11, paras 34–37 and ANETT, above n 78, paras 34–35. This test, however, is said to rest on three principles: non-discrimination, Cassis de Dijon and market access. It is unclear what Cassis de Dijon stands for other than a test for indirect discrimination (unjustified unequal treatment as a consequence of a rule that applies to everyone equally in law, but with which not everyone finds it equally easy to comply). Nevertheless, I do concur with Purnhagen (if for different reasons, explained above), that distinguishing between ‘product requirements’ and (‘rules relating to) selling arrangements’ is otiose. 81 Case C-55/94 Reinhard Gebhard v Consiglio dell’Ordine degli Avvocati e Procuratori di Milano EU:C:1995:411, para 37, emphasis added. 82 Case C-212/06 Gouvernement de la Communauté française and Gouvernement wallon EU:C:2008:178, para 52, emphasis added.

52  Stefan Enchelmaier how big or small a given restriction is. The panel discussion at the College following the first round of presentations at the ‘Internal Market 2.0’ conference (College of Europe, February 2019) confirmed that there is a whole spectrum of opinions on the question. Eric White, author of a classic article paving the way towards Keck,83 eloquently lampooned those who were in search of a ‘grand theory of everything, one single formula across the freedoms’. Against such an endeavour, he pointed to the different wordings of the provisions. He emphasised that the freedoms address different problems, provide for different exceptions, and have spawned different secondary law for their harmonisation. He finally highlighted the development of individuals’ status from that of workers to that of citizens. The latter point was succinctly put by another prominent scholar: ‘people and products are simply different’.84 The last observation is incontrovertible. Less clear, however, is how it translates into legal terms: does it refer to the scope of the freedoms, to restrictions imposed on them, or to the justification for any restrictions? Goods have no right to take up employment, services cannot be archaeological treasures, companies do not carry infectious diseases, workers cannot be subject to prudential supervision. The various freedoms each raise their own problems, which call for tailored solutions. They all have their Eigengesetzlichkeit, they are, to an extent, to be approached separately from one another. Nevertheless, under section II above we had seen that there are as many commonalities as there are differences, especially with regard to the question what amounts to a restriction of the freedoms. This should come as no surprise: Article 26(2) TFEU lists all freedoms equally as means to the one end of creating the internal market. On that focal point, they all converge. The textual analysis of the subsequent provisions of the Treaty reveals that this is no mere metaphor. Eric White articulates the suspicion of the common lawyer against daring but bloodless abstractions. His misgivings have a long tradition. ‘Every judgment’, we are told by an eminent judge at the turn of the last century, must be read as applicable to the particular facts proved … since the generality of the expressions which may be found there are not intended to be expositions of the whole law, but governed and qualified by the particular facts of the case in which such expression are to be found. [… A] case is only authority for what it actually decides. I entirely

83 Above, n 8. 84 Nic Shuibhne, above n 40, 35. Similarly, P Oliver and W Roth, ‘The Internal Market and the Four Freedoms’ (2004) 41 Common Market Law Review 407, 439: ‘in so far as natural persons are concerned, the Community has moved on from treating them merely as units of production or other economic units to considering them as human beings’. One may wonder when exactly the Community did ever not so consider natural persons, but that it now does is undeniable. Earlier (at 408), the same authors had argued that, ‘the citizen’s right to purchase the washing-powder of his choice and the corresponding right of another individual to purvey it, appear trivial in comparison with human rights such as are bound up with the free movement of workers; but it must not be overlooked that the free movement of goods can be linked to the freedom of expression, for example where restriction are placed on the importation or exportation of books, magazines or video cassettes’.

Development of the Free Movement Principles  53 deny that it can be quoted for a proposition that may seem to follow logically from it. Such a mode of reasoning assumes that the law is necessarily a logical code, whereas every lawyer must acknowledge that the law is not always logical at all.85

There is much good sense in the judges’ proceeding cautiously, not articulating far-reaching principles, lest they have to row back later. One may wonder, however, how far this reticence can be driven. Would we go so far as to say that Cassis de Dijon applies only to blackcurrant fruit liqueur, and that we may not extend it to products which the Court had not in later judgments expressly subjected to whatever we think is Cassis’ underlying principle? In the same spirit, one might argue that as there are so many different ways of killing a person, there is simply no point in even trying to arrive at a single, uniform definition of ‘murder’.86 Instead, we may console ourselves that, the reality of nuance does not rule out the light that comes from generalisation. Everything, to be sure, is more complex than appears. Every person, every event is unique. Even so, some effort must be made to simplify, to find patterns. Otherwise, we have nothing but a grab bag of unrelated data.87

To put it differently, ‘concepts without perceptions are empty, perceptions without concepts are blind’.88 We must not waste our time playing with glass beads, but equally we must not lose sight of the wood for all the trees.

B.  Keck and the Freedom to Provide Services The judgment in Deliège offers an example of convergence as it plays out under the freedom to provide services, Article 56 TFEU. The case concerned the rules of the Belgian judo federation, under which the professional judoka, Ms Deliège, had failed to qualify for the coming world cup in Paris. She argued that the rules curtailed her freedom to provide services. The Court, however, found that the

85 Quinn v Leathem [1901] 1 AC 495 (HL) 506 (Lord Halsbury LC). A more recent echo of this can be found in Lord Salmon’s judgment in DPP v Majewski [1977] AC 443, 482: ‘[I]n strict logic this view cannot be justified. But this is the view that has been adopted by the common law of England, which is founded on common sense and experience rather than strict logic’. 86 White, above n 8, 253, 254 discusses the question ‘whether the laws against murder and violence and the carrying of offensive weapons are compatible with the Treaty only to the extent that they are justified under Art 34’, because ‘these laws undeniably restrict the demand for, and therefore the sales of and also ultimately imports of guns and knives’. Would a different definition apply in case of the services offered by a hit-man from another Member State? 87 D Landes, The Wealth and Poverty of Nations (London, 1998) 416. 88 I Kant, Kritik der reinen Vernunft, I Heidemann (ed) (Stuttgart, Phillip Reclam Jun, 1966) 120 (75 of the 2nd edn Riga 1787 – my translation). Kant continues, ‘[d]aher ist es ebenso notwendig, seine Begriffe sinnlich zu machen (d.i. ihnen den Gegenstand in der Anschauung beizufügen), als seine Anschauungen sich verständlich zu machen (d.i. sie unter Begriffe zu bringen)’: it is equally necessary to make one’s concepts comprehensible (ie, to add to them perceptions of specific examples) as it is to make one’s perceptions intelligible (that is, to bring them under concepts), ibid.

54  Stefan Enchelmaier selection rules did not determine the conditions governing access to the labour market for professional sportsmen and -women, and did not contain nationality clauses limiting the number of nationals from other Member States who may participate in a competition. Also, although such selection rules inevitably had the effect of limiting the number of participants in a tournament, this limitation was inherent in the conduct of an international high-level sports event, which necessitated certain selection rules or criteria. Such rules were not therefore, in themselves, to be regarded as constituting a restriction on the freedom to provide services.89 In this, the three Keck criteria are recognisable: the rules contained no nationality clause, ie, they were indistinctly applicable. The technique of judo, applied by all fighters, did not require (or even permit) any adaptation by athletes coming from different Member States. The rules thus had the same factual impact on all fighters. Lastly, there was no bar to access to the labour market. A select few did qualify and were sent to fight at the tournament. There was no bar to market access for Ms Deliège either. If she failed to qualify for one professional tournament, she could still participate in others for which she made the grade: someone always did. Such rules of qualification are moreover necessary to spare the audience the spectacle of professional fighters wasting their time on dilettantes. The mere limitation of the number of fighters to the very best at any given moment did not, therefore, constitute a restriction under Article 56, and hence needed no justification.

C.  Keck and the Free Movement of Workers Within the scope of Article 45 TFEU, Lehtonen provides an illustration of the Keck tests in action. In issue were rules of the Belgian professional basketball association that allowed transfers of players between Belgian clubs only up to a certain number of weeks before the start of the season. For players joining a Belgian team from a club in another Member State, the transfer window was more generous, ie, they could join until fewer weeks before the season began. Mr Lehtonen was a player from Finland who was fielded by his new Belgian club after the deadline applicable to this transfer had passed. This provoked sanctions by the basketball league authorities. The transfer rules, Mr Lehtonen and his club argued, were in breach of Article 45. The Court found that there was no direct or indirect discrimination, as stricter transfer deadlines applied to players (presumably, in the majority Belgian nationals) transferring between Belgian clubs. The rules imposed, however, a bar to market access, if only for the current season, on Lehtonen as a player coming from another Member State. The transfer rules therefore required a justification.90



89 Joined 90 Case

Cases C-51/96 and C-191/97 Deliège EU:C:2000:199, paras 61, 64. C-176/96 Lehtonen EU:C:2000:201, paras 48–51.

Development of the Free Movement Principles  55

D.  Bosman and Alpine Investments: Keck Abjured? Deliège and Lehtonen demonstrate that the same test can yield different results, depending on the facts to which it comes to apply. In that respect, Lehtonen is the parallel case to Cassis (justification required, albeit not for an uneven factual impact but because of the presence of a – temporary, only for the current season – bar to market access), whereas Deliège is the counterpart to Keck (no justification required in the absence of any legal or factual inequality, and of a bar to market access). By the same token, the finding of a restriction and the consequent need for a justification do not indicate that the test has not been applied, even less that is not applicable in the first place: such a finding is but one possible outcome of its application. It is entirely to be expected that some cases, sooner or later, will yield one result or the other, always depending on the specifics of their factual backdrop. To put it in a different way: the rules of adding and subtracting are always the same, but depending on the numbers that go into the calculation, these rules sometimes yield a positive sum, sometimes a negative. It is thus nothing unusual when the Court declined to follow the argument, raised in the proceedings in Bosman, that transfer rules of similar effect to the ones assessed in Lehtonen amounted to the equivalent to rules on ‘selling arrangements’, and were therefore exempt from scrutiny under Article 45.91 This does not mean that the discrimination-cum-universal-bans test is irrelevant for the free movement of workers. It merely means that because Mr Bosman was not allowed to take up employment (even though that ban was not unconditional or definitive), the transfer rules required a justification. The same pattern can be observed in the Golden shares judgments, as has been argued elsewhere: the Court did not there ‘refuse to apply Keck’ because there was a bar to acquiring certain shares altogether, as well as of percentages of other, ordinary shares.92 Again, the outcome must not be confused with the test, a negative outcome especially not with the test not being at work. Exactly this confusion of the result with the underlying test distorts the discussion of another case, Alpine Investments.93 In that judgment, the Court allegedly 91 Case C-415/93 Union royale belge des sociétés de football association and others v Bosman and others EU:C:1995:463, para 103: ‘It is sufficient to note that, although the rules in issue in the main proceedings apply also to transfers between clubs belonging to different national associations within the same Member State and are similar to those governing transfers between clubs belonging to the same national association, they still directly affect players’ access to the employment market in other Member States and are thus capable of impeding freedom of movement for workers. They cannot, thus, be deemed comparable to the rules on selling arrangements for goods which in Keck and Mithouard were held to fall outside the ambit of Art 34 of the Treaty’. 92 S Enchelmaier, ‘The Awkward Selling of a Good Idea, or a Traditionalist Interpretation of Keck’ (2003) 22 Yearbook of European Law 249, 318–319; S Enchelmaier, ‘The ECJ’s Recent Case Law on the Free Movement of Goods: Movement in All Sorts of Directions’ (2007) 26 Yearbook of European Law 115, 150–153. 93 Case C-384/93 Alpine Investments v Minister van Financiën EU:C:1995:126.

56  Stefan Enchelmaier barred the way to a uniform assessment under all freedoms by declining to transplant the ideas developed in Keck to the freedom to provide services. For a start, the prohibition of cold calling in issue in that case was not a prohibition of Alpine Investments’ services.94 The company was a provider of financial brokerage, not an advertising agency. Phone calls made in one’s own name to find customers are not (separately) remunerated (as required by Article 57). They would thus not constitute services, anyway. What is more, the Dutch company was not altogether excluded from the Belgian brokerage market. It merely had to ask potential clients whether they were happy to discuss the proposal on a separate occasion, rather than spring an offer on them on first contact. Alpine Investment’s Belgian competitors were not subject to the Dutch prohibition which, therefore, imposed a burden on the company that its competitors were spared. It was unusual that the restriction was imposed by the ‘exporting’ Member State. Nevertheless, just as providers, recipients and services can move in either direction between Member States, so each of the Member States involved needs to justify any restrictions that it imposes on the freedom. Because of the uneven impact of the Dutch rule on Dutch and Belgian companies active on the Belgian market, respectively, the situation was not as in Keck, but as in Cassis. The prohibition on cold calling made the provision of services from the Netherlands into Belgium more difficult than the provision of the same services within the Netherlands alone. On the Belgian market, Dutch providers faced competition from Belgian providers, who were allowed to cold call. On the Dutch market, by contrast, no one enjoyed the advantage over another competitor of being allowed to use this sales technique.95 The Court’s reasoning is not immediately obvious because the judgment flips between considering the two markets. The national court had expressly addressed the ‘export’ perspective, that is, the situation on the Belgian market: ‘providers established in the Member States where the potential recipients reside are not necessarily subject to the same prohibition or in any event not on the same terms’.96 The Court replied that the prohibition could constitute a restriction because it, ‘deprives the operators concerned of a rapid and direct technique for marketing

94 Contra, for instance, Nic Shuibhne, above n 40, 239, 240, 245: ‘non-discriminatory exit block’; 248: ‘prevention of access’. A similar situation obtained in Case C-500/06 Corporación Dermoestética v To Me Advertising Media EU:C:2008:421, where the prohibition was imposed not on the service (cosmetic surgery) as such, but on advertisements for it. In Italy advertisements for cosmetic surgery were not allowed on national television stations but were permissible on local or regional stations. 95 Moreover, the Dutch providers would not have been able to turn, under Art 56, against any advantages that their Belgian competitors might have enjoyed on the Dutch market (because, say, the Dutch authorities found it more difficult in fact to police the prohibition vis-à-vis providers established abroad). The Dutch providers would in this paradigm have found themselves in a purely internal situation, and thus unable to invoke this freedom of the internal market. 96 Alpine Investments, above n 93, para 26.

Development of the Free Movement Principles  57 and for contacting potential clients in other Member States’.97 In other words, there is an inequality of permissible sales techniques, to the disadvantage of the Dutch providers. Against this, the Dutch and British governments raised the ‘Keck objection’, to wit, that the prohibition amounted to ‘mere selling arrangements’.98 The Court conceded that, ‘neither [the prohibition’s] object nor its effect is to put the national market at an advantage over providers of services from other Member States’.99 This presumably refers to the Dutch (‘national’) market. Then, however, the Court reverts to the ‘export’ perspective. It goes on to refute the governments’ objection because the prohibition … imposed by the Member State in which the provider of services is established … affects not only offers made by him to addressees who are established in that State … but also offers made to potential recipients in another Member State.100

This reiterates that the Dutch prohibition puts Dutch providers at a disadvantage, compared with their local rivals on the market in the other Member State involved in the provision of the services in question. The rule applied to everyone who wanted to provide brokerage services into and out of the Netherlands. In fact, however, it made Dutch providers’ services more cumbersome on the Belgian market than it was for Belgian competitors to provide their services there. The disadvantage arises indirectly as the combined effect of the prohibition in the Netherlands, and the permissibility (the absence of a prohibition) in Belgium. The situation is thus as in Cassis, not as in Keck. Alpine Investment’s Belgian operations (its ‘access’ to that market) were ‘directly’ affected because the company had hitherto engaged in cold calling (rather than just contemplated it for the future). Again, the outcome in this particular case does not indicate that the test applied was different from that applied in another case, where it yielded a different outcome, let alone that there should be different tests. To devise one test applicable to all freedoms is worthwhile for another reason, too. The requirement to have an establishment on national territory as a precondition for offering insurance services in that Member State (as in German insurances)101 is a restriction of the positive freedom to provide services:102 Article  56 gives a right to provide services to any national of a Member State who is established in a Member State ‘other than’ that of the person for whom the services are intended. At the same time, forcing someone to establish themselves on national territory restricts that person’s negative freedom of establishment, ie,



97 ibid,

para 28. para 33. 99 ibid, para 35. 100 ibid, para 38. 101 Case 205/84 Commission v Germany (German insurances) EU:C:1986:463. 102 ibid, para 52. 98 ibid,

58  Stefan Enchelmaier their right not to establish themselves there (or not to choose a subsidiary over a branch or an agency, say, ie, not to choose a particular mode of establishment).103 If one freedom under one Treaty article is the flipside of another freedom under a different article, it would be paradoxical to have different tests as to whether there is a restriction and whether the Member State has discharged its burden of proving a justification for that restriction. A similar case can be made (and has been made elsewhere) for similar conflicts between other freedoms.104

E.  Moped Trailers: ‘Market Access’ as the New Universal Formula? Meanwhile, the Court seems to have embarked on a renewed quest for a uniform test, in the course of which it discovered ‘market access’ as the one formula for all freedoms. While the phrase is not new – it already occurs in Keck, paragraph 17 – the judgment in Moped trailers105 is commonly seen to elevate it to a central position.106 The judgment does not, however, make much effort to formulate a coherent explanation of this move. The central passage, paragraphs 33–36, consists of unconnected quotes from Dassonville, Cassis and Keck. These are drawn together by a mere ‘consequently’ at the beginning of paragraph 37, which then (again without explanation) introduces an unattributed excerpt from Groenveld,107 lifts part of the motives (namely, paragraph 35) straight into the new formula, and concludes with the entirely nondescript ‘[a]ny other measure which hinders access of products originating in other Member States to the market of a Member State’. This last part is then used to decide the case, but in a way that does not break the formula down into readily applicable definitions. The fundamental problem with this is that ‘market access’ is nothing but ‘free movement between Member States’. The creation of a market ‘without internal

103 Case 270/83 Commission v France (Avoir fiscal) EU:C:1986:37, para 22: ‘The fact that insurance companies whose registered office is situated in another Member State are at liberty to establish themselves by setting up a subsidiary in order to have the benefit of the tax credit cannot justify different treatment. The second sentence of the first paragraph of Art 49 expressly leaves traders free to choose the appropriate legal form in which to pursue their activities in another Member State, and that freedom of choice must not be limited by discriminatory tax provisions’. 104 Enchelmaier, ‘The ECJ’s Recent Case Law on the Free Movement of Goods’ above n 92, 147–49. 105 Commission v Italy (Moped trailers), above n 11. 106 See, for instance, Nic Shuibhne, above n 40, 236–242; Snell, ‘The notion of market access, above n 58, passim; Purnhagen, above n 80. 107 Case C-15/79 PB Groenveld BV v Produktschap voor Vee en Vlees EU:C:1979:253; the unattributed quote is taken from para 7: ‘[Art 35 TFEU] concerns national measures which have as their specific object or effect the restriction of patterns of exports and thereby the establishment of a difference in treatment between the domestic trade of a Member State and its export trade in such a way as to provide a particular advantage for national production or for the domestic market of the state in question at the expense of the production or of the trade of other Member States’ (emphasis added).

Development of the Free Movement Principles  59 frontiers’, and the abolition of ‘restrictions’ to the four freedoms are meant to bring about the legal conditions for market access. The relevant question is not whether Member States may restrict market access. It is trite to say that they must not. The question is, rather, what they must avoid lest they breach the prohibition, namely discrimination and (unjustified) universal bans. To make ‘access’ into a criterion will inevitably lead to finding any regulation a restriction. This, in turn, calls for some sort of limitation, and before we know it we are back with Sunday trading.

F.  Carpenter: Human Rights to Unify the Freedoms? Another emerging theme across several freedoms is the role of human rights in the assessment. The most general proposition in that respect is found in the judgment in Carpenter. The central passage is this: Mr Carpenter is exercising the right freely to provide services guaranteed by Article 56 TFEU … It is clear that the separation of Mr and Mrs Carpenter would be detrimental to their family life and, therefore, to the conditions under which Mr Carpenter exercises a fundamental freedom. That freedom could not be fully effective if Mr Carpenter were to be deterred from exercising it by obstacles raised in his country of origin to the entry and residence of his spouse. A Member State may invoke reasons of public interest to justify a national measure which is likely to obstruct the exercise of the freedom to provide services only if that measure is compatible with the fundamental rights whose observance the Court ensures.108

The problem with this analysis is that Mr Carpenter’s provision of services from a base in, say, London was equally made more difficult with regard to clients in Paris or Dublin as in Glasgow or Belfast. He had to replace the childcare that Mrs Carpenter had hitherto afforded him, no matter where he provided his services. In other words, his provision of services between the UK and other Member States was no more affected than within the UK alone. This would take the situation out of the scope of a provision that aims at protecting those crossing the borders in the course of an economic activity against disadvantages that those do not suffer whose activity is confined within one and the same Member State. Instead, the Court in the second sentence of the above quote argues, in so many words, that the restriction of Mr Carpenter’s right to family life equals a restriction of his freedom to provide services. In the third sentence, the Court confirms the traditional position that locates human rights at the justification stage of the assessment. This is where Article 36 EEC had placed the protection of human life. As a result, there is strictly no need in the Court’s scheme for the freedom to provide services: the right to family life is at the beginning and at the end of the assessment. Mr Carpenter’s provision of services is merely the factual backdrop against which the restriction of that right arises.

108 Case

C-60/00 Carpenter EU:C:2002:434, paras 37, 39, 40.

60  Stefan Enchelmaier In that respect, any other freedom would also do, or even none. In accordance with Article 20(2)(a) TFEU, citizens of the Union have the right to move and reside freely within the territory of the Member States. By virtue of national law, citizens have the right to reside within their own Member State anyway. One should think that the specific ‘value added’ by EU citizenship would be to grant citizens the right to move to another Member State and to reside there. Residence followed by movement from another Member State by nationals of another Member State would thus be protected: ‘first move, then reside’. In Ruiz Zambrano,109 however, the Court severed movement and residence, and grounded residence alone anywhere in the territory of the EU in citizenship of the Union. When the Treaty in Article 20(2)(a) grants Union citizens the right to reside ‘freely’, it would follow from Carpenter that this means that they are allowed to invoke all the rights that the ECHR and now also the Charter of Fundamental Rights of the EU grants them. Invoking one of the four freedoms of the internal market besides would be optional. Most recently, the Court has attributed (horizontal) applicability between private parties to the Charter.110 As a result, Union citizens’ legal position is now comprehensively protected by the Charter. Whatever one may think of this outcome and the way the Court reached it,111 it will simply mean that the Court will have to ‘reinvent the wheel’ under the Charter. The rights under the Charter may, according to Article 52(1), be limited in pursuance of ‘objectives of general interest recognised by the Union or the need to protect the rights and freedoms of others’. Then again, ‘[r]ights recognised by this Charter for which provision is made in the Treaties shall be exercised under the conditions and within the limits defined by those Treaties’, paragraph (2). Carpenter hence remains an aberration after all. Human rights are not a panacea for unifying the assessment under all freedoms. Lastly, it will be noticed that in the above excerpts from Carpenter, the Court uses ‘deterrence’ as a criterion for ascertaining a restriction of the freedom to provide services. Any encumbrance imposed by national law is apt to deter somebody or other from extending their provision of services into that Member State. As a consequence, any and all national legislation requires a justification, not just a rule that burdens those who engage in cross-border transactions more than those who provide their services exclusively in the Member State in which they are established. As a starting point, this would sweep away all national regulation of the markets in goods and services, and would selectively let back in those which

109 Case C-34/09 Ruiz Zambrano EU:C:2011:124. 110 Case C-414/16 Egenberger EU:C:2018:257; this has to be seen in conjunction with Case C-122/17 Smith v Meade EU:C:2018:631, in which the Court held that individuals could not inter se invoke the supremacy of directives over conflicting national law, with the exception of Directive 2015/1535. 111 For a critical discussion of the jurisprudence beginning with Case C-144/04 Mangold EU:C:2005:709, see S Enchelmaier, The Magical Mystery of Words: ‘Direct Effect’ and All That (forthcoming, 2020).

Development of the Free Movement Principles  61 Member States were able successfully to justify. This is again the Sunday trading trap.112 The discrimination-cum-universal-bans test from Keck deals with this in a straightforward way.

VI. Conclusion The development of the free movement principles over time shows an early concentration on the concept of ‘discrimination’. This was not surprising given the wording of the Treaty and of early secondary law on the free movement of workers. The principles first articulated in this area were in the course of the 1970s transferred to the freedom to provide services and to the free movement of goods. This culminated in the landmark judgment in Cassis de Dijon in 1979. The pattern of assessment devised in this period worked well throughout most of the 1980s, but ended in the confusion of the Sunday trading cases at the close of that decade. The Court in its judgment in Keck in 1993 saw clearly what had gone wrong, and offered all the elements of a rational and uniform test. The central insight was that the discrimination assessment had to be complemented by a prohibition of universal bans, but that any neutral (neither directly nor indirectly discriminatory) measures that merely reduced turnover across the board did not require a justification. This is gleaned from the description in Article 26(2) of the internal market as an ‘area without internal frontiers’. The problem of Keck was the insecure and somewhat infelicitous presentation of this new approach. The central passage contains two redundancies, and crucially, introduced the nebulous term ‘(rules relating to certain) selling arrangements’. Neither then, nor subsequently did the Court even attempt a definition. This led to lively discussions among commentators. Right from the start, the confusion was avoidable because a closer look at the text could have revealed that the purportedly central classifications of ‘product requirements’ versus ‘selling arrangements’ did not actually determine the questions the Court asked when assessing cases under Article 34; they merely marked two different outcomes under the same questions. These questions were entirely sufficient to structure a rational assessment under other freedoms as well, as exemplified by Deliège and Lehtonen from, respectively, the free movement of workers and the freedom to provide services. More unnecessary confusion came from the fallacy, prompted by passages in Alpine Investments, that a negative outcome (that the rules under scrutiny in that case were not tantamount to rules relating to certain ‘selling arrangements’) indicated that the test articulated in Keck could not be applied in the first place. 112 This trap was summarised by Tizzano AG in Case C-442/02 Caixa-Bank France v Ministre de l’Economie, des Finances et de l’Industrie EU:C:2004:586, para 63, to the effect that the Treaty aims at an internal market in which conditions are similar to those of a single market and where operators can move freely, not at a market without any rules at all – or rather, one in which rules are prohibited as a matter of principle, except if specifically justified.

62  Stefan Enchelmaier While Alpine Investments hinged on the uneven factual impact of the rules in issue, those in Bosman imposed a universal ban and were for that reason not excused a justification. Nevertheless, enough uncertainty seemed to surround these questions for the Court to continue looking for a general formula. It tried at least two: the first was the elevation of ‘market access’ to be the decisive criterion, culminating in the 2009 judgment in Moped trailers. This, however, is little more than wordplay: ‘market access’ is merely another way of saying, ‘free movement between Member States’. We already knew that that must not be impeded. The question is not whether, but how. Fortunately, the old tools – the discrimination test and the prohibition of universal bans (‘internal frontiers’) – are still in perfect working order. The second new approach, if it was meant to be one, in Carpenter was the translation of internal market questions into human rights questions. This is again no more than a reformulation. The conflicts and the mechanisms for their solution are ultimately the same, and the Charter expressly recognises the four freedoms’ traditional province. Nevertheless, the excitement about the potential for human rights to be deployed in economic relationships can only increase with the Court’s very recent attribution of horizontal effect to the Charter of Fundamental Rights of the EU in Egenberger. In all, therefore, the solid early developments of the legal principles governing free movement were followed by decades of experimentation, much of it productive, some less so.

Annex Schematic Overview Treatment of imports in law

Effect of the rule on domestic / imported goods

Justification

Dassonville

Different (distinctly applicable)

Different

Art 36

Cassis de Dijon

Same (indistinctly applicable)

Different (‘dual burden’, adaptation required: composition, labelling, promotion schemes)

Art 36 + ‘mandatory requirements’

Keck

Same (indistinctly applicable)

Same (no adaptation costs) + no (‘substantial’?) bar to market access (= universal ban)

Not required

Development of the Free Movement Principles  63

64

3 In Search of the Limits of Article 30 of the EEC Treaty Revisited ERIC WHITE

I. Introduction ‘In Search of the Limits of Article 30 of the EEC Treaty’, published in the Common Market Law Review in 1986, was a response to the expansive interpretation given by the Court of Justice to the concept of ‘quantitative restrictions and all measures having equivalent effect’ in what is now Article 34 of the Treaty on the Functioning of the European Union (TFEU). The Court had reformulated this concept in Dassonville to cover ‘all trading rules enacted by Member States that are capable of hindering, directly or indirectly, actually and potentially, intracommunity trade’. This had culminated in the Sunday Trading cases, where the Court held that restrictions on the opening hours of shops were measures of equivalent effect that needed to be justified in order to be consistent with the Treaty. After an extensive review of the case law, ‘In Search of the Limits’ concluded notably that: • The Dassonville formula could not be taken literally. Requiring all economic regulation to be justified and proportionate leads, as Miguel Poiares Maduro put it later in 1998 in his book We the Court, to a ‘process of Europeanisation of regulation in the Common Market through the judicial process’. • In relation to indistinctly applicable measures, what is now Article 34 TFEU only applies to barriers to trade that arise out of the application of rules relating to the characteristics of products (thus preventing the product from benefiting in the Member State of importation of the advantages arising from the different legal and economic environment prevailing in the Member State of exportation) and not therefore to rules that regulate the circumstances of sale in the Member State of importation. • The same limitation of the scope of Article 34 TFEU must also apply to non-discriminatory restrictions on the use of products (so that, to take one example, the Reinheitsgebot may be applied to prevent the use of cereals other than barley in the production of beer in Germany even though it cannot be

66  Eric White applied to prevent the sale of beers legally produced and marketed in other Member States). • Regulations on the sale and use of goods that effectively prevent all sale or use of a product do fall under Article 34 TFEU as they are more properly considered quantitative restrictions (the quantity being zero) rather than measures of equivalent effect. • Article 34 TFEU applies only to goods legally produced and marketed in a Member State and not to third-country goods in free circulation. • In this article, the author of ‘In Search of the Limits’ reviews his original conclusions in the light of some subsequent developments over the last 30 years.

II.  The Case Law Since ‘In Search of the Limits’ The following rather selective review of the cases is designed to illustrate the evolution on the case law over the period.

A.  Case C-145/88 Torfaen (Sunday Trading)1 The first case, or rather series of cases, that needs to be mentioned arose out of a concerted attempt by large UK retailers, especially do-it-yourself stores, to challenge the restrictions on Sunday trading that then applied in the UK. These retailers argued that the restriction on Sunday trading reduced their sales, including of products imported from other Member States and thus fell within the Dassonville formula of trading rules that are capable of hindering, directly or indirectly, actually and potentially, intra-Community trade. These cases were the inspiration for ‘In Search of the Limits’ and the European Commission defended in its observations to the Court a position along the lines of that set out in that article. The Court however, in a rather short judgment, stuck to its predominant case law of the time and held that the ‘obstacle to Community trade’ within the meaning of the Dassonville formula was not compatible with what is now Article 34 TFEU unless it ‘did not exceed what was necessary in order to ensure the attainment of the objective in view and unless that objective was justified with regard to Community law’.2 The Court went on to hold that the objective (‘to ensure that working and non-working hours are so arranged as to accord with national or regional sociocultural characteristics’) could be justified3 but left it to the national court to

1 Case

145/88 Torfaen Borough Council v B&Q plc EU:C:1989:593. 12 of the judgment. 3 Para 14 of the judgment. 2 Para

The State of the Case Law on Article 34 TFEU  67 decide4 whether the restrictive effect of the measure ‘exceeds the effects intrinsic to trade rules’,5 that is, to consider whether it was proportionate.

B.  Case C-292/92 Hünermund6 It was not until 1993 that the Court started to read the Dassonville formula in a less all-encompassing way. Hünermund concerned an advertising restriction imposed on pharmacists. The referring court assumed that this constituted a measure of equivalent effect and asked whether it was justified. Advocate General Tesauro made a careful review of the case law and the literature and examined the different effects of the various divergences in trading rules. He came to the conclusion that ‘measures, whose subject is the manner in which trading activity is carried on, are in principle to be regarded as falling outside the scope of Article 30’.7 This, he explained, was because such rules were ‘not liable to make access to the market less profitable for the operators concerned and thus, indirectly, to make access more difficult for the products in question’.8 He therefore called upon the Court to change its case law – ‘and, for this to be useful, to do so clearly and explicitly’.9 The Court agreed but did so by relying on its judgment rendered during the period between the publication of the Advocate General’s Opinion and the judgment in in Hünermund. It is to this key judgment which we now turn.

C.  Joined Cases C-267/91 and C-268/91 Keck and Mithouard10 It is in its famous judgment in Keck that the Court definitively broke with its previous case law. In doing so, it declined to follow the reasoning suggested in the two Advocate General opinions delivered in the case but adopted instead the reasoning that Advocate General Tesauro had developed a month earlier in Hünemund. Messrs Keck and Mithouard were being prosecuted for selling merchandise at a loss in contravention of French law. The referring court invoked a range of Treaty 4 Para 16 of the judgment. 5 Para 15 of the judgment. 6 Case 292/92 Ruth Hünermund and others v Landesapothekerkammer Baden-Württemberg EU:C:1993:932. 7 Para 25 of the opinion. 8 Para 25 of the opinion. 9 Para 26 of the opinion. 10 Joined Cases C-267/91 and C-268/91 Criminal proceedings against Bernard Keck and Daniel Mithouard EU:C:1993:905.

68  Eric White provisions but the Court quickly cut the range of issues to those relating to the free movement of goods. The Court recognised that rules such as those at issue could reduce imports11 but announced that [i]n view of the increasing tendency of traders to invoke Article 30 of the Treaty as a means of challenging any rules whose effect is to limit their commercial freedom even where such rules are not aimed at products from other Member States, the Court considers it necessary to re-examine and clarify its case-law on this matter.12

After reaffirming its judgment in Cassis de Dijon on barriers to trade arising out of differences in national legislation relating to product characteristics, the Court adds that contrary to what has previously been decided, the application to products from other Member States of national provisions restricting or prohibiting certain selling arrangements is not such as to hinder directly or indirectly, actually or potentially, trade between Member States within the meaning of [Dassonville], so long as those provisions apply to all relevant traders operating within the national territory and so long as they affect in the same manner, in law and in fact, the marketing of domestic products and of those from other Member States.13

The Court goes on to explain why this is so in the following terms: [T]he application of such rules to the sale of products from another Member State meeting the requirements laid down by that State is not by nature such as to prevent their access to the market or to impede access any more than it impedes the access of domestic products.14

The distinction between product characteristics and selling arrangements and the explanation that the latter do not restrict access to the market while the former do are exactly those suggested by Advocate General Tesauro in his opinion in Hünermund a month before and in In Search of the Limits several years earlier.

D.  Commission v Greece (Infant Formula Milk) In case C-391/92 Commission v Greece,15 the Court held that a requirement that infant formula milk could only be sold in pharmacies did not come within the scope of Article 34 TFEU because it was a selling arrangement of the kind described in Keck.16



11 Para

13 of the judgment. 14 of the judgment. 13 Para 16 of the judgment (emphasis added). 14 Para 17 of the judgment (emphasis added). 15 Case C-391/92 Commission v Greece EU:C:1995:199. 16 Para 13 of the judgment. 12 Para

The State of the Case Law on Article 34 TFEU  69 The Commission had tried to distinguish Keck by arguing that Greece did not manufacture infant formula milk.17 The Court held that the applicability of Article 30 of the Treaty to a national measure for the general regulation of commerce, which concerns all the products concerned without distinction according to their origin, cannot depend on such a purely fortuitous factual circumstance, which may, moreover, change with the passage of time. If it did, this would have the illogical consequence that the same legislation would fall under Article 30 in certain Member States but fall outside the scope of that provision in other Member States. The situation would be different only if it was apparent that the legislation at issue protected domestic products which were similar to processed milk for infants from other Member States or which were in competition with milk of that type.18

E.  Selected Subsequent ‘Unfair Competition’ Cases In De Agostini,19 the Court held that advertising was a selling arrangement within the meaning of the Keck principle but that the outright ban in Sweden of television advertising aimed at children under 12 could come within the scope of [Article 34 TFEU] ‘unless it is shown that the ban does not affect in the same way, in fact and in law, the marketing of national products and of products from other Member States’.20 The requirement of non-discrimination or neutrality is also present in Keck and one can wonder about the instruction to the national court to require the defender of the legality of the rule to prove this neutrality. It is by now well established that if the European Commission were to bring infringement proceedings against the Member State in question, it would be required to prove all elements of its case (as, for example, in the case of Commission v Greece discussed above).21 In Case C-254/98 TK Heimdienst,22 the Court held that the requirement for Austrian butchers, bakers and grocers wishing to sell their wares door-to-door to have a permanent establishment in the same or a neighbouring administrative district to be a measure of equivalent effect to a quantitative restriction despite recognising it as a selling arrangement. The rationale of the Court is that such a measure lacks neutrality23 since, even if an establishment in a neighbouring district of a neighbouring Member State might be sufficient, traders from elsewhere in the EU would not be able to provide their goods door-to-door in Austria24 and also



17 Para

7 of the judgment. 17 and 18 of the judgment. 19 Case C-34/95 De Agostini EU:C:1997:344. 20 Para 44 of the judgment. 21 See para 19 of the judgment in Commission v Greece. 22 Case C-254/98 TK Heimdienst EU:C:2000:12. 23 Para 25 of the judgment. 24 Para 28 of the judgment. 18 Paras

70  Eric White that the measure does not allow traders from other Member States to provide their goods door-to-door in central areas of Austria.25

F.  Use Restrictions The application of the Keck formula in the case law was considered by many to be giving rise to unpredictability.26 It was however cases concerning use restrictions that gave the opportunity for a clarification of the guiding principle. The first, that we will discuss is Mickelsson and Roos,27 a reference for a preliminary ruling concerning restrictions on where jet skis (referred to as ‘personal watercraft’ in the judgment) could be used on inland waterways in Sweden. It was argued in that case that these restrictions created an obstacle to the marketing of such products in Sweden. The case took an extraordinary length of time to decide, evidencing perhaps an intensive debate. Advocate General Kokott had delivered an opinion28 in December 2006 in which she observed that ‘arrangements for use and selling arrangements … are comparable in terms of the nature and the intensity of their effects on trade in goods’29 and ‘it therefore appears logical to extend the Court’s Keck case-law to arrangements for use and thus to exclude such arrangements from the scope of Article 28 EC’.30 The Advocate General was however conscious that restrictions on use could amount to a virtual general prohibition if use was restricted to such an extent that little use was possible. In such a case, access to the market would be prevented and Article 34 TFEU would apply.31 That was however a matter for the national judge to assess in light of the facts. It is not clear, however, whether the Court accepted the Advocate General’s ‘logical’ conclusion that an approach similar to that in Keck should apply. Although the Court refers to a hindering of access to the market,32 it goes on to approach the question from the point of view of justification holding that the measures would be compatible if regulations were adopted that allowed use of such craft ‘within a reasonable period’.33

25 Para 30 of the judgment. 26 Unpredictability is arguably an inevitable by-product of evolving case law. A large part of the criticism was more motivated by a desire to return to the pre-Keck days when all national trading rules needed to be justified under Art 34 TFEU. 27 Case C-142/05 Åklagaren v Percy Mickelsson and Joakim Roos EU:C:2009:336, judgment of 4 June 2009. 28 Opinion of 14 December 2006, EU:C:2006:782. 29 Para 52 of the opinion. 30 Para 55 of the opinion 31 Para 67 of the opinion. 32 Para 28 of the judgment. 33 Para 44 of the judgment.

The State of the Case Law on Article 34 TFEU  71 The Court paid more attention to the problem in its judgment in Italian Trailers,34 whose gestation overlapped that in Mickelsson and Roos. That case concerned the prohibition contained in the Italian Highway Code on motorcycles towing trailers. The Commission argued in an infringement proceeding that this constituted a measure of equivalent effect to a quantitative restriction since it created an obstacle to the marketing of such products in Italy. The case was originally assigned to a chamber but then, after a first Advocate General’s opinion, was referred to the Grand Chamber and all Member States were specifically requested to provide observations on whether the principles underlying the Keck judgment should also apply to use restrictions. This led to a second Advocate General’s opinion. Although this has not been widely noted, the Court effectively held in this case that non-discriminatory (in law and effect) use restrictions do not come within the scope of Article 34 TFEU, since it dismissed the action insofar as concerned trailers not specially designed for motorcycles.35 It limited the case therefore to trailers specially designed for motorcycles, for which the prohibition in the Italian Highway Code was tantamount to an absolute prohibition of the product. What Italian Trailers highlights is a change of emphasis from the classification of a measure (as relating to product characteristics, selling arrangements or use restrictions) to its effects (restricting access to the market).36 A consideration of whether particular measures restrict access to the market was present in Hünemund and Keck and also in ‘In Search of the Limits’ but served more as a means of arriving at the conclusion on selling arrangements than as a test in itself. Some have asked whether Italian Trailers means ‘Good-Bye Keck?’37 The answer is clearly ‘no’, not only because Italian Trailers did not concern selling arrangements but also because the Court continues to apply the test in Keck in its case law.

G.  Hallmarking (Commission v Czech Republic)38 A judgment relevant to the last of the conclusions from ‘In Search of the Limits’ noted above arose recently from an infringement proceeding concerning the Czech Republic. The Commission claimed in that proceeding that a refusal of the Czech assay office to recognise hallmarks on precious metals from affixed by a third-country branch of the assay office of a Member State was a measure of equivalent effect to a quantitative restriction.

34 Case C-110/05 Commission v Italy EU:C:2009:66. 35 Para 53 of the judgment. 36 See esp paras 56 and 58 of the judgment. 37 P Pecho, ‘Good-Bye Keck?: A Comment on the Remarkable Judgment in Commission v Italy, C-110/05’ (2009) 36 Legal Issues of European Integration 257. 38 Case C-525/14 Commission v Czech Republic EU:C:2016:714.

72  Eric White The Court took the opportunity to clarify the significance for third-country goods of the condition that a product must be lawfully manufactured and marketed in another Member State in order for the prohibition of measures of equivalent effect to quantitative restrictions between Member States to apply to it. It recalled that, in accordance with Article 28(2) TFEU, Article 34 TFEU applied to goods originating in third countries which are in free circulation in any of the Member States (that, is once all import formalities are complied with and all customs duties paid). In other words, they are assimilated to goods manufactured in the EU.39 However, marketing is a stage subsequent to manufacture and lawful import does not mean that a product can be placed on the market.40 It would appear to follow from this that the benefit of Article 34 does not automatically extend to third-country goods. It is only market access from one Member State to another that is protected. The conclusion would also seem to follow from the definition of the internal market in Article 26(2) TFEU as an area without internal frontiers. The actual outcome of the infringement proceeding was that the Czech rules infringed Article 34 TFEU but this was only as a result of the scope of the proceeding being limited to goods legally marketed in another Member State to the exclusion of those coming directly from third countries.41 It has been argued that the mutual recognition principle resulting from Article 34 TFEU cannot be reserved for the goods of Member States and must be applied to the goods of all other World Trade Organization (WTO) Member States since otherwise there would be a breach of the most favoured nation (MFN) principle in Article I of the General Agreement on Tariffs and Trade (GATT) by the Member States as well as various provisions of the WTO Agreement relating to technical regulations. This raises difficult issues that have not yet been explored in WTO jurisprudence.42 The usual defence is that the EU is a customs union and as such is covered by the exception for customs union and free trade agreements in Article XXIV GATT 1994. In Turkey–Textiles43 the WTO Appellate Body adopted a rather restrictive approach to Article XXIV and in particular held that it only provided a defence for measures that were ‘necessary’ for the creation of the customs union. The EU may find itself arguing that the limitation of Article 34 TFEU to trade ‘between Member States’ and the consequent limitation of its benefit to goods originating in other Member States and third-country goods legally imported and marketed in other Member States was ‘necessary’ for the formation of the EU. As noted in ‘In Search of the Limits’, Article XXIV:8(a) GATT 1994 only requires the removal

39 Para 37 of the judgment 40 Para 38 of the judgment. 41 Para 40 of the judgment. 42 L Bartels, ‘The Legality of the EC Mutual Recognition Clause under WTO Law’ (2005) 8 Journal of International Economic Law 691. 43 WTO case DS34, Turkey – Restrictions on Imports of Textile and Clothing Products.

The State of the Case Law on Article 34 TFEU  73 of restrictive regulation of commerce on substantially all the trade in products originating in the customs union. There is however an additional argument that the EU could deploy. The EU is a member of the WTO in its own right, as are its Member States. Although there is no declaration of competence, there needs to be a division of WTO rights and obligations between the EU and its Member States since otherwise this will give rise to overlapping and potentially conflicting rights and obligations. EU competence is such a variable concept capable of many variations that it can hardly serve this purpose. The better approach is, it is submitted, to consider the EU responsible at the WTO for all measures that arise from EU law and that the Member States are responsible for those measures that do not arise from EU law. Since the mutual recognition principle (and indeed more generally the free movement of goods resulting from Article 34 TFEU) is a creature of EU law, it is the EU and not its Member States that have the obligation to respect the MFN principle in this regard. Thus the EU has to accord the benefit of mutual recognition to all WTO Members if and to the extent that it (the EU) accords this treatment to another third country (and is not excused from doing so by Article XXIV GATT 1994).

III. Conclusion Over the last 30 years there have been many judgments and even more debate about the limits to Article 34 TFEU.44 Although all cases are fact specific and the case law relating to a vague yet fundamental treaty provision will inevitably have many conflicting strands, it is submitted that some more clarity has been achieved. First, Article 34 TFEU is not a provision that generally limits economic regulation by the Member States by requiring all such measures to have a justification recognised as valid in EU law and to be neutral and proportionate. Attempts to challenge non-protectionist measures that do not create an obstacle to the realisation of an internal market will fail. Second, the better test of whether a measure comes within the scope of Article 34 TFEU relates to the effects of the measure as limiting market access rather than the categorisation of the measure as relating to product characteristics, selling arrangements or even use restrictions. The categorisation of the measure as relating to product characteristics or to selling or use arrangements remains however a good guide as to whether it comes within the scope of Article 34 TFEU. Third, although this is likely to be even more contested than the previous points, the market access facilitated by Article 34 TFEU does not apply automatically and in the same way to third-country goods. They must first be not only legally imported but also legally placed on the market of a Member State before 44 No attempt is made in this short chapter to discuss all the case law and learned articles on the subject. It is truly vast.

74  Eric White they can benefit from the principles established for goods legally produced and marketed in a Member State. Only then does the market integration logic of the Treaty apply to require their assimilation to goods produced and marketed in the Member States for the purposes of Article 34 TFEU. However, as Judge Rosas put it in his article commenting on Italian Trailers ‘Life after Dassonville and Cassis: Evolution but No Revolution’: ‘[t]hose who have hoped that the ECJ has already said everything that can be said on [Article 34 TFEU] will be disappointed’.45

45 A Rosas, ‘Life after Dassonville and Cassis: Evolution but No Revolution’ in MP Maduro and L Azoulai (eds), The Past and Future of EU Law: The Classics of EU Law Revisited on the 50th Anniversary of the Rome Treaty (Oxford, Hart Publishing, 2010).

4 Internal Market Dynamics: On Moving Targets, Shifting Contextual Factors and the Untapped Potential of Article 3(3) TEU INGE GOVAERE

I. Introduction The internal market is one of the few EU concepts which is expressly defined in the Treaties. All is not, however, thereby said and done. Paradoxically, it will be argued, the internal market is one of the most prominent features of the European integration process but its specific internal and external dynamics are probably also among the least well understood. Article 26 (2) of the Treaty on the Functioning of the European Union (TFEU) stipulates unequivocally that: ‘The internal market shall comprise an area without internal frontiers in which the free movement of goods, persons, services and capital is ensured in accordance with the provisions of the Treaties’. This definition, inserted by the Single European Act, goes beyond the Rome Treaty concept of ‘common market’ which already covered the four freedoms without, however, initially going so far as to abolish the physical internal frontiers. The conceptual change from ‘common’ to ‘internal’ market serves as a perfect illustration that the four freedoms were always and inherently instrumental to respond to shifting contextual factors underlying the European integration process. When important lessons needed to be drawn from the oil crisis of the 1970s and the ensuing return to nationalism and protectionism,1 this resulted in a double restriction of national sovereignty in terms of veto right2 and exercising control over intra-EU borders.3 1 See the argumentation put forward by the Commission in ‘Completing The Internal Market’, White Paper from the Commission to the European Council (Milan, 28–29 June 1985) COM (1985) 0310_FIN. 2 By passing to qualified majority voting instead of unanimity voting in the Council for internal market measures, see the current Art 114 TFEU. 3 By the abolishment of physical border controls between the Member States.

76  Inge Govaere The ensuing ‘internal market’ Treaty definition as formulated in the mid1980s has, however, never been modified since.4 Nor have the detailed Treaty provisions relating to each of the four freedoms, and which are instrumental to interpreting Article 26 TFEU,5 been fundamentally redrafted over time, other than the Maastricht Treaty rewording of the free movement of capital.6 Of course it could be mentioned that the Maastricht Treaty additionally inserted provisions on EU citizenship.7 According to the Court of Justice of the European Union (CJEU) EU Citizenship is now the ‘fundamental status’ of persons, which allows also for non-economically active EU citizens and their family members to derive rights of movement and residence in other Member States.8 The Citizenship Directive 2004/389 provides the flanking legislative framework to exercise the rights to free movement of EU citizens and their family members, whereas the Services Directive 2006/123 does so for the exercise of certain services and establishment rights.10 Moreover the Amsterdam Treaty inserted the flanking measures 4 The Lisbon Treaty only removed the long overdue reference to the 1992 deadline for the establishment of the internal market. On the 1992 objective, see J Schwarze, I Govaere, F Helin and P Van den Bossche (eds), The 1992 Challenge at National level (Baden Baden, Nomos Verlagsgesellschaft, 1990). 5 The importance of having the internal market definition in the Treaties should also not be overestimated. In the Wijsenbeek case related to the free movement of persons prior to the Amsterdam Treaty, the CJEU held that the possibility to also effectively invoke the fundamental freedoms in order to oppose the application of national law is neither automatic nor unconditional, as it is expressly made conditional upon the existence of more detailed Treaty provisions, Case C-378/97 Wijsenbeek EU:C:1999:439. 6 The Maastricht Treaty expressly introduced the free movement of capital and payments in the relations between the Member States but also with respect to third countries, see currently Arts 63–66 TFEU. For the immediate impact thereof on the post-Maastricht case law, see L Flynn, ‘Coming of Age: The Free Movement of Capital Case Law 1993–2002’ (2002) 39 Common Market Law Review 773. 7 The Maastricht Treaty introduced the provisions on EU citizenship, currently Arts 20–21 TFEU. 8 Case C-184/99 Grzelczyk EU:C:2001:458. For follow-up case law and analysis, see for instance H Kroeze, ‘The Substance of Rights: New Pieces of the Ruiz Zambrano Puzzle’ (2019) 44 European Law Review 238; P Neuvonen, ‘EU citizenship and its “very specific” essence: Rendón Marin and CS’ (2017) 54 Common Market Law Review 1201; F Wollenschläger, ‘A New Fundamental Freedom beyond Market Integration: Union Citizenship and its Dynamics for Shifting the Economic Paradigm of European Integration’ (2011) 17 European Law Journal 1; D Kochenov, EU Citizenship and Federalism: The Role of Rights (Cambridge, Cambridge University Press, 2017); A Tryfonidou, The Impact of Union Citizenship on the EU’s Market Freedoms (London, Bloomsbury, 2016); N Nic Shuibhne, ‘Limits rising, duties ascending: The changing legal shape of Union citizenship’ (2015) 52 Common Market Law Review 889; N Nic Shuibhne, ‘Case C-434/09, Shirley McCarthy v Secretary of State for the Home Department, Judgment of the Court of Justice (Third Chamber) of 5 May 2011; Case C-256/11, Dereci and others v Bundesministerium für Inneres, Judgment of the Court of Justice (Grand Chamber) of 15 November 2011’ (2012) 49 Common Market Law Review 349. 9 Or EU Citizenship Directive, Directive 2004/38/EC of the European Parliament and of the Council of 29 April 2004 on the right of free movement of persons [2004] OJ L229/35 (Corrigenda [2005] OJ L197/34). 10 Directive 2006/123 of the European Parliament and of the Council of 12 December 2006 on services in the internal market [2006] OL L376/36. For the early discussions of this rather controversial directive, at least in its draft stage see, inter alia, S D’Acunto, ‘Directive services (2006/123/CE): radiographie juridique en dix points’ (2007) 2 Revue du droit de l’Union européenne 261; C Barnard, ‘Unravelling the services Directive’ (2008) 45 Common Market Law Review 323; G Davies, ‘The Services Directive: extending the country of origin principle and reforming public administration’ (2007) 32 European Law Review 232; I Govaere, ‘De Bolkestein-richtlijn in de contex van het acquis communautaire:

Internal Market Dynamics  77 long needed to establish the external borders for persons, albeit subject to opt outs.11 Yet all those measures are merely complementary instruments which do not fundamentally alter the essence of the internal market concept as initially conceived in the Treaties. It will be argued in this chapter that it would be wrong to infer from the above historical brushstroke that the very ‘internal market’ concept itself is therefore necessarily static, outdated or out of touch with reality. Attention will first briefly be drawn to the ‘external dynamics’ of the internal market, or the necessarily dynamic interpretation of the internal market concept to underpin new policy targets and trigger judge-made solutions in response to changing contextual factors. Then the focus will be on the ‘internal dynamics’ of the internal market, or the so far untapped potential of Article 3(3) of the Treaty on European Union (TEU) and its wider implications. Much has been said and written already about the still ‘missing’ reference to, for instance, the Digital Single Market in the Lisbon Treaty.12 To the contrary, little or no attention has been devoted in case law or doctrine to the single most visible change introduced by the Lisbon Treaty to the internal market in the form of the redefinition of its objectives.13 Whereas before the internal market was mainly reactive and instrumental to achieving peace and stability, its objective is now drafted in a much more purposeful and proactive manner. Still, the pre-Lisbon Tobacco Directive test14 continues to be applied unchallenged post-Lisbon. Considering the ‘purposive’ method of interpretation of EU law used by the CJEU,15 it is rather surprising that such an apparent and important contextual legal change, and its potentially important implications for other EU policies, have been largely ignored and underexplored.

evolutie of revolutie?’ [2006] Tijdschrift voor Sociaal Recht 3. For its implementation, M Mirschberger, U Stelkens and W Weiss, The Implementation of the EU Services Directive: Transposition, Problems and Strategies (The Hague, TMC Asser Press, 2015); M Klamert, ‘Of Empty Glasses and Double Burdens: Approaches to Regulating the Services market à propos the Implementation of the Services Directive’ (2010) 37 Legal Issues of Economic Integration 111. 11 D Hanf and R Munoz (eds), La libre circulation des personnes: États des lieux et perspectives (Brussels, PIE-Peter Lang, 2007); J Shaw, ‘Between Law and Political Truth? Member State Preferences, EU Free Movement Rules and National Immigration Law’ (2015) 17 Cambridge Yearbook of European Legal Studies 247. 12 See also Part III of this volume. 13 The potential implications of the Lisbon Treaty were extensively analysed in the various contributions to the book, I Govaere (ed), Moderne Interne Markt Voor De Praktijkjurist (Mechelen, Kluwer, 2012); see in particular for an early appraisal of the new internal market objectives post-Lisbon, I Govaere, ‘De Lissabon Internemarktdoelstelling en de ‘horizontale’ Burger– en Dienstenrichtlijnen: implicaties voor de Belgische Rechtsorde’ in I Govaere, Moderne Interne markt voor de praktijkjurist (Mechelen, Kluwer, 2012). 14 See below, III.A. 15 L Mackenzie-Stuart, The European Communities and the Rule of Law (London, Stevens, 1977) esp at 77, where he defines it as follows: ‘The use by the Court of the purposive approach – seeking out the object of the text in dispute and trying to give practical effect to it’. On the purposive method as used by the CJEU see also K Lenaerts, ‘Interpretation and the Court of Justice: A Basis for Comparative Reflection’ (2007) 41 International Lawyer 1011.

78  Inge Govaere

II.  External Dynamics of the Internal Market: On Moving Targets and Shifting Contextual Factors Shifting contextual factors have always and systematically led to a renewed questioning of the coverage and importance of the four freedoms, also in the absence of corresponding Treaty change. This is most visible in the types of cases and new issues put before the CJEU and also by the national courts.16 To a large extent contextual factors have also dictated the internal market priority policy targets of the European Union and have thus determined how the Treaty provisions are to be more proactively applied and further regulated.17 A dynamic internal market is thus largely dependent on judge-made appraisal and policy-oriented application of the more static Treaty framework. Such shifting contextual factors can be EU-specific or global, as is respectively illustrated below.

A.  EU-Specific Contextual Factors A typical example of an EU-specific contextual factor which led to judge-made solutions and policy response would be enlargement.18 By way of illustration, the important contextual change caused by the enlargement of the EU to include Spain, thus changing the territorial scope of application of the internal market rules, triggered the so-called ‘strawberries’ or ‘road blockades’ case.19 The CJEU for the first time had to clarify the scope of application of the rules on the free movement of goods to tackle restrictions caused by private actors, in casu the French farmers, where before only positive state measures were held to be targeted by the rules on the free movement of goods. A judge-made solution in furtherance of the internal market objective was found in a combined reading of Article 34 TFEU with Article 4(3) TEU, extending the scope of the application of Article 34 TFEU to also include the failure of a Member State to take appropriate measures to ensure the free movement of goods. This judge-made solution still stands today but it was necessarily refined in later case law to take into consideration the impact thereof on the respect for fundamental rights, in particular the right to demonstrate.20 16 See for instance below the Tom Kabinet case on the issue of digital exhaustion. 17 See for instance the Services Directive 2006/123, above n 10, which covers the four GATS modes of supplies, thus including services and establishment, in one EU secondary law instrument. 18 Of course enlargement is also a Treaty modification, but mainly affecting the territorial scope of application rather than the substance. 19 Case C-265/95 Commission v France (Strawberries) EU:C:1997:595. 20 Case C-112/00 Schmidberger EU:C:2003:333. For an analysis see for instance A Biondi, ‘Free Trade, A Mountain Road and the Right to Protest: European Economic Freedoms and Fundamental Individual Rights’ (2004) 1 European Human Rights Law Review 51; J Morijn, ‘Balancing Fundamental Rights and Common Market Freedoms in Union Law: Schmidberger and Omega in the Light of the European Constitution’ (2006) 12 European Law Journal 15.

Internal Market Dynamics  79 The immediate policy response to the Road Blockades case in the form of ‘Regulation 2679/98 on the functioning of the internal market in relation to the free movement of goods among the Member States’, however, proved to be much less performant in practice.21 It is not really the ‘rapid intervention mechanism’ it was initially meant to be, although it is still in force today and currently under revision.22

B.  Global Contextual Factors Over the past decades in particular two newly emerging global contextual factors have impacted greatly on the internal market. On the one hand, there was the globalisation of trade including all goods, services and intellectual property, culminating in the setting up of the World Trade Organization of which the EU and its Member States are founding members. On the other hand, the technological revolution triggered the information society which would fundamentally and durably alter the functioning of societies, individuals, businesses and markets. Both factors combined entailed that information society services rapidly became the most important trade commodities in need of an adequate and appropriate regulatory framework worldwide but also for the EU. However, the EU Treaty framework was not amended accordingly, and also not by the Lisbon Treaty, thus causing much pressure to rest on EU courts and EU policymakers to find an adequate response on the basis of the existing Treaty framework. This is most illustrative in relation to the digital sphere. It is striking to find that, whereas the creation of the ‘Digital Single Market’ has become one of the

21 Council Regulation (EC) No 2679/98 of 7 December 1998 on the functioning of the internal market in relation to the free movement of goods among the Member States [1998] OJ L337/8, the so-called ‘rapid intervention mechanism’, proved to be much less performant in practice, not least because the finally approved text agreed upon by the Council was a watered down version of the original Commission proposal. See, inter alia, A Mattera, ‘Un instrument d’intervention rapide pour sauvegarder l’unicité du marché intérieur: le règlement 2679/98’ (1999) 2 Revue du Marché Unique Européen 29. 22 On 11 October 2019 the Commission published an ‘Evaluation of Regulation (EC) 2679/98 on the functioning of the internal market in relation to the free movement of goods among the Member States’, see: op.europa.eu/en/publication-detail/-/publication/0efcaca3-ee2d-11e9-a32c-01aa75ed71a1/. At page 12 it states the recommendations ‘to ensure a more effective implementation of the Regulation: –

Better definition of “obstacles”, the timing of communications (“immediately” and “as soon as possible”) and “necessary and proportionate measures” to ensure the correct application of the Regulation. – Introduction of a monitoring mechanism to check Member States’ compliance with their obligations to both inform the Commission and take the necessary and proportionate steps to ensure the free movement of goods in the event of an obstacle, and thus ensure better enforcement of the Regulation. – Introduction of a unified digital solution with real time information accessible to businesses and national associations to speed up information exchange and reduce the asymmetry of information. – Establishment of a transparent, non-discriminatory and effective procedure to compensate economic operators for any loss or damage related to the obstacles’.

80  Inge Govaere priority targets of the Commission over the past decade,23 the Lisbon Treaty not even once mentions the word ‘digital’. Does this then imply that, as a matter of principle, the existing internal market rules and principles should apply, possibly with slight variations where necessary to accommodate the specificities of the digital world? Or should the Digital Single Market be singled out and be subject to a distinct set of rules, and thus be treated as some kind of new ‘fifth freedom’ albeit one not expressly mentioned in the Treaties? The answer as to whether such a new given requires an extension or instead a deepening of the internal market is not always clear and simple and may moreover vary over time. It is interesting in this respect to recall the Commission’s 2007 communication on ‘A Single Market for 21st Century Europe’, where the term ‘e-Internal Market’ was used and the proposal made to ‘promote free movement of knowledge and innovation as a ‘fifth freedom’ in the single market’.24 Whereas the 19 February 2020 Communication on ‘Shaping Europe’s digital future’ lists, among the key actions to be undertaken by the Von der Leyen Commission, ‘new and revised rules to deepen the Internal Market for Digital Services,25 without any reference to such a fifth freedom. The CJEU was also to some extent faced with such an underlying existential question in relation to the digital in the Tom Kabinet judgment of December 2019.26 In this much anticipated and long-awaited judgment, in particular after the former UsedSoft case which was specifically in relation to computer programmes,27 the CJEU firmly reject the application of the principle of digital exhaustion in relation to e-books. Interestingly, the CJEU pointed out that books and e-books ‘cannot be considered equivalent from an economic and functional point of view’.28 It follows 23 Commission Communication, ‘A Digital Single Market Strategy for Europe’ COM(2015) 192 final. On this issue see also ch 11 by Claire Bury and Irene Roche-Laguna, detailing the 2015 European Commission’s Digital Single Market (DSM) Strategy. 24 Commission Communication, ‘A single market for 21st century Europe’ COM(2007) 724 final, 8–9; the Commission pointed out: ‘The single market can be a platform to stimulate innovation in Europe. It encourages the spread of new technologies across the EU. It lends itself to networks – virtual and real – and fosters the development of a sophisticated logistics sector allowing for integrated management of the flows of goods, energy, information, services and people’. 25 COM(2020) 67 final, 12; the phrase continues as follows: ‘by increasing and harmonising the responsibilities of online platforms and information service providers and reinforce the oversight over platforms’ content policies in the EU’. 26 Case C-263/18 Tom Kabinet EU:C:2019:1111. See for instance L Oprysk, ‘“Digital” Exhaustion and the EU (Digital) Single Market’ in T-E Synodinou, P Jougleux, C Markou and T Prastitou (eds), EU Internet Law in the Digital Era: Regulation and Enforcement (Springer, International Publishing, 2020) 161–80. 27 Case C-128/11 UsedSoft EU:C:2012:407. On the speculations about the wider importance of this first case for digital exhaustion, see for instance E Linklater, ‘UsedSoft and the Big Bang Theory: Is the e-Exhaustion Meteor about to Strike?’ (2014) 5 Journal of Intellectual Property, Information Technology and E-Commerce Law 1. 28 In Case C-263/18, above n 26, para 85, the CJEU pointed to the following distinction between books and e-books, warranting also a distinction in terms exhaustion of rights: ‘The supply of a book on a material medium and the supply of an e-book cannot, however, be considered equivalent from an economic and functional point of view. As the Advocate General noted in point 89 of his Opinion, dematerialised digital copies, unlike books on a material medium, do not deteriorate with use, and used copies are therefore perfect substitutes for new copies. In addition, exchanging such copies

Internal Market Dynamics  81 from this that it cannot be taken for granted that the same internal market principles will always and necessarily apply to both the material and the digital alike. Yet whether and to what extent this really singles out the Digital Single Market as compared to the four traditional freedoms still remains to be further fine-tuned.

III.  Internal Dynamics of the Internal Market: The Untapped Potential of Article 3(3) TEU Post-Lisbon In a post-Lisbon setting, most attention has gone to the above-mentioned queries about the dynamic use of more static detailed Treaty provisions to respond to shifting contextual factors whilst shaping EU policy and practice. All too often one has remained oblivious, however, of the fact that the Lisbon Treaty does not remain static at all in so far as the internal market is concerned. On the contrary, the newly drafted Article 3(3) TEU has introduced great potential for modernisation of the very internal market concept through a revolutionary redefinition of its objectives. As will first be illustrated below, instead of the pre-Lisbon ‘reactive’ and ‘instrumental’ internal market objective, the Lisbon Treaty has quite radically opted for a much more outspoken ‘proactive’ and ‘purposeful’ definition of the internal market objectives. How does that potentially affect the pre-Lisbon Tobacco Directive test? The question then arises of how such a prominent primary law contextual change has been addressed in post-Lisbon internal market case law, in particular considering the ‘purposive’ (or teleological) method of interpretation of EU law traditionally used by the CJEU.29 In so doing, the potential implications of such a more proactive and purposeful internal market on other EU important and horizontal policy objectives will also briefly be addressed.

A.  A ‘Proactive’ and ‘Purposeful’ Instead of the Former ‘Reactive’ and ‘Instrumental’ Internal Market Objective: Quid the Tobacco Directive Test? From 1957 until 2009, the Treaties continued to express, albeit with slight modifications over time, the original vision of a common/internal market which is inherently and necessarily ‘instrumental’ and ‘reactive’. Instrumental, in the sense that the establishment of the internal market was initially not conceived as an objective as such. Creating economic interdependence by merging the national requires neither additional effort nor additional cost, so that a parallel second-hand market would be likely to affect the interests of the copyright holders in obtaining appropriate reward for their works much more than the market for second-hand tangible objects, contrary to the objective referred to in paragraph 48 of the present judgment’. 29 See also the introduction above, Mackenzie-Stuart, above n 15.

82  Inge Govaere markets as much as possible into one common market was rather considered to be a necessary instrument to achieve the higher political objectives of the European integration process, in particular maintaining peace and stability on the European continent. This instrumental approach is immediately apparent from the wording of Article 2 of the (founding) 1957 Rome Treaty, which specified the European Community’s objectives as follows: The Community shall have as its task, by establishing a common market and progressively approximating the economic policies of Member States, to promote throughout the Community a harmonious development of economic activities, a continuous and balanced expansion, an increase in stability, an accelerated raising of the standard of living and closer relations between the States belonging to it.30

The pre-Lisbon approach to the internal market was also clearly meant to be reactive in nature. The focus of the activities listed in Article 3 TEC to achieve the objectives of Article 2 TEC firmly lay in the elimination or prohibition of existing or identifiable barriers to trade between the Member States which might jeopardise the establishment of the internal market. Article 3 of the (founding) 1957 Rome Treaty specified as follows: For the purposes set out in Article 2, the activities of the Community shall include, as provided in this Treaty and in accordance with the timetable set out therein (a) the elimination, as between Member States, of customs duties and of quantitative restrictions on the import and export of goods, and of all other measures having equivalent effect; (c) the abolition, as between Member States, of obstacles to freedom of movement for persons, services and capital.31

This instrumental and reactive approach was not surprisingly also fully endorsed by the CJEU in its pre-Lisbon case law. In the well-known Tobacco Directive case 30 Emphasis added. Art 2 TEC was slightly amended over time; the last pre-Lisbon version (Nice Treaty) stated as follows: ‘The Union shall set itself the following objectives: – to promote economic and social progress and a high level of employment and to achieve balanced and sustainable development, in particular through the creation of an area without internal frontiers, through the strengthening of economic and social cohesion and through the establishment of economic and monetary union, ultimately including a single currency in accordance with the provisions of this Treaty’. 31 Emphasis added. Art 3 TEC was slightly amended over time; the last pre-Lisbon version (Nice Treaty) stated as follows: ‘(1) For the purposes set out in Article 2, the activities of the Community shall include, as provided in this Treaty and in accordance with the timetable set out therein: (a) the prohibition, as between Member States, of customs duties and quantitative restrictions on the import and export of goods, and of all other measures having equivalent effect; (c) an internal market characterised by the abolition, as between Member States, of obstacles to the free movement of goods, persons, services and capital.’

Internal Market Dynamics  83 of 5 October 2000,32 the CJEU emphasised the importance of reading the principle of conferral33 together with the ‘express wording’ of the internal market related Treaty provisions. It pointed in particular to the wording of the abovementioned Article 3(c) TEC combined with the internal market definition (now in Article 26 TFEU, see above), to hold that the internal market harmonisation provision of Article 100a(1) TEC (now Article 114 TFEU) could not be construed as conferring a general competence to regulate the internal market upon the EU.34 Instead, the CJEU put forward the following test to assess the validity of an EU internal market harmonisation measure: the aim must be limited ‘to prevent the emergence of future obstacles to trade resulting from multifarious development of national laws. However, the emergence of such obstacles must be likely and the measure in question must be designed to prevent them’.35 This Tobacco Directive test to delineate EU competence in the field of the internal market has been repeatedly and consistently applied since then, not just in case law but also, unsurprisingly so, by the EU legislature.36 The Lisbon Treaty, however, radically departs from the earlier treaties by opting for a strong proactive and purposeful wording of the internal market objective. The former Article 2 TEC is now repealed and replaced, in substance, by Article 3 TEU.37 Article 3(3) TEU reformulates specifically the internal market objective as follows: The Union shall establish an internal market. It shall work for the sustainable development of Europe based on balanced economic growth and price stability, a highly competitive social market economy, aiming at full employment and social progress, and a high level of protection and improvement of the quality of the environment. It shall promote scientific and technological advance.

32 Case C-376/98 Federal Republic of Germany v European Parliament and Council of the European Union EU:C:2000:544. 33 Now in Art 5 TEU. 34 Case C-376/98, above n 32, see paras 82–83, where the CJEU ruled as follows: ‘(82) Under Article 3(c) of the EC Treaty (now, after amendment, Article 3(1)(c) EC), the internal market is characterised by the abolition, as between Member States, of all obstacles to the free movement of goods, persons, services and capital. Article 7a of the EC Treaty (now, after amendment, Article 14 EC), which provides for the measures to be taken with a view to establishing the internal market, states in paragraph 2 that that market is to comprise an area without internal frontiers in which the free movement of goods, persons, services and capital is ensured in accordance with the provisions of the Treaty; (83) Those provisions, read together, make it clear that the measures referred to in Article 100a(1) of the Treaty are intended to improve the conditions for the establishment and functioning of the internal market. To construe that article as meaning that it vests in the Community legislature a general power to regulate the internal market would not only be contrary to the express wording of the provisions cited above but would also be incompatible with the principle embodied in Article 3b of the EC Treaty (now Article 5 EC) that the powers of the Community are limited to those specifically conferred on it’. 35 Case C-376/98, above n 32, see para 86. 36 For a critique, see S Weatherill, ‘The Limits of Legislative Harmonization Ten Years after Tobacco Advertising: How the Court’s Case Law has become a ‘Drafting Guide’’ (2011) 12 German Law Journal 827. 37 Table of Equivalences to the Lisbon Treaty, see [2012] OJ C326/1, 368.

84  Inge Govaere It shall combat social exclusion and discrimination, and shall promote social justice and protection, equality between women and men, solidarity between generations and protection of the rights of the child. It shall promote economic, social and territorial cohesion, and solidarity among Member States. It shall respect its rich cultural and linguistic diversity, and shall ensure that Europe’s cultural heritage is safeguarded and enhanced.

Upon reading the very first short sentence of Article 3(3) TEU, it is immediately striking that the establishment of the internal market is an objective in itself in a post-Lisbon setting and thus is no longer merely instrumental in serving another stipulated (political) objective. Besides thus becoming ‘purposeful’ in itself, the wording used also expresses a strong assertion that the internal market ‘shall be established’ and thus ‘proactively’ pursued. Moreover, Article 3(3) TEU dictates that such a purposeful internal market shall also proactively pursue other higher and horizontal Treaty objectives. The Lisbon Treaty thus provides the necessary, but seemingly also compulsory, legal framework for the establishment of a more proactive internal market policy which positively intertwines with other EU policies.38 It cannot go unnoticed that this echoes the provisions of general application as listed in Articles 7–13 TFEU.39 At the same time, the Lisbon Treaty also fundamentally alters the pre-Lisbon appraisal in terms of conferral of competence to the EU specifically in relation to the internal market. The former Article 3(1) TEC, which conferred on the EU the competence to remove obstacles to the internal market, is repealed and held to be replaced in substance by Articles 3–6 TFEU.40 Interestingly, Article 4(2)(a) TFEU now expressly provides that the EU and the Member States share competence with respect to ‘the internal market’. It does so in a general manner, without further qualification or limitation. According to Article 2(2) TFEU this then unequivocally means that ‘the Union and the Member States may legislate and adopt legally binding acts in that area’.41 An important exception thereto is laid down in Article 3(b) TFEU, which stipulates that the EU has an exclusive competence to establish ‘the competition rules which are necessary for the functioning of the internal market’. Yet this only further reinforces rather than weakens the general

38 See also below, III.B.ii. 39 Art 7 TFEU in particular reads: ‘The Union shall ensure consistency between its policies and activities, taking all of its objectives into account and in accordance with the principle of conferral of powers’. For an analysis, see N Nic Shuibhne, ‘Deconstructing and reconstructing Article 7 TFEU’ in F Ippolito, ME Bartoloni and M Condinanzi (eds), The EU and the Proliferation of Integration Principles under the Lisbon Treaty (London, Routledge, 2019). For an analysis of the individual horizontal provisions as listed in Arts 8–13 TFEU, see also the other contributions to this edited volume. 40 Table of Equivalence to the Lisbon Treaty, see [2012] OJ C326/1, 369. 41 It continues to read: ‘The Member States shall exercise their competence to the extent that the Union has not exercised its competence. The Member States shall again exercise their competence to the extent that the Union has decided to cease exercising its competence’.

Internal Market Dynamics  85 regulatory competence conferred on the EU to act in relation to the internal market in a post-Lisbon setting.42 The Lisbon Treaty thus revolutionises the primary law framework against which the more detailed internal market provisions need to be assessed in terms of both its objectives and competences conferred on the EU. Yet rather than shaking to the core the pre-Lisbon internal market case law and practice, it will be illustrated in the next section that this huge potential has to a large extent remained untapped.

B.  Post-Lisbon Case Law on Article 3(3) TEU It is rather surprising to find that the above apparent and important contextual change brought by the Lisbon Treaty to the internal market legal framework, has been largely ignored and underexplored until now. Two reasons might have pointed in a different direction and will be explored in the post-Lisbon internal market case hereunder. First there is the so-called ‘purposive’ or teleological method of interpretation of EU law used by the CJEU, as mentioned before,43 whereby EU law provisions are usually interpreted and given a useful effect in view of the underlying Treaty objectives. When the latter changes so, the logic goes, should then also the former. A pre-Lisbon example of such a purposive interpretation of internal market provisions is to be found in the Viking case,44 where the CJEU attached great importance to the wording of Articles 2 and 3 TEC, to put social policy in the balance with internal market policy.45 The second is the heavy reliance on the, now repealed, wording of Article 3(3) TEC in combination with the principle of conferral by the CJEU in the Tobacco 42 According to Art 2(2) TFEU this means that ‘only the Union may legislate and adopt legally binding acts, the Member States being able to do so themselves only if so empowered by the Union or for the implementation of Union acts’. 43 See above, Introduction. 44 Case C-438/05 International Transport Workers’ Federation and Finnish Seamen’s Union v Viking Line EU:C:2007:772. On the convergence between internal market law and fundamental rights, see for instance F de Cecco, ‘Fundamental Freedoms, Fundamental Rights, and the Scope of Free Movement Law’ (2014) 15 German Law Journal 383; N Nic Shuibhne, ‘The Social Market Economy and Restriction of Free Movement Rights: plus c’est la même chose?’ (2019) 57 Journal of Common Market Studies 111. 45 Case C-438/05, above n 44, see paras 78–79: ‘It must be added that, according to Article 3(1)(c) and (j) EC, the activities of the Community are to include not only an “internal market characterised by the abolition, as between Member States, of obstacles to the free movement of goods, persons, services and capital”, but also “a policy in the social sphere”. Article 2 EC states that the Community is to have as its task, inter alia, the promotion of “a harmonious, balanced and sustainable development of economic activities’ and ‘a high level of employment and of social protection”’. Since the Community has thus not only an economic but also a social purpose, the rights under the provisions of the Treaty on the free movement of goods, persons, services and capital must be balanced against the objectives pursued by social policy, which include, as is clear from the first paragraph of Art 136 EC, inter alia, improved living and working conditions, so as to make possible their harmonisation while improvement is being maintained, proper social protection and dialogue between management and labour.

86  Inge Govaere Directive case to elaborate the validity test for EU harmonisation measures. Is reference to such pre-Lisbon case law and internal market test still good law today? By way of example, in the Daiichi Sankyo case the CJEU expressly acknowledged in relation to Article 207 TFEU that ‘in view of that significant development of primary law, the question of the distribution of the competences of the European Union and the Member States must be examined on the basis of the Treaty now in force’.46 The expressly drawn consequence with respect to the concept of common commercial policy was that, as from the entry into force of the Lisbon Treaty, important and authoritative pre-Lisbon case law could simply no longer continue to be invoked successfully. Contrary to such a rather logical dynamic case law responding to the internal dynamics of Treaty amendments, post-Lisbon case law has so far devoted little or no attention to the new purposeful and proactive formulation of the internal market objective in Article 3(3) TEU.47 Perhaps this is simply due to the fact that the dynamic challenge was not often raised by the parties.48 Before turning to that internal market specific case law, first attention will briefly be given to the importance of Article 3(3) TEU specifically in relation to competition policy as well as environmental protection and social and linguistic policy as acknowledged by the CJEU.

46 In Case C-414/11 Daiichi Sankyo EU:C:2013:520, para 48 the CJEU held: ‘In view of that significant development of primary law, the question of the distribution of the competences of the European Union and the Member States must be examined on the basis of the Treaty now in force (see, by analogy, Opinion 1/08 [2009] ECR I-11129, paragraph 116). Consequently, neither Opinion 1/94 ([1994] ECR I-5267), in which the Court established in relation to Article 113 of the EC Treaty which provisions of the TRIPs Agreement fell within the common commercial policy and hence the exclusive competence of the Community, nor the judgment in Merck Genéricos – Produtos Farmacêuticos, defining, at a date when Article 133 EC was in force, the dividing line between the obligations under the TRIPs Agreement assumed by the European Union and those remaining the responsibility of the Member States, is material for determining to what extent the TRIPs Agreement, as from the entry into force of the FEU Treaty, falls within the exclusive competence of the European Union in matters of the common commercial policy’. 47 Art 3(3) TEU was expressly referred to in relation to free movement, but only in relation to intraMember States double taxation treaties, Case C-648/15 Austria v Germany (double taxation treaties) EU:C:2017:664. See para 26: ‘That is manifestly so in the present case, in the light of the beneficial effect of the mitigation of double taxation on the functioning of the internal market that the European Union seeks to establish in accordance with Article 3(3) TEU and Article 26 TFEU. As the European Commission observed, in essence, in its communication to the European Parliament, the Council and the European Economic and Social Committee of 11 November 2011, entitled ‘Double Taxation in the Single Market’ (COM(2011) 712 final), 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’. 48 In Case C-152/17 Consorzio Italian Management e Catania Multiservizi EU:C:2018:264, the Art 3(3) TEU question was held to be inadmissible. At para 23, the CJEU held: ‘It should be noted, in that regard, that the order for reference gives no explanation of the relevance of the interpretation of Article 3(3) TEU or Articles 26, 57, 58 and 101 TFEU for the resolution of the dispute in the main proceedings’.

Internal Market Dynamics  87

i.  The Importance of Article 3(3) TEU for Competition and Horizontal Provisions The express referral to Article 3(3) TEU in competition cases is made by the CJEU mainly to emphasise that the Lisbon Treaty does not modify the prior understanding that the competition rules are and remain an essential complement to the internal market. Uncertainty had arisen in that respect by the Lisbon Treaty’s perceived ‘relegation’ of competition from the Treaty objectives to Protocol (No 27) on the internal market and competition. The CJEU instead opts for a combined reading of Article 3(3) TEU with Protocol (No 27) which, it is stressed, is an integral part of the Treaties by virtue of Article 51 TEU and ensures that the internal market is to include a system ensuring that competition is not distorted.49 Article 3(3) TEU is also expressly invoked by the CJEU with reference to the other and more horizontal EU policies. In relation to environmental protection, post-Lisbon case law systematically reads the other Treaty provisions, in particular Article 37 of the Charter of Fundamental Rights of the European Union and Article 191(2) TFEU, together with Article 3(3) TEU to underscore that ‘EU policy on the environment aims at a high level of protection’.50 There are also several judgments on linguistic policy, whereby the CJEU systematically refers to a combined reading of both the fourth subparagraph of Article 3(3) TEU and Article 22 of the Charter of Fundamental Rights of the European Union to underline that, as a matter of principle, ‘the Union is committed to the preservation of multilingualism’.51

49 See for instance Case C-52/09 TeliaSonera Sverige EU:C:2011:83, para 20: ‘Article 3(3) TEU states that the European Union is to establish an internal market, which, in accordance with Protocol No 27 on the internal market and competition, annexed to the Treaty of Lisbon (OJ 2010 C 83, p 309), is to include a system ensuring that competition is not distorted’; or Case C-93/17 Commission v Greece (state aid) EU:C:2018:903, para 122: ‘The rules which are the subject of Decision 2009/610 and the judgment establishing the failure to fulfil obligations are the expression of one of the essential missions conferred on the European Union by virtue of Article 3(3) TEU, that is the establishment of the internal market and Protocol (No 27) on the internal market and competition, which, in accordance with Article 51 TEU, is an integral part of the Treaties and under which the internal market concludes a system guaranteeing that competition is not distorted’. 50 See for instance Case C-197/18 Wasserleitungsverband Nördliches Burgenland and Others EU:C:2019:824, para 49; Case C-723/17 Craeynest and Others EU:C:2019:533, para 33; Case C-128/17 Poland v Parliament and Council EU:C:2019:194 paras 127–30. At para 130, the CJEU clarifies: ‘Article 52(2) of the Charter provides that rights recognised by the Charter for which provision is made in the Treaties are to be exercised under the conditions and within the limits defined by those Treaties. Such is the case with Article 37 of the Charter, which is essentially based on Article 3(3) TEU and Articles 11 and 191 TFEU (judgment of 21 December 2016, Associazione Italia Nostra Onlus, C-444/15, EU:C:2016:978, paragraph 62)’. 51 See for instance Case C-147/13 Spain v Council EU:C:2015:299, para 42; Case C-202/11 Las EU:C:2013:239, para 26; Case C-377/16 Spain v Parliament EU:C:2019:249, para 36. For a case where the CJEU only refers to Art 3(3) TEU in this respect, see Case C-73/17 France v Parliament (Exercice du pouvoir budgétaire) EU:C:2018:787, para 41: ‘The Parliament is therefore required to act in this area with all the attention, rigour and commitment which such a responsibility demands (see, to that effect, judgment of 13 December 2012, France v Parliament, C-237/11 and C-238/11, EU:C:2012:796, paragraph 68), which presupposes that the parliamentary debate and vote be based on a text that

88  Inge Govaere Interestingly, the CJEU also seems to endorse a more purposeful reading of Article 3(3) TEU in so far as social protection is concerned. In the Test-Achats case, it ruled that: While Article 157(1) TFEU establishes the principle of equal treatment for men and women in a specific area, Article 19(1) TFEU confers on the Council competence which it must exercise in accordance, inter alia, with the second subparagraph of Article 3(3) TEU, which provides that the European Union is to combat social exclusion and discrimination and to promote social justice and protection, equality between men and women, solidarity between generations and protection of the rights of the child, and with Article 8 TFEU, under which, in all its activities, the European Union is to aim to eliminate inequalities, and to promote equality, between men and women. In the progressive achievement of that equality, it is the EU legislature which, in the light of the task conferred on the European Union by the second subparagraph of Article 3(3) TEU and Article 8 TFEU, determines when it will take action, having regard to the development of economic and social conditions within the European Union. However, when such action is decided upon, it must contribute, in a coherent manner, to the achievement of the intended objective, without prejudice to the possibility of providing for transitional periods or derogations of limited scope.52

In other cases the CJEU continues to refer expressly to the above-mentioned preLisbon Viking ruling as still good law in a post-Lisbon context, to underscore that, as is apparent from Article 3(3) TEU, the European Union is not only to establish an internal market but is also to work for the sustainable development of Europe, which is based, in particular, on a highly competitive social market economy aiming at full employment and social progress, and it is to promote, inter alia, social protection.53

It is thus apparent that in a post-Lisbon setting, Article 3(3) TEU is heavily relied upon to reinforce the application of the horizontal provisions of Articles 7–13 TFEU and the Charter, either on their own or in combination with the internal market. This is in line with the above assessment that Article 3(3)TEU provides the necessary legislative framework for the establishment of an internal market policy which positively intertwines with other EU policies, thereby reinforcing the provisions of general application.

has been made available to the Members in good time and been translated into all the official EU languages. The European Union is committed to multilingualism, the importance of which is stated in the fourth subparagraph of Article 3(3) TEU (see, to that effect, judgment of 5 May 2015, Spain v Council, C-147/13, EU:C:2015:299, paragraph 42, and of 6 September 2017, Slovakia and Hungary v Council, C-643/15 and C-647/15, EU:C:2017:631, paragraph 203)’. 52 Case C-236/09 Association belge des Consommateurs Test-Achats and Others EU:C:2011:100, paras 19–21. In casu the CJEU ruled that ‘Article 5(2) of Council Directive 2004/113/EC of 13 December 2004 implementing the principle of equal treatment between men and women in the access to and supply of goods and services is invalid with effect from 21 December 2012’, as it considered there to be a risk that EU law would permit the derogation from the equal treatment of men and women, to persist indefinitely (see para 31). 53 Case C-201/15 AGET Iraklis EU:C:2016:972, para 76.

Internal Market Dynamics  89

ii.  The Importance of Article 3(3) TEU for Internal Market Provisions In view of the above, it is all the more striking to find that the fundamentally changed internal market objective, from reactive and instrumental to proactive and purposeful, is not at all addressed by the CJEU in relation to the establishment of the internal market as such. In fact, it would appear that the very first sentence of Article 3(3) TEU has so far never been considered by the CJEU in post-Lisbon case law. Not even in relation to the interpretation of specific internal market provisions, such as Article 114 TFEU. Instead, the CJEU simply points back to pre-Lisbon rulings as the basis for a consistent line of case law which still stands good post-Lisbon. It nonetheless seems highly problematic for the CJEU to bluntly do so without explaining the underlying rationale. How does this square with the reformulation of the internal market objective, but also with the abovementioned repeal of former Article 3 TEC which was very much the anchorage of the pre-Lisbon case law? Unfortunately no such insights are currently being offered, as the post-Lisbon Tobacco cases amply illustrate.54 For instance the 2016 Phillip Morris case could have been a perfect occasion for the CJEU to clarify the impact of the Lisbon Treaty amendments on the internal market concept, in the same vein as the abovementioned Daiichi Sankyo case was used by the CJEU to explain the impact of the Lisbon Treaty on the concept of common commercial policy. Instead, the CJEU simply turned to the pre-Lisbon Tobacco Directive test as the point of departure of its analysis. It recalled that for the EU to be able to act under Article 114 TFEU ‘a mere finding of disparities between national rules is not sufficient’, but that there also has to be an ‘obstruction of the fundamental freedoms’ or a need ‘to prevent the emergence of future obstacles to trade’.55 With the repeal of the former Article 3 TEC and the express conferral of shared competence for the internal market in Article 4 TFEU, it seems difficult to anchor such restrictive conditions in the Treaties post-Lisbon. The continuation of the ‘obstacles’ test also does not seem to follow from the wording Article 114 TFEU itself, as it more generally provides for the possibility to adopt EU ‘measures for the approximation of the provisions

54 Case C-358/14 Poland v Parliament and Council EU:C:2016:323; Case C-547/14 Philip Morris EU:C:2016:325. 55 Case C-547/14, above n 54; see in particular paras 58–60: ‘In that regard, while a mere finding of disparities between national rules is not sufficient to justify having recourse to Article 114 TFEU, it is otherwise where there are differences between the laws, regulations or administrative provisions of the Member States which are such as to obstruct the fundamental freedoms and thus have a direct effect on the functioning of the internal market (see, to that effect, judgments in Germany v Parliament and Council, C-376/98, EU:C:2000:544, paragraphs 84 and 95; British American Tobacco (Investments) and Imperial Tobacco, C-491/01, EU:C:2002:741, paragraphs 59 and 60; Arnold André, C-434/02, EU:C:2004:800, paragraph 30; Swedish Match, C-210/03, EU:C:2004:802, paragraph 29; Germany v Parliament and Council, C-380/03, EU:C:2006:772, paragraph 37; and Vodafone and Others, C-58/08, EU:C:2010:321, paragraph 32).

90  Inge Govaere laid down by law, regulation or administrative action in Member States which have as their object the establishment and functioning of the internal market’.56 The broader the scope of the internal market measures, the higher will of course be the potential impact on other important and horizontal EU policy objectives, such as public health. This does not however mean that such objectives need to clash. Earlier in the pre-Lisbon Tobacco Directive case and later restated in the post-Lisbon Phillip Morris ruling,57 the CJEU pointed to the convergence imposed by Article 114(3) TFEU which explicitly requires that in achieving EU harmonisation, a high level of, inter alia, protection of human health should be guaranteed.58 Article 3(3) TEU now makes an even clearer and more imperative link between the pursuit of the internal market and other higher objectives.59 As already developed elsewhere,60 the Lisbon Treaty, seems to provide the necessary legislative framework for such a more proactive internal market policy which positively intertwines with other policies.61 The big challenge in so doing would be (59) It is also settled case-law that, although recourse to Article 114 TFEU as a legal basis is possible if the aim is to prevent the emergence of future obstacles to trade as a result of divergences in national laws, the emergence of such obstacles must be likely and the measure in question must be designed to prevent them (judgments in British American Tobacco (Investments) and Imperial Tobacco, C-491/01, EU:C:2002:741, paragraph 61; Arnold André, C-434/02, EU:C:2004:800, paragraph 31; Swedish Match, C-210/03, EU:C:2004:802, paragraph 30; Germany v Parliament and Council, C-380/03, EU:C:2006:772, paragraph 38; and Vodafone and Others, C-58/08, EU:C:2010:321, paragraph 33). (60) The Court has also held that, provided that the conditions for recourse to Article 114 TFEU as a legal basis are fulfilled, the EU legislature cannot be prevented from relying on that legal basis on the ground that public health protection is a decisive factor in the choices to be made (judgments in British American Tobacco (Investments) and Imperial Tobacco, C-491/01, EU:C:2002:741, paragraph 62; Arnold André, C-434/02, EU:C:2004:800, paragraph 32; Swedish Match, C-210/03, EU:C:2004:802, paragraph 31; and Germany v Parliament and Council, C-380/03, EU:C:2006:772, paragraph 39)’. 56 Interestingly, in para 63, the CJEU does look at the wording of Art 114 TFEU to rule that ‘by using the words “measures for the approximation” in Article 114 TFEU, the authors of the Treaty intended to confer on the EU legislature a discretion, depending on the general context and the specific circumstances of the matter to be harmonised, as regards the method of approximation most appropriate for achieving the desired result, in particular in fields with complex technical features (judgments in Germany v Parliament and Council, C-380/03, EU:C:2006:772, paragraph 42, and United Kingdom v Parliament and Council, C-270/12, EU:C:2014:18, paragraph 102. It was thus open to the EU legislature, in the exercise of that discretion, to proceed towards harmonisation only in stages and to require only the gradual abolition of unilateral measures adopted by the Member States (judgment in Rewe-Zentral, C-37/83, EU:C:1984:89, paragraph 20)’. 57 Case C-547/14, above n 54, para 61. 58 For an interesting analysis of that case law, see for instance V Delhomme, ‘Between Market Integration and Public Health: The Paradoxical EU Competence to Regulate Tobacco Consumption’ (2018) College of Europe series Research Papers In Law 1/2018. 59 On the link between internal market and higher objectives, see also B de Witte, ‘A competence to protect: the pursuit of non-market aims through internal market legislation’ in P Syrpis, The Judiciary, the Legislature and the EU Internal Market (Cambridge, Cambridge University Press, 2012); B de Witte, ‘Non-market values in internal market legislation’ in N Nic Shuibhne (ed), Regulating the Internal Market (Cheltenham, Edward Elgar, 2006). 60 For a detailed analysis, see I Govaere, ‘Modernisation of the internal market: potential clashes and crossroads with other policies’ in M De Vos (ed), European Union Internal Market and Labour Law: Friends or Foes? (Cambridge, Intersentia, 2009); for a slightly amended version see I Govaere, ‘The Future Direction of the EU Internal Market: On Vested Values and Fashionable Modernism’ (2009) 16 Columbia Journal of European Law 67. 61 See also ch 16 by Sacha Garben in this volume.

Internal Market Dynamics  91 to deal with the issues up front and to provide the necessary crossroads between the different policies with different institutional implications.62 It is also crucial to actively search for complementarities and to establish sound and enduring criteria for reaching an acceptable balance, where needed, between the internal market and such other horizontal objectives.63 It cannot in itself be a reason, however, to refrain from revisiting the very internal market concept as such.

IV. Conclusion Much has already been said in relation to the internal market but, paradoxically, too much has also been left unspoken. Of all the Treaty concepts, the internal market is probably the most EU-specific, the most important and the most stable. As such, it was and continues to be crucial to maintain and further the European integration process, to the extent that the latter would probably not be viable without the former. A crucial feature thereto is the constantly revisited and dynamically applied internal market concept both by the CJEU and EU policymakers to tackle shifting contextual factors, whether EU-specific or global. It thus responds elastically to external dynamics. A strong dynamic internal market however also necessitates critical reflection about its own internal dynamics as a concept. To constantly question why and how certain rules and tests, such as the Tobacco Directive test, were developed in the first place allows for the concept itself to be invigorated when the underlying factors that led to those answers change. A fortiori when a moving target is expressly endorsed by Treaty amendment it would seem that a thorough reflection of former practice is called for. Surprisingly enough, radical amendment of the internal market objectives by the Lisbon Treaty, from a reactive and instrumental to a proactive and purposeful approach, has not triggered such a reflective response. It appears that, without due consideration, the pre-Lisbon Tobacco Directive test simply continues to be applied largely unchallenged post-Lisbon. As such the great potential of Article 3(3) TEU post-Lisbon to trigger renewed internal dynamics of the internal market concept remains largely underexplored and untapped.

62 See also AG Sharpston, opinion delivered on 11 April 2019, Case C-482/17 Czech Republic v Parliament and Council EU:C:2019:321, para 40: ‘The powers to regulate the internal market touch on a plethora of other matters. The scope of Article 114 TFEU therefore has particular significance for Member States for two main reasons. First, whether acts based on that provision encroach upon matters for which the Treaties do not confer competence on the EU legislature (as the Czech Republic argues here) is a sensitive issue. Second, EU acts are adopted under that provision following the ordinary legislative procedure (thus, employing qualified majority voting), which means that no one Member State is able to veto any proposed act’. 63 For an example, see Case C-482/17, above n 62, esp paras 52–53, where the CJEU assessed the balance between the free movement of goods and security guarantees.

92

part ii The Four Freedoms

94

5 The Classic Freedom? The Free Movement of Goods: Old Doctrines, New Cases and Contemporary Reflections ELEANOR SPAVENTA*

I. Introduction It is undeniable that the free movement of goods, and especially the Court’s interpretation of measures having equivalent effect to quantitative restrictions on imports, has laid some of the founding pillars of the internal market. Here, it is sufficient to recall the ruling in Cassis de Dijon which, by imposing on Member States the principle of mutual recognition of regulatory standards, allowed intraState trade to develop without the need for harmonisation.1 The Commission, consequently, could focus its energies to propose common standards only where those were necessary for the achievement of the internal market by dismantling obstacles to trade which would be justified by public policy concerns, and especially public health and consumer protection.2 If the foundations were laid in the 1970s, in the 1980s the Court had the chance to develop a rich body of case law on regulatory barriers and potential justifications through the mandatory requirements doctrine. In the 1990s, the attention shifted to the boundaries of the free movement of goods, ie, to where exactly the line should be drawn between an extensive interpretation of EU law and the need and will to preserve at least some national regulatory autonomy. Here, we see an intense scholarly debate on how to limit the reach of the Dassonville formula so as to ensure that interstate tradeneutral measures, ie, those rules which did not specifically affect imports or the * I am grateful to the organisers and participants in the conference; the usual disclaimer applies. 1 Case 120/78 Rewe-Zentral AG v Bundesmonopolverwaltung für Branntwein (Cassis de Dijon) EU:C:1979:42. 2 Communication from the Commission concerning the consequences of the judgment given by the Court of Justice on 20 February in Case 120/78 (‘Cassis de Dijon’) [1980] OJ C256/2.

96  Eleanor Spaventa flow of goods within the EU, would not be subjected to the proportionality assessment required by the mandatory requirements doctrine.3 The perceived risk in that respect is that the judicature (national and EU) would substitute its own assessment of the wisdom of given regulatory policies for that of the democratically accountable decision-maker, with all the problems that that entails not only for governance but also in relation to access to expertise. In the second part of the 1990s – following the clarification in Keck to the effect that certain rules (certain selling arrangements) would only be caught by Article 34 of the Treaty on the Functioning of the European Union (TFEU) if directly or indirectly discriminatory, there is a 180-degree shift in the debate so that the issue becomes whether the introduction of a presumption that certain rules do not interfere with interstate trade might not leave some barriers undetected.4 The early 2000s see the Court wavering between various tests in relation to Article 34 TFEU: market access, discrimination and a combination of the two. This culminates in Commission v Italy with a new category of rules, ie, those regulating the use of a product, which are subjected to a broad market access test.5 Whereas we will look in more detail at the various tests in section I, what is important to note here is that despite a certain lack of clarity in the dicta of the Court, the sheer number of cases have provided national courts with a reasonable idea of the boundaries of Article 34 TFEU. And yet, this does not mean that there are no challenges ahead – simply that those have evolved with the development of the internal market. In particular, in this contribution and after having recalled the scope of application of Articles 34 and 35 TFEU, I will turn to discuss what I consider 3 See, eg, EL White, ‘In Search of the Limits to Article 30 of the EEC Treaty’ (1989) 26 Common Market Law Review 235; L Gormley, ‘Recent case law on the free movement of goods: some hot potatoes’ (1990) 27 Common Market Law Review 825; A Arnull, ‘What shall we do on Sunday?’ (1991) 16 European Law Review 112; K Mortelmans, ‘Article 30 of the EEC Treaty and Legislation Relating to Market Circumstances: Time to Consider a New Definition?’ (1991) 28 Common Market Law Review 115. 4 AG Jacobs’ opinion in Case C-412/93 Société d’Importation Edouard Leclerc-Siplec v TF1 Publicité SA and M6 Publicité SAI EU:C:1994:393; S Weatherill, ‘After Keck: some thoughts on how to clarify the clarification’ (1996) 33 Common Market Law Review 885; C Barnard, ‘Fitting the remaining pieces into the goods and persons jigsaw?’ (2001) 26 European Law Review 35; N Nic Shuibhne, ‘The free movement of goods and Article 28 EC: an evolving framework’ (2002) 27 European Law Review 408. 5 Case C-110/05 Commission v Italy (Moped trailers) EU:C:2009:66; and Case C-142/05 Åklagaren v Percy Mickelsson and Joakim Roos EU:C:2009:336. Those cases have provoked a fair amount of academic analysis; see E Spaventa, ‘Leaving Keck Behind?: The Free Movement of Goods after the Rulings in Commission v Italy and Mickelsson and Roos’ (2009) 35 European Law Review 914; P Wennerås and K Bøe Moen, ‘Selling Arrangements, Keeping Keck’ (2010) 35 European Law Review 387; S Enchelmaier, ‘Moped Trailers, Mickelsson & Roos, Gysbrechts: the ECJ’s Case Law on Goods Keeps on Moving’ (2010) 29 Yearbook of European Law 190; P Oliver, ‘Of Trailers and Jet Skis: Is the Case Law on Article 34 TFEU Hurtling in a new Direction?’ (2010) 33 Fordham International Law Journal 1423; L Gormley, ‘Free Movement of Goods and their Use: What is the Use of it?’ (2010) 33 Fordham International Law Journal 1589; T Horsley, ‘Unearthing buried treasure: Art, 34 TFEU and the exclusionary rules’ (2012) 37 European Law Review 734; MS Jansson and H Kalimo, ‘De minimis meets ‘market access’: transformations in the substance – and the syntax – of EU free movement law?’ (2014) 51 Common Market Law Review 523; L Gormley, ‘Inconsistencies and misconceptions in the free movement of goods’ (2015) 40 European Law Review 925; R Schütze, ‘Of types and tests: towards a unitary doctrinal framework for Article 34 TFEU?’ (2016) 41 European Law Review 826.

The Free Movement of Goods  97 to be the more problematic areas. In this respect I will focus on the procedural constraints imposed on regulators by Article 34 TFEU and on the problem arising from an abstract and detailed proportionality assessment in relation to devolved administrations.

II.  The Free Movement of Goods between Keck and Market Access The free movement of goods, and especially Article 34 TFEU, have been the focus of much academic debate and the case law in this field has been determinant in shaping internal market law more generally. In this respect, and amongst the various turns in the case law, two core principles have remained constant; first, any rule which is directly or indirectly discriminatory favouring domestic production is always caught by the Treaty, be it a restriction on imports or a restriction on exports.6 Second, in the case of Article 34 TFEU, the underlying principle is that of mutual recognition of regulatory standards established in Cassis de Dijon,7 pursuant to which once goods are lawfully produced and marketed in one Member State they can be sold in every EU marketplace unless there is a mandatory requirement of public interest which justifies a restriction to movement.8 With time, it will become clear that Cassis introduces an absolute and non-rebuttable presumption of effect on intra-EU trade of product requirements, ie, those rules which affect the physical qualities of the product: on the one hand, those rules are very restrictive because they might require the manufacturer to have a separate line of production to comply with different regulatory standards in different Member States; on the other hand, the mutual trust which lies at the foundation of the internal market and other areas of EU integration means also trusting that the regulatory standards imposed by other Member States are sufficient to protect consumers and public health. For this reason, Member States can impose their own regulatory standards only to the extent to which that is absolutely necessary (strict proportionality test) to pursue a legitimate public interest.

6 Consistent case law, see eg, Case 113/80 Commission v Ireland (Irish souvenirs) EU:C:1981:139; Case 72/83 Campus Oil EU:C: 1984:256; Case 15/79 Groenweld EU:C:1979:253; Case C-209/98 FFAD (Sydhavnens Sten & Grus) EU:C:2000:279; indirect discrimination has only more recently been brought within the scope of Art 35 TFEU; see Case C-205/07 Lodewijk Gysbrechts EU:C:2008:730. However, it should be considered that we have comparatively little case law on Art 35 TFEU and, in the writer’s opinion, this is the reason for the slow evolution. 7 Case 120/78 Rewe-Zentral AG, above n 1. 8 See also Regulation 764/2008 laying down procedures relating to the application of certain national technical rules to products lawfully marketed in another Member State and repealing decision 3052/95 [2008] OJ L218/21.

98  Eleanor Spaventa At the opposite end of the spectrum, we also know that, after the Sunday Trading cases,9 the Court is reluctant to interfere with some regulatory choices which are considered a matter for national discretion: opening times, mandatory rest periods, planning restrictions10 and the like do not fall within the scope of Article 34 TFEU11 (they might however fall within the scope of another internal market freedom) and therefore do not need to be justified unless they are directly or indirectly discriminatory.12 It should be remembered though that we do not have any case law on ‘extreme measures’ such as for instance rules that would only allow the opening of shops for a very limited amount of time during the night: it is very unlikely that the Court would not subject those to scrutiny. In this respect, it will be interesting to see whether any challenge to national rules will arise from the response to the COVID-19 pandemic. As is well known, many Member States have introduced very restrictive measures that mandate the closure of most retail shops, bar those selling essential goods. The question, and not a merely academic one, is how those measures should be qualified: on the one hand, as mentioned above rules relating to opening times are selling arrangements and therefore only fall within the scope of Article 34 TFEU to the extent to which they are directly or indirectly discriminatory (which would not be the case here). But, on the other hand, the severity of those measures clearly interferes with the functioning of the internal market; whereas they would most likely be justified on public health grounds,13 their severity fully justifies the Court of Justice’s oversight, which should be exercised also having regard to fundamental rights and respect for basic democratic accountability. That said, there is a range of rules in between the two extremes (product requirements and selling arrangements) in relation to which the Court has found itself wavering between different tests and presumptions: those are rules concerning advertising, rules restricting long-distance sales and especially internet sales, and rules regulating, rather than altogether prohibiting, the use of a product. 9 Case 145/88 Torfaen Borough Council v B&Q plc EU:C:1989593; Case C-332/89 Conforama EU:C:1991:94; Case 306/88 Rochdale Borough Council v SJ Anders EU:C:1992:510; Case C-169/91 Council of the City of Stoke on Trent and Norwich City Council v B&Q EU:C:1992:519. 10 See, eg, Case C-69/93 Punto Casa EU:C:1994:226; Case C-418/93 Semeraro Casa Uno EU:C:1996:242; Case C-387/93 Banchero EU:C:1995:439; Case C-401/92 Tankstation EU:C:1994:220; Case C-20/03 Burmanjer EU:C:2005:307; Case C-441/04 A-Punkt EU:C:2006:141 11 There was some confusion about whether price fixing rules fell in principle within the scope of Art 34 TFEU; it now seems that this is always the case, although whether market access discrimination is presumed or must be proven is still uncertain. For the former see, eg, Scottish Whisky, also discussed below; Case C-148/15 Deutsche Parkinson Vereinigung eV EU:C:2016:776. 12 The Court has confirmed its case law on Certain Selling Arrangements also after the ruling in Commission v Italy; see, eg, Case C-221/15 Etablissements Fr Colruyt NV EU:C:2016:704, on rules prohibiting the sale of tobacco products at a price lower than that indicated by the revenue stamp; Case C-198/14 Visnappu EU:C:2015751, in relation to a discriminatory licensing regime. 13 This would of course depend on the details and coherence of the measures adopted, although I think it is safe to say that it would take extreme and unreasonable measures for the presumption of public health justification to be reversed in this instance.

The Free Movement of Goods  99

A.  Advertising, Long Distance Selling Rules and Rules Regulating Use In the Keck ruling the Court held that certain selling arrangements fell within the scope of EU law only insofar as directly or indirectly discriminatory.14 It is well known that the Keck categorisation stemmed from Advocate General Tesauro’s opinion in the parallel Hünermund case:15 the rules in that case concerned a prohibition on advertising of quasi-pharmaceutical products near pharmacies, their advertisement being otherwise allowed; for this reason those rules created no real disruption to trade in general and to intra-Union trade more specifically. The factual background was then deemed by AG Tesauro the perfect occasion for rethinking the case law on Article 34 TFEU: after all, academic opinion almost unanimously condemned the Sunday Trading case law,16 as it was seen as an undue interference with the national regulatory autonomy of the Member States not justified by the aims of Article 34 TFEU. It is in this context that the Advocate General suggested the distinction between different types of rule: product requirements and rules concerning the manner in which trading activity is carried out. The former, as mentioned above, are very trade restrictive and are presumed to affect intra-EU trade so that they always have to be justified pursuant to the mandatory requirements doctrine. However, rules regulating when and how a product can be sold should be presumed to be intra-EU trade neutral, so that they should fall within the scope of Article 34 TFEU only to the extent to which they affect imported products more than domestic products. This distinction between product requirements and selling arrangements was then adopted by the Court in the Keck case which was decided a few months after Tesauro’s opinion in Hünermund. The ruling in Keck then created a system of presumptions which proved very useful especially for national courts and rulemakers, increasing legal certainty and delimiting more clearly the boundaries of the Dassonville formula.17

14 If a rule, which would be otherwise classed as a selling arrangement, requires the product to be amended in any way (labelling, content etc), it falls within the scope of Art 34 TFEU and must be justified. See, eg, Case C-368/95 Vereinigte Familiapress Zeitungsverlags- und vertriebs GmbH v Heinrich Bauer Verlag EU:C:1997:325; and Case C-244/06 Dynamic Medien Vertriebs GmbH EU:C:2008:85; for a good example of a truly indirectly discriminatory selling arrangement see Case C-104/07 Commission v Germany EU:C:2008:492 (Geographical proximity of pharmacies to supply medicinal products to hospitals). 15 AG Tesauro’s opinion in Case C-292/92 Hünermund and others EU:C:1993:863; as is well known, AG Tesauro extensively drew from an article by EL White (then at the Commission’s legal service), ‘In Search of the Limits to Article 30 of the EEC Treaty’ (1989) 26 Common Market Law Review 235. 16 See, eg, White, above n 15; Gormley, ‘Recent case law on the free movement of goods’, above n 3; Arnull, above n 3; Mortelmans, above n 3. 17 See the constant decrease in preliminary references in relation to the definition of an MEE postKeck. This is not to say however that the Keck ruling was necessarily welcomed; see, eg, AG Jacobs’ opinion, above n 4; Weatherill, ‘After Keck, above n 4; Barnard, above n 4; Nic Shuibhne, above n 4.

100  Eleanor Spaventa And yet, unfortunately, since the Hünermund case concerned advertising, these types of rule were included in the ‘certain selling arrangements’ category which in principle is excluded from the reach of Article 34 TFEU. Yet, and it hardly bears stressing, advertising is a crucial component of market penetration – indeed it might be less costly for manufacturers of branded products to stick a label on, than to devise different advertising strategies for each of the Member States.18 This of course created a problem for the Court after Keck – to limit the reach of Article 34 TFEU to directly or indirectly discriminatory rules also in relation to advertising would potentially deprive traders of a fundamental tool to gain market share; and yet to rethink Keck would pose its own problems in terms of both consistency and credibility. The solution to this conundrum has been rather confused: a very broad interpretation of indirect discrimination, which did not require the trader to discharge any burden of proof and, in extreme cases, ignored concrete evidence as to the neutral effects of the advertising restrictions,19 coupled with references to market access. The combination of the two then lead to a new, and more specific form of discrimination – market access discrimination:20 here, the Court has decided that domestic products inherently enjoy some advantages both in relation to being more familiar to consumers,21 and in relation to access to a retail network.22 For this reason, there is a further presumption that certain rules affect the ability of imported products to access the domestic market in a more severe way than they do domestic products, which, as mentioned above, have inherent advantages simply by reason of being home-produced. As a result, a total ban on advertising is always considered inherently indirectly discriminatory since consumers are ‘naturally’ more familiar with domestic products, whereas partial advertising bans are assessed on their merits, depending on the form that the restriction takes.23 Similarly, restrictions on long-distance sales, now most commonly internet sales, are considered both as indirectly discriminatory and a restriction on market access, the reasoning being that out-of-state traders might benefit particularly from modalities of sale that do not require a fixed or permanent infrastructure.24 18 See, eg, Case C-239/02 Douwe Egberts NV v Westrom Pharma NV, Christophe Souranis EU:C:2004:445, esp para 52. 19 Case C-405/98 Konsumentombudsmannen (KO) v Gourmet International Products EU:C:2001:135; advertising is also always a service so rules on advertising even if and when they are not caught by Art 34 TFEU are caught by Art 56 TFEU. In the context of services, restrictions on advertising of services is also caught by Art 56 TFEU, see for instance Case C-339/15 Luc Vanderborght EU:C:2017:335. 20 See also Case C-148/15 Deutsche Parkinson Vereinigung eV, above n 11, where fixed price for medicinal products was found to be a MEE because out of state pharmacies selling by mail order relied on price competition as a market penetration tool more than traditional pharmacies which rely on providing advice; the rules were not examined under the CSA category. 21 See the presumption in Case C-405/98 Konsumentombudsmannen (KO), above n 19. 22 Case C-322/01 Deutscher Apothekerverband v 0800 DocMorris NV and Jacques Waterval EU:C:2003:664; Case C-108/09 Ker-Optika bt v ÀNTSZ Dél-dunántúli Regionális Intézete EU:C: 2010:725. 23 Case C-405/98 Konsumentombudsmannen (KO), above n 19. 24 Case C-322/01 Deutscher Apothekerverband, above n 22.

The Free Movement of Goods  101 Overall, and although there is no clear test, it is not so difficult to discern which of those rules falls within the scope of Article 34 TFEU; in practice, the Court balances the restrictive effect of the rule on trade (rather than cross-border trade) vis-à-vis other interests, including the interest in maintaining national regulatory sovereignty, so that whether a selling arrangement is caught or not depends on where that balance is struck. Furthermore, even if the rule at issue in a given case is not necessarily problematic, if the ‘genus’ of rules might affect the internal market in the future (such as is the case with advertising, internet or long-distance selling rules) then it will be brought within the scope of Article 34 TFEU so that the door is left open for a case-by-case assessment of different regulatory frameworks of similar types of rules. In this respect, it could be wondered whether we should revisit the distinction, drawn before Keck by Mortelmans, between static and dynamic rules,25 the latter, like rules on internet sales inherently of interest to the internal market. In any event, and even before Commission v Italy, the case law on Article 34 TFEU can be understood as introducing a system of presumptions which clarify the reach of the Dassonville formula. In Commission v Italy, and Mickelsson and Roos,26 the Court adds another category of rules, those regulating the ‘use’ of a product. It might be recalled that Commission v Italy concerned rules restricting the towing of trailers only to motor vehicles, to the exclusion of motor bikes and the like, whereas Mickelsson and Roos related to rules restricting the use of personal watercrafts to certain waterways. The Court could have simply held those rules as falling within the scope of the Dassonville formula; instead it decided, after confirming the Keck dichotomy, to use a different formula to the effect that the rules at issue fell within the scope of Article 34 TFEU because they constituted a barrier to market access. And overall, the idea behind creating a separate category, consistent with the system of presumptions, is that it alerts national courts to the potential restrictive effects of certain rules: there cannot be any doubt, in this respect, that there is little point importing and/or selling a product if it cannot be used by consumers. Rules prohibiting use are as akin as quantitative restrictions as rules prohibiting sales27 since they prevent a market from existing altogether: it is not so surprising then that they should be justified. All in all, the case law on Article 34 TFEU with all its nuances is better understood, in the writer’s opinion, as establishing a system of rebuttable and nonrebuttable presumptions. Thus: • Directly discriminatory rules are always caught by Articles 34 (and 35) TFEU and justifiable (usually) only pursuant to Article 36 TFEU.28 25 Mortelmans, above n 3. 26 Case C-110/05 Commission v Italy (Moped trailers), above n 5; and Case C-142/05 Åklagaren v Percy Mickelsson and Joakim Roos, above n 5. 27 See, eg, Case C-293/94 Brandsma EU:C:1996:1996:254. 28 Some confusion arises in relation to environmental protection which is obviously a legitimate public interest but which is not included in the exhaustive list of Art 36 TFEU; see, eg, Case C-379/98 PreussenElektra EU:C:2001:160; and Case C-573/12 Vindkraft AB EU:C:2014:2037.

102  Eleanor Spaventa • Product requirements (all rules concerning the physical qualities of a product) are caught by Article 34 TFEU as there is an absolute presumption that they affect intra-Community trade, and must be justified pursuant to mandatory requirements of public interest and/or Article 36 TFEU.29 This means that it is not open to a Member State to rebut the presumption of effect on intercommunity trade. • Certain selling arrangements on the other hand benefit from a rebuttable presumption of compatibility with Article 34 TFEU, unless they are deemed directly or indirectly discriminatory. Crucially no ‘hard evidence’ is required, leaving some interpretative leeway to the judiciary in assessing the restrictive effect of the rule before deciding whether to declare those indirectly discriminatory and in need of justification. Furthermore, what is relevant in those cases is not so much discrimination on grounds of place of production of the good in question, but rather market access discrimination. • Rules banning advertising and rules regulating long-distance sales are presumed to have an effect on intra-EU trade and must be justified. • Rules prohibiting the use of a product must be justified because they affect market access; rules regulating use must be satisfied if they have an effect on market access. • All other rules that hinder directly or indirectly, actually or potentially intraEU trade, or rules that affect market access, need to be justified.30

B.  The Reach of Article 35 TFEU Before turning our attention to the more interesting developments of the last decade or so, and for the sake of completeness, we should briefly recall the extent to which barriers to the export of products are caught by the Treaty, stressing how (and why) the reach of the two provisions is specular rather than similar. In this respect it is useful first to recall some of the theories that explain the reach of the free movement of goods: in particular, a very helpful conceptualisation of the free movement of goods is that put forward by Bernard.31 He argued that Keck established a presumption of repartition of regulatory competences so that, lacking EU harmonising rules, the State where the product is manufactured has full regulatory competence over the manufacturing process, whereas the State where the product is sold has full regulatory competence regarding the modalities of sale. Leaving aside

29 Public health for instance is both a mandatory requirement of public interest and a legitimate ground of derogation under the TFEU. 30 Spaventa, ‘Leaving Keck Behind?, above n 5. 31 N Bernard, ‘La libre circulation des marchandises, des personnes et des services dans le Traité CE sous l’angle de la compétence’ (1998) 33 Cahiers de droit eropéen 11; and N Bernard, Multi Level Governance in the European Union (London, Kluwer Law International, 2002) esp ch 2.

The Free Movement of Goods  103 the nuances in the more recent case law outlined above, this model is still useful to understand the different reach of Article 35 TFEU. Here, if the Member State of production has regulatory competence, it is clear that its manufacturing rules should not be subjected to the proportionality assessment in the same way as they would if they were imposed by the Member State of import. Or else, the effect of the internal market would be to subject all rules to scrutiny, risking regulatory vacuums and significantly eroding regulatory diversity amongst the Member States. For this reason, the Court limited the reach of Article 35 TFEU only to those rules which would give an advantage to domestic products to the detriment of goods for export.32 And, of course, those rules are not very easy to come by as Member States have very little incentive, in normal times, to reduce their exports.33 More recently, however, the Court has rightly expanded the scope of Article 35 TFEU so that an advantage to domestic production is no longer a prerequisite for the rule to be construed as a barrier to exports (at least for rules other than product requirements). Thus, for instance in New Valmar34 the case related to a contract between an Italian and Belgian manufacturer for the supply of toys; the contract was regulated by Italian law and the forum was Belgian. The Italian company failed to pay for the goods and alleged that the invoices sent to it in Italian were null and void under Belgian law which provided for all invoices drawn by companies established in the Flemish speaking region in Belgium to be written in Dutch. The Court found that the rules at issue constituted a restriction pursuant to Article 35 TFEU on the grounds that they were more likely to affect traders in other Member States since they would be less likely to understand Dutch.35 There was no mention 32 Case 15/79 Groenveld, above n 6; Case C-205/07 Lodewijk Gysbrechts EU:C:2008730; and see more recently Case C-169/17 Asociación Nacional de Productores de Ganado Porcino v Administración del Estado EU:C:2018:440. 33 Although of course this is not always true; see for instance opinion in Case C-648/18 Autoritatea Naţională de Reglementare în Domeniul Energiei (ANRE) v Societatea de Producere a Energiei Electrice în Hidrocentrale Hidroelectrica SA EU:C:2020:256, judgment still pending at the time of writing; see also the export restrictions imposed by some Member States on sanitary equipment during the COVID-19 crisis and the Commission’s response to allow Member States to block the export of certain sanitary products outside the EU, see Commission Implementing Regulation (EU) 2020/402 of 14 March 2020 making the exportation of certain products subject to the production of an export authorization [2020] OJ L771/1. The Commission’s strategy was therefore extremely EU centric (keep goods flowing within accept restrictions towards outside) and, whereas perhaps politically justified, has been criticised both on moral and on practical grounds, see, eg, CP Bown, ‘EU limits on medical gear exports put poor countries and Europeans at risk’ Peterson Institute for International Economics, available at: www.piie. com/blogs/trade-and-investment-policy-watch/eu-limits-medical-gear-exports-put-poor-countriesand; and André Sapir, ‘What the EU should do and not do on trade in medical equipment’ (Bruegel, 25 March 2020), available at: www.bruegel.org/2020/03/what-the-eu-should-do-and-not-do-on-tradein-medical-equipment/ has to a certain extent paid off and France and Germany have lifted (at the time of writing) restrictions; it is open to debate whether the latter would be justified under Art 36 TFEU; see recently Case C-222/18 VIPA Kereskedelmi és Szolgáltató Kft v Országos Gyógyszerészeti és Élelmezésegészségügyi Intézet EU:C:2019:751. 34 Case C-15/15 New Valmar BVBA v Global Pharmacies Partner Health Srl EU:C:2016:464. 35 Language requirements have always been considered as falling within the scope of internal market rules; see in relation to labelling for example; and in in relation to rules similar to those at issue in New Valmar in the context of employment contracts see Case C-202/11 Las EU:C:2013:239.

104  Eleanor Spaventa of the need to establish an advantage to domestic traders: and in fact, it is only in relation to product requirements that the need to establish an advantage to domestic goods should be proven, exactly because of what was mentioned above in relation to repartition of regulatory competences. In relation to other rules what should matter is the existence of a barrier to exports, without the need to prove discrimination. Here, just consider a ban on internet sales, which would clearly have a significant effect on the export of goods regardless of any discrimination.36

III.  The Frontiers of the Free Movement of Goods: Duty to Act and Good Administration We have given a brief outline of the scope of Articles 34 and 35 TFEU in their more traditional sense, ie, when rules imposed by a Member State on imported, exported or transit goods amount to a measure having equivalent effect to a quantitative restriction. But the reach of the free movement of goods provisions is broader: in particular, those provisions might also impose on Member States a duty to act, as well as constraints on the way new rules are adopted. We will examine the two in turn. The introduction of a positive duty to act to safeguard the free movement provisions arises from the now distant Commission v France;37 there the Court held that France had breached Article 34 TFEU because of its (wilful) failure to protect importers from the unlawful action of French farmers seeking to prevent Spanish strawberries from reaching French markets. The interpretation in Commission v France was not altogether surprising given the background to the case: the unlawful action of the French farmers had been protracted in time and France had chosen to compensate the Spanish importers rather than prosecute the unlawful actions of the farmers. Furthermore, it is an established principle in fundamental rights law that the State’s failure to protect rights might trigger its responsibility. However, the case also sparked some worries amongst prescient Member States that such an interpretation could affect the protection of fundamental rights, and especially freedom of expression and freedom to strike, by imposing on national authorities a duty to act to prevent those exercising a fundamental right from interfering with Articles 34 and 35 TFEU. It is sufficient in this respect to remember

36 In the context of services see Case C-384/93 Alpine Investments EU:C:1995:126 where the Court accepted that extra-territorial rules disciplining the way the service was also to be provided in Member States other than that of establishment fell within the scope of Art 56 TFEU. 37 Case C-265/95 Commission v France (riots) EU:C:1997:595; the ruling was then ‘codified’ in Regulation 2679/98 on the functioning of the internal market in relation to the free movement of goods among the Member States [1998] OJ L337/8, esp Art 1(2); crucially the Regulation also provides that the Regulation cannot be interpreted as affecting fundamental rights recognised in Member States, including the freedom to strike. However, and strangely, this Regulation was never referred to by the Court in subsequent case law and especially in the case of Schmidberger, below n 39.

The Free Movement of Goods  105 that, inevitably, strike action has an impact on the economic performance of the employer: in the case of manufacturing this might well impact on the possibility to export products or import components, hence engaging the Treaty free movement of goods provisions. In order to address this problem, and so as to balance competing interests, the Council adopted Regulation 2679/98 which on the one hand codified the principle pursuant to which State inactivity might determine an obstacle to the free movement of goods pursuant to the Treaty, but on the other provided that the Regulation itself could not be interpreted as ‘affecting in any way the exercise of fundamental rights as recognised in Member States, including the right or freedom to strike’.38 The Regulation set out procedural requirements, and yet it immediately identified the potential conflict between national constitutional rights and the free movement provisions. Furthermore, the Regulation clearly privileged the former over the latter, pre-skewing the balance in favour of national (not European) fundamental rights. This notwithstanding, in Schmidberger39 the Court took a different approach, one that eventually led to the more contentious Viking and Laval rulings.40 In particular, the Court accepted that failure by the Austrian authorities to stop a demonstration on the Brenner motorway could be construed as a barrier to the free movement of goods. Whereas the barrier was justified, the Schmidberger case constitutes a turning point in internal market law: up until that case, the free movement provisions had the effect of either enhancing fundamental rights by allowing individuals to invoke EU fundamental rights in relation to acts of Member States limiting their free movement rights;41 or the free movement provisions had a neutral effect on national rights, by allowing Member States to plead the protection of domestic fundamental rights within the context of the public policy justification.42 In Schmidberger the Court chooses a different interpretative path both by qualifying the exercise of a constitutional fundamental right as a barrier to trade and by justifying said restriction with reference to EU fundamental rights rather than domestic constitutional rights; in so doing, the balancing between conflicting rights and the ensuing proportionality assessment are brought within the Court of Justice’s jurisdiction.43 Furthermore, and not surprisingly given the path chosen, there is no reference to Regulation 2679/98: as mentioned above that Regulation refers to domestic fundamental rights rather than EU fundamental rights, indicating a hierarchy of values whereby (non-economic) domestic

38 Regulation 2679/98, above n 37, Art 2, emphasis added. 39 Case C-112/00 Schmidberger EU:C:2003:333; see also Case C-573/12 Vindkraft AB, above n 28, para 74. 40 Case C-438/05 International Transport Workers’ Federation and Finnish Seamen’s Union v Viking Line EU:C:2007:772; Case C-341/05 Laval un Partneri EU:C:2007:809. 41 Case C-260/89 ERT EU:C:1991:254. 42 Case C-36/02 Omega EU:C:2004:614. 43 E Spaventa, ‘Federalisation Versus Centralisation: Tensions in Fundamental Rights Discourse in the European Union’ in M Dougan and S Currie (eds), 50 Years of the European Treaties (Oxford, Hart Publishing, 2009).

106  Eleanor Spaventa fundamental rights should not be sacrificed on the altar of the free movement of goods. Such an approach however would have been inconsistent with established principles elaborated by the Court, especially in relation to the supremacy of EU law even vis-à-vis constitutional rights. The second important development also arises from the imposition of positive duties upon Member States, this time in their capacity as lawmakers:44 in Commission v Germany,45 the Court found that an otherwise legitimate rule, aimed at switching beverages from disposable to reusable bottles, was disproportionate since it did not afford a sufficient transitional period to enable producers to adapt to legislative changes. This interpretation then led to a more politically sensitive line of cases, in relation once again to the Brenner motorway.46 Here Austria was aiming to switch transport of goods from wheels to rail in order to reduce pollution, which in some areas was way above EU permitted levels, being therefore inconsistent with both European law and EU international commitments to reduce greenhouse emissions. However, and this was the Commission’s concern, the Brenner motorway is the main corridor for the transport of goods from southern Europe to the north. In the first case, the Commission also successfully challenged the new rules on the grounds of the very short transition period (two months) which was ‘clearly insufficient to allow the operators concerned to adapt to the new circumstances’.47 In the follow-up case, the Commission challenged the amended rules which ranged from variable speed limits to night travel bans and traffic bans for Euro II and III lorries. As in the first case, the Austrian government was under an EU law duty to reduce emissions, and it was universally accepted that those measures were in principle justifiable on environmental protection grounds. What was at issue then was the proportionality of the measures: here, and without going into the details of the case, what is interesting is the fact that in order to prove that the measures were disproportionate, the Commission proposed an alternative regulatory framework to that adopted by Austria and that, in accepting it, the Court endorsed the rather simplistic view according to which when regulating there is one single interest to be taken into account and an optimal and superior regulatory framework. This said, with those cases the circle is completed so that the free movement provisions impose a duty to abstain from enacting quantitative restrictions and measures having equivalent effect (however defined); a duty to protect the effective enjoyment of the free movement rights; and a duty to take into due consideration the impact on EU right-holders of otherwise legitimate rules, and to minimise disruption and facilitate transition to new regulatory frameworks. In this way

44 In the field of services see Case C-98/14 Berlington Hungary Tanácsadó és Szolgáltató kft EU:C:2015:386. 45 Case C-463/01 Commission v Germany (reusable packaging) EU:C:2004:797. 46 Case C-320/03 Commission v Austria (ban on heavy lorries) EU:C:2005:684; and Case C-28/09 Commission v Austria EU:C:2011:854. 47 Case C-320/03 Commission v Austria, n 46 above (ban on heavy lorries), para 90.

The Free Movement of Goods  107 then, the free movement of goods provisions (and the other internal market rights) are framed as constitutional fundamental rights, and benefit from the hermeneutic tools developed in that context. Whereas at first sight this approach might seem entirely coherent and reasonable, it is not without its conceptual and practical problems: conceptually because it steers decision-making towards technocraticism and away from the reality of political compromise, contributing to the narrative, dear to the Court and to a certain extent characterising the European integration project at many levels, that the choice between different regulatory frameworks is technical rather than inherently political. Practically because it overestimates the possibility to carry out extensive impact assessments, and it curtails policy experimentation. Furthermore, and as noted by Weatherill in the context of justifications,48 the Court holds an optimistic view of the overall ability of the legislature to ensure the coherence of regulatory frameworks, a view which fails to take into account the reality of democratic systems which are dynamic in nature, with changing constituencies and priorities. Closely linked to the extension of the reach of Articles 34 and 35 TFEU to affect the way national authorities adopt regulatory frameworks is a more recent problem, brought about by the Scotch Whisky case.49 It might be recalled that this case related to the Scottish government’s then proposed legislation providing minimum price per alcohol unit. The aim of the new framework, which affects disproportionately lower-priced alcoholic beverages, was to reduce alcohol-related deaths and health problems caused by excessive alcohol consumption; the issue in the case was whether the rules were proportionate to the stated aim or whether there was a less intrusive way to achieve the same aim. In particular, the Court found that taxation would be, in principle, a less restrictive way to achieve the same aim since it would not have an impact on the formation of prices, whilst leaving the final assessment to the national court. What is interesting in relation to the case before the Court of Justice is the fact that all parties agreed that, for the purposes of the case, and in line with established case law, Scotland was to be treated as if it were the United Kingdom. This approach is entirely consistent with international law practice: however, the same approach is more debatable when transferred point blank to the realm of the proportionality assessment, especially when the latter includes a detailed examination of regulatory alternatives. In particular, in the Scotch Whisky case the issue, one which was described by the national court as the ‘elephant in the room’,50 was that whereas the Scottish government had devolved jurisdiction in the field of health, it did not have competence in 48 S Weatherill, ‘Justification, Proportionality and Consumer Protection’ in P Koutrakos, N Nic Shuibhne and P Syrpis (eds), Exceptions from EU Free Movement Law (Oxford, Hart Publishing, 2016) 236, esp section VII. 49 Case C-333/14 The Scotch Whisky Association and Others v The Lord Advocate and The Advocate General for Scotland EU:C:2015:845. 50 The Scotch Whisky Association and Others v The Lord Advocate, Court of Session [2016] CSIH 77, available at: www.scotcourts.gov.uk/search-judgments/judgment?id=9a1821a7-8980-69d2-b500ff0000d74aa7.

108  Eleanor Spaventa the field of taxation. The assessment of regulatory alternatives therefore became an exercise in legal fiction – the measures examined by the Court of Justice as being suitable regulatory alternatives were measures that the regulator could not in fact adopt. As said above, the Court of Justice left it to the national court to determine the compatibility of the rules with Article 34 TFEU and eventually the Supreme Court found in favour of the Scottish government and the minimum unit pricing rules have been in force for some time;51 they also had some effect in reducing alcohol consumption. Yet, the problem remains: how should we treat devolved administrations in those instances? If on the one hand it is crucial that the free movement provisions, and the Treaties more generally, apply vertically to all emanations of the State, some thoughts should be given to ensuring that such application does not upset the constitutional balance of competence established at national level. This, in the writer’s opinion means that the policy alternatives available to devolved administrations should be assessed in relation to the devolved administrations’ actual competences so that if the rules pursue a legitimate aim and are neither protectionist nor discriminatory, then they should be assessed only in relation to alternatives available to that policymaker.

IV.  Concluding Remarks The internal market has naturally evolved in the past 60 or so years: there are new challenges, old problems and some considerable progress. There can hardly be any doubt that the ‘classic freedom’, Article 34 TFEU, is the framework within which the Court was able to develop and refine its internal market jurisprudence. Whilst there is no denying that nowadays the primary role has been taken by the free movement of services, the tools elaborated within the context of the free movement of goods are still useful to understand the internal market and the possible different interpretative avenues open to the Court. And, whereas it is true that the case law has pretty much settled, this does not mean that there are no challenges ahead: in particular, and as highlighted above, the extensive duties imposed on lawmakers in choosing the correct regulatory policy are perhaps somehow abstract, not taking into consideration the fact that regulation is necessarily an outcome of political choices which should be given the correct weight by the judiciary. This does not mean that the judiciary should not keep a watchful eye, or that the role of judicial review should be curtailed. Rather it means acknowledging that rule-making is not a neutral exercise, that it is deeply affected by a series of considerations, that democracy is by definition, and at its best, a complex exercise which has to take into account a multitude of interests and preferences so that



51 The

Scotch Whisky Association and Others v The Lord Advocate (2017) UKSC 76.

The Free Movement of Goods  109 often the end result is not necessarily the ‘first best’ (determination of which any event entails a political assessment). Furthermore, in evaluating policy alternatives the judiciary should take care in not stifling regulatory experimentation since, of course, the effect of existing and established regulatory choices can be ‘measured’ whereas the effect of new regulatory avenues can only be speculated upon. More generally the considerations rehearsed above are of relevance for a broader reflection on the present challenges to international trade and globalisation from both the political left and right: from the left, a deep scepticism towards a technocratic model which is often seen to pursue deregulation for deregulation’s sake at the expense of other competing interests. From the populist right, a deep scepticism of ‘international’ cooperation and a preference for nationalist/ sovereignty narratives. And those challenges are only going to increase following two unprecedented shocks: Brexit and the COVID-19 crisis.

110

6 The ‘Social Freedom’? The Free Movement of Persons in EU27 NIAMH NIC SHUIBHNE

I. Introduction In EU free movement law, the extent to which the objective of constructing an effective internal market supports an appropriate, never mind ambitious, level of social protection provokes intensive analysis – and concern. Adjudicating between the freedom extended to service providers, on the one hand, and the protection of posted as well as local workers in that context, on the other, provides the archetypal example of the register of conflict that typifies the market/social enquiry. The judgments of the Court of Justice in Viking Line and Laval, in particular, represent challenges attributable in part to the structure of free movement law – which asks that restrictions of the Treaty freedoms be justified on public interest grounds – but also reflective of working out the values that matter (the most) in the shaping of the internal market.1 More recently, there are some indications that meaningful integration of internal market and social goals is not just wanted but also possible.2 This chapter is

1 Case C-438/05 Viking Line EU:C:2007:772; and Case C-341/05 Laval EU:C:2007:809. These judgments also raised but did not explore questions about how EU standards intersect with those set by other actors, eg, on ILO standards, see S Garben, ‘The European Pillar of Social Rights: An Assessment of its Meaning and Significance’ (2019) 21 Cambridge Yearbook of European Legal Studies 101, 119–22. 2 eg, in 2016, the Commission published a proposal (COM(2016) 128 final) to amend Directive 96/71 on the posting of workers in the framework of the provision of services ([1997] OJ L18/1); Directive 2018/977 was adopted on 28 June 2018 ([2018] OJ L173/16) and must be transposed into national law by 30 July 2020. See further, the contribution to this volume by B de Witte (ch 7). There has been more mixed reaction to the broader ‘balancing’ exercise articulated by the Court of Justice in Case C-201/15 AGET Iraklis EU:C:2016:972, paras 71–78, including the statement at para 77 that ‘[s]ince the European Union thus has not only an economic but also a social purpose, the rights under the provisions of the Treaty on the free movement of goods, persons, services and capital must be balanced against the objectives pursued by social policy’. For analysis of these developments, see, eg, ACL Davies, ‘How has the Court of Justice changed its management and approach towards the social acquis?’ (2018) 14 European Constitutional Law Review 154; and D Schiek, ‘Towards More Resilience for a Social EU – the Constitutionally Conditioned Internal Market’ (2017) 13 European Constitutional Law Review 611.

112  Niamh Nic Shuibhne rooted in the same impulse for better integration of different objectives but it reverses the premises of the central question, ie, rather than asking if or to what extent social objectives are properly accommodated by EU free movement law, has the development of EU social policy properly accommodated the distinctive free movement-shaped dimension of the EU’s internal market? It is argued that asking the question in this way still exposes a problematically non-integrated set of objectives, but that weakness is characterised and sustained in this ‘reverse’ context by a dynamic of disconnect rather than conflict. The chapter first outlines the fused nature of economic and general free movement rights when defining the scope of ‘persons’ for the purposes of EU internal market law (section II), using the examples of underpinning foundations, legislative frameworks, and concepts and principles of interpretation. In that sense, the fact or prospect of cross-border mobility – rather than any of its purposes more specifically – defines the distinctiveness of the internal market. At the same time, however, sharp legal distinctions attach to the activity that is then undertaken (or not) in the host State and these distinctions can cause the lines of social protection to be redrawn. In section III, the example of indexing exported family benefits is then used to contemplate whether the withdrawal of the UK from the European Union would shift the coordinates of the market/social debate in any important respects. Less positively, the indexation example reinforces an instrumental or political quality of social protection in free movement law; more positively, however, it shows that legally questionable political choices can be corrected. In section IV, the drafting, proclamation and implementation of the European Pillar of Social Rights (EPSR) are briefly presented. The EPSR institutes a significant new phase in EU social policy, but it is undeniably ‘mobility light’ in terms of both its vision and its substance. In particular, it will be shown that outcomes or situations produced precisely because of the exercise of free movement rights are not acknowledged within the framework of the EPSR. It must be recognised that a more overtly mobility-attuned social policy risks entrenching the view that economic and social interests inevitably come into conflict, and that when they do, economic interests inevitably prevail. It raises constitutional complexities that coalesce around uneven competence distribution for market and social regulation respectively. It also raises normative questions that connect not just to the accommodation but also to the promotion of free movement as the overriding internal market ‘good’. However, the perspective focused on here is more functional: in essence, can EU social policy really be effective if it overlooks the effects and implications of free movement, of the EU-bespoke dimension of the internal market around which social policy should place a critical frame of protection? The example of retaining the status of worker in a host Member State is used in section IV to illustrate more concretely why disconnecting rather than integrating economic and social policy threads is sub-optimal. That discussion confirms that long-standing systemic vulnerabilities of EU social policy – notably, asymmetries of competence and a reluctance to assign responsibility for action – were not reconceived

Free Movement of Persons: The ‘Social Freedom’?  113 through the adoption of the Pillar, therefore most likely prolonging disconnect between the social and the mobility dimensions of the internal market.3 Echoing the view of the Commission, the chapter therefore concludes that transformative progress – exhibiting more meaningful integration of the internal market’s social and economic dimensions – ultimately ‘requires more political courage and determination than 25 years ago, and greater efforts than ever to close the gap between rhetoric and delivery’.4

II.  The ‘Free Movement of Persons’ and the EU Internal Market ‘Who’ are we talking about when we address the free movement of ‘persons’ in the context of the EU’s internal market? In its 2018 Communication on ‘The Single Market in a changing world’, the Commission observed that 17 million Union citizens live or work in another Member State but that just over half (9.5 million) of them are economically active there; it then explained that: The intra-Union mobility of people has significantly increased over the last decade … In addition, since the start of the Erasmus Programme, over 9 million citizens have been able to spend time in another country to study, to learn or to train. While progress in these areas has been impressive, the figures remain low for a continent of more than 512 million inhabitants. They should be seen in the context of specific factors such as language and differences in social systems which are difficult to overcome and which mean that labour mobility is always likely to remain lower in the Union than in other integrated markets.5

The gap between the relatively low extent to which what we might think of as more ‘intensively’ exercised free movement rights – ie, excluding temporary or virtual movement for the provision or receipt of services, for example6 – are exercised in practice and the perception that such freedom is exercised at or close to overwhelming levels is returned to in section III below, in the context of the UK’s withdrawal from the European Union. For present purposes, the key point 3 Reluctant engagement with the social rights protected by the Charter of Fundamental Rights provides a related but distinct forum of disconnect; see S Robin-Olivier, ‘Fundamental Rights as a New Frame: Displacing the Acquis’ (2018) 14 European Constitutional Law Review 96; including the suggestion that ‘[t]he Court of Justice seems more at ease forcing Member States to respect even extensively construed provisions of a directive than it is with wielding fundamental rights with uncertain force and content, and seems furthermore to be held back by the reluctance, if not frontal opposition, expressed by some of the Member States’ (104). 4 European Commission Communication, ‘The Single Market in a Changing world: A unique asset in need of renewed political commitment’ (COM(2018) 772 final, 1. 5 COM(2018) 772 final, 3. 6 Though the Commission does note that ‘[a]round 2 million citizens are cross-border daily commuters who work or study in one country but live in another’ (ibid, referring to ‘Boosting growth and cohesion in EU border regions’ COM(2017) 534).

114  Niamh Nic Shuibhne concerns the Commission’s concurrent reflection on economically active and economically inactive Union citizens in its framing of the internal market. This point is not about whether Union citizenship is a form of market citizenship or similar genealogical debates; but, more functionally, about recognising the contribution to the market that all mobile Union citizens make – irrespective of the purpose or purposes for which they moved in the first place. EU free movement law exhibits the same sense that rights attached to different purposes are, at once, both different and entangled. At one level, the free movement rights conferred on all Union citizens by Article 21 of the Treaty on the Functioning of the European Union (TFEU) are conceptually and functionally distinct from the free movement rights associated with the economic freedoms (Articles 45, 49 and 56 TFEU). Union citizenship is considered to represent a considerable qualitative step forward in that it separates that freedom from its functional or instrumental elements (the link with an economic activity or attainment of the internal market) and raises it to the level of a genuinely independent right inherent in the political status of the citizens of the Union.7

That idea is captured by the method used by the Court of Justice to determine which Treaty provision applies in a given situation, ie, the right to move and reside conferred on Union citizens by Article 21 TFEU is considered to find ‘specific expression’ in Articles 45,8 499 and 5610 TFEU. Thus, while a Member State national who resides in another Member State may be present there on several parallel premises – for example, someone may be working while studying, and also the family member of another Union citizen residing there – it is not necessary to consider the interpretation of Article 21 where the facts of the dispute fall within the scope of one of the economic freedoms.11 The rights conferred by Article 21 TFEU are therefore residual in nature, which means, conversely, that the rights conferred by the economic freedoms are assessed first.12 In reality, though, the relationship between citizenship-based or general free movement rights and rights based on the Treaty’s economic freedoms is messier and more complex. In other words, even for internal market purposes, economic and general free movement rights cannot always be neatly dissociated: rather, their essential premises are broadly fused. This point can be illustrated by considering

7 AG Ruiz-Jarabo Colomer in Case C-386/02 Baldinger EU:C:2003:671, para 25 of the opinion. 8 eg, Case C-100/01 Oteiza Olazabal EU:C:2002:712, para 26. 9 eg, Case C-193/94 Skanavi and Chryssanthakopoulos EU:C:1996:70, para 22. 10 eg, Case C-76/05 Schwarz and Gootjes-Schwarz EU:C:2007:492, para 34. 11 eg, Case C-420/15 U EU:C:2017:408, paras 13–19. For a rare exception, analysing Art 21 TFEU ‘irrespective of whether the person concerned engaged in an economic activity’ (in this case, the provision of services under Art 56 TFEU), see Case C-208/09 Sayn-Wittgenstein EU:C:2010:806, paras 39–42. 12 eg, Case C-520/04 Turpeinen EU:C:2006:703, para 13; Case C-367/11 Prete EU:C:2012:668, para 20. In preliminary reference proceedings under Art 267 TFEU, the Court has recast the referring court’s question(s) to frame the dispute around the correct Treaty provision where this has been deemed necessary; see, eg, Case C-632/13 Hirvonen EU:C:2015:765, paras 20–22.

Free Movement of Persons: The ‘Social Freedom’?  115 the examples of legal foundations, legislative frameworks, and concepts and principles of interpretation. First, on fused legal foundations, the right to move and reside is the most developed legal dimension of Union citizenship. It is therefore not surprising that the free movement of persons in a broader sense is drawn from the shaping of citizenship mobility rights, and vice versa. For example, Articles 20 and 21 TFEU refer to conditions and limits ‘defined by the Treaties’ (Article 20) or ‘laid down in the Treaties’, the most basic example of which concerns the limitation of free movement on the grounds of public policy, public security and public health – all of which are provided for explicitly in the Treaty in the context of the economic freedoms. Further reflecting this pattern of fusion, Recital 2 of Directive 2004/38 states that ‘[t]he free movement of persons constitutes one of the fundamental freedoms of the internal market’. In its Communication on the Council’s common position on the draft Directive, the Commission had considered that ‘’[t]he free movement of citizens constitutes one of the fundamental freedoms of the internal market and is at the heart of the European project’.13 The Court adopted that interpretation when it stated that citizenship of the Union confers on each citizen a primary and individual right to move and reside freely within the territory of the Member States, subject to the limitations and restrictions laid down by the Treaty on the functioning of the European Union and the measures adopted for their implementation, freedom of movement for persons being, moreover, one of the fundamental freedoms of the internal market, which was also reaffirmed in Article 45 of the Charter of Fundamental Rights of the European Union.14

Similarly, in Metock, when rejecting a requirement of prior lawful residence in a Member State for the family members of Union citizens, the Court based its reasoning in part on the fact that such a requirement ‘would not be compatible with the objective set out in Article 3(1)(c) EC of an internal market characterised by the abolition, as between Member States, of obstacles to the free movement of persons’, confirming that ‘[e]stablishing an internal market implies that the conditions of entry and residence of a Union citizen in a Member State whose nationality he does not possess are the same in all the Member States’.15 In a different context, the challenges of unpicking the market-rooted and citizenship-rooted dimensions of the free movement of persons are apparent in the application and interpretation of free movement concepts for the context of the EEA.16 13 ‘Communication from the Commission to the European Parliament and the Council on guidance for better transposition and application of Directive 2004/38/EC on the right of citizens of the Union and their family members to move and reside freely within the territory of the Member States’ COM(2009) 313 final, 3 (emphasis added). 14 Case C-162/09 Lassal EU:C:2010:592, para 29 (emphasis added). See similarly, Case C-434/09 McCarthy EU:C:2011:277, para 27. 15 Case C-127/08 Metock and Others EU:C:2008:449, para 68 (emphasis added). 16 See further, H Haukeland Fredriksen and CNK Franklin, ‘Of pragmatism and principles: the EEA Agreement 20 years on’ (2015) 52 Common Market Law Review 629, esp section 2.2; and N Rennuy and P van Elsuwege, ‘Integration without membership and the dynamic development of EU law: United Kingdom v Council (EEA) (2014) 51 Common Market Law Review 935, esp section 5.3.

116  Niamh Nic Shuibhne Second, considering fused legislative frameworks, Directive 2004/3817 repealed and replaced almost all the legislation that dealt with free movement rights on a sectoral basis, with the objective of establishing a single system applicable to any free movement context. However, for the right to reside in another Member State for more than three months, important distinctions between different categories of citizen are retained. In particular, Article 7 of the Directive distinguishes between the economically active – workers and self-employed persons have a right to reside in another Member State with no further conditions attached (Article 7(1)(a)) – and the economically autonomous. For the latter, a general right to move and reside is provided for in Article 7(1)(b) but it is subject to conditions on having sufficient resources to avoid becoming a burden on the social assistance system of the host Member State and comprehensive sickness insurance cover.18 If these conditions are not fulfilled, the citizens in question do not reside in the host State on the basis of the Directive and may not, in consequence, claim equal treatment with host State nationals for matters that fall within the scope of the Treaty.19 In contrast, the economically active do not have to be economically autonomous. In particular, workers and self-employed persons may claim social and tax advantages in the host State on a basis of equal treatment with national workers, which includes a right of recourse to the host State’s social assistance system.20 Moreover, EU workers are required to meet the definition of worker in EU law only,21 which may differ from, and displace, minimum hours or income thresholds that might apply to the determination of work(er) under national rules. The broader significance of an internal market that legally privileges activity over autonomy is returned to below. For now, it should be noted that, notwithstanding the adoption of the Directive, residence rights can still be based on Regulation 492/2011, Article 10 of which confers on the children of migrant workers a right to remain in the host State to complete their education even if the worker no longer resides there.22 Crucially, residence rights based on Article 10 are not subject to the

17 Directive 2004/38/EC on the right of citizens of the Union and their family members to move and reside freely within the territory of the Member States [2004] OJ L158/77. 18 Residence rights for students (Art 7(1)(c)); for the family members of Union citizens who are themselves Member State nationals (Art 7(1)(d)); and for family members who are not nationals of a Member State (Art 7(2)) are also provided for in the Directive. 19 Art 24(1) of Directive 2004/38 confirms that ‘all Union citizens residing on the basis of this Directive in the territory of the host Member State shall enjoy equal treatment with the nationals of that Member State within the scope of the Treaty’; see further, Case C-333/13 Dano EU:C:2014:2358, paras 68–69. Specific exceptions from equal treatment, which concern social assistance (in the first three months of residence and for jobseekers) and maintenance aid for studies, are then established in Art 24(2). 20 eg, Case 249/83 Hoeckx EU:C:1985:139, para 22; Art 7(2) of Regulation 492/2011 on freedom of movement for workers within the Union [2011] OJ L141/1, provides that EU workers ‘shall enjoy the same social and tax advantages as national workers’ in the host State. 21 eg, Case C-46/12 LN EU:C:2013:97, para 47. 22 A residence right for the child’s primary carer is also extended in these circumstances; see, eg, Case C-310/08 Ibrahim EU:C:2010:89, para 50.

Free Movement of Persons: The ‘Social Freedom’?  117 conditions for lawful residence specified in Article 7 of Directive 2004/38.23 But when those conditions are not met during the period of ‘Regulation residence’, the persons concerned are not then eligible for a right of permanent residence under Article 16 of the Directive.24 The conditions for lawful residence in Directive 2004/38 thereby ripple outwards and fuse with residence statuses drawn from other sources of EU law; the same phenomenon is seen when its provisions are applied ‘by analogy’ to residence rights based directly on the citizenship provisions of the Treaty rather than on secondary EU law.25 The conditions in Article 7 have also been imprinted onto residence tests applied for the purposes of social security benefits within the scope of Regulation 883/2004.26 Finally, third, on fused concepts and principles of interpretation, there are multiple instances of cross-fertilisation between economic rights and general rights across the span of EU free movement law. At a general level, the implications of Union citizenship augment the interpretation of the economic freedoms,27 an influence that is usually positive in the sense that Union citizenship, as a legal concept, enhances or expands the rights of the individual concerned. But that is not always the case. For example, Advocate General Wathelet has criticised how ‘genuine link’ analysis, which seeks to measure the sufficiency of an individual’s integration into the host State in order to determine entitlement to claims for financial assistance made there, has been transposed from citizenship law to the free movement of workers.28 And even a ‘positive’ influence is open to criticism on the grounds that appropriating concepts developed for general free movement rights can undermine the integrity of a framework designed specifically for the exercise of economic activity.29 Hybrid concepts assembled from both economic and general free movement rights provide another interesting example of conceptual and interpretative fusion. For example,30 the providers and recipients of services constitute an 23 eg, Case C-480/08 Teixeira EU:C:2010:83, para 70. 24 eg, Case C-529/11 Alarape and Tijani EU:C:2013:290, paras 36–37. Art 16(1) of Directive 2004/38 establishes that ‘Union citizens who have resided legally for a continuous period of five years in the host Member State shall have the right of permanent residence there’. Permanent residence status confers, in particular, extensive equal treatment rights and enhanced protection against expulsion. 25 eg, for family member residence rights where a Union citizen returns to their home State following the exercise of free movement rights see, eg, Case C-673/16 Coman EU:C:2018:385, para 39; Case C-230/17 Deha Altiner and Ravn EU:C:2018:497, para 27. Such rights are based directly on Art 21 TFEU. 26 Case C-308/14 Commission v UK EU:C:2016:436; for criticism of that extension, see M Cousins, ‘The Baseless Fabric of this Vision: EU Citizenship, the Right to Reside and EU Law’ (2016) 23 Journal of Social Security Law 89; C O’Brien, ‘The ECJ Sacrifices EU Citizenship in Vain: Commission v. United Kingdom’ (2017) 54 Common Market Law Review 209. Regulation 883/2004 on the coordination of social security systems [2011] OJ L166/1. 27 eg, Case C-291/05 Eind EU:C:2007:771, para 32, where the Court stated that its interpretation of Art 45 TFEU and related legislation was ‘substantiated by the introduction of the status of citizen of the Union’. 28 AG Wathelet in Case C-238/15 Bragança Linares Verruga and Others EU:ECLI:2016:389, paras 34 ff of the opinion. 29 eg, AG Geelhoed in Case C-212/05 Hartmann EU:C:2006:615, paras 28–41 of the opinion. 30 Other contributors to this volume also address hybrid concepts or statuses: see esp the chapters by B de Witte on posted workers (ch 7) and V Hatzopoulos on ‘dependent contractors’ in the gig economy (ch 10).

118  Niamh Nic Shuibhne exception to the basic distinction established by Article 7 of Directive 2004/38 noted above, in the sense that they do engage in economic activity but are nevertheless required to meet the conditions of economic autonomy in Article 7(1)(b) to secure a right of residence in another State for more than three months. The position of jobseekers is even more complicated. Their right to reside in another State to look for work is based on Article 45 TFEU; rights to equal treatment in the host State are then constructed on a hybrid basis that draws from both Articles 45 and 21 TFEU;31 and their position with respect to fulfilling the conditions for lawful residence in Article 7 of the Directive is overridden by a derogation in Article 14(4)(b).32 Across the various examples of fusion offered above, a linking theme concerns the expected internal market emphasis on mobility – on ensuring the effectiveness of free movement rights irrespective of the purpose or purposes for which the movement itself was undertaken. Where Union citizenship and the facilitation of the internal market mutually reinforce that objective, then the legal foundations, legislative frameworks and/or concepts and principles of interpretation that apply are mutually reinforced in turn. However, that does not mean that the purposes for which movement was undertaken then become flattened or irrelevant. On the contrary, the legal positions of the economically active citizen – who is not required to be economically autonomous – and the economically inactive citizen – who must be economically autonomous – are materially different for the purposes of lawful residence status and equal treatment protection, which are conditioned differently precisely because of perceived respective contributions to the internal market. The mobile jobseeker, who has the potential to make a market contribution through future employment, is the halfway figure in this picture and is, accordingly, protected in a halfway sense: entitled to equal treatment for financial benefits that facilitate access to the host State labour market;33 but not if the ‘predominant function’ of such benefits ‘is in fact to cover the minimum

31 The substantive scope of these rights has narrowed over time: compare Joined Cases C-22/08 and C-23/08 Vatsouras and Koupatantze EU:C:2009:344 with Case C-67/14 Alimanovic EU:C:2015:597. Raising the potential for a more positive perspective on this case law when viewed not from the perspective of the individual claimants but from the angle that it can ‘read as re-locating or re-empowering domestic authorities to shape the contours of their social assistance system’, see E Muir, ‘Drawing Positive Lessons From the Presence of “The Social” Outside of EU Social Policy Stricto Sensu’ (2018) 14 European Constitutional Law Review 75, 80–81. 32 Art 14(2) of the Directive provides that ‘Union citizens and their family members shall have the right of residence provided for in [Art 7] as long as they meet the conditions set out therein’. Art 14(4) then provides: ‘By way of derogation from [para 2] … an expulsion measure may in no case be adopted against Union citizens or their family members if: … (b) the Union citizens entered the territory of the host Member State in order to seek employment. In this case, the Union citizens and their family members may not be expelled for as long as the Union citizens can provide evidence that they are continuing to seek employment and that they have a genuine chance of being engaged’. These tests are drawn from Case C-292/89 Antonissen EU:C:1991:80. 33 Case C-138/02 Collins EU:C:2004:172, para 63.

Free Movement of Persons: The ‘Social Freedom’?  119 subsistence costs necessary to lead a life in keeping with human dignity’, ie, not if they constitute social assistance.34 Free movement status is therefore delineated not by (substantive) financial need but by (formal) activity, actual or potential, undertaken in the host State. In that light, for those who do not fall within the scope of the economic freedoms, the extent to which the free movement of persons is integrated with EU social policy becomes pivotal. In other words, where the market-driven framework of free movement law does not provide protection for Union citizens in need, what does EU social policy then offer? This question is picked up in section IV below. First, aspects of the UK’s withdrawal from the European Union are examined to consider if the free movement of persons may be more the ‘political freedom’ than the ‘social freedom’ in reality.

III.  The ‘Political Freedom’? The Free Movement of Persons in EU27 At the time of writing, the UK had not yet withdrawn from the European Union. Yet in several fields, preparatory work for policy change or development was already articulated around EU27 only – including all the work leading to the proclamation of the European Pillar of Social Rights. For example, in April 2017, the European Commission framed its modelled scenarios on the future of ‘social Europe’ around the anticipated reality but also advocated a shift in approach post-Brexit: As the Europe of 27 looks to shape its future, the discussion on the social dimension of our Union is timely and essential. In recent years, Europe has been busy with ‘firefighting’, responding to one crisis after another. Now is the time to draw lessons and to open a new chapter. For this, we must take a longer-term perspective and confront the more profound transformations in our economy and society.35

The specific example of indexation of exported child benefits will now be examined to contemplate whether the social dimension of the free movement of persons would be any different in EU27. Amending Regulation 883/2004 to accommodate the possibility of indexing exported family benefits according to ‘the conditions of the Member State where

34 Case C-67/14 Alimanovic, above n 31, para 45. Social assistance for the purposes of Directive 2003/38 is defined as ‘all assistance introduced by the public authorities, whether at national, regional or local level, that can be claimed by an individual who does not have resources sufficient to meet his own basic needs and the needs of his family and who, by reason of that fact, may become a burden on the public finances of the host Member State during his period of residence which could have consequences for the overall level of assistance which may be granted by that State’ (Case C-140/12 Brey EU:C:2013:565, para 61). 35 European Commission, ‘Reflection paper on the social dimension of Europe’ COM(2017) 206, 6–7.

120  Niamh Nic Shuibhne the child resides’36 was one of the proposals agreed in the February 2016 Decision on a new settlement for the United Kingdom within the European Union. The motivation behind the proposal, alongside limits on access to ‘in-work benefits’ for ‘newly arriving EU workers’ was presented as follows: Different levels of remuneration among the Member States make some offers of employment more attractive than others, with consequential movements that are a direct result of the freedom of the market. However, the social security systems of the Member States, which Union law coordinates but does not harmonise, are diversely structured and this may in itself attract workers to certain Member States. It is legitimate to take this situation into account and to provide, both at Union and at national level, and without creating unjustified direct or indirect discrimination, for measures limiting flows of workers of such a scale that they have negative effects both for the Member States of origin and for the Member States of destination.37

For child benefits specifically, it was proposed that an indexation option would first apply ‘only to new claims made by EU workers in the host Member State. However, as from 1 January 2020, all Member States may extend indexation to existing claims to child benefits already exported by EU workers’.38 The Commission clarified that the conditions of the Member State where the child resides ‘include the standard of living and the level of child benefits applicable in that Member State’.39 The result of the referendum in the UK in June 2016 dissolved the legal force of the February 2016 agreement. But the possibility of indexing exported child benefits outlived the specifics of the Brexit process. The Commission had already been working on a revision of Regulation 883/2004 since 2014, but the proposal and impact assessment were not published until December 2016, ie, after the February 2016 Decision as well as the outcome of the referendum in the UK. The proposal indicated that ‘a significant minority of Member States delegations favoured different coordination of benefits intended to replace income during child-raising periods’.40 However, only the option of ‘no export’ was ‘disregarded due to legal reasons’.41 Specifically on indexation, the Commission made an evidence-based rather than legal determination, pointing out that ‘[t]he impact of the export of child benefits on total expenditure is quite limited for most of the Member States under the current rules … A change to another option has on average no significant impact

36 Decision of the Heads of State or Government, meeting within the European Council, concerning a new settlement for the United Kingdom within the European Union [2016] OJ C691/1 Section D, para 2(a). 37 Decision of the Heads of State or Government, Section D. 38 Decision of the Heads of State or Government Section D, para 2(a). 39 European Commission, Annex V, Decision of the Heads of State or Government. In para 2(a) of the main Decision, it was stated that ‘[t]he Commission does not intend to propose that the future system of optional indexation of child benefits be extended to other types of exportable benefits, such as old-age pensions’. 40 COM(2016) 815 final, 6. 41 Report on export of family benefits included in the Impact Assessment, SWD(2016) 460 final, 30 and 42.

Free Movement of Persons: The ‘Social Freedom’?  121 on the public spending on family benefits’.42 On the balance of evidence gathered, the Commission press release issued on the publication of the proposal concluded that ‘[n]o indexation of child benefits is foreseen’ since ‘[l]ess than 1 per cent of child benefits in the EU are exported from one Member State to another’.43 The story developed further through a Commission press release issued on 24 January 2019, which confirmed that a letter of formal notice had been sent to Austria with respect to ‘new legislation under which EU citizens working in Austria face an indexation of their family benefits and family tax reductions when their children reside abroad’.44 That legislation, which came into effect on 1 January 2019, makes family benefits and family tax reductions paid for children residing in another Member State dependent on the costs of living of that Member State. This means that many EU citizens, who work in Austria and contribute to its social security and tax system in the same way as local workers, would receive fewer benefits only because their children are living in another Member State.

The Commission reiterated that proposals to allow indexation of family benefits for children residing abroad were rejected in the process of revising Regulation 883/2004. It then commented on the legal dimension of the infringement, stating that EU rules on the coordination of social security systems in Regulation 883/2004 ‘do not allow a Member State to reduce cash benefits granted to persons insured under its legislation solely because they or their family members reside in another Member State’. It confirmed that the provisions of Regulation 883/2004 also prohibit discrimination on grounds of nationality. Any reduction of family benefits solely because the children reside abroad, breaches the EU rules on social security as well as the principle of equal treatment of workers who are nationals of another Member State as regards social and fiscal advantages.

It also invoked Article 7(2) of Regulation 492/2011, which was simply not mentioned at all in the 2016 proposal for indexation of child benefits. The legal demands of non-discrimination on nationality grounds seem obvious in this connection, but were utterly suppressed for the facilitation of UK-specific demands in 2016.45 Similarly but in more of a moral sense, in her comment on the Austrian infringement, the Commissioner for Employment, Social Affairs, Skills and Labour Mobility stated that [o]ur single market is based on fairness and equal treatment. There are no second-class workers in the EU. When mobile workers contribute in the same way to a social security system as local workers, they should receive the same benefits, also when their children live abroad. There are no second-class children in the EU. 42 ibid, 66. 43 European Commission, press release of 13 December 2016, IP/16/4301. 44 European Commission, press release of 24 January 2019, IP/19/462. The Commission has since confirmed that it has decided to refer Austria to the Court of Justice in this infringement procedure (press release of 14 May 2020, IP/20/849). 45 See further, N Nic Shuibhne, ‘Reconnecting the Free Movement of Workers and Equal Treatment in an Unequal Europe’ (2018) 43 European Law Review 477.

122  Niamh Nic Shuibhne There were no comparable statements in the 2016 Decision. It is too simplistic to attribute the Commission’s repositioning from acquiescence to fact-based analysis to protection of foundational legal principles solely to the UK’s repositioning at the edges of EU influence as it neared and awaited the completion of its own EU exit. But surely the example of indexation shows that a diminution of responsiveness to Brexit concerns cannot be entirely irrelevant either as the Commission restores a bolder approach to the social dimension of integration from the embers of concession. In particular, in its 2018 Communication on ‘The Single Market in a Changing World’, the Commission commented that: The Single Market is a powerful engine driving the Union’s competitiveness and the prosperity of its inhabitants. It fulfils an important societal function by creating a shared living space based on common rules for over 512 million Europeans. According to the spring 2018 Eurobarometer, 82% of Union citizens support the freedom to live, work, study and do business in other Member States. This is the highest level of support for any Union policy. In addition, the external dimension of the Single Market is also generating economic and social benefits as it has given the Union unique leverage in international trade negotiations and is an asset that attracts foreign investment and talent.46

That is a remarkable statistic and yet it feels far too often that policymaking is driven disproportionately by the 18 per cent. Tracking the story of indexation over recent years demonstrates, at worst, that the free movement of persons is the political rather than the social freedom: that it is instrumentalised to meet political ends, and often to address perceived more than actual concerns around which the law then bends. A re-emboldened Commission is now evident, however, as are the restoration of evidence-based policy change and the willingness to be bound by fundamental principles of EU law – all of which are essential for ethically robust regulation. Speeches that set out the vision of and work ahead for the Union also underline the importance of social cohesion as a fundamental objective alongside formidable challenges such as climate change and the digital economy in a reconfigured global order.47 And even in the Brexit context, the UK’s political rejection of the free movement of persons does not displace, for the EU side at least, EU standards on social protection remaining relevant through the premise of the ‘level playing field’ at the heart of the intended future partnership.48 46 European Commission, ‘The Single Market in a Changing World’, above n 4, 2; referring to Standard Eurobarometer 89, Spring 2018. See also, M Ferrera, ‘The Contentious Politics of Hospitality: Intra-EU Mobility and Social Rights’ (2016) 22 European Law Journal 791, 797–802. 47 See, eg, the speech by Michel Barnier at the College of Europe in Natolin, ‘Europe after Brexit’, 29 March 2019, available at: ec.europa.eu/commission/publications/speech-michel-barnier-collegeeurope-natolin-europe-after-brexit_en. 48 See esp Art 79 of the political declaration setting out the framework for the future relationship between the European Union and the United Kingdom [2019] OJ C66I/185: ‘[t]he future relationship must ensure open and fair competition. Provisions to ensure this should cover state aid, competition, social and employment standards, environmental standards, climate change, and relevant tax matters, building on the level playing field arrangements provided for in the Withdrawal Agreement and commensurate with the overall economic relationship’.

Free Movement of Persons: The ‘Social Freedom’?  123 Do these commitments translate into a more integrated economic/social understanding of the EU’s internal market? Even more fundamentally, do they adequately reflect ‘the desire to protect the dignity and autonomy of individuals as well as social justice’, which must be ‘the driving force for legal and policy change’ in any dimension of EU law?49

IV.  The ‘Social Freedom’? Through evaluating the extent to which the free movement of persons can claim to be the ‘social freedom’ of the EU’s internal market, three linked structural vulnerabilities emerge over the discussion in this section: first, a prevailing register of disconnect rather than conflict; second, the blocking quality of asymmetries of competence; and, third, a lack of clarity around responsibility for action. The analysis is undertaken through the prism of the EPSR. In particular, the extent to which the objectives of the Pillar either translate to or more directly confront the social challenges generated by the mobility dimension of the internal market is investigated, using the example of retaining worker status to illustrate the challenges of market/social disconnect more concretely. The plea is for more consciously joined-up thinking across the span of EU activities, reflecting ultimately Article 7 TFEU’s statement that ‘[t]he Union shall ensure consistency between its policies and activities, taking all of its objectives into account and in accordance with the principle of conferral of powers’. The social dimension of the free movement of persons manifests in different ways from a legal perspective but these share in common critical questions about the (limited) competence of the Union. Considering, first, a more indirect dimension, social concerns have long infused regulation of the free movement of persons. In fact, what might now be described using the language of mainstreaming or policy integration50 was already evident in very early legislation and case law on the economic freedoms; considering, for example, the legislative provision made for the family members of Member State nationals moving to another State to undertake economic activity, notwithstanding complete silence on family

49 Claire Kilpatrick, Elise Muir and Sacha Garben, ‘From Austerity Back to Legitimacy? The European Pillar of Social Rights: A Policy Brief ’ (EU Law Analysis, 20 March 2017), available at: eulawanalysis.blogspot.com/2017/03/from-austerity-back-to-legitimacy.html. The authors caution that ‘if the message is blurred by economic arguments in support for change, or made subject to economic conditions, or wishing away hard choices between the economic and the social, or attributing Social Europe’s malaise to new technologies and platforms, the message and its delivery will be imperilled’. 50 See esp Art 9 TFEU: ‘In defining and implementing its policies and activities, the Union shall take into account requirements linked to the promotion of a high level of employment, the guarantee of adequate social protection, the fight against social exclusion, and a high level of education, training and protection of human health’. See also, Art 3(3) TEU (returned to below).

124  Niamh Nic Shuibhne members in the Treaty;51 or how formative case law on the coordination of social security sidestepped requiring a connection between the claim made and the performance of economic activity, so long as a cross-border connection could be established.52 But alongside this grafting of social protection onto free movement rights, there was also the sense that the aims of respecting (EU) free movement rights and protecting (national) social rights were inevitably pitched against each other, even when presented as aspirations for better balance.53 Moreover, the opening up over time of new legal routes for enhanced social protection in the free movement context has not seen the depth of progress that might have been expected. For example, with the notable exception of respect for family life (Article 7) and guarantees on the best interests of the child (Article 24), there has been relatively little engagement with the social rights protected by the Charter of Fundamental Rights in case law on the free movement of persons.54 Furthermore, references to social protection added to the Treaty provisions on free movement have been more concerned with securing maximum participation for the Member States in any further legislative initiatives.55 Second, as the Union acquired more direct competence in the field of social policy, those who moved to other Member States necessarily benefited from, for example, improvements in working conditions and in equal pay linked to EU legislation and case law. But they would have benefited from the same advances in standards had they never moved in the first place. With respect to responsibility for action, while the Union and the Member States share competence for social policy (Article 4(2) TFEU), specific constraints are placed on Union competence

51 Council Directive 68/360/EEC of 15 October 1968 on the abolition of restrictions on movement and residence within the Community for workers of Member States and their families [1968] L257/13; Regulation 1612/68/EEC on freedom of movement for workers within the Community [1968] OJ L257/13; and Directive 73/148/EEC on the abolition of restrictions on movement and residence within the Community for nationals of Member States with regard to establishment and the provision of services [1973] OJ L172/14. 52 eg, Case 31/64 Bertholet EU:C:1965:18; Case 33/64 van Dijk EU:C:1965:19; Case 44/65 Singer EU:C:1965:109; Case 27/69 SNCFL EU:C:1969:56. 53 eg, consider this statement in the 1997 High Level Panel report on the free movement of persons: ‘For the Member States, the days when one thought one could protect the employment and social advantages of nationals at the expense of migrants living precariously on the margin are over. In the Community, migration is independent of the economic and social short-term situation and it must be regulated in accordance with criteria which foster initiative, innovation and competitiveness. In such a context, the solution of employment problems calls for solidarity from all, but not for protectionist withdrawal’ (Report of the High Level Panel on the free movement of persons chaired by Mrs Simone Veil, presented to the Commission on 18 March 1997, 13). 54 eg, Art 34(3) of the Charter provides that ‘[i]n order to combat social exclusion and poverty, the Union recognises and respects the right to social and housing assistance so as to ensure a decent existence for all those who lack sufficient resources, in accordance with the rules laid down by Union law and national laws and practices’; but this provision has only been engaged to date in Case C-571/10 Kamberaj EU:C:2012:233 (in the context of Directive 2003/109 concerning the status of third-country nationals who are long-term residents [2003] OJ L16/44). 55 See esp Arts 21(3) TFEU and 48 TFEU.

Free Movement of Persons: The ‘Social Freedom’?  125 in Title X of the Treaty, noting in particular the limits expressed across different parts of Article 153 TFEU.56 For present purposes, this configuration of EU and Member State competences is taken as a neutral ‘fact’ but it is acknowledged that how competence is distributed for EU social policy is widely contested.57 The cornerstone development in the field is the European Pillar of Social Rights, solemnly proclaimed in November 2017 by the European Parliament, Council and Commission.58 In March 2016, the Commission launched a consultation on this initiative.59 In January 2017, the European Parliament presented an ambitious vision in response, calling on the Commission to build on the review of the social acquis and of EU employment and social policies as well as on the outcomes of the 2016 public consultation by making proposals for a solid European Pillar of Social Rights (EPSR) that is not limited to a declaration of principles or good intentions but reinforces social rights through concrete and specific tools (legislation, policy-making mechanisms and financial instruments), delivering a positive impact on people’s lives in the short and medium term and enabling support for European construction in the 21st century by effectively upholding the Treaties’ social objectives, supporting national welfare states, strengthening cohesion, solidarity and upward convergence in economic and social outcomes, ensuring adequate social protection, reducing inequality, achieving long overdue progress in reducing poverty and social exclusion, facilitating national reform efforts through benchmarking and helping to improve the functioning of the Economic and Monetary Union (EMU) and of the EU’s single market.60

56 eg, Art 153(1) TFEU provides: ‘With a view to achieving the objectives of Art 151, the Union shall support and complement the activities of the Member States in the following fields: (a) improvement in particular of the working environment to protect workers’ health and safety; (b) working conditions; (c) social security and social protection of workers; (d) protection of workers where their employment contract is terminated; (e) the information and consultation of workers; (f) representation and collective defence of the interests of workers and employers, including co-determination, subject to paragraph 5; (g) conditions of employment for third-country nationals legally residing in Union territory; (h) the integration of persons excluded from the labour market, without prejudice to Art 166; (i) equality between men and women with regard to labour market opportunities and treatment at work; (j) the combating of social exclusion; (k) the modernisation of social protection systems without prejudice to point (c). Art 153(4) should also be noted: ‘The provisions adopted pursuant to this Article: shall not affect the right of Member States to define the fundamental principles of their social security systems and must not significantly affect the financial equilibrium thereof; shall not prevent any Member State from maintaining or introducing more stringent protective measures compatible with the Treaties’. 57 eg, a more positive perspective can be found in the argument that ‘[w]hile these limitations may seem regrettable to those in favour of Social Europe, it can also be argued that the authority to take decisions of such re-distributive impact are rightly withheld from the EU institutions, in light of their limited democratic legitimacy’ (S Garben, ‘The Constitutional (Im)balance between “the Market” and “the Social” in the European Union’ (2017) 13 European Constitutional Law Review 23, 26). 58 For general analysis of the Pillar, see S Garben, ‘The European Pillar of Social Rights: Effectively Addressing Displacement?’ (2018) 14 European Constitutional Law Review 210. 59 COM(2016) 127 final, the results of which were published in April 2017 (SWD(2017) 206 final). 60 European Parliament, Resolution of 19 January 2017 on a European Pillar of Social Rights, 2016/2095(INI).

126  Niamh Nic Shuibhne In April 2017, the Commission published a reflection paper on the social dimension of Europe61 as part of its broader initiative on the future of Europe,62 modelling various scenarios in which the extent and reach of Union action in the field of social policy differed in intensity. In the same month, it also published its Recommendation on the European Pillar of Social Rights, including a proposed text organised around 20 headline principles.63 As noted above, the final text of the Pillar was formally ‘proclaimed’ in November 2017.64 Though not a legally binding text – as Garben emphasises, ‘the Pillar is best understood not as a single, contained legal measure but instead as a more fluid, political initiative that is still developing’65 – the EPSR is nevertheless establishing itself as a centripetal point of reference for a wide range of activities and initiatives. Three overarching objectives define its three chapters: equal opportunities and access to the labour market; fair working conditions; and social protection and inclusion. In terms of realising these objectives, the EPSR is conceived as a joint Union/Member State endeavour: Delivering on the European Pillar of Social Rights is a shared political commitment and responsibility. The European Pillar of Social Rights should be implemented at both Union level and Member State level within their respective competences, taking due account of different socio-economic environments and the diversity of national systems, including the role of social partners, and in accordance with the principles of subsidiarity and proportionality.66

Similarly, the Commission has acknowledged the reality of EU competence limits in the field of social policy, observing that [m]ost of the tools required to deliver on the Pillar are in the hands of local, regional and national authorities, as well as the social partners, and civil society at large. The European Union – and the European Commission in particular – can help by setting the framework, giving the direction and establishing a level-playing field, in full respect of the specificities of national circumstances and institutional set-ups.67

These statements are particularly important for present purposes, since they require us to reflect on where responsibility for implementing the principles of the EPSR does – and should – lie in the specific context of free movement, especially since the EU legislator has a clear mandate to regulate the free movement of

61 European Commission, ‘Reflection paper on the social dimension of Europe’ COM(2017) 206. 62 European Commission, ‘White Paper on the Future of Europe: Reflections and Scenarios for the EU27 by 2025’ COM(2017) 2025. 63 European Commission, Recommendation on the European Pillar of Social Rights C(2017) 2600 final. See further, Commission Communication, Establishing a European Pillar of Social Rights’ (2017) 250 final. 64 The text of the EPSR is available at: ec.europa.eu/commission/sites/beta-political/files/socialsummit-european-pillar-social-rights-booklet_en.pdf. 65 Garben, ‘The European Pillar of Social Rights’, above n 58, 211. 66 European Pillar of Social Rights, recital 17 of the preamble. 67 COM(2017) 250 final, para 1.

Free Movement of Persons: The ‘Social Freedom’?  127 persons in the internal market, but less straightforward capacity to settle questions at the intersection of free movement and social protection. As noted earlier, social policy advances at a general level benefit all participants in the internal market just as social challenges also impede them; but what of the specific advances and challenges produced precisely because of the exercise of free movement rights, which would simply never have arisen had the persons concerned remained in their home States (where responsibility for their social protection is then more straightforward)? In fact, there was little reflection on this question in the documents that shaped the EPSR. At one level, this is an understandable aspect of the measure’s wider focus: specifically, ‘[t]he Pillar is not aimed at fixing any problems with the internal market’.68 But disconnect between social policy and the specific challenges produced by the exercise of free movement rights undercuts the purpose of the policy integration method deepened through the Lisbon amendments in particular. Article 9 TFEU was already mentioned; it requires the taking into account of ‘the guarantee of adequate social protection’ and ‘the fight against social exclusion’ in ‘defining and implementing [the Union’s] policies and activities’.69 Commitments the other way around – ie, requiring reflection on the social dimensions of internal market policy – are mainly found in Article 3 of the Treaty on European Union (TEU) and especially in the idea of ‘a highly competitive social market economy, aiming at full employment and social progress’ (Article 3(3) TEU) and sitting alongside the requirement to establish an internal market in the first place. The Communication from the Commission launching the pre-EPSR public consultation did not mention free movement at all – any references to mobility were about occupational, not internal market, mobility.70 In its January 2017 Resolution, the European Parliament did consider the free movement dimension to some (though still quite limited) extent. For example, one of several objectives expressed in paragraph 3 of the Resolution was that the EPSR should ‘facilitate free movement of workers in a deeper and fairer European labour market’. In paragraph 16, addressing the rights of people with disabilities, reference was made to ‘guaranteeing free movement and the transferability of services between EU Member States’. And more substantively, three paragraphs of the Resolution engaged with free movement-specific challenges in a section entitled ‘labour mobility’, raising questions that included the need for ‘a smoothly functioning system of social security

68 Garben, ‘The European Pillar of Social Rights’, above n 58, 228. 69 On the requirements of Art 9 TFEU more specifically, see further ME Bartolini, ‘The Horizontal Social Clause in a Legal Dimension’ in F Ippolito, ME Bartoloni and M Condinanzi (eds), The EU and the Proliferation of Integration Principles under the Lisbon Treaty (London, York, Routledge, 2019) 83. 70 COM(2016) 127 final; see similarly, COM(2017) 250 final. There was some reflection on free movement in the Commission’s ‘Reflection paper on the social dimension of Europe’ (COM(2017) 206), but mainly in the context of one of the more restrictive scenarios, ie, limiting the social dimension of EU law to free movement.

128  Niamh Nic Shuibhne coordination’, ‘the importance of cohesion policy and other instruments for territorially balanced economic development’, and the need ‘to take into account the social impact of mobility on the increasing number of transnational families’.71 These are precisely the kinds of questions that an integrated rather than disconnected approach would – should – confront. Moving from questions about framing to questions about implementing, the Commission has already presented a slew of initiatives that have either been introduced or completed against the headline 20 principles that the EPSR articulates.72 While free movement-linked initiatives are not particularly visible as part of that work, looking more closely demonstrates that there have, in fact, been some significant developments. For example, attached to principle 1 (education, training and lifelong learning), a proposal to double Erasmus funding for the 2021–27 budgetary period indicates that ‘[t]he focus of the new Programme will be on inclusiveness, and to reach more young people from disadvantaged backgrounds’.73 The revised Directive on the posting of workers,74 which establishes the principle of equal pay for equal work at the same place, is linked to principle 3 (equal opportunities). Reform of Regulation 883/2004 on social security coordination75 is connected to principles 12 (social protection) and 18 (long-term care). And most directly, the

71 These paragraphs provide in full that the Parliament: ‘27. Emphasises that free movement of people is one of the greatest achievements of the EU and that free movement of workers is a cornerstone of the internal market, which plays an important role in enhancing convergence and integration among Member States; underlines that mobility within the EU is an opportunity and a fundamental right whose exercise must be supported, including through a smoothly functioning system of social security coordination; calls for proper implementation and enforcement of EU rules regarding mobility of workers and cross-border provision of services; calls, moreover, for support to be provided for labour mobility through adequate language training at all levels of education, improved comparability of education systems and recognition of professional qualifications, readily available information on the rights and duties of mobile workers, and measures ensuring decent working conditions and effective cooperation among public employment services across Europe; notes that mobility should not be the result of inadequate employment opportunities or inadequate social protection in workers’ home regions as prolonged workforce outflows may hinder economic convergence; points therefore to the importance of cohesion policy and other instruments for territorially balanced economic development; considers that labour mobility should not be abused to undermine host countries’ social standards through fraud or circumvention of law; highlights the fact that mobile workers are usually net contributors to host countries’ public budgets; calls for adequate investments in public services in areas experiencing population increases and points to the support which the European Social Fund can provide in this respect; 28. Calls for the EU and the Member States to take into account the social impact of mobility on the increasing number of transnational families, eg by providing leave opportunities to care for a family member in another country and ensuring the transferability and comparability of education systems in terms of the mobility of school-age children; 29. Calls on the Member States to make it mandatory for employers to provide an employment contract in a language known by EU mobile citizens, in order to make the labour contract understandable for workers’. 72 See esp European Commission, ‘One year on: European Pillar of Social Rights’, available at: ec.europa.eu/commission/sites/beta-political/files/european_pillar_one_year_on.pdf. 73 COM(2018) 367 final; see also, European Commission press release of 30 May 2018, IP/18/3948. 74 See further, the chapter in this volume by B de Witte (ch 7). 75 COM(2016) 815 final; see further, section III above.

Free Movement of Persons: The ‘Social Freedom’?  129 creation of a European Labour Authority76 seeks ‘to contribute to ensuring fair labour mobility in the Internal Market’ and, more specifically, aims at: [i]mproving access to information by individuals and employers about their rights and obligations in the areas of labour mobility and social security coordination as well as access to relevant services; [s]trengthening operational cooperation between authorities in the cross-border enforcement of relevant Union law, including facilitating joint inspections; [and] [p]roviding mediation and facilitating solutions in cases of disputes between national authorities and cross-border labour market disruptions, such as a restructuring of companies affecting several Member States.77

Importantly, the initiatives summarised just above do exemplify an approach of integrated policy objectives at one level. The problem is that they go nowhere near far enough. The example of retaining worker status for the purposes of free movement law usefully illustrates the persisting disconnect between social protection objectives and the context of mobility more concretely. Article 7(3) of Directive 2004/38 outlines the situations in which the status of worker or self-employed person is retained under EU law with reference to residence rights in Article 7(1)(a), and they concern temporary incapacity to work because of illness or accident, involuntary employment (with retention of worker status for different time periods depending on the period of economic activity in the host State), or embarking on vocational training. The crucial point about retaining worker or self-employed status in these circumstances is that residence rights in the host State under Article 7(1)(a) remain unconditional with respect to financial resources and sickness insurance and they sustain equal treatment with host State nationals for all social and tax advantages (including eligibility for social assistance). Moreover, the capacity to acquire permanent residence rights under Article 16 of the Directive after five years of continuous and legal residence is then undisturbed. In Gusa, the Court equated the protection offered to workers and self-employed persons under Article 7(3) of the Directive, confirming that the expression ‘involuntary unemployment’ may, depending on the context in which it is used, refer to a situation of inactivity due to the involuntary loss of employment following, for example, a dismissal, as well as, more broadly, to a situation in which the occupational activity, whether on an employed or self- employed basis, has ceased due to an absence of work for reasons beyond the control of the person concerned, such as an economic recession.78

76 Linked to principles 5 (secure and adaptable employment) and 12 (social protection). 77 COM(2018) 131 final, 2. At the time of writing, the legislative process that underpins the creation of the Authority had just been formally completed (see Regulation 2019/1149/EU of 20 June 2019 establishing a European Labour Authority [2020] OJL186/21) and it was anticipated that the Authority will reach full operational capacity by 2024. 78 Case C-442/16 Gusa EU:C:2017:1004, para 31. See further on these questions, the Commission’s ‘Proposal for a Council recommendation on access to social protection for workers and the selfemployed’ COM(2018) 132 final, which is explicitly linked to principle 12 of the EPSR.

130  Niamh Nic Shuibhne But what happens when worker status is lost in situations that are not addressed by Article 7(3) of the Directive? In Saint Prix, the Court of Justice ruled that a Union citizen who ceased working in the host State ‘because of the physical constraints of the late stages of pregnancy and the aftermath of childbirth’ retained the status of worker on the basis of Article 45 TFEU ‘provided she returns to work or finds another job within a reasonable period after the birth of her child’.79 The Court ruled out the application of Article 7(3) on the facts of the case, but it rejected the proposition that Article 7(3) ‘lists exhaustively the circumstances in which a migrant worker who is no longer in an employment relationship may nevertheless continue to benefit from that status’.80 It advised that [i]n order to determine whether the period that has elapsed between childbirth and starting work again may be regarded as reasonable, the national court concerned should take account of all the specific circumstances of the case in the main proceedings and the applicable national rules on the duration of maternity leave.81

Article 8(1) of Directive 92/95 refers to ‘a continuous period of maternity leave of a least 14 weeks’ but that measure does not provide guidance for situations that involve more difficult circumstances such as the illness of the mother or baby; or consider the support that might be needed for the purposes of finding new employment where relevant. Moreover, the claimant in Saint Prix was working once again by the time her case was considered by the Court of Justice, her claim for retrospective payment of income support then facilitated for the national authorities since, by then, she already fulfilled the Court’s test. But how can national authorities resolve claims when it is logically impossible for the pregnant worker or new mother to show that they have already returned to work or found another position ‘within a reasonable period’? The approach taken by the Court constructs an ex post framework for demonstration of retained worker status, whereas income support claims will normally need to be determined prior to that. The Saint Prix example exposes social protection fault-lines in the free movement context on its own terms. But consider the framework for retention of worker status also against one of the most developed (in terms of legal initiative) aspects of the EPSR. Principle 9 of the Pillar (‘work–life balance’) includes the statement that ‘[w]omen and men shall have equal access to special leaves of absence in order to fulfil their caring responsibilities and be encouraged to use them in a balanced way’. The Commission also proposed a directive on work–life balance for parents and carers, with specific reference to the objective of equality between men and

79 Case C-507/12 Saint Prix EU:C:2014:2007, para 47. Confirmed for Art 49 TFEU in Case C-544/18 Dakneviciute, EU:C:2019:761. 80 Saint Prix (ibid), para 38. 81 ibid, para 41; referring to Council Directive 92/85/EEC on the introduction of measures to encourage improvements in the safety and health at work of pregnant workers and workers who have recently given birth or are breastfeeding [1992] OJ L348/1.

Free Movement of Persons: The ‘Social Freedom’?  131 women in this context, which was adopted in May 2019.82 Article 9(1) requires that ‘Member States shall take the necessary measures to ensure that workers with children up to a specified age, which shall be at least eight years, and carers, have the right to request flexible working arrangements for caring purposes’ but also provides that ‘[t]he duration of such flexible working arrangements may be subject to a reasonable limitation’. Under Article 9(2), employers ‘shall consider and respond to requests for flexible working arrangements referred to in paragraph 1 within a reasonable period of time, taking into account the needs of both the employer and the worker’ and ‘shall provide reasons for any refusal of such a request or for any postponement of such arrangements’. Article 12 then outlines procedural protections and elements of the burden of proof for situations of dismissal. The protection conceived in the Directive is important and welcome, but it is undeniably limited in both temporal and substantive senses. Most critically for present purposes, it also ignores mobility-specific implications of altered economic activity circumstances because of caring responsibilities. At a general level, for example, the Directive does not address whether changes in the nature of work would affect fulfilment of the conditions that define worker status under Article 45 TFEU (and, in turn, Article 7(1)(a) of the Directive). Where caring responsibilities exceed the boundaries protected by the Directive, what are the implications for carers in a host Member State? Would they retain the status of worker under EU law if they have to cease work or self-employed activity by extended analogy with Saint Prix? If they do, how would the obligation to return to work ‘within a reasonable period’ be assessed outwith the context of pregnancy and maternity? What support, financial or otherwise, would remain available to them when not working? The latter point also connects to the strongly expressed statement on ‘minimum income’ in the EPSR (principle 14) ie, ‘[e]veryone lacking sufficient resources has the right to adequate minimum income benefits ensuring a life in dignity at all stages of life, and effective access to enabling goods and services’. Would responsibility for realising that objective be recognised in the context of caring responsibilities even if the status of worker or self-employed person is lost? These are just some of the questions that a joined-up reflection on the free movement and social protection dimensions of these situations produces. As seen in section II, EU free movement law rules out entitlement to subsistence income in the host Member State unless the person concerned is working or selfemployed there (or a right of permanent residence has already been acquired). But the example of retaining worker or self-employed status shows that there is a

82 COM(2017) 253 final; Directive 2019/1158/EU on work–life balance for parents and carers ([2019] OJ L118/79). The new Directive, which repeals Council Directive 2010/18/EU on parental leave ([2010] OJ L68/13), must be transposed by the Member States before 2 August 2022. On the reconciliation of work and caring responsibilities generally, and the procedural nature of duties in that context in particular, M Bell, ‘Adapting Work to the Worker: The Evolving EU Legal Framework on Accommodating Worker Diversity’ (2018) 18 International Journal of Discrimination and the Law 124, 134–36.

132  Niamh Nic Shuibhne far wider spectrum of experience between the ‘optimal’ free movement status of being in work or self-employment (or otherwise financially self-sufficient), on the one hand, and being in need of support because of not having worked or been selfemployed at all, on the other. Importantly, it also demonstrates that focusing only on claims for social assistance – critical though they are in so many situations – far too narrowly reduces the intersection of social protection and free movement law, especially when addressed through the profoundly reductive discourse of social ‘tourism’. If the range of situations made more visible through association with the EPSR highlights even that point, it is already some considerable progress. We will continue to wonder at a general level if the EPSR is mainly relevant as a unifying reference point for scattered activities that achieve incremental or sectoral progress or if it also has the potential to drive significant change. The more pressing point for present purposes is the awkward and unproductive disconnect between the level of ambition conveyed by the EPSR, on the one hand, and the persisting structural vulnerabilities about allocating responsibility for meaningful protection of social rights in the context of free movement, on the other. The value of surmounting a disconnected EU social policy is widely acknowledged. For example, Muir advocates ‘look[ing] at the process of European integration as profoundly reflexive and constantly inviting a reconsideration of pre-established parameters’ in distinction from ‘reason[ing] in binary terms seeing economic integration and social protection, or anti-discrimination and social rights, or again European and State intervention as mutually exclusive’.83 For Schiek, ‘[i]f If the EU is to regain social legitimacy while continuing economic integration, it must pursue economic and social integration at European and national levels as an interconnected endeavour’.84 The register of disconnect is emphasised here not to detract from novel and progressive elements of ongoing work linked to the EPSR or unduly to prioritise for the Pillar the responsibility of realising a high level of social protection in internal market law.85 It is not, in other words, a selfish consuming of the EPSR by what some would see as the already overprivileged sector of EU free movement law. It is explored instead to underline the intractable asymmetries that define through obstruction the progression of EU-led betterment of life: whether we are speaking about asymmetries in connection with the competence to regulate, or mobility patterns, or levels of (in)equality, or social protection standards, or political perspectives. Conversely, the legal framework of the internal market works best precisely when the aims it is deployed to achieve pursue symmetry: harmony, levelness, uniformity. Thus, while much recent work on the social dimension of

83 Muir, above n 31, 77. 84 Schiek, above n 2, 612–13. 85 After all, as Muir underlines, ‘[t]he process of European internal market and economic governance, as well as the driving forces behind EU law-making on fundamental rights, are characterised by a uniquely powerful and sophisticated web of legal tools’ (Muir, above n 31, 95).

Free Movement of Persons: The ‘Social Freedom’?  133 the internal market rightly confronts changing labour market realities,86 perhaps insufficient attention is being paid to asymmetrical internal market realities. Nevertheless, and reflecting the ‘turn’ identified also in section III, a more overt and confident emphasis on the integration of all elements of the internal market – in both its static and mobility-linked senses – is discernible in the recent work of the Commission, which acknowledged in its 2018 Communication on ‘The Single Market in a changing world’ that the positive impact of the Single Market has not been spread evenly and that not all citizens are in a position to benefit from its freedoms. There is a clear need to address the concerns of citizens in regions with high unemployment or facing structural changes. An increase in labour mobility or market integration can lead to problems with income and job security. To do this, the Union has taken action to make its workforce more resilient to changes on the labour market, for example through reskilling or retraining, and by putting in place strong labour market standards. The recent revision of the legislation on posted workers, for example, provides strengthened protection and in particular puts in place the principle of equal pay for equal work in the same place. The European Pillar of Social Rights reflects a common understanding of European social standards, sets out an agenda to deliver new and more effective rights for citizens and workers, and addresses emerging social and demographic challenges and the changing world of work.87

In the same Communication, the Commission – emphasising the need for ‘willingness to follow through’ when it comes forward with new proposals – argued that [w]e are too often confronted with a situation where the consensus which appears to exist at the highest level on the need to deepen the Single Market is not matched by a political willingness to adopt the concrete measures that the Commission proposes and that would make a difference, or to transpose and implement measures which have already been agreed.88

Similarly, Garben suggests that the EU will have to engage in a more difficult, and more self-critical exercise, in which the internal market, budgetary balance and social justice are not presented as noncontroversial, apolitical issues that benefit citizens, workers and businesses alike, but instead are being recognised for the sensitive, political issues that they are.89

86 eg, in its Statement of 21 June 2018 (18/4222), commenting on its proposal for more transparent and predictable working conditions across the EU (COM(2017) 797 final), the Commission emphasises that ‘the proposal creates new minimum standards to ensure that all workers, including those on atypical contracts, benefit from more predictability and clarity as regards their working conditions. [It] updates and replaces the 1991 Written Statement Directive (91/533/EEC), which no longer captures changing labour market realities, in particular the new forms of work that have developed in recent years’. 87 COM(2018) 772 final 4 (emphasis added). 88 COM(2018) 772 final 1. 89 Garben, ‘The European Pillar of Social Rights’, above n 58, 230.

134  Niamh Nic Shuibhne Unfortunately, in many respects, the ambition of the EPSR – at least on paper – seems to fall back too quickly on the social policy-familiar approach of persuasion through soft law rather than determination by hard(er) law: in large part, precisely because of the thin political willingness highlighted explicitly by the Commission. The power of persuasion should not be underestimated. For example, linking back to the significance of establishing rather than conjecturing the social costs of EU policies raised in section III above and arguing for a ‘better approach to social issues through interdisciplinarity’, Robin-Olivier suggests ‘[a]usterity measures for which no proof has been offered of their necessity, for instance, could be more easily contested if indicators of inequality and poverty were systematically used to monitor their impact and to address the adverse (social) impact of transformations of the law’, locating a role for the Pillar as reference point for that purpose.90 Garben makes another critical point, arguing that the Pillar ‘increases the cost of opposing or down-levelling social initiatives for all institutions that have “solemnly” proclaimed their attachment to these values, which includes the Member States in the Council’.91 Benedi Lahuerta and Zbyszewska also point out that ‘the wording of the principles is quite commanding for a soft law instrument’.92 Like the Charter of Fundamental Rights before it then – and before that measure acquired binding legal effect through the Lisbon Treaty – the value of entrenching the EPSR in legal consciousness if not legal effect, through consistent articulation of and engagement with its principles in legal as much as political contexts, should not be ruled out.93 But the fact remains that only a more widely shared commitment from all actors, across all levels of EU governance, will progress the integration of economic and social objectives in markedly harder and more meaningful ways. The sense of things thus far is that the Pillar is proving effective for identifying challenges that need to be addressed; but perhaps less effective for identifying how to address them.

V. Conclusion This chapter has acknowledged the significance of the European Pillar of Social Rights as a milestone in the framing and also the progressing of social policy at EU level, but questioned the extent to which social challenges that result directly 90 Robin-Olivier, above n 3, 109 and 112. 91 Garben, ‘The European Pillar of Social Rights’, above n 58, 7. 92 S Benedi Lahuerta and A Zbyszewska, ‘EU equality law after a decade of austerity: On the Social Pillar and its transformative potential’ (2018) 18 International Journal of Discrimination and the Law 163, 180. 93 The EPSR has yet to gain a foothold in EU litigation consciousness, however; at the time of writing, it has been referred to in just one opinion (AG Wahl in Case C-33/18 Čepelnik EU:C:2018:311, para 57) and has been invoked, alongside several provisions of the Treaties and of the Charter of Fundamental Rights, in just one preliminary reference (Joined Cases C-789/18 and C-790/18 Corte dei Conti and Others, which concerned various questions on the nature of the employment relationship and on working conditions).

Free Movement of Persons: The ‘Social Freedom’?  135 from the exercise of the free movement rights attached to the EU’s internal market are adequately acknowledged or confronted. The conventional framework of free movement law – premised on equality of treatment and uniform application of EU rights – is not, on its own terms, designed to address the complex questions of asymmetry (both competence-related and substantively) that emerge not when the ‘conflict’ between free movement rights and social protection is raised but, instead, the ‘disconnect’ between social protection and free movement rights is foregrounded. There needs, first, to be more consistent recognition that the free movement of persons raises specific, transnationally generated challenges that require, in turn, transnationally generated solutions. Second, these challenges should not be recognised and will not be effectively addressed in the universe of internal market law in isolation. Critically, this is not a claim for limitless free movement rights – or indeed limitless social protection – in all circumstances. But pretending that the free movement of persons exists in a disconnected, self-referential regulatory space and failing, in consequence, to take a harder look at how responsibilities should be shared for the social challenges produced by and particular to the mobility dimension of the internal market – the precisely novel dimension of the EU’s internal market – is not the answer. A more integrated approach undoubtedly calls for, in the Commission’s words, greater ‘political courage’. The malleability of the supposedly protective legal framework was negatively exposed through the process of the UK’s decision to withdraw from the European Union. However, following that story over time shows that instrumentalising the free movement of persons for political purposes can also be corrected.

136

7 The Freedom to Provide Services: The Controversial Freedom? BRUNO DE WITTE

I.  Introduction: A Fragmented Field The second part of the title of this chapter is inspired by events that took place earlier in this century. In 2004, the European Commission presented a draft Directive on services in the internal market, which aimed at producing a radical acceleration in the creation of the internal market for services. The ‘Bolkestein draft’ (as it became popularly known) created an unexpectedly massive upheaval, contributing even, according to some accounts, to the rejection of the Constitutional Treaty in the French referendum of 2005. Eventually, the Directive was enacted in 2006 with a more modest content that no longer produced the important change promised by the Commission’s original draft. Now, more than 10 years after these events, does this freedom still deserve to be called the controversial one among the four market freedoms? One may doubt it. One of the other freedoms, namely the free movement of persons, has become very central (and ‘controversial’) in the Brexit debate, whereas the political controversy surrounding the adoption of the Services Directive1 has subsided. Political controversies did arise more recently for one or other aspect of the freedom to provide services (for example, the regulation of the posting of workers, and the question whether platform services such as Uber and Airbnb may fully benefit from this freedom), but it is probably no longer appropriate to qualify this market freedom as an especially controversial one. The account proposed in this chapter focuses precisely on the recent evolution, since the time that services became so controversial; the chapter proposes a global view of EU services law as it has developed from 2010 until today. The year 2010 is chosen as the starting point for two reasons. First, the implementation of the Services Directive had to happen by 28 December 2009 (the instrument had been 1 The Services Directive will be spelt with capitalised initials throughout this chapter, so as to distinguish it from other, more specific, directives regulating the provision of services.

138  Bruno de Witte adopted by 2006 but left a very long transposition time to the Member States). More than 10 years have passed, so that one can now draw some first conclusions as to the amount and nature of the changes that this Directive brought to the field of EU services law. Second, the Lisbon Treaty entered into force in December 2009 as well. Although that Treaty did not modify the Treaty chapter about the freedom to provide services (which, rather amazingly, reads still as it was drafted at the time of the adoption of the EEC Treaty), it did contain the promise of some change in this field, through its insistence on the importance of non-market values, and through its new provisions dealing with public services. The overall perspective adopted in this chapter is that EU services law is a fragmented legal field. A recently published textbook noted that, following the adoption of the Services Directive, ‘services now fall broadly into one of three categories: those governed by the Directive, those governed by specific legislation and those governed by the case law and Article 56 TFEU directly’.2 In fact, the ‘specific legislation’ can be subdivided into two categories with different characteristics: sectoral regulation (dealing with an economically distinctive service sector, such as banking or audiovisual services); and horizontal regulation (dealing with a particular regulatory aspect of a whole range of service sectors, such as electronic commerce or posting of workers). In addition, the legal regime of services of general interest and the EU’s external policy on services present very distinctive features too. All in all, this chapter distinguishes between six different fields of EU services law: the free movement of services under primary law; the Services Directive; sectoral regulation; horizontal regulation; services of general interest; and external relations. Each field is marked by the existence of its own set of legal sources, by the dominant role of particular legal actors, and by the existence of different policy issues and debates. They are, one could say with some exaggeration, different legal worlds. In fact, the fragmentation goes even further as every one of the sector-specific regimes constitutes a legal field with its own characteristics, where the regulation of cross-border services is closely linked with other regulatory concerns. For example, EU health law has gradually emerged as a separate field of law,3 of which the cross-border mobility of patients (based on the free movement of services) is one particular aspect. As a preliminary illustration of the fragmentation thesis, one can compare the application of the Treaty freedom to provide services with the regulation of banking services. The former is a legal arena used by individuals (mostly business firms) and national and EU courts, where the main substantive issue is the tension between the freedom to conduct a transnational business as against the regulatory autonomy of states. The field of banking services, by contrast, is one in

2 D Chalmers, G Davies and G Monti, European Union Law, 4th edn (Cambridge, Cambridge University Press, 2019) 774. 3 See the reconstruction of this evolution by T Hervey, ‘Telling Stories about European Union Health Law: The Emergence of a New Field of Law’ (2017) 15 Comparative European Politics 352.

The Freedom to Provide Services  139 which secondary EU law forms the main source of law (rather than the Treaty on the Functioning of the European Union (TFEU)), and the legal issues concern the interpretation and application of this, often very detailed, regulatory framework. The main actors in this sub-field are commercial banks, national and European regulatory agencies, and a specific European supervision system, the Single Supervisory Mechanism (SSM). One could add a seventh field to the list above, namely digital services. These are indeed ‘services’ in the legal sense (although sometimes ancillary to the sale of goods or movement of persons), but are now situated in a separate policy domain (and they are discussed in a separate chapter of this volume). ‘Internal Market’ and ‘Digital Single Market’ are nowadays listed as different priorities in the Commission’s work programmes,4 and they institutionally correspond to separate Directorates General of the Commission, namely DG GROW and DG CONNECT. Some parts of the sectoral services Regulation have been absorbed into the digital agenda, as for example the regulation of telecommunication services.5 In legal terms, though, regulating the digital market is not a separate competence. The frequent use of Article 114 TFEU as the legal basis for EU digital market measures creates some ambiguity as to whether this is regulation of services or (also) something else.6 The ‘catch-all’ scope of Article 114 is convenient, since it allows for the adoption of regulatory regimes that deal with the provision of services but also, at the same time, with other dimensions of the internal market. For example, the Regulation on the cross-border portability of access to online services7 seems to concern essentially the provision of services; yet, the preamble of the Regulation also refers to the aim of facilitating the free movement of persons, even though what is meant here is the temporary movement of persons outside their normal place of residence rather than free movement in the sense of the TFEU. In fact, the question whether ‘catch all’ Article 114 is the appropriate legal basis for a piece of EU law, rather than the specific legal basis for facilitating the provision of services, does not have much practical importance since the legislative procedure for the adoption of the measure is the same anyway (namely, the ordinary legislative procedure).

4 See, for instance, ‘Commission Work Programme 2019 – Delivering what we promised and preparing for the future’ COM(2018) 800 of 23 October 2018. The document distinguishes, among the Commission’s 10 priorities, ‘a connected Digital Single Market’ from ‘A deeper and fairer Internal Market with a strengthened industrial base’. 5 See Directive 2018/1972 establishing the European Electronic Communications Code (recast) [2018] OJ L321/36. This Directive replaces the ‘telecom package’ of four directives originally enacted in 2002, but is now firmly anchored in the Digital Single Market agenda. 6 See, as examples of legislation based on Art 114 Treaty on the Functioning of the European Union (TFEU) where the services component may seem predominant, Regulation 1128/2017 on cross-border portability of online content services in the internal market [2017] OJ L168/1; Directive 2018/1972; and Regulation 2019/1150 on promoting fairness and transparency for business users of online intermediation services [2019] OJ L186/57. 7 See above, n 6.

140  Bruno de Witte The, unavoidably quite synthetic, examination of the six fields listed above in sections III–VIII of this chapter will be preceded, in section II by a general impression of the state of play at the end of 2009, the starting date of our examination.

II.  The State of EU Services Law at the End of 2009 An account of the EU law of services, published in 2008, stated that, in this area, ‘legislative activity was patchy, and the task of bringing coherence to this aspect of the Treaty was left to the Court of Justice’. Moreover, the same authors argued, ‘there has been little by way of academic discussion and analysis in comparison with the very extensive work undertaken on goods or persons’.8 Most legal scholarship, at the time, dealt with the evolution in the case law of the Court interpreting the EC Treaty provisions on services,9 with some rare works examining EU services law in a more comprehensive way,10 alongside a specialised literature discussing the legislative measures dealing with specific services, such as banking and insurance, telecommunications and audiovisual services. The impact of EU law on services of general interest had also become a matter of interest,11 although most of the scholarship focused on the application of State aid law rather than the freedom to provide services; one major exception to this was the healthcare sector, as the case law of the Court of Justice of the European Union (CJEU), starting from the late 1990s, had brought cross-border access to healthcare within the scope of free movement of services. The main novelty at the time was the adoption, in 2006, of the Services Directive. The idea for such a cross-sectoral legislative instrument had not been contemplated in the framework of the ‘1992’ project for completing the internal market. The Commission’s White Paper on completing the internal market,

8 D Edward and N Nic Shuibhne, ‘Continuity and Change in the Law Relating to Services’ in A Arnull, P Eeckhout and T Tridimas (eds), Continuity and Change in EU Law: Essays in Honour of Sir Francis Jacobs (Oxford, Oxford University Press, 2008) 243. 9 A Biondi, ‘Recurring Cycles in the Internal Market: Some Reflections on the Free Movement of Services’ in A Arnull, P Eeckhout and T Tridimas (eds), Continuity and Change in EU Law: Essays in Honour of Sir Francis Jacobs (Oxford, Oxford University Press, 2008); S O’Leary and J Fernandez Martin, ‘Judicially-Created Exceptions to the Free Provision of Services’ in M Andenas and WH Roth (eds), Services and Free Movement in EU Law (Oxford, Oxford University Press, 2002); S Enchelmaier, ‘Always at Your Service (Within Limits): The ECJ’s Case Law on Article 56 TFEU (2006–11)’ (2011) 36 European Law Review 615; V Hatzopoulos, ‘The Court’s Approach to Services (2006–2012)’ (2013) 50 Common Market Law Review 459. 10 V Hatzopoulos, Regulating Services in the European Union (Oxford, Oxford University Press, 2012); C Calliess and S Korte, Dienstleistungsrecht in der EU (München, Beck, 2011). 11 M Krajewski, U Neergaard and J van de Gronden (eds), The Changing Legal Framework for Services of General Interest in Europe: Between Competition and Solidarity (The Hague, TMC Asser Press, 2009); JV Louis and S Rodrigues (eds), Les services d’intérêt économique général et l’Union européenne (Bruxelles, Bruylant, 2006); D Gallo, I servizi di interesse economico generale. Stato, mercato e welfare nel diritto dell’Unione europea (Milano, Giuffrè, 2010).

The Freedom to Provide Services  141 of 1985, is better remembered for what it said about goods than about services, perhaps because of its innovative and eye-catching emphasis on the removal of ‘physical barriers’ to trade (ie, border controls), which is more relevant to goods than to services. However, the main legal and economic content of the White Paper related to what was rather vaguely called the ‘removal of technical barriers’, and under that heading due attention was given to the common market for services. The Commission in fact stated at the time that ‘trade in services is as important for an economy as trade in goods’.12 In its White Paper, the Commission had proposed legislation in a selected number of crucial service sectors, namely financial services, transport, broadcasting and a rather vaguely defined group of communication services.13 In addition, ‘horizontal’ legislation to facilitate the supply of a range of different services was proposed under two separate headings: that of public procurement and that of recognition of diplomas. The Commission’s action programme was duly carried out in the following years, leading to the adoption prior to the 1 January 1993 deadline of a set of banking and insurance directives, directives liberalising the various modes of transport (land, water, air), the ‘Television without Frontiers’ Directive, a directive on public procurement, and two directives for recognition of diplomas. The legislative approach was not exhausted by the 1 January 1993 deadline. In fact, a fairly large number of services directives were adopted in the 1990s and early 2000s, including the telecommunications ‘package’ of four directives adopted in 2002,14 the Directive on electronic commerce 2000/31,15 and a whole host of financial services directives that quickly developed into a highly complex self-contained regime.16 In addition, the posted workers Directive was adopted in 1996.17 This was formally part of services law in terms of its legal basis, but it stood out from the previously mentioned instruments in that its primary aim was not so much the facilitation of cross-border service provision but rather its regulation

12 ‘Completing the Internal Market’, White Paper from the Commission to the European Council, 14 June 1985 COM(85) 310, 26. 13 ibid, 26–32. 14 A package of four directives dealing with telecommunications was adopted in 2002, all of which were published in [2002] OJ L108. They liberalised the communications markets but also imposed detailed harmonisation of the sector. See A de Streel, ‘Current and Future European Regulation of Electronic Communications: A Critical Assessment’ (2008) 32 Telecommunications Policy 722. 15 See the commentary published shortly after its adoption by A Lopez-Tarruella, ‘A European Community Regulatory Framework for Electronic Commerce’ (2001) 38 Common Market Law Review 1337. 16 Financial services became the object of a separate Commission action plan in 1999 COM(1999) 232 of 11 May 1999, and have since then led a legal life of their own. On the financial services directives, both pre- and post-1992, see V Hatzopoulos, Le principe communautaire d’équivalence et de reconnaissance mutuelle dans la libre prestation de services (Bruxelles, Bruylant, 1999) 413–50; E Wymeersch, ‘The Future of Financial Regulation and Supervision in Europe’ (2005) 42 Common Market Law Review 987; D Chalmers, C Hadjiemmanuil, G Monti and A Tomkins, European Union Law, 1st edn (Cambridge, Cambridge University Press, 2006) ch 18. 17 Directive 96/71 concerning the posting of workers in the framework of the provision of services [1997] OJ L18/1.

142  Bruno de Witte from the perspective of the social protection of workers. From the point of view of the firms providing cross-border services, this Directive made them comply with two different sets of labour law rules (those of their country of establishment and of the country where they post workers), which is why some critics argued at the time that this Directive did not, in fact, facilitate the free movement of services (as its legal basis requires), but hindered it.18 It took some time, after the 1992 programme, for the idea of a general directive on services to germinate in the Brussels corridors. The Conclusions of the Lisbon European Council, in March 2000, had requested a ‘strategy for the removal of barriers to services’ to be set out by the end of 2000. In response to this, the Commission produced on 29 December 2000 a strategy paper for services in which, for the first time, the idea of a general directive on services was canvassed, but only as one possible route for future action.19 It took some more time, until 13 January 2004, before the Commission produced its draft Directive on Services in the Internal Market, after it had received approving nods from the Council and the Parliament and from a number of interest groups. In that so-called ‘Bolkestein draft’, the Commission proposed a regulatory programme applying to the whole range of services (rather than a single sector) without attempting to list those services. The regulatory balance was decidedly tilted towards deregulation and full-scale liberalisation of trade in services. The basic principle underlying the text was that the laws of the host country would not apply to cross-border service provision, but instead the laws of the home country in which the service provider was established. The draft Directive did not aim at total liberalisation since it provided for measures of so-called horizontal harmonisation that come under the heading ‘quality of services’ and aim at the protection of consumers of commercial services. However, the Commission’s draft did not propose any harmonisation related to non-market concerns. In fact, it would have been difficult to address those nonmarket concerns in a horizontal directive dealing with the whole range of services, given that those concerns tend to be service-specific (for example: the concern for the administration of justice in relation to legal services, or the concern for cultural diversity in relation to broadcasting services, etc). The well-considered regulatory mix that characterised the earlier sector-specific directives was absent this time. This important regulatory shift remained unexplained in the Bolkestein draft. It presupposed a leap of faith which the Commission was prepared to make but which it could not convince the other political actors (represented in the Council and the Parliament) to make. The latter were simply not convinced that there was a huge untapped potential of cross-border provision of services that would be

18 For discussion of this point (and of the overall content of the Directive), see P Davies, ‘Posted Workers: Single Market or Protection of National Labour Law Systems?’ (1997) 34 Common Market Law Review 571. 19 ‘An Internal Market Strategy for Services’ COM(2000) 888 of 29 December 2000.

The Freedom to Provide Services  143 unleashed by a complete shift to home country control, whereas they did see the many regulatory drawbacks of such a shift. In that context, services became for the first time (but only briefly) a matter of high politics, as the European Council itself intervened to support the Parliament’s objections to the Commission’s original plan.20 The end version, as adopted in 2006, was more modest than the Commission had originally planned.21 Existing, more specific, services legislation (telecommunications, financial services, transport, information society, posting of workers, etc) was not affected by the new Directive, and many as yet unregulated service sectors (such as health, social services or gambling) were excluded from its scope. The main remaining coverage is business services, and travel, recreational and personal services, and even there the host countries were still allowed to regulate incoming services on a number of public interest grounds.

III.  The Fundamental Freedom to Provide Services In parallel to the ‘1992’ legislative programme, the European Court of Justice had refined its own approach to the free movement of services. In 1991, shortly before the completion of the internal market programme, it had clearly stated that the Treaty requires the abolition of any restriction, even if it applies without distinction to national providers of services and to those of other Member States, when it is liable to prohibit or otherwise impede the activities of a provider of services established in another Member State where he lawfully provides similar services.22

The existence of unnecessary ‘double burdens’ (in the country of origin and in the country of service provision) is at the heart of the Court’s scrutiny. This broad definition of the Treaty obligation imposed on States was compensated by the fact

20 European Council of 23–24 March 2006, Presidency Conclusions, para 57. The political history of the Services Directive is recounted by A Crespy, Qui a peur de Bolkestein? Conflit, résistances et démocratie dans l’Union européenne (Paris, Economica, 2012). See also J Flower, ‘Negotiating European Legislation: The Services Directive’ (2007) 9 Cambridge Yearbook of European Legal Studies 217; MD Jensen and P Nedergaard, ‘From “Frankenstein” to “Toothless Vampire”? Explaining the Watering Down of the Services Directive’ (2012) 19 Journal of European Public Policy 844; and B de Witte, ‘Setting the Scene: How Did Services Get to Bolkestein and Why?’ (2007) EUI Working Papers LAW 2007/20. 21 Directive 2006/123 on services in the internal market [2006] OJ L376/36. The Directive was abundantly discussed in the literature at the time of its adoption. Among many other contributions, see: C Barnard, ‘Unravelling the services Directive’ (2008) 45 Common Market Law Review 323; S Evans, ‘The Services Directive: (Too) Great Expectations? An Initial Overview of the Rights and Obligations under the Services Directive’ in J van den Gronden (ed), The EU and WTO Law on Services (Alphen aan den Rijn, Kluwer Law International, 2009) 7; U Neergaard, R Nielsen and L Roseberry (eds), The Services Directive: Consequences for the Welfare State and the European Social Model (Copenhagen, DJOF Publishing, 2008). 22 Case C-76/90 Säger v Dennemeyer EU:C:1991:331, para 12. The Säger definition was never overruled and often confirmed in quasi-identical terms until today; see, for instance, Case C-91/13 Essent Energie Productie EU:C:2014:2206, para 44.

144  Bruno de Witte that national restrictions, if indistinctly applied, may be kept in place if they are necessary to achieve one of a long list of mandatory requirements. The mandatory requirements test, as well, was firmly put in place by the early 1990s. The judicial approach was inherently open-ended, and therefore fit to identify new forms of trade in services as well as newly emerging impediments to such trade; it also covered gaps left even where internal market legislation had been adopted.23 In the years after the ‘completion’ of the 1992 single market programme, the number of CJEU judgments dealing with the freedom to provide services increased.24 It is not entirely clear whether this trend indicated a growing awareness of long-existing restrictions to services trade, or whether it reflected an increase of such obstacles in Member State laws and regulations. In some ways, the Court acted as a regulator: in service domains which were not the object of EU legislation, or in which that legislation was insufficiently clear, the Court developed the applicable legal regime in a piecemeal, case-by-case, manner.25 Today, the freedom to provide and receive services, now laid down in Article 56 TFEU, is still the main provision of primary EU law dealing with services. However, its central role in legal and policy practice seems to have declined over the past 10 years. One could say that it now merely plays a residual role, filling the gaps left by the Services Directive and by the increasing number of specific legislative acts. It has been convincingly argued that many questions which would have been resolved on the basis of the fundamental freedoms 40 years ago are subject to EU harmonisation nowadays. It is on these areas of substantive harmonisation by the EU legislature on which the ECJ concentrates and where it develops its vision of good society.26

One remaining ‘gap’ is the legal regulation of gambling and gaming: in the absence of EU legislation, and given the wide variety of national laws regulating these services, the role of Article 56 TFEU remains central. Preliminary references on this subject were quite frequent in the first decade of this century27 and have continued to be lodged, although at a slower pace than before.28 Significantly, the

23 eg, there were a number of cases, in the first years of the century, about the posting of workers which the Court examined on the basis of the EC Treaty because the Posted Workers Directive did not, for one reason or another, apply to the facts of the case. 24 V Hatzopoulos and TU Do, ‘The Case Law of the ECJ concerning the Free Provision of Services: 2000–2005’ (2006) 43 Common Market Law Review 923, 923. 25 For a comprehensive analysis of such judge-made regulation across a number of service domains, see V Hatzopoulos, Regulating Services in the European Union, above n 10, ch 5. 26 D Thym, ‘The Constitutional Dimension of Public Policy Justifications’ in P Koutrakos, N Nic Shuibhne and P Syrpis (eds), Exceptions from EU Free Movement Law (Oxford, Hart Publishing, 2016) 184. 27 See the overview of the case law by S Van den Bogaert and A Cuyvers, ‘“Money for Nothing”: The Case Law of the EU Court of Justice on the Regulation of Gambling’ (2011) 48 Common Market Law Review 1175; and F de Witte, ‘The Constitutional Quality of the Free Movement Provisions: Looking for Context in the Case Law on Article 56 TFEU’ (2017) 42 European Law Review 313, 331–36. 28 Among the main cases of the past 10 years are Case C-156/13 Digibet EU:C:2014:1756 and Case C-98/14 Berlington Hungary EU:C:2015:386.

The Freedom to Provide Services  145 Commission decided, in 2017, to close its ongoing infringement procedures in this area, as it no longer considered this to be a priority question.29 Apart from the gambling cases, occasional references reach the Court of Justice, on the initiative of service providers who challenge national rules before national courts, or who are the object of criminal or administrative sanctions against which they defend themselves by invoking the freedom to provide services. The main legal issues are the usual ones: issues about the scope of application of Article 56 TFEU (is there a service provided for remuneration? is there a cross-border restriction?) and issues about the proportionality of restrictive national measures. However, and strikingly, we have not seen any new landmark cases about Article 56.30 There have been a few Grand Chamber judgments dealing with the Treaty freedom to provide services but, on a closer look at them, it is not always clear why a Grand Chamber ruling was deemed appropriate, given that no major issues of principle seemed at stake. One such case dealt with an obligation imposed by Hungarian law on foreign-based advertisers to lodge a tax declaration which domestic advertisers did not have to produce. This was held to be compatible in principle with Article 56, but the sanction for non-compliance was held to be excessive and therefore in breach of Article 56.31 Two Grand Chamber judgments dealt with the exclusion of ‘purely internal’ situations from the scope of application of Article 56,32 and in some other Grand Chamber judgments the freedom of services was discussed by the Court but was not really the central issue of the case.33 Generally speaking, the Court seems to leave considerable deference to choices made by national governments and to the fact finding by national courts.34 There are roughly equal numbers of judgments in which the Court finds national measures compatible or incompatible with Article 56. So, all in all, the view that the Court of Justice still functions as an ‘engine of liberalisation’35 seems less persuasive than before, at least in the field of services.

29 European Commission, ‘Monitoring the application of Union law – Annual report 2017’ COM(2018) 540 of 12 July 2018, 5. 30 One may note, for example, that a recent textbook account of the freedom to provide services, containing some 24 extracts from CJEU judgments, has only one extract from a post-2009 judgment, namely the Visser case which is in fact about establishment rather than services (Chalmers, Davies and Monti, above n 2, ch 17). 31 Case C-482/18 Google Ireland EU:C:2020:141. See also the similarly unexciting facts of the Grand Chamber judgment in Case C-179/14 Commission v Hungary EU:C:2016:108 and of the Grand Chamber judgment in Case C-33/17 Cepelnik EU:C:2018:896. 32 Case C-268/15 Ullens de Schooten EU:C:2016:874; and Case C-591/15 The Gibraltar Betting and Gaming Association Limited and The Queen EU:C:2017:449. 33 Case C-403/08 Football Association Premier League and Others EU:C:2011:631; and Case C-591/17 Austria v Germany EU:C:2019:504. 34 This corresponds to the findings, relating to free movement law in general, of J Zglinski, ‘The Rise of Deference: The Margin of Appreciation and Decentralized Judicial Review in EU Free Movement Law’ (2018) 55 Common Market Law Review 1341. 35 N Dunne, ‘Liberalisation and the Pursuit of the Internal Market’ (2018) 43 European Law Review 803, 809–10.

146  Bruno de Witte

IV.  The General Services Directive The Services Directive has a broad coverage, as it applies to both the ‘provision of services’ and the ‘right of establishment’, as these are defined by the TFEU. As regards the provision of services (in the sense of the TFEU), the central substantive rule of the Directive is in Article 16. Article 16(1) states that ‘Member States shall respect the right of providers to provide services in a Member State other than that in which they are established’. This promise of sweeping liberalisation is immediately tempered by the next sentence which makes clear that the host States may continue to apply their laws and regulations, as long as these are non-discriminatory, necessary and proportional – as the Court of Justice has consistently held in its case law on services. Article 16 can thus be seen as a restatement of existing courtmade law. Paragraph 2 of Article 16 adds a number of prohibited requirements which, again, probably correspond to the Court’s views of what is permissible, but the specification brings more clarity. Paragraph 3 of Article 16 then specifies the grounds on which the host countries can base their proportionate requirements, namely: reasons of public policy, public security, public health, the protection of the environment, and rules on employment conditions. This list is much shorter than the list of mandatory requirements that the Court of Justice has come to recognise in the course of the years. Neither the text of the Directive nor its preamble explain whether (and if so, why) the Directive attempted to modify the Court’s case law on this point. Indeed, the preamble gives the impression that no change was intended since it defines, in recital 40, the concept of ‘overriding reasons relating to the public interest’ with reference to the full list of mandatory requirements recognised by the Court of Justice, without adding that the crucial Article 16 only refers to some of those ‘overriding reasons’, leaving the rest in a legal limbo.36 So, although the ‘revolution’ proposed in the Bolkestein draft was turned back by the Council and Parliament, the laborious compromise reached during the co-decision procedure could nevertheless have produced (albeit perhaps unintentionally) a slight deregulatory shift compared with existing services law. Even if the Services Directive does not radically change EU services law, compared with the pre-existing interpretation of primary law by the CJEU, that does not make it a useless instrument. It was rightly noted that ‘a mere change in the design of a legal provision without touching upon its content, namely from established case law of the ECJ to codification in EC secondary legislation, can have far-reaching economic implications’.37 Indeed, the text of the Directive, and especially its transposition

36 For further discussion of this point see M Klamert, ‘Way To Go? More on the Services Directive and the Fundamental Freedoms’ (2009) 64 Zeitschrift für Öffentliches Recht 335, 349 ff; and J Snell, ‘Freedom to Provide Services in the Case Law and in the Services Directive: Problems, Solutions and Institutions’ in U Neergaard, R Nielsen and L Roseberry (eds), The Services Directive: Consequences for the Welfare State and the European Social Model (Copenhagen, DJØF Publishing, 2008) 187–91. 37 H Baldinger and N Maydell, ‘Legal and Economic Issues in Completing the EU Internal Market for Services: An Interdisciplinary Perspective’ (2009) 47 Journal of Common Market Studies 693, 714.

The Freedom to Provide Services  147 into national law, makes the legal rules more visible for both the service operators and the national administrations. In particular, the Directive requires states to put in place ‘points of single contact’ that help providers to complete administrative formalities, and also organises detailed administrative cooperation between the Member State authorities.38 Gareth Davies went so far as saying that ‘the Directive is not of great substantive import, but is primarily a communicative measure, which in turn may make the substantive rules on free movement of services – which are greatly composed of mutual recognition – effective’.39 However, it is difficult to measure whether such an economic effect has taken place, ie, whether there is an increase in cross-border services in the European Union and, if so, whether that increase is due, in some part, to the Services Directive. The Commission published in 2013 a series of reports assessing the implementing measures taken by each Member State, but since then little evidence has emerged about the concrete impact of the Directive on the cross-border provision of services. Despite this absence, the Commission sought to strengthen the practical impact of the Directive by proposing, in January 2017, a ‘Services Package’, consisting of three horizontal legislative measures: (i) the creation of a European services e-card, delivered by the home country, which would allow its holder to skip some of the administrative hurdles in the host States so as to ensure speedy access to the services market of the host State; (ii) the introduction of a proportionality test to be accomplished by the Member States prior to adopting or amending their rules on professional services; and (iii) a stricter notification requirement for all draft national laws affecting the provision of services.40 The proportionality test Directive was adopted by Council and Parliament in 2018.41 It codifies the way in which the Court of Justice assesses the proportionality of restrictive State measures and aims to make the States self-enforce that test prior to the adoption of new laws and regulations. The other two proposals of the ‘Services Package’ have not yet been adopted, and the proposal for the ‘services e-card’ was even roundly rejected by the European Parliament. Immediately after the adoption of the Services Directive, it had been predicted that, given the lack of clarity of many of its key provisions, things would have to be clarified through litigation, and the Court of Justice would be called to play an active role in this respect.42 However, we have seen relatively few preliminary references or infringement cases about the Services Directive.43 One of the few important 38 See respectively Arts 5–8 and 28–36. 39 G Davies, ‘Trust and Mutual Recognition in the Services Directive’ in I Lianos and O Odudu (eds) Regulating Trade in Services in the EU and the WTO: Trust, Distrust and Economic Integration (Cambridge, Cambridge University Press, 2012) 100. 40 The three legislative proposals were laid down in COM(2016) 821–24. 41 Directive 2018/958 on a proportionality test before adoption of new regulation of professionals [2018] OJ L173/25. 42 Snell, above n 36, 197. 43 For discussion of the CJEU case law about the Services Directive, see P Delimatsis, ‘From Sacchi to Uber: 60 Years of Services Liberalization, Ten Years of the Services Directive in the EU’ (2018) 37 Yearbook of European Law 188, 229–48.

148  Bruno de Witte developments in the case law relates to the establishment chapter, as the Court of Justice held in the Visser judgment of 2018 that the provisions of that chapter also apply to purely internal situations, ie, to companies starting a new service activity within their own country.44 In contrast, the chapter dealing with the provision of services presupposes a cross-border element, as the right protected by Article 16 of the Directive is the freedom to provide services between EU countries. One of the reasons why there has been relatively little litigation about the Directive might be the fact that it includes very few harmonisation measures. Harmonisation of national laws (which typically applies to both cross-border and internal services) is left for specific EU legislation in which the EU institutions can decide on the most appropriate mix between the aim of liberalising the European services market and that of regulating the operation of that market for public interest reasons. In other words, sectoral regulation (to which the next section turns) has remained as useful and necessary as before the adoption of the Services Directive.

V.  Sectoral Regulation The sectoral logic of EU services law was already at work in the original text of the EEC Treaty, when transport (at that time probably considered the most important cross-border service) was excluded from the scope of the services chapter, and was covered by a Treaty chapter all of its own. This separate treatment of the transport sector was never reversed by later Treaty amendments, so that still today transport services are subject to their own specific rules of primary and secondary EU law, and neither the general Treaty provisions on services nor the Services Directive apply to them.45 This special status played a crucial role in the recent Uber case.46 The Court’s decision to qualify Uber services as predominantly transport services rather than information society services meant that neither the e-commerce Directive nor the general freedom to provide services applied and that Member States (or cities) could continue to apply restrictive regulations, which might otherwise have fallen foul of the freedom to provide services. For a long time, the legislative approach to services remained sector-specific. The Commission identified particular service sectors, in which national regulation was particularly intrusive and/or in which EU-based liberalisation promised to bring particularly important economic benefits. In those selected domains, the Commission each time proposed a particular mix of liberalisation and harmonisation (or, to use largely equivalent terms, deregulation and re-regulation), which 44 Joined Cases C-360/15 and C-31/16, X and Visser EU:C:2018:44. Another recent judgment relating specifically to the establishment chapter of the Directive is Case C-393/17 Kirschstein EU:C:2019:563. 45 See respectively Art 58 (1) TFEU and Art 2(2)(d) of the Services Directive. For the current state of play in EU transport law, see P Mendes de Leon, ‘Transport Policy’ in PJ Kuijper et al (eds), The Law of the European Union, 5th edn (Alphen aan en Rijn, Kluwer Law International, 2018). 46 Case C-434/15 Elite Taxi v Uber Systems Spain EU:C:2017:981.

The Freedom to Provide Services  149 then became the object of debate within the Council and the European Parliament, leading to directives striking a particular regulatory balance. The principle of home country control was often adopted in those instruments, in order to remove the requirement of an authorisation by the host State, which has traditionally been a major impediment to the cross-border provision of services.47 However, the adoption of that principle was counterbalanced by an amount of harmonisation which reduced the discrepancy between the laws of the home and host countries. Indeed, the fact that the EU instruments, in terms of their legal basis, should aim at facilitating cross-border provision of services is not an impediment to the adoption of common rules, in those instruments, that apply to all services in that sector of activity, whether cross-border or not. In that sense, sector-specific legislation usually has a broader personal scope than the Treaty freedom, as it applies not only to cross-border provision but also to domestic services. The balance struck by the EU legislator was (and is) sometimes criticised from both sides of the ideological spectrum, for being too deregulatory or (on the contrary) for being too interventionist. Some scholars have argued that, taken as a whole, internal market legislation post-1992 was marked by ‘an “advocacy alliance” of civic interest groups, stringent-standard producers, several member state governments, the Parliament, and often the Commission in favour of more stringent standards’,48 whereas other commentators have found a strong pro-market bias in some or most of this services legislation. This ideological battle was fought on a case-by-case basis with regard to each particular sector in which the Commission had set the policy agenda by identifying a need for common European rules. This model, of home State control combined with a degree of substantive harmonisation, was in place in a number of sectors by 2010, namely audiovisual services, banking and insurance, and transport.49 Since 2010, no new sectors have been opened up for the country-of-origin approach,50 and in those sectors where it was well established, the trend has been towards more substantive harmonisation rather than deregulation. This logic was taken furthest, in recent years, in the banking sector. This was one of the first economic sectors in which the country-oforigin approach was introduced.51 However, the legislative regime of that sector has

47 See KE Sørensen, ‘The Country-of-Origin Principle and Balancing Jurisdiction between Home Member States and Host Member States’ (2019) 30 European Business Law Review 37. 48 AR Young, ‘The Single Market – A New Approach to Policy’ in H Wallace, W Wallace and M Pollack (eds), Policy-Making in the European Union, 5th edn (Oxford, Oxford University Press, 2005) 107. 49 For an overview of the legislative evolution across service sectors up until 2010, see Hatzopoulos, Regulating Services in the European Union, above n 10, ch 6. 50 According to Sørensen, above n 47, 43, ‘most secondary law using the country-of-origin principle in recent years has been adopted to refine and develop the principle’s application in areas which were already dedicated to the use of the principle’. 51 See, for those early developments, E Lomnicka, ‘The Home Country Control Principle in the Financial Services Directives and the Case Law’ in M Andenas and WH Roth (eds), Services and Free Movement in EU Law (Oxford, Oxford University Press, 2002); and M Ortino, ‘The Role and Functioning of Mutual Recognition in the European Market of Financial Services’ (2007) 56 International & Comparative Law Quarterly 309.

150  Bruno de Witte undergone a profound transformation as a result of the financial crisis starting in 2008. There is now very detailed substantive harmonisation of all financial services activities (not just the cross-border activities) and the substantive regulation is accompanied by the creation of new European-level actors in charge of day-today regulation and supervision of the banks: the European Banking Authority, the Single Supervisory Mechanism and the Single Resolution Board. This rapid evolution has resulted in ‘Banking Union’ becoming a separate policy field from EU services law, and indeed a specialised field of academic writing.52 Even though one could say that ‘the essence of the banking union is merely and simply the required positive integration to make the single market for banking services function properly’,53 this is no longer the way in which this area of EU law is understood by the political actors. In terms of legal practice, we can see that some banking legislation is still based on a service-specific legal basis, but other elements (such as the banking resolution mechanism) have Article 114 TFEU as their broader legal basis.54 In policy terms, the regulation of financial services now tends to be presented by the Commission as separate from the single market strategy and to form part, instead, of a new policy domain called ‘Capital Markets Union/Banking Union’.55 The other service sector in which EU legislative activity has continued unabated is audiovisual services. The original ‘Television without Frontiers’ Directive from 198956 had been amended in 1997 and again in 2007, followed by a codification in 2010 under the name ‘Directive on audiovisual media services’.57 A new revised version was adopted in 2018.58 The most striking thing in the evolution of the legislative regime over the years is the constant adaptation to the changing economic and technical reality. The Directive used to apply only to traditional over-the-air television, but was gradually expanded to new forms of diffusion as a result of what was called ‘media convergence’,59 and the latest 2018 revision even extended the regulatory scope to video-sharing platforms.60 52 Among the main recent studies of this legal field, see N Moloney, EU Securities and Financial Markets Regulation (Oxford, Oxford University Press, 2014); F Marcucci (ed), L’Union bancaire (Bruxelles, Bruylant, 2016); MP Chiti and V Santoro (eds), The Palgrave Handbook of European Banking Union Law (Cham, Springer International, 2019). 53 J Pelkmans, ‘Why the Single Market Remains the EU’s Core Business’ (2016) 39 West European Politics 1095, 1107. 54 Compare Directive 2014/49 on deposit guarantee schemes [2014] OJ L173/149 (legal basis: Art 53(1) TFEU) with Directive 2014/59 establishing a framework for the recovery and resolution of credit institutions and investment firms [2014] OJ L173/190 (legal basis: Art 114 TFEU). 55 See, in this sense, Annex 1 of the Commission Communication, ‘The Single Market in a changing world: A unique asset in need of renewed political commitment’ COM(2018) 772. 56 Directive 89/552 [1989] OJ L298/23. 57 Directive 2010/13 [2010] OJ L 95/1. 58 Directive 2018/1808 [2018] OJ L303/69. 59 See, for a discussion of the expanded scope of the 2007 version of the Directive, R Craufurd Smith, ‘Media Convergence and the Regulation of Audiovisual Content: Is the European Community’s Media Services Directive Fit for Purpose?’ (2007) 60 Current Legal Problems 238. 60 L Woods, ‘Video-sharing Platforms in the Revised Audiovisual Media Services Directive’ (2018) 23 Communications Law 127. For a complete discussion of the 2018 reform, see JM Martinez Otero, ‘Un nuevo marco regulador para el sector audiovisual en Europa: La Directiva 2018/1808 en el contexto

The Freedom to Provide Services  151 Throughout its existence, this piece of legislation remained based on the internal market objective of facilitating the provision of media services from one European country to the others, but it also pursued, from the beginning, several non-market regulatory objectives by means of minimum harmonisation of national media laws. Guaranteeing free movement is of course a central concern of the Directive and, indeed, constitutes the threshold condition for the legitimate use of Articles 53(1) and 62 TFEU as its legal basis, but the true nature and main characteristic of the Directive is that it enacts a complex and hard-fought balance between facilitating the operation of a Europe-wide market for media services and imposing common regulatory standards on those services. The Directive always had a cultural policy function, as it seeks to promote the diffusion of audiovisual content of European origin by means of diffusion quotas or financing obligations for media companies.61 Apart from its cultural policy provisions, the Directive contains a large number of other norms addressing non-market concerns, including: a duty to protect minors against unsuitable media content; a provision authorising States to impose free-to-air access to major events; a provision that broadcasting and online media should not contain hate speech; and that their services should be accessible to people with a visual or hearing disability, etc.

VI.  Horizontal Regulation Services, both cross-border and domestic, are regulated by numerous EU legislative acts that are not sector-specific. Much of that legislation applies to service providers as to other economic operators, and service firms are often not singled out in the text of those instruments. This is the case with copyright legislation, consumer protection, data protection, anti-discrimination and employment legislation, to name just the main ones. In addition, there is horizontal internal market regulation in which services are expressly regulated together with other market freedoms; this is the case with the EU rules on professional qualifications that govern the access to the market of professionals both through freedom of establishment and through freedom to provide services. Also, the general legal framework of standardisation now applies to services as well as to goods,62 but in the absence of specific EU legislation giving legal force to the standards adopted by the relevant bodies for a particular type of service, standardisation has not had a measurable impact yet on the free movement of services.63 The EU’s public de la convergencia mediatica y el mercado unico digital’ (2019) 63 Revista de Derecho Comunitario Europeo 539. 61 See R Craufurd Smith, ‘EU Media Law: Cultural Policy or Business as Usual?’ in V Vadi and B de Witte (eds), Culture and International Economic Law (London, Routledge, 2015). 62 The scope of the standardisation regime was extended to services by Regulation 1025/2012 on European standardisation [2012] OJ L316/12. 63 See B van Leeuwen, European Standardisation of Services and its Impact on Private Law (Oxford, Hart Publishing, 2017).

152  Bruno de Witte procurement regime, as last revised in 2014, comprises a general directive that covers the procurement of services alongside the procurement of goods, but also a special (but similar) regime for the procurement by public utilities, as well as a special regime for concessions, the latter two being legislation that applies specifically to the domain of services.64 The E-commerce Directive is a particularly important piece of horizontal legislation applying specifically to services.65 It applies to a particular mode of delivery of services (namely by electronic means)66 but is horizontal, in that it applies to all market sectors where that mode of delivery is used. Compared with the general services Directive, it stands out by regulating both the provider’s home State (which is not allowed to require prior authorisation for the service) and the receiving State (which is only allowed to restrict the service under narrowly defined conditions). The effective application of the Directive is supported by an obligation for the Member States to notify the European Commission of new rules affecting information society services.67 The question whether platform services are covered by the e-commerce Directive became very controversial in recent years.68 The key issue, in this respect, is whether the information society service offered by the platform is sufficiently distinguishable from the underlying ‘real life’ service (transportation, accommodation or other) so as to justify its protection by the market-friendly rules of the e-commerce Directive. In two landmark cases, the Court of Justice held that the activity of Uber should not be classified as an information society service but rather as an urban transportation service,69 whereas in the Airbnb case, it held that the services offered by that company were indeed information society services that could be distinguished from the underlying home-sharing service, so that Airbnb could invoke the e-commerce Directive against restrictive measures taken by the French authorities.70 This is an area of the law that is still in flux. The adoption of a dedicated legislative instrument for

64 Respectively: Directive 2014/24 [2014] OJ L94/65; Directive 2014/25 [2014] OJ L94/243; and Directive 2014/23 [2014] OJ L94/1. On the current public procurement regime, see S Schoenmaekers, ‘Public Procurement’ in PJ Kuijper et al (eds), The Law of the European Union, 5th edn (Alphen aan den Rijn, Kluwer Law International, 2018). 65 Directive 2000/31 on certain legal aspects of information society services, in particular electronic commerce, in the internal market [2000] OJ L178/1. 66 It applies to ‘any service normally provided for remuneration, at a distance, by electronic means and at the individual request of a recipient of services’ (Directive 2000/31, Art 2(a)). 67 Directive 98/34 laying down a procedure for the provision of information in the field of technical standards and regulations and of rules on information society services [1998] OJ L204/37. See the recent enforcement of that notification obligation in Case C-299/17 VG Media EU:C:2019:716. 68 For general discussion of the application of EU law to platform services, see V Hatzopoulos and S Roma, ‘Caring for Sharing? The Collaborative Economy under EU Law’ (2017) 54 Common Market Law Review 81. 69 Case C-434/15, above n 46. See the case comment by M Finck, ‘Distinguishing Internet Platforms from Transport Services: Elite Taxi v Uber Spain’ (2018) 55 Common Market Law Review 1619. 70 Case C-390/18 Criminal proceedings against X EU:C:2019:1112.

The Freedom to Provide Services  153 platform services might make sense,71 but the Commission has not, so far, moved in that direction. Even more controversial was the recent adoption of another horizontal instrument, namely the revised posted workers Directive. The original enactment of that Directive, in 1996, was, already then, seen as a controversial use of internal market competences to favour (also) social policy ends. After the restrictive interpretation given to the Directive in Laval72 and some later CJEU cases, a vivid debate in the literature emerged as to whether this Directive, as interpreted by the Court, was causing regulatory competition between labour law systems rather than ensuring effective protection of the posted workers.73 Trade unions in Western European countries, in particular, argued that the Directive offers an incentive to ‘social dumping’, especially since the ‘big bang’ enlargement of the EU in 2004 which widened the wage gap between European countries and thus also increased the competitive advantage of service providers from low-wage countries. A first legislative intervention occurred in 2014, when a directive was adopted aimed specifically at ensuring the effective enforcement of the posted workers Directive.74 One of its main aims was to define ‘genuine establishment’ and ‘genuine posting’ so as to combat abuses by companies locating abroad with the sole purpose of circumventing domestic employment law. After continuing pressure from the side of the European Parliament and from some governments of host countries, the Juncker Commission decided to engage in a substantive revision of the 1996 Directive.75 Next to some other changes, the most important addition was the rule that posted workers should receive equal remuneration with the host country workers (and not just the minimum wage, as under the original directive).

71 For an argument in this sense, see M Inglese, ‘The Collaborative Economy Legal Conundrum: A Way Forward Through Harmonization’ (2018) 45 Legal Issues of Economic Integration 375. See also M Finck, ‘Digital Co-regulation: Designing a Supranational Legal Framework for the Platform ­Economy’ (2018) 43 European Law Review 47. 72 Case C-341/05 Laval un Partneri EU:C:2007:809. 73 Among the contributions to that rich debate, see S Deakin, ‘Regulatory Competition after Laval’ (2009) 10 Cambridge Yearbook of European Legal Studies 581; O De Schutter, ‘Transborder Provision of Services and ‘Social Dumping’: Rights-Based Mutual Trust in the Establishment of the Internal Market’ in I Lianos and O Odudu (eds) Regulating Trade in Services in the EU and the WTO: Trust, Distrust and Economic Integration (Cambridge, Cambridge Universoty Press, 2012) 349; C Kilpatrick, ‘Internal Market Architecture and the Accommodation of Labour Rights’ in P Syrpis (ed), The Judiciary, the Legislature and the EU Internal Market (Cambridge, Cambridge University Press, 2012); S Garben, ‘The Constitutional (Im)balance between ‘the Market’ and ‘the Social’ in the European Union’ (2017) 13 European Constitutional Law Review 23, 33–41. 74 Directive 2014/67 on enforcement of the posted workers directive [2014] OJ L159/11. This instrument is part of a wider EU move to curb the abusive exercise of internal market freedoms by so-called letterbox companies; see KE Sørensen, ‘The Fight against Letterbox Companies in the Internal Market’ (2015) 52 Common Market Law Review 85. 75 European Commission, ‘Proposal for a Directive amending Directive 96/71 concerning the posting of workers in the framework of the provision of services’ COM(2016) 128 of 8 March 2016.

154  Bruno de Witte Although the new text would ostensibly benefit the workers from low-wage countries (as their salaries during the time of posting would increase), the proposal raised the opposition precisely from the low-wage countries of Central and Eastern Europe, as they were concerned that their firms would lose business opportunities on Western European markets, especially in the construction sector. After the Commission tabled its proposal, a group of national parliaments from Central and Eastern European countries waved a ‘yellow card’ for breach of the principle of subsidiarity, but the Commission (acting under pressure from Western European countries) decided to confirm its proposal as it stood. The co-decision procedure moved on, and the revision of the Directive was finally agreed by Council and European Parliament at the end of May 2018, after some limited changes to the Commission’s original draft.76 As the Central European states who opposed the revision could not muster a blocking minority to stop the approval of the text in the Council, most of them agreed with the lightly amended final version. Only Hungary and Poland voted against adoption of the Directive and brought actions for annulment of the new Directive, arguing, among other things, that it infringed the Treaty freedom to provide services.77 This case will provide the Court of Justice with an opportunity to consider the extent to which internal market legislation can serve to regulate (rather than deregulate) the operation of the market for services.

VII.  Services of General Interest The provision of public services in the Member States became gradually impacted, over the years, by EU law rules. A crucial provision in this respect was (and is) Article 106 TFEU (formerly Article 86 EC) that brings what the Treaty text calls ‘services of general economic interest’ within the scope of application of EU law, whilst allowing the Member States to keep them under special regimes. The battle over the limits of Member State autonomy in organising their public services was fought mainly in the area of State-aid law, where the financial support to public services came under close scrutiny by the Commission; but Article 106 also affects the application of free movement law, including the law on the free movement of services.78 There is an important distinction in this respect between (i) the public

76 Directive 2018/957 [2018] OJ L173/16; for discussion of the negotiation process and the content of the new directive, see R Zahn, ‘Revision of the Posted Workers Directive: A Europeanisation Perspective’ (2017) 19 Cambridge Yearbook of European Legal Studies 187; P Van Nuffel and S Afanasjeva, ‘The Posting of Workers Revised: Enhancing the Protection of Workers in the Cross-border Provision of Services’ (2018) 3 European Papers 1401. 77 Case C-620/18 Hungary v European Parliament and Council (pending case). 78 For a sketch of the early stages of interference by EU law with the autonomy of the Member States in this field, see D Damjanovic and B de Witte ‘Welfare Values and Welfare Integration under the Lisbon Treaty’ in U Neergaard, L Roseberry and R Nielsen (eds), Integrating Welfare Functions into EU Law: From Rome to Lisbon (Copenhagen, DJØF Publishing, 2009); and V Hatzopoulos, Regulating Services in the European Union, above n 10, ch 2.

The Freedom to Provide Services  155 utilities (such as gas, electricity, telecommunications and postal services) whose economic character is undisputed and that were the object of liberalisation, in part through the case law of the Court of Justice, but above all through specific secondary legislation; and (ii) the welfare and social services, such as healthcare, education, social assistance and social housing. The economic character of the latter category is much less obvious, and the question has often arisen whether free movement law (and competition law) should apply to them at all. The Court’s case law had brought parts of the welfare sector within the scope of application of internal market law. This was the case, most prominently, with health services. Although the Court acknowledged that the regulation of access to medical services was part of the social security system, this fact alone was not enough to keep this sector of activity outside the scope of internal market law. In its early patient mobility cases, the Court held that ‘Community law does not detract from the powers of the Member States to organise their social security systems’ but immediately added that the ‘Member States must nevertheless comply with Community law when exercising those powers’ and that ‘the special nature of certain services does not remove them from the ambit of the fundamental principle of freedom of movement’.79 The Court pursued a consistent line of treating patients going abroad to seek medical help as ‘receivers of services’, who can invoke the Treaty provisions on the free movement of services, even when the service provided to them is part of the social security system and even when it is provided without any direct payment by the receiver. It is enough that ‘someone’ remunerates the service, even if that someone is a public healthcare fund. At the same time, the CJEU has been rather lenient at the justification stage, by accepting that those restrictions may be justified if they are required to preserve the financial balance of a country’s social security system.80 The Court thus accepted an economic interest as a good reason to restrict a fundamental freedom, which is rather unusual in its internal market case law.81 Public services other than healthcare were not so ‘fully’ covered by internal market law.82 The CJEU found a restriction of the freedom to provide services only where the national welfare system was organised in such a way that it provides services which could equally be offered by a commercial provider based in 79 Case C-158/96 Kohll v Union des caisses de maladie EU:C:1998:171, respectively paras 17, 19 and 20. 80 See, eg, Case C-368/98 Vanbraekel and Others EU:C:2001:400, para 47. For a comprehensive view of the early case law, see V Hatzopoulos, ‘Health Law and Policy: The Impact of the EU’ in G de Búrca (ed), EU Law and the Welfare State: In Search of Solidarity (Oxford, Oxford University Press, 2005); and T Hervey, ‘The Current Legal Framework on the Right to Seek Health Care Abroad in the European Union’ (2007) 9 Cambridge Yearbook of European Legal Studies 261. 81 N Nic Shuibhne and M Maci, ‘Proving Public Interest: The Growing Impact of Evidence in Free Movement Case Law’ (2013) 50 Common Market Law Review 965, 997 ff. 82 In fact, the line of case law on patient mobility has been criticised for its contribution to the ‘blurring of the economic and welfare spheres at the EU legal level and hence to the legal uncertainties … on the application of the EU market rules to the Member States’ welfare regimes’. (D Damjanovic, ‘The EU Market Rules as Social Market Rules: Why the EU Can Be a Social Market Economy’ (2013) 50 Common Market Law Review 1685, 1701.)

156  Bruno de Witte another EU State. In such cases, a national policy choice to exclude the operation of the market may be challenged on internal market grounds. An example of this approach is the Kattner judgment from 2009.83 That case concerned a German law providing for the compulsory affiliation of employers in a particular economic sector to a social insurance scheme for accidents at work, based on the principle of solidarity. The Court considered that this scheme restricted the possibility for foreign insurance companies to offer contracts covering some of the risks covered by the scheme, and that it therefore created a restriction to the freedom to provide services. In the next step of the reasoning, when examining possible justifications for restrictions to free movement, the Court invariably accepts welfare policy objectives as overriding reasons in the public interest that may offer a justification,84 but then the final outcome of the assessment depends, as usual in internal market cases, on the proportionality of the concrete national measure.85 All the developments described so far had taken place before 2010. By that date, the battle over the liberalisation of the utilities was over. Their liberalisation was accompanied by detailed rules for ‘universal service’ provision to consumers, a cross-sectoral legal concept that was developed in the 1990s, first used in European regulation of telecommunication, and later extended to postal services and electricity.86 This regulatory approach seeks to ensure that the creation of an open market with undistorted competition preserves, at least to some extent, the public interest mission of those services.87 However, the application of internal market to welfare services remained controversial. As the Monti Report of 2010 observed, ‘the place of public services within the single market has been a persistent irritant in the European public debate’.88 Welfare and social services were entirely excluded from the scope of application of the Services Directive.89 The Lisbon Treaty could

83 Case C-350/07 Kattner Stahlbau EU:C:2009:127, paras 74–83. 84 Kattner judgment, above n 82, para 85; similarly, in the case of a social housing scheme, Case C-567/07 Woningstichting Sint Servatius EU:C:2009:593, para 31; and Joined Cases C-197/11 and C-203/11 Libert and Others EU:C:2013:288, paras 52 and 67. 85 A recent case in which the Court found the national restriction to be disproportional is Case C-171/17 Commission v Hungary EU:C:2018:888 (monopoly entrusted to a domestic mobile payment system found to be a violation of Art 56 TFEU). 86 H-W Micklitz, ‘Universal Services: Nucleus for a Social European Private Law’ in M Cremona (ed), Market Integration and Public Services in the European Union (Oxford, Oxford University Press, 2011). For an overview of universal service obligations in the various sectors see W Sauter, Public Services in EU Law (Cambridge, Cambridge University Press, 2014) ch 6. 87 H Schweitzer, ‘Services of General Economic Interest: European Law’s Impact on the Role of Markets and of Member States’ in M Cremona (ed), Market Integration and Public Services in the European Union (Oxford, Oxford University Press, 2011) 60. 88 A new strategy for the single market at the service of Europe’s economy and society. At the service of Europe’s economy and society, report by Mario Monti to the President of the European Commission, 9 May 2010, 73. 89 The Services Directive of 2006 does not apply to healthcare services, to ‘social services relating to social housing, childcare and support of families and persons permanently or temporarily in need’, nor to ‘non-economic services of general interest’. In this way, most of the public services are excluded from its scope of application (Directive 2006/123, Art 2(2). But see the more detailed analysis of this question by U Neergaard, ‘Services of General (Economic) Interest and the Services Directive – What is

The Freedom to Provide Services  157 have inaugurated a change here. A new Protocol No 26 on Services of General Interest was added to the Treaties, whose Article 2 states in strong wording that ‘the provisions of the Treaties do not affect in any way the competence of Member States to provide, commission and organise non-economic services of general interest’.90 However, the new post-Lisbon constitutional framework has, so far, not had a direct impact on the case law of the CJEU. The Court’s earlier position that some welfare and social services, despite their predominantly non-economic character, are subject to the free movement rules, has not been overruled. A consequence of the fact that certain welfare services fall within the scope of internal market law is that they also fall within the competence of the EU to adopt harmonisation measures to ensure the smooth functioning of the internal market. This consequence has been drawn most clearly, again, in the health sector. Building on the CJEU’s complex case law about cross-border access to medical services, the European Union adopted in 2011 a Directive on cross-border patient rights which aimed at codifying the principles developed by the Court in its case law as well as providing procedural guidance on how to apply those principles.91 This is, in legal terms, a piece of internal market legislation but it does impact on the organisation of national healthcare systems in a manner which could seem at odds with the Treaty article on public health which states that ‘Union action shall respect the responsibilities of the Member States for the definition of their health policy and for the organisation and delivery of health services and medical care’.92 In the following years, the Court continued along the lines of its earlier case law, though there are surprisingly few recent cases on the implications of the Treaty freedom93 and even fewer about the Directive for patient mobility. The Left Out, Why and Where to Go?’ in U Neergaard, R Nielsen and L Roseberry (eds), The Services Directive: Consequences for the Welfare State and the European Social Model (Copenhagen, DJØF Publishing, 2008). 90 Protocol no 26 on services of general interest [2016] OJ C202/307. The Lisbon Treaty also added a new sentence to Art 14 TFEU creating a legal basis for EU legislation harmonising the principles and conditions for the provision of services of general economic interest. This would allow for the adoption of EU-wide ‘universal service’ standards for utilities not yet regulated by EU law but the Commission has, so far, shown no interest in using this legal basis. For discussion of the new primary law provisions added by the Lisbon Treaty, see C Wehlander, Services of General Economic Interest as a Constitutional Concept of EU Law (The Hague, TMC Asser Press, 2016) 80–91. 91 Directive 2011/24 on the application of patients’ rights in cross-border healthcare [2011] OJ L88/45. See, for an analysis of this instrument: S de la Rosa, ‘The Directive on Cross-Border Healthcare or the Art of Codifying Complex Case Law’ (2012) 49 Common Market Law Review 15; W Sauter, ‘Harmonization in Health Care: The EU Patients’ Rights Directive’ in B Cantillon, H Verschueren and P Ploscar (eds), Social Inclusion and Social Protection in the EU: Interactions between Law and Policy (Antwerp, Intersentia, 2012). For the legislative history of the adoption of the Directive, see A de Ruijter, EU Health Law & Policy: The Expansion of EU Power in Public Health and Health Care (Oxford, Oxford University Press, 2019) ch 5. 92 Art 168(7) TFEU. 93 See the discussion of the recent case law by R Cisotta, ‘Limits to Rights to Health Care and the Extent of Member States’ Discretion to Decide on the Parameters of their Public Health Policies’ in FS Benyon (ed), Services and the EU Citizen (Oxford, Hart Publishing, 2013); by T Hervey and J McHale, European Union Health Law: Themes and Implications (Cambridge, Cambridge University Press, 2015) ch 4; and by Chalmers, Davies and Monti, above n 2, 767–73.

158  Bruno de Witte feared disruption of national health systems by the conditional free movement of patients has not happened, as mobility of patients between countries remains a limited phenomenon.94

VIII.  Services in External Relations Law Regulation of services in the context of the EU’s external relations is a separate policy arena from its internal services law. Since much of this regulation is contained in international agreements concluded by the EU, the institutional balance in deciding these rules is different from internal law-making; in particular, the European Parliament is not directly involved in the formulation of the norms, only in their ex-post approval. The role of the CJEU and national courts in enforcing external services law is also more limited than for internal services law. This is partly due to the CJEU’s own choice when it decided that the World Trade Organization (WTO) Agreements (including GATS) were not justiciable and that, in particular, European or third-country service providers could not invoke WTO law to challenge incompatible Union or Member State laws. More recently, since 2008, the text itself of the free trade agreements concluded by the EU excludes their direct effect, so that these agreements cannot serve as an effective legality constraint in the EU legal order. This is the case with the Free Trade Agreement with Korea, the Economic Partnership Agreement with the CARIFORUM States, the Trade Agreement with Colombia and Peru, and those concluded with Ukraine and with Singapore.95 A recent example is Article 30.6 of the Trade Agreement with Canada (CETA), stating that ‘nothing in this Agreement shall be construed as conferring rights or imposing obligations on persons other than those created between the Parties under public international law, nor as permitting this Agreement to be directly invoked in the domestic legal systems of the Parties’.96 Similarly, Article 23.5 of the EU–Japan Economic Partnership Agreement, states that ‘Nothing in this Agreement shall be construed as conferring rights or imposing obligations on persons’.97 This recent practice may seem problematic in light of the EU’s constitutional commitment to the rule of law. According to the CJEU, the EU legal order is based on the rule of law, meaning that all EU legal acts are subject to judicial

94 ‘Report from the Commission on the operation of Directive 2011/24/EU on the application of patients’ rights in cross-border healthcare’ COM(2018) 651 of 21 September 2018; H Vollaard, ‘Patient Mobility, Changing Territoriality and Scale in the EU’s Internal Market’ (2017) 15 Comparative European Politics 435; in the same sense, see (writing in 2013) J Baquero Cruz, ‘The Case Law of the European Court of Justice on the Mobility of Patients: An Assessment’ in FS Benyon (ed), Services and the EU Citizen (Oxford, Hart Publishing, 2013). 95 A Semertzi, ‘The Preclusion of Direct Effect in the Recently Concluded EU Free Trade Agreements’ (2014) 51 Common Market Law Review 1125. 96 Comprehensive Economic and Trade Agreement between Canada, of the one part, and the European Union and its Member States, of the other part [2017] OJ L11/23. 97 Agreement between the European Union and Japan for an Economic Partnership [2018] OJ L330/3.

The Freedom to Provide Services  159 review of their compatibility with higher law.98 How to explain then the situation whereby precise provisions of an international agreement (which is ‘higher law’ in the EU legal order) cannot serve for reviewing internal EU (or Member State) acts? A further constitutional change in this domain is the fact that the Lisbon Treaty finally ended the long discussion as to whether services trade was part of the EU’s exclusive competence to conduct a common commercial policy.99 It is now clear that the European Union has such exclusive competence. The scope of this exclusive power was further clarified in a CJEU judgment of 2013.100 The European Union is actively seeking to facilitate the international trade in services, even though it has shifted its focus from the multilateral to the bilateral arena. The only ‘active’ negotiation in the WTO framework is that on electronic commerce,101 and the mega-regional TiSA negotiations came to a (provisional?) halt at the end of 2016. By contrast, the European Union is pursuing the conclusion of a series of bilateral preferential trade agreements which always include substantial chapters on services. The EU is generally keen to liberalise services trade, as it is an exporter in most service domains, but it also seeks to shield many service sectors from international trade liberalisation.102 At the end of the day, important portions of the services sector are excluded from the free trade agreements, a fact which may be problematic from a WTO law perspective, since Article V GATS allows for preferential trade agreements (derogating from the most-favoured-nation rule of GATS) only when these agreements have ‘substantial sectoral coverage’.103 Services of general interest remain a highly contentious issue in the EU’s free trade negotiations, maybe more so, today, than in internal services law. The EU has used different approaches for shielding public services from the scope of free trade obligations.104 Audiovisual services and ‘services supplied under governmental authority’ (a term familiar from GATS law rather than from internal EU law) are invariably excluded from the scope of the agreements, and a ‘public utilities exemption’ is often included in the part of the agreement in which the EU’s substantive commitments are listed. In addition to the free trade agreements, the Union also acts by means of specific international agreements dealing with specific services, such as the European 98 Case 294/83 Parti écologiste Les Verts v European Parliament EU:C:1986:166, para 23. 99 M Cremona, ‘The Internal Market and External Economic Relations’ in P Koutrakos and J Snell (eds), Research Handbook on the Law of the EU’s Internal Market (Cheltenham, Edward Elgar, 2017) 481–83. 100 Case C-137/12 Commission v Council EU:C:2013:675. 101 See European Commission ‘Recommendation for negotiating directives on electronic commerce’ COM(2019) 165 of 2 April 2019, followed by the EU ‘Proposal for WTO disciplines and commitments relating to electronic commerce’, 26 April 2019, WTO Document INF/ECOM/22. 102 See P Delimatsis, ‘The Evolution of the EU External Trade Policy in Services – CETA, TTIP, and TiSA after Brexit’ (2017) 20 Journal of International Economic Law 583. 103 On this question, see J Jacobsson, Preferential Services Liberalization: The Case of the European Union and Federal States (Cambridge, Cambridge University Press, 2019). 104 L Pedreschi, ‘Balancing Efficacy with Policy Space: The Treatment of Public Services in EU Trade Agreements’ (2017) 20 Journal of International Economic Law 627, 634 ff.

160  Bruno de Witte Convention on the legal protection of conditional access.105 It also adopts unilateral regulation of the activity of third-country service providers. The most egregious example of this is probably the adoption of equivalence decisions for financial services. A large number of EU legislative acts dealing with financial services provide for Commission implementing decisions which extend equivalence to individual third countries for the services concerned. By July 2019, the Commission had taken more than 280 equivalence decisions for more than 30 countries.106

IX. Conclusion The argument proposed in this chapter is that EU services law is best understood as a container of many different legal fields with their own rules, actors and controversies. The direct effect of the Treaty freedom to provide services is still occasionally used by the CJEU and by national courts to disapply national measures unduly restricting cross-border services, but this is no longer the main area of legal action. The Services Directive fleshes out the Treaty freedom in greater detail but without adding much substantive detail, and its scope is limited as it does not apply to the service sectors that are subject to specific EU legislation. In fact, one might argue that the most important parts of EU services law, in terms of day-to-day application and also in terms of litigation, are now the sectoral regulations applying in a number of key service domains, and the horizontal regulations of specific forms of delivery of services, in particular electronic commerce and the posting of workers. The balance between deregulation and re-regulation is situated differently in the different service domains. As a consequence, the normative evaluation of the action of the Court of Justice and of the EU’s political institutions is necessarily bound to be context-specific.107 Overall assessments of services law being clearly neo-liberal or excessively interventionist are therefore not convincing. The tension between market liberalisation and the protection of non-market values remains a central feature of services law but it does not necessarily play out in the traditional form of ‘EU market making versus Member State policy autonomy’. As some service sectors (such as banking and telecommunications) are now entirely Europeanised, the regulatory debate is played out entirely at the European level. As for the age-old debate about the preservation of (national) public services against marketisation, it has now become less virulent in internal services law where a modus vivendi seems to have been found, but it is still very much alive in the context of the EU’s external services policy. 105 Council Decision 2015/1293 on the conclusion of the European Convention on the legal protection of services based on, or consisting of, conditional access [2015] OJ L199/3. 106 Communication from the Commission, ‘Equivalence in the area of financial services’ COM(2019) 349 of 29 July 2019, 2. This document also lists (at 8–9) the equivalence decisions adopted in 2018 and the first half of 2019. 107 The point is made, in the context of an analysis of the Court’s performance in interpreting Art 56 TFEU, by de Witte, above n 27.

8 Free Movement of Capital and Protection of Social Objectives in the EU: Critical Reflections on the Case Law Regarding Golden Shares and Privatisations ILEKTRA ANTONAKI

I. Introduction Capital liberalisation has undoubtedly brought significant benefits to the global economy, by improving economic efficiency, allocating efficiently global resources, promoting foreign investment and supporting the transfer of technology and innovation around the globe.1 However, it is often associated with increased financial instability, rising levels of unemployment and income inequality.2 In view of the possible adverse consequences of free capital flows, a contemporary economic school of thought has emerged departing from an unconditional adherence to unfettered capital liberalisation and advocating a more restrictive approach regarding international capital flows in order to prevent the risks of financial instability

* This contribution is based on the author’s doctoral dissertation ‘Privatisations and golden shares: Bridging the gap between the State and the market in the area of free movement of capital in the EU’ (The Meijers Research Institute and Graduate School of the Leiden Law School of Leiden University, 2019) defended on 3 September 2019 at Leiden University. 1 J Ostry, A Ghosh and A Korinek, ‘Multilateral Aspects of Managing the Capital Account’ (2012) IMF Staff Discussion Note; M Obstfeld, ‘International Finance and Growth in Developing Countries: What Have We Learned?’ (2009) 56 IMF Staff Papers 63; K Rogoff, ‘International Institutions for Reducing Global Financial Instability’ (1999) 13 Journal of Economic Perspectives 21. 2 A Charlton, ‘Capital Market Liberalization and Poverty’ in JA Ocampo and JE Stiglitz (eds), Capital Market Liberalization and Development (Oxford, Oxford University Press, 2008); JE Stiglitz et al, Stability with Growth: Macroeconomics, Liberalization and Development (Oxford, Oxford University Press, 2006); D Rodrik, ‘Who Needs Capital-Account Convertibility?’ in S Fischer et al (eds), Should the IMF Pursue Capital-Account Convertibility? (Essays in International Finance No 207, May 1998) International Finance Section, Department of Economics, Princeton University, 1998.

162  Ilektra Antonaki and social inequality.3 This economic thinking is to some extent embraced also by international economic organisations such as the International Monetary Fund (IMF) and the Organisation for Economic Co-operation and Development (OECD) which, in view of the controversy surrounding the impact of free capital flows on social inequality as well as the inconclusive economic evidence regarding their contributory effect to economic growth, are of the opinion that capital liberalisation improves economic efficiency only if it is complemented by measures of prudential supervision and regulation, sound macroeconomic policies and transparency.4 The negative social consequences of unfettered capital flows have caused a rising backlash against globalisation, which is often accompanied by measures of economic protectionism. In the field of trade, protectionist barriers on imports and exports of goods are counterproductive and they have a negative impact on growth and employment.5 In the field of financial integration, however, defining what constitutes a protectionist measure and assessing its impact on economic growth and employment is a more difficult exercise. In the EU, golden shares and forms of public ownership are often regarded as protectionist measures aimed at protecting national industries against hostile takeovers, from foreign or even domestic competitors. However, although in some cases golden shares in strategic privatised undertakings might be considered to be forms of economic protectionism, they should not always be regarded as restrictions on the fundamental freedoms. While measures which prohibit the acquisition of shares by foreign investors constitute discriminatory restrictions on the free movement of capital, other measures affecting the voting rights of shareholders, the composition of the board of directors or the decision-making process within an undertaking without discriminating between foreign and domestic investors might not restrict the free movement of capital. It is true that Member States have used golden shares as a means of protecting national industries against hostile takeovers from foreign or even domestic competitors. For instance, in the seminal Volkswagen case,6 the special rights of the Federal State of Germany and the Land of Lower Saxony in Volkswagen effectively prevented the acquisition of total control of the company by the rival automobile industry Porsche.7 However, the adoption of a very broad interpretation of the 3 J Ostry, P Loungani and D Furceri, ‘Neoliberalism: Oversold?’ (2016) 53 Finance & Development 38; T Piketty, Capital in the Twenty-First Century (Cambridge, MA, The Belknap Press of Harvard University Press, 2014); JE Stiglitz, ‘Capital Market Liberalization, Economic Growth and Instability’ (2000) 28 World Development 1075, 1076. 4 IMF, The Liberalization and Management of Capital Flows – An Institutional View (2012) 10–12; OECD, The OECD’s Approach to Capital Flow Management Measures Used with a Macro-Prudential Intent (2015) report to G20 Finance Ministers, 2015; OECD, Getting the most out of International Capital Flows (2011). 5 C Lagarde, ‘Fix the Roof While the Window of Opportunity is Open: Three Priorities for the Global Economy’, speech at the University of Hong Kong, IMF, 2018. 6 Case C-112/05 Commission v Germany EU:C:2007:623. 7 W-G Ringe, ‘Case C-112/05, Commission v Germany (“VW law”), Judgment of the Grand Chamber of 23 October 2007, nyr’ (2008) 45 Common Market Law Review 37.

Free Movement of Capital and Social Objectives  163 notion of ‘capital restrictions’ under Article 63 of the Treaty on the Functioning of the European Union (TFEU) and the treatment of golden shares and forms of public ownership as inherently incompatible with the internal market raises significant concerns, first, regarding the division of competences between the EU and the Member States in the fields of corporate governance and property ownership systems, and second, regarding the protection of public interest objectives. Drawing from the international controversy surrounding the costs and benefits of unfettered capital liberalisation, this contribution starts from the premise that a certain degree of regulation in the area of capital movements is necessary in order to prevent the risk of financial instability, reduce income inequality and protect legitimate social objectives. On the basis of this theoretical claim, it seeks to address the risks of a very broad interpretation of ‘capital restrictions’ under Article 63 TFEU by identifying the legal tools under the existing legal framework and the possible modifications in the judicial interpretation of the free movement provisions that will ensure respect for the division of competences between the EU and the Member States in the fields of corporate governance and property ownership, and will allow sufficient room for reconciling economic integration with societal values. In this respect, the chapter focuses on the golden shares and privatisations case law and advances certain adjudicative methods, which could allow Member States to determine their property ownership and corporate governance systems without imposing protectionist obstacles on foreign investment. In particular, it puts forward two main suggestions: first, the recalibration of the capital restrictions test in the golden shares case law by reference to a Keck-inspired notion of ‘investment arrangements’; and second, the rediscovery of the principle of neutrality under Article 345 TFEU as a legal provision shielding national decisions to maintain public ownership of undertakings from internal market scrutiny. It argues that through this ‘velvet revolution’ in the context of negative integration, the interpretation of the free movement of capital in the EU can respect the delicate balance of competences between the EU and the Member States and facilitate the reconciliation of the capital freedom with public interest objectives.

II.  Free Movement of Capital and Golden Shares A.  The Use of Golden Shares in Privatised Undertakings in the EU The first part of this chapter focuses on golden shares, ie, the special rights that Member States maintain in strategically sensitive privatised companies (telecommunications, electricity and energy production and distribution, postal services, car industries, etc).8 These special rights allow the State to control changes in 8 Oxera, Special rights of public authorities in privatised EU companies: the microeconomic impact, Report prepared for the European Commission, 2005.

164  Ilektra Antonaki ownership and/or veto certain strategic decisions in order to prevent hostile takeovers, to guarantee the provision of services of general interest, to safeguard public security and other public interest objectives.9 The use of golden shares became a widespread phenomenon in Europe in the 1990s and early 2000s. The privatisation of State-owned companies stimulated the growth and development of capital markets in Europe by creating an equity culture.10 However, the State’s involvement in the competitive market through golden shares was perceived as a threat or a hindrance to the emergence of a fully competitive market for corporate control.11 The establishment of special shareholding in national privatised champions was regarded by the Commission as an expression of economic protectionism and therefore incompatible with the free movement of capital.12 This gave rise to long-running litigation between the Commission and the Member States before the Court of Justice.13 The Member States argued that the special rights in privatised undertakings did not derogate from ordinary company law and were in any case necessary in order to pursue public interest objectives, such as safeguarding security of energy supplies or guaranteeing a service of general interest. The Court adopted a rigorous interpretation of the fundamental freedoms provisions ruling in all but one case14 that the special shareholding that Member States retain in privatised undertakings constitutes a restriction on the free movement of capital because of its deterrent effect on foreign investment. The adjudicative model of the Court is based on the ideological premise that golden shares are inherently incompatible with the fundamental freedoms and is characterised by a strikingly broad interpretation of the notion of ‘capital restrictions’. This broad interpretation carries the risk that it might turn into a ‘quality control’ of domestic company law, whereby the Court is called upon to review not 9 European Commission, Special rights in privatized companies in the enlarged Union – a decade full of developments, Commission Staff Working Document, 2005. 10 E Szyszczak, ‘Golden Shares and Market Governance’ (2002) 29 Legal Issues of Economic Integration 255, 260. 11 ibid, 261. 12 J Rickford, ‘Protectionism, Capital Freedom and the Internal Market’ in U Bernitz and W-G Ringe (eds), Company Law and Economic Protectionism: New Challenges to European Integration (Oxford, Oxford University Press, 2010) 55; J Rickford, ‘Free movement of capital and protectionism after Volkswagen and Viking Line’ in M Tison et al (eds), Perspectives in Company Law and Financial Regulation: Essays in Honour of Eddy Wymeersch (Cambridge University Press, 2009) 62. 13 Case C-58/99 Commission v Italy EU:C:2000:280; Case C-367/98 Commission v Portugal EU:C:2002:326; Case C-483/99 Commission v France EU:C:2002:327; Case C-503/99 Commission v Belgium EU:C:2002:328; Case C-463/00 Commission v Spain EU:C:2003:272; Case C-98/01 Commission v United Kingdom EU:C:2003:273; Case C-174/04 Commission v Italy EU:C:2005:350; Joined Cases C-282/04 and C-283/04 Commission v The Netherlands EU:C:2006:608; Case C-112/05 Commission v Germany EU:C:2007:623; Case C-274/06 Commission v Spain EU:C:2008:86; Case C-207/07 Commission v Spain EU:C:2008:428; Case C-326/07 Commission v Italy EU:C:2009:193; Case C-171/08 Commission v Portugal EU:C:2010:412; Case C-543/08 Commission v Portugal EU:C:2010:669; Case C-212/09 Commission v Portugal EU:C:2011:717; Case C-244/11 Commission v Greece EU:C:2012:694; Case C-95/12 Commission v Germany EU:C:2013:676. 14 Case C-503/99 Commission v Belgium, above n 13.

Free Movement of Capital and Social Objectives  165 only questions of cross-border discrimination or market access, but also any type of company law arrangements.15 This may lead to a situation where different types of company law rules (ranging from various anti-takeover mechanisms to the German co-determination rule) might be considered to be liable to deter acquisitions of shares by foreign investors and therefore to hinder capital movements.16 The rigorous application of the free movement of capital in the golden shares case law implicitly favours the constitutional foundations of liberal market economies as opposed to coordinated market economies17 and promotes through negative integration a corporate governance system, which endorses the principle of proportionality between ownership and control and the principle of shareholders’ primacy.18 This in turn might lead to an erosion of the varieties of capitalism that exist in Europe19 and to a judicially driven convergence into a specific model of market economy, which might give rise to concerns regarding the respect of the division of competences between the EU and the Member States. The implications of these rulings are profound, as they affect the structure and the organisation of national corporate governance systems, one of the core aspects of national economic policy and one of the main factors determining the type of capitalism established at national level. This section focuses on three legal issues arising from the golden shares case law: horizontality, restrictions and proportionality. First, with respect to the question of horizontality, it is argued that while it is true that the horizontal effect of Article 63 TFEU might interfere with the principle of private autonomy, reasons of legal consistency and equality between the State and the private parties as market participants require that the free movement of capital should in principle be horizontally applicable. Second, with respect to the question of capital restrictions, it

15 W-G Ringe, ‘Company Law and Free Movement of Capital’ (2010) 69 Cambridge Law Journal 378, 400. 16 ibid. See also M Andenas, T Gütt and M Pannier, ‘Free Movement of Capital and National Company Law’ (2005) 16 European Business Law Review 757; S Grundmann and F Möslein, ‘The Golden Share – State Control in Privatised Companies: Comparative Law, European Law and Policy Aspects’ (2001–02) 4 European Banking and Financial Law Journal – Euredia 623. 17 The structural bias of negative integration in favour of liberal market economies as opposed to social market economies has been thoroughly addressed by eminent scholars of the Max Planck Institute for the Study of Societies; see FW Scharpf, ‘The asymmetry of European integration o: why the EU cannot be a “social market economy”’ (2010) 8 Socio-Economic Review 211; M Höpner and A Schäfer, ‘Integration among unequals: How the heterogeneity of European varieties of capitalism shapes the social and democratic potential of the EU’ (2012) MPIfG Discussion Paper 12/5 Cologne, Max Planck Institute for the Study of Societies; M Höpner and A Schäfer, ‘A New Phase of European Integration: Organized Capitalisms in Post-Ricardian Europe’ (2007) MPIfG Discussion Paper 07/4; M Höpner and A Schäfer, ‘Embeddedness and Regional Integration. Waiting for Polanyi in a Hayekian Setting’ (2012) 66 International Organization 429. 18 C Gerner-Beuerle, ‘Shareholders Between the Market and the State. The VW Law and other Interventions in the Market Economy’ (2012) 49 Common Market Law Review 97. 19 P Hall and D Soskice, Varieties of Capitalism: The Institutional Foundations of Comparative Advantage (Oxford, Oxford University Press, 2001); J Snell, ‘Varieties of Capitalism and the Limits of European Economic Integration’ (2012) 13 Cambridge Yearbook of European Legal Studies 415.

166  Ilektra Antonaki is argued that the Court should reconsider the strikingly broad interpretation of the notion of capital restrictions. Drawing inspiration from the Keck concept of ‘selling arrangements’ as developed in the field of free movement of goods, it is suggested that the transposition of a similar concept in the field of free movement of capital in the form of ‘investment arrangements’ can offer room for State participation in the market for the purposes of pursuing public interest objectives. Third, with respect to the question of proportionality, it is argued that a proportionality assessment based on the ‘good governance’ model and the procedural standards of legal certainty can guarantee an objective judicial review that respects national regulatory diversity in a sensitive field that is not harmonised at the EU level, while at the same time ensuring compliance with procedural requirements that create a transparent legal framework protecting the legitimate expectations of foreign investors and providing effective judicial remedies against arbitrary decisions of public authorities.

B.  The Horizontal Application of Article 63 TFEU The doctrine of horizontal direct effect of Treaty provisions in EU law originates from the concept of horizontality or third-party effect (Drittwirkung) of the fundamental rights enshrined in national constitutions. In particular, in constitutional law there is a long-standing debate regarding the scope of application of the fundamental rights protected under the constitution. Over the years two main approaches to human rights protection have been developed: on the one hand, the vertical theory holds that fundamental rights can be invoked by individuals only against the State; on the other hand, the horizontal theory expands the scope of application of fundamental rights so as to cover relations between private individuals as well.20 The theoretical underpinnings of the two theories reflect different political philosophies and different views about the interconnection between the public and private sphere. In particular, the vertical approach to human rights protection is inspired by the political and economic theory of classical liberalism21 and is premised on a strict distinction between the public and private sphere.22 It proclaims 20 For an overview of the rich constitutional debate regarding the dichotomy between vertical and horizontal approaches to human rights protection, see: D Beyleveld and SD Pattinson, ‘Horizontal applicability and horizontal effect’ (2002) 118 Law Quarterly Review 623; HWR Wade, ‘Horizons of horizontality’ (2000) 116 Law Quarterly Review 217; J Morgan, ‘Questioning the True Effect of the Human Rights Act’ (2002) 22 Legal Studies 259; G Phillipson and A Williams, ‘Horizontal Effect and the Constitutional Constraint’ (2011) 74 Modern Law Review 878; S Gardbaum, ‘The “Horizontal Effect” of Constitutional Rights’ (2003) 102 Michigan Law Journal 387. 21 D Dyzenhaus, ‘The New Positivists’ (1989) 39 University of Toronto Law Journal 361. 22 M Hunt, ‘The “horizontal effect” of the Human Rights Act’ [1998] Public Law 423, 424; A Clapham, Human Rights in the Private Sphere (Oxford, Clarendon Press, 1996); W Marshall, ‘Diluting Constitutional Rights: Rethinking State Action’ (1985) 80 Northwestern University Law Review 558; C Sustein, The Partial Constitution (Cambridge, MA, Harvard University Press, 1998); E Chemerinsky, ‘Rethinking State Action’ (1985) 80 Northwestern University Law Review 503; B Slattery, ‘Charter of Rights and

Free Movement of Capital and Social Objectives  167 values such as private autonomy, privacy and market efficiency23 and advances the view that the most crucial function of a constitution is to provide the law for the lawmaker, not for the citizen. Therefore, relations between private individuals must remain outside the reach of human rights protection.24 At the other end of the spectrum, advocates of the horizontal approach to human rights protection show greater commitment to social democratic norms, which regard the State as having not only negative obligations, which essentially require the State not to interfere with the exercise of individual rights, but also positive obligations, which require national authorities to take all necessary measures to safeguard human rights.25 The State must ensure that fundamental rights are fully protected within its jurisdiction not only in State–citizen relations but also in relations between private individuals, either by making sure that the legislature enacts laws protecting human rights and abolishing discrimination or by allowing the judiciary to apply human rights horizontally.26 The question of horizontality in EU law has been addressed in several landmark judgments of the Court of Justice. Rulings such as Defrenne,27 Walrave and Koch,28 Bosman,29 Angonese,30 Viking and Laval,31 Fra.bo,32 AMS,33 Egenberger34 and Bauer35 have extended the application of the fundamental freedoms and the fundamental rights to horizontal situations and have sparked an animated debate among scholars.36 Freedoms – Does it Bind Private Persons?’ (1985) 63 Canadian Bar Review 148; A Butler, ‘Constitutional Rights in Private Litigation: A Critique and Comparative Analysis’ (1993) 22 Anglo-American Law Review 1. 23 S Gardbaum, ‘The Place Of Constitutional Law in the Legal System’ in M Rosenfeld and A Sajó (eds), The Oxford Handbook of Comparative Constitutional Law (Oxford University Press, 2012) 177. 24 ibid. 25 For the distinction between negative and positive obligations in Human Rights Law see D Shelton and A Gould, ‘Positive and Negative Obligations’ in D Shelton (ed), The Oxford Handbook of International Human Rights Law (Oxford, Oxford University Press, 2013); JF Akandji-Kombe, Positive obligations under the European Convention on Human Rights – A guide to the implementation of the European Convention on Human Rights Human Rights Handbooks, No 7, Council of Europe (2007); H Breakey, ‘Positive Duties and Human Rights: Challenges, Opportunities and Conceptual Necessities’ (2015) 63 Political Studies 1198. 26 M Tushnet, ‘The issue of state action/horizontal effect in comparative constitutional law’ (2003) 1 International Journal of Constitutional Law 79, 90. 27 Case 43–75 Defrenne EU:C:1976:56. 28 Case 36–74 BNO Walrave and Koch EU:C:1974:140. 29 Case C-415/93 Bosman EU:C:1995:463. 30 Case C-281/98 Angonese EU:C:2000:296. 31 Case C-438/05 Viking Line EU:C:2007:772; Case C-341/05 Laval EU:C:2007:809. 32 Case C-171/11 Fra.bo SpA EU:C:2012:453. 33 Case C-176/12 Association de médiation sociale EU:C:2014:2. 34 Case C-414/16 Egenberger EU:C:2018:257. 35 Case C-569/16 Stadt Wuppertal v Maria Elisabeth Bauer EU:C:2018:871. 36 E Frantziou, ‘The Horizontal Effect of the Charter of Fundamental Rights of the EU: Rediscovering the Reasons for Horizontality’ (2015) 21 European Law Journal 657; S Van den Bogaert, ‘Horizontality: The Court Attacks?’ in C Barnard and J Scott (eds), The Law of the Single European Market: Unpacking the Premises (Oxford, Hart Publishing, 2002); H Schepel, ‘Constitutionalising the Market, Marketising the Constitution, and to Tell the Difference: On the Horizontal Application of the Free Movement Provisions in EU Law’ (2012) 18 European Law Journal 177; R Babayev, ‘Contractual Discretion and

168  Ilektra Antonaki In the area of free movement of capital, the Court has not dealt explicitly with the question of the horizontal application of Article 63 TFEU. It has been argued that the majority of the capital case law concerns the field of taxation, a quintessentially vertical issue, and this is why the thorny question of horizontality of the free movement of capital has not been answered until now.37 Furthermore, in the field of golden shares, national governments acting in their capacity as shareholders have tried to escape from judicial scrutiny by arguing that the contested measure does not constitute a ‘State measure’ and therefore does not fall within the scope of the free movement of capital, an argument which implies that Member States oppose the idea of granting horizontal effect of Article 63 TFEU. For instance, in the Dutch golden shares case,38 the Dutch government argued that the golden shares in the privatised companies Koninklijke KPN NV and TPG NV could not qualify as ‘State measures’, as they did not hold them in their capacity as a public authority, but instead as a private shareholder without departing from normal national company law. However, until now, the Court has rejected this argument on the basis of a broad interpretation of the notion of ‘State measure’ covering all measures taken by public authorities even when acting in their private capacity as shareholders. This has allowed the Court to treat these situations as vertical and to review the compatibility of those measures with the fundamental freedoms, without having to give an explicit answer to the question whether Article 63 TFEU is horizontally applicable. Scholars have expressed different views regarding the horizontal effect of Article 63 TFEU, both in relation to the question whether Article 63 TFEU applies when the State acts as a shareholder/market participant under private law (extended vertical effect) and in relation to the stricto sensu horizontal effect of Article 63 TFEU covering acts of private parties. In more detail, Jonathan Rickford and Wolf-Georg Ringe embrace the extended vertical effect of Article 63 TFEU, arguing that the obligation not to obstruct capital movements binds Member States not only when they exercise their sovereign powers under public law, but also when they act in their private capacity under private law.39 At the same time, they express doubts as to whether private parties engaging for purely private purposes in conduct which falls short of discrimination are bound by the free movement the Limits of Free Movement Law’ (2015) 23 European Review of Private Law 875; A Dashwood, ‘Viking and Laval: Issues of Horizontal Direct Effect’ (2007) 10 Cambridge Yearbook of European Legal Studies 525; S Prechal and S De Vries, ‘Seamless web of judicial protection in t he internal market?’ (2009) 34 European Law Review 5; A Hartkamp, ‘The Effect of the EC Treaty in Private Law: On Direct and Indirect Horizontal Effects of Primary Community Law’ (2010) 3 European Review of Private Law 529; S Enchelmaier, ‘Horizontality: the application of the four freedoms to restrictions imposed by private parties’ in P Koutrakos and J Snell (eds), Research Handbook on the Law of the EU’s Internal Market (Cheltenham, Edward Elgar Publishing, 2017). 37 C Barnard, The Substantive Law of the EU: The Four Freedoms, 5th edn (Oxford University Press, 2016) 526. 38 Joined Cases C-282/04 and C-283/04 Commission v The Netherlands, above n 13. 39 Rickford, ‘Free movement of capital and protectionism after Volkswagen and Viking Line’, above n 12, 83–84; Ringe, ‘Company Law and Free Movement of Capital’, above n 15, 378, 396.

Free Movement of Capital and Social Objectives  169 of capital. Conversely, Andrea Biondi argues that if every action attributable to the State is caught by capital scrutiny, this leaves very limited leeway for Member States to organise their economic and industrial policies and to pursue public policy objectives.40 Inspired by State-aid law, he suggests that this risk might be prevented through the adoption of a ‘private market investor’ principle, whereby State conduct is immune from scrutiny if proven to be subject to the normal rules of the operation of the market.41 Carsten Gerner-Beuerle is of the opinion that the vertical effect of Article 63 TFEU covers not only regulatory acts of the State (ius imperii), but also acts taken by the State in the course of its operation as a normal shareholder without ius imperii.42 However, he argues that in the latter scenario, only measures discriminating between domestic and foreign investment should fall within the scope of the free movement of capital. In relation to the horizontal application of Article 63 TFEU, he expresses his fear that if the Court accords horizontal effect to Article 63 TFEU, this could potentially open the door to an avalanche of claims challenging any type of control enhancing mechanisms contained in a company’s articles of association as capital restrictions.43 In the same vein, Harm Schepel argues that the recognition of horizontal effect of Article 63 TFEU together with the broad interpretation of capital restrictions will most likely place all company law structures deviating from the one share/one vote principle and from the principle of shareholders’ primacy under judicial review for conformity with the free movement of capital.44 The aforementioned scholarly contributions reveal the intricate legal issues arising from a possible horizontality of the free movement of capital in view of the multilevel regulatory regime governing national corporate governance systems (consisting of not only national legislation but also of private instruments, such as shareholders’ agreements, articles of association, codes of conduct, etc). It is submitted here that if one accepts that the free movement of capital binds the State even when it acts as a shareholder (extended vertical effect), reasons of legal consistency and equality between market participants require that the same should apply to private action. Private parties are important market operators (in many instances much more important than the State) and their actions can have far-reaching economic and social implications. Therefore, they should be bound by the fundamental freedoms, especially when they are in a dominant position and can exercise quasi-regulatory powers over other market participants.45 40 A Biondi, ‘When the State is the Owner – Some Further Comments on the Court of Justice “Golden Shares” Strategy’ in U Bernitz and W-G Ringe (eds), Company Law and Economic Protectionism: New Challenges to European Integration (Oxford, Oxford University Press, 2010) 97. 41 ibid, 101. 42 Gerner-Beuerle, above n 18, 97, 131. 43 ibid, 133. 44 Schepel, ‘Constitutionalising the Market’, above n 36, 177, 193. 45 cf R Babayev, ‘Private Autonomy at Union Level: On Article 16 CFREU and Free Movement Rights’ (2016) 53 Common Market Law Review 979, in which he argues that if the free movement provisions

170  Ilektra Antonaki It is true that a horizontal application of Article 63 TFEU could lead to a ‘quality control’ review of not only public but also private company law arrangements. A possible recognition of the horizontal effect of Article 63 TFEU would allow individuals to challenge a wide spectrum of private instruments, thus expanding even more the already broad scope of measures affecting capital movements. This might have an intrusive effect into the sphere of private autonomy. However, the potential interference with the principle of private autonomy could be prevented either through a delineation of the scope of capital restrictions or through a balancing exercise between economic freedoms and social values at the justifications or proportionality level. The recognition of horizontal effect of the free movement of capital can promote the effectiveness of EU law, especially in a field dominated by private corporations and financial institutions which can exercise quasi-regulatory powers over other market participants. This is all the more true if one accepts the theory of horizontality of fundamental rights which regards private individuals as not existing in a legal vacuum, but as part of a solidaristic society in which individuals are responsible for their actions towards other individuals.

C.  Restricting Capital Restrictions In the golden shares case law, the Court has adopted a broad interpretation of the notion of capital restrictions encompassing all measures liable to deter foreign investors from acquiring shares in the undertakings concerned. This over-inclusive definition is reminiscent of the broad Dassonville46 definition of measures having an effect equivalent to quantitative restrictions in the area of free movement of goods. The Member States have consistently expressed their strong opposition to the strikingly broad interpretation of capital restrictions, arguing that it essentially deprives them of their right to choose and determine their corporate governance regime in accordance with their industrial policy and the structure of their economy. Among their various defences, some Member States attempted to draw a parallel between golden shares and the concept of ‘selling arrangements’ as developed in Keck,47 arguing that a similar approach should apply in relation to the special shareholding retained in privatised companies. This argument was

were attributed full horizontal effect this would negate the freedom to conduct a business under Art 16 of the Charter. To resolve this conflict, he puts forward a ‘double proportionality’ test: on the one hand, when there is no ‘dominance’ (ie, when the private actors in question do not exert power, the outcome of which other private actors are not able to avoid), the ‘minor’ interference with free movement rights cannot justify the ‘major’ interference with the freedom to conduct a business; on the other hand, when there is ‘dominance’ (ie, when the private actors impose obligations which other private actors cannot avoid), the ‘major’ interference with the free movement rights can justify the ‘minor’ interference with the freedom to conduct a business. 46 Case 8–74 Dassonville EU:C:1974:82, para 5. 47 Joined Cases C-267/91 and C-268/91 Keck and Mithouard EU:C:1993:905.

Free Movement of Capital and Social Objectives  171 invoked by Spain,48 the UK49 and Portugal.50 Although in principle the Court did not exclude the possibility of transposing the concept of ‘selling arrangements’ into the free movement of capital, it nevertheless rejected it on the facts, stating that the measures at issue were not comparable to the rules concerning ‘selling arrangements’. The Court found that while the contested measures were not discriminatory, they nonetheless affected the position of a person acquiring a shareholding and were thus liable to deter foreign investors from making such investments and consequently affected their access to the market.51 The underlying rationale of the interpretation of the Court is that golden shares, by their very nature, are inherently incompatible with EU law. This is a rather political argument, which stems from the general ideological premise that State participation in the market should be limited and reflects a political choice favouring the model of liberal market economy as opposed to the model of coordinated market economy. Indeed, in the golden shares case law, the Court seems to endorse the principle of proportionality between corporate ownership and control, implying that any shareholder should in principle own the same fraction of cash flow rights and voting rights. Any structure or mechanism that derogates from this principle should be disapplied as it discourages investors from acquiring shares in the undertaking concerned. According to this reasoning, the involvement of the State in the management of the company through the vehicle of golden shares is capable of depressing the value of the shares, thus reducing the attractiveness of an investment in that company.52 The reason is that the State might exercise its special rights in order to pursue public interest objectives, ‘which might be contrary to the economic interests of the company’.53 However, Article 63 TFEU should not be regarded as a carte blanche to abolish all national company law requirements that derogate from the principle of proportionality between ownership and control or that do not adhere completely to shareholders’ primacy on the assumption that they render investment less attractive.54 The choice of corporate governance model is a political question, which depends on various socio-economic factors that differ significantly among 48 Case C-463/00 Commission v Spain, above n 13. 49 Case C-98/01 Commission v UK, above n 13. 50 Case C-171/08 Commission v Portugal, above n 13; Case C-543/08 Commission v Portugal, above n 13; Case C-212/09 Commission v Portugal, above n 13. 51 See for instance Case C-98/01 Commission v UK, above n 13, para 47. 52 Case C-171/08 Commission v Portugal, above n 13, para 54. 53 Joined Cases C-282/04 and C-283/04 Commission v The Netherlands, above n 13, para 30. A similar view was expressed by Advocate General Maduro in Federconsumatori, where he opined that public ownership of shareholding does not reduce the attractiveness of investing in the company concerned, only insofar as the State respects the normal rules of operation of the market with a view to maximising its return on investment. See opinion of Advocate General Maduro in Joined Cases C-463/04 and C-464/04 Federconsumatori and Others EU:C:2006:524, paras 25–26. 54 H Schepel, ‘Of Capitalist Nostalgia and Financialisation: Shareholder Primacy in the Court of Justice’ in C Joerges and C Glinski (eds), The European Crisis and the Transformation of Transnational Governance (Oxford, Hart Publishing, 2014).

172  Ilektra Antonaki Member States and should be distinguished from the objective of market integration pursued by the Treaties.55 While it is true that Member States should refrain from adopting measures that obstruct the functioning of the internal market, they should in principle be able to define their market economy model, participate in the market and protect societal values. A judicially driven convergence into the model of liberal market economies can barely be squared with the wording and the spirit of the Treaties, let alone with the international debate about the possible contributory effect of capital liberalisation to the rising levels of income inequality. Such a convergence raises significant concerns regarding the division of competences between the EU and the Member States in the fields of corporate governance and property ownership systems and the protection of public interest objectives. This by no means should be interpreted as leaving room for suspicious State interferences that could lead to serious market abuses and distortions of competition. The Court should not refrain from disapplying national measures that obstruct capital flows and undermine the process of financial integration. However, it should respect the division of competences between the EU and the Member States and the discretion of a Member State to determine the corporate governance regime of its national economy, and it should intervene only when the national choices seriously undermine the objectives of market integration either by discriminating against foreign investors or by hindering the market access of foreign investors in derogation from ordinary company law. Its prime objective should not be the maximisation of shareholders’ profit, but the establishment of a well-functioning and integrated market economy, which respects the national policy choices as expressed through various mechanisms of national company law.56 On the basis of the aforementioned considerations and taking into account the evolution of the interpretation of restrictions in the field of free movement of goods, it is suggested that a refined test could be adopted in order to delineate in a more consistent manner the scope of capital restrictions: 1.

2.

Discrimination on grounds of nationality. Measures that discriminate between domestic and foreign investors constitute restrictions on the free movement of capital, which can be justified only by the objectives provided for in the Treaty. Derogation from ordinary company law. Measures that do not discriminate on grounds of nationality may be divided into two sub-categories: a. Non-discriminatory measures that (i) do not derogate from ordinary company law and (ii) respect the principle of equality between the State and private parties as market participants without granting any privilege

55 Gerner-Beuerle, above n 18, 97, 126. 56 ibid, 108. However, the author admits that company law is a highly complex field of law with many different variations, exceptions and special provisions, making the determination of what complies with or derogates from ordinary company law a rather intricate and challenging exercise.

Free Movement of Capital and Social Objectives  173 to the State should be regarded as ‘investment arrangements’,57 which – just like the Keck ‘selling arrangements’ in the field of free movement of goods – fall outside the scope of Article 63 TFEU. These rules merely structure the market and the corporate governance regime of a Member State, without hindering foreign investment. b. Non-discriminatory measures that (i) derogate from ordinary company law and/or (ii) are available only to the State (to the exclusion of private investors) constitute restrictions on the free movement of capital prohibited under Article 63 TFEU. It can then be examined whether they can be justified by the Treaty-based derogation grounds or overriding reasons in the public interest in accordance with the principle of proportionality. The derogation from ordinary company law criterion stems from the golden shares case law and ensures that neither the State nor private parties depart from the normal application of national company law in order to avail themselves of undue advantages that infringe the national legislative framework and distort the market for corporate control. This way, both the principle of legal certainty and the discretion of Member States in organising their corporate governance systems are respected. Private market operators are expected to abide by the binding rules of national company law while at the same time exercising their private autonomy under certain optional choices allowed under the national legislative framework. In the same vein, the State, in its capacity as a shareholder, is expected to act in accordance with the binding rules of national company law while at the same making use of the options available under the national corporate governance regime. However, it should be noted that although the non-derogation from ordinary company law might be a necessary condition for the exclusion of a certain measure from the scope of capital restrictions, it is nevertheless not a sufficient one. What is also needed is a guarantee that the rules of national company law from which the measure in question does not derogate, create a level playing field for all market participants, both public and private. In other words, it is necessary to ensure that the measure at stake respects the principle of equality between the State and private parties as market participants and does not grant an undue advantage to the State to the detriment of private operators. The refined test on capital restrictions would allow the Court to exercise thorough oversight over national rules that unjustifiably grant privileges to the State, while at the same time ensuring that it does not overstep the boundaries of its judicial powers by implicitly imposing a specific corporate governance regime on the Member States. It would allow a better drawing of the dividing line between the EU and the Member States and would respect the preserved competence of the Member States to determine their corporate governance systems and protect



57 Biondi,

above n 40, 96.

174  Ilektra Antonaki public interest objectives, while at the same time remaining within the contours of the internal market and preventing protectionist measures which derogate from ordinary company law and hinder the market access of foreign investors.

D.  Procedural Proportionality and Legal Certainty The golden shares case law provides a typical example of rigorousness in the application of the principle of proportionality. The Court requires a high degree of evidence in order to establish the legal claim that the contested special rights are justified and proportionate to the objective pursued. The intensity of the judicial review comes as no surprise as it reflects the underlying rationale that golden shares are inherently incompatible with free movement of capital and it confirms the generally strict jurisprudential approach to proportionality in internal market litigation. What is more interesting, however, is the additional step that emerges from the proportionality assessment of golden shares and characterises the capital case law in general. This additional step examines the compatibility with the principle of legal certainty and it is governed by a procedural rather than substantive rationality. Procedural elements of proportionality such as policy coherence, consistency and transparency are discerned in other fields of internal market case law, primarily those touching upon morally contentious issues. In these fields, the ‘procedural proportionality’ test is used to rationalise the process of national legislation and is more respectful of the normative policy aims of the Member States.58 The ‘procedural proportionality’ test is juxtaposed with the ‘substantive proportionality’ test, which seeks to rationalise the content of national legislation, leaving little normative leeway for Member States. Floris de Witte argues that for morally and ethically contentious questions, the ‘procedural proportionality’ test is better suited, as it respects the substance of national moral and ethical choices and focuses only on teasing out discriminatory or protectionist biases by assessing the normative coherence of national policies, the consistent application of sanctions and the legislative transparency.59 The main field in which the ‘procedural proportionality’ test has been widely applied is gambling, a field in which the process of market liberalisation has not been very successful and Member States continue to enjoy a wide margin of discretion in defining their regulatory framework. In fact, it has been argued that the relaxation of the internal market requirements in the gambling cases comes down to virtual dismantling.60 In its gambling case law,61 the Court has granted 58 F De Witte, ‘Sex, Drugs & EU Law: The Recognition of Moral and Ethical Diversity in EU Law’ (2013) 50 Common Market Law Review 1545, 1566. 59 ibid, 1573. 60 S Van den Bogaert and A Cuyvers, ‘“Money for Nothing”: The Case law of the EU Court of Justice on the Regulation of Gambling’ (2011) 48 Common Market Law Review 1175, 1208. 61 Case C-243/01 Gambelli and Others EU:C:2003:597; Case C-338/04 Placanica EU:C:2007:133; Case C-42/07 Liga Portuguesa EU:C:2009:519; Case C-316/07 Markus Stoß EU:C:2010:504; Case C-46/08

Free Movement of Capital and Social Objectives  175 a general margin of appreciation to the Member States in choosing how to regulate gambling services and has accepted a whole array of public interest objectives capable of justifying restrictions on the fundamental freedoms (such as the fight against gaming addiction, the reduction of gambling opportunities, the fight against crime, the general need to preserve public order, protection of consumers of games of chance against fraud on the part of operators, etc). At the same time, it has refined the suitability test so as to include the requirement of consistency: restrictions on gaming activities are suitable to achieve their public interest objectives only insofar as these objectives are being pursued ‘in a consistent and systematic manner’.62 The ‘procedural proportionality test’ has also been described as a ‘good governance’ model, ie, an adjudicative model capable of ensuring the ‘social responsiveness’ and ‘social legitimacy’ of EU free movement law.63 Applying this model in the Volkswagen case,64 Jotte Mulder argues that the judicial scrutiny of the Volkswagen law should have been based on principles of good governance (ie, coherence and transparency of the decision-making process, availability of judicial remedies, and broader coherence of the national policy) and not on substantive efficiency.65 The application of the ‘good governance model’ would not necessarily have led to a different outcome in the case, but it would have allowed diverging corporate governance models to coexist in accordance with good governance standards, ie, it would have respected national regulatory diversity while stimulating EU procedural unity.66 A proportionality assessment based on the procedural standards of the legal instruments granting golden shares in privatised undertakings can indeed guarantee an objective judicial review that respects national regulatory diversity in a sensitive field that is not harmonised at the EU level, while at the same time ensuring compliance with procedural requirements that create a transparent legal framework protecting the legitimate expectations of foreign investors and providing effective judicial remedies against arbitrary decisions of public authorities. The adoption of a good governance adjudicative model shifts the discussion away from politically and ideologically fraught questions to procedural questions, which can be objectively reviewed by the Court. The Belgian case,67 which is in fact the only case where the Court found that the golden shares in question were justified and

Carmen Media Group Ltd EU:C:2010:505; Case C-64/08 Engelmann EU:C:2010:506; Case C-347/09 Dickinger EU:C:2011:582; Case C-186/11 Stanleybet International Ltd EU:C:2013:33; Case C-336/14 Sebat Ince EU:C:2016:72; Case C-375/14 Laezza EU:C:2016:60. 62 Case C-169/07 Hartlauer EU:C:2009:141, para 55; Case C-42/07 Liga Portuguesa, above n 61, para 61. 63 J Mulder, ‘Responsive Adjudication and the ‘Social Legitimacy’ of the Internal Market’ (2016) 22 European Law Journal 597, 599. 64 Case C-112/05 Commission v Germany, above n 6. 65 Mulder, above n 63, 597, 616. 66 ibid. 67 Case C-503/99 Commission v Belgium, above n 13.

176  Ilektra Antonaki proportionate, offers an example of how the principle of legal certainty can operate in the golden shares case law so as to respect the regulatory autonomy of the Member States in the field of corporate governance and property ownership while at the same time applying procedural standards that can guarantee the protection of foreign investment. The judicial scrutiny of procedural requirements such as compliance with the principle of legal certainty, transparency, consistency of national policies and availability of effective judicial remedies falls undoubtedly within the competence of the Court and can effectively promote procedural uniformity in the EU, which can in turn significantly facilitate cross-border investment. In fact, it can be argued that the existence of procedural discrepancies and the lack of transparency often pose more obstacles to cross-border capital flows than the mere participation of the State in the shareholding of privatised undertakings. This is why the application of the ‘good governance model’ in the proportionality assessment of the golden shares case law offers a better adjudicative approach, one that provides all the procedural guarantees necessary for the promotion of foreign investment, while at the same time respecting the discretion of the Member States in the field of corporate governance and promoting a ‘pluralism of values’ in the EU.68

III.  Free Movement of Capital and Privatisations The second part of this chapter focuses on the pre-privatisation phase and the role that EU law plays in the decision of a national government to privatise or nationalise an undertaking – as opposed to the first part of this chapter which focused with golden shares, ie, the special shareholding that a Member State decides to retain once it has privatised an undertaking, thus in the post-privatisation phase. In the pre-privatisation phase, the decision to nationalise or privatise an undertaking touches upon the determination of the national property ownership system, which under Article 345 TFEU remains a national competence. Article 345 TFEU stipulates that ‘the Treaties shall in no way prejudice the rules in Member States governing the system of property ownership’. This provision enshrines the principle of neutrality of the Treaties with respect to the property ownership systems of the Member States. The agnosticism in relation to the regime of property ownership means that the mere fact that an economic activity is carried out by public or private undertakings is not contrary to the Treaties.69 Consequently, the Treaties, in principle, do not preclude the nationalisation or privatisation of undertakings.

68 N Nic Shuibhne, ‘Margins of appreciation: national values, fundamental rights and EC free movement law’ (2009) 34 European Law Review 230, 255. 69 P Craig and G De Búrca, EU Law: Text, Cases, and Materials (Oxford, Oxford University Press, 2011) 1073.

Free Movement of Capital and Social Objectives  177 In formal terms, Article 345 TFEU protects the competence of the Member States to choose between public or private ownership systems.70 This allows Member States to organise independently their national economies, thus giving room for a harmonious coexistence of varieties of capitalism in Europe. The symbolic and perhaps enigmatic nature of this provision has given rise to various interpretations depending on the role of the principle of neutrality in excluding or not excluding certain national provisions from the scrutiny of the fundamental freedoms.71 In particular, three interpretations can be distinguished: (1) the maximalist interpretation; (2) the reductionist interpretation; and (3) the sword interpretation.

A.  The Maximalist Shield Interpretation According to the maximalist shield interpretation, all national measures regulating the property ownership systems of undertakings are shielded from the scrutiny of the four freedoms. This interpretation was expressed by Advocate General Ruiz-Jarabo Colomer in his opinion in the first generation of golden shares cases, which concerned restrictions in the acquisition of shares imposed by legislative provisions that conferred special rights to Portugal, France and Belgium in privatised undertakings active primarily in the energy sector. In his view, the expression ‘system of property ownership’ contained in Article 345 TFEU refers not only to the civil rules concerning property relations but also to any interventionist measure which allows the State to contribute to the organisation of the national economy.72 This includes not only rules, which determine the public or private ownership status of undertakings, but also rules on special rights (golden shares) that Member States retain in privatised strategically important undertakings, as public interventionist means of implementing national economic policy objectives.73

70 For a comprehensive analysis of Art 345 TFEU see P Van Cleynenbreugel, ‘No privatisation in the service of fair competition? Article 345 TFEU and the EU market–state balance after Essent’ (2014) 39 European Law Review 264; B Akkermans and E Ramaekers, ‘Article 345 TFEU (ex Article 295 EC), Its Meanings and Interpretations’ (2010) 16 European Law Journal 292; F Losada Fraga and others, ‘Property and European Integration: Dimensions of Article 345 TFEU’ (2012) 148 Tidskrift utgiven av Juridiska föreningen i Finland 203; W Devroe, ‘Privatizations and Community Law: Neutrality Versus Policy’ (1997) 34 Common Market Law Review 267; R Kovar, ‘Nationalisations – privatisations et droit communautaire’ in J Schwarze (ed), Discretionary Powers of the Member States in the Field of Economic Policies and their Limits under the EEC Treaty: Contributions to an International Colloquium of the European University Institute held in Florence on 14–15 May 1987 (Baden-Baden, Nomos, 1988); T Papadopoulos, ‘Privatized Companies, Golden Shares and Property Ownership in the Euro Crisis Era: A Discussion after Commission v Greece’ (2015) 12 European Company and Financial Law Review 1. 71 Van Cleynenbreugel, above n 70. 72 Opinion of Advocate General Colomer in Cases C-367/98, C-483/99 and C-503/99 Commission v Belgium, Commission v France and Commission v Portugal EU:C:2001:369, paras 54–56. 73 ibid, paras 61 and 91.

178  Ilektra Antonaki

B.  The Reductionist Shield Interpretation According to the reductionist shield interpretation, Article 345 TFEU creates a shield against internal market scrutiny only with respect to measures intrinsically connected with the ownership system established in a Member State (such as a prohibition of privatisation), but not with respect to measures that only indirectly relate to that ownership system (such as golden shares). This interpretation is supported by the Commission, which is of the opinion that Article 345 TFEU is not relevant to golden shares, since they do not concern the public or private ownership system of a company, but rather certain rights and powers relating to the distribution of property ownership of that company between private persons.74 The rationale behind this argument is that once a Member State has implemented a privatisation scheme, it must ensure that the privatised undertaking operates in accordance with the principles of the model of liberal market economy. Therefore, in the Commission’s view, golden shares are not covered by the presumption of validity of Article 345 TFEU. However, when it comes to stricto sensu property ownership rules, the Commission remains neutral: it ‘has always scrupulously ensured neutrality in its dealings with different forms of ownership’75 and ‘has nothing to say on whether companies responsible for providing general interest services should be public or private and is therefore not requiring privatisation’.76 This interpretation was also supported by Advocate General Jääskinen in the Essent case, arguing that in the case of golden shares, the State has already exercised its sovereign right in favour of privatisation, and therefore the privileged treatment it retains for itself within an essentially private property ownership system is not exempted from the Treaty provisions on fundamental freedoms.77 However, in the Essent case, the Dutch legislation at issue established a prohibition of privatisation of the energy distribution system operators. This prohibition was flowing ineluctably from the public ownership of the Dutch energy distribution system and as such it was covered by the presumption of validity of Article 345 TFEU.78 To the extent that this prohibition constituted an intrinsic consequence of the ownership system chosen by the Dutch State, it could not be regarded as an obstacle to the free movement of capital in so far as this system is not discriminatory. On the other hand, any restrictive indirect consequences other than those directly and inevitably stemming from the public or private system of ownership were to be subject to the fundamental freedoms of the Treaty.79

74 Case C-503/99 Commission v Belgium, above n 13, paras 41–42. 75 European Commission (1991) 24 (7/8) Bulletin of the European Commission section 1.2.75. 76 European Commission, ‘Services of General Interest in Europe’ [1996] OJ C281/3, 5. 77 Opinion of Advocate General Jääskinen in Joined Cases C-105/12 to C-107/12 Staat der Nederlanden v Essent NV EU:C:2013:242, para 46. 78 ibid, para 49. 79 ibid, para 42.

Free Movement of Capital and Social Objectives  179

C.  The Sword Interpretation According to the sword interpretation, Article 345 TFEU does not exclude any national measure regarding property ownership from the scrutiny of the fundamental freedoms. The sword interpretation corresponds to the least possible degree of State intervention in the market and it is the one adopted by the Court. While recognising the competence of the Member States to determine their property ownership systems, the Court subjects the exercise of this competence to the rules of the internal market. In particular, in the Essent case, the Court held that the fact that the prohibition of privatisation of the Dutch energy distribution system fell within the scope of Article 345 TFEU did not exclude the application of the free movement of capital under Article 63 TFEU.80 Such a prohibition constituted an ‘impediment’ for the purposes of Article 63 TFEU since, first, it was liable to prevent or limit the acquisition of shares in the undertakings concerned or to deter investors of other Member States from investing in their capital81 and, second, it constituted an obstacle to the raising of capital by the undertakings concerned, since the acquisition of shares was being restricted.82 However, the sword interpretation effectively deprives Member States of the possibility to opt for a public property ownership system. The presumption of neutrality under Article 345 TFEU is transformed into ‘a presumption of illegality of State intervention in the market’83 and the mere fact that the State owns shares in a company is regarded as a restriction on the free movement of capital.84 While it is true that the European Economic Constitution favours economic freedom, liberalisation and deregulation, it nonetheless respects the competence and the discretion of the Member States to choose between public and private property. EU law is neutral with respect to privatisations and nationalisations. The fact that Article 345 TFEU is a symbolic provision does not mean that it is deprived of any legal content. Its symbolic importance lies in the fact that it reflects the fundamental compromise between State and market that was achieved in the post-war construction of Europe. If a prohibition of privatisation is regarded as a restriction on the free movement of capital, this is liable to render Article 345 TFEU devoid of any legal content and meaning whatsoever. Therefore, it is argued that from a legal point of view the most consistent interpretation of Article 345 TFEU is the reductionist shield interpretation according to which measures intrinsically linked

80 Joined Cases C-105/12 to C-107/12 Essent NV EU:C:2013:677, para 37. 81 ibid, para 41. 82 ibid, para 42. 83 E Szyszczak, The Regulation of the State in Competitive Markets in the EU (Oxford, Hart Publishing, 2007) 14. 84 J Snell, ‘Economic Justifications and the Role of the State’ in P Koutrakos, N Nic Shuibhne and P Syrpis (eds), Exceptions from EU Free Movement Law: Derogation, Justification and Proportionality (Oxford, Hart Publishing, 2016) 25.

180  Ilektra Antonaki with the national property ownership system, such as a prohibition on privatisation, should be excluded from the scope of the free movement provisions by virtue of Article 345 TFEU.

D.  The Compromise: The Use of Article 345 TFEU as a Justification Aware of the sensitive nature of the case and the possible consequences of its judgment, the Court tried to moderate its approach in the Essent case: with an enigmatic and ambiguous language it implicitly accepted that Article 345 TFEU could be used as a justification ground for restrictions on capital movements. Even though in earlier case law it had explicitly excluded the possibility of invoking Article 345 TFEU in order to justify a restriction on the free movement of capital,85 it nonetheless explained [t]hat does not however mean that the interest underlying the choice of the legislature in relation to the rules on the public or private ownership of the electricity or gas distribution system operator may not be taken into consideration as an overriding reason in the public interest.86

The Court emphasised that the Essent case should be clearly distinguished from the golden shares case law, as the former related to ‘an absolute prohibition of privatisation’, whereas the latter ‘concerned restrictions created by privileges which the Member States attached to their position as a shareholder in a privatised undertaking’.87 For these reasons, the Court held that the reasons underlying the choice of the rules of property ownership adopted by the national legislation within the scope of Article 345 TFEU constitute factors which may be taken into consideration as circumstances capable of justifying restrictions on the free movement of capital.88

The use of Article 345 TFEU as a justification ground could be regarded as opening the door for the use of economic considerations for the purpose of justifying restrictions on the fundamental freedoms. According to settled case law, purely economic grounds cannot serve as justification for obstacles prohibited by the Treaty (known as the ‘economic rule’).89 However, the Court has accepted that a restriction on free movement is lawful whenever a measure is based equally on

85 Case C-171/08 Commission v Portugal, above n 13, para 64. 86 Joined Cases C-105/12 to C-107/12 Essent NV, above n 80, para 53. 87 ibid, para 54. 88 ibid, para 55. 89 Case 7–61 Commission v Italy (pork) EU:C:1961:31 329; Case 72/83 Campus Oil EU:C:1984:256, paras 35–36; Case C-398/95 SETTG EU:C:1997:282, paras 22–23; Case C-35/98 Verkooijen EU:C:2000:294, paras 47–48; Case C-309/99 Wouters and others EU:C:2002:98, para 52; Case C-201/15 AGET Iraklis EU:C:2016:972, para 72.

Free Movement of Capital and Social Objectives  181 economic grounds and on grounds which are admissible under the Treaty.90 This is known as the ‘further purpose’ doctrine, under which it is exceptionally possible to justify protectionist measures if, apart from their economic objective, they pursue simultaneously another non-economic objective.91 Furthermore, the ‘further purpose doctrine’ has recently been expanded so as to cover not only objectives with an economic dimension which fall within the ambit of one of the Treaty derogations, but also objectives with an economic dimension which fall within the ambit of an overriding reason in the public interest. In AGET Iraklis, the Court acknowledged that ‘considerations connected with the maintenance of employment may, under certain circumstances and conditions, be acceptable justifications’.92 These considerations include the ‘protection of workers’ recognised as an overriding reason in the public interest as well as ‘the encouragement of employment and recruitment which, being designed in particular to reduce unemployment, constitutes a legitimate aim of social policy’.93 Drawing from this ruling, in a recent case concerning the re-privatisation of the Portuguese airline TAP, Advocate General Campos Sanchez-Bordona explained that the ‘economic rule’ ‘does not mean that the Court of Justice is impervious to Member States’ economic interests when these are put forward as grounds for hindering freedom of establishment in pursuit of a legitimate social policy objective’.94 It has been argued that economic considerations expressly provided for by the Treaty, such as services of general economic interest under Article 106 TFEU, are not caught by the ‘economic rule’, as the Treaty itself allows their use as justification grounds by the Member States.95 The same could be argued with respect to Article 345 TFEU. Indeed, the property ownership systems of the Member States necessarily entail economic considerations, which in principle cannot be used as justification grounds according to the ‘economic rule’. However, the fact that this is an explicit provision of the Treaty renders the ‘economic rule’ inapplicable. Therefore, Article 345 TFEU can be used as a justification ground despite the fact that it inevitably serves economic interests. Furthermore, the choice of national property ownership system does not only pursue economic objectives, but also other objectives of a non-economic nature that could perhaps fall within the scope of the Treaty derogations. Accordingly, on the basis of the ‘further purpose’ doctrine, Article 345 could be used as a justification ground. This was in essence the approach that was implicitly chosen in Essent, where the Court ruled that reasons underlying

90 Case 72/83 Campus Oil, above n 89, paras 35–36. 91 S Arrowsmith, ‘Rethinking the approach to economic justifications under the EU’s free movement rules’ (2015) 68 Current Legal Problems 307, 320. 92 Case C-201/15 AGET Iraklis, above n 89, para 75. 93 ibid, paras 73–74. 94 Opinion of Advocate General Campos Sánchez-Bordona in Case C-563/17 Associação Peço a Palavra EU:C:2018:937, para 95. 95 P Oliver, ‘When, if ever, can restrictions on free movement be justified on economic grounds’ (2016) 41 European Law Review 147.

182  Ilektra Antonaki the choice of a public ownership system under Article 345 TFEU could be taken into consideration as circumstances capable of justifying restrictions on the free movement of capital resulting from an absolute prohibition of privatisation. It should be acknowledged that the approach that was ultimately chosen by the Court in the Essent case appears to be an important compromise aiming to strike a delicate balance between the fundamental freedoms and the public ownership of undertakings intended to safeguard public interest objectives. The departure from a strict application of the ‘economic rule’ was perhaps inevitable in light of the fact that there are certain economic objectives which are so inextricably linked with or dictated by social objectives that they cannot be separated from each other. The use of Article 345 TFEU as a justification for the restriction imposed on the free movement of capital by the prohibition of privatisation can be regarded as a recognition by the Court of the exceptionally sensitive nature of the political choice between public and private ownership. In light of the politically sensitive nature of the case, the Court tried to mitigate the effects of its judgment by ruling that the reasons underlying the choice of a public ownership system under Article 345 TFEU could be used as a justification for the capital restriction imposed by the prohibition of privatisation. This can be regarded as a compromise aiming at reconciling capital liberalisation with public interest objectives. The underlying rationale of the legal reasoning in Essent is the acknowledgment that in certain sensitive sectors public ownership can protect essential social objectives, which would otherwise be threatened by market forces. This is linked to the overarching theme of this contribution, which is the fundamental clash between economic freedoms and social values and the efforts to strengthen the social component of the European edifice and achieve a ‘social market economy’, as enshrined in Article 3(3) of the Treaty on European Union (TEU). However, it should be noted that the distinction between the use of Article 345 TFEU as an exemption from the scope of capital restrictions and the use of Article 345 TFEU as a justification is of paramount importance and can have significant consequences for the outcome of the case. This is because in the first scenario the national measure is excluded altogether from internal market scrutiny, whereas in the second scenario the measure constitutes a restriction, which then needs to be balanced against the conflicting public interest objective subject to a strict proportionality assessment. Therefore, from a legal consistency perspective, a measure such as a prohibition of privatisation touches upon the essence of the national competence to opt for a public ownership system and, therefore, should be excluded from the scope of the internal market provisions according to the reductionist shield interpretation.

IV. Conclusion The protection of public interest objectives in the privatisations and golden shares case law relates to the broader discussion about the so-called ‘social deficit’ in

Free Movement of Capital and Social Objectives  183 Europe and the fundamental clash between economic freedoms and social values. The discussion about the ‘social deficit’ of the EU has attracted anew significant scholarly attention in the aftermath of the euro crisis.96 Indeed, the debate regarding the strengthening of the social dimension of the European integration project has grown even more in scale as a result of the disrupting social and economic effects of the austerity policies that were adopted as a response to the financial crisis. In the aftermath of the crisis, the need to protect public interest objectives against unfettered market forces has become pressing and has mobilised support for projects aiming at reinforcing the social component of the EU, such as the recent initiative of the Commission for a European Pillar of Social Rights.97 In this context, the overarching theme of this chapter has been the delicate balancing exercise between economic freedoms and social values, especially in times of ideological contestation over the social face of Europe. For this purpose, the chapter has drawn from the theoretical underpinnings of ‘social market economy’ enshrined in Article 3(3) TEU and has attempted to provide a normative assessment of the clash between economic and social objectives in the field of free movement of capital and more specifically in the case law concerning golden shares and privatisations. In a nutshell, the chapter has put forward two main modifications in the judicial interpretation of the free movement provisions which could help address the aforementioned concerns regarding the division of competences and the protection of societal values: first, the recalibration of the ‘capital restrictions’ test in the golden shares case law by reference to a Keck-inspired notion of ‘investment arrangements’; and second, the rediscovery of the principle of neutrality under Article 345 TFEU as a legal provision which shields national decisions to maintain public ownership in undertakings from internal market scrutiny. Seen from a broader perspective, the two main proposals of this chapter attempt to bridge the gap between the State and the market and to create sufficient room for the State to engage in economic activities as a market participant. It is argued that the State should in principle be able to participate in the market under equal conditions with other private market operators and that forms of public ownership of undertakings or special shareholding in privatised undertakings should not be regarded as inherently problematic from the perspective of EU internal market law. The participation of the State in the market might even be desirable in some cases in order to pursue vital public interest objectives, such as the protection of 96 A Hinarejos, The Euro Area Crisis in Constitutional Perspective (Oxford, Oxford University Press, 2015); C Kilpatrick, ‘Are the bailout measures immune to EU Social challenge because they are not EU Law?’ (2014) 10 European Constitutional Law Review 393; A Poulou, ‘Financial assistance conditionality and human rights protection: What is the role of the EU Charter of Fundamental Rights?’ (2017) 54 Common Market Law Review 991; B De Witte and C Kilpatrick, ‘A comparative framing of fundamental rights challenges to social crisis measures in the Eurozone’ (2014) 1 European Journal of Social Law 2. 97 European Commission, ‘Recommendation on the European Pillar of Social Rights’ C(2017) 2600 final; European Commission, ‘Communication of 26 April 2017 establishing a European Pillar of Social Rights’ COM(2017) 250 final; S Garben, ‘The European Pillar of Social Rights: Effectively Addressing Displacement?’ (2018) 14 European Constitutional Law Review 210.

184  Ilektra Antonaki workers or the provision of services of general interest. Overall, the interpretative approaches suggested in this contribution aim at promoting respect for the delicate balance of competences between the EU and the Member States and can facilitate the reconciliation of the capital freedom with public interest objectives. The reinterpretation of the free movement of capital in the privatisations and the golden shares case law will allow the EU to follow the contemporary economic school of thought advocating a more restrictive approach regarding international capital flows in order to prevent the risks of financial instability and social inequality. At the same time, it will reinforce the social dimension of the European integration project at this critical juncture of ideological contestation over the future of Europe.

part iii The Internal Market in Digital Times

186

9 Single Market 2.0: The European Union as a Platform ANDREA RENDA

Often defined as ‘unfinished business’, the European Union’s market integration process appears to have become more fragile than ever at the beginning of the new decade. By 2010, the Monti Report denounced the existence of a single market ‘fatigue’, which made it difficult to complete the market integration process, especially in most difficult areas such as services.1 Today, Brexit potentially threatens the future attractiveness of the single market, by depriving the Union of its third largest economy and leading to an unprecedented thorn in the EU’s side, as a Member of the Union sets sail. At the same time, the post-Brexit single market may become more cohesive and ambitious, as one of the most reluctant Member States leaves the group: the Union may also have the opportunity to rediscover some of the features of continental Europe’s legal and economic traditions, from Civil Law rules to state-led industrial policy, which faced obstacles when the UK was in the Union.2 Against this background, the challenges for the single market project do not end with Brexit. To the contrary, EU policymakers are confronted with a frustrating prospect: as they try to complete the single market, technological evolution is pushing the frontier of integration further, requiring new efforts and policies to fully achieve the desired goal. In particular, the digital transformation is changing the traditional, textbook economics of market integration, based on tenets such as economies of scale and the four freedoms. The rise of the digital economy requires a radical change in the policies for the single market, as well as in the trade policies that underpin the whole market integration process. Trends such as the virtualisation, servitisation and platformisation of the economy (as described below), coupled with the rise of the Internet of Things and Artificial Intelligence, 1 See A new strategy for the Single Market. At the service of Europe’s economy and society Report to the President of the European Commission José Manuel Barroso, by Mario Monti, 9 May 2010. 2 D Kalff and A Renda, ‘Hidden Treasures. Mapping Europe’s Sources of Competitiveness Advantage in Doing Business’ (2019) CEPS Monograph; and A Bradford, The Brussels Effect. How the European Union Rules the World (Oxford, Oxford University Press, March 2020).

188  Andrea Renda make market integration at once more appealing and increasingly challenging for EU policymakers, projecting the single market into a complete new dimension, in which the ‘Fifth Freedom’ (the free circulation of non-personal data) is intertwined with new concerns with the need to protect fundamental rights, and at the same time secure Europe’s technological sovereignty. As EU institutions were struggling to complete the ambitious Digital Single Market strategy formulated by the Juncker Commission, technology has changed so fast that brand new policy initiatives are needed: the Von der Leyen Commission seems to have marked a significant change towards a more assertive and futureoriented approach to digital policy. The new pillars of the Single Market 2.0 are not focused any more on platform regulation, data protection and the free flow of nonpersonal data; while these remain very important pillars, the future of the single market will require that the whole internal market is seen as a layered ecosystem, in which infrastructure, rules, protocols and standards become a platform for large and small companies to develop value-added solutions to the benefit of all European consumers. This ‘EaaP’ (Europe as a Platform) approach may also induce a change of terminology: what used to be mutual recognition will now mostly be related to interoperability; what used to be subsidiarity is translated into a choice between centralised, distributed and decentralised governance; open interconnection becomes ‘open Application Programming Interfaces’ (Open API), and is applied far beyond network industries; and the free circulation of people is enhanced with a strong digital identity and verification layer. Importantly, especially after the appointment of Commissioner Thierry Breton, the Single Market 2.0 is becoming the locus of data spaces and ecosystems as the basic pillars of the future EU ‘competitive sustainability’ agenda. No doubt, achieving the Single Market 2.0 requires strong political commitment and will not prove easy, if the foundations of economic integration remain as fragile as they are today. The COVID-19 pandemic has shown very clearly the difficulty of keeping the EU’s ‘crown jewel’ alive in times of despair and shortage of essential goods, such as medical protective equipment. The deterioration of the rule of law in some of the Member States (for example, Poland, Hungary), the suspension of Schengen, the overall lack of solidarity between Member States and the temptation to use digital technology (for example, in the form of contact-tracing apps) in a way that jeopardises fundamental rights, are only a few examples of a more general crisis of the European integration project, which the von der Leyen Commission has barely managed to contain. At the same time, the COVID-19 pandemic is also accelerating the digital transformation, with the online economy becoming paramount for economic and social relations: this, in turn, makes the development of rules and standards for a trustworthy single market even more urgent.3 3 See Communication from the Commission to the European Parliament, the European Council, the Council, the European Central Bank, the European Investment Bank and The Eurogroup, ‘Coordinated Economic Response To The Covid-19 Outbreak’ COM(2020)112 final, 13 March 2020.

Single Market 2.0: The EU as a Platform  189 Can code succeed where law and politics have so far partly failed? In the remainder of this chapter, I outline the possible contours of a ‘Single Market 2.0’. Section I below describes the current and upcoming waves of digital transformation as featuring very different economic paradigms and consequences for EU policy. Section II discusses the reconfiguration of the single market as a layered ecosystem and describes the current Digital Single Market Strategy and the upcoming initiatives outlined by the European Commission, and partly affected by the COVID-19 pandemic. Section III provides a layered architecture for the Single Market 2.0 and outlines possible further avenues to speed up reform in a way that is consistent with all the cornerstones of the EU strategy for growth and sustainable development.

I.  Two Waves of Internet Evolution Information technology (IT) experts traditionally approach architectural problems through modular structures, by distinguishing different layers and components of complex system goods and working through different options regarding their interoperability.4 Different choices in this respect lead to more closed architectures (for example, the early Apple Macintosh), or more open architectures, in which different layers and components can be produced according to standard specifications by more than one firm (for example, the early Microsoft Windows). The history of IT also suggests that no governance architecture is fully open, and typically one or more layers become dominated by one or a few players due to the emergence of network externalities, which lead to ‘winner-take-all’ effects and often highly concentrated market structures within some of the layers.5 The internet is itself depicted as a complex layered architecture, based on a massive physical layer, encompassing fixed and wireless communications systems, submarine cables and satellite systems, massive internet exchanges and data centres. The internet age constituted a major enhancement of the original layered architecture of the personal computer, with the introduction of globalscale networking possibilities between computers and similar devices at the physical layer, and the definition of open protocols and standards that defined the traffic rules for data at the so-called ‘logical layer’. Open standards such as the Internet Protocol and the File Transfer Protocol defined the characteristics of the internet as a means of communication: as observed in the 1990s by Lawrence 4 See RN Langlois, ‘Modularity in technology and organization’ (2002) 49 Journal of Economic Behavior & Organization 19; RN Langlois and P Robertson, ‘Networks and innovation in a modular system: lessons from the microcomputer and stereo component industries’ (1992) 21(4) Research Policy 297. See also H Chesbrough, ‘Towards a dynamics of modularity: a cyclical model of technical advance’ in A Prencipe and M Hobday (eds), The Business of Systems Integration (Oxford, Oxford University Press, 2004); and H Chesbrough, Open Innovation (New York, Free Press, 2003). 5 C Shapiro and H Varian, Information Rules. A Strategic Guide to the Network Economy (Boston, MA, Harvard Business School Press, 1999).

190  Andrea Renda Lessig, in a digital environment, ‘code, not law, defines what’s possible’.6 In the case of the early internet, the openness and neutrality of the standards defined at the logical layer determined the rise of the internet as a formidable means of communication between peers, and a vehicle for permissionless innovation, in which at once users could preserve their anonymity (‘nobody knows you are a dog’, as in Peter Steiner’s well-known New Yorker cartoon in 1993); and developers could work on their applications without having to seek anyone’s permission to reach the marketplace. That said, the internet exacerbated the features of the original layered architecture of the personal computer: the introduction of a new layer, with an end-to-end7 architecture and a growing ability to support data flows, generated an explosion in the amount of data available to end users, which in turn converted into a ‘poverty of attention’.8 The firms that managed to capture a significant share of that scarce resource (user attention), ended up becoming trillion-dollar corporations, or (as they are sometimes called in the economics literature) ‘superstar firms’. Accordingly, the rise of the digital, interconnected economy has also coincided with a rapid, necessary transformation of the ecosystem dynamics, which would soon come back to haunt the early creators of the internet, confronting them with a rising market concentration, a declining neutrality, and gradual attempts to depart from the original end-to-end design.9 This is what I call the ‘first wave of transformation’ of the internet.

A.  The First Transformation of the Internet: Virtualisation, Servitisation and Platformisation During the past two decades, the internet has undergone a swift transformation, which led to the emergence of a more diverse layered architecture, and very peculiar forms of governance that hardly existed in the ‘bricks and mortar’ world. Understanding them is useful to identify the direction taken by the EU institutions in trying to reap the benefits of the digital ecosystem, and at the same time minimising its associated risks. First, the emergence of an end-to-end infrastructure, fuelled by growing computing capacity, high-capacity networks and wireless connectivity, greatly 6 L Lessig, Code: And other Laws of Cyberspace (New York, Basic Books, 1999). 7 D Autor, D Dorn, LF Katz, C Patterson and J Van Reene, ‘The Fall of the Labor Share and the Rise of Superstar Firms’ (2020) 135 Quarterly Journal of Economics 645. 8 HA Simon, ‘Designing Organizations for an Information-Rich World’ in M Greenberger, Computers Communication, and the Public Interest (Baltimore, MD, The Johns Hopkins Press, 1971) 40–41. 9 MA Lemley and L Lessig, ‘The End of End-to-End: Preserving the Architecture of the Internet in the Broadband Era’ (2001) 48 UCLA Law Review 925. A Renda, ‘Competition, Neutrality and Diversity in the Cloud, Communications & Strategies’ (2012) 85 1st Quarter 23; A Renda, ‘Net Neutrality and Mandatory Network-Sharing: How to disconnect the continent’ (2013) CEPS Policy Briefs; and A Renda, ‘Antitrust, regulation and the “neutrality trap”’ (2015) CEPS Special Report 104.

Single Market 2.0: The EU as a Platform  191 expanded the possibilities for the digital economy to permeate the economy. The emergence of cloud computing made it possible for small companies to avoid buying or leasing hardware and downloading software and applications: these traditional transactions were replaced by ‘everything as a service’, which led to enormous advantages both for individuals and businesses. The transition towards a ‘cloud era’ allowed personal devices to become increasingly agile, while users were able to access hardware and software located in the cloud, as well as retrieve their files from cyberspace. Put more simply, a limitless ‘office LAN’ where the main server was not located downstairs, but potentially on the other side of the globe.10 This was the realisation of the so-called ‘age of access’ already evoked by scholars in the 1990s. An age in which products and services are dematerialised to an extent that ownership and property rights become less important, and access rights become gradually more dominant.11 The progress observed in ubiquitous connectivity and in compression techniques, coupled with enhanced possibilities to capture end users’ attention, has gradually led to the emergence of access-based services. These include a variety of new business models, from pure streamingbased content access services (Netflix, Spotify) to intermediate forms (Apple Music + iTunes + Apple TV) which contemplate both ownership and access; and the so-called ‘sharing economy’, based on a combination of network effects, granularity and reputational effects (for example, Airbnb, Uber). Many of these services rely on the ‘cloud’ as a key resource for virtual access and use of IT resources.12 Thanks to the end-to-end nature of the internet, the digitisation of information and ever-increasing capacity and connectivity, many parts of the economy 10 Cloud computing is a general-purpose technology of the IT field which became widely available in the late 2000. Vaquero et al define it as ‘a large pool of easily usable and accessible virtualized resources (such as hardware, development platforms and/or services). These resources can be dynamically reconfigured to adjust to a variable load (scale), allowing also for an optimum resource utilization. This pool of resources is typically exploited by a pay-per-use model in which guarantees are offered by the Infrastructure Provider by means of customized Service Level Agreements’. See L Vaquero, L RoderoMerino, J Caceres and M Lindner, ‘A Break in the Clouds: Towards a Cloud Definition’ (2009) 39 Computer Communication Review 50. Cloud architectures are conceived to be very simple for end users but feature a very complex architecture ‘behind the curtains’. As an example, Apple’s iCloud allows the syncing of various devices with the cloud, such that the end user always enters the same environment regardless of the device used to connect to the network. Similar strategies have been pursued for the end user market by Google (Android), Microsoft (Azure) and Amazon (AWS). The most widely acknowledged taxonomies of cloud computing are those that relate to the basic cloud ‘modes’ (ie Public, Private, Hybrid) and the main cloud ‘types’ (ie, Saas, AaaS, IaaS, PaaS). The provision of platform as a service (PaaS), for example, leaves more control of the configuration to the client than mere application as a service (AaaS) or software as a service (SaaS) modes. At the same time, private clouds are certainly more customised to the client’s needs than hybrid or public clouds, which however enjoy clear economies of scale. 11 See, eg, RW Gomulkiewicz, ‘The License is the Product: Comments on the Promise of Article 2B for Software and Information Licensing’ (1998) 13(3) Berkeley Technology Law Journal 891. 12 A specific case is 3D printing, which leads to a de-materialisation of the product, but rather than its remote access, entails a remote re-production of the product. This changes the role of the players active in the production cycle: in 3D printing, the borderline between manufacture and service provision is blurred due to uncertainty as to who should be assumed to be the manufacturer of the product, particularly when a 3D printer has been used somewhere in the value chain.

192  Andrea Renda have gradually transitioned towards virtualisation and servitisation, two intertwined phenomena that made the economy more agile, but also came with a price to pay in terms of inequality.13 The ‘uberisation’ of many sectors such as passenger transport, accommodation, childcare, handyman jobs, IT work (for example, Mechanical Turk, Upwork) and low-skilled jobs and many other markets has led to extreme situations in which humans, themselves, are offered ‘as a service’.14 The end-to-end, digitised nature of the internet also determined the rise of peer-to-peer interaction in various forms. The transition towards an access-based economy initially affected audiovisual content, creating significant disruption (first with peer-to-peer file-sharing, later with streaming-based services that almost restore the industry’s profitability). Later, the rise of the collaborative economy reached unprecedented levels: in 2019 Bank of America Merrill Lynch valued the worldwide sharing economy at US$250 billion and further estimated that US$6 trillion in commerce could be disrupted by the sharing economy across sectors such as transportation, travel, food, retail and the media. This, representing approximately 8 per cent of global GDP, is supported, inter alia, by the fact that eight of the world’s 10 largest start-ups based on valuation are in fact sharing economy businesses. This is likely to be further exacerbated by the impact of the COVID-19 pandemic, which led to an explosion in the demand for online services. The proliferation of information sources and the need to store and retrieve data for a multitude of user transactions also led to the rise of fierce competition for user attention (so-called ‘competition for eyeballs’). The explosion of internet traffic in the 1990s and 2000s, powered by parallel streams of evolving technologies (data storage, broadband communications, data compression, innovation in traffic management) led to an emerging need for solutions that would reduce complexity: this solution was spontaneously developed by market forces, and mostly took the form of industry convergence towards a limited number of de facto industry standards at the higher layers of the architecture.15 Today, the digital ecosystem has 13 The servitisation (or ‘servicification’) of the economy is a well-known process that largely pre-dates the internet era. Janja Hojnik reminds that the de-industrialisation of developed economies started in the 1950s and the value added by manufacturing as a percentage of GDP is now below 15% in most OECD countries, and that ‘economic studies show that servitization is one of the economic megatrends of modern society, along with globalization, encompassing a broad range of business models that are currently occurring on the market’. See J. Hojnik, ‘The servitization of industry: EU law implications and challenges’ (2016) 53 Common Market Law Review 1575. 14 See J Prassl, Humans as a Service: The Promise and Perils of Work in the Gig Economy (Oxford, Oxford University Press, 2018). 15 Examples of de facto ICT industry standards in the pre-Internet age include Lotus 123, WordPerfect and other applications based on the original IBM PC architecture and the MS-DOS. Later, Windows 3.1 and Windows 95 (which ushered the Internet age) became widely diffused de facto industry standards. The case of Microsoft Windows is perhaps the most telling in the evolution that the ICT went through during the 1990s: the modular architecture of the personal computer entailed the existence of one layer (at the time, the OS layer), which would end up being essential in terms of connecting hardware with software and determining the compatibility and interoperability requirements of the whole system.

Single Market 2.0: The EU as a Platform  193 evolved into a much more diverse environment: the original open internet architecture coexists with various multisided platforms, which coordinate, steer and manage the innovation taking place at the higher layer of the internet architecture. This phenomenon, often called ‘platformisation’, bears far-reaching consequences for innovation, competition and public policy.16 While a full review of the impact of platformisation on public policy would go beyond the scope of this chapter, it is worth looking at the peculiar governance features of platforms, as well as their impact on value distribution in the digital economy. A close look suggests that they are different from the traditional corporations. They are, indeed, a hybrid between the firm and the market, compared with the traditional distinction made in social sciences.17 So-called multisided platforms are hierarchical structures serving various categories of user, in which most of the traditional activities of a firm are outsourced, automated, or ‘heteromated’ (automated through the help of third parties).18 A good example is Uber, a ride-hailing platform that matches independent contractors (drivers) with end users (passengers), in which the management and control of the former is largely done through a combination of algorithms and user reviews. The corporate structure and size of the firm in terms of factors of production (capital and labour) is extremely small compared with the number of transactions that the platform generates and profits from. For example, Uber has approximately 20,000 employees, but ‘employs’ more than four million drivers. The peculiar structure of platforms, their relatively small size and the powerful network externalities that sustain their competitive position in the market contribute to strengthening their bargaining power vis-à-vis all categories of user. Such stronger market power increasingly contrasted with a regulatory approach largely left these powerful intermediaries untouched, as a legacy of the early days of the ‘neutral’ internet. The platformisation of the internet had little to do with the early vision of a neutral, end-to-end ‘network of networks’, open to all for permissionless innovation. Rather, it led to the rise of new gatekeepers, which occupy an almost unattackable position and continue to reap a large share of the value associated with the transactions they not-so-neutrally orchestrate. Not surprisingly, the need to govern the humongous amount of data their ecosystems generate also led these companies to increasingly invest gigantic sums in data-hungry machine learning systems, which came to dominate the digital environment as well as the AI landscape. From Netflix’s recommendation system to Google’s search engine, enormous investment has reinforced and ring-fenced the market positioning of 16 See A Gawer, Platforms, Markets and Innovation (Cheltenham, Edward Elgar, 2009). 17 See RH Coase, ‘The Nature of the Firm’ (1973) 4(16) Economica, New Series 386. OE Williamson, The Economic Institutions of Capitalism (New York, The Free Press, 1985). OE Williamson, ‘Public and Private Bureaucracies: A Transaction Cost Economics Perspective’ (1999) 15 Journal of Law, Economics, & Organization 306. OE Williamson, ‘Transaction-Cost Economics: The Governance of Contractual Relations’ (1979) 22 The Journal of Law & Economics 233. 18 HR Ekbia and BA Nardi, Heteromation, and Other Stories of Computing and Capitalism (Cambridge, MA, MIT Press, 2017).

194  Andrea Renda these platforms, enabling them to personalise product offerings, prices and conditions to an extent that no competitor could even come close to attaining. Virtualisation, servitisation and platformisation have been extremely important drivers of change of the internet ecosystem in the past two decades and have led to a massive change in the economy, largely consisting in a re-intermediation, rather than a dis-intermediation, of previous market transactions. Where the new business models have preserved the original end-to-end nature of the internet, new governance forms such as distributed ledger technologies have started to surface, thanks to technological advances, cost reductions and the possibility to rely on peer-to-peer computing and the power of direct network externalities. Where companies have departed from the original architecture, multisided markets have replicated the characteristics of one-to-many communications systems, such as television (for example, Netflix), and have largely profited from indirect network externalities. The thirst for personal data, coupled with the users’ relative lack of awareness of the value of the data they shared and contributed to the working of large-scale algorithms, led these digital intermediaries to prosper by shrinking their size compared with the traditional firm, externalising most functions, reaping advertising benefits and enjoying a largely unregulated space. Not surprisingly, this first wave of digital transformation created important tensions among regulators, in particular in the European Union. The more digital transformation was permeating traditional markets, the more the differences in the regulatory treatment of incumbent players and digital firms started to tilt the market balance in favour of the latter. The greater the diffusion of data-driven business models, the greater the tensions in terms of data protection, and the loss of control of personal data for end users. The stronger the centripetal forces unleashed by network externalities and platformisation, the greater the polarisation of market power and profits in the hands of a fistful of companies.19 The more multisided platforms conquered the market, the more precarious most workers’ conditions became, the more obscure the algorithmic practices behind their selection and reward, and the weaker their access to social dialogue. The greater the imbalance between firm size and overall profits, the more evident the need for digital taxation based on the place where value is created and revenues are reaped, rather than the place where the digital company has its headquarters. The reaction of the European Union to this first wave of internet transformation was initially very slow, then gradually more assertive and determined. The General Data Protection Regulation, which entered into force in May 2018, is the poster child of a generation of reforms that have attempted to restore balance in the digital ecosystem, by establishing principles such as ‘data minimisation’ and ‘user control over data’, which were echoed by legislation in many other legal systems, from Brazil to Japan and California. A regulation on the free flow of nonpersonal data tried to couple restrictive rules on personal data with expansive rules

19 See

T Philippon, The Great Reversal (Cambridge, MA, Harvard University Press, 2019).

Single Market 2.0: The EU as a Platform  195 on non-personal data flows, so far with little impact. Besides antitrust investigations and fines against giants like Google, new regulations expanded the remit of competition rules in areas such as the relationship between platforms and businesses (P2B), echoing national rules on abuse of economic dependence and abuse of superior bargaining power. After interventions on the tax side to counter tax rebates for digital giants such as Apple (in Ireland), work on a web tax has started to take shape, alternating with international efforts in the context of the Organisation on Economic Co-operation and Development (OECD). And the first attempts to attribute entitlements over data emerged in the field of agriculture, where a code of conduct seeks to empower farmers in reaping value from their data.20 And the approach to promoting competition in regulated sectors started to move from traditional access obligation related to infrastructure, towards mandatory interoperability obligations, initially imposed on industry incumbents, rather than on tech giants: the case of the Second Payment Services Directive (PSD2) portrays the awakening of the EU institutions to the use of Application Programming Interfaces as a powerful way to impose interoperability obligations, with the (hopefully meaningful) consent of the data subject whose credit data are now moved to third-party providers. Against this background, the Juncker Commission also worked behind the scenes to speed up the reflection on AI rules, on the impact of the digital transformation on labour markets, as well as on the achievement of interoperability between administrations at the EU, national and local level. All these initiatives took the form of expert reflection (through ad hoc high-level groups) or voluntary frameworks (as in the case of the ‘Programme on Interoperability solutions for public administrations, businesses and citizens’, ISA2), but had the merit to pave the way for an expected acceleration in the coming years, with more binding initiatives. The urgency of a reaction to the evolution of the digital ecosystem was felt more strongly by Member States, leading the European Commission in some cases to adopt initiatives in the attempt to stop the proliferation of inconsistent rules at the national level (for example, for web taxes, as well as for AI). Overall, the EU’s reaction to the first wave of transformation of the internet arrived too late, leaving the EU behind other superpowers in terms of preparedness and adaptation to the new paradigms of the digital economy. As also acknowledged by the European Commission, the United States and China have clearly taken the lead on the cloud-dominated, platform-dominated, machinelearning-led digital environment that has emerged in the past two decades, and only started to concretely react halfway through the Juncker years. The lack of anticipatory policymaking, for an inevitably cumbersome bloc of 28 states, created a significant lag between the digital transformation and the EU’s policy response. That said, once the EU institutions managed to concretely respond, their ability to embed policy proposals into a concrete, solid and comprehensive set 20 See A Renda, N Reynolds, M Laurer and G Cohen, ‘Digitising Agrifood’ (2019) CEPS and BCFN Joint Monograph.

196  Andrea Renda of principles led to the emergence of the first real corpus of legal rules aimed at creating a more socially and economically sustainable environment for the internet age. The complete absence of similar rules in most other legal systems made the EU a real pioneer in this policy domain, despite the rather timid approach adopted in several areas. Anu Bradford even sees a magnified ‘Brussels effect’ on global digital policy compared with the many other areas in which the EU already exerts a significant ‘normative power Europe’.21 This, in turn, encouraged the new European Commission to consider adopting a more assertive approach to digital policy, which culminated in a completely new stance on the single market.

B.  The Second Transformation Wave: The Internet of Things, the Rise of Artificial Intelligence and the Emergence of Distributed Architectures The von der Leyen Commission had, by the end of 2019, taken stock of Europe’s competition positioning in the digital environment with a degree of despair. Statements on Europe’s lag compared with the increasingly battling United States and China proliferated: the world is dominated by a fistful of cloud operators, most of which are American, none of which are European; there are no European companies among the top 20 global tech firms; the data train has left the station, as more than 90 per cent of the data in the Western world are stored in the United States; China is going to dominate 5G, as the United States rules the world on platforms and applications; Europe will never manage to match the level of AI investment of the United States and China. However, digital technology never stands still. As policymakers struggle to address the ‘pacing problem’ and respond to the first wave of digital transformation, the evolution of technology is already paving the way for a transformation in the digital environment. And indeed, the next generation of internet transformation may create new opportunities for Europe to regain its role in the global competition for digital solutions. In particular, the next few years will mark the blossoming of the Internet of Things (IoT). The physical layer will indeed be enriched by the availability of smart, cyber-physical objects, which can enable decentralised data production, communication and processing, requiring storage and intelligence to be increasingly distributed. The projections of the Internet of Things are breathtaking, with the number of connected objects poised to skyrocket to one trillion by 2035.22 More specifically, IoT systems are essentially organised around four main (sub-) layers: directly attached to the ‘things’ are sensors, antennas and actuators, which 21 See Bradford, above n 2; and A Renda, ‘Making sense of the “Geopolitical Commission” Insights from the TRIGGER project’ (2020) CEPS Policy Brief. 22 D Gros, ‘Global Trends to 2035: Economy and Society, Report for the European Parliament’ (2019), available at: www.europarl.europa.eu/RegData/etudes/STUD/2018/627126/EPRS_STU (2018)627126_EN.pdf.

Single Market 2.0: The EU as a Platform  197 can take a wide variety of forms; these devices must be connected to a network layer, which allows the aggregation and basic control of data; above these layers (or, as commonly said, above the ‘edge’) is a first layer of intelligence (Edge IT), which provides analytics functions and pre-processing of data; and the cloud, in which data are stored, analysed and processed for ultimate action and decisionmaking, mostly through AI. The IoT revolution will lead to an expansion of the possibility to automate complex processes, but will also require that both data and AI are kept as close as possible to the ‘things’: for example, autonomous vehicles cannot rely on a basic ‘sensor-to-cloud-and-back’ model of thinking, since the fact that data have to travel long distances would generate latency (for every 100 miles, an estimated 0.82 milliseconds). Looking at different network topologies, the immediate alternative to centralised IoT systems would be the implementation of intelligent solutions closer to things, and in particular ‘at the edge’. While a fully decentralised system would entail ‘embedded AI’ in each of the connected objects, and would therefore be too costly using current technologies, most market analysts consider the so-called Edge/Cloud model to be the most interesting paradigm for the most sophisticated IoT use cases in the near future.23 In an edge/cloud model, local computing, storage and networking resources are provided close to IoT devices, and the data generated can be stored and pre-processed by the local edge cloud and only a small volume of processed data are eventually sent to central data centres.24 As already explained, the conventional cloud models will remain viable for a number of uses. However, several emerging applications would strongly require an edge/cloud architecture: such a solution can offer important cost savings on top of a more distributed structure.25 Performing computations at the network edge has several advantages: (i) the volume of data needed to be transferred to a central

23 A related term, ‘fog computing’, describes an architecture where the ‘cloud is extended’ to be closer to the IoT end-devices, thereby improving latency and security by performing computations near the network edge. So fog and edge computing are related, the main difference being about where the data is processed: in edge computing, data is processed directly on the devices to which the sensors are attached (or on gateway devices physically very close to the sensors); in fog computing, data is processed further away from the edge, on devices connected using a LAN. See A Renda and M Laurer, ‘IoT4SDGs. What can the Digital Transformation and IoT achieve for Agenda 2030?’ (2020) Joint CEPS-Hitachi Report. 24 Several attempts have been made at operationalising an edge/cloud model in the past few years. They include projects such as Cloudlet, Nebula, Femtocloud, HomeCloud and Fog Computing. Each of these alternatives has pros and cons, and as occurs for transmission protocols, different solutions may fit different use cases. For example, the implementation of Network Functions Virtualisation and Software-Defined Networking can provide numerous advantages to the dynamic management of edge/ cloud systems, especially in the context of 5G deployment. 25 In 2015, for example, David Floyer studied the data management and processing costs of a remote wind farm using a cloud-only system versus a combined edge/cloud system. The wind farm consisted of several data producing sensors and devices such as video surveillance cameras, security sensors, access sensors for all employees and sensors on wind-turbines. The edge/cloud system turned out to be 36% less expensive and the volume of data required to be transferred was observed to be 96% less, compared with the cloud-only system. D Floyer, ‘The Vital Role of Edge Computing in the Internet of Things’ (2015), available at: wikibon.com/the-vital-role-of-edge-computing-in-the-internet-of-things.

198  Andrea Renda computing location is reduced because some of it is processed by edge devices; (ii) the physical proximity of edge devices to the data sources makes it possible to achieve lower latency which improves real-time data processing performance; (iii) for the case of data that still must be processed remotely, edge devices can be used to discard personally identifiable information (PII) prior to data transfer, thus enhancing user privacy and security; (iv) decentralisation can make systems more robust by providing transient services during a network failure or cyberattack; and (v) edge computing increases scalability by expanding computing capacity through a combination of edge and IoT devices. In non-technical terms, it is possible to state that while the AI revolution is the ‘brain’ of the digital ecosystem, and the cloud governs its central nervous system, the IoT will represent the limbs and muscles, and the edge will act as the peripheral nervous system, requiring a degree of automation and fast thinking at the local level. This will require a number of important changes in the way the digital economy functions and is organised: small data become more important than big data; data storage occurs mostly at the edge and in devices; and more distributed architectures are possible, with possible consequences also for competition. While the edge/cloud architecture will become prominent in the next few years, the Business-to-Consumer (B2C) world will also see a possible acceleration, in particular through the emergence of more distributed and decentralised architectures, enabling the possibility to store data locally, and in a more privacypreserving way. The emergence of new paradigms for restoring user control over personal data, from the IHAN project proposed by the Finnish innovation agency SITRA to the approach proposed by the MyData movement and by Tim Berners Lee’s Solid project, there is a wealth of ideas for enabling the creation of ‘data trusts’ and other intermediaries, able to help end users easily manage their data by selectively sharing the information they need to share, without losing control of their diffusion in obscure secondary markets. Most importantly, the new wave of digital transformation promises to achieve very substantial progress in the domain of public services, and more generally in the role of government. Governments that manage to collect good quality data will be able to develop APIs and share them with small and large corporations, researchers and other organisations for the development and provision of valueadded services. The new age of ‘Government as a platform’, which saw Estonia as a pioneer with its X-Road ecosystem (now also available in Finland), promises to revolutionise the relationship between public authorities and citizens, and places governments in the driving seat when it comes to securing trust in technology. The upcoming information-rich age will require a strong layer of trusted intermediaries, in charge of verifying the trustworthiness of data flows: suffice it to consider the fast development of ‘deepfakes’, which contribute to an already rich repertoire of disinformation tools by making it almost impossible to distinguish between real and fake audiovisual content. All these developments will determine the final departure of the internet from its original design, and also away from its current architecture. As occurs in the

Single Market 2.0: The EU as a Platform  199 evolution of complex organisms in biology, here too the internet will have to accommodate increased complexity due to the coexistence of very different uses, including low-latency industry services enabling control through digital replicas (or ‘twins’), immersive holographic presence and so-called ‘multi-sense media’, alongside more traditional data flows. Inevitably, the giant technology companies of today may have an advantage in conquering those spaces, which explains to a large extent why they continue investing huge sums in R&D. At the same time, the competitive space is open to new players, and even more to investment by private and public institutions that, from the ‘real economy’, seek the achievement of a more balanced and sustainable internet architecture. This is where the European Union may have an advantage over other superpowers, and even vis-à-vis the current tech giants.

C.  Wrapping Up: An Evolving Ecosystem The past few years have been characterised by the rise of a new wave of technological developments, which promise to revolutionise the digital economy, bringing it towards an era dominated by dramatically superior computing power and connectivity speeds; a skyrocketing number of cyber-physical objects connected to the internet (the so-called IoT, powered by nano-technology and by 5G wireless broadband connectivity); and the pervasive spread of AI into almost all aspects of personal and professional life. This new stack will be composed of powerful hardware, including faster processors (mostly a combination of CPUs, GPUs and TPUs); distributed computing capacity through edge (or fog) computing; new, distributed and decentralised platforms such as blockchain, able to keep audit trails of transactions and other asset-backed values; and a pervasive presence of AI-enabled solutions, mostly in the form of data-hungry techniques such as smart analytics, deep learning and reinforcement learning.26 Focusing on all layers of this emerging stack is extremely important when it comes to scaling up these technologies to the benefit of society: merely focusing on one element, such as AI or blockchain, would not harness the full potential of this emerging world. Figure 1 above portrays the evolution of the technology stack. The Internet of Things (IoT) layer generates an unprecedented amount of data, requiring sensor technology, nano-tech, enhanced connectivity through 5G or satellite, and devices like drones or robots, able to generate live data remotely.27 Regardless of the way in which data are generated, stored and exchanged, the use of AI will be ubiquitous in 26 See A Renda, ‘Artificial Intelligence: Ethics, Governance and Policy Challenges’ (2019) CEPS Monograph. 27 Data can be stored in various ways, including through remotely accessible, cloud-enabled solutions; through distributed databases; or through distributed ledger technologies such as blockchain. Some of these technologies are key enablers of value chain integrity, monitoring and trust, since they produce ‘audit trails’ that enhance the verifiability of transactions and contractual performance across the value chain.

200  Andrea Renda Figure 1  The old versus new digital technology stack Users

Content layer

Content/Data

Application layer

Apps/Services

Logical layer

AI

Physical (transport) layer

Fixed

Mobile

Open Internet/platforms

Blockchain DApps Blockchain Protocols

Logical layer/Internet Protocols Other

Infrastructure-Connectivity Internet of Things

Source: Author.

most supply chains. At the top of the supply chain, end users very often constitute the ‘weakest’ link, which requires the provision of adequate skills in using digital technologies. Although no real estimate of the combined impact of these technologies on the future economy exists, several studies have already been published on the economic impact of AI, as well as on the impact of IoT in specific sectors. For example, recent reports by Accenture/Frontier Economics, McKinsey and PWC conclude that AI will be a game changer for total factor productivity and growth, by gradually rising as a third pillar of production, together with labour and capital. PWC concluded that by 2030, global GDP will be 14 per cent higher due to AI development and diffusion;28 the Accenture study finds that growth rates will be doubled by 2035 thanks to AI.29 The latter study also shows an industry-by-industry breakdown, which includes agriculture, forestry and fisheries: this sector is expected to more than double its growth rate by 2030, from 1.3 per cent to 3.4 per cent on a yearly basis thanks to AI. Distributed ledger technologies are expected to complement these developments by solving several market failures along supply chains, as well as empowering end users in their consumption choices; some commentators go beyond these expectations and foresee a revolutionary impact of blockchain in many sectors, but this chiefly depends on whether more decentralised architectures will prove scalable over time.

II.  The Single Market 2.0 as a Layered Ecosystem The previous section embarked on a short description of three decades of digital transformation, with a clear objective: to show that technological evolution often 28 PWC, ‘Sizing the Prize. What’s the real value of AI for your business and how can you capitalise?’ (2019) PWC Analysis, available at: www.pwc.com/gx/en/issues/analytics/assets/pwc-ai-analysis-sizingthe-prize-report.pdf. 29 M Purdy and P Dougherty, ‘Why Artificial Intelligence is the Future of Growth’ (2017) Accenture/Frontier Economics Report, available at: www.accenture.com/t20170927T080049Z__w__/ usen/_acnmedia/PDF-33/Accenture-Why-AI-is-the-Future-of-Growth.PDFla=en).

Single Market 2.0: The EU as a Platform  201 opens up new possibilities, which must be accounted for in the definition of public policies. In this context, the EU is faced with the herculean task of shifting gear towards a more assertive and enlightened policy for the Digital Single Market, in a way that preserves its values and at the same time boosts its competitiveness and sustainability over the coming years. This will require a thorough redefinition of the strategy for the single market, to preserve at once its role of the ‘workhorse’ of European integration, as well as driver of EU international actorness and competitiveness. The European Commission has acknowledged on many occasions that there can be no ‘geopolitical Commission’ without a strong single market at home: even strategies aimed at orienting digital technology towards human rights and sustainability require a significant market positioning to be credible in internal fora. As already mentioned, the ‘new’ European Commission made a strong commitment to the digital transition, positioning the digital environment as a critical infrastructure, and postulating the need for digital sovereignty and the twin transition (green and digital) as cornerstones of its new geopolitical mission. The adoption of a comprehensive, ambitious data strategy and the White Paper on Artificial Intelligence paved the way for new policy developments that aim, for the first time, at anticipating future market developments by achieving a first-mover advantage in the forthcoming second wave of digital transformation. As clearly stated by Commissioner Thierry Breton, the European Commission expects that the next years will mark a transition towards more distributed, localised data storage, at the edge and in devices as opposed to the cloud. While cloud operators will remain important, they will be part of a more complex architecture, as we described in section I.B. This represents at once a challenge and an opportunity: on the one hand, it is clear that Europe has missed the first train (the B2C wave), and this may deprive it of the resources, the skills and the industrial base needed to jump into the second; on the other hand, the B2B sector is not (yet) dominated by the large tech giants, and requires the orchestration of resources from legacy industrial sectors, such as manufacturing, automotive, pharma, energy, in which European corporations are often market leaders. Similarly, the importance of data flows in (and for) government requires action to avoid public institutions ending up relying on non-EU players for cloud and software solutions: digital sovereignty thus calls for new initiatives aimed at creating a full European technology stack. As explained by the European Commission, this assertive approach can and should be coupled with concrete commitments to use digital technology ‘for good’, in particular in support of the green transition: this requires, for example, that data centres become carbon neutral by 2030, and that the IT equipment used in Europe aligns with the requirements of the circular economy.30

30 See the European Commission’s Communication, Shaping Europe’s Digital Future, 19 February 2020, available at: ec.europa.eu/info/sites/info/files/communication-shaping-europes-digital-futurefeb2020_en_4.pdf.

202  Andrea Renda The COVID-19 pandemic created a double challenge to the original commitments of the von der Leyen Commission. On the one hand, the need for concrete actions to strengthen Europe’s digital economy has become even stronger, as most economic and social activities move online: an investment in the IoT and the Edge IT layers appears urgent to say the least, in the attempt to transform suffering economic sectors, facing an unprecedented nosedive in production and turnover. On the other hand, the upcoming depression may lead those sectors into dire straits, and thereby to the impossibility of embarking in such an ambitious transformation. Despite the crisis situation, at the time of writing EU institutions have confirmed their intention to proceed as planned on the twin transition: the Council conclusions of 26 March 2020 urged the Commission to ‘get back to a normal functioning of our societies and economies and to sustainable growth, integrating inter alia the green transition and the digital transformation, and drawing all lessons from the crisis’.31 The Commission’s response on the twin transition placed the single market at the forefront, advocating the acceleration towards a ‘more circular, climate neutral and modernised economy’. Current emphasis is however being placed on incremental measures such as reinforcing digital skills and making new tools and resources available to SMEs, on the reduction of red tape, the strengthening of enforcement in the traditional single market. Only time will tell whether Breton’s original ides of a more forward-looking approach to the Digital Single Market will survive the COVID-19 crisis. In order for Europe to realise its ambition to preserve and further nurture its actorness and competitiveness at the global level, it is important that the Commission realises that the ‘data train’ has not left the station, and that Europe has a unique window of opportunity to rely on the new digital transition to boost its weight in the global economy. This requires a number of initiatives, aimed at rebalancing (or restoring) competition in B2C digital markets; creating and promoting efficient data spaces in the B2B domain; paving the way for a distributed and reliable single market for services; enabling a network of interoperable administrations and leverage eIdentity as a means to empower citizens and consumers; and leading the world in the responsible development and use of digital technologies. Below, I briefly elaborate on each of those pillars.

A.  Rebalancing Competition and Value Allocation in the B2C Domain: From Competition Policy to the Federated Cloud Infrastructure It has become relatively uncontroversial that the ‘first wave’ of digital transformation has created positions of excessive power, transforming a fistful of digital 31 See Joint Statement of the Members of the European Council, 26 March 2020, available at: www. consilium.europa.eu/media/43076/26-vc-euco-statement-en.pdf.

Single Market 2.0: The EU as a Platform  203 players into gatekeepers of the internet, and leading to an excessive concentration of value generated by the digital economy.32 There are many ways in which the EU institutions can try to rebalance this situation. First, the Commission is likely to reform its competition rules to acknowledge the existence of specific situations in which digital players, regardless of market definition, occupy a position that grants them ‘intermediary’ power, mostly fuelled by network effects and data availability. This move, which would echo the recent debate on the German draft Digitalisation Bill, might lead to considering specific remedies, such as the imposition of mandatory interoperability for specific data sets, to enable competitors to deploy similar services. This is likely at least for specific data sets which can be considered to be essential to develop services of general interest, and would be a re-proposition of the essential facilities doctrine, which dates back more than two decades in EU competition law, to cases like Magill, IMS Health, and Microsoft.33 Second, in the B2C domain, EU institutions could decide to go beyond data access and interoperability obligations, and adopt policies aimed at returning control of their data to end users, or even treat data ‘as labour’ whenever possible, as advocated recently by the Report of the High Level Expert Group on the Impact of the Digital Transformation on EU Labour Markets.34 This would lead to forms of remuneration from digital platforms to end users, which may take various forms, including the provision of free services, or most likely a web tax, which seems even more likely in the aftermath of the COVID-19 pandemic.35

32 See M Mazzucato, The Value of Everything. Making and Taking in the Global Economy (Penguin Books, 2018). 33 See A Renda, ‘Competition–regulation Interface in Telecommunications. What’s left of the Essential Facilities Doctrine’ (2010) 34(1–2) Telecommunications Policy 23. Based on this approach, whenever a dominant market player holds an asset of information that is essential for competitors to viably compete in the relevant market, and refusal to provide access to this information is likely to either lead to the exit, or even prevent the growth of, ‘as efficient’ or even ‘not yet as efficient’ competitors, then competition law may provide for compulsory access remedies. Much in the same vein, the German government is now imposing compulsory access obligations to tech giants for specific data sets. In a recent paper for the European Commission’s DG COMP, Jacques Crémer, Yves-Alexandre de Montjoye and Heike Schweitzer echo this view by observing that ‘the ability to use data to develop new, innovative services and products is a competitive parameter whose relevance will continue to increase’; and clarified that ‘in a number of settings, data access will not be indispensable to compete, and public authorities should then refrain from intervention. In other settings, however, duties to ensure data access – and possibly “data interoperability” – may have to be imposed’. The paper correctly points out that a ‘broader diffusion of data is not always desirable, either from a social welfare or from a competition perspective’ due to privacy concerns; and that in addition to data interoperability, in some cases full protocol interoperability may be needed for competitors to be able to compete on an equal footing. 34 See the Final report of the High-Level Expert Group on the Impact of the Digital Transformation on EU Labour Markets, April 2019, available at: ec.europa.eu/digital-single-market/en/news/ final-report-high-level-expert-group-impact-digital-transformation-eu-labour-markets. 35 In that case, the tax would be based on the consideration that the digital platforms derive (some would say, extract) value from the end users, who provide data in exchange for being part of the platform: the main theoretical argument in favour of such a form of redistribution is the ‘collective action problem’ faced by end users, who are structurally unable to place a price on the data they provide, while these data, once aggregated, become extremely valuable to the platform. This form of positive externality could be seen as the market failure that a web tax, or any other form of redistribution, would seek to remedy.

204  Andrea Renda This approach, however, would not lead to the creation of more competition in the market, or possibly even the entry of European players in the B2C segment. Another area that will prove important to rebalance the bargaining position of market players and users in the ecosystem is the effective implementation of recent EU rules on the unfair distribution of the contractual surplus among parties in a commercial relationship. So-called ‘platform-to-business’ (P2B) practices have been subject to specific regulation in the European Union, where a specific observatory on online platforms has been created in order to monitor the application of the regulation. The EU P2B regulation introduces a ban on certain unfair practices (for example, no more sudden, unexplained account suspensions, plain and intelligible terms and advance notice for changes, greater transparency and mandatory disclosure for a range of business practices) and new dispute resolution possibilities.36 P2B rules echo similar legislation that many countries have enacted in the domains of abuse of economic dependency, abuse of superior bargaining power, contract law, or unfair competition. In Germany, the draft Digitalisation Bill echoes existing provisions on ‘relative market power’ vis-à-vis smaller enterprises, by removing any reference to the business size and adding that relative market power shall also be assumed for ‘undertakings acting as intermediaries on multisided markets insofar as undertakings are dependent on their intermediary services with regard to access to supply and sales markets in such a way that sufficient and reasonable alternatives do not exist’ (Section 20, paragraph 1 GWB-Draft).37 Apart from reforming existing rules, the Commission will also act to restore digital sovereignty in the ecosystem, in particular by seeking the creation of a federated cloud infrastructure, operating under rules and protocols that embed strict data protection and governance requirements. This will most likely take inspiration from the GAIA-X project, initiated by France and Germany and already including more than 120 partners. GAIA-X can be seen as the quintessential panEuropean approach to the future of the single market: rather than representing a single player competing with the US tech giants, GAIA-X is a federated data infrastructure, open to small and large companies, which attempts at once to level the playing field, and to embed in the cloud specifications the compatibility with key European provisions on security, data protection, openness and transparency, interoperability and trust. In particular, interoperability is sought at three different infrastructure levels: network, data and service, in a way that resembles the 36 In Australia, similar concerns were expressed during the ACCC Digital Platforms Inquiry. Among those practices, some may also lead the digital platform to favour its own products, or ‘preferred’ customers on the business side, possibly through algorithmic ranking and product placement choices. 37 Relative market power is to be assumed if undertakings depend on access to data to enter a market; in addition, hampering rivals’ attainment of positive network effects can constitute an abuse if this is capable of triggering the tipping of a market: this includes a case in which a platform adopts measures measures to disable data portability or interoperability along with exclusivity clauses or tying practices, to the extent that such measures create a ‘serious risk of a considerable restriction of competition on the merits’.

Single Market 2.0: The EU as a Platform  205 approach to interoperability adopted by the European Commission to enhance exchanges between EU and national administrations (see below). While still in its infancy, the GAIA-X project represents a gateway towards a next generation approach to technology-enabled policymaking at the EU level. It is, in particular, a clear response to the emerging edge/cloud infrastructure (as acknowledged in the GAIA-X White Paper);38 and a sign of the increased awareness, in the European Commission, of the revolutionary potential of the second wave of digital transformation: centralised governance does not mean market concentration, especially if a common environment for cloud services is coupled with clear data interoperability and portability rules. In this respect, the emphasis on common standards and a thicker (three-level) interoperability framework embeds the key features of the Single Market 2.0 approach, in conjunction with a long-standing approach of the European Commission.

B.  Securing the B2B Domain: The End of Open Data Based on our description of the EU strategy above, it is inevitable that the strongest effort of the new Commission will be concentrated in the B2B domain, where the emergence of the IoT and the edge/cloud infrastructure powered by 5G and other forms of connectivity call for a new approach to industrial policy. This seems to lead to a combination of data strategy and industrial policy, both adopted by the von der Leyen Commission under the auspices of Commissioner Breton in the first 100 days of the Commission’s mandate. The data strategy announces the objective to create a single European data space and couple it with measures aimed at ensuring that by 2030, the EU’s share of the data economy corresponds to its economic weight (‘not by fiat but by choice’, the Commission adds). The idea of creating a ‘genuine single market for data’ leads to an upgrade of the ‘free flow of non-personal data’ approach that emerged during the Juncker Commission. Even if the Commission is very cautious not to venture into too assertive statements, it emerges clearly that in the B2B domain, the age of ‘open data’, free-flowing information as a means to the promotion of innovation is definitely over. The need to avoid capture of industrial data by large tech giants, and imbalances in the distribution of revenues along the value chain, leads the Commission to propose the creation of a two-layer architecture, with an overarching single European data space and a number of domain-specific or mission-specific data spaces. The stated reasons for this move are the fragmentation between national data policies, and the persistence of significant constraints on all types of data flows; the existence of imbalances of market power; problems of data quality and interoperability; lack of adequate provisions for data governance; 38 See Project GAIA-X. ‘A Federated Data Infrastructure as the Cradle of a Vibrant European Ecosystem’, available at: www.bmwi.de/Redaktion/EN/Publikationen/Digitale-Welt/project-gaia-x. pdf?__blob=publicationFile&v=4.

206  Andrea Renda and a collection of other problems on both the supply and the demand side of data, including security aspects, regulatory certainty and skills. The result is the proposed creation of a series of large pools of data in specific domains, combined with the technical tools and infrastructures necessary to use and exchange data, as well as appropriate governance mechanisms. These pools (renamed ‘data spaces’) require the adoption of a horizontal framework complemented by sectoral legislation for data access and use, and mechanisms for ensuring interoperability, and must be developed in full compliance with data protection rules and according to the highest available cybersecurity standards. Such a framework will be adopted by the end of 2020, and will need to be complemented by policies that stimulate the use of data and demand for services enriched with data.39 Apart from the governance aspects of data space management, which are still unknown, it is clear that data spaces are a key component of the Commission’s new vision for data-driven industrial policy, and aim at realising at once a rebalancing effect (keep entitlements over data in the hands of industrial players) and a repatriation effect (ensure that data are stored and managed according to European rules, and preferably in the territory of the EU). The data spaces proposed by the Commission in its data strategy, as already mentioned, are in some cases cross-sectoral, in others more sector-specific. Among the cross-sectoral ones are a ‘Green Deal data space’, which is expected to mobilise public and private data to help achieve Europe’s environmental goals, even by creating a digital twin of the Earth; a Common European skills data space, aimed at reducing skills mismatches in the labour market; and European data spaces for public administration, aimed at strengthening data exchanges, promoting transparency and accountability, fighting corruption, and enabling GovTech solutions. More sectoral solutions are devoted to manufacturing, mobility, health, finance, energy and agriculture. The data spaces approach must be analysed in conjunction with the Industrial Strategy communications, adopted by the Commission in March 2020.40 The strategy, though mostly focused on the real economy and on related topics such as innovation and entrepreneurship as well as industry alliances and the ‘analogue’ single market, places strong emphasis on the concept of industrial ecosystems, which are not clearly defined in the Communication. One of the most crucial aspects of the new ecosystems approach is whether they will be defined as coinciding with industry sectors (for example, aviation), or in a mission-oriented way (for example, mobility). Data spaces have been defined according to the latter approach, but the COVID-19 crisis may call for a more sector-specific approach to industry support and aid. The new approach to the Single Market 2.0 would 39 See the Commission’s Work Programme 2020, available at: ec.europa.eu/info/publications/2020commission-work-programme-key-documents_en. 40 See Communication from the Commission to the European Parliament, the European Council, the Council, the European Economic and Social Committee and the Committee of the Regions, ‘A New Industrial Strategy for Europe’ COM(2020)102 final, 10 March 2020.

Single Market 2.0: The EU as a Platform  207 preferably preserve the mission-oriented approach, as well as its vocation towards sustainable development in a technology-neutral way.

C.  A Distributed Single Market for Services Above the infrastructure and data governance layers, the Single Market 2.0 will enable different forms of governance, from centralised to more distributed and decentralised. One interesting characteristic of the digital economy, in this respect, is that the traditional barriers to entry in the form of economies of scale are less likely to materialise, especially in the service economy. The servitisation of the economy describe in section I above makes it possible for very small businesses to use powerful IT equipment, rent drones and even tractors or other costly machinery ‘as a service’. The rise of 3D printing decentralises the production and assembling of products, drastically reducing costs. For example, in agriculture communities of farmers can then access modern technology at a fraction of the cost, and only for the time needed.41 However, as already recalled these barriers to entry are being replaced by less tangible ones, notably represented by data access and management, as well as related skills. The new technological frontier, coupled with interoperability rules, will make further servitisation possible, alongside the adoption of more distributed governance structures, thereby enabling greater market competition. Industrial economics will fundamentally change, unveiling the possibility to address issues in a centralised, distributed or decentralised way, as depicted in Figure 2 below. In some cases, a centralised architecture (for example, exclusively cloud-based) may emerge, thereby triggering network externalities and winner-take-all competition; in other situations, a decentralised structure may lead to edge/cloud architectures and federated structures organised along a limited number of ‘supernodes’; and finally, with technological development even fully distributed structures, with service provision among peers, become increasingly possible. In more concrete terms, this may mean that in some markets, a large number of small corporations may end up competing for services, by using data made available through widespread interoperability requirements. This comes very close to the IHAN model proposed by SITRA, which de facto re-proposes the PSD2 ‘open API’ approach on a wider scale, commoditising data availability (hopefully, implementing privacy-preserving arrangements) and laying the foundations for a more competitive, pluralistic single market for services. Maximising data availability will also mean liberating all possible sources of data flow, including the use of public sector information by business (G2B); sharing and use of privately held data by other companies (B2B); the use of privately held data by government authorities when appropriate and desirable (B2G); and data-sharing between public

41 See

Renda et al, above n 20.

208  Andrea Renda Figure 2  Centralised, decentralised and distributed computing

Link Station Centralised (A)

Decentralised (B)

Distributed (C)

Source: Truong et al (2016).42

authorities (G2G). In this respect, all other pillars of the proposed new framework for the Single Market 2.0 are essential to feed the new services market. For example, public administrations (as defined in more detail below) could enable innovation by acting as platforms and offering open APIs to citizens and businesses, thereby significantly lowering the data barrier to entry. Widespread, privacy-compatible data availability throughout the Union can also contribute to the environment by enabling more localised solutions. In a nutshell, the Single Market 2.0 would be more integrated (through data flows), competitive (through data interoperability), decentralised (through lower data barriers to entry) and environmentally sustainable (through lower transport costs, as well as carbon-neutral data centres). Such a vision requires, inevitably, the support of modernised legal rules. In particular, the scope of most product liability regimes does not include intangible goods, implying that cases of inadequate services, careless advice, erroneous diagnostics and flawed information are as such not covered. A comparable situation exists in the field of product safety regulation, which so far has not been accompanied by a regulatory framework in the field of safety of services. In all these fields, the EU acquis appears far from complete, and will require more attention in the years to come. The upcoming Digital Service Act should fill this gap by clarifying the conditions for the liability of online service providers and intermediaries. The needed update of the Product Liability Directive will most certainly entail a revision of key definitions such as ‘product’ and ‘producer’, as well as rules on certain practices adopted by service providers vis-à-vis end users, such as price personalisation through automated decision-making and profiling. In other words, the

42 N Truong, U Jayasinghe, T-W Um and M Gyu, ‘A Survey on Trust Computation in the Internet of Things’ (2016) 33 The Journal of Korean Institute of Communications and Information Sciences (J-KICS) 10.

Single Market 2.0: The EU as a Platform  209 body of rules that supports the new vision of the single market should be able to generate sufficient trust among end users, as will be explained in more detail in section II.E below.

D.  Digital Government and eIdentity Data availability and user-centric innovation in the future Single Market 2.0 would greatly benefit from a proactive role of government in the generation, collection, protection and provision of data. This can occur through ad hoc data trusts, or simply by public administrations acting as orchestrators of the data economy. The European Commission has gone a long way in creating the preconditions for the interoperability between administrations in Europe, in particular through its ISA2 programme, which led to the development of reusable building blocks, available to national and local governments on a purely voluntary basis. However, that framework has led to a very low uptake, and would need to be converted into a much more concerted action for the development of digital government and ‘government as a platform’ approaches in Europe: this includes measures included in the semester, in the InvestEU and Digital Europe programmes, and dedicated support through the newly created DG REFORM. The swift transition towards digital government solutions is even more urgent since the Ministerial Declaration on e-Government in Tallinn on 6 October 2017, in which the ministers in charge of e-Government policy from 32 countries of the European Union and the European Free Trade Area (EFTA) unanimously committed to the vision laid out in the EU e-Government Action Plan 2016–2020 and in the new European Interoperability Framework that public administrations and public institutions in the EU should be open, efficient and inclusive, providing borderless, interoperable, personalised, user-friendly, end-to-end digital public services to all citizens and businesses – at all levels of public administration.43 This includes, inter alia, the development of more efficient and user-centric digital services; a call on the EU institutions to develop more interoperable, efficient, open and transparent administrative procedures to best serve their citizens and interoperate with all levels of government.44 The ISA2 programme will leave, as an important legacy, a layered approach aimed at building the foundations of the Single Market 2.0 by ensuring a high level of legal, organisational, semantic and technical interoperability between administrations. EU institutions should take the time line of the Tallinn declaration seriously: they agreed, at the end of 2017, to achieve six targets within five years: digital by default and inclusiveness; the application of the ‘once only principle’; to secure trusted electronic identification and trust services for electronic transactions in the internal market; to enable the possibility for people and businesses to access 43 See: ec.europa.eu/digital-single-market/en/news/ministerial-declaration-egovernment-tallinndeclaration. 44 ibid.

210  Andrea Renda personal data held by the public administrations; to integrate instruments and a call to public authorities to oblige cross-border interoperable solutions compatible with European frameworks and standards; and to improve the digital leadership skills and IT education at every level of the public administrations.45 Among these commitments, particularly important for the development of the whole single European data space, will be the area of digital verification, encompassing both digital identify (eID) and electronic trust services (eTS), altogether subsumed under an EU framework for authentication in digital transactions (eIDAS). eIDAS sets the standards and criteria for simple electronic signature, advanced electronic signature, qualified electronic signature, qualified certificates and online trust services. Furthermore, it rules electronic transactions and their management. Among other benefits, it fully recognises digital means of verification that are considered to be equivalent to physical presence. In doing so, it lays the foundations for the creation of the Single European Data space. The system could in the future be updated to include other means of verification (for example, fingerprint scan) and create a new system for certification of digital ID in the Single Market 2.0. It should also be complemented by trustless (or better, trustenhancing) mechanism for time-stamping, origin-stamping and other transaction authentication methods offered by Distributed Ledger Technologies (including fully privacy-preserving ways such as zero knowledge proofs).

E.  Leading the World on Digital Technology ‘for Good’ As already remarked in the previous section, a key pillar of the emerging Single Market 2.0 will necessarily have to be the trustworthiness of its infrastructure, protocols, rules and services. In this respect, the EU has the responsibility and the opportunity to lead the world in the development of a trusted digital environment and has already started to do so in the domain of AI. In the AI field, the European Commission (backed by an ad hoc High Level Expert Group (AI HLEG), advocated the transition towards ‘Trustworthy Artificial Intelligence’, defined as AI that meets three cumulative requirements: legal compliance, ethical alignment and socio-technical robustness. The AI HLEG identified four key principles (defined as ethical ‘imperatives’) for Trustworthy AI: the respect for human autonomy, the prevention of harm, fairness, and explicability.46 These principles were further broken down into seven requirements, which were then operationalised into an Assessment List on Trustworthy AI (AL TAI). This strategy arrived

45 See: joinup.ec.europa.eu/collection/nifo-national-interoperability-framework-observatory/news/ tallinn-egov-declaration. 46 Importantly, contrary to what typically occurs in similar documents, the list did not include an imperative to ‘do good’, or the so-called ‘beneficence’ principle, which had been included in earlier drafts of the Guidelines. See A Jobin, M Ienca and E Vayena, ‘The global landscape of AI ethics guidelines. Nature Machine Intelligence’ (2019) (ISSN 2522-5839).

Single Market 2.0: The EU as a Platform  211 at a turning point with the White Paper adopted on 19 February 2020 (together with the new EU strategy for data), in which the von der Leyen Commission announced the adoption of a flexible, agile regulatory framework limited to ‘highrisk’ applications, in sectors such as healthcare, transport, police and the judiciary, and focusing on provisions related to data quality and traceability, transparency and human oversight. A legislative initiative on AI is now expected by the end of 2020, as outlined in the Work Programme of the European Commission, which envisages a follow-up to the White Paper, including on safety, liability, fundamental rights and data. The EU agenda on AI inspired many other countries and international organisations, including the OECD principles on AI, the G20 human-centred AI Principles, as well as the ‘AI for good’ within the International Telecommunications Union. This revived the EU’s actorness in the digital technology space, where it is now a recognised standard-setter, and perhaps the only superpower able to credibly orchestrate a dialogue on responsible uses of digital technologies (starting with AI). In order to sustain high standards in this domain, as already recalled, the EU will need to rely on a vibrant single market, to be leveraged through extraterritorial rules to avoid EU products being outcompeted by non-European, less sustainable standards.

III.  Conclusion: Towards a New Architecture for the Single Market 2.0 The five pillars of the Single Market 2.0 illustrated in section II are certainly a nonexhaustive account of the possible future of Europe’s ‘crown jewel’. Compared with the architecture of the internet shown in Figure 1 above, the new Single Market would look more articulate, but not necessarily more complex from an end user perspective. As shown in Figure 3, the Single Market 2.0 would have an extensive infrastructure layer, composed of connected things, the edge IT layer, various connectivity protocols (including 5G and many others) and the federated cloud infrastructure. Above that layer, the traditional logical layer would be flanked by a legal, identity/trust and semantic interoperability layer, which will then support the emerging cross-cutting and sector-specific data spaces. These will cover most of the Single Market 2.0 from a B2B perspective, whereas the B2C will be more similar to the current internet ecosystem. An important role could be played by G2C and G2B services with public administration becoming real catalysts of social innovation.47 This, inevitably, is only a sketch of the main components of the future Single Market 2.0, which will certainly be even more complex. The role of technologies



47 See

also AI HLEG, Policy and Investment Recommendations, June 2019.

212  Andrea Renda Figure 3  A sketched architecture of Single Market 2.0 End users

Sectoral data spaces (B2B) Skills

Open Internet + DLTs

Digital platforms (B2C)

Green Deal

Digital government (G2C, G2B)

Administration

Personal Data Control (e.g. IHAN, MyData)

Cross-cutting data-spaces Artificial intelligence

Legal and semantic interoperability Digital Verification layer

Federated Cloud (e.g. GAIA-X) Connectivity (4G/5G, Zigbee, LoRa, SigFox, etc.) EdgeIT Internet of Things

Source: Author.

such as blockchain and other Distributed Ledger Technologies (DLTs) in enabling transactions and fostering the so-called token economy is one of many aspects that will emerge over the next few years, and which are difficult to fully anticipate at the time of writing. The digital environment, as I have tried to explain throughout these pages, is an ever-changing multilayered ecosystem, which becomes thicker as time goes by, and where the technological evolution defines the frontier of what is possible in a constantly evolving way. The EU finds itself in a relatively favourable position at least in some of the layers portrayed in Figure 3. This is the case for 5G connectivity, an area in which EU companies like Nokia and Ericsson are rivalled (and complemented) by Chinese and South Korean players, and are way ahead of their North American counterparts in the development of a federated cloud, thanks to the initiative of Member States; in the availability of a pan-European framework for digital verification, and soon also in the development of a governance framework for data spaces. The EU is also leading the world in responsible AI, in trusted blockchain applications and in a number of industrial B2B applications, from manufacturing to healthcare. Against this background, the looming economic recession, the oscillating trust and commitment of Member States and the lack of agility of EU institutions may stand in the way of a rapid shift towards a more vibrant, futureproof single market. Absent political commitment and a good dose of enlightened policymaking, the European Union risks also missing the second wave of digital transformation: new waves will certainly come, yet catching up may prove increasingly challenging for the ones that lag two (or more) generations behind.

10 The Internal Market and the Online Platform Economy VASSILIS HATZOPOULOS

I. Introduction The platform economy has been termed as the absolute social, economic and legal disruption. Based on tripartite relationships, whereby an electronic platform matches consumers’ demand with specific services (or goods) offered by its affiliated ‘prosumers’ (producing consumers) or professionals, the platform economy disrupts traditional legal categories. Three-party contracts for distinctbut-interdependent services sit uneasily between traditional contract and tort law, while pushing for fresh legal categories under employment law. The fact that prosumers are non-professionals renders consumer protection law inapplicable, while the power hidden in algorithms and the possibilities they offer for market parallelism may be calling for a revision of basic competition law principles and instruments. At the same time, the manipulation of big data, the ensuing ‘profiling’ of users and the commercial exploitation thereof make the new EU General Data Protection Regulation (GDPR) already look outdated. What exactly is covered under the term ‘platform economy’ is unclear. A search in the web suggests that platforms are active in the following areas:1 (a) online marketplaces, such as ebay and Amazon; (b) App stores, such as Google Play Store; (c) collaborative economy platforms, such as Uber and Airbnb; (d) video-sharing platforms, such as Youtube and Dailymotion; (e) payment systems such as Paypal and ApplePay; (f) internet search engines, such as Google and Bing; (g) specialised search engines such as Tripadvisor and Yelp; (i) location-based business directories, such as Google Maps; and (h) news aggregators, such as Google News. The European Commission, however, in its 2016 Communication on The Platform Economy only considered categories (a), (b) and (f), above, as well as social media (such as Facebook) and online advertising services.2 Then, again, the more 1 See, eg: www.thegeniusworks.com/2018/01/the-platform-revolution-how-network-based-marketsare-transforming-the-rules-of-business/platform-categories-and-companies-900x501/. 2 COM(2016) 288.

214  Vassilis Hatzopoulos recent draft Regulation on fairness and transparency for business users of online platforms,3 only applies to categories (a) and (b) and to social media. In view of the uncertain – and tentatively extremely wide – scope of the ‘platform economy’ and in order to be able to develop the legal issues raised within a specific context, what follows will focus on those platforms which produce the most disruptive effects for the EU internal market: collaborative platforms.4 The definitions of the collaborative economy vary considerably: some insist on platform intermediation, others on the facilitation of peer interactions, and others on the lack of ownership transfer. Taking into account the definition given by the Commission in its 2016 European Agenda for the Collaborative Economy,5 the term ‘collaborative economy’ includes platforms which facilitate (a) access as opposed to transfer of ownership and (b) the conclusion of a transaction (a contract) between two other parties (ie, a tripartite relationship), and where (c) the parties are primarily – but not exclusively – peers. Based on the above definition, the following characteristics can be identified, such as ‘disrupting’ the traditional economy. First, these are two-sided markets which produce network effects or externalities, both at the user level (the more users in a platform, the more useful the platform for each user) and at the metadata level (the more data there is in the algorithm, the more efficient it becomes). This allows platforms, to develop super-dominant positions, having the characteristics of superstar economics. Second, matching is ensured by the use of algorithms producing search rankings and recommendations tailored to user preferences. Third, in order to produce maximum results and better matching, improve user experience, and reduce transaction costs, platforms proceed to the profiling of each individual user through data collected, purchased or acquired. Therefore, data – both personal and mass (big data) – are essential for the platforms. Fourth, platforms make possible transactions between peers. The introduction into the economic circuit of ‘prosumers’ and the creation of new markets that depend exclusively on the platform, is a real legal disruption. Finally, fifth, the booming economy of the platforms would not have been possible in the absence of selfregulatory and reputation-rating systems. It is by these means that collaborative platforms convince perfect strangers to ‘trust’ and share their cars, rooms, sofas and beds. This peer review system has been particularly useful to the collaborative economy, reducing the inherent risks of dealing with strangers.6 All the above disruptions reflect on the EU internal market and affect the way it operates. So much so, that the need for fresh EU-wide regulation should to be discussed, before emitting some concluding thoughts. 3 COM(2018) 238. 4 For a first discussion of the issues raised by the collaborative economy under EU Law, see V Hatzopoulos and S Roma, ‘Caring for Sharing? The Collaborative Economy under EU law’ (2017) 54 Common Market Law Review 81; and more extensively see V Hatzopoulos, The Collaborative Economy and EU Law (Oxford, Hart Publishing, 2018). 5 COM(2016) 356 final. 6 On this issue see below, under III.A.ii.

The Internal Market and Online Platform Economy  215

II.  The EU Internal Market Disrupted by Platforms A.  Who Does What: Legal Qualifications of the Parties The first legal difficulty raised in the context of the collaborative economy is the qualification of the relationship between the parties involved. This qualification is essential not only in order to know against whom and on what basis (contractual or tortuous) will an injured party be able seek redress, but also to know whether labour law is applicable and, further, to know whether the platform needs an authorisation related to the underlying service. There are at least four (or five) different scenarios that depict the relationship described above: a.

The platform, while creating new offers/demand through its intermediation, offers only electronic (intermediation) services to the parties without directly intervening in the underlying service contract concluded by them. This is the approach constantly advocated by the platforms. b. The platform controls the definition and content of services or providers of such services (through the conclusion of a contract of employment or else), to such an extent that it must also qualify as the provider of the underlying service (approach followed by the Court of Justice of the European Union (CJEU) in Uber Spain, for which see below, under II.B ‘Market Access’). c. The platform does not create new offer/demand. Rather, it is (yet) another means of communication for the services already offered by established providers to the general public (such as, for example, Booking.com for the Hilton chain of hotels). d. The platform and the service providers are so closely linked that they form a single contractual entity, which in tort qualifies as a ‘joint enterprise’. e. The last scenario is ‘artificial’ in the sense that it requires the legislator (European or national), adopting special rules governing, inter alia, the rights and obligations of the three (or more) parties involved. This approach has been already followed by the European legislator in the case of the Package Travel Directive, another two-sided market.7 In its 2016 Communication on the Collaborative Economy,8 the Commission retains a view very favourable for platforms, since it proposes three cumulative criteria that could show that the collaborative platform also provides the underlying service, namely that the platform imposes (and not ‘proposes’) a price, decides certain key points of the contractual relationship and owns the assets to provide 7 European Parliament and Council Directive (EU) 2015/2302 of 25 November 2015 on package travel and linked travel arrangements, amending Regulation (EC) No 2006/2004 and Directive 2011/83/EU of the European Parliament and of the Council and repealing Council Directive 90/314/ EEC (Package Travel Directive) [2015] OJ L326/1. 8 COM(2016) 356 final.

216  Vassilis Hatzopoulos the services. However, the possession of the assets by the platform contradicts two concepts forming an integral part of the definition of collaborative economy, ie, the sharing of goods and services and the two-sided nature of the platforms. The approach of the Commission criticised by the doctrine, is contradicted by the Court in its Uber cases, where the Court retained that the platform (a) creates a new market until then nonexistent, (b) over which it exercises ‘considerable influence’ and accepted that the platform is part of a ‘global’ service composed of both an electronic intermediation part and a part constituted by the underlying activity.

B.  Market Access Strictly related to the legal qualification of the relations between the three (or more) parties participating in the collaborative economy (above), is the question of market access for the different parties. Unfettered market access for economic operators is the quintessential requirement stemming from the four internal market freedoms.9 As the Court has recently held, in relation to services: [T]hat requires, above all, the elimination of obstacles which are encountered by providers in becoming established in the Member States, whether in their own Member State or in another Member State, and which are liable to affect adversely their ability to supply services to recipients located throughout the European Union.10

However, the conditions under which platforms and, to a lesser degree prosumers, have access to the market of one Member State, then to that of all the others, are far from clear. This question arises in different terms for platforms and prosumers, depending on whether they are established in a Member State or not; most interesting is the case of platforms established in the EU. This question was raised, inter alia, in the two Uber cases decided by the CJEU.11 The Court had to rule whether Uber is just an electronic intermediation service (therefore being subject to the extremely favourable conditions of the E-Commerce Directive),12 or whether, on the contrary, it is a transportation service (in need of the relevant authorisations and permits). This is a key issue for all platforms as the E-Commerce Directive could offer privileged market access: platforms that can avail themselves of this Directive do not require prior authorisation in their Member State of origin (Article 4(1)) and automatically benefit from the internal market clause (Article 3(2)) whereby other (host) Member States cannot 9 See on this issue, among many, AL Sibony, ‘Can market access be taken seriously?’ (2012) 2 Revue Européenne du droit de la consommation 323. 10 Joined Cases C-360/15 and C-31/16 College van Burgemeester v X and Visser EU:C:2018:44, para 105. 11 Case C-434/15 Elite Taxi v Uber Systems Spain EU:C:2017:981; Case C-320/16 Uber France EU:C:2018:221; for a discussion of these cases see V Hatzopoulos, ‘After Uber Spain: The EU’s Approach on the Sharing Economy in Need of Review?’ (2019) 44 European Law Review 89. 12 Directive 2000/31/EC of the European Parliament and of the Council of 8 June 2000 on certain legal aspects of information society services, in particular electronic commerce, in the Internal Market [2000] OJ L178/1.

The Internal Market and Online Platform Economy  217 systematically impose any obstacle to the activity of platforms in their territories.13 The E-Commerce Directive is a lex specialis and applies only to information society services, ie, ‘any service normally provided for remuneration, remotely by electronic means and at the individual request of a recipient of services’. Looking back at the different scenarios discussed above, platforms could claim market access under the E-Commerce Directive under scenario (a) and scenario (c), ie, in cases where they act only as electronic intermediaries and are not involved in the provision of the underlying services. As soon as the platform becomes directly involved in the provision of such services, it ceases to be a digital service provider and should be described as an ‘offline’ service provider. As a result, the platform would lose the privileged status offered by the E-Commerce Directive. ‘Plan b’ for the platforms would be to claim their access to the market according to the rules of the Services Directive. They will thus be able to benefit from a liberal and transparent framework for obtaining required authorisations (Articles 9–15) and from rules facilitating the transnational provision of services (Article 16). This presupposes, of course, that the activity in question does not fall within one of the many exclusions and exceptions of the Directive, interpreted erratically by the Court of Justice. In this regard, one cannot help but note the Grupo Itevelesa judgment, cited by the Court in Uber Spain: in this earlier judgment the Court concluded that the technical inspection activity of vehicles, even if not a transport service per se, must be excluded from the scope of the Directive as it constitutes a ‘service in the transport sector’. If the reasoning of this judgment is to be pursued, any ancillary activity offered by the platforms, in order to facilitate the provision of the underlying services (such as logistical support and other facilities offered by the electronic intervention) would transform the platforms as providers of the underlying services. They would therefore be subject to the more stringent relevant rules and forego any benefit connected to their ‘platform nature’. ‘Plan C’, for platforms not falling under the scope of the Services Directive, is to invoke the substantive rights deriving directly from Articles 56 and 57 of the Treaty on the Functioning of the European Union (TFEU), with the principles of transparency and non-discrimination developed by the Court these last years – a case law which, as explained elsewhere,14 may be source of important safeguards and (even) rights.15

13 The extremely favourable conditions of market access offered to platforms by the said directive have been extensively discussed by AG Szpunar in his opinion of 30 April 2019 in Case C-390/18 Airbnb Ireland EU:C:2019:336. 14 See V Hatzopoulos, ‘The Allocation of Limited Authorisations under EU Internal Market Rules’ in P Adriansee, F van Omeren, W de Ouden and J Wolswinkel (eds), Scarcity and the State (Cambridge, Intersentia, 2016); see also, by the same author, ‘Des marchés publics à la délivrance des autorisations: Spill-over all over?’ in I Govaere and D Hanf (eds), Scrutinizing Internal and External Dimensions of European Law: Les dimensions internes et externes du droit européen à l’épreuve – Liber Amicorum Paul Demaret, Vol 1 (Bruxelles, PIE-Peter Lang, 2013); and, more extensively, by the same author, ‘Du principe de non-discrimination (au niveau européen) au principe de bonne administration (au niveau national) ?’ (2016) 50 Cahiers de droit européen 311. 15 See, in particular, Joined Cases C-72/10 and C-77/10 Costa and Cifone EU:C:2011:699.

218  Vassilis Hatzopoulos Finally, the least favourable scenario for the platforms is to escape entirely from the application of Union law and to have to comply with the divergent legislations of the Member States. This is the fate that the Court reserved for Uber in its judgments. Uber Spain, the first occasion on which the CJEU took position on the collaborative economy, concerned specifically the issue of market access. It was followed by Uber France SAS,16 which bore on the criminal liability of Uber for failing to comply with the rules applicable on transportation undertakings.17 In both judgments the Court found against the platform. In Uber Spain, the Court held that all platforms offer an intermediation service. Under specific circumstances, however, [t]hat intermediation service must thus be regarded as forming an integral part of an overall service whose main component is a transport service and, accordingly, must be classified not as ‘an information society service’ within the meaning of … Article 2(a) of Directive 2000/31 …, but as ‘a service in the field of transport’ within the meaning of Article 2(2)(d) of Directive 2006/123.18

According to the judgment in Uber Spain the platform is itself participating in the underlying service when two conditions are simultaneously met:19 the platform (a) should be making possible the provision of that service, thus having a market-maker or ‘gatekeeper’ function; and (b) the platform should be exercising a ‘decisive influence’ over different attributes of this new service. The latter is true, according to the Court, when the platform: (1) organises the characteristics and functions of this new service; (2) selects the providers (drivers) according to its own criteria; (3) fixes the conditions of service delivery, (4) most importantly the price, (5) which it receives directly from the users before returning a part to the providers; (6) (indirectly) controls the quality of the service (vehicles, drivers, conduct) through a rating application; and (7) reserves to itself the right to ultimately exclude those providers which it does not deem fit.20 As noted elsewhere,21 this judgment only offers a solution applicable to the specific transportation platform and to those following the same business model, but not to other transportation platforms, such as, for example, Blablacar, let alone platforms in other segments of the market. The case, currently pending, before the CJEU concerning the status of Airbnb, will serve as a test-bed for many of these criteria; this time, however, the question is not whether the platform is active in

16 Case C-320/16 Uber France, above n 11. 17 Both have been preceded by the Order in Case C-526/15 Uber Belgium v Taxi Radio Bruxellois EU:C:2016:830, where the Court rejected as inadmissible the question whether Uber drivers should be considered to be in an employment relationship with the platform. 18 Para 40. 19 The fact that the two conditions are cumulative has been clearly spelled out in Uber France, above n 11, para 21. 20 Para 39. 21 Hatzopoulos, ‘After Uber Spain’, above n 11.

The Internal Market and Online Platform Economy  219 the underlying market (ie, short-term accommodation), but rather whether it is in the same market as traditional intermediaries (ie, real estate agents).22 The answer to each specific case will necessarily reflect the facts and the precise question put to the Court, but most importantly, it will vary depending on the exploitation model followed by each individual platform; and there are as many exploitation models as there are platforms! A last issue in relation to market access, is to know who will have the right to claim the application of favourable EU rules, where applicable. In traditional (two-sided) services the Court has held that EU rules can be claimed (a) by service providers in order to facilitate their activity;23 (b) by service recipients in order for them to have unfettered access to the service;24 and (c) by service recipients in order to facilitate the provision to them of a service by a non-national provider.25 In the tripartite relationship developing in the collaborative context it should be accepted that the platform is also able to claim the protection of EU free movement rules, not only against any limitation of its own freedom to provide (intermediation) services to the two sides of its market, but also against measures restricting these two parties (providers of the underlying service and consumers) from transacting with one another. Hence, for instance, Airbnb should have the right to invoke Articles 56 et seq TFEU and the Services Directive against national measures limiting to 90 the nights that an owner can offer his apartment through the platform.

C.  Consumer Protection European consumer protection law only applies to B2C (business-to-consumer) relationships, not B2B (business-to-business) or P2P (peer-to-peer) relationships – central to the collaborative economy. If the supplier is qualified as a ‘trader’ the rules will apply; if, on the contrary, the supplier is qualified as a ‘prosumer’, their relations with consumers will be governed only by traditional contract and/or tort law. This creates a potentially very important gap in the field of consumer protection as, in those cases where the providers of the underlying services are not themselves traders, the consumers will be devoid of any claim (other than based on contract or, worse, tort) against the prosumer who has made them suffer damage. Not only will such claims be unable to benefit from the favourable rules on jurisdiction and

22 Case C-390/18 Airbnb Ireland, above n 13 pending; for a brief discussion of the hearing of this case see: blogdroiteuropeen.com/2019/01/17/laudience-de-de-la-cjue-dans-laffaire-airbnb-irelandc390-18-continuite-ou-rupture-avec-larret-uber-par-edoardo-stoppoini/; AG Szpunar in his opinion of 30 April 2019 proposes that, contrary to Uber, Airbnb is an information society service provider. 23 Case 33/74 Van Binsbergen EU:C:1974:131. 24 Case 286/82 Luisi & Carbone EU:C:1984:35. 25 Case C-215/01 Schnitzer EU:C:2003:662.

220  Vassilis Hatzopoulos applicable law foreseen in favour of consumers,26 but they shall also be subject to ‘traditional’ burden of proof standards, and will require a violation of a contractual term, or fault. Most importantly, such claims shall be addressed against individuals who may lack any financial surface and may be unable to offer any redress to the aggrieved consumer. In most cases, the underlying service providers offer occasional collaborative services, in their spare time, in addition to their main job (or unemployment benefits);27 as a result, collaborative activity does not become the main ‘profession, enterprise, craft or profession’ of the individuals concerned. It could be otherwise if the collaborative activity was part of the wider scope of the professional activity of the person concerned, in which case it could be said that it constitutes a diversified expression of the profession exercised (for example, hoteliers promoting their rooms on Airbnb or licensed cab drivers driving for Uber). This would also be true for the inactive, unemployed or underemployed who, after a while, become economically dependent on their collaborative activity. Inspired by the practices of several Member States, the Commission in its 2016 Communication suggested the use of thresholds as criteria for the qualification of a ‘trader’: (a) the frequency of services, ie, whether the services are offered regularly or on a purely marginal or accessory basis; (b) the motive for profit, as opposed to the objective of exchanging assets or skills; and (c) the income generated by the activity concerned, and if it is higher/lower than those of other activities undertaken by the same person. It is important to note that these criteria only serve to clarify cases at the two extremes of the spectrum of possibilities; they remain of little use in most routine situations and would deserve further clarification. The Court in a recent judgment concerning the occasional sales of goods (in the analog economy) has established a long set of criteria which should be met for an individual to qualify as a trader;28 the Court, however, has cut short the 26 Under the Rome I Regulation (European Parliament and Council Regulation (EC) No 593/2008 of 17 June 2008 on the law applicable to contractual obligations (Rome I) [2008] OJ L177/6) and the Brussels I Regulation (European Parliament and Council Regulation (EU) No 1215/2012 of 12 December 2012 on jurisdiction and the recognition and enforcement of judgments in civil and commercial matters (Brussels I Regulation (recast)) [2012] OJ L351/1). 27 See, eg, the data published in Commission Staff Working Document ‘European agenda for the collaborative economy – supporting analysis’ SWD(2016) 184 final, 37–38. 28 Case C-105/17 Kamenova EU:C:2018:808, para 38, according to which ‘the referring court will, in particular, have to verify whether the sale on the online platform was carried out in an organised manner, whether that sale was intended to generate profit, whether the seller had technical information and expertise relating to the products which she offered for sale which the consumer did not necessarily have, with the result that she was placed in a more advantageous position than the consumer, whether the seller had a legal status which enabled her to engage in commercial activities and to what extent the online sale was connected to the seller’s commercial or professional activity, whether the seller was subject to VAT, whether the seller, acting on behalf of a particular trader or on her own behalf or through another person acting in her name and on her behalf, received remuneration or an incentive; whether the seller purchased new or second-hand goods in order to resell them, thus making that a regular, frequent and/or simultaneous activity in comparison with her usual commercial or business activity, whether the goods for sale were all of the same type or of the same value, and, in particular, whether the offer was concentrated on a small number of goods’.

The Internal Market and Online Platform Economy  221 importance of its judgment by making clear that these criteria are ‘neither exhaustive nor exclusive’,29 and that the qualification of a person as a trader should be carried out on a case-by-case basis,30 since this is a functional concept.31 It is clear from the above that the situation of the consumer depends on many fine legal qualifications, nourishing legal uncertainty. This is only made worse by the application of divergent national laws in the field of contract and tort law as well as intermediaries’ liability. Hence, several commentators put forward the idea of a ‘Platform Directive’,32 which would clarify the relations between the platform, final consumer and the supplier of the underlying service, irrespective of whether the latter is a professional or a prosumer. Further, it could set rules for the choice of jurisdiction and applicable rules in case of disputes. Such a directive could be in the interests of consumers, but also of platforms which, in view of the Court’s judgment in Uber Spain, may be held fully liable for the underlying service. A further issue is that of the content of consumer protection rules applicable to the collaborative economy. EU consumer protection law has been criticised as being overly formalistic and heavily dependent on an ‘information overload’, supposedly to the benefit of the ‘average consumer’.33 In the collaborative economy, however, the information which is most relevant for consumers is voluntarily provided by the platforms themselves in the form of reputational ratings, in order to build the trust on which collaborative transactions between unknown parties rest. Moreover, since platforms rely on a detailed ‘profiling’ of their users in order to perform their matching function, it is worth questioning whether the ‘average consumer’ is the appropriate benchmark,34 or whether smarter consumer protection rules, in the form of individualised default rules or else,35 also taking into account the hindsight offered by the behavioural analysis of the law, would be more appropriate.36 29 ibid, para 39. 30 ibid, para 37. 31 ibid, para 35. 32 See, eg, E Terryn, ‘The sharing economy in Belgium – a case for regulation?’ (2016) 5 Journal of European Consumer and Market Law 45, 50. 33 O Bar-Gil and O Ben-Sahar, ‘Regulatory Techniques in Consumer Protection: A critique of European Consumer Contract Law’ (2013) 50 Common Market Law Review 109; A Tor, ‘Some Challenges Facing a Behaviorally-Informed Approach to the Directive on Unfair Commercial Practices’ in T Tόth (ed), Unfair Commercial Practices: The Long Road to Harmonized Law Enforcement (Budapest, Pázmány Press, 2014); and more recently, G Helleringer and AL Sibony, ‘European Consumer Protection Through the Behavioral Lense’ (2017) 23 Columbia Journal of European Law 607. 34 See R Incadrona and C Poncibό, ‘The average consumer, the unfair commercial practices directive and the cognitive revolution’ (2007) 30 Journal of Consumer Policy 21. 35 See CR Sunstein, ‘Impersonal Default Rules vs Active Choices vs Personalized Default Rules: A Triptych’ (2012) (unpublished manuscript, on file with author), available at: nrs.harvard.edu/urn3:HUL.InstRepos:9876090; see also A Porat and LJ Strahilevitz, ‘Personalizing Default Rules and Disclosure with Big Data’ (2014) 112 Michigan Law Review 1417. 36 On the ways behavioural insights can be useful for law, see the classic A Tor, ‘The Methodology of the Behavioral Analysis of Law’ (2008) 4 Haifa Law Review 237, available at: weblaw.haifa.ac.il/ he/Faculty/Tor/Publications/The%20Methodology%20of%20the%20Behavioral%20Analysis%20 of%20Law.pdf; see also Helleringer and Sibony, above n 33.

222  Vassilis Hatzopoulos

D.  Data Protection Platforms collect and process a large amount of data concerning age, gender, residence, employment, professional qualifications, diet or food preferences, fitness, medical treatments, location and economic status on both sides of the two-sided market, in order to optimise the matching between platform users. These data are downloaded by users either in order to access the services of the platform (mandatory data), or to obtain the underlying service that corresponds to their profile (usage data). They can also be registered by the platforms, by tracking the IPs of their users, using cookies or gaining direct access to their devices (photos, addresses, etc). The data thus collected are used by the platforms to provide the best (matching) service to their users; data are also sold to other platforms or other types of user. In this way, data constitute a currency: they are offered by the users in return for services rendered by the platform (freemium), while representing a commodity on which platforms base their market power and which they occasionally sell. Rather than considering them as a bargaining chip, EU law apprehends and protects personal data and the right to privacy from a fundamental rights perspective, notably in Articles 7 and 8 of the EU Charter of Fundamental Rights, Article 16(1) TFEU, Article 8 of the European Convention on Human Rights and Council of Europe Convention 108. The GDPR,37 together with the (still draft) ‘e-Privacy Regulation’ replacing Directive 2002/58/EC,38 set out strict rules and severe penalties for the protection of data and privacy. However, few of the rules contained therein seem to take into account the way data is used in the collaborative economy. For example, consent given to platforms is rarely as explicit and specific as the rules require; nor is it preliminary and sufficiently informed. Moreover, sensitive data, such as disabilities, medications, test results, allergies, mental health information, drug use, etc are regularly processed by platforms; other data, such as the itinerary often followed by Uber customers may be revealing of their sexual preferences and lives, as the notorious Uber’s ‘Rides of Glory’ scandal has shown.39 Furthermore, the way in which the right to be forgotten – recognised by the Court and endorsed by the GDPR – should apply on data which have served as ‘freemium’ for services already consumed, raises salient questions of contract law. Last but not least, according to the rules mentioned above, no decision affecting individuals should be taken automatically on their profiling

37 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 [2016] OJ L119/1. 38 COM(2017) 010 final. 39 D Perry, ‘Sex and Uber’s “Rides of Glory”: The company tracks your one-night stands – and much more’ (The Oregonian/Oregon Live, 20 November 2014), available at: www.oregonlive.com/today/ index.ssf/2014/11/sex_the_single_girl_and_ubers.html.

The Internal Market and Online Platform Economy  223 alone. Yet, in practice, countless decisions are made by algorithms based on automated processing. Combined with personal data, non-personal data may also be used against the interests of data subjects. Uber has revealed that customers having low battery level on their devices are willing to pay up to 10 times more than usual for a ride; and Uber does have access to battery data! For the time being, the EU only captures non-personal data with a view to ensuring their free movement within the EU, and not from the perspective of protecting individuals.40

E.  Labour Law The level of remuneration of ‘workers’ in the collaborative economy depends on whether the service is totally digitalised (text editing, translation, coding, design, etc) and subject to global competition or, on the contrary, its delivery has a specific geographic location (driving, housing, etc). Even in the second case, where hourly remuneration may be higher than that of workers in the ‘analog’ market, gig-workers must bear the costs of infrastructure, maintenance and insurance and work without social security, paid leave or other social benefits. To ensure decent pay, individuals whose living depends on the collaborative economy must work at least 12 hours a day, without a fixed or regular schedule and compete for each minute of their work. Their work is characterised by a very high degree of specialisation, the isolated execution of specific mini-tasks distanced from the corresponding bigger project, the execution terms of each task being managed and controlled by algorithms and other automated means (cameras, click-counts, etc); by the very important information asymmetries existing between each individual gig-worker and the platform; by precariousness, uncertainty and all related health problems (stress, depression, etc); in addition to all disorders (musculoskeletal, visual fatigue, etc) inherent in online work.41 40 Regulation (EU) 2018/1807 of the European Parliament and of the Council of 14 November 2018 on a framework for the free flow of non-personal data in the European Union [2018] OJ L303/59. 41 See, eg, M Schmid-Drüner, ‘The situation of workers in the collaborative economy’ (2016) Employment and Social Affairs, European Parliament, available at: www.europarl.europa.eu/RegData/etudes/ IDAN/2016/587316/IPOL_IDA(2016)587316_EN.pdf; WP De Groen and I Maseli, ‘The Impact of the Collaborative Economy on the Labour market’ (2016) Centre for European Policy Studies (CEPS) Special Report, 20 available at: ec.europa.eu/DocsRoom/documents/16953/attachments/1/translations; C Codagnore, F Abadie and F Biagi, ‘The Future of Work in the “Sharing Economy”: Market Efficiency and Equitable Opportunities or Unfair Precarisation?’ (2016) EU Commission JRC Science for Policy Report No 27913, 4, available at: publications.jrc.ec.europa.eu/repository/bitstream/JRC101280/ jrc101280.pdf; and OECD, ‘New Forms of Work in the Digital Economy’ (2016) 22; and for a more recent view, from the other side of the globe see P Gahan, J Healy and D Nicholson, ‘Technology, the Digital Economy and the Challenge for Labour Market Regulation’ Australian Senate Select Committee on the Future of Work and Workers, Submission 111 – Attachment 3, available at: at: www.aph. gov.au/Parliamentary_Business/Committees/Senate/Future_of_Work_and_Workers/FutureofWork/ Submissions?main_0_content_1_RadGrid1ChangePage=6_20.

224  Vassilis Hatzopoulos In view of the above, some authors treat gig-workers of ‘Lumpen-cognitariat et salariat algorithmique’42 while others talk about ‘taylorism revamped’.43 Taylorism is being pushed to unprecedented lengths by the use of technology and algorithms, which operate as substitutes to direct managerial control and create power asymmetries between the platform and the worker. Such – typically opaque – algorithms may be used, as is the case with Uber,44 in order to assign work (as they perform the matching), in order to fix prices (for example, in case of surge) and in order to evaluate the worker in a semi-automated manner, on the basis of users’ ratings, acceptance/cancellation rates and other, more opaque criteria.45 The automatic ‘termination’46 of workers when their ratings fall below a certain level, or else ‘firing by algorithm’,47 is the most extreme manifestation of algocracy.48 The fact that platform work (or gig-work) sidesteps traditional categories of labour law has the consequence that important pieces of EU (and national, for that matter) legislation concerning the protection of workers, such as, inter alia, the Working Time Directive (WTD),49 the Fixed-Term Work Directive50 and the Part-Time Work Directive51 – and occasionally also the Temporary Work Agency Directive,52 risk being left without application. A proposal put forward by the Commission in order to bring within the definition of ‘worker’ many of those engaged through platforms, is still in the pipeline.53

42 Expression borrowed by Olivier Ertsscheid, ‘Du digital labor à l’uberisation du travail’ (Numerique 25 January 2016), available at www.inaglobal.fr/numerique/article/du-digital-labor-l-uberisationdu-travail-8747#intertitre-5. 43 MA Cherry, ‘Beyond Misclassification: The Digital Transformation of Work’ (2016) 37 Comparative Labor Law & Policy Journal 544; MA Cherry, Saint Louis U Legal Studies Research Paper No 2016-2; available at: ssrn.com/abstract=2734288. 44 Codagnore, Abadie and Biagi, above n 41, 40. 45 It has been documented, for example, that Uber in San Francisco requires drivers to have cancellation rates below 5% and an acceptance rate of 90%, see Codagnore, Abadie and Biagi, above n 41, 39. 46 As the word is appropriately used in the contract signed between Uber and its drivers – a brush of humour from Uber’s lawyers, a tribute to the ex-governor of California, or a truthful expression of the platform’s attitude towards its workers? 47 Cherry, above n 43, Saint Louis U Legal Studies Research Paper 22. 48 A Aneesh, ‘Global Labour: Algocratic Modes of Organization’ (2009) 27 Sociological Theory 347; it is interesting to note, from an etymological point of view, that although the term algorithm sounds as being of Greek origin, it comes in fact from the medieval Latin word ‘algorismus’, itself inspired from the name of the Persian mathematician Al-Khwārizmī; in Greek, the word ‘algos’, stands for ‘pain’, ‘sufferance’ or ‘evil’. 49 European Parliament and Council Directive 2003/88/EC of 4 November 2003 concerning certain aspects of the organization of working time (Working Time Directive – WTD) [2003] OJ L299/9 Art 2(1). 50 Council Directive 1990/70/EC of 28 June 1999 concerning the framework agreement on fixedterm work concluded by ETUC, UNICE and CEEP [1999] OJ L175/43. 51 Council Directive 97/81/EC of 15 December 1997 concerning the Framework Agreement on part-time work concluded by UNICE, CEEP and the ETUC [1998] OJ L14/9. 52 European Parliament and Council Directive 2008/104/EC of 19 November 2008 on temporary agency work [2008] OJ L327/9. 53 ‘Proposal for a Directive of the European Parliament and of the Council on transparent and predictable working conditions in the European Union’ COM(2017) 797; political agreement was reached in February 2019.

The Internal Market and Online Platform Economy  225 In the meantime, gig-workers see their status determined by national courts, applying different rules and reaching divergent results. Thus, to remain only within European territory, the Central London Labour Court, then the Appeals Labour Court, and finally the London Court of Appeals held in the Aslam, Farrar et al v Uber case that Uber drivers are workers (an intermediate category between independent contractors and employees).54 The Labour Court of Paris (Prud’hommes) ruled in the case of Le Cab55 (a competitor of Uber) that the drivers were employees, but more recently the same Court considered in the Menard case56 that Uber drivers were independent. The Paris Commercial Court concluded the same way in the Viacab case.57 The controversy in France came to a (provisional) end with the recent judgment of the Cour de cassation in Uber France, where the French Supreme Court found against the platform.58 The Commercial Court of Madrid considered that the drivers of Blablacar (who enjoy a degree of autonomy much higher than that of the Uber drivers) were not employees either.59 In the meantime, in Uber Spain,60 a case concerning market access and not employment, the CJEU found that the platform exerts ‘a decisive influence’ over the drivers. Could this mean that there is an employer–employee relationship?

III.  The EU Internal Market in Need of Regulation for Platforms All the above issues, together with others pertaining to competition law not discussed here,61 create great legal uncertainty in the internal market. Such uncertainty is further nurtured by dissonant or even opposing administrative measures, decisions by independent authorities and bodies (such as, for example, consumer protection bodies, competition authorities, etc) and courts. National legislators have hesitantly started adopting measures, thus pushing fragmentation even further. The above are signs that self-regulation, as practised by the platforms so far, has brought the collaborative economy to its present stage of fruition, but may no more govern it effectively under its current expansion, both in terms of sectors and of users. Self-regulation, however, needs be taken into account by any initiative of traditional top-down regulation.

54 Aslam, Farrar et al v Uber, Court of Appeal [2018] EWCA Civ 2748, Case n A2/2017/3467. 55 Arrêt du 20 décembre 2016, n° 14/16389. 56 Arrêt du 29 janvier 2018, n° F 16/11460 Menard c Uber. 57 Arrêt du 30 janvier 2017, n° 2014054740 Viacab c Uber. 58 Arrêt du 4 mars 2020, n° 374 FP-P+B+R+I Uber France and Uber BV c X. 59 Arrêt du 2 février 2017, 6/2017 Confebus v BlablaCar ES:JMM:2017:6. 60 Discussed above II.B; one is reminded that a case specifically concerning the working status of Uber drivers has been rejected as inadmissible on technical grounds by the Court; see above, n 11. 61 But see Hatzopoulos, The Collaborative Economy and EU Law, above n 4, ch 4.

226  Vassilis Hatzopoulos

A.  The Limits of Self-Regulation The collaborative economy is based on trust in platforms and peers. While peer reviews and reputation rating systems are the flagship instruments for building trust, other (more general) measures of self-regulation also play an important role.

i.  Regulation by the Platforms Platforms often establish quality standards for service providers, in order for the latter to have access to the former. For example, transport platforms (such as Uber) impose conditions on the types of car (type, age, etc) and the services that must be offered (water bottles, internet connection, Bluetooth connection, etc). As a result, the proposed service becomes standardised and the expectation of consumers tends to correspond to the service actually offered by each platform. By the same logic, but pushing further, Airbnb next to its regular hosts and apartments, distinguishes experienced and reliable ‘superhosts’, as well as apartments with a certain degree of architectural/stylistic interest and a certain level of comfort, offered under the banner ‘Airbnb plus’. Other examples of self-regulation include the performance of background checks (concerning in particular previous convictions, etc), identification of the users based on official documents, as well as personality verification and mentoring of new service providers. Big data analytics are also used to control transactions and block or report suspicious activity (such as fake reviews or even money laundering) that are then investigated and evaluated by human controllers. Finally, the insurance coverage offered by platforms to users minimises uncertainty and increases digital trust. Other voluntary actions, such as Airbnb’s anti-discrimination policy, or other measures put in place by the platforms in favour of their ‘employees’ (in the form of minimum monthly wages, health insurance, etc) contribute to increase confidence in the digital economy. In total, it can be seen that the platforms impose rules or standards that affect one of four areas: the conditions of access to the activity (personal qualifications, absence of criminal record); the conditions of exercise of the activity (quality of service, use of infrastructure or specific logos); the supervision of the performance of the activity (by algorithms or cameras as well as the extraction of statistics on the performance of each provider); and finally, the adoption of insurance policies. By both their object and their content, private regulations are very similar to traditional public regulations; the question then arises as to the type of interest they pursue. While it is true that platforms have an interest in enhancing the trust of their users (both providers and consumers), they are much less inclined to also protect the interests of third parties (for example, residents of an Airbnb apartment), which may be subject to negative externalities or, more generally, the public interest.

The Internal Market and Online Platform Economy  227

ii.  Regulation by the Peers: Reputation Rating The difficulty of distinguishing good from bad quality also exists in the traditional economy and has been famously explained by Akerlof in his market for ‘lemons’.62 The reputation rating system used by the platforms to a large extent resolves this problem: through the evaluations and the reputation score potential customers obtain information on the quality of products or services before consuming them. Thus, the reputation of a supplier in the collaborative economy is the equivalent of the brand name or trademark in the traditional economy. Reputation thus acquires a significant market value, which, however, depends exclusively on the arrangements of each platform. The mechanisms for establishing reputation can be qualitative (in the form of comments), numerical (in the form of star attribution), or a combination thereof. Although scoring mechanisms are of paramount importance, both to inform peers and to allow platforms to make decisions, they raise several issues: • Peers are typically able to evaluate the ‘visible’ part of the service (for example, the cleanliness of an apartment) not the ‘invisible’ which, however, may be much more important (for example, fire safety). • The ex post evaluation of a service cannot always prevent the worst-case scenario from happening, as can real ex ante regulation. • Peer reviews are not always objective and are often subject to both the charm of personal and informal contact and the fear of retaliation assessments; a study showed that 95 per cent of Airbnb ratings gave four stars or more.63 • Ratings can be manipulated by both platforms and their competitors. • Peer-based reputation systems are confronted with the phenomenon of cold start, ie, the difficulty for new entrants to obtain consumer confidence and market share. In this sense, ratings are an important barrier to entry. • The opposite problem is also present, namely reputation-skimming: a provider who has accumulated a good reputation ‘burns’ his last consumers before flying off to some exotic destination. • Assessments can be fully defamatory or amount to hate speech. Different types of solutions – all imperfect – may counter the above problems. A first category of measures, to be adopted by each individual platform, could aim in framing in a strict and transparent manner the way that peer reviews operate and the impact they are likely to have on providers. Thus, only actual users of a service (as opposed to mere visitors etc), who are real people (as opposed to robots) and who can be identified and (if need be) traced, should be allowed. 62 G Akerlof, ‘The Market for “Lemons”: Quality Uncertainty and the Market Mechanism’ (1970) 84 Quarterly Journal of Economics 488. 63 G Zervas and D Prosperio, ‘A First Look at Online Reputation on Airbnb, Where Every Stay is Above Average’, available at: people.bu.edu/zg/publications/airbnbreviews.pdf.

228  Vassilis Hatzopoulos Further, it should be clear whether the reviews are reproduced as such or if, on the contrary, they are manipulated by some algorithm. In this regard, French decree 2017-1436 ‘on the information obligations relating to online consumer evaluations’ provides a good illustration:64 platforms must offer information on the existence and characteristics of a control system of the evaluations; the date of publication of the evaluation and the date of the commented experience; the evaluation criteria; the existence (or not) of counterpart in exchange for the evaluation offered; the duration during which the evaluation will remain published, then stored by the platform; the possibility of contacting the peer who did the review; the possibility and conditions of modification of the evaluation; and the reasons which could justify that an evaluation is not posted. A second category of measures, to be adopted by the platforms in consultation with each other, or by standardisation bodies, concerns the parameterisation of the evaluation. Thus, evaluations in each business sector should be based on common characteristics and similar criteria and/or scales, so that it is widely possible to compare/combine results across platforms; this would also have a pedagogical effect for users who would become more proficient evaluators. Finally, a third category of solutions lies with platforms that deal specifically with the reputation of others. These reputation aggregators operate through reputational data aggregations and rely on social networks or centralised collection of reputation information from parties across multiple platforms. They may be either sector-specific, as for example, Tripadvisor, or global, as for example, Neputation, PeerIndex. The use of such platforms would promote multi-homing, ie, the mobility of providers between platforms – a valuable means for enhancing competition between platforms. All these proposals, however, present only imperfect solutions to the problems mentioned above. Thus, the question of the desirability of a regulatory intervention concerning reputation rating – and more generally the collaborative economy – becomes increasingly acute.

B.  The Need for Some ‘External’ Regulation? Several questions are raised in relation to ‘external’ regulation in the collaborative economy, in particular concerning its desirability in a field which has thrived specifically because of the absence of any regulation: its timing, ie, whether it should occur immediately or later, once a critical mass of platforms and users has been developed; whether it should operate ex ante (through authorisations, etc) or ex post (through insurance, etc); whether it should be based on a top-down, bottom-up or participatory process; whether it should occur at the local, national or supranational level. 64 JORF n° 233 5/1/17, available at: www.legifrance.gouv.fr/eli/decret/2017/9/29/ECOC1716649D/ jo/texte.

The Internal Market and Online Platform Economy  229 Any effort to regulate the collaborative economy should ‘internalise’ two sets of constraints, legal and political. From a legal point of view constraints may come from domestic law, both constitutional (notably the principle of nondiscrimination, equal treatment and freedom of enterprise), as well as contract and tort law. Constraints may also arise from EU law, including the free movement and competition rules; additional limits may arise from the competence-share between the EU and its Member States, and their exercise by the former, leading to the pre-emption of the latter. A third source of limits to Member States’ regulatory freedom may come from international trade agreements, such as the GATS: a Member State which has offered ‘market access’ commitments in a sub-sector without any limitations, will be precluded from introducing new restrictive legislation even if such legislation is needed to frame the collaborative economy. In addition to legal constraints, there are also political constraints; it is sufficient to recall, in this respect, the (violent) reactions against Uber. The vision proposed here is that, especially after the judgments of the CJEU in Uber Spain and Uber France, the need for regulation at the European level is present and pressing. By refusing to recognise the specific characteristics of the collaborative economy, the CJEU is turning on its head the regulatory gap that so far has worked in favour of platforms to their detriment, and in favour of their consumers and ‘incumbents’. This result runs counter to the Commission’s declared intention to foster the development of online platforms and the collaborative economy, as part of the Digital Single Market strategy.65 It is also opposed to the EU’s aspiration to catch up with the United States in the development of the platform economy. Based on its findings in Uber Spain, where it admitted that the platform participates in the underlying (transportation) service, the Court in Uber France held the platform criminally liable for not possessing the relevant administrative authorisations. The preliminary question was referred to the Court by the TGI Lille only a few months after the French Conseil d’Etat had issued a judgment relieving Uber from any charges.66 Hence, the risk of divergent solutions and regulatory fragmentation is present, not only between different Member States, but also between the jurisdictions of one and the same Member State (since the TGI Lille will have to follow the CJEU and rule against the Conseil d’Etat). The Conseil d’Etat has expressed its preference for some kind of regulation, in its 2017 annual study on ‘uberisation’:67 it put forward the idea of regulatory action at the European level, combining hard and soft law. The Conseil d’Etat also discussed the idea of submitting the collaborative economy to special – more

65 See the 2016 Communications on Digital Platforms COM(2016) 356 and on the Collaborative Economy COM(2016) 356. 66 CE 9 mars 2016 Uber France ea n° 388213, 388343, 388357. 67 Puissance publique et plateformes numériques: accompagner l’ ‘ubérisation’, available at: www.conseil-etat.fr/Decisions-Avis-Publications/Etudes-Publications/Rapports-Etudes/ Etude-annuelle-2017-Puissance-publique-et-plateformes-numeriques-accompagner-l-uberisation.

230  Vassilis Hatzopoulos flexible – rules; an idea previously mooted by the Italian Competition Authority in a position paper published in September 2015, in which it suggested that Uber and its drivers should be considered tertius genus compared with existing categories.68 Similarly, the OECD suggested a model of cooperative (shared) responsibility, whereby ‘the burden of offering a reasonable level of consumer protection is a shared one’, with practices that include ‘effective reputation, monitoring and feedback mechanisms, educating and informing peers, to make sure peers properly inform other peers, establish secure ways of communication and mediation, and make sure peers can be identified in case matters go wrong’.69 Legislators in the Member States have taken legislative initiatives, often in divergent directions. Thus, in France, the 2016 Digital Republic Act,70 supplemented by three decrees of 2017,71 put the emphasis on transparency and fairness without interfering too much with the actual activities of the platforms. The Italian legislator, on the other hand, in its legislative proposal 3564/2016 is much more intrusive, since it gives the Autorita Guarante della Concorrenza e del Mercato (AGCM) the responsibility of approving the platforms’ General Policy Statement in which they are supposed to set out their general contract terms and conditions, pricing rules, reputation evaluation systems etc, while it also entrusts it (the AGCM) with overseeing their operation. The same proposal requires ‘foreign’ platforms, for tax reasons, to have a secondary establishment in Italy. This proposal raises several questions, in particular as regards its compatibility with EU law: first, the system of prior authorisation must be sufficiently justified under Articles 49 and 56 TFEU and, more demandingly, under Article 9 of the Services Directive. Second, the requirement of a secondary establishment on the national territory is a condition expressly ‘blacklisted’ under Article 14 of the Service Directive, and may in no way be justified.72 These national initiatives come to occupy the regulatory space left by the European legislator. The void left at the European level corresponds to the deliberate choice of the Commission to offer only guidelines and let the collaborative economy run on its own dynamism. The void, however, is only partial. In addition to the two 2006 Communications on online platforms in general,73 and the collaborative economy in particular,74 other recently adopted texts create a broad regulatory environment for the collaborative economy.

68 For a discussion of this position and other developments in Italian law, see A de Franceschi, ‘The Adequacy of Italian Law for the Platform Economy’ (2016) 5 Journal of European Consumer and Market Law 56. 69 Organisation for Economic Co-operation and Development (OECD), ‘Protecting Consumers in Peer PlatformMarkets,ExploringtheIssues’(2016)OECDDigitalEconomyPapersNo253,23,availableatwww.oecdilibrary.org/science-and-technology/protecting-consumers-in-peer-platform-markets_5jlwvz39m1zw-en. 70 Law 2016-1321 of 7 October 2016. 71 Decrees 2017-1434, 2017-1435 and 2017-1436 of 29 September 2017. 72 Case C-360/15 Gemeente Amersfoort v X et Visser EU:C:2018:44. 73 COM(2016) 288. 74 COM(2016) 356.

The Internal Market and Online Platform Economy  231 First, there is the old E-Commerce Directive75 as well as the new ‘e-commerce’ rules of the Digital Single Market package, notably on geo-blocking,76 on crossborder delivery of parcels77 and on digital contracts.78 Second, the Fairness and Transparency Regulation for business users using online intermediation services would fill an important gap by regulating the relationship between platforms and their professional service providers (hence P2P relations, as opposed to B2C relations already covered by the consumer laws).79 Third, the Regulation on the free flow of non-personal data, shall complement the GDPR in facilitating the transfer of data.80 Lastly, the Directive on transparent and predictable working conditions covers several forms of flexible, occasional and on-demand work, such as that offered through platforms, thus addressing some of the labour law issues raised above.81 The adoption of the above texts makes, to a large extent, the ‘platform directive’ proposed by part of the doctrine redundant.82 However, there would still be areas in which the EU could, through more or less binding rules, offer added value: by simplifying and standardising the general terms and conditions of platforms active in similar areas; by making clear how data protection rules apply and, where appropriate, by establishing different levels of protection, represented by different pictograms; by establishing personalised consumer protection rules, based on the profiling of each individual; by guiding the platforms in (geographic) areas where their activity is most useful for the public interest (for example, Uber could offer an efficient substitute for urban–suburban transport in outlying areas, where public transport is constantly in need of subsidisation); more generally, by adjusting 75 Directive 2000/31/EC of the European Parliament and of the Council of 8 June 2000 on certain legal aspects of information society services, in particular electronic commerce, in the Internal Market [2000] OJ L178, 1–16. 76 Regulation (EU) 2018/302 of the European Parliament and of the Council of 28 February 2018 on addressing unjustified geo-blocking and other forms of discrimination based on customers’ nationality, place of residence or place of establishment within the internal market and amending Regulations (EC) No 2006/2004 and (EU) 2017/2394 and Directive 2009/22/EC [2018] OJ L601/1. 77 Regulation (EU) 2018/644 of the European Parliament and of the Council of 18 April 2018 on cross-border parcel delivery services [2018] OJ L112, 19–28; and Commission Implementing Regulation (EU) 2018/1263 of 20 September 2018 establishing the forms for the submission of information by parcel delivery service providers pursuant to Regulation (EU) 2018/644 of the European Parliament and of the Council [2018] OJ L238/65. 78 Directive (EU) 2019/770 of the European Parliament and of the Council of 20 May 2019 on certain aspects concerning contracts for the supply of digital content and digital services [2019] OJ L136/1; and Directive (EU) 2019/771 of the European Parliament and of the Council of 20 May 2019 on certain aspects concerning contracts for the sale of goods, amending Regulation (EU) 2017/2394 and Directive 2009/22/EC, and repealing Directive 1999/44/EC [2019] OJ L136/28. 79 Regulation (EU) 2019/1150 of the European Parliament and of the Council of 20 June 2019 on promoting fairness and transparency for business users of online intermediation services [2019] OJ L186/57. 80 Regulation (EU) 2018/1807 of the European Parliament and of the Council of 14 November 2018 on a framework for the free flow of non-personal data in the European Union [2018] OJ L303/59. 81 Directive (EU) 2019/1152 of the European Parliament and of the Council of 20 June 2019 on transparent and predictable working conditions in the European Union [2019] L186/105. 82 eg, Terryn, above n 32, 50.

232  Vassilis Hatzopoulos collaborative economy services to serve as substitutes for traditional services of general economic interest. In the meantime, the Commission has set up an observatory on the economics of online platforms in order to better understand the challenges ahead.83

IV. Conclusion The platform economy questions established legal categories both under national and under EU law. EU internal market law is being affected in many ways, from core market access conditions and workers’ rights to consumer and data protection. In all these areas the platform economy either stretches the scope and content of existing rules or, more radically, operates in a legal vacuum. The regulatory indeterminacy which has benefited platforms and has allowed them to thrive in the early phases of their development, is now reaching its limits as the size of the platform economy grows by the day. Organised opposition by the incumbents, unhappy user experiences, unexpected negative externalities and even scandals, touching upon life, integrity, dignity and other fundamental rights of users on both sides of the two-sided market, have backfired against platforms and have prompted divergent regulations and/or judicial reactions in different states. Hence, the risk of fragmentation in the internal market along national, and even local, lines is now as real as ever. Against this danger, it is suggested here that the EU should move on from its initial wait-and-see attitude and intervene more decisively, at least in relation to ‘horizontal’ matters of the platform economy, such as transparency, fairness, non-discrimination and data protection (problems exacerbated by the use of algorithms and big data), as well as in the area of consumer protection and reputation management. The adoption of such rules, principles or standards should lead to the issuance of some kind of ‘passport’ allowing startups and other European platforms (typically small or medium in size) to have unfettered access in all the Member States.84

83 Commission Decision COM(2018) 2393 final of 26 April 2018. 84 Of course, compliance with local regulations on zoning, traffic regulation, opening hours, etc, would still have to be complied with.

11 ‘Tinkering or Fundamental Overhaul?’ The Past, the Present and the Future of the Digital Single Market CLAIRE BURY AND IRENE ROCHE LAGUNA*

I. Introduction A.  DSM as Compared with Previous Initiatives The Digital Single Market (DSM) was the second of the Juncker Commission’s 10 priorities.1 This guaranteed regular attention from the European Council to ensure progress and in the context of the Inter-Institutional Agreement, yet did not facilitate any shortcuts as regards impact assessments or the full rigours of the legislative process. Compared with previous Commission initiatives – such as the 2010 Digital Agenda for Europe2 – the DSM followed a very focused approach with a limited number of legislative and non-regulatory measures aimed primarily at making the market work for business and for consumers. It combined short-term, targeted initiatives of great interest to consumers – for example, the end of roaming charges, portability of online content services, geoblocking – with more far-reaching proposals with broader regulatory implications, such as the European Electronic Communication Code, the Cybersecurity Act, the modernisation of the Copyright framework, and the first regulatory instrument targeting online platforms on fairness and transparency for business users of online intermediation services. At the end of the Commission’s mandate, 28 out of 30 legislative measures3 have been finally adopted by the co-legislators, which * The content of this chapter does not reflect the official opinion of the European Commission. Responsibility for the information and views expressed therein lies entirely with the authors. 1 See: ec.europa.eu/commission/priorities/digital-single-market_en. 2 Communication from the Commission ‘A Digital Agenda for Europe’ COM(2010) 245 final, 19 May 2010, available at: eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=COM:2010:0245:FIN:EN:PDF. 3 On top of the measures mentioned in this chapter, the DSM contained other initiatives (see Annex II) such as the Wifi4EU Regulation, the .eu top domain Regulation, the 700MHz Decision, VAT rules on

234  Claire Bury and Irene Roche Laguna is a relatively impressive scorecard. But of course, beyond the mere number of measures adopted, this result needs to be measured against the original level of ambition and objectives set by the Commission as well as the effective implementation of the measures by Member States, assisted by the Commission.

B.  Why a DSM Strategy? Citizens travel more, it is easier to sell and buy online or offer services cross-border: in 2018, 78 per cent of internet users played or downloaded games, images, films or music;4 and 17.2 per cent of EU SMEs sold online,5 compared with 14 per cent in 2014. It is also easier to provide services from a different MS via the internet: 8.38 per cent of EU SMEs sold cross-border in 2018,6 compared with 6.49 per cent in 2014. These are not necessarily impressive figures. The European Union has always lagged behind its potential as regards the digital economy. Achieving its potential in terms of online and cross-border exchanges is both a long-standing objective for Europe (in 2011, the Commission identified the digital single market as a lever to boost growth and strengthen confidence)7 and a trigger for new challenges that needed to be addressed. While some barriers cannot be solved so easily, at least not through regulatory means (language, culture, education), others still need to be dismantled to allow EU companies to compete with the current giants amongst the digital economies such as the US and China. By creating a connected DSM, the Commission estimated up to €250 billion of additional economic growth in Europe could be achieved.8 Only time and the effective enforcement of the adopted measures will tell whether this ambitious objective will, in fact, be attained.

C.  Evolution or Revolution? From the legal and regulatory perspective, the Commission’s approach in the DSM has been one of evolution, rather than revolution. The main motivation of the Commission’s measures was facilitating exercise of single market freedoms and e-commerce and e-books, parcel delivery Regulation, Marrakesh Treaty, single digital gateway or the non-cash payments Directive. On the other hand, it deals with additional initiatives in the digital field that were proposed (and eventually adopted) later. 4 Digital Economy and Society Index 2018, indicator: 3a2 Music, Videos and Games, available at: ec.europa.eu/digital-single-market/en/desi). 5 Indicator 4b1: SMEs Selling Online. 6 Indicator 4b3: Selling Online Cross-border. 7 Single Market Act – ‘Twelve levers to boost growth and strengthen confidence – Working together to create new growth’ (COM(2011) 206). 8 J-C Juncker, ‘A New Start for Europe: My Agenda for Jobs, Growth, Fairness and Democratic Change’, Opening Statement in the European Parliament Plenary Session (Strasbourg, 15 July 2014), available at: ec.europa.eu/commission/sites/beta-political/files/juncker-political-guidelines-speech_en.pdf.

‘Tinkering or Fundamental Overhaul?’  235 avoiding the reappearance online of new barriers, that had been more or less successfully erased in the offline world, for example through use of geo-blocking techniques. After all, the DSM is only the translation, into the online environment, of the single market rules and principles that have been the fundamentals of the European Union legal order from its outset. Against this background, the DSM’s objective is to tackle areas where adaptation to new digital challenges were identified, using a problem-based approach. This does not mean that the DSM’s impact is minor in nature: it is fair to say that the existing legal framework has been modernised to a significant extent. Some areas, such as rules on sales of digital content9 and business relations on platforms10 established new rules and involved innovative legal approaches. The Electronic Communications Code (the Code)11 is based on the existing framework for electronic communication, and familiar concepts derived, for example, from competition rules, but it also integrates some new concepts such as connectivity and co-investment aimed at stimulating the critical mass of private sector investment which will be needed to upgrade the EU’s digital infrastructure. In areas such as portability of content12 and geo-blocking,13 measures serve rather to clarify how consumers can benefit from existing freedoms to provide cross-border (digital) services. When identifying bottlenecks for the smooth development of digital business models and opportunities, the Commission found examples of rules which might have made sense in the offline world, but have less relevance online. For instance, while in the offline world it makes sense to require bookkeepers to retain invoices and documentation in the territory of a Member State for potential inspections, the translation of these rules to mandatory localisation requirements for data (e-invoices) would represent a clear restriction to the cross-border movement of data. The Free Flow of Data Regulation was adopted to avoid such rules and ensure that any restrictions on free movement of data in the EU are based only on public security grounds.14

9 Directive (EU) 2019/770 of 20 May 2019 on certain aspects concerning contracts for the supply of digital content and digital services and Directive (EU) 2019/771 of 20 May 2019 on certain aspects concerning contracts for the sale of goods, amending Regulation (EU) 2017/2394 and Directive 2009/22/EC, and repealing Directive 1999/44/EC [2019] OJ L136/1. 10 Regulation (EU) 2019/1150 of 20 June 2019 on promoting fairness and transparency for business users of online intermediation services [2019] OJ L186/57. 11 Directive (EU) 2018/1972 of 11 December 2018 establishing the European Electronic Communications Code [2018] OJ L321/36. 12 Regulation (EU) 2017/1128 of 14 June 2017 on cross-border portability of online content services in the internal market [2017] OJ L168/1. 13 Regulation (EU) 2018/302 of 28 February 2018 on addressing unjustified geo-blocking and other forms of discrimination based on customers’ nationality, place of residence or place of establishment within the internal market and amending Regulations (EC) No 2006/2004 and (EU) 2017/2394 and Directive 2009/22/EC [2018] OJ L601/1. 14 Art 4(1) of Regulation (EU) 2018/1807 on a framework for the free flow of non-personal data in the European Union ([2018] OJ L303/59) establishes that ‘Data localisation requirements shall be prohibited, unless they are justified on grounds of public security in compliance with the principle of proportionality’.

236  Claire Bury and Irene Roche Laguna

D.  Differences in Regulating the Online World On the other hand, there are some circumstances that are specific to the online environment which cannot be extrapolated from the offline world. For instance, in the offline world no one would contest that only a judge is empowered to determine whether information or activity is illegal. However, in the online world illegal content is transmitted so quickly that its removal depends on the necessary intervention of private actors (online platforms, trusted flaggers) to react swiftly through informal notice-and-action procedures.15 It is, however, necessary to ensure that such private decisions can be challenged in court.16 A further challenge which has often arisen during the negotiation of DSM measures is whether start-ups or SMEs should be exempted from regulation so that their capacity to innovate and take risks is not unduly affected. Existing SME exemptions in the single market are based on turnover and number of employees,17 but such criteria are not adapted to the online world where companies with relatively few employees can reach and have an impact on large audiences. In this context, the legislator has taken careful account of the fact that the online space is characterised by a few ‘tech giants’, in particular in the context of the platform economy. If the legislator wants to create a Digital Single Market that facilitates and provides incentives for the take up of online services by (European) providers, it needs to set lighter regimes for small enterprises, without however creating a glass ceiling or a disincentive to grow. It remains for the next Commission to ascertain whether ‘GAFA’18 specific legislation is now needed and if so what criteria should be used for regulatory purposes.

E.  A Level Playing Field Online and Offline in the EU? Finally, the DSM has also been characterised by a constant effort to create and maintain a more level playing field between more traditional operators and online service providers. New business models and new services have been created and offered on top of traditional ones, creating some competitive distortions. In doing so, there have been difficulties in adapting the regulatory framework to these new businesses while recognising their different nature and capacities. For instance, the new Audio-Visual Media Services Directive (AVMSD)19 imposes certain 15 ‘Given that fast removal of or disabling of access to illegal content is often essential in order to limit wider dissemination and harm, those responsibilities imply inter alia that the service providers concerned should be able to take swift decisions as regards possible actions with respect to illegal content online’ (Commission Recommendation of 1 March 2018 on measures to effectively tackle illegal content online, C(2018) 1177 final). 16 Judgment of 27 March 2014, C-314/12 UPC Telekabel Wien EU:C:2014:192, para 57. 17 Commission Recommendation of 6 May 2003 concerning the definition of micro, small and medium-sized enterprises. 18 Term commonly used to refer to tech giants, such as Google, Amazon, Facebook and Apple. 19 Directive (EU) 2018/1808 of 14 November 2018 amending Directive 2010/13/EU on the coordination of certain provisions laid down by law, regulation or administrative action in Member

‘Tinkering or Fundamental Overhaul?’  237 obligations on video-sharing platforms, such as YouTube, which are similar, but not the same, as on video-on-demand or traditional broadcasting services. The abovementioned Electronic Communications Code20 extends some consumer rules to ‘over-the-top’ (OTT) services, number-independent interpersonal communication services which are functional substitutes of voice calls or SMS. In parallel, the Court of Justice of the EU ruled that UberPop is a transport service, rather than an information society service,21 thus acknowledging the disruptive nature of digital actors in traditional offline markets. In a certain way, the Court also contributed to levelling the playing field in the transport sector. A further hurdle has been to ensure positive synergies and careful delimitation of the application of the competition rules and internal market regulation. This was most obvious during the work on the Regulations on portability,22 geo-blocking,23 the broadcasting transmissions Directive,24 the platforms-tobusiness Regulation25 and the Electronic Communications Code.26 The challenges faced in the digital age by classical competition tools has been also identified as an issue for study by the Juncker Commission to inform potential further steps.27

F.  A Level Playing Field at International Level? The DSM gives the EU a firmer platform for working to craft and influence the international regulatory agenda (the EU as a normative super power), not only in fora such as the Internet Governance Forum but also through its bilateral relationship with its neighbours such as the Western Balkans and the Mediterranean and regions such as Africa and Asia. Discussions in the framework of the World Trade Organization towards e-commerce trade rules could prove soon whether this is the case.28 Section II gives a rapid overview of what has been achieved so far: the success stories and the necessary trade-offs/compromises made – and new challenges encountered – during the negotiations. As the digital environment changes at incredible speed, regulating the DSM is a moving target: section II presents the States concerning the provision of audiovisual media services (Audiovisual Media Services Directive) in view of changing market [2018] OJ L303/69. 20 Directive (EU) 2018/1972, above n 11. 21 Judgment of the Court of 20 December 2017, Case C-434/15 Asociación Profesional Élite Taxi v Uber Systems Spain SL EU:C:2017:981. 22 Regulation (EU) 2017/1128, above n 12. 23 Regulation (EU) 2018/302, above n 13. 24 Directive (EU) 2019/789 laying down rules on the exercise of copyright and related rights applicable to certain online transmissions of broadcasting organisations [2019] OJ L130/82. 25 Regulation (EU) 2019/1150, above n 10. 26 Directive (EU) 2018/1972, above n 11. 27 J Crémer, Y-A de Montjoye and H Schweitze, ‘Competition Policy for the digital era – Final report’, available at: ec.europa.eu/competition/publications/reports/kd0419345enn.pdf. 28 Negotiations to put in place global rules on electronic commerce were launched by the EU and 48 members of the World Trade Organisation on 25 January 2019. Joint statement, available at: trade. ec.europa.eu/doclib/docs/2019/january/tradoc_157643.pdf.

238  Claire Bury and Irene Roche Laguna emerging gaps that are being identified and the areas which are likely to be the focus of attention of the Von der Leyen Commission which took office on 1 December 2019.

II.  The Past and the Present: DSM Strategy Communication 2015 In 2014, when the Juncker Commission took office, the Digital Single Market was the second in its 10 announced priorities.29 This was translated in May 2015 into a Communication on the DSM Strategy,30 where the Commission presented its ‘roadmap’ for the DSM. The Strategy presented three main pillars that ensure: (1) a real DSM from which consumers and businesses can benefit; (2) fast and efficient digital networks that facilitate the DSM; and (3) delivery of the growth potential of the European Digital Economy.

A.  Pillar I: Better Access for Business and Consumers The first point of focus is access to goods and services online, tackling the existing cross-border barriers and differences between offline and online trade. These measures included new rules to ban unjustified geo-blocking, further harmonisation of contract rules and cross-border cooperation among consumer protection authorities, new rules on parcel delivery, reforms of the copyright regime and broadcasting services and a simplification of VAT rules when selling cross-border. Assessing the rules after completion of negotiations, the balance of activity has arguably been on the side of improving consumer law, rather than on business access. The problem-based approach – ie, modifying existing rules where possible, creating new rules only where necessary – could be seen as perpetuating a patchwork of rules, rather than creating a more coherent regulatory landscape.

i.  Sales of Goods and Digital Content As regards sales of goods, the Commission proposal initially focused exclusively on online sales,31 but the negotiations showed that separating the offline and online

29 Juncker, above n 8. 30 Communication from the Commission, ‘A Digital Single Market Strategy for Europe’ COM(2015) 192 final. 31 European Commission’s ‘Proposal for a directive concerning contracts for the supply of digital content’ COM(2015) 634; and ‘Proposal for a directive on certain aspects concerning contracts for the online and other distance sales of goods’ COM(2015) 635.

‘Tinkering or Fundamental Overhaul?’  239 regulatory frameworks could be seen as artificial. The resulting rules32 achieve a targeted full harmonisation33 of consumer rights as regards rules on conformity, legal guarantee period, amongst others, and deal with problems arising in areas such as the Internet of Things (digital embedded content in physical goods) or ‘free’ services where the user ‘pays’ with personal data.

ii.  Better Consumer Protection Users need to be confident that their rights as consumers are equally protected online as offline. With that in mind, the DSM included a strong consumer angle which started with the adaptation of rules governing cooperation among consumer protection authorities to the online world (CPC Regulation),34 allowing national authorities to remove sites or social media accounts where scams have been identified or to trace the identity of rogue online traders. The Commission also issued guidance on the Unfair Commercial Practices Directive,35 then subsequently (in a proposal which came late in the Commission’s mandate) it proposed a New Deal for Consumers36 which, among other elements,37 will further strengthen consumer rights online through transparency obligations for online marketplaces and search tools.38

iii. Geo-Blocking39 This flagship initiative dealt with a typical source of consumers’ frustrations within the single market: when access to websites in another Member 32 Directive (EU) 2019/770 and Directive (EU) 2019/771, above n 9. 33 While both directives are based on the principle of maximum harmonisation, some room is foreseen for Member States to go beyond the Directive, in particular as regards specific remedies, liability claims against a third party, the consequences of a failure to supply digital content or a digital service due to force majeure, rights of parties to withhold the performance or to extend the application of the Directive to other contracts or to non-consumers (such as governmental organisations, start-ups or SMEs). 34 Regulation (EU) 2017/2394 of 12 December 2017 on cooperation between national authorities responsible for the enforcement of consumer protection laws and repealing Regulation (EC) No 2006/2004 [2017] OJ L345/1. 35 Commission Staff Working Document Guidance on the implementation/application of Directive 2005/29/EC on Unfair Commercial Practices (SWD/2016/0163 final), accompanying the Communication from the Commission on ‘A comprehensive approach to stimulating cross-border e-Commerce for Europe’s citizens and businesses’ COM(2016) 320 final. 36 The ‘New Deal for Consumers’, accompanied with a Commission’s Communication, included two different legislative proposals: for a Directive amending Council Directive 93/13/EEC of 5 April 1993, Directive 98/6/EC, Directive 2005/29/EC and Directive 2011/83/EU as regards better enforcement and modernisation of EU consumer protection rules, and for a Directive on representative actions for the protection of the collective interests of consumers, and repealing Directive 2009/22/EC. 37 The New Deal also proposes effective penalties for violations of EU consumer law (to not less than 4% cent of the trader’s annual turnover), or a proposal on representative actions for the protection of the collective interests of consumers. 38 As regards the first the proposal has been adopted: Directive No 2019/2161 [2019] OJ L328/7; the work on the second – Representative Actions – was completed on 22 June 2020 (not yet published). 39 Regulation (EU) 2018/302, above, n 13.

240  Claire Bury and Irene Roche Laguna State is blocked, or the user is treated differently depending on his or her IP address (geo-discrimination). The scope of the measure as adopted is limited to preventing geo-blocking and geo-discriminaton as regards online sales of goods, certain electronically supplied services and services provided in a physical location, together with non-discrimination with regard to payment methods. In view of the expectations of consumers, the measure has been criticised for not covering audiovisual services and/or copyright-protected content. As part of the final compromise, the co-legislators request the Commission to undertake a review of the Regulation by 23 March 2020 to analyse the need to expand its scope.40

iv. Portability41 While the Geo-blocking Regulation does not cover audiovisual, copyrightprotected content, the Commission proposed a targeted initiative to broaden users’ access to online content services in the EU: now, Europeans can fully use their online subscriptions to films, sports events, eBooks, video games and music services when travelling within the European Union in the same way they access them at home. This has become even more important since 15 June 2017, when the new roaming rules entered into force42 and the possibility for consumers to have cheaper access to data as part of their mobile subscriptions.

v.  Broadcasting Transmissions43 The third instrument whose purpose is to dismantle existing and new barriers to the cross-border access to online content and services is the so-called reform of the Satellite and Cable Directive.44 However, the Commission proposal, which sought to apply the country of origin principle to all programmes to assist in clearing rights that broadcasters may need if they wish to offer their broadcasts on their ancillary online services (mainly simulcasting and catch-up services), was considerably watered down during the negotiation phase. The proposed Regulation was 40 Art 9 of the Geo-blocking Regulation (‘review clause’) establishes that the first evaluation report (due to be presented by 23 March 2020, but not yet available) ‘shall be carried out, in particular, with a view to assessing the scope of this Regulation, as well as the extent of the prohibition laid down in point (b) of Art 4(1) and whether this Regulation should also apply to electronically supplied services the main feature of which is the provision of access to and use of copyright protected works or other protected subject matter, including the selling of copyright protected works or protected subject matter in an intangible form, provided that the trader has the requisite rights for the relevant territories’. 41 Regulation (EU) 2017/1128, above n 12. 42 Regulation (EU) 2017/920 of 17 May 2017 amending Regulation (EU) No 531/2012 as regards rules for wholesale roaming markets [2017] OJ L147/1. 43 Directive (EU) 2019/789, above n 24. 44 Council Directive 93/83/EEC of 27 September 1993 on the coordination of certain rules concerning copyright and rights related to copyright applicable to satellite broadcasting and cable retransmission [1993] OJ L248/15.

‘Tinkering or Fundamental Overhaul?’  241 transformed into a Directive and its effective application delayed (after a long transposition deadline).45 In terms of scope, it was limited to radio and television programmes which are news and current affairs or fully financed own productions of the broadcasting organisation. Television broadcasts of sport events are excluded from the scope of application of the country of origin principle. The status quo has been further safeguarded: existing contracts will remain unaffected for a period of four years from the entry into force of the Directive and the extension of the two additional types of TV programme will only be reviewed six years after the entry into force.

vi. Copyright46 All the political attention around this initiative has focused on the new press publishers’ neighbouring right (Article 15) and the so-called ‘value gap’ (Article 17, formerly known and heavily debated as Article 13), but it contains other key elements to adapt traditional copyright rules to the digital era. Cross-border access to data and content will be facilitated by new exceptions on text and data mining for research and educational purposes, on preservation of cultural heritage or on use of out-of-commerce works; furthermore, new transparency rules will allow for better determination of fair compensation in current and future value chains. User-generated content enters for the first time as a legal concept that merits protection under the new copyright rules. As regards Article 17, it makes an online content-sharing service provider liable for an act of communication to the public when it gives the public access to copyright-protected works uploaded by its users. However, the provider might be exempted from liability if it demonstrates that it made best efforts to license the content and ensure the unavailability of copyrightprotected content, without however impeding the availability of legal content already authorised or protected by exceptions and limitations. This Article should not lead to general monitoring obligations. When transposing the Directive, Member States will have to resolve this legal ‘Sudoku’ without going beyond or falling short of respecting this provision. The next Commission will certainly need to assist Member States (and stakeholders) in striking a difficult balance.

B.  Pillar II: Digital Networks and Services The second pillar of the DSM focused on the need to create trustworthy and reliable networks and services, while encouraging both innovation and competition. 45 While the Commission proposal provided an application deadline of six months after publication, the final Directive establishes a deadline of 24 months for Member States to transpose the new rules into their national law. 46 Directive (EU) 2019/790 of 17 April 2019 on copyright and related rights in the Digital Single Market and amending Directives 96/9/EC and 2001/29/EC [2019] OJ L130/92.

242  Claire Bury and Irene Roche Laguna New developments such as cloud computing, big data tools and the Internet of Things require innovations in the electronic communications sector, together with privacy rules and high levels of data protection. On the other hand, these new developments (and new actors) had raised concerns regarding an unlevel playing field that required corrections: incumbent actors were subject to existing obligations, while new actors or business models were capturing the users’ attention, slowly becoming substitutes.

i.  The Electronic Communications Code (EECC ‘the Code’)47 The last major update of the European telecom rules was in 2009, and the rules needed upgrading to be fit for the digital era. The Code was built on the embers of the so-called Connected Continent (or Telecoms Single Market (TSM)) proposal presented in the dying days of the Barroso II Commission. The Commission’s original TSM proposal was scaled back and limited mainly to rules on net neutrality and roaming. The Code can be viewed as a phoenix rising from the ashes of the TSM proposal. The Code codifies existing directives from 2002 (Framework, access, universal service, authorisation) which liberalised the sector and increased competition. Certain rules are extended to OTT providers, and it introduces a new – functional – definition of what constitutes an electronic communication service. The Code introduces a new connectivity objective in addition to the current objectives of promotion of competition, development of the internal market and promotion of the interests of the citizens of the Union. The new objective is articulated in terms of outcome: promote connectivity and access to, and take-up of, very highcapacity networks, including fixed, mobile and wireless networks, by all citizens and businesses in the Union. More specifically, the new rules incentivise investment in very high-capacity broadband networks and support 5G deployment in a competitive environment. The Code also has a strong consumer angle: it ensures that all citizens have access to affordable communication services, including adequate broadband access. It increases consumer protection and security for users and facilitates regulatory intervention. Furthermore, it introduces more harmonisation of rules on public warnings (‘reverse 112 system’) to citizens by text message in case of imminent or developing emergencies and disasters (by June 2022). The accompanying BEREC Regulation caps intra-EU calls and SMS from 15 May 2019, introducing an important deliverable for the final user.48

47 Directive (EU) 2018/1972, above n 11. 48 Regulation (EU) 2018/1971 of 11 December 2018 establishing the Body of European Regulators for Electronic Communications (BEREC) and the Agency for Support for BEREC (BEREC Office), amending Regulation (EU) 2015/2120 and repealing Regulation (EC) No 1211/2009 [2018] OJ L321/1.

‘Tinkering or Fundamental Overhaul?’  243

ii.  Online Platforms It is often said that platforms live comfortably in an unregulated environment, but the truth is that the EU institutions have been very busy applying existing rules and creating new ones for online platforms. By 2015,49 the Commission centred its work on two main areas: the liability of online intermediaries and the central ‘gatekeeper’ role that platforms were increasingly playing, in particular as regards business users. As regards the first of these areas, the Commission opted to maintain the E-Commerce Directive,50 in force since 2001, after a public consultation that concluded that its rules were still fit for purpose, but needed clarification. The Directive harmonises the exemption of liability for online services that host third-party content, to the extent that they do not have actual knowledge of their existence and, as soon as they are made aware, take expeditious action to disable access to it (commonly known as ‘notice-and-action’ procedures). Furthermore, the Directive prohibits Member States from imposing on such services general obligations to monitor the content they host. This Directive has been considered crucial for the development of online services in Europe, and the risk was high of opening long and difficult negotiations with uncertain results. Instead, the Commission opted for a sectoral approach (in copyright and AVMSD) and soft law: proposing an interpretation of the rules in the current environment through a Communication and later a Recommendation, and conducting dialogues with the biggest platforms to tackle illegal content online. However, this approach did not diminish the pressure on the liability regime. Platforms’ own efforts to tackle illegal content were often considered insufficient and offered unequal results. After exploring cooperation paths through the EU Internet Forum,51 the Commission proposed a Regulation on terrorist content online52 to channel the cooperation between enforcement authorities and online services. Other initiatives (Goods’ package,53 explosives precursors,54 the new Deal)55 have imposed different levels of ‘duties of care’56 on online marketplaces. Looking back, 49 Communication from the Commission, ‘A Digital Single Market Strategy for Europe’, above n 30. 50 Directive 2000/31/EC of 8 June 2000 on certain legal aspects of information society services, in particular electronic commerce, in the Internal Market [2001] OJ L178/1 (E-commerce Directive). 51 Launched on 3 December 2015 and bringing together governments, Europol and technology companies to counter terrorist content (see: europa.eu/rapid/press-release_IP-15-6243_en.htm). 52 ‘Proposal for a Regulation on preventing the dissemination of terrorist content online – A contribution from the European Commission to the Leaders’ meeting in Salzburg on 19–20 September 2018’ COM(2018) 640 final. 53 Regulation (EU) 2019/515 of 19 March 2019 on the mutual recognition of goods lawfully marketed in another Member State and repealing Regulation (EC) No 764/2008 [2019] OJ L91/1. 54 Regulation (EU) 2019/1148 of 20 June 2019 on the marketing and use of explosives precursors, amending Regulation (EC) No 1907/2006 and repealing Regulation (EU) No 98/2013 [2019] OJ L186/1 (not yet published). 55 See above, n 36. 56 A classical tort law concept, ‘duty of care’ has been used in this context as a legal obligation of means imposed on platforms due to the intermediary role played by them, but that has to be limited to what can be reasonably expected from them, without imposing a general monitoring obligation (see also Recital 48 of the E-commerce Directive).

244  Claire Bury and Irene Roche Laguna we could ask ourselves whether the Commission should have been more ambitious and tried to maintain a horizontal regime that ensures both a clean and a free internet for users and a horizontal and predictable legal framework for service providers. The next Commission will probably have to face the important challenge of modernising these rules and bringing consistency among the different sectoral solutions adopted so far, without changing the very nature of the internet ecosystem. Second, as regards the ‘love–hate’ relationship between intermediation services offered by platforms and the businesses that use them to scale and have access to markets, the EU has adopted a Regulation57 which introduces transparency requirements and aims to create a fair relationship, based on a balanced application of contractual terms and quick and accessible dispute settlement mechanisms. It is important to stress that these principles and obligations solely protect business users, while private users will be subject to consumer protection rules.58 Together with the proposed Regulation, the Commission set up an Observatory on the platform economy to increase its understanding of these business models and their impact on the single market.59

iii.  Audiovisual and Media Services Since 2007, when the AVMSD was adopted,60 the media landscape had shifted dramatically. While ‘classical’ broadcasters and video-on-demand providers were subject to strict rules on advertising and promotion of European works, the consumer shifted towards other sources of consumption online (in particular video-sharing platforms, such as YouTube, and some social media). The new rules61 extend the objectives of the AVMSD (in particular on protection of users against hate speech and minors against harmful content, including terrorist content and on advertising) to these new business models, while fully acknowledging that these platforms do not play an editorial role in the content they offer. The rules applying to platforms tread a fine line, imposing certain obligations62

57 Regulation (EU) 2019/1150, above, n 10. 58 The ‘New Deal for consumers’ (see above, n 36) modifies Directive 2005/29/EU and 2011/83/EU to introduce transparency requirements with a view to inform consumers using online marketplaces about the main parameters determining ranking of offers, and whether they enter into a contract with a trader or a non-trader (such as another consumer). 59 Commission Decision of 26 April 2018 on setting up the Group of experts for the Observatory on the Online Platform Economy, C(2018) 2393 final. 60 Directive 2007/65/EC of 11 December 2007 amending Council Directive 89/552/EEC on the coordination of certain provisions laid down by law, regulation or administrative action in Member States concerning the pursuit of television broadcasting activities [2007] OJ L332/27. This Directive, which entered into force on 19 December 2007, substantially amended the ‘Television without Frontiers Directive’ and renamed it ‘Audiovisual Media Services Directive’ (AVMSD). 61 Directive (EU) 2018/1808, above n 19. 62 Following Art 28b, video-sharing platform providers need to take appropriate measures to protect minors from harmful content and the general public from hate speech or terrorist content. To do so,

‘Tinkering or Fundamental Overhaul?’  245 and allowing for implementation by co-regulation, while ensuring compatibility with the existing limitation of liability for intermediary services. Furthermore, the cornerstone of the regulation of audiovisual and media services – and information society services – is maintained: the country of origin. Finally, it strengthens the obligations of video-on-demand services (such as Netflix) to promote European works. The new Directive also introduces requirements concerning independence of national regulators.

iv.  Trust and Security Trust and security in digital services were not a prominent part of the original roadmap of the DSM, but the Commission needed to adapt quickly to existing cyber threats and other risks to democracy, such as organised disinformation. As work progressed on the DSM and as a ‘spin-off ’ of the Directive on security of network and information systems (NIS Directive), in April 2019 the EU adopted a European ICT security framework through the EU Cybersecurity Act,63 and reviewed the organisation of ENISA, which increased powers in particular as regards the creation of the first voluntary EU cybersecurity certification framework for ICT products. During the dying days of its mandate, the Juncker Commission had to react to measures adopted by the United States against the Chinese manufacturer Huawei from supplying 5G mobile networks: the Chinese government could allegedly use the telecommunications giant for espionage and could present a potentially grave national security risk. In order to ensure a high level of cybersecurity of 5G networks across the EU, the Commission recommended to Member States a set of operational steps and measures and called for coordination of the national risk assessments and responses to this potential threat.64 Following the adoption of an Action Plan,65 the Commission set up a Code of Practice on Disinformation bringing together commitments by platforms, leading social networks, advertisers and the advertising industry. Beyond disinformation, the European elections in 2019 saw a joint effort by the Commission and Member States to coordinate to ensure the security of the election process.66

the Directive specifies a series of appropriate measures, such as adapting and enforcing their terms and conditions, establishing flagging mechanisms for users or parental control systems (among others). 63 Regulation (EU) 2019/881 of 17 April 2019 on ENISA (the European Union Agency for Cybersecurity) and on information and communications technology cybersecurity certification and repealing Regulation (EU) No 526/2013 (Cybersecurity Act) [2019] OJ L151/15. 64 Commission Recommendation of 26 March 2019 on Cybersecurity of 5G networks, C(2019) 2335 final. 65 Joint Communication of 5 December 2018, Action Plan against Disinformation, JOIN(2018) 36 final. 66 ‘Communication from the Commission of 12 September 2018 on Securing free and fair European Elections’ COM(2018) 637 final.

246  Claire Bury and Irene Roche Laguna

v.  GDPR and e-Privacy The GDPR67 entered into application in 2018, and is already showing its effects through fines and procedures opened against big platforms and social networks. At the same time in January 2019 the Commission proposed a revision of the e-Privacy Directive68 through a Regulation69 that adapts its scope in line with the new telecom rules, which extends the scope of certain obligations to OTT services. Furthermore, among other adaptations of the existing rules (for instance on spam), the so-called ‘cookie provision’, which has resulted in an overload of consent requests for internet users, would be streamlined and more user-friendly. However, this proposal has been the subject of a huge debate between the co-legislators and the object of intense lobbying. Nonetheless the rules on confidentiality of communications need to be upgraded so the co-legislators will at some point have to decide on a valid solution for this area, in line with the GDPR standards. Finally, a high level of data protection and privacy should not be at odds with the public need for ensuring a speedy and effective use of electronic evidence in national criminal investigations. With this in mind, the Commission proposed new rules which will make it easier and faster for police and judicial authorities to access the electronic evidence they need in investigations to catch and convict criminals and terrorists.70 The negotiations are running in parallel with international negotiations with the US and within the framework of the Council of Europe on cross-border access to electronic evidence.71

C.  Pillar III: Maximising the Growth Potential of the Digital Economy Through this pillar, the Commission set the ambitious goal of increasing investments on digital infrastructure, hardware and software, applications and data,

67 Regulation (EU) 2016/679 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 (General Data Protection Regulation) [2016] OJ L119/1. 68 Directive 2002/58/EC of 12 July 2002 concerning the processing of personal data and the protection of privacy in the electronic communications sector [2002] OJ L201/37 (Directive on privacy and electronic communications). 69 ‘Proposal for a Regulation concerning the respect for private life and the protection of personal data in electronic communications and repealing Directive 2002/58/EC (Regulation on Privacy and Electronic Communications)’ COM(2017) 10 final. 70 ‘Proposal for a Regulation of the European Parliament and the Council on European Production and Preservation Orders for electronic evidence in criminal matters’ COM(2018) 225 final; and ‘Proposal for a Directive laying down harmonised rules on the appointment of legal representatives for the purpose of gathering evidence in criminal proceedings’ COM(2018) 226 final. 71 The European Commission proposed. on 5 February 2019. to start international negotiations on cross-border access to electronic evidence, by means of two sets of negotiating directives, one for negotiations with the United States and one for the Second Additional Protocol to the Council of Europe ‘Budapest’ Convention on Cybercrime.

‘Tinkering or Fundamental Overhaul?’  247 digitising existing sectors. This would offer unprecedented opportunities in sectors such as transport, health, energy or urban planning. Fulfilling the potential of the digital economy depends on digitisation of industrial sectors for goods and services. But it also depends above all on the availability of data to develop new services and to develop and deploy technologies such as artificial intelligence. The Commission set out the challenges connected to the use of and access to data in its Communications on the Data Economy. Two legislative initiatives were intended to advance the EU’s efforts towards becoming a data economy.

i.  Free Flow of (Non-Personal) Data The first of these concerned the free flow of data (FFD).72 The Commission recognised the essential role of data as a resource for economic growth, job creation and societal growth – and therefore that restrictions on the free movement of data should be abolished, as a sort of a ‘fifth fundamental freedom’ of the EU single market, alongside workers, goods, services and capital. While the free movement of personal data was ensured through the GDPR, there were no rules for non-personal data or mixed data sets. Such rules are fundamental, as innovative applications and new digital services depend on availability of data for companies, customers and public authorities. For that reason, the Commission issued guidance (as mandated in the Regulation) on the interaction of this instrument with the GDPR, especially as regards data sets composed of both personal and non-personal data.73 Following the FFD Regulation, restrictions on the location of data for storage or processing purposes are banned, unless they are justified on grounds of public security in compliance with the principle of proportionality, as they constitute a barrier to free movement of data. Furthermore, the Regulation ensures availability of data stored abroad for competent authorities (similar to the US Cloud Act). Finally, it requires codes of conduct that ensure data portability for professional users.

ii.  PSI Directive The Review of the Public Sector Information Directive (PSI Directive):74 the now renamed Directive on ‘Open Data’ and the re-use of Public Sector Information introduces the concept of ‘High Value Data sets’, defined as documents the reuse

72 Regulation (EU) 2018/1807, above n 14. 73 ‘Communication from the Commission to the European Parliament and the Council – Guidance on the Regulation on a framework for the free flow of non-personal data in the European Union’ COM(2019) 250 final. 74 Directive (EU) 2019/1024 of 20 June 2019 on open data and the re-use of public sector information [2019] OJ L172/56.

248  Claire Bury and Irene Roche Laguna of which is associated with important benefits for the society and economy. They need to be available free of charge, in machine-readable formats, provided via Application Programming Interfaces (APIs) and, where relevant, as bulk download. However, during the negotiations this ambition was somewhat reduced, as the final text allows certain public sector bodies to delay the free availability of data.

iii.  Data Economy In addition to these legislative proposals, the Commission’s Data Economy Communication75 opened important workstreams for reflection in areas where legislative intervention was not yet justified, but where the Commission identified the need to gain further intelligence and improve dialogues among stakeholders. This is especially needed to avoid uncoordinated national regulatory interventions creating legal fragmentation that would undermine advances in cross-border data access. In particular, further thinking is needed in the following areas. Private data sharing in Business-to-Business (B2B) and Business-to-Government (B2G) situations: an increasing amount of data is often created in an automated way by machines or processes based on emerging technologies, such as the Internet of Things. The Commission suggests a number of draft principles for companies to consider when drafting relevant contracts, such as transparency, shared value creation, avoiding distortions of competition and data lock-in effects. Codes of conduct seem the best way, at this stage, to improve private data sharing, but more may be needed. As regards B2G, the Commission has created an Expert Group on access to and reuse of private sector data by public bodies for public interest purposes to explore best practices and self-regulatory solutions under a principlebased framework. Data ownership: who is the owner of machine-generated non-personal data sets? This question remains unresolved, and indeed one solution would be to grant a right to use and authorise the use of non-personal data to the ‘data producer’, ie, the owner or long-term user of the generating device. Such an approach would not be without controversy, and focusing on issues linked to access to data may deliver more speedy results. Liability: while rules on product liability at EU level are well developed, their application to the data economy in relation to IoT is not always self-evident as design errors and malfunctioning in the transmission of data are always possible, leading to damage in the offline world. Users and manufacturers of devices need legal certainty as to their potential liability. The Commission, after consultation with stakeholders, will prepare the ground for future action, if needed.

75 Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, ‘Towards a common European data space’ COM(2018) 232 final.

‘Tinkering or Fundamental Overhaul?’  249 Portability and interoperability: users buy different devices, or use different applications, depending on their needs and rapid apparition of improved systems. The GDPR already provides for the users’ right to ‘port’ their personal data from one system to the next. Such obligations do not exist for non-personal data, nor rules for the interoperability of different platforms. Under the FFD Regulation, the Commission encouraged codes of conduct to identify best practices in this regard. Finally, growth is also achieved through investment in technology developments. Together with a path to digitise European industry,76 the Commission has centred financial efforts on two different strands: on the one hand, channelling, coordinating and co-financing top-of-the-range exascale supercomputers77 through the European High-Performance Computing Joint Undertaking;78 on the other hand, the Commission proposed the creation of a European cybersecurity competence centre with a related network of national coordination centres.79

III.  The Future of the DSM: New Technological Developments and Possible Regulatory Gaps – What Next? The DSM is a moving target as much as the evolving technologies: despite the very ambitious agenda, the work is not yet accomplished. It is clearly for the next Commission to determine its priorities. However, the Commission’s experience during this mandate should inform that process. Some of the rules that have been adopted are designed to be just a first step and will be followed by further assessments of the need for additional rules. As explained above, the geo-blocking and P2B regulations are two such examples. In other cases, the Commission has opted for soft tools, but is closely monitoring developments – such as the collaborative economy, enforcement of intellectual property rights, hate speech online, protection of minors, disinformation. Codes of conduct with the platforms and the advertising industry in several of these areas have delivered some results. Final decisions on issues such as regulation of online political advertising and disinformation will depend on how effective the soft measures are judged to have been. 76 The Commission mobilised €500 million in a pan-EU network of digital innovation hubs (centres of excellence in technology) where businesses can obtain advice and test digital innovations, set up large-scale pilot projects to strengthen the IoT, advanced manufacturing and technologies and focus investments in EU’s public–private partnerships. 77 On 7 June 2019, the Commission announced eight sites in eight Member States to host worldclass supercomputers: Sofia (Bulgaria), Ostrava (Czechia), Kajaani (Finland), Bologna (Italy), Bissen (Luxembourg), Minho (Portugal), Maribor (Slovenia) and Barcelona (Spain). 78 Council Regulation (EU) 2018/1488 of 28 September 2018 establishing the European High Performance Computing Joint Undertaking [2018] OJ L252/1. 79 ‘Proposal for a Regulation establishing the European Cybersecurity Industrial, Technology and Research Competence Centre and the Network of National Coordination Centres’ COM(2018) 630 final.

250  Claire Bury and Irene Roche Laguna The predominance of business models based on free services but fuelled by advertising incentives (clickbait) coupled with extensive harvesting of personal data then deployed for targeting and manipulating users has led to considerable pressure for regulatory measures in response. Did the Juncker Commission underestimate the need for measures to enhance trust and transparency in online services? Or rather, did it react quickly to adapt its regulatory framework to new challenges? The next Commission might need to consider reopening the E-Commerce Directive. The ‘problem-based’ approach and the changes which have ensued through the inclusion of video-sharing platforms in the AVMSD, the Copyright Directive’s provisions on the value gap and the draft Regulation on the dissemination of terrorist content online have led to a complex framework to face this task. The fundamental controversial issue remains the liability regime and in particular the extent to which online intermediaries should be compelled to take proactive action and how to frame their ‘duties of care’ and liability. The Court of Justice will undoubtedly have a say on this, with three pending cases on the liability of social networks for defamation,80 video-sharing platforms81 and cloud services for copyright.82 The Court is likely to rule on these cases by end 2019 early 2020. There may also be a case for setting out a clearer framework for the governance of private content moderation to ensure that there is at least an independent review of the mechanisms which platforms put in place. The ‘sectoral approach’ to platforms has led to a number of definitions which are designed to capture, in one way or another, one or other service provided by the omnipresent platforms (see Annex I). It will continue to be important to focus on the particular service to be regulated, rather than targeting the provider of services. Experience has shown that online businesses evolve quickly and tend to provide services which can be very different in nature. For instance, a platform such as Facebook is now an information society service under the E-commerce Directive, an online content sharing service provider under the Copyright Directive, a provider of an intermediation service under the P2B Regulation, a video-sharing platform under the AVMSD, a hosting service provider under the Terrorist content proposal, an online marketplace under the New Deal, an online content service under the Portability Regulation; if Messenger is considered, it is even an electronic communication service under the Code. As regards platforms, competition law, data access issues and consumer protection should be addressed in an integrated way. Will the Commission blaze a trail in building a regulatory framework which enables business and society to meet the challenges of a ‘trustworthy’ Artificial Intelligence, as it has arguably done in the case of GDPR? In any event, the Commission should strive to set out a third way of doing policy which is humancentric, ethical and founded on respect for fundamental rights, distinct from both

80 Case

C-18/18 Glawischnig-Piesczek v Facebook ECLI:EU:C:2019:821. C-682/18 YouTube. 82 Case C-683/18 Elsevier. 81 Case

‘Tinkering or Fundamental Overhaul?’  251 a laissez-faire approach and a top-down controlled model. Shouldn’t it then also facilitate the more widespread use of decentralised ledger technology (Blockchain) by addressing the compatibility of that technology with existing regulatory requirements and the feasibility of smart contracts? This does not necessarily mean new rules are needed but clarity of the regulatory framework in this respect could be a competitive advantage in helping companies take appropriate risks and make investments in further technological development. The targeted modernisation of the Copyright acquis was one of the most challenging endeavours of the DSM project. Will the new Commission embark on further legislative action to tackle potential barriers built by still significant differences in copyright protection across Member States? Geo-blocking rules will have to be reviewed, to assess to what extent they can apply to copyright protected content, and in this context the Commission will need at least to reflect on the territorial limitations of copyright licences. Meanwhile, the Multi-Annual Financial Framework (MFF)83 recognises the need for the EU to invest more in digital technologies such as Artificial Intelligence, Super and Quantum Computing and state-of-the-art cybersecurity. The EU needs to develop its strategic resilience and autonomy in these fields. Last but not least, the proof of the DSM will be in its effects on the ground. After such intense legislative activity, the Commission needs to put on its hat of ‘guardian of the Treaties’. This means intensive work to ensure full implementation and proactive engagement and cooperation with Member States to ensure that these rules translate into rights for businesses and consumers in their daily lives. This might also be the moment to evaluate existing achievements and consider developing a digital ‘refit’ of all existing single market legislation to test that it is fit for purpose in the digital age.

IV. Conclusion ‘Europe will not be made all at once, or according to a single plan. It will be built through concrete achievements which first create a de facto solidarity’. Thus spoke the Schuman declaration: its purpose was to create an incipient single market for steel and coal under one common High Authority that confronted Germany and France with a de facto solidarity. Alfonso Mattera, one of the ‘fathers’ of the single market, also understood the single market as an instrument of peace. The same applies for the Digital Single Market, which is a translation of these de facto solidarities into the online world. More than ever in the online world, where information and data know no borders and can spread in a millisecond, Member States must continue working together to face common challenges. In so doing they should keep in mind that economic growth and technological development are not objectives per se, but must serve the wellbeing of EU citizens and the pursuit of their common values. 83 On 2 May 2018, the Commission proposed a long-term budget for the period from 2021 to 2027, not yet adopted.

252  Claire Bury and Irene Roche Laguna It can be said with confidence that the Digital Single Market Strategy was ambitious and has succeeded in covering a very broad part of the digital economy. This chapter seeks to demonstrate that, despite fast and furious activity by the Commission – using its monopoly of initiative – and by the co-legislators, the Digital Single Market is a moving target. Twenty years ago there was no Facebook, Google was born only in 1998; 10 years ago, only experts knew about ‘artificial intelligence’; until 2018 no one had heard of ‘deepfakes’. Technology is evolving incredibly fast, and the law can only follow and seek to keep pace with developments. After five intense years, the Commission can look back and proudly declare that it has achieved much, but it certainly cannot sit back and relax. As section III above shows, several outstanding challenges are already clear. No doubt during the next legislative period, it will require considerable ingenuity on the part of the Commission, together with a continued strong political commitment of the co-legislators, to ensure that the EU remains at the forefront of regulating in the online word.

Annex I: Definitions of Service Providers under the Different DSM Initiatives Legal instrument

Definition

Directive (EU) 2019/790 (DSM Copyright Directive)

‘online content-sharing service provider’ means a provider of an information society service of which the main or one of the main purposes is to store and give the public access to a large amount of copyright-protected works or other protected subject matter uploaded by its users, which it organises and promotes for profit-making purposes.

New Deal for Consumers (proposal)

‘online marketplace’ means a service which allows consumers to conclude distance contracts with other traders or consumers using software, including a website, part of a website or an application that is operated by or on behalf of the trader.

Regulation (EU) 2019/1148 on the marketing and use of explosives precursors

‘online marketplace’ means a provider of an intermediary service that allows economic operators on the one side, and members of the general public, professional users, or other economic operators, on the other side, to conclude transactions regarding regulated explosives precursors via online sales or service contracts, either on the online marketplace’s website or on an economic operator’s website that uses computing services provided by the online marketplace. (continued)

‘Tinkering or Fundamental Overhaul?’  253 Annex I  (Continued) Legal instrument

Definition

Regulation (EU) 2019/1150 ‘online intermediation services’ means services which on promoting fairness and meet all of the following requirements: transparency for business (a) they constitute information society services within users of online intermediation the meaning of point (b) of Article 1(1) of Directive services (EU) 2015/1535 of the European Parliament and of the Council1; (b) they allow business users to offer goods or services to consumers, with a view to facilitating the initiating of direct transactions between those business users and consumers, irrespective of where those transactions are ultimately concluded; (c) they are provided to business users on the basis of contractual relationships between the provider of those services and business users which offer goods or services to consumers. Proposal for a Regulation on Terrorist Content Online

‘hosting service provider’ means a provider of information society services consisting in the storage of information provided by and at the request of the content provider and in making the information stored available to third parties.

Directive (EU) 2018/1808 on Audiovisual and Media Services

‘video-sharing platform service’ means a service as defined by Articles 56 and 57 of the Treaty on the Functioning of the European Union, where the principal purpose of the service or of a dissociable section thereof or an essential functionality of the service is devoted to providing programmes, user-generated videos, or both, to the general public, for which the video-sharing platform provider does not have editorial responsibility, in order to inform, entertain or educate, by means of electronic communications networks within the meaning of point (a) of Article 2 of Directive 2002/21/EC and the organisation of which is determined by the video-sharing platform provider, including by automatic means or algorithms in particular by displaying, tagging and sequencing. (continued)

254  Claire Bury and Irene Roche Laguna Annex I  (Continued) Legal instrument

Definition

Regulation (EU) 2017/1128 on portability of online content

‘online content service’ means a service as defined in Articles 56 and 57 TFEU that a provider lawfully provides to subscribers in their Member State of residence on agreed terms and online, which is portable and which is (i) an audiovisual media service as defined in point (a) of Article 1 of Directive 2010/13/EU, or (ii) a service the main feature of which is the provision of access to, and the use of, works, other protected subject matter or transmissions of broadcasting organisations, whether in a linear or an on-demand manner.

E-evidence (proposal)

‘service provider’ means any natural or legal person that provides one or more of the following categories of services: (a) electronic communications service as defined in Article 2(4) of [Directive establishing the European Electronic Communications Code]; (b) information society services as defined in point (b) of Article 1(1) of Directive (EU) 2015/1535 of the European Parliament and of the Council for which the storage of data is a defining component of the service provided to the user, including social networks, online marketplaces facilitating transactions between their users, and other hosting service providers; (c) internet domain name and IP numbering services such as IP address providers, domain name registries, domain name registrars and related privacy and proxy services. (continued)

‘Tinkering or Fundamental Overhaul?’  255 Annex I  (Continued) Legal instrument Digital tax (proposal)

Definition ‘digital services’ means services which are delivered over the internet or an electronic network and the nature of which renders their supply essentially automated and involving minimal human intervention, and impossible to ensure in the absence of information technology, including in particular: (a) the supply of digitised products generally, including software and changes to or upgrades of software; (b) services providing or supporting a business or personal presence on an electronic network such as a website or a webpage; (c) services automatically generated from a computer via the internet or an electronic network, in response to specific data input by the recipient; (d) the transfer for consideration of the right to put goods or services up for sale on an internet site operating as an online market on which potential buyers make their bids by an automated procedure and on which the parties are notified of a sale by electronic mail automatically generated from a computer; (e) Internet Service Packages (ISP) of information in which the telecommunications component forms an ancillary and subordinate part, in other words packages going beyond mere internet access and including other elements such as content pages giving access to news, weather or travel reports, playgrounds, website hosting, access to online debates or any other similar elements; (f) the services listed in Annex II.

Annex II: DSM Initiatives Initiative

Publication OJ

Entry into force

Entry into application

Pillar I – Better Access for Business and Consumers Regulation (EU) 2017/2394 (CPC)

27 December 2017

16 January 2018

17 January 2020

Regulation 2018/302 (Geo-blocking)

2 March 2018

22 March 2018

3 December 2018

Directive (EU) 2017/2455 (VAT for supplies of services and distance sales of goods)

29 December 2017

18 January 2018

(several)

(continued)

256  Claire Bury and Irene Roche Laguna Annex II  (Continued) Initiative

Publication OJ

Entry into force

Entry into application

Regulation (EU) 2017/2454 (VAT – administrative cooperation)

29 December 2017

18 January 2018

1 January 2021

Regulation (EU) 2018/644 (Parcel delivery)

2 May 2018

22 May 2018

22 May 2018

Directive (EU) 2018/1713 (VAT e-books)

14 November 2018

4 December 2018

4 December 2018

Directive (EU) 2019/770 (digital content contracts)

22 May 2019

11 June 2019

1 July 2021

Directive (EU) 2019/771 (sale of goods)

22 May 2019

11 June 2019

1 July 2021

Regulation (EU) 2017/1128 (Portability of online content)

30 June 2017

20 July 2017

20 March 2018

Regulation (EU) 2017/1563 (Marrakesh Treaty)

20 September 2017

10 October 2017

12 October 2018

Directive (EU) 2017/1564 (Marrakesh Treaty)

20 September 2017

10 October 2017

12 October 2018

Directive (EU) 2019/789 (Broadcasting)

17 May 2019

6 June 2019

7 June 2021

Directive (EU) 2019/790 (DSM Copyright)

17 May 2019

6 June 2019

7 June 2021

Regulation (EU) 2018/1724 Single Digital Gateway

21 November 2018

12 December 2018 12 December 2020/2022/2023

Pillar II – Digital Networks and services Decision (EU) 2017/899 (700 MHz)

25 May 2017

14 June 2017

30 June 2020 (continued)

‘Tinkering or Fundamental Overhaul?’  257 Annex II  (Continued) Initiative

Publication OJ

Entry into force

Entry into application

Regulation (EU) 2017/920 (Wholesale roaming)

9 June 2017

12 June 2017

15 June 2017

Regulation (EU) 2017/1953 (Wifi4EU)

1 November 2017

4 November 2017

4 November 2017

Directive (EU) 2018/1972 (EECC Code)

18 December 2018

21 December 2018 21 December 2020

Regulation (EU) 2018/1972 (BEREC)

18 December 2018

21 December 2018 Direct/15 May 2019

Directive (EU) 2018/1808 (AVMSD)

28 November 2018

18 December 2018 19 September 2020

Regulation (EU) 2019/1150 (P2B)

11 July 2019

31 July 2019

Regulation (EU) 2018/1725 (Data protection for EU Institutions)

21 November 2018

12 December 2018 Direct/12 December 2019

Regulation (EU) 2019/517 (.eu)

29 March 2018

18 April 2018

13 October 2022

Regulation (EU) 2019/881 (Cybersecurity Act)

7 June 2019

27 June 2019

Direct/27 June 2021

Directive (EU) 2019/713 (non-cash payments)

10 May 2019

30 May 2019

31 May 2021

11 July 2020

e-Privacy European Cybersecurity Competence Network Pillar III – Maximising the growth potential of the Digital Economy Regulation (EU) 28 November 2018 2018/1807 (Free flow of non-personal data)

18 December 2018 28 May 2019

(continued)

258  Claire Bury and Irene Roche Laguna Annex II  (Continued) Initiative

Publication OJ

Entry into force

Entry into application

Council Regulation (EU) 2018/1488 (High-performance computing)

28 September 2018

8 October 2018

Direct

Open Data Directive (EU) 2019/1024 (PSI)

26 June 2019

16 July 2019

16 July 2021

part iv Critical Reflections on the Internal Market and its Future

260

12 The Fundamental Question of Minimum or Maximum Harmonisation STEPHEN WEATHERILL

I. Introduction This chapter focuses on the choice between maximum and minimum harmonisation in the internal market. At its heart lie two simple questions. First, what space, if any, is allowed for national regulatory choices in areas where the EU has adopted secondary legislation? Second, what space, if any, should be allowed for national regulatory choices in areas where the EU has adopted secondary legislation? These are simple questions – but the answers are not so simple. The search for answers reveals ambiguities about the legal, political and economic character of the internal market, while also providing a basis for reflection on the priorities that exists among the EU’s several objectives. Maximum harmonisation is better suited to impose common standards as the foundation of the internal market. It promotes uniformity and centralisation. Minimum harmonisation is better suited to promote regulatory diversity and to protect local autonomy. Neither model is right or wrong: the choice depends on what one wishes to achieve. Both models of harmonisation have a role to play within the EU, but this chapter argues for a case-by-case assessment of which is the more appropriate in the particular circumstances which create a demand for EU legislative action.

II.  The Treaty Superstructure The areas in which competence is shared between the EU and the Member States are listed in Article 4(2) of the Treaty on the Functioning of the European Union (TFEU). Article 6 TFEU sets out the areas in which the EU shall support, coordinate or supplement the actions of its Member States. The broad intent is plain: both EU and Member States have a role to play in these areas. The more specific

262  Stephen Weatherill questions of design are more awkward. Exactly what should be the relationship between EU action in these fields and action by the Member States? What is the effect of the adoption of EU rules on residual national competence to act in the relevant area? Most pertinent of all to the enquiry pursued in this chapter, may a Member State add stricter or otherwise more wide-ranging rules on top of the agreed EU rule? Where the matter falls within the scope of exclusive EU competence, as provided for by Articles 2(1) and 3 TFEU, such questions are readily answered by pointing to the exclusion of State competence to act, but where the matter falls within the scope of shared or supporting competence the luxury of simplicity is not available. Some Treaty provisions equipping the EU with a legislative competence are strikingly more helpful and explicit than others. They make plain in one way or another that EU legislative measures adopted pursuant to them do not exclude State competence to act in a more interventionist manner. Several of the Treaty provisions governing supporting competences provide examples. According to Article 6(c) TFEU the EU enjoys a supporting competence in the matter of culture. This is amplified in Article 167 TFEU, which limits action by the Union to ‘encouraging cooperation between Member States and, if necessary, supporting and supplementing their action’ in defined areas, and confines legislative action to the adoption of incentive measures by the Council and Parliament, which expressly excludes any harmonisation of laws, and to Council recommendations. This reveals that the relevant Treaty provision, which was first added to the Treaty superstructure with effect from 1993, was written with conscious care in order to preserve a high level of autonomy for Member State action in the field of culture even where the EU has acted. The EU’s role is similarly carefully restricted in its competence to address sport. Article 6(e) TFEU lists (inter alia) sport as a supporting competence enjoyed by the EU, and the amplifying Article 165 TFEU confines action to incentive measures adopted by the Parliament and Council, excluding any harmonisation of laws, and to Council recommendations. These provisions were added only in 2009, and they betray a readiness to equip the EU with an explicit competence in the field of sport for the first time, but on terms that are drawn carefully and narrowly, and which do not displace the primary role of the public authorities in the Member States and the governing bodies which run sport. Areas of shared competence commonly grant the EU the possibility of a higher profile and stronger legislative presence than is envisaged in the case of supporting competences. However, they are still marked by drafting which aims carefully to assert the limits of EU action, and consequently to preserve areas of State autonomy. Consumer protection offers a good example. It was added as an explicit legislative competence with effect from 1993 and it is today listed as a shared competence in Article 4(2)(f) TFEU. Article 169(2)(b) TFEU adds that in order to promote the interests of consumers and to ensure a high level of consumer protection, the Council and Parliament may adopt measures which ‘support, supplement and monitor the policy pursued by the Member States’. Article 169(4) TFEU adds that

Minimum or Maximum Harmonisation  263 such measures ‘shall not prevent any Member State from maintaining or introducing more stringent protective measures’, although it is added that such measures must be compatible with the Treaties, which means, inter alia, they must not be unjustified barriers to inter-State trade, and that the Commission shall be notified of them. Article 168 TFEU contains a similar but not identical set of provisions which govern public health, including an explicit exclusion of harmonisation of laws in its fifth paragraph. Some Treaty provisions grant the EU legislative powers which are written with less aggressive restraint, but which still show an overt concern to preserve space for stricter national rule-making even after the EU has made its legislative choice. Environmental protection, first introduced as an explicit EU legislative competence with effect from 1987, is a shared competence. The scope of permitted EU legislative action is defined relatively broadly by Articles 191 and 192 TFEU. These provisions are strikingly more generous and flexible in their grant of legislative competence in the field of environmental protection than Articles 168 (public health) and 169 (consumer protection). Article 193 TFEU adds that ‘The protective measures adopted pursuant to Article 192 shall not prevent any Member State from maintaining or introducing more stringent protective measures’. This is to identify the EU as a maker of uniform rules which also serve as minimum rules, in the sense that they grant flexibility to the Member States to adopt a higher level of protection than that mandated by the EU. Article 193 adds, in congruence with the case of consumer protection under Article 169, that such stricter national measures must be compatible with the Treaties,1 and that they shall be notified to the Commission. A similar pattern is visible in the matter of social policy. This is a competence shared between the EU and the Member States, and Article 153(4) TFEU provides that EU legislative intervention ‘shall not prevent any Member State from maintaining or introducing more stringent protective measures compatible with the Treaties’. Member States may not apply rules which are lower than the standard stipulated by the EU measure but they may choose to set rules that apply a higher standard. The EU sets the floor but not the ceiling. The model varies in detail, sector-by-sector, legal base by legal base. It has emerged incrementally, as successive revisions of the Treaty have added new legislative competences to the lists now found in Articles 2–6 TFEU but typically on terms that reveal the nervousness of the Member States about the consequences of reallocating regulatory competence. So the EU is empowered – but only so far, and only according to provisions that (in different ways) protect and preserve space for national choices. Norbert Reich spotted the trend as long ago as 1992: ‘the more competences the [Union] is acquiring, the less exclusive will be its jurisdiction’.2 1 On the notion of more stringent measures of environmental protection, see, eg, Case C-6/03 Deponiezweckverband Eiterköpfe EU:C:2005:222; and Case C-100/08 Commission v Belgium EU:C:2009:537. 2 N Reich, ‘Competition between Legal Orders: A New Paradigm of EC Law’ (1992) 29 Common Market Law Review 861, 895.

264  Stephen Weatherill We might today say that these provisions are concrete applications of the general notion of subsidiarity. Locking the entire EU in to a single agreed standard from which any deviation is forbidden (except by revision of the EU legislation itself) is vulnerable to the criticism that its admitted contribution to uniformity and certainty comes at a heavy cost. The appeal of a model of rule-making which does not involve a complete transfer of regulatory responsibility from State to EU level once the EU has legislated is that it preserves room for national choices, thereby to encourage learning, regulatory emulation and innovation. Rule-making which confines the EU to a status as a minimum rule-maker enhances flexibility, pluralism and multilevel governance. The question, however, is how this applies to the internal market, if at all.

III.  Harmonisation in the Internal Market Article 4(2)(a) lists the internal market as an area of shared competence. It is Article 114 TFEU which is the key Treaty provision dealing with the harmonisation of national laws in order to create and maintain an internal market in the EU. Article 114 is fundamentally different from the carefully written legal bases governing matters such as consumer protection, public health and environmental protection considered above. Article 114 is not defined by reference to any particular sector. Article 114 is instead defined by the functional imperative of pursuit of an integrated trading space stretching across the territories of all the Member States. Any national rule may be subjected to the discipline of legislative harmonisation pursuant to Article 114 whatever its shape, form, colour or intent provided only that such harmonisation makes a true contribution to the good functioning of the internal market3 – excepting only those particularly sensitive matters of policy set out in Article 114’s second paragraph. The current question, however, is not the scope of Article 114 but rather its effect on residual national competence to act in areas that have been subjected to EU-level harmonisation. And here too Article 114 differs from the sector-specific shared and supporting competences addressed above. Article 114 is less openly generous to national competence to act after the EU has acted. Article 114 does not explicitly allow Member States to adopt stricter rules than the agreed EU rule. It contains only a limited procedure found in Article 114(4)–(9), which permits Member States to apply stricter standards than are mandated by the agreed EU rule, but only under defined and carefully limited preconditions and only after securing Commission approval. Article 114 (4)–(9) offers a safety valve to a Member State wanting to insist on a higher standard than that agreed at EU level, but it is a far more tightly 3 eg, in particular Case C-376/98 Germany v Parliament and Council EU:C:2000:544; Cases C-154/04 and C-155/04 Alliance for Natural Health EU:C:2005:449; Case C-58/08 Vodafone, O2 et al v Secretary of State EU:C:2010:321; Case C-380/03 Germany v Parliament and Council EU:C:2006:772.

Minimum or Maximum Harmonisation  265 controlled concession than that typically found in the sector-specific Treaty provisions considered above, and consequently it has been relied on only rarely. Article 114 attaches a higher value to regulatory uniformity than do the other legal bases. This, one might assume, is dictated by the needs of the internal market – a common market needs common rules. So this brings us to the simple question which lies at the heart of this chapter. Is there any scope for individual measures of legislative harmonisation to offer more generous safety valves than Articles 114(4)–(9) allow? Put another way, minimum rule-making is not explicitly envisaged by Article 114, but is it nonetheless an available model in designing secondary legislation adopted pursuant to it? The question is whether a measure of legislative harmonisation adopted under Article 114 may itself provide that stricter national measures are permitted. Inspection of legislative practice yields a simple answer – yes! But this needs fuller elaboration, because the picture is mixed, as explained in the next section, and moreover it is contested both constitutionally and in policy terms, as explained in the two sections that follow the next. Only once this enquiry is complete may attention turn to assessing policy and competing priorities, within which an argument will be sustained to the effect that both maximum and minimum harmonisation have a place in the development of the EU’s internal market, but that preconceptions about which is better which are divorced from contextspecific assessment should be avoided.

IV.  The Practice of Legislative Harmonisation A.  Terminology and Description Legislative harmonisation, as a legal technique, involves two distinct issues. First, identification of the material scope covered by the EU measure and, second and logically subsequent, the effect on residual national competence of the EU measure within that defined material scope. It is, to use, conventional jargon, about first determining the scope of the occupied field and then, second, determining the extent of pre-emption within that occupied field. The phenomenon has, admittedly, attracted a proliferation of descriptors in EU academic writing and judicial practice. In particular, ‘full’ and ‘exhaustive’ harmonisation are sometimes used to describe a measure’s comprehensive coverage of a particular matter, an issue of material scope, and at other times to describe the measure’s exclusion of any scope for different national rules, which is a matter associated with pre-emption.4 4 See AG Szpunar in Case C-105/17 Kamenova EU:C:2018:378, paras 33–35. More fully, eg, J Bauerschmidt, ‘Die Sperrwirkung in Europarecht’ (2014) 49 Europarecht 277; R Schütze, From Dual to Co-Operative Federalism: the Changing Structure of European Law (Oxford, Oxford University Press, 2009); A Arena, ‘Exercise of EU Competences and Pre-Emption of Member States’ Powers in the Internal and the External Sphere: Towards Grand Unification?’ (2016) 35 Yearbook of European Law 28.

266  Stephen Weatherill However, provided one is attentive to this terminological trap and focuses instead on the distinct concepts, they are capable of being, and should be, kept separate. Identifying the material scope of a measure of EU harmonisation is commonly awkward, and may require the assistance of the Court in so far as the legislative text is (deliberately or by accident) imprecise. Many of these rulings tend to be detailed and of no particular significance beyond the particular legislative text at stake5 though on occasion more high-profile disputes pop up, such as whether Uber’s activities fall within the material scope of EU secondary legislation.6 But it is the second question, that which asks how much scope remains for national regulatory choices within the material scope properly defined, which is the focus of this chapter. And I shall describe a measure which asserts the EU as the exclusive site of rule-making within the material scope as maximum harmonisation, while, by contrast, EU measures which permit scope for Member States to prefer stricter rules above the agreed EU norm will be called minimum measures. Here too sometimes the intended effect of legislative harmonisation is clear, sometimes the matter requires judicial elucidation.7 But the basic models available are not difficult to grasp.

B.  Maximum Harmonisation …. A measure of maximum harmonisation requires Member States to permit products or services that comply with the harmonised standards to be marketed on their territory without any restriction and, as a corollary, they are also required to prohibit the marketing of products or services that do not so comply. It means, in particular, that a Member State may not have recourse to Article 36 TFEU (where goods markets are concerned) to justify measures that are stricter than those foreseen by the EU measure. Over the years the Court has frequently had occasion to insist on this point.8 The EU, by legislating, has ‘occupied the field’: it has 5 eg, Case C-242/17 LEGO SpA EU:C:2018:804 (Directive 2009/28 on promotion of use of energy from renewable sources); Case C-35/13 Assica EU:C:2014:306 (Regulation 2081/92 on protection of geographical indications); Case C-484/09 Ferreira Santos EU:C:2011:158 (Directives 72/166, 84/5 and 90/232 on insurance against civil liability in respect of the use of motor vehicles; Case C-215/17 Nova Kreditna Banka Maribor EU:C:2018:901 (Directive 2003/98 on re-use of public sector information). 6 In Case C-434/15 Uber Systems Spain EU:C:2017:981 and Case C-320/16 Uber France EU:C:2018:221 the Court placed Uber’s activities beyond the material scope of Directives 2000/31 and 2006/123, with the consequence that national regulation subject to EU free movement law prevails. 7 eg, Case C-128/94 Hans Hoenig v Stockach EU:C:1995:341; Case 150/88 Provide EU:C1989:594; Case C-169/89 Gourmetterie Van den Bergh EU:C:1990:227; Case 60/86 Commission v UK EU:C:1988:382; Joined Cases C-281/03 and C-282/03 Cindu Chemicals BV et al EU:C:2005:549; Case C-374/05 Gintec International Import-Export GmbH EU:C:2007:654. 8 eg, Case 35/76 Simmenthal EU:C:1976:180; Case 190/87 Oberkreisdirektor v Moormann EU:C:1988:424; Case 60/86 Commission v UK, above n 7; Case 148/78 Ratti EU:C:1979:110; Case C-374/05 Gintec International, above n 7; Case C-44/01 Pippig Augenoptik v Hartlauer EU:C:2003:205; Case 246/80 Broekmeulen v Huisarts Registratie Commissie EU:C:1981:218.

Minimum or Maximum Harmonisation  267 ‘pre-empted’ national measures. The regime can be altered only from within the EU legislative process. In fact, at a constitutional level, what has happened is that a matter, the regulation of the internal market, which is in principle a competence shared between the EU and its Member State,9 has in the particular area covered by the relevant piece of secondary legislation been converted for the time being by legislative act into an area of exclusive EU competence, subject to the possibility that the EU might subsequently adjust its rules to release space once again for national choices. This is why Article 36 TFEU is disabled. Whatever interests might have motivated national rules in the areas subjected to harmonisation – public health, public morality and so on – are now dealt with exclusively by the EU. Such maximum harmonisation is commonplace in fixing standards governing composition in goods markets. So, to take an example that is gloriously trivial yet entirely typical of this foundation stone of the internal market, adding milk precludes use of the description ‘egg liqueur’, because the relevant EU rule, Regulation 110/2008, regulates the matter exhaustively and milk is not on the list of permitted ingredients of such a beverage.10 For cosmetics in the internal market, Regulation 1223/2009 governs the matter through its Article 9 entitled Free movement.11 This stipulates that ‘Member States shall not, for reasons related to the requirements laid down in this Regulation, refuse, prohibit or restrict the making available on the market of cosmetic products which comply with the requirements of this Regulation’.12 Maximum harmonisation is not limited to product standards. Directive 2005/29 on unfair commercial practices in business-to-consumer relationships takes as its legal base Article 95 EC, which is today in amended form Article 114 TFEU.13 The Directive requires that, in short, fair practices be permitted throughout the EU and that unfair ones be prohibited, and it includes a common EU-wide definition of unfairness. Article 4 contains its so-called Internal market clause: ‘Member States shall neither restrict the freedom to provide services nor restrict the free movement of goods for reasons falling within the field approximated by this Directive’. There is no legislative ambiguity here and consequently the Court has had no difficulty in confirming that stricter national rules within the material scope of the Directive are not permitted.14 This in turn accentuates the importance of determining just what falls within its material scope, a matter which does involve ambiguity.15 The aim is a level regulatory playing field in the internal market within which legal and commercial certainty is enhanced while also ensuring that the regulatory concerns that previously provoked national rules are absorbed into and addressed 9 Art 4(2)(a) TFEU. 10 Regulation 110/2008 [2008] OJ L39/16, as applied in Case C-462/17 Tänzer and Trasper GmbH EU:C:2018:866. 11 [2009] OJ L342/59. 12 Previously Directive 76/768, which generated one of the Court’s early interpretative rulings insisting on the exclusion of stricter national rules within the material scope, Case 150/88 Provide, above n 7. 13 [2005] OJ L149/22. 14 eg, Case C-261/07 VTB-VAB NV EU:C:2009:244; Case C-206/11 Georg Köck EU:C:2013:14. 15 eg, Case C-391/12 RLvS Verlagsgesellschaft EU:C:2013:669.

268  Stephen Weatherill in a uniform way at EU level, and only at EU level. This is the virtue of maximum harmonisation.

C.  … But also Minimum Harmonisation Article 114 does not always generate maximum rules. As a matter of empirical observation the EU’s legislative acquis is home to a great deal of minimum harmonisation, where the EU harmonised standard sets the floor but it is open to each Member States to choose to raise the ceiling. At its most generous, this allows stricter rules subject only to the control exercised by primary law, most obviously the control exercised by free movement law and the prohibition against nationality-based discrimination. A more managed version would permit stricter rules but only in circumstances defined and limited by the EU measure itself. In appreciating the prevailing patterns of minimum harmonisation, it is important to realise that Directive 2005/29 on unfair commercial practices, mentioned above,16 is abnormal. Much of the harmonised acquis pertaining to the protection of the economic interests of consumers is built on an assumption that minimum harmonisation shall be the norm. So, for example, the constitutional basis for Directive 93/13 on Unfair Terms in Consumer Contracts is Article 100a, today in amended form Article 114 TFEU.17 The Directive is designed to replace diverse national techniques in the field with a common EU standard for testing the fairness of terms in consumer contracts that have not been individually negotiated. This is explained in the Directive’s Preamble as a means to improve the functioning of the internal market. However, it is expressly recognised by Article 8 of the Directive that ‘Member States may adopt or retain the most stringent provisions compatible with the Treaty in the area provided by this Directive, to ensure a maximum degree of protection for the consumer’. So, for example, the Directive’s Annex includes a ‘grey list’ of ‘terms which may be regarded as unfair’,18 but Article 8 of the Directive leaves room for the conversion of some or all of the ‘grey list’ into a ‘black list’ of prohibited terms, should a State so choose.19 Directive 99/44 on consumer sales and guarantees includes a functionally equivalent minimum proviso to the effect that Member States ‘may adopt or maintain in force more stringent provisions, compatible with the Treaty in the field covered by this Directive, to ensure a higher level of consumer protection’.20 So minimum directives do not set a single rule as both floor and ceiling, but rather only a floor. Member States may maintain stricter rules, up to the ceiling set 16 Above, n 13. 17 [1993] OJ L95/29. 18 Art 3(3), emphasis added. 19 eg Case C-119/15 Biuro podróży ‘Partner’ sp. z o.o. sp.k. v Dąbrowie Górniczej EU:C:2016:987: ‘It is apparent from the provisions of Dir 93/13, in particular Article 8 thereof, that the Member States may draw up lists of terms deemed to be unfair’ (para 36). 20 [1999] OJ L171/12.

Minimum or Maximum Harmonisation  269 by EU primary law. Like the safety valves found in Articles 114(4)–(9), this model reflects the notion that harmonisation need not involve the automatic subordination of national choices about market regulation to the dictates of trade liberalisation. But minimum harmonisation of this type is more flexible in the leeway it offers to States than the relatively tightly defined and Commission-supervised provisions of Articles 114(4)–(9). This legislative practice, manifest in Directives 93/13 and 99/44 and several others in the consumer acquis adopted pursuant to (what is today) Article 114 TFEU, aligns the harmonised consumer law acquis with practice in the sphere of environmental protection, where, as explained above, the Treaty base itself contains a constitutional direction that minimum rule-making is the EU norm. It also aligns the EU’s activities in the consumer field driven in the name of market-making harmonisation with the (in practice little used) consumerspecific Treaty base found in Article 169(4) TFEU which, also as mentioned above, locates in the Treaty an explicit preference for minimum rule-making by the EU. And in fact these analogies are still closer since the amendments made to both Directives 93/13 and 99/44 by Directive 2011/83,21 which added to both Directives a new provision, Article 8a in both instances, requiring that stricter provisions shall be notified to the Commission, which mirrors the management procedure envisaged under Articles 193 and 169 TFEU.22 A minimalist explanation for this convergence may be found in the suggestion that the areas in which minimum harmonisation has achieved a high profile are areas that are not really at the core of the integrative process. Allowing diversity in levels of consumer protection directed at economic interests and standards of environmental protection may have some incidental distortive effect on trade patterns, but such rules do not directly affect the composition of products where legislative pre-emption is the norm, subject only to the scarcely used Article 114(4)–(9) procedure. A connected explanation holds that some of the relevant EU legislative measures date from before the conferral of an explicit legal base governing environmental protection (which occurred in 1987 by virtue of the Single European Act) and consumer protection (1993, the Maastricht Treaty). The EU legislature borrowed the Treaty-conferred competence to harmonise laws in order to express a political preference for the development of a legislative programme of environmental and consumer protection at a time when the Treaty conferred no adequate competence in that field. In truth minimum harmonisation formed part of the political bargain that has characterised the growth of the harmonised legislative acquis affecting protection of the economic interests of consumers. Directives of this type have strained the outer limits of the principle of conferral, but the faint whiff of constitutional insincerity was ignored amid prevailing political support for the development of a consumer policy for the EU which also guaranteed to

21 [2011] OJ L304/64. 22 Are stricter provisions enforceable if left unnotified? Probably, but the matter is entertainingly arguable, cf Case C-194/94 CIA Security International SA EU:C:1996:172.

270  Stephen Weatherill Member States scope to maintain stricter protective rules under the minimum formula.23 A more positive account insists that minimum harmonisation has true virtue. By locating responsibility at both EU and Member State level even after the EU has legislated, it opens up space for regulatory experimentation, learning and development. Moreover, minimum harmonisation embraces diversity and differentiation as in themselves valuable within an EU dedicated to more than economic integration and inapt for reduction to a ‘one-size-fits-all’ model. There is ample evidence that consumer preferences, tastes and capabilities differ in Europe.24 Minimum harmonisation has the capacity to reflect the reality of legal and cultural heterogeneity between the Member States. Minimum harmonisation stands for regulatory pluralism in the internal market. The precise shape of that diversity will be determined by the extent to which Member States choose to add an extra layer of regulation, which will doubtless vary sector by sector, as States respond to competing pressures. In some respects the advantages of minimum harmonisation are the disadvantages of maximum harmonisation, and vice versa, but the underlying story is that the choice of model of harmonisation is not simply a technical matter of market-making, it is intensely political. One really would not expect defining egg liqueur to be treated in the same way as assessing the fairness of terms in consumer contracts. A narrower version of minimum harmonisation also exists. It is a model which sets a floor of EU rules, which Member States must implement and enforce, while also permitting stricter national rules, but only in circumstances defined and limited by the EU’s legislative measure. This is visible in the Services Directive, 2006/123.25 Its Chapter V, ‘Quality of Services’, includes in Articles 22–27 a set of requirements that must be met by providers which fall within the material scope of the Directive. But compliance with these terms does not guarantee access to markets in all the Member States. In the matter of establishment, the Directive’s Article 15, ‘Requirements to be Evaluated’, envisages that an array of listed requirements which restrict access to a service activity or the exercise of it may be applied by the regulating Member State on condition that they are non-discriminatory, justified by an overriding reason relating to the public interest, and proportionate. In the matter of services, Article 16(3) carves out scope for requirements justified for reasons of public policy, public security, public health or the protection of the environment. This, then, is a model which compares with the Directives affecting the economic interests of consumers which were noted above, but the room left to Member States is more constrained.

23 See S Weatherill, EU Consumer Law and Policy (Cheltenham, Elgar European Law, 2012) ch 3; S Weatherill, Contract Law of the Internal Market (Cambridge, Intersentia, 2016) chs 3, 6. 24 See especially (and influentially) T Wilhelmsson, ‘The abuse of the confident consumer as a justification for EC consumer law’ (2004) 27 Journal of Consumer Policy 317. 25 [2006] OJ L376/36.

Minimum or Maximum Harmonisation  271 However, the reason for excluding maximum harmonisation runs in common. It lies in the political salience of choosing between models of harmonisation. There is nothing mundanely technical about this. The original ‘Bolkestein draft’ of what would become the Services Directive proposed that compliance with the Directive’s stipulated quality standards should be of itself enough to guarantee access to the markets of all the Member States.26 This led to fierce political resistance, ‘Polish plumbers’, tortuous negotiation and the compromises found in the finally adopted and more nuanced Directive. Diminution in the measure’s material scope27 and the derogations contained in Articles 15 and 16, permitting stricter national rules, were both necessary precisely because of intense political disquiet about surrendering regulatory authority to the EU.28 It was a rejection of the political viability of maximum harmonisation in this sector.

D.  Does the EU Legislature Enjoy a Free Choice between Models? The plain fact is that the legislative record discloses instances of both minimum and maximum harmonisation. It appears to be a legislative choice, dictated by the context and subject matter. But – is this correct? And should it be correct? Legislative practice cannot override constitutional fundamentals, so a first question is whether this is consistent with the Court’s view of Article 114. Then the second question concerns the normative desirability of minimum harmonisation. This involves assessment of the attitude of the Commission. Ultimately I believe the model of minimum harmonisation does and should survive this twin-track critical scrutiny, but the position of the Court and the Commission now needs to be examined.

V.  Is there a Constitutional Objection to Minimum Harmonisation? As is so common in discussing the constitutional character and limits of legislative harmonisation, the enquiry leads to the Court’s decision in the first

26 COM(2004) 2. 27 See exclusions and limits in Arts 1(2)–(7), 2(2)–(3). This itself has – thematically – provoked litigation, eg, Case C-57/12 Femarbel EU:C:2013:517 (healthcare services); Joined Cases C-340/14 and C-341/14 Trijber and Marmsen EU:C:20156:641 (transport services); Case C-33/17 Čepelnik EU:C:2018:896 (labour law). 28 See, eg, J Loder, ‘The Lisbon Strategy and the politicization of EU policy-making: The case of the Services Directive’ (2011) 18 Journal of European Public Policy 566; J van de Gronden and H de Waele, ‘All’s well that bends well? The Constitutional Dimension to the Services Directive’ (2010) 6 European Constitutional Law Review 397.

272  Stephen Weatherill Tobacco Advertising case.29 As is well known, the Court in that case annulled Directive 98/43 as an improper exercise of the legislative competence to harmonise conferred (at the time) by Articles 57(2), 66 and 100a EC, which are today, after amendment, Articles 53(2), 62 and 114 TFEU. Directive 98/43 was a measure of legislative harmonisation affecting both product and services markets associated with the advertising of tobacco products but the Court concluded that – in short – the measure went too far and pursued public health objectives which were insufficiently associated with any demonstrated connection to the functioning of the internal market. The Court accepted that harmonisation of laws governing advertising of tobacco products in magazines, newspapers and periodicals would be valid as a contribution to improving the functioning of the market for press products30 – and this was exactly the model of the subsequently adopted and narrower replacement Directive 2003/33 which itself survived challenge before the Court in the second Tobacco Advertising case.31 But, as the Court put it in the first of the cases, the ‘prohibition of advertising on posters, parasols, ashtrays and other articles used in hotels, restaurants and cafés, and the prohibition of advertising spots in cinemas, prohibitions’ did not help to facilitate trade in such products.32 In similar vein it concluded that the Directive did not actually contribute to eliminating appreciable distortions of competition. Differences between national rules on tobacco advertising may give rise to appreciable distortions of competition, and the Court noted in particular the distortive consequences of variation between national rules governing sponsorship of sports events by tobacco companies, but here too it reached the conclusion that the Directive went too far. This condemned it as invalid. However, the Court added a further objection to the Directive. It ‘does not ensure free movement of products which are in conformity with its provisions’.33 Its Article 5 permitted Member States to apply stricter requirements concerning the advertising or sponsorship of tobacco products as they deem necessary to guarantee the health protection of individuals; and it contained no provision ensuring the free movement of products which conform to its provisions. The Court’s treatment of this matter stretches over only five paragraphs of the judgment34 and it is clearly not essential to the finding that the Directive is not valid. But these – unhappily underdeveloped – observations are capable of being read as constitutionally fatal to minimum harmonisation. The Court’s criticism of the Directive for failing to ensure free movement of goods and services which comply with its provisions appears to mean, turned round, that a harmonisation Directive must ensure free movement of products and services which comply with



29 Case

C-376/98 Germany v Parliament and Council, above n 3. 97–98. 31 Case C-380/03 Germany v Parliament and Council, above n 3. 32 Case C-376/98 Germany v Parliament and Council, above n 3, para 99. 33 ibid, para 101. 34 ibid, paras 101–05. 30 Paras

Minimum or Maximum Harmonisation  273 its provisions. A measure of minimum harmonisation cannot do that, because its very nature is to leave space for stricter (but not less strict) rules chosen in individual Member States. In the second Tobacco Advertising case the Court found the more tightly drawn replacement Directive 2003/33 to be valid, and part of its analysis included approval of the presence of a clause guaranteeing that complying products would be entitled to be marketed throughout the internal market.35 In BAT, which concerned another tobacco-related measure of harmonisation, Directive 2001/37 on the presentation and sale of tobacco products (dealing most conspicuously with health warnings), the Court similarly went out of its way to explicitly contrast the Directive with that at stake in the first Tobacco Advertising case on the basis that it, unlike the annulled Directive 98/43, contained a provision which guaranteed the free movement of products which comply with its requirements. This, the Court remarked, ensured the Directive exerts ‘its full effect in relation to its object of improving the conditions for the functioning of the internal market’.36 The assumption is that harmonisation must be maximum, not minimum. The matter resurfaced once again more recently in Philip Morris Brands.37 This concerned aspects of the validity of Directive 2014/40, the successor to Directive 2001/37 which had been at stake in BAT. The particular objection of present relevance was to Article 24(2) of the Directive. Article 24 is entitled ‘Free Movement’. Its first paragraph stipulates that Member States may not prohibit or restrict the placing on the market of tobacco or related products which comply with the Directive, but this is stated to be subject to paragraphs 2 and 3. Paragraph 2 directs that this shall not affect the right of a Member State to maintain or introduce further requirements, applicable to all products placed on its market, in relation to the standardisation of the packaging of tobacco products, where it is justified on grounds of public health, taking into account the high level of protection of human health achieved through this Directive.

The Court interpreted this to refer to room for further rules in areas that lie beyond the harmonised regime. It is, in short, treated as an expression of the limits of the material scope of the Directive, which covers most, but not all, of the matters associated with packaging. The Court excluded any possibility of an interpretation which would allow further rules on packaging within the scope of the harmonised regime. Such an understanding would, it remarked, undermine the harmonisation effected by the Directive and would render Article 24(2) incompatible with the Directive.38 The Court went so far as to contrast its view of the Directive



35 ibid,

paras 73–74. C-491/01 R v Secretary of State ex parte BAT and Imperial Tobacco EU:C:2002:741, para 74. 37 Case C-547/14 Philip Morris EU:C:2016:325. 38 ibid, paras 71–72. 36 Case

274  Stephen Weatherill with that taken in the first Tobacco Advertising case, because here, but not there, Member States may not prevent, on grounds relating to the aspects of packaging harmonised by the Directive, the marketing of products which comply with the requirements laid down by the Directive.39 The Court therefore appears to treat Article 114(4) et seq as the only permitted route to derogation from a harmonised standard.40 According to this approach ‘minimum harmonisation’ is not allowed: or at least it needs to be confined to circumstances where a State chooses to insist on compliance with stricter rules only in application to that State’s own suppliers operating within the domestic market. The problem is that another stream of the Court’s case law shows no such constitutional anxiety about the concept of minimum harmonisation. This is especially visible in cases concerning the harmonisation of measures that protect the economic interests of consumers, where the Court has not even hinted at objection to the model of ‘minimum harmonisation’. Buet is the landmark case in this vein.41 The Court ruled that although Directive 85/577 on ‘doorstep selling’ required only that a consumer be allowed a right to cancel the contract concluded on the doorstep within a defined period of time, a complete ban imposed by France on doorstep selling was not excluded by the Directive. The Directive was a minimum measure: it provided explicitly in Article 8 that the Directive ‘shall not prevent Member States from adopting or maintaining more favourable provisions to protect consumers’. In Buet the Court took this at face value. It did not question whether harmonisation could validly permit a State to demand more than compliance with the harmonised regime.42 The only control asserted over the French measure was that it must comply with the Treaty rules on free movement, and on this point the Court found the ban justified as a measure designed to protect vulnerable consumers. Buet predates the first Tobacco Advertising case. It is ignored in the first Tobacco Advertising case. In a clutch of rulings since the first Tobacco Advertising case, the Court has addressed measures of minimum harmonisation, commonly concerning the protection of the economic interests of consumers, and it has shown no suspicion of constitutional distaste for the chosen model. The Court follows the logic of Buet: if the Directive permits stricter rules, then the only control which is exercised over the national choice is that of primary law, most prominently free movement law and non-discrimination. So, for example, the Court has shown not a shred of anxiety about assessing national rules governing unfair terms in consumer contracts which are stricter than those mandated by Directive 93/13, happily referring to the escape route created by its Article 8. In Biuro podróży ‘Partner’ it noted that the Directive requires that 39 ibid, para 82. 40 ibid, para 71. 41 Case 382/87 Buet EU:C:1989:198. 42 For the same approach to the same Directive, see Case C-441/04 A-Punkt Schmuckhandels EU:C:2006:141.

Minimum or Maximum Harmonisation  275 terms falling within its scope be assessed to determine whether they are fair, but, citing Article 8, it expressed no objection to national practices that treat particular types of term as automatically unfair.43 In Jana Pereničová, Vladislav Perenič v SOS financ spol. s r. o. it ruled that a Member State may choose to provide that a contract concluded with a consumer by a trader which contains one or more unfair terms is to be void as a whole where that will ensure better protection of the consumer, even though that is not required under the Directive.44 The Court explained that Article 8 of Directive 93/13 ‘carried out only a partial and minimum harmonisation of national legislation concerning unfair terms, while allowing Member States the option of giving consumers a higher level of protection than that for which the directive provides’.45 Directive 93/13’s minimum formula seems to be entirely acceptable to the Court. One final confirmation is supplied by María Victoria González Sánchez v Medicina Asturiana.46 This judgment determines that the Product Liability Directive, 85/374, which is a measure of harmonisation adopted pursuant to what is now Article 115 TFEU, excludes the possibility for the maintenance of stricter national rules within its scope. But the Court explicitly contrasted the Directive with Directive 93/13’s express authorisation of more stringent provisions in matters in respect of which it makes provision.47 It suggested no anxiety about the validity of that clause. It is the same story elsewhere in the acquis. DocMorris concerned internet sales across borders of medicinal products for human use.48 Dutch traders faced restrictions in Germany (a long-standing and continuing problem).49 The Court found that the German intervention fell within the scope of Directive 97/7, which mandated particular formalities associated with distance selling to consumers, but it did not ban such marketing. Its Article 14 allowed Member States to ‘introduce or maintain, in the area covered by this Directive, more stringent provisions compatible with the Treaty, to ensure a higher level of consumer protection’, and it explicitly envisaged the possibility to ban distant selling of certain goods, ‘particularly medicinal products’. So compliance with the Directive does not mean automatic access to the regulated market. The Court found the German rules to be compatible with the Directive, provided they conformed to primary law (free movement) which took the Court into a careful examination of justification, involving a distinction between non-prescription and prescription medicines. As in Buet, there is no hint in the judgment that the Court finds anything constitutionally troubling in a measure of harmonisation which leaves room for 43 Case C-119/15 Biuro podróży ‘Partner’ sp. z o.o. sp.k, above n 19. 44 Case C-453/10 Jana Pereničová, Vladislav Perenič v SOS financ spol. s r. o. EU:C:2012:144. See similarly Case C-484/08 Caja de Ahorros y Monte de Piedad de Madrid v Asociación de Usuarios de Servicios Bancarios (Ausbanc) EU:C:2010:309. 45 Case C-453/10 Jana Pereničová, Vladislav Perenič, above n 44, para 34. 46 Case C-183/00 María Victoria González Sánchez v Medicina Asturiana EU:C:2002:255. 47 ibid, para 27. 48 Case C-322/01 DocMorris EU:C:2003:664. 49 cf Case C-148/15 Deutsche Parkinson Vereinigung EU:C:2016:776.

276  Stephen Weatherill stricter rules to be selected by Member States which will lead to obstacles to trade even where the terms of the Directive have been met. The Court in DocMorris expressly noted that the legislature had made this choice.50 Herbert Karner takes a similarly recepitve and wholly uncritical approach to Article 7 of Directive 84/450 on misleading advertising which states that the Directive is not to preclude Member States from retaining or adopting provisions with a view to ensuring more extensive protection.51 Stricter rules must comply with the rules of primary law and so fall to be checked against the demands of Article 34 TFEU in so far as they obstruct inter-State trade in goods, but they are not ruled out as a matter of principle. Directive 2005/60 on money laundering, a harmonisation measure affecting both goods and services markets, contains a standard minimum formula permitting Member States to apply stricter provisions in the field covered by the Directive.52 The dispute in Jyske Bank involved Jyske, a branch of the Danish bank NS Jyske Bank established in Gibraltar, and the authorities in Spain, where Jyske provided services and had been condemned to pay fines for breach of the applicable rules.53 The Court noted the existence of the minimum clause and showed no distaste for it. Safe Interenvios also concerns the Money Laundering Directive and takes the same uncritical approach to the minimum clause as Jyske Bank, which is cited in the judgment.54 In Rina Services the Court, asked to provide an interpretation of Directive 2006/123, the Services Directive, explicitly contrasted the provisions of the Directive which do not permit Member States to justify stricter rules from those that do so permit, without even hinting that any constitutional objection may be raised to the validity of this particular instance of (managed) minimum harmonisation.55 As noted above, the negotiation of the Services Directive involved a high level of political contestation and the derogations permitting stricter rules contained in the final version were integral to what the Member States and the European Parliament were prepared to tolerate. It would be astonishing were this to be ruled constitutionally impermissible: the EU would rightly be condemned for its inflexibility and incapacity to take into account the prevailing plurality of multilevel regulatory concerns. Some of the cases, such as those provoked by Directive 93/13, concern directives which are ‘pure’ minimum harmonisation, while others, such as Rina Services, concern a more limited version of minimum harmonisation in which Member State autonomy is preserved within tighter limits than simply by reference to the Treaty as a control. Some of the cases involve what one might term process rules – that is, rules which are stricter than the Directive but which do not prevent access 50 ibid, para 110. 51 Case C-71/02 Herbert Karner EU:C:2004:181. See also C-44/01 Pippig Augenoptik, above n 8. 52 [2005] OJ L309/15. 53 Case C-212/11 Jyske Bank EU:C:2013:270. 54 Case C-235/14 Safe Interenvios EU:C:2016:154. 55 Case C-593/13 Rina Services EU:C:2015:399. See similarly Joined Cases C-360/15 and C-31/16 X BV EU:C:2018:44.

Minimum or Maximum Harmonisation  277 to the regulated market but rather concern only the conditions that apply on it. That would be true of Directive 93/13 (and in the background would be scepticism that such consumer protection measures were ever mainly or at all about marketmaking anyway).56 But cases such as Doc Morris, Jyske Bank and Rina Services cannot be easily explained away on this basis: they involve directives which envisage stricter rules which keep out a complying product or service. What a mess! Some of these rulings are hard to read, especially on the point whether material scope or pre-emptive effect is being examined with a view to determining the space left for national measures. The sense that the Court is not fully aware of what it is doing is accentuated by close inspection of the ruling in Safe Intervenios. The italics are added, … [79] In providing that the Member States may adopt or retain in force stricter provisions in the field covered by the Money Laundering Directive to prevent money laundering and terrorist financing, Article 5 of that directive does not grant the Member States a power or obligation to legislate by virtue of EU law, but merely, unlike the provisions laid down in Chapter II of the directive, recognises the power which the Member States enjoy under national law to provide for such stricter provisions outside the framework of the regime established by the directive.57

… to show that the Court slips unhelpfully and apparently unwittingly from the matter of minimum rule-making to that of defining the measure’s material scope. The fundamental problem is that in none of the group of cases beginning with the first Tobacco Advertising case did the Court attempt to explain how and why its ruling could fit with the long-established acceptance in Buet of stricter rules above the harmonised norm. And in the cases of the other type, which are wholly receptive to the validity of the model of minimum harmonisation, the Court does not mention the first Tobacco Advertising case at all, nor does it seem the issue was even addressed. The best attempt to make sense of this mess is to take the view that what matters is the purpose and context of each individual EU measure. There should be no general rule opposing (or requiring) either model, maximum or minimum. It should depend on the object and purpose pursued by the particular measure under review. Normally the Court is attentive to this need for contextual review.58 The problem is that in Tobacco Advertising and related rulings its broader formula seems antagonistic to the very concept of minimum harmonisation. Given the accumulation of legislative practice and case law elsewhere which is wholly inconsistent with such a broad view, we need to adopt a narrower view of the case. So the first Tobacco Advertising case is best understood as a case in which the Court took the view that a market access clause was required to make constitutional sense of that particular Directive. It is an exaggeration to regard it as a ruling which rejects in principle the model of minimum harmonisation, even if one must confess

56 cf

above, n 23. 79. 58 cf above, n 7. 57 Para

278  Stephen Weatherill that it (and the subsequent case law culminating in Philip Morris) are written in exactly that exaggerated style. This understanding then allows cases such as Buet and Rina Services to be read as rooted in the nature and purpose of the particular EU measures that formed the dispute’s subject matter. As measures of minimum harmonisation they improved the regulatory environment of the internal market by setting a floor which replaced the previous completely unharmonised pattern, which eliminates the danger of a race to the bottom in sectors where this is plausible, while also promoting regulatory objectives including consumer protection which are recognised by EU law both at EU level (in the form of the minimum harmonised standards) and at national level (in the form of scope for selected stricter rules). The dual function of harmonisation is well understood: this is the dual function of minimum harmonisation! On this approach ‘minimum harmonisation’ is capable of being treated as valid even if the scope for stricter national rules which it assumes prevents it serving as a means to establish a perfectly even layer of regulation for the entire internal market. It is harmonisation – in the sense that all Member States must meet the standard mandated, and may not drop below it. It is floor, but not ceiling – unlike maximum harmonisation, which is both. But both are harmonisation. In Philip Morris the Court had to accept that, because the material scope of the Directive is not comprehensive, the Directive does not guarantee that products whose packaging complies with the requirements of the Directive may move freely on the internal market, but it treated this as the inevitable consequence of the method of harmonisation chosen by the EU legislature.59 But the measure, though partial in coverage, ‘undeniably offers advantages for the functioning of the internal market, since, whilst it does not eliminate all obstacles to trade, it does eliminate some’.60 It is, as the Court put in Swedish Match, a subsequent ruling also dealing with Article 24 of Directive 2014/40, ‘harmonisation in stages’.61 The same is true of minimum harmonisation! It is for the legislature to choose which model is more suitable in particular political and economic circumstances.

VI.  Is there a Policy Objection to Minimum Harmonisation? In a communication released in July 2001 the Commission launched a programme of inquiry into the proper role of the EU in the field of contract law.62 The trajectory of the Commission’s subsequent work has been twisted, and has led up dead ends and false options, but a key part of the Commission’s assumption throughout 59 Case C-547/14 Philip Morris, above n 37, paras 79–80. 60 ibid, para 81. 61 Case C-151/17 Swedish Match EU:C:2018:938, para 73. See also Case C-220/17 Planta TabakManufaktur EU:C:2019:76, para 44. 62 COM(2001) 398.

Minimum or Maximum Harmonisation  279 has been that the model of minimum harmonisation which has long underpinned the contract law acquis – mainly consumer contract law – is inapt to realise the full benefits of the internal market. It leads to fragmentation of the regulatory landscape in so far as Member States take the opportunity to apply stricter rules. The Commission was able to provide evidence for this.63 The inquiry quickly led the Commission to a high level of antipathy directed at the minimum model and to a strong preference for maximum rules, as one of the means to achieve the core aim, ‘a significantly higher degree of coherence in European contract law’.64 The Commission’s Consumer Policy Programme for 2002–06 captured the mood. It advocated ‘moving away from the present situation of different sets of rules in each Member State towards a more consistent environment for consumer protection across the EU’.65 This entailed a policy preference for maximum rather than minimum harmonisation. Its Green Paper on the ‘Review of the Consumer Acquis’ released in February 2007 concluded what the Commission described as the ‘diagnostic phase’ of the review of contract law, and, in forward looking mode, it set out the main options for reform.66 It was striking, though little emphasised in the document, that the orthodox pre-existing model of minimum harmonisation was not even listed as a possible option for the future. Maximum, rather than minimum, harmonisation was on the march. The Commission’s principal explanations for its policy preference were all devoted to the quest to improve the functioning of the internal market. Uniformity rather than diversity is the guiding vision of maximum harmonisation in the internal market. The Commission enjoyed successes in propelling its preferred design through the EU’s legislative process. Directive 2002/65 on distant selling of financial services was the first measure in the orthodox heartland of EU harmonisation activity affecting consumer contract law to reflect the Commission’s re-focused policy preference.67 The Directive applies a model of maximum harmonisation to the field covered by the Directive. Directive 2005/29 concerning unfair commercial practices, mentioned above,68 is probably the most high profile success of

63 COM(2004) 651. See also Commission Report on implementation of Directive 97/7 on distance contracts COM(2006) 514; Commission Communication on the implementation of Directive 99/44 COM(2007) 210 and, much more recently on Directive 99/44, the Study on the costs and benefits of minimum harmonization under the Consumer Sales and Guarantees Directive prepared by ICF and published by the Commission in March 2017, EUR 2017.2801. 64 D Staudenmayer, ‘The Place of Consumer Contract Law within the Process on European Contract Law’ (2004) 27 Journal of Consumer Policy 269, 277. See also D Staudenmayer, ‘European Contract Law – What Does It Mean and What Does It Not Mean?’ in S Vogenauer and S Weatherill (eds), The Harmonisation of European Contract Law: Implications for European Private Law, Business and Legal Practice (Oxford, Hart/IECL, 2006). Staudenmayer was a (highly impressive and able) Commission official closely involved in the project. 65 COM(2002) 208, para 3. 66 COM(2006) 744. 67 [2002] OJ L271/16. 68 Above, n 13.

280  Stephen Weatherill the maximum-minded Commission. It secured agreement on a measure which allocates to EU level exclusively the decision on whether a practice is fair (and therefore must be allowed) or unfair (and therefore must be suppressed). Directive 2005/29 leaves no scope for Member States to choose to establish stricter standards within the field occupied by the Directive.69 The whole field of market regulation which falls within the scope of Directive 2005/29 has been completely transferred to EU competence. Claims to prefer a higher standard of consumer protection may not even be advanced. Justifications for trade restrictions previously available under primary law to a regulating State pursuant to Articles 36, 52 and 62 TFEU are simply no longer on the table. The values that might previously have animated a claim to justify a national measure have been transplanted to the Directive alone and the survival of a practice which falls within the material scope of the Directive depends exclusively on measuring its conformity with the benchmark standard of fairness set by the Directive. This is the character of maximum harmonisation. It disables Member State competence to do anything other than apply the EU’s harmonised rules. The evenness of the resultant regulatory landscape is the point: this is a model driven ruthlessly by the desire to release the economic benefits of the internal market. But there are costs. The principal anxiety is that the maximum model, in contrast to the more flexible consequences that apply in a case of minimum harmonisation, dictates that the nature and level of (re-)regulatory protection within the internal market is determined exclusively by the EU. Dissatisfaction with the EU standard – whether on paper or as interpreted from time to time by the Court of Justice, or both – must be addressed by seeking to provoke legislative reform at EU level. This is a model of the internal market which is prone to rigidity and not receptive to local regulatory variation at all. The question was how far the Commission could push its policy preference in favour of maximum harmonisation before political resistance was encountered. Not long, it proved. The story of Directive 2011/83 on consumer rights illuminates the political sensitivity of this debate about models of harmonisation. In October 2008, as a result of its review of the consumer acquis, the Commission adopted a proposal for a Directive on Consumer Rights which would have replaced Directive 85/577 on doorstep selling, Directive 97/7 on distance selling, Directive 93/13 on unfair terms and Directive 99/44 on consumer sales with what it described as ‘horizontal instrument’.70 All those four Directives contained the ‘minimum formula’, but the proposal was to convert them into a single measure of maximum harmonisation, under a claim that this shift would deliver an improved, more ‘coherent’ legal infrastructure for the internal market. In the particular matter of doorstep selling, distance selling, unfair terms and consumer 69 Its Art 3(9) makes small (but politically salient) exceptions: in relation to financial services and immovable property, Member States may impose requirements which are more restrictive or prescriptive than the Directive. 70 COM(2008) 614.

Minimum or Maximum Harmonisation  281 sales, it would lock in the Member States to regulate in the way and at the level demanded by the EU’s ‘horizontal instrument’, with no scope for locally determined divergence. But the Commission’s 2008 proposal suffered heavy criticism, not least though not only for its commitment to the maximum model of harmonisation which, it was protested, caused costs in the shape of unduly suppressed national diversity.71 The political climate was unwelcoming too. Voices in the Parliament, especially those belonging to Socialist and Green MEPs, were largely hostile to maximum harmonisation on the basis that it risks upsetting established standards of protection at national level and shutting down any scope for reform at national level.72 National Parliaments objected too, using the reasoned opinion procedure and framing objections to the suppression of stricter national rules as hostile to the dictates of subsidiarity.73 This led to political compromise, which was given shape in late 2011 with the adoption of Directive 2011/83 on consumer rights.74 Four had become two. Directive 2011/83 is a measure of maximum harmonisation which excludes stricter national initiatives in the area which it covers.75 This is clear and explicit, as the Court has had opportunity to underline.76 However, Directive 2011/83’s material scope is only slightly larger than that which was previously covered by Directive 85/577 on doorstep selling and Directive 97/7 on distance contracts, the two Directives which are overhauled and repealed by Directive 2011/83. Directive 93/13 and Directive 99/44 were excluded from the final agreed text of Directive 2011/83, and they remain untouched as persisting measures of minimum harmonisation, subject only to the above-mentioned obligations imposed on Member States to notify the Commission where they elect to impose stricter rules.77 So both Directive 93/13 and Directive 99/44 withstood the Commission’s campaign to convert them from the model of minimum harmonisation to that of maximum harmonisation. The lesson holds that ‘coherence’ and the appeal to a better functioning internal market was not enough to win the day. Maximum harmonisation represents a vertical transfer of authority from Member States to the EU: it may well sustain a more competitive internal market but it does so at the expense of local autonomy, and the more politically sensitive and socially cherished the subject matter which is debated, the less likely is it that appeal to the economic benefits of the internal market will persuade Member States and the Parliament that the 71 See, eg, H-W Micklitz and N Reich, ‘Crónica de una Muerte Anunciada: the Commission Proposal for a Directive on Consumer Rights’ (2009) 46 Common Market Law Review 471; M Faure, ‘Towards a Maximum Harmonization of Consumer Contract Law?!?’ (2008) 15 Maastricht Journal 433; V Mak, ‘Review of the Consumer Acquis: Towards Maximum Harmonization?’ (2009) 17 European Review of Private Law 55. 72 S Weatherill, ‘The Consumer Rights Directive: how and why a quest for “coherence” has (largely) failed’ (2012) 49 Common Market Law Review 1279. 73 COM(2010) 547, 9. 74 Above, n 21. 75 Art 4. 76 eg, Case C-332/17 Starman AS EU:C:2018:721. 77 Arts 32, 33 Directive 2011/83, above n 21.

282  Stephen Weatherill constitutional and regulatory adjustment demanded by the maximum model is worth it. Directive 2011/83 was adopted only by placating the Member States and the Parliament by abandoning the attempt to persuade them to locate exclusively at EU level the competence to regulate unfair terms and consumer sales, just as the Services Directive 2006/123 could be adopted only by surrendering the ambition to achieve a complete transfer of regulatory responsibility from Member State to EU level. The debate about choosing between models of harmonisation will and should continue, but the Commission will be well advised to conduct it in relation to individual measures,78 with particular current relevance to its preoccupation with the development of the Digital Single Market. Any attempt to assert a general normative preference for maximum harmonisation is politically incautious. The sharply political context of choosing among models of harmonisation and divergent views about the nature of the internal market more generally is plainly visible. This is more than technical market-making! And so it is important that preconceptions about which model is ‘better’ should be avoided. Both types of harmonisation need to be available for selection by the EU legislature.

VII.  Policy and Priorities In making the choice to which this chapter is dedicated, there are two competing positions. Either … the concept of ‘minimum harmonisation’ is an oxymoron! That is: how can harmonisation possibly operate by setting minimum standards? Minimum EU standards – which permit Member States to set higher standards if they so choose – are incompatible with the very nature and purpose of harmonisation, which is to set common standards. The proper functioning of the EU’s internal market demands that harmonisation set standards which dictate how that market shall be regulated in the same way across the territory of all the Member States. Without that necessary regulatory uniformity, there is no common economic space. Regulatory disharmony destroys the common foundations of the internal market. So there can be no room allowed to Member States to set standards which depart from the EU’s harmonised norm. There can be no minimum harmonisation in the internal market. Or … the concept of ‘minimum harmonisation’ is a beautiful compromise! Minimum EU standards – which permit Member States to set higher standards if they so choose – are common standards, which contribute to the better functioning of the internal market by ensuring that all Member States regulate on the same platform. But, by serving as floor but not ceiling, they leave room for Member States to set stricter standards. This allows for diversity, founded on the common 78 cf Study on the costs and benefits of minimum harmonization under the Consumer Sales and Guarantees Directive, above n 63.

Minimum or Maximum Harmonisation  283 platform, which promotes regulatory experimentation and learning. Moreover, it ensures that the EU is not exclusively responsible for standards-setting. Instead Member States can improve on the choices made by the EU by enacting stronger standards of consumer protection, social policy or environmental protection – should they so wish. So minimum harmonisation accommodates both concerns for market-making and other regulatory values – all of which are part of the EU’s aims and objectives as expressed in Articles 2 and 3 TEU. At heart is a choice of priorities. Maximum harmonisation is better suited to impose common standards as the foundation of the internal market. Ruling out space for stricter national rules brings benefits consequent on the uniformity promoted at EU level, leading to a simplification of the regulatory landscape across the internal market. That is doubtless the ‘harmony’ seductively located within the discourse of harmonisation. It is an agenda of centralisation. A maximum model radically shifts the pattern of European law-making away from a cooperative model towards a more hierarchical pattern. Minimum harmonisation is better suited to promote regulatory diversity and a race to the top in protective standards. It is attuned to expression of local autonomy. It avoids the ‘paralysis of state action’ to which Advocate General Villalón drew attention in an opinion concerned with air transport and environmental law, but which contains astute observations of general application about the stifling peril of maximum harmonisation as a regulatory model.79 Neither perspective is right or wrong: the choice depends on what one wishes to achieve. The Commission’s review of contract law revealed much. Its aim to achieve ‘a significantly higher degree of coherence in European contract law’80 would be achieved at EU level under a model wedded to maximum harmonisation, but at national level the consequence would likely be much less coherence, because established rules that are found to fall within the material scope of the EU measure must be set aside unless they are precisely in line with the EU system (on paper and as interpreted by the Court), to the detriment of accumulated local preferences. The more ‘coherent’ the EU regime, the less coherent and the less autonomous national systems become. From this perspective there is a respectable case to be made against coherence at EU level – at least, one must appreciate that pursuit of coherence at EU level cannot possibly be cost-free. This is a lesson that extends far beyond contract law. It opens a window on a much deeper political and cultural enquiry into the preferred relationship between the EU and its Member States – how much should be done in common, how much diversity should be tolerated, even encouraged?81 The question is which model of harmonisation should prevail? The answer is – it depends. Some matters may be thought appropriate 79 Case C-120/10 European Air Transport SA EU:C:2011:94. The Court’s ruling (EU:C:2011:556) is much narrower in scope and ambition. 80 Staudenmayer, ‘The Place of Consumer Contract Law within the Process on European Contract Law’, above n 64. 81 See, eg, N Boeger, ‘Minimum harmonisation, free movement and proportionality’ in P Syrpis (ed), The Judiciary, the Legislature and the EU Internal Market (Cambridge, Cambridge University Press,

284  Stephen Weatherill for a single meaning across the EU – safety, probably; egg liqueur, possibly. Other matters, particularly of a more culturally or socially sensitive nature, may more readily resist the pull of maximum harmonisation on the basis that Europe should cherish and promote its historically determined diversity. Both the Services Directive and the Directive on Consumer Rights offer insights into the prevailing political sensitivities. Economic integration is and always has been at the forefront of EU activity, and it has been compellingly argued that the single market ‘remains the EU’s core business’,82 but there is much more to the EU’s mission than trade, as a glance at Articles 2 and 3 TEU reminds.

VIII. Conclusion Minimum harmonisation represents a realistic attempt to accommodate diverse national tradition and to promote scope for dynamic regulatory innovation within the process of integration. In the development of the regulatory framework that underpins the internal market there are constitutional commitments that favour minimum rule-making over maximum measures contained in EU primary law, most prominently in Articles 12, 114(3), 168(1) and 169(1) TFEU and Articles 35, 37 and 38 of the Charter. These several provisions assert the EU’s role as a proper source of regulatory protection, but it does not and should not follow that it should be the exclusive source. A model built not only on EU rulemaking but also on national level contributions above an EU minimum offers a compelling alternative. In fact the stakes are high enough to justify reliance on Article 4(2) TEU as a provision that favours minimum over maximum harmonisation, for only by foreclosing the pre-emptive effect of EU law is space left for the expression of national identity in its many diverse forms across the Member States. So there is a strong case to be made for regarding maximum harmonisation as applicable only exceptionally, where its use has been carefully justified in prevailing sector-specific conditions. This debate is of immense practical significance in the detail of law-making, but it also reveals much about the choices to be made about the vertical distribution of competences within the EU. Wider still, it illuminates the values of the EU, seeking to accommodate a balance between market-making uniformity and variety across a wide sweep of multifunctional activity.

2012); S Weatherill, ‘Maximum versus Minimum Harmonization: Choosing between Unity and Diversity in the Search for the Soul of the Internal Market’ in N Nic Shuibhne and L Gormley (eds), From Single Market to Economic Union: Essays in Memory of John A Usher (Oxford, Oxford University Press, 2012). 82 J Pelkmans, ‘Why the Single Market Remains the EU’s Core Business’ (2016) 39 West European Politics 1095.

13 From Supranationality to Managing Diversity: A (Re-)New(ed) Paradigm for the Establishment of the Internal Market? KAI P PURNHAGEN

I. Introduction EU law shall establish an internal market.1 While this mandate is undisputed, there is debate about the design of such an internal market. It is further disputed by which means such an internal market should be established. Debates concern questions such as how to find a balance between social and market provisions,2 the level (for example, maximum or minimum) and type (for example, mutual recognition or standardisation, positive or negative) of harmonisation,3 and which is the right balance of regulation between innovation and protection.4 This contribution looks into questions regarding the level of harmonisation and consequently the relationship between the EU, Member States and individuals in the EU. It will investigate first whether we can observe a shift in internal market policy moving from the establishment of a supranational institution to managing diversity.5 Second, if so, 1 Art 3(3) sentence 1 TEU. 2 S Garben, ‘The Constitutional (Im)balance between “the Market”’ and “the Social” in the European Union’ (2017) 13 European Constitutional Law Review 23. 3 See S Weatherill, ch 12 in this volume; W Kerber and R Van den Bergh, ‘Mutual Recognition Revisited: Misunderstandings, Inconsistencies, and a Suggested Reinterpretation’ (2008) 61 Kyklos 447; K Purnhagen and J Wesseler, ‘Maximum vs Minimum Harmonization: What to expect from the institutional and legal battles in the EU on Gene editing technologies?’ (2019) Pest Management Science, 2310 available at: onlinelibrary.wiley.com/doi/epdf/10.1002/ps.5367. 4 K Garnett, G Van Calster and L Reins, ‘Towards an innovation principle: an industry trump or shortening the odds on environmental protection?’ (2018) 10 Law, Innovation and Technology 1. 5 I have, together with A Afilalo and D Patterson, already made this claim, albeit in embryonic form, in A Afilalo, D Patterson and K Purnhagen, ‘Statecraft, the Market State and the Development of European Legal Culture’ in G Helleringer and K Purnhagen (eds), Towards a European Legal Culture (Munich, CH Beck, 2014) 279.

286  Kai P Purnhagen what are the consequences of such a shift for EU law? I acknowledge that this debate is not new, but had been discussed in similar ways before.6 However, the context has changed. On the one hand the body of EU law has meanwhile accumulated to an extent not comparable to previous years.7 On the other hand, Brexit and the rise of nationalism across the EU spark questions as to whether the relationship between EU law, Member State sovereignty and individuals needs to be evaluated afresh. Departing from the core of the internal market clause, which I understand to be the establishment of the internal market in Article 3(3) sentence 1 of the Treaty on European Union (TEU), this contribution first outlines the legal framework the Treaties provide for the question of the design of the internal market and which forms and values need to be integrated. Subsequently, it will assess selected areas of secondary legislation as well as court judgments to determine whether, within this legal framework, a switch can be observed to give preference to the management of diversity rather than to ‘one-size-fits-all’ harmonisation. Finally, consequences for EU law of such a switch will be outlined.

II.  The Requirement to Establish the Internal Market Article 3(3) sentence 1 TEU stipulates that ‘(t)he Union shall establish an internal market’. Articles 26(1) and 114(1) of the Treaty on the Functioning of the European Union (TFEU) make certain that such an internal market is to be achieved via ‘measures’, of which most are enumerated in Article 288 TFEU. Article 114(1) TFEU stipulates that the ‘approximation of the provisions laid down by law, regulation or administrative action in Member States’ would be the governing mode for the establishing of the internal market. According to which end these measures shall work, or in other words, how the internal market that is to be achieved shall look like is not clearly elaborated on in the Treaties. Article 28(2) TFEU famously stipulates that ‘(t)he internal market shall comprise an area without internal frontiers in which the free movement of goods, persons, services and capital is ensured in accordance with the provisions of the Treaties’. Beyond that there is debate about the goals this internal market aims to achieve,  the arguments brought forward are mostly depending on the political background of the respective scholar.8 The Treaties provide a catalogue of certain 6 W Kerber, ‘Interjurisdictional Competition within the European Union’ (2000) 23 Fordham International Law Journal 217; N Reich, ‘Competition between Legal Orders: A New Paradigm of EC Law’ (1992) 29 Common Market Law Review 861, 895. 7 By 1998 Majone had termed the body of EU law ‘impressive’, see G Majone, ‘Europe’s ‘Democratic Deficit’: The Question of Standards’ (1998) 4 European Law Journal 13. 8 See, eg, R Barents, ‘Constitutional Horse Trading: Some Comments on the Protocol on the Internal Market and Competition’ in M Bulterman, L Hancher, A McDonnell and H Sevenster (eds), Views of European Law from the Mountain (Alphen aan den Rijn, Kluwer Law International, 2009); M Bartl, ‘Internal market rationality: In the way of re‐imagining the future’ (2018) 24 European Law Journal 99.

From Supranationality to Managing Diversity  287 objectives, which to an uninformed observer look more like a ‘wish list’ from which the legislator or interpreter can choose from when determining the content of the internal market. Most prominently is the list stipulated in Article 3(3) sentences 2–5 TEU, which provides the legal interpreter and the law-maker alike with a menu to choose from when deciding on the means of legislation or interpretation. Additionally, Protocol (No 27) on the Internal Market and Competition adds that the activities necessary to achieve the objectives of the Community include ‘a system ensuring that competition in the internal market is not distorted’. When weighing these objectives, primary internal market regulation provides some help.9 First and foremost, Article 51 TEU stipulates that Protocols to the Treaties are an integral part of the Treaties. Hence Protocol (No 27) is applied at the same level as Article 3(3) TEU.10 Title II of the TFEU contains a catalogue of provisions which have general application. Some of them reiterate the objectives mentioned in Article 3(3), sentence 2–5, such as the principle of equal treatment of women in Article 8 TFEU.11 Others provide more flesh on the bones of these objectives, such as that the objective to combat social inclusion and discrimination (Article 3(3) 1st Alt TEU) may be understood as ‘to combat discrimination based on sex, racial or ethnic origin, religion or belief, disability, age or sexual orientation’ (Article 10 TFEU). Other provisions, such as Article 9 TFEU, qualify the respective objective, such as that a ‘high level’ of education needs to be achieved. Incidentally the TFEU thereby also prioritises objectives such as education over other objectives, as a prerequisite of ensuring a ‘high level’ may mean accepting a lower level of realisation of other objectives in case of conflict. In Test-Achats,12 the European Court of Justice (ECJ) decided upon the operationalisation of this relationship between the different and at times contradictory objectives in the Treaty in view of gender equality: while the general policy norm in Article 157(1) TFEU confines the EU legislator to progressively achieve equal treatment of men and women by making use of the competence norm in Article 19(1) TFEU, Article 3(3) TFEU and Article 8 TFEU confers on the EU legislator the right to ‘determine when it will take action, having regard to the development of economic and social conditions within the European Union’.13 Once it has decided to take action, however, the EU legislator ‘must contribute, in a coherent manner, to the achievement of the intended objective, without prejudice to the possibility of providing for transitional periods or derogations of limited scope’.14 Whether this judgment from the specific area of equal treatment can

9 See on the relationship between Art 3(3) TFEU and the provisions in Title II of the TFEU, Case C-236/09 Association belge des Consommateurs Test-Achats ASBL, Yann van Vugt, Charles Basselier v Conseil des ministres EU:C:2011:100, paras 19–21. 10 Case C-496/09 European Commission v Italian Republic EU:C:2011:740, para 60. 11 Case C-236/09 Association belge des Consommateurs Test-Achats, above n 9, para 19. 12 ibid, above n 9. 13 ibid, para 20. 14 ibid, para 21.

288  Kai P Purnhagen be extrapolated to other areas of internal market law is questionable,15 not least because other policies explicitly provide for provisions which allow for transitional measures16 and derogations of limited scope.17 Within the various areas of internal market law, such as the four free movement provisions, competition law, and agriculture, each of these objectives are also given special attention. Article 39 TFEU, for example, stipulates which objectives in which form are applicable to the Common Agricultural Policy (CAP). Despite the fact that the rules of the CAP have developed largely in topical isolation from other provisions of internal market law18 and frequent emphasis on their exceptional character,19 Article 38(1)2 TFEU highlights that the CAP is nonetheless part of the internal market.20 The Court has established that the EU legislature and institutions have a broad discretion in pursing the objectives of the CAP21 and no hierarchy between the CAP’s objectives exists.22 Another example where the Treaties clearly favour certain objectives over others is Article 114 TFEU. Article 114(3) TFEU binds the Commission to take in its proposal as a base a high level of protection concerning health, safety, environmental protection and consumer protection, thereby implicitly requiring the Commission to take precedence of these objectives over other internal market objectives. The Court has explicitly endorsed such a priority of public health over other objectives of internal

15 K Lenaerts, ‘The Court’s Outer and Inner Selves: Exploring the External and Internal Legitimacy of the European Court of Justice’ in M Adams, H de Waele, J Meeusen and G Straetmans (eds), Judging Europe’s Judges (Oxford, Hart Publishing, 2013) 26, also emphasised that the special reasoning in this case largely relies on the policy choices in the equal treatment provisions. 16 See, eg, Art 191(2) TFEU. 17 See, eg, Art 114(4–8) TFEU. 18 J Usher, EC Agricultural Law 2nd edn (Oxford, Oxford University Press, 2001) 1 ff. 19 Case C-671/15 Président de l’Autorité de la concurrence v Association des producteurs vendeurs d’endives (APVE) and Others EU:C:2017:860, para 37; E-J Mestmäcker and H Schweizer, Europäisches Wettbewerbsrecht, 2nd edn (München, CH Beck, 2004) ch 1, para 77: ‘Die Landwirtschaftspolitik ist ein Ausnahmebereich. Er rechtfertigt keine Analogie für andere Wirtschaftsbereiche oder andere Zuständigkeiten der Gemeinschaft’. [Own translation: ‘Agricultural policy is an exception. It does not justify an analogy for other economic sectors or other Community competences’.] In the follow-up edition, the authors provide a bit more nuance, but conclude that, despite of recent development in the agricultural market it has remained an exceptional sector, see E-J Mestmäcker and H Schweizer, Europäisches Wettbewerbsrecht, 3rd edn (München, CH Beck, 2014) ch 1, para 54. 20 K Purnhagen, ‘The End of Agricultural Exceptionalism in EU Free Movement Law and Competition Law after Lisbon?’ (2019) Wageningen Working Papers of Law and Governance 3/2019, available at: papers.ssrn.com/sol3/papers.cfm?abstract_id=3391134. 21 Case 203/83 Asteris EU:C:1985:357; Case C-127/07 Arcelor Atlantique and Lorraine and Others EU:C:2008:728; Case C-373/11 Panellinios EU:C:2013:567. 22 Case 71/74 Frubo EU:C:1975:61, paras 25/26 is sometimes interpreted in a way as to give priority to the protection of farmers’ income to the other goals mentioned in Art 39 TFEU. This is mainly due to the fact that the German translation classifies this goal as ‘Hauptziel’ [main goal], while the English language version only speaks of ‘the first two objectives’; such as the Italian version: ‘le due finalita principali’ [the first two objectives]. It is hence probably due to a translation error that this aim has been stipulated over time as the main aim; see also K Purnhagen and H Schebesta, ‘A Case at the Frontiers of Market Access, Freedom of Goods, The Common Agricultural Policy and Science in Court: Reflections on Scotch Whisky Association’ (2017) 42 European Law Review 420, 428.

From Supranationality to Managing Diversity  289 market law even beyond the application of Article 114 TFEU.23 Some observers may even interpret a ruling in a manner that the Court has established such a priority also for the objective of consumer protection.24 According to Article 101 TFEU EU competition law primarily aims at the ‘prevention, restriction or distortion of competition within the internal market’, while Article 102 TFEU targets the abuse of a dominant market position. Unsurprisingly this reiterates the internal market goal ‘hidden’25 in Protocol No 27 to safeguard a system of undistorted competition in the internal market, which the Court has applied, in accordance with Article 51 TEU, in conjunction with the internal market definition of Article 3(3) TEU.26 Whether and to what extent the competition law provisions also aim at the realisation of other internal market policy objectives, such as consumer protection and sustainability, is subject to debate in the literature27 and was seldom touched upon by the EU institutions.28 While these provisions allow for some guidance to determine which policy objectives to realise when ‘establishing’ the internal market, they do not provide any guidance as to ‘how’. Particularly in a multilevel system such as the EU, it is important to investigate at which level the internal market should be ‘established’. This is a question of competence. The internal market is, according to Article 4(2)(a) TFEU, an area of shared competence, that means according to Article 4(1) TFEU it ‘shall share competence with the Member States where the Treaties confer on it a competence’. Internal market policy is explicitly mentioned in Article 4(2)(a) TFEU. However, as the definition of the internal market has been substantially widened in Article 3(3ff) TEU and the common market as a legal concept has vanished from the Treaties, this internal market concepts encroaches on other competence areas as well. This does not cause any problem as long as the areas overlapping the new internal market concept are also assigned to the area of shared competence, such as for example consumer protection (Article 4(2)(f) TFEU) 23 Case C-183/95 Affish BV v Rijksdienst voor de Keuring van Vee an Vlees EU:C:1997:373, para 43; approved Case C-221/10 P Artedogan GmbH v European Commission EU:C:2012:216, para 99. 24 See, more elaborately, K Purnhagen, ‘Beyond Threats to Health: May Consumers’ Interests in Safety Trump Fundamental Freedoms in Information on Foodstuffs? Reflections on Berger v Freistaat Bayern’ (2013) 38 European Law Review 711 ff. 25 eg, Barents, above n 8, 125: ‘The requirement of undistorted competition has been hidden from the uninformed reader’. 26 Case C-496/09 European Commission v Italian Republic, above n 10, para 60. 27 See, eg, AD Chirita, ‘Undistorted, (Un)Fair Competition, Consumer Welfare and the Interpretation of Article 102 TFEU’ (2010) 33 World Competition: Law and Economics Review 417; K Cseres, ‘Towards an Economic Model of Social Justice: The Role of Competition Law’ in H-W Micklitz (ed), The Many Concepts of Social Justice in European Private Law (Cheltenham, Edward Elgar, 2011); A Gerbrandy, ‘Solving a Sustainability-Deficit in European Competition Law’ (2017) 40 World Competition: Law and Economics Review 539; G Monti and J Mulder, ‘Escaping the Clutches of EU Competition Law’ (2017) 42 European Law Review 635; H Schweitzer, ‘Competition Law and Public Policy: Reconsidering an Uneasy Relationship – The Example of Article 81’ in J Drexl, L Idot and J Monéger (eds), Economic Theory and Competition Law (Cheltenham, Edward Elgar, 2009); E van der Zee, ‘Quantifying Benefits of Sustainability Agreements Under Article 101 TFEU’ (2020) 43 World Competition: Law and Economics Review 189; G Monti, ‘Four Options for a Greener Competition Law’, Journal of European Competition Law & Practice, lpaa007, https://doi.org/10.1093/jeclap/lpaa007, Online First. 28 One example is the CECED Agreement. See Case IV.F.1/36.718, CECED [2000] OJ L187/47.

290  Kai P Purnhagen or agriculture and fisheries (Article 4(2)(d) TFEU). However, in particular, competition law (Article 3(1)(b) TFEU) and the common fisheries policy (Article 3(1)(d) TFEU) are listed as exclusive competences, which allow for a different standard to apply when evaluating at which level of the internal market the respective policy should be regulated. For these areas of the internal market, decisions have to be taken at the Union level. For the rest of the internal market, primary law leaves a leeway to legislators and judges to decide at which level and to what extent market establishment should be conducted. However, assigning the internal market to the area of shared competence already indicates that, as a default, the establishment of the internal market rather asks for the ‘management of interdependence’29 than, for example, providing ‘one-size-fits-all’ solutions via harmonisation.

A. Legislator Article 114 TFEU is the paramount competence norm for the establishment of the internal market. Despite the fact that the internal market belongs to the area of shared competence, Article 114(4–8) TFEU stipulates that, if Member States would like to deviate from a harmonisation measure by the European Parliament and the Council, they need to notify the Commission. One may hence argue that Article 114 TFEU presupposes maximum harmonisation as a default. In an area of maximum harmonisation, the Member States may not deviate from the common European rule, neither below, nor above. Paraphrasing Weatherill’s words: EU law provides both floor and ceiling.30 Indeed, in the early days of internal market law, most secondary legislative acts were geared towards harmonisation of national laws, understood more or less as equalisation of national laws to establish the internal market.31 Product standards were ‘Europeanised’, so goods could flow with the market without being disturbed by legal barriers.32 One standard to rule them all.33 In this world of maximum harmonisation social concerns are exclusively dealt with at the EU level with the result that Article 36 TFEU is factually disabled.34 The wording of Article 114(1) TFEU, namely that the ‘approximation of the provisions laid down by law, regulation or administrative action in Member States’ is the governing mode for internal market establishment still pays witness to the fact that in those days harmonisation meant in fact equalisation of Member State laws at a certain level. However, it has been pointed out that nowadays 29 S Weatherill, Law and Values (Oxford, Oxford University Press, 2016) 3. 30 S Weatherill, ch 12 in this volume. 31 See for the area of foods E Vos and F Wendler, ‘Food Safety Regulation at the EU level’ in E Vos and F Wendler (eds), Food Safety Regulation in Europe. A Comparative Institutional Analysis (Antwerp, Intersentia, 2006) 74. 32 A McGee and S Weatherill, ‘The Evolution of the Single Market: Harmonisation or Liberalisation’ (2019) 53 Modern Law Review 582. 33 Afilalo, Patterson and Purnhagen, above n 5. 34 S Weatherill, ch 12 in this volume.

From Supranationality to Managing Diversity  291 this cannot be verified in the practice of legislation any longer.35 Acts based on Article 114 TFEU follow minimum harmonisation policies, providing only a floor of harmonisation.36 The ceiling can still vary according to Member States’ desires. Moreover, a reading of Article 114 TFEU favouring maximum harmonisation necessarily creates tension with the fact that the internal market is part of the area of shared competence.37

B. Judges EU law, and the level of harmonisation, is often determined by the interpretation of the provisions by the Court of Justice of the European Union (CJEU). The CJEU’s mandate is, according to Article 19(1), sentence 2 TEU determined to ‘ensure that in the interpretation and application of the Treaties the law is observed’. This paragraph, jointly applied with other principles of EU law such as the effet utile, provided the EU judiciary with the power to initiate far-reaching decisions.38 From this power, the Court has developed principles of EU law,39 and as part of these principles fundamental rights,40 it has stipulated the EU’s autonomy vis-à-vis the national order,41 and it has declared EU law’s supremacy.42 It has created (State) liability norms,43 criteria for measures having equivalent effect within the meaning of the provisions concerning the free movement of goods,44 and the notion of direct effect.45 Why was it that the Court could be assigned such a wide mandate? For many, the Court went too far by declaring that so many areas belong at the EU level. It was perceived as activist,46 as a political actor,47 35 G Howells and N Reich, ‘The current limits of European harmonisation in consumer contract law’ (2011) 12 ERA Forum 39. 36 Exceptions make a rule: some markets such as foods or chemicals are subject to ever more systematised regulation at the EU level, see K Purnhagen, Systematization in EU Product Safety Law (Dordrecht, Springer, 2013). In areas such as foods, even fierce critics of maximum harmonisation admit that it may exceptionally make sense, see M Faure, ‘The Economics of Harmonization of Food Law in the EU’ in H Bremmers and K Purnhagen (eds), Regulating and Managing Food Safety in the EU (New York, Springer Nature, 2018) 283. 37 On the constitutional validity of these strategies see S Weatherill, ch 12 in this volume. 38 See for a more elaborative analysis in connection to legal theory Purnhagen, Systematization in EU Product Safety Law, above n 36, 145. 39 See more comprehensively T Tridimas, The General Principles of EU Law, 3rd edn (Oxford, Oxford University Press, 2020). 40 Case 29/69 Erich Stauder v City of Ulm ECLI:EU:C:1969:57. 41 Case 75/63 Hoekstra v Bedrijfsvereniging Detailhandel [1964] ECR 177; see on this effect Afilalo, Patterson and Purnhagen, above n 5. 42 Case 6/64 Costa/ENEL EU:C:1964:66; Case 11/70 Internationale Handelsgesellschaft EU:C:1970:114. 43 Case C-6/90 and C-9/90 Francovich EU:C:1991:428. 44 Case 8/74 Dassonville EU:C:1974:82. 45 Case 26/62 van Gend & Loos EU:C:1963:1. 46 See, for assessments of this claim, I Solanke, ‘“Stop the ECJ?” An Empirical Assessment of Activism at the Court’ (2011) 17 European Law Journal 764; A Grimmel, ‘“This is not Life as it is Lived Here”. The European Court of Justice and the Myth of Judicial Activism’ (2014) 7 European Journal of Legal Studies 61. 47 S Schmidt, The European Court of Justice and the Policy Process: The Shadow of Case Law (Oxford, Oxford University Press, 2018).

292  Kai P Purnhagen or as a vehicle for particular interest groups to pursue their agendas via Court decisions.48 For others, the same Court represented itself as the only EU institution with a clear integrationist agenda,49 the engine of European integration,50 which drives EU integration without steering it.51 The law-creating function of the CJEU had been described as a necessary requirement to build a supranational legal order.52 Indeed, in the early years EU law was a fragmented system, which needed to assert itself in an area between national and international law.53 In such incomplete legal systems, non-court institutions often lack the regulatory potential they need to solve problems and/or are blocked by disputes over competences.54 In the absence of other possibilities, the theory of constitutionalism assigns courts to be those that have the means to fill such regulatory gaps, creating a constitution of inter- or supranational law.55 It is within this concept that the rather expansionist early judgments of the Court need to be viewed. ‘The unwillingness- (sic!) and unableness of Member States to fulfil their obligations to integrate via a political process was substituted for integration via judicial control’.56 However, such constitutionalism by judgment has come under fierce attack in international law.57 At the EU level, however, if we look at recent cases such as Wightman58 or Confederation Payenne,59 the Court has been much more reluctant in proclaiming

48 U Haltern, ‘Integration through Law’ in A Wiener and T Diez (eds), European Integration Theory (Oxford, Oxford University Press, 2004) 177 ff. 49 S Saurugger and F Terpan, The Court of Justice of the European Union and The Politics of Law (London, Palgrave Macmillan, 2017) 208–16. 50 U Everling, ‘Die Zukunft der europäischen Gerichtsbarkeit in einer erweiterten Europäischen Union’ (1997) 32 Europarecht 398 ff. 51 C-D Ehlermann, ‘The European Communities, its Law and Lawyers’ (1992) 29 Common Market Law Review 218; G Mancini, ‘The Making of a Constitution for Europe’ (1989) 26 Common Market Law Review 595 ff; JHH Weiler, ‘Journey to an Unknown Destination. A Retroperspective and Prospective of the European Court of Justice in the Area of Political Integration’ (1993) 31 Journal of Common Market Studies 417. 52 JHH Weiler, ‘The Community System. The Dual Character of Supranationalism’ (1982) Yearbook of European Law 1981 267 ff; JHH Weiler, ‘The Transformation of Europe’ (1991) 100 Yale Law Journal 2403. 53 Lenaerts, above n 15, 15,16; Purnhagen, Systematization in EU Product Safety Law, above n 36, 144–45. 54 A-M Slaughter, ‘A Global Community of Courts’ (2003) 44 Harvard International Law Journal 191, esp 193 ff; E-U Petersmann, ‘How to Constitutionalize International Law and Foreign Policy for the Benefit of Civil Society?’ (1998–99) 20 Michigan Journal of International Law 1. 55 D Caruso, ‘Private Law and State-Making in the Age of Globalization’ (2006) 38 New York University Journal of International Law and Politics 29 ff. 56 Purnhagen, Systematization in EU Product Safety Law, above n 36, 158. 57 See for WTO law F Ronchetti and K Purnhagen, ‘Can the EU’s Reform Proposal for the WTO Dispute Settlement Understanding break the impasse at the WTO?’ (Winter 2019) 27 California International Law Journal 11. 58 Case C-621/18 Wightman and Others v Secretary of State for Exiting the European Union EU:C:2018:999. 59 Case C-528/16 Confédération paysanne and Others v Premier Ministre and Ministre de l’agriculture, de l’agroalimentaire et de la forêt EU:C:2018:583.

From Supranationality to Managing Diversity  293 great integrationist judgments. Rather, the Court based the outcome of its judgments more prominently than usual on the respect for national sovereignty and the management of the interdependence of Member States.60 If one had paid close attention to the literature on EU law, this may not come as a surprise. Since the beginning of the millennium, single voices in the literature have cautiously questioned whether in an EU legal system with an ever growing body of law the narrative of judicial constitutionalisation would still hold value.61 In 2013, just little time before Professor Koen Lenaerts took office as the President of the Court of Justice of the European Union, he reiterated some of these voices and announced a paradigm shift at the CJEU.62 He proclaimed, in the establishment phase of the EU, the Court had filled the gaps of the EU constitution. This was a clear endorsement of the role the previously voiced theory of constitutionalism had provided to the Court. However, Professor Lenaerts moved on to explain that, in his opinion, the constitution had by now been established at EU level. Elsewhere I have noted that the more the EU legal system develops, the more the role of the judiciary changes from a ‘constructor’ or ‘creator’ of the law to an ‘applier’.63 ‘The most logical reason for such an increasingly restrained role of the ECJ would therefore be that Member States in fact recaptured their role as political actors’ in areas where they are clearly assigned competences.64 In areas where the EU has established competences to shape politics via law, the Court’s jurisdiction will and/or should be ‘less exclusive’.65 Indeed, Professor Lenaerts endorses this view voiced earlier as he noted that, after the Constitution is established, the Court now moves on into a more governing mode, paying more deference to individuals, Member States and EU desires.66 Without a doubt, a court decides cases on individual bases67 and each judge is independent. The academic voice of a single professor, even if he is President of the CJEU, shall not be overrated in terms of impact on the day-to-day work of an EU institution such as the CJEU. However, a piece published by an established

60 See for the judgment in Confederation paysanne K Purnhagen, ‘How to manage the Union’s diversity: The case on the regulation of New Plant Breeding Technologies Confédération paysanne and Others’ (2019) 56 Common Market Law Review 1379; for the case of Wightman, see M Bartl and C Carr, ‘Law And Politics After Wightman: Taking Stock Of Neo-Formalism in the EU’ in L de Almeida, M Cantero, M Durovic and KP Purnhagen (eds), The Transformation of Economic Law (Oxford, Hart Publishing, 2019). 61 J Hunt, ‘The End of Judicial Constitutionalization’ (2007) 3 Croatian Yearbook of European Law and Policy 155; Purnhagen, Systematization in EU Product Safety Law, above n 36, 146–47. 62 Lenaerts, above n 15, 16. 63 Purnhagen, Systematization in EU Product Safety Law, above n 36, 146–47. 64 ibid, 158. 65 Reich, above n 6. 66 Lenaerts, above n 15, 16. 67 J Lindeboom, ‘Interpreting the EU Internal Market’ in F Amtenbrink, G Davies, D Kochenov and J Lindeboom, The Internal Market and the Future of European Integration (Cambridge, Cambridge University Press, 2019) 93 ‘(T)he Court decides cases, it does not write textbooks’.

294  Kai P Purnhagen professor a couple of years before he takes office as President, where he delineates his view of the role of the Court may indicate how he is willing to give shape to his office, reach decisions and talk to fellow judges during his mandate. After all, presidents are also elected based on the academic merits of the candidate. It can therefore be assumed that there is some endorsement of such fundamental issues on the bench. Hence, if one endorses Professor Lenaerts’ view that the constitution has now been established, Article 19(1), sentence 2 TEU also needs to be interpreted in a different light. Following this view, the ‘law’ that is to be observed by the Court does not refer any longer to a fragmented EU constitution, which translates into a mandate of the Court to fill these gaps. Rather, the ‘law’ is now established, and more emphasis needs to be given to balance the different Member State requirements, individual rights of EU citizens and the demands of the EU institutions.

C.  Preliminary Conclusion We may conclude that the Treaties provide ample leeway to determine how the various actors exercise their duty to ‘establish’ the internal market. While the internal market has not been defined strictu sensu, both the TEU and the TFEU provide some guidelines as regards its features and according to which features shall be given more weight than others. Article 114 TFEU, in addition, likewise provides guidelines, which are however not followed strictu sensu in practice. A general trend towards more decentralised solutions, which are paying more witness to the fact that diversity within the EU is more appreciated, is being observed. For the judiciary, its major mandate follows from Article 19(1), sentence 2 TEU. Within this mandate, the Court had in the past exercised an extensive interpretation of this provision in line with what would be expected from a ‘constitutional’ court. Strong voices have, however, called this view into question and advocated for a more restrained court, which focuses more on deference to Member States’ interests and individual rights protection. Based on this analysis one may already form the hypothesis that the aim of the establishment of the internal market is changing from constitutionalisation of the general supranational EU order to a mixed approach: as the Court is more deferential to Member State interests, one may foresee that when Member States are the addressees of EU law, EU law is being interpreted in a way that manages their diversity. When individuals are addressed as rights holders, either directly or indirectly, however, internal market law empowers them increasingly not only as market participant but as individual rights-holders. This may result in stronger harmonisation at EU level in terms of individual rights protection, while also managing the diversity needed to safeguard the respective different individual rights and individual interests as long as they are protected by the EU legal order. What still needs to be developed further, and is missing from this picture, are the cases where EU law adresses and governs action of EU institutions.

From Supranationality to Managing Diversity  295

III.  From Supranationality to the Management of Diversity? This section will establish further evidence that the notion to ‘establish’ the internal market is changing. To this end I will critically investigate several prominent areas of secondary internal market law (A) and adjudication (B) to investigate whether and how such a switch materialises. I will show that based on the critical analysis of these areas one may indeed illustrate a shift in market establishment towards diversity management in both legislation and adjudication as more respect for national space and more respect for different individual rights and preferences are pronounced.

A.  Secondary Legislation This section investigates whether the announced turn towards more respect for national space and individual rights protection can be observed in secondary legislation. As reference areas, I will look into the areas of unfair commercial practices law and risk regulation. Both areas are of special interest, as both areas touch heavily upon individual rights, both have been at the forefront of maximum harmonisation policies in areas where national regulatory interests remain high, and individual preferences are likely to diverge.

i.  Unfair Commercial Practices Law In order to respect national sovereignty over private law and diverging consumer preference, EU unfair commercial practices law had traditionally adopted a minimum harmonisation approach. In the literature, there has been wide consensus that this is the preferred method of integration, as it touches upon a number of issues which should better be regulated locally.68 However, in the early 2000s the Commission announced a switch in harmonisation strategy towards maximum harmonisation in these areas, which has triggered widespread scepticism in academia.69 Indeed, this strategy had been adopted in many areas of consumer contract law.70 In the legislative procedure, however, this strategy had been watered 68 T Wilhelmsson, ‘Full harmonisation of consumer contract law?’ (2008) 16 Zeitschrift fűr Europäisches Privatrecht 225. 69 M Faure, ‘Towards a Maximum Harmonization of Consumer Contract Law?!?’ (2008) 15 Maastricht Journal of European and International Law 433; O Osuji, ‘Business-to-consumer Harassment, Unfair Commercial Practices Directive and the UK – A Distorted Picture of Uniform Harmonisation?’ (2011) 34 Journal of Consumer Policy 437; Wilhelmsson, above n 68; M Durovic, European Law on Unfair Commercial Practices (Oxford, Hart Publishing, 2016) 14. 70 E Hondius, ‘The proposal for a European directive on consumer rights: A step forward’ (2010) 18 European Review of Private Law 103; C Twigg-Flesner, ‘Time to do the job properly – the case for a new approach to EU consumer legislation’ (2010) 33 Journal of Consumer Policy 355; M Loos,

296  Kai P Purnhagen down to what elsewhere was called a ‘targeted harmonisation’ or ‘limited full harmonisation’ approach in order to respect diversity and differences in Member States.71 In unfair commercial practices law the legislator largely succeeded in implementing a maximum harmonisation strategy.72 Two waves of harmonisation may be distinguished.73 The ‘first wave’ of unfair commercial practices law encompasses Directive 2005/29/ EC on unfair commercial practices (UCPD);74 the second wave in particular Regulation (EU) No 1169/2011 on Food Information to Consumers (FIR).75 In both waves, the maximum harmonisation strategy largely succeeded on paper; however, the rules have built-in mechanisms to account for the management of diversity, jeopardising the maximum harmonisation approach. Within the first wave, as also stipulated by recital 18 UCPD the protective level, according to which the unfairness of a practice is determined, still rests on the average consumer test.76 This test provides Member States’ leeway to determine whether the practice in question is deceptive according to the respective Member State requirement.77 Hence, the management of diversity of different Member States requirements was built into the terminology of the UCPD act itself. Consumer protection in the UCPD furthermore rests on an information-based regulatory approach, which has largely neglected substantive individual consumer protection. Provision of information as such was deemed the most appropriate means to protect consumers and realise market integration at the same time.78 Diversity has been managed largely in neglect of individual consumer rights, giving more weight to the virtue of internal market integration as a value in itself. ‘Full harmonisation as a regulatory concept and its consequences for the national legal orders: the example of the Consumer Rights directive’ (2010) Centre for the Study of European Contract Law Working Paper Series 2010/03. 71 See for a summary of the discussion Howells and Reich, above n 35. 72 Durovic, above n 69; B Keirsbilck, The New European Law of Unfair Commercial Practices and Competition Law (Oxford, Hart Publishing, 2001) 182–92. 73 See H Schebesta and K Purnhagen, ‘Island or Ocean: Empirical Evidence on the Average Consumer Concept in the UCPD’ (2020) 42 European Review of Private Law 299. 74 Directive 2005/29/EC of the European Parliament and of the Council of 11 May 2005 concerning unfair business-to-consumer commercial practices in the internal market and amending Council Directive 84/450/EEC, Directives 97/7/EC, 98/27/EC and 2002/65/EC of the European Parliament and of the Council and Regulation (EC) No 2006/2004 of the European Parliament and of the Council (Unfair Commercial Practices Directive). 75 Regulation (EU) No 1169/2011 of the European Parliament and of the Council of 25 October 2011 on the provision of food information to consumers, amending Regulations (EC) No 1924/2006 and (EC) No 1925/2006 of the European Parliament and of the Council, and repealing Commission Directive 87/250/EEC, Council Directive 90/496/EEC, Commission Directive 1999/10/EC, Directive 2000/13/EC of the European Parliament and of the Council, Commission Directives 2002/67/EC and 2008/5/EC and Commission Regulation (EC) No 608/2004. 76 Osuji, above n 69. 77 See on the different levels of this test, Schebesta and Purnhagen, ‘Island or Ocean’, above n 73. 78 E Steindorff, EG Vertrag und Privatrecht (Baden-Baden, Nomos, 1996) 195 ff; JA Usher, ‘Disclosure Rules (Information) as a Primary Tool in the Doctrine on Measures Having an Equivalent Effect’ in S Grundmann, W Kerber and S Weatherill (eds), Party Autonomy and the Role of Information in the Internal Market (Berlin, de Gruyter, 2001) 152–53.

From Supranationality to Managing Diversity  297 Secondary acts from the second wave have likewise adopted a maximum harmonisation approach. As Article 38(1) FIR, for example, makes clear, ‘Member States may not adopt nor maintain national measures unless authorised by Union law. Those national measures shall not give rise to obstacles to free movement of goods, including discrimination as regards foods from other Member States’. However, Article 39 and Article 45 FIR provide a procedure which stipulates under which conditions Member States may deviate from the maximum harmonisation provisions. Hence, management of diversity of Member States’ desires is legally proceduralised in these provisions. Regarding individual rights protection, the benchmark according to which the protective level is determined can be interpreted as being more tainted with individual rights protection. While the average consumer is still the benchmark according to which the protective level of the individual is being determined in most ‘second wave’ legislation acts, they also describe the targeted individual as a more vulnerable one than the typical average consumer benchmark, often implicitly drawing on insights from consumer behaviour science.79 Article 5(2) of the Nutrition and Health Claims Regulation,80 for example, explicitly stipulates that ‘the use of nutrition and health claims shall only be permitted if the average consumer can be expected to understand the beneficial effects as expressed in the claim’. Article 32(3) of the Regulation of Organic Production and Labelling of Organic Products (ROPL)81 stipulates that certain information on organic foods ‘shall be marked in a conspicuous place in such a way as to be easily visible, and shall be clearly legible and indelible’. Hence, these passages take into account research on the limited processing capacities of consumers. Article 32(2) sub-para 4 ROPL requires further that ‘(t)he words “EU” or “non-EU” shall not appear in a colour, size and style of lettering that is more prominent than the name of the product’. This passage takes into account that the size and placement of pictures on food packs can determine consumers’ choice. Under Article 35(1)(d) FIR, a nutrition declaration may be given ‘by other forms of expression and/or presented using graphical forms or symbols in addition to words or numbers provided that the following requirements are met’, namely if ‘they are supported by scientifically valid evidence of understanding of such forms of expression or presentation by the average consumer’. These provisions bear witness to the fact that more recent EU unfair commercial practices

79 See K Purnhagen and H Schebesta, ‘Food Labelling for Consumers – EU Law, Regulation and Policy Options, Study requested by the PETI committee of the European Parliament’ PE 608.871, April 2019, 35–38, available at: www.europarl.europa.eu/RegData/etudes/STUD/2019/608871/IPOL_ STU(2019)608871_EN.pdf. 80 Regulation (EC) No 1924/2006 of the European Parliament and of the Council of 20 December 2006 on nutrition and health claims made on foods ([2006] OJ L404, 9) last amended by Commission Regulation (EU) No 1047/2012 of 8 November 2012 [2012] OJ L310/36 (Nutrition and Health Claims Regulation). 81 Regulation (EU) 2018/848 of the European Parliament and of the Council of 30 May 2018 on organic production and labelling of organic products and repealing Council Regulation (EC) No 834/2007 [2018] OJ L150/1.

298  Kai P Purnhagen legislation does not target an average consumer image any longer, but rather takes into account consumers’ deficits in understanding and processing information, and hence arguably gives more weight to their individual rights of protection. In the case of Articles 30(1)(a) and 35(1)(d) FIR, the FIR makes direct reference to consumer studies to substantiate the use of pictograms and symbols. This appears to accommodate the more modern version of the behaviourally informed understanding of the average consumer that arose in about 2005.82

ii.  Risk Regulation In the area of risk regulation, which covers health, safety, and the environment, after decades of legislative and judicial expansion of the EU’s regulatory scope one may observe a move to more respect for national diversity in line with the move to the management of diversity. In the EU, the area of risk regulation is governed by the triangle of risk assessment, risk management and risk communication.83 As a matter of principle, risk assessment is in most of these areas, in particular in EU food law, to be conducted primarily at EU level while risk management remains at national level.84 In line with the adoption of the supply chain approach,85 however, parts of risk analysis lie with the food chain actors, mainly food business operators, in the form of mandatory hazard analysis and critical control points (HACCP) controls.86 The implementation of this risk regulation approach, which started at around 2000, has an in-built mechanism to manage the diversity of Member States’ interests in the Union at risk management level while keeping a required level of harmonisation in the area of risk analysis. Likewise, the area of risk regulation has been one of the first to realise individual protection via secondary legislation in the area of health. These attempts have famously been struck down by the ECJ in the Tobacco judgment, where it for the first and until now only time, had declared secondary legislation to be not in compliance with a competence norm of the Treaty.87 The Court, however, had already given the protection of health a prominent position in EU internal market law as it had stipulated in 1995 that whenever

82 G Howells ‘The Potential and Limits of Consumer Empowerment by Information’ (2005) 32 Journal of Law and Society 349. 83 See, inter alia, Art 3 No 10 Regulation (EC) No 178/2002 of the European Parliament and of the Council of 28 January 2002 laying down the general principles and requirements of food law, establishing the European Food Safety Authority and laying down procedures in matters of food safety [2002] OJ L31, 1–24. 84 Vos and Wendler, above n 31; Faure, ‘The Economics of Harmonization of Food Law in the EU’, above n 36. 85 J Vapnek, ‘Legislative Implementation of the Food Chain Approach’ (2007) 40 Vanderbilt Journal of Transnational Law 987. 86 See for an overview of the legislative package: ec.europa.eu/food/safety/biosafety/food_hygiene/ legislation_en. 87 See S Weatherill, ‘The Limits of Legislative Harmonization Ten Years after Tobacco Advertising: How the Court’s Case Law has become a “Drafting Guide”’ (2011) 12 German Law Journal 827.

From Supranationality to Managing Diversity  299 a measure is intended to guarantee the protection of public health ‘it must take precedence over economic considerations’.88 Interestingly, when risk regulation comes into contact with other internal market regulation such as agricultural law, which follows a different harmonisation strategy, tensions are inevitable. For example, while the food market in Europe has for quite some time been ruled by the Cassis de Dijon principle of conditional mutual recognition,89 measures to establish the Common Market Organisation in certain sectors such as spirits and wine have established standardised rules on features of foodstuffs which directly jeopardise this principle.90 Likewise, diversity arguments from the application of the proportionality principle in the free movement of goods impact the static application of the provisions on Common Market Organisations within the CAP.91

B. Adjudication This section investigates whether the announced turn towards more respect for national space and individual rights protection can also be observed in the case law of the CJEU. As reference cases, I will look into major recent case law, which had to deal in particular with issues of diverging national interest and/or with the protection of individual rights at the EU level. Regarding the former, I will look into the recent landmark cases of Wightman and others and Confédération paysanne and Others. Regarding the latter, I will investigate the case of Åklagaren v Hans Åkerberg Fransson as a representative case of the Court’s (re)new(ed) approach.

i.  Wightman and Others The judgment in Wightman and Others may be the judgment which best illustrates the Court’s switch towards more deference to Member States’ interests.92 The English Court of Session had submitted the following question: where, in accordance with Article 50 TFEU, a Member State has notified the European Council of its intention to withdraw from the European Union, does EU law 88 Case C-183/95 Affish BV v Rijksdienst voor de Keuring van Vee an Vlees, above n 23, para 43; approved Case C-221/10 P Artedogan GmbH v European Commission, above n 23, para 99. 89 See S Weatherill, ‘Why There Is No “Principle of Mutual Recognition” in EU law (and Why that Matters to Consumer Lawyers)’ in K Purnhagen and P Rott (eds), Varieties of European Economic Law and Regulation (Dordrecht, Springer Science, 2014). 90 Most illustratively, Arts 2(1)(c) and 6(2) in connection with No 34 Annex II, which directly regulate the minimum alcohol content of Crème de Cassis, and hence jeopardise the conditional mutual recognition regime established in internal market law by the Cassis de Dijon judgment. I am indebted to S Weatherill who pointed this out to me. 91 Purnhagen, ‘The End of Agricultural Exceptionalism in EU Free Movement Law and Competition Law after Lisbon?’, above n 20. 92 Case C-621/18 Wightman and Others, above n 58.

300  Kai P Purnhagen permit that notice to be revoked unilaterally by the notifying Member State; and, if so, subject to what conditions and with what effect relative to the Member State remaining within the European Union? While the answer to this question required the interpretation of Article 50 TFEU, the questions incidentally touched upon core questions of the EU as a legal order, its relationship to Member States’ interests and to the international legal order. Can a Member State, once it has given away sovereignty to a supranational institution such as the EU, claim it back? Under which conditions? International law, EU law, national law? Ultimately, this is a question of politics and law’s conception of the EU as a supranational institution.93 Advocate General Campos Sánchez-Bordona has proposed to answer this question in a manner taking into account the hybrid nature of such a withdrawal request. On one hand, the withdrawal request is granted by EU law, hence the procedures and legal values of EU law apply to this request. On the other, the withdrawal request can be viewed as an exercise of any State’s capacity to withdraw as vested in Article 42(2) of the Vienna Convention on the Law of Treaties. Hence, according to the AG’s view, the Vienna Convention on the Law of Treaties (VCLT) can provide interpretative guidance for the interpretation of Article 50 TFEU.94 By establishing this caveat, the AG had placed the withdrawal request within the scope of supranational EU law, carrying, however, a strong facet of intergovernmentalism by introducing the VCLT as interpretative yardstick. This notion towards more deference to Member States’ interests in the withdrawal situation at the cost of supranationality is even more pronounced as the AG qualifies the withdrawal request from the EU as an expression of the State’s sovereignty, conditional upon a State’s constitutional requirements.95 Member State interest is implemented via rules of interpretation from international and national laws of supranational rules. How deference to Member States’ interests has been the guiding line of this opinion becomes particularly evident in the AG’s submission that membership in the EU includes the principle of respect for the constitutional identity of the Member State, which supports taking into account a change in the sovereign will of the departing State.96 This submission is difficult to bring in line with contemporary theories of statecraft, which presuppose shared sovereignty as the default.97 However, this view has also been contested in the literature, which described the relationship between the EU and its Member States in a different way: Member States did not give up their sovereignty by joining the EU but simply chose to exercise their sovereignty in a different way.98 It is with this view that the 93 See also Bartl and Carr, above n 60. 94 Opinion of AG Campos Sánchez-Bordona in Case C-621/18 Wightman and Others v Secretary of State for Exiting the European Union EU:C:2018:978, para 82. 95 ibid, paras 91–93. 96 ibid, paras 130–32. 97 D Patterson and A Afilalo, The New Global Trading Order (Cambridge, Cambridge University Press, 2008) 38. 98 M Avbelj, ‘Supremacy or Primacy of EU Law – (Why) Does it Matter?’ (2011) 17 European Law Journal 744.

From Supranationality to Managing Diversity  301 reasoning in the judgment Wightman and Others sits more easily. Interestingly, the AG, however, categorises the withdrawal request as a unilateral act, which is governed by principles of good faith and sincere cooperation.99 The Court has taken the approach of deference to Member States’ interests at the expense of supranationality and shared sovereignty further. The Court emphasises the sovereignty of a State in deciding to retain its status as a Member State.100 Relying strongly on the notion of Member State will, albeit framed in EU law jargon such as the principle of an ever closer union among the peoples of Europe, and the values of liberty and democracy,101 the Court notes that a State cannot be forced to accede to the Union.102 A contrario, Member States cannot be forced to withdraw unwillingly. In other words: the Member State’s will determines membership of the Union and the amount of sovereignty shared. In contrast to the reasoning of the AG this would not stem from international law, but rather from the named EU law principles. It is hard to imagine how the CJEU could be more deferential to Member States’ interests at the expense of supranationality. How the rationale in this judgment impacts the EU legal order may be illustrated best if contrasted with earlier judgments on the matter from the heyday of EU constitutionalism. Compare in particular the wording of summary 3 in Costa ENEL: By contrast with ordinary international treaties, the EEC Treaty has created its own legal system which, on the entry into force of the Treaty, became an integral part of the legal systems of the Member States and which their courts are bound to apply. By creating a community of unlimited duration, having its own institutions, its own personality, its own legal capacity and capacity of representation on the international plane and, more particularly, real 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 have thus created a body of law which binds both their nationals and themselves. The integration into the laws of each Member State of provisions which derive from the Community and more generally the terms and the spirit of the Treaty, make it impossible for the States, as a corollary, to accord precedence to a unilateral and subsequent measure over a legal system accepted by them on a basis of reciprocity. Such a measure cannot therefore be inconsistent with that legal system. The law stemming from the Treaty, an independent source of law, could not because of its special and original nature, be overridden by domestic legal provisions, however framed, without being deprived of its character as Community law and without the legal basis of the Community itself being called into question. The transfer by the States from their domestic legal system to the Community legal system of the rights and obligations arising under the Treaty carries with it a permanent limitation of their sovereign rights.103



99 Opinion

of AG Campos Sánchez-Bordona, above n 94, paras 148 and 156. C-621/18 Wightman and Others, above n 58, paras 59 and 60. 101 ibid, paras 61–63. 102 Consolidated Version of the Treaty on European Union [2008] OJ C115/13 (TFEU) Art 49. 103 Case 6-64 Flaminio Costa v ENEL EU:C:1964:66. 100 Case

302  Kai P Purnhagen In this case, the Court has been particularly clear that transfer of sovereignty is ‘permanent’, and that the Treaty provision cannot be ‘overridden by domestic legal provisions, however framed’. When contrasting these lines to Wightman, representing the new managerial area of EU adjudication with Costa/ENEL from the previous area, I cannot help but express the feeling that something very important got lost.

ii.  Confédération paysanne and Others Confédération paysanne and Others serves as another example of the Court’s new approach to more deference for Member States interests.104 In contrast to Wightman, the case does not operate at a conceptual level, but rather illustrates how this new approach is being operationalised at the day-to-day level of judicial decision-making. The case concerned a preliminary reference procedure regarding, among others, questions concerning the interpretation of the technical terms ‘genetically modified organism’ (GMO) and ‘mutagenesis’ in Directive 2001/18/EC.105 New technologies had emerged on the market, which would technically fall under the respective provisions, but had not been known at the time the Directive came into force. Neither Member States nor the Commission have so far updated the Directive. In the past, Member States could not find consensus on any question in relation to genetically modified organisms.106 Hence, the judgment touches upon a sensitive political issue in the Member State. Probably as a result of this, both Advocate General Bobek and the Court have approached the question not from a legal, but rather from a managerial perspective.107 Neither the AG nor the CJEU answered the legal question of whether these new technologies fall within the scope of the Directive. Rather, the AG emphasised the minimum harmonisation character of the respective provisions, leaving it to Member State to decide if and how they wished to include these new technologies under the scrutiny of the Directive. The CJEU capitalised in the specifics of the requirements of the preliminary reference procedure, by which facts are established by the referring national court. It took the review of the risk assessment of the referring court for granted, which had established that these new technologies are similar to GMOs which are already under the scrutiny of the Directive. As a consequence, it applied the Directive without questioning the factual assessment of new technologies by the referring court. Both the AG and the Court have hence applied the maximum possible legal amount of deference to Member

104 Case C-528/16 Confédération paysanne and Others, above n 59. 105 Directive 2001/18/EC of the European Parliament and of the Council of 12 March 2001 on the deliberate release into the environment of genetically modified organisms and repealing Council Directive 90/220/EEC – Commission Declaration [2001] OJ L106, 1–39. 106 RD Smart, M Blum and J Wesseler, ‘EU member states’ voting for authorizing genetically engineered crops: a regulatory gridlock’ (2015) 64 German Journal of Agricultural Economics 244. 107 Purnhagen, ‘How to manage the Union’s diversity’, above n 60.

From Supranationality to Managing Diversity  303 States’ interests: the AG by empowering the national legislator and the CJEU by empowering national Courts.108

iii.  Åklagaren v Hans Åkerberg Fransson In the judgment in Åklagaren v Hans Åkerberg Fransson109 the Swedish Haparanda tingsrätt had submitted questions concerning the scope of application of the Charter of Fundamental Rights. Among the many interesting facets of the case, the judgment also established that the individual rights protection provided by the Charter also extends to national provisions which do not directly transpose EU legislation into national law, but which ‘nonetheless remain in direct and close connection with European law by way of ensuring its effective application in the national legal system’.110 This judgment came as a surprise to many, as Article 51(1) of the Charter stipulates that ‘(t)he provisions of this Charter are addressed to the institutions and bodies of the Union with due regard for the principle of subsidiarity and to the Member States only when they are implementing Union law’. In Åklagaren v Hans Åkerberg Fransson the Court interpreted this provision in a rather expansionist manner. Conventional wisdom would probably have assigned such an expansionist interpretation to key EU provisions such as supremacy and fundamental freedoms. One is hence attempted to view Åklagaren v Hans Åkerberg Fransson as a legal innovation to protect the individual more strongly at EU level. This would play well into the argument made earlier that since 2015 the Court pays more attention to the protection of the individual at the EU level. However, empirical analysis of Court decisions has illustrated that expansionist interpretation to protect the individual is nothing new to the Court. Rather, the Court had always used the tool of effet utile to protect the individual in cases concerning provisions of direct effect, supremacy and human rights.111

C.  Preliminary Conclusion In the investigated legislative acts as well in the case law one can indeed identify a move towards more deference to Member States’ interests and an emphasis on the protection of the individual. Admittedly, the investigated measures and acts are few, however, these are also areas of big impact. Further empirical research would 108 ibid. 109 Case C-617/10 Åklagaren v Hans Åkerberg Fransson EU:C:2013:105. 110 M Safjan, ‘The Horizontal Effect of Fundamental Rights in Private Law – On Actors, Vectors, and Factors of Influence’ in K Purnhagen and P Rott (eds), Varieties of European Economic Law and Regulation (Dordrecht, Springer Science, 2014) 143. 111 See 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’ (2015) 8 European Journal of Legal Studies 18, 41 ff.

304  Kai P Purnhagen need to be conducted to substantiate this claim.111a I will, however, continue with the established tendency that an emphasis towards more ‘managing of diversity’ can indeed be observed.

IV.  Consequences for EU Law In the areas observed one may identify a preference for more deference of EU law for Member States’ interests and an increasing concern for the protection of individual rights at EU level. Whether this applies to other areas as well needs further investigation.111b However, assuming that it does, this switch of paradigm in the interpretation and legislation of EU internal market law begs the question whether the expected consequences also require legal responses. This section will hence first outline the likely effects of such an increase of deference to Member States’ interests and individual rights protection in Europe. Subsequently, it will investigate how EU law should respond to such a switch.

A.  Consequences of Managing Diversity The consequences of maintaining divergent legal systems in the EU have been studied most extensively in economics. It has been observed that the maintenance of diversity is often the most costly form of regulation among several other regulatory options.112 At the same time, the resulting competition between Member States is expected to deliver better outcomes over time,113 as countries where policy adjustments are less costly are expected to grow faster than others.114 In addition, as the economics of federalism have observed, diverse Member State systems also widen the choice options of individuals and are hence more likely to provide the framework for each EU citizen to maximise their preferences.115 Interestingly, here the ‘managing’ approach is at odds with the other facet of the new internal market approach, namely increasing protection of individual rights. If this were understood as a general priority of consumer rights to safety over other internal market goals, for example, this would result in the harmonisation of regulation at the highest conceivable level of protection, jeopardising diversity. In this sense, it has been rightly observed that the provision of diversity of products forms an ‘economic 111a See for and excellent analysis in the area of EU law of fundamental freedoms J Zglinski, ‘The Rise of Deference: The Margin of Appreciation and Decentralized Judicial Review in EU Free Movement Law’ (2018) 55 Common Market Law Review 1341. 111b ibid. 112 Purnhagen and Wesseler, above n 3. 113 Kerber, above n 6. 114 J Wesseler, R Jongeneel and K Purnhagen, ‘Bioeconomy Economics and Policies’ in L Dries, W Heijmans, J Wesseler and K Purnhagen (eds), EU Bioeconomy Economics and Policies (Cham, Palgrave Macmillan, 2019) 8. 115 R van den Bergh, ‘Subsidiarity as an economic demarcation principle and the emergence of European Private Law’ (1998) 15 Maastricht Journal of European and Comparative Law 192.

From Supranationality to Managing Diversity  305 demarcation principle’ of internal market law.116 However, if costs of safety are bigger than those of diversity a tipping point may be observed, where standardisation is preferable. This has been suggested in the area of the food market, where the mutual recognition procedure as applied in practice could have delivered more efficient outcomes.117 Very often the provision of such diversity is connected to the fear that when unregulated it may, comparable to the ‘market for lemons’,118 trigger a race to the bottom of standards. However, the practice of internal market law has not endorsed this criticism raised in theoretical economics:119 in Drei Glocken the Court acknowledged evidence that in spite of increasing liberalisation of the pasta market, the market shares of pasta made from wheat of higher quality increased.120 Likewise, other studies have shown that higher standards combined with bigger market share can create a race to the top.121 Leaving decisions to Member States may also hinder the achievement of certain policy choices. Politically, placing more decisions with Member States makes certain that it leaves decisions of socio-economic life with democratically elected representatives who have the respective cultural background, thereby strengthening democracy. Within the framework of EU law, Member States may implement measures according to their socio-economic needs. Depending on the named socioeconomic factors, national institutions may dial up or dial down the stringency of their oversight accordingly. Ultimately it may well be, however, that one model will be preferred by most actors on the internal market, resulting in de facto harmonisation. When diversity is managed through respect for national procedures such as it is in the preliminary reference procedure, however, the danger exists that national judgments establishing facts will de facto be applicable across the EU.122 In cases where the establishment of facts involves highly controversial political issues this approach may not be a preferable option, as it ultimately favours the judicial solutions of one Member State over others.123

B.  Consequences for EU Law From a governance perspective, an approach which manages diversity at EU level creates a network of different Member States and individuals, each of which can be viewed as ‘institutions’. Elsewhere, such a network in an internal market has 116 ibid. 117 Faure, ‘The Economics of Harmonization of Food Law in the EU’, above n 36. 118 G Akerlof, ‘The Market for “Lemons”: Quality Uncertainty and the Market Mechanism’ (1970) 84 Quarterly Journal of Economics 488. 119 See K Purnhagen, ‘The Virtue of Cassis de Dijon 25 years later – It is not Dead, It Just Smells Funny’ in K Purnhagen and P Rott (eds), Varieties of European Economic Law and Regulation (Dordrecht, Springer Science, 2014) 325. 120 Case 407/85 Drei Glocken GmbH v USL Centro-Sud EU:C:1988:401, para 27. 121 A Bradford, ‘The Brussels Effect’ (2012) 107 Northwestern University Law Review 1. 122 Purnhagen, ‘How to manage the Union’s diversity’, above n 60. 123 ibid.

306  Kai P Purnhagen been described as ‘polycentric’.124 Such institutional networks require governance devices to steer the web of relationships if a certain outcome is desired.125 In this sense, competition among legal orders in the EU do not operate in a void. Rather, they work towards the achievement of certain principles and values as outlined at the beginning of this chapter. It is from this rationale that EU law derives its purpose within the concept of management of diversity: ‘The basic purpose (of EU law, addendum KP) is the management of interdependence’.126 Searching for a role for EU law to play we can meaningfully conclude that legal principles which particularly fulfil the aim to manage interdependence have the potential to become more prominent in EU law. AG Campos Sánchez-Bordona in his thoughtful opinion in Wightman has already identified the principles of good faith and sincere cooperation127 as principles which govern such interdependence. In EU legal terminology, several principles exist which qualify for such steering principles of jurisdictional competition. Most of them, interestingly, have received increasing attention lately in both the jurisprudence of the CJEU and scholarly literature. Subsequently, I will discuss their application and strengths and weaknesses.

i.  Subsidiarity Principle The principle of subsidiarity is often mentioned first in political debates when it comes to describing the limits of the power of the EU. The principle is mentioned in Article 5(1), sentence 1 and 5(3) TEU as a limit to Union action and is further elaborated on in Protocol 2 on the application of the principles of subsidiarity and proportionality. The subsidiarity principle has traditionally played only a minor role in the jurisprudence of the CJEU. For example, it had stipulated that the principle of subsidiarity is sufficiently taken into account if a measure of secondary legislation fulfils the requirements of Article 114 TFEU.128 After having been buried rather ineffectively in the Treaties for some time, the principle has gained attention in recent years and has become quite a fashionable topic. First attempts to ask for a more prominent function were declared in the context of the drafting of the Treaty of Lisbon, which has received considerable critical attention.129

124 J van Zeben and A Bobic, ‘Polycentricity and the Internal Market’ in J van Zeben and A Bobic (eds), Polycentricity in the European Union (Cambridge, Cambridge University Press, 2019). 125 J van Zeben, ‘Polycentricity as a Theory of Governance’ in J van Zeben and A Bobic (eds), Polycentricity in the European Union (Cambridge, Cambridge University Press, 2019); F Cafaggi, ‘Contractual Networks and Contract Theory: A Research Agenda for European Contract Law’ in F Cafaggi (ed), Contractual Networks, Inter-Firm Cooperation and Economic Growth (Cheltenham, Edward Elgar, 2011) 76–77; T Haryanto and KP Purnhagen, ‘Governing Nature Conservation in Political “Hotbeds”: A Contractual Approach’ (2019) 25 Hastings Environmental Law Journal 143. 126 Weatherill, Law and Values, above n 29, 3. 127 Opinion of AG Campos Sánchez-Bordona, above n 94, paras 148 and 156. 128 Case 377/98 Netherlands v Parliament and Council EU:C:2001:523, para 32. 129 G Davies, ‘Subsidiarity: the wrong idea, in the wrong place, at the wrong time’ (2006) 43 Common Market Law Review 63; R Schütze, ‘Subsidiarity after Lisbon: reinforcing the safeguards of federalism?’ (2009) 68 Cambridge Law Journal 525.

From Supranationality to Managing Diversity  307 After the measures introduced through the Treaty of Lisbon had (foreseeably) little effect, President Juncker in 2017 established a task force to reflect on the principles on proportionality and subsidiarity.130 The Commission’s report fed into a Commission Communication in 2018, which had the aim to turn the principle of subsidiarity into a more effective tool.131 The initial attempts have likewise received critical attention.132 It is doubtful whether the proposed measures will indeed give the subsidiarity principle more bite. Maybe the problem with the focus on subsidiarity and why it does not really take a grip is that it has a very limited application. The principle of subsidiarity only covers legislation and, to a limited extend, also adjudication, and thereby the least controversial part of European integration.133 Furthermore, it only applies to questions of substantive law in the EU/Member State relationship, leaving procedural law and the individual aside. Hence, in order to govern the shift towards more management of diversity other principles may play a more prominent role in the near future.

ii.  Principle of Proportionality The principle of proportionality has always been assigned a significant role in EU law as both, a general principle of EU law134 and a limit to Union action.135 There is some debate regarding the methodology behind the Court’s exercise of the proportionality test.136 Either way, the standard formula applied by the CJEU to test whether a provision of Union law is consonant with the principle of proportionality is whether the means it employs to achieve the aim correspond to the importance of the aim and whether they are necessary for its achievement.137 In this sense, the proportionality principle governs both the allowed intensity of

130 Decision of the President of the European Commission on the establishment of a Task Force on Subsidiarity, Proportionality and ‘Doing Less More Efficiently’ C(2017) 7810. 131 Communication from the Commission to the European Parliament, the European Council, the Council, the European and Social Committee and the Committee of the Regions, ‘The principles of subsidiarity and proportionality: Strengthening their role in the EU’s policymaking’ COM(2018) 703 final. 132 Vincent Delhomme, ‘How to turn subsidiarity into an effective tool?’ (European Law Blog, 19 March 2019). 133 S Garben, ‘Restating the Problem of Competence Creep, Tackling Harmonisation by Stealth and Reinstating the Legislator’ in S Garben and I Govaere (eds), The Division of Competences between the EU and the Member States: Reflections on the Past, the Present and the Future (Oxford, Hart Publishing, 2017). 134 ECJ 8/55, Fédération Charbonnière ECLI:EU:C:1956:11, para 311. 135 Art 5(1) sentence 1 and (3, 4) TEU; Protocol (2) on the application of the principles of subsidiarity and proportionality; Art 52(1) of the Charter of Fundamental Rights. 136 Some criticism concerns the standards applied, see BVerfG, judgment of the Second Senate of 05 May 2020 – 2 BvR 859/15, paras 1–237; other criticism points out that the sequence of the test may be different, see K Purnhagen and E van Kleef, ‘Commanding to ‘Nudge’ via the Proportionality Principle?’ in H Bremmers and K Purnhagen, Regulating and Managing Food Safety in the EU (Dordrecht, Springer Science, 2018) 160. 137 See, eg, Case 66/82 Fromançais SA v Fonds d’orientation et de régularisation des marchés agricoles (FORMA) ECLI:EU:C:1983:42, para 8.

308  Kai P Purnhagen an intervention and whether the intervention is fit for purpose. It hence has the potential to play a major role in managing diversity particularly between Member States and Union action, and intra-Union action. In this sense, the Court has in recent years reinvestigated the proportionality principle in particular in the area of free movement laws in the following way:138 in recent case law the Court opened up the scope of fundamental freedoms, widening the ‘net’ with which the rules on the free movement of goods catch national provisions.139 In a multilevel system such as the EU, an increase in importance of the proportionality test means that a widening of the scope at EU level puts pressure for justification on Member States,140 as the proportionality review is applied at Member State level. A necessary conundrum to a widening of the scope of free movement of goods is hence an increase in importance of the proportionality test.141 Framing the exercise of such a proportionality test with EU law hence carries the potential to also serve as a governing tool to manage diversity. Plenty of possibilities have been brought forward regarding how such a framing of diversity can be realised.142 While the CJEU has not yet taken this approach with full force,143 the Court increasingly pays attention to the standard of review it applies to a Member State’s exercise of the proportionality test.144 However, as regards the potential to serve as a governance tool for managing diversity, it shares the same fate as the subsidiarity principle. It is likewise linked to legislation and, to a limited extent, to adjudication and the EU/Member State relationship, and hence covers only a small part of the problems addressed.

iii.  Principle of Conferral Unlike the prominent treatment of the principle of conferral that some textbooks on EU law make us believe, the principle of conferral has had very little role to play

138 See B van Leeuwen, ‘Rethinking the Structure of Free Movement Law: The Centralisation of Proportionality in the Internal Market’ (2017) 10 European Journal of Legal Studies 235, 259–60; K Purnhagen, ‘Keck is Dead, Long Live Keck? How the Court of Justice tries to avoid a Sunday trading saga 2.0’ in F Amtenbrink, G Davies, D Kochenov and J Lindeboom (eds), The Internal Market and the Future of European Integration, Essays in Honour of Laurence W Gormley (Cambridge, Cambridge University Press, 2019). 139 M Jesse, ‘What about Sunday Trading …? The Rise of Market Access as an Independent Criterion under Article 34 TFEU’ (2012) 2 European Journal of Risk Regulation 437, 441. 140 Purnhagen, ‘Keck is Dead, Long Live Keck?’, above n 138; N Nic Shuibhne and M Maci, ‘Proving Public Interest: The Growing Impact of Evidence in Free Movement Case Law’ (2013) 50 Common Market Law Review 965. 141 LW Gormley, ‘Reasoning Renounced? The Remarkable Judgment in Keck & Mithouard’ (1994) 5 European Business Law Review 63, 67; van Leeuwen, above n 138. 142 T Tridimas, ‘Constitutional Review of Member State Action: The Virtues and Vices of an Incomplete Jurisdiction’ (2011) 9 I·CON 737; K Purnhagen and H Schebesta, ‘A Case Moving at the Frontiers of Market Access, Freedom of Goods, the Common Agricultural Policy and Science in Court’ (2017) 42 European Law Review 420, 431. 143 Lindeboom, above n 67, 99. 144 Purnhagen, ‘Keck is Dead, Long Live Keck?’, above n 138.

From Supranationality to Managing Diversity  309 in EU law in action. It is enshrined in Article 5(1), sentence 1 TEU as a governance tool to limit the Union’s competences. So far the Court has only once made use of the principle of conferral to declare a measure to be not in line with the EU Treaties.145 Member State courts, however, frequently considered making use of the principle of conferral in their jurisprudence to define acts ultra vires.146 While this illustrates the status quo, it is questionable if it will remain this way. Providing the principle of conferral with a stronger constitutional value could indeed be a path worth considering. However, it shares the same fate as the subsidiarity principle, as it is likewise linked to legislation and, to a limited extent, to adjudication and the EU/Member State relationship. It hence covers only a small part of the problems addressed.

iv.  Principle of Solidarity At about the same time that scholarly and legislative attention started to be more devoted to subsidiarity, scholars147 and the Court148 have been emphasising solidarity as a legal principle of EU law.149 Unlike the principles of conferral and subsidiarity, the principle of solidarity carries a bigger potential to serve as a meaningful legal governance tool for managing diversity in the EU as it has a wider scope of application: it may potentially apply to many kinds of EU integration measures, not only legislation and adjudication. It covers not only the EU/Member State relationship but also the relationship between Member States, individuals and procedural law. Substantively, unlike the principle of conferral and the subsidiarity principle, it is still very much an ‘abstract concept’150 in the sense that the European judiciary still has the chance to fill it with meaning in view of the challenges in managing diversity. 145 See Weatherill, ‘The Limits of Legislative Harmonization Ten Years after Tobacco Advertising’, above n 87. 146 See, inter alia, P Craig, ‘The ECJ and ultra vires action: A conceptual analysis’ (2011) 48 Common Market Law Review 395; Z Kühn, ‘Ultra Vires Review and the Demise of Constitutional Pluralism: The Czecho–Slovak Pension Saga, and the Dangers of State Courts’ Defiance of EU Law’ (2016) 23 Maastricht Journal of European and Comparative Law 185. 147 C Newdick, ‘Citizenship, Free Movement and Health Care: Cementing Individual Rights by Corroding Social Solidarity’ (2006) 43 Common Market Law Review 1645, 1666; C Barnard, ‘EU-Citizenship and the Principle of Solidarity’ in M Dougan and E Spaventa (eds), Social Welfare and EU Law (Oxford, Hart Publishing, 2005). 148 Case C-209/03 The Queen, on the application of Dany Bidar v London Borough of Ealing and Secretary of State for Education and Skills ECLI:EU:C:2005:169, para 56; opinion of Advocat General Maduro in Case C-281/06 Hans-Dieter Jundt and Hedwig Jundt v Finanzamt Offenburg ECLI:EU:C:2007:590, para 19; opinion of Advocate General Jacobs in Case C-147/03 Commission v Austria ECLI:EU:C:2005:40, para 53; opinion of Advocate General Geelhoed in Case C-209/01 The Queen, on the application of Dany Bidar v London Borough of Ealing and Secretary of State for Education and Skills ECLI:EU:C:2004:715, paras 24, 31. 149 See A Wunder, ‘The Usage of Solidarity in the jurisdiction of the ECJ’ in L de Almeida, M Cantero, M Durovic and KP Purnhagen (eds), The Transformation of Economic Law (Oxford, Hart Publishing, 2019). 150 ibid.

310  Kai P Purnhagen In EU law, the value of the principle of solidarity increased with each Treaty amendment.151 Nowadays, the principle of solidarity is in most of the cases pronounced as solidarity between Member States:152 In this sense it either refers to a political process conducted in a manner conducive to solidarity153 or it is related, as in Articles 122, 194 of the treaty, to the distribution of resources among the member states. The ties between the member states and the EU are specially emphasised in the solidarity clause in Article 222 of the Lisbon treaty.154

In the sense of managing diversity, the Treaty indeed recognises this principle as a major governance tool. However, the Treaty also mentions this principle as a governing tool between individuals in the EU. In the operative part of the Lisbon Treaty, Article 3 particularly mentions the fostering of solidarity among generations and the use of solidarity155 to encourage mutual respect among peoples.156

v.  Increasing Use of Findings from Behavioural Science The focus on individual protection at the EU level may result in levelling the legislative protection for individuals across the Union and thereby interfering with the different stages of economic development in the Union and the different values across the Union which have developed from different levels of legal protection and social custom.157 Hence, when increasingly protecting individuals at the EU level, the EU law-makers needs to know where and when to intervene with different preferences in the Union to avoid resistance against overregulation. In this sense, knowing how people react to certain types of intervention can prove to be essential for legitimately and responsively regulating individual rights at the EU level.158 In areas such as EU food labelling law where the EU increasingly regulates vulnerability levels at the EU level, it indeed already makes reference to insights from behavioural science.159 Furthermore, the line of argumentation of the CJEU, when interpreting provisions more in line with individual rights protection at the EU level, increasingly follows the line of argumentation of insights gained from behavioural science.160 151 ibid. 152 ibid. 153 Art 24(3), Art 31, Art 67(2), Art 80. 154 Wunder, above n 149. 155 Art 3(3 II). 156 Art 3(3 III). 157 See for this effect on the example of harmonisation of EU product safety law, McGee and Weatherill, above n 32; Purnhagen, ‘Beyond Threats to Health’, above n 24. 158 K Purnhagen, ‘Why Do We Need Responsive Regulation and Behavioural research in EU Internal Market Law?’ in K Mathis (ed), European Perspectives on Behavioural Law and Economics (Dordrecht Springer Science, 2015). 159 Purnhagen and van Kleef, above n 136, 153–54; Schebesta and Purnhagen, ‘Island or Ocean’, above n 73. 160 H Schebesta and K Purnhagen, ‘The Behaviour of the Average Consumer: A little Less Normativity and a Little More Reality in the Court’s Case Law? Reflections on Teekanne’ (2016) 41 European Law Review 590.

From Supranationality to Managing Diversity  311

V. Conclusion This contribution observed two research questions. First, is there a paradigm shift in internal market policy moving from the establishment of a supranational institution to managing diversity? Second, if so, what are the consequences for EU law? I critically investigated internal market theory, legislation and adjudication. Based on the theory, legislation and cases examined, I concluded that such a shift may indeed be observed. Based on doctrinal analysis, I unfolded how such a shift may affect European Union law. While traditionally the principles of proportionality, subsidiarity and conferral would be designed to govern such a switch, their application is limited mainly to legislation. Instead, the principle of solidarity is a much better fit to govern such an internal market approach, as its application is much wider. With regard to governing the turn towards more individual protection at EU level, insights from behavioural science can help in determining where individual protection can meaningfully be achieved at EU level, and where such protection would be better decentralised.

312

14 The Internal Market in its Historical Context. Has the ECJ ‘Over-Constitutionalised’ the Internal Market? PETER BEHRENS

My contribution will highlight and criticise a recent attack on the alleged ‘overconstitutionalisation’ of the internal market, an attack that has been launched with particular emphasis and widespread attention in Germany. Before going into details, some preliminary remarks about the ‘constitutionalisation’ of the internal market may be useful for a better understanding of its meaning, of the critique thereof and of my counter-critique. Looking at the concept of the ‘internal market’ in its historical context, we should, first, keep in mind that the term was not initially part of the Rome Treaty (EEC Treaty) which instead used the term ‘common market’. The term ‘internal market’ was first coined by the European Court of Justice (ECJ) in its Gaston Schul judgment of 5 May 1982,1 a case about the imposition by a Member State of importation of value-added tax on imported goods supplied by a private person from another Member State where the goods have already been subject to value-added tax by the Member State of exportation. According to the Court, this constituted unlawful internal taxation in excess of the tax imposed on similar domestic products within the meaning of Article 95 EEC (now Article 110 of the Treaty on the Functioning of the European Union (TFEU)). The Court based its holding on the statement that the concept of a common market involved ‘the elimination of all obstacles to intra-Community trade in order to merge the national markets into a single market bringing about conditions as close as possible to those of a genuine internal market’ (paragraph 33). This was the birth of the notion ‘internal market’ which, since the Single European Act of 1986, became an integral part of all later Treaty versions.

1 Case

15/81 Gaston Schul EU:C:1982:135.

314  Peter Behrens We should, furthermore, keep in mind that the ‘internal market’ is not limited to the set of economic freedoms which grant market participants free access to Member States’ markets. These freedoms apply irrespective of the nature and the substance of cross-border transactions, ie, irrespective of whether they involve the trading in goods or services, the movement of self-employed persons or workers, the movement of capital or cross-border payments. The system of freedoms is in fact all-inclusive, because according to Article 57 TFEU, if a transaction is not covered by one of the other freedoms, the freedom to provide services applies as a default rule. This system of freedoms does therefore grant the elimination of all obstacles to intra-Union trade in order ‘to merge the national markets into a single market bringing about conditions as close as possible to those of a genuine internal market’. However, from the very beginning the ‘common market’ and later the ‘internal market’ included a second pillar which was and still is the ‘system of undistorted competition’ which includes the prohibition of cartels, the prohibition of any abuse of a dominant position on the internal market, the control of mergers, the control of State aids and the requirement of non-discrimination in public procurement. Even though the Lisbon version of the Treaty on European Union removed the ‘system of undistorted competition’ from the definition of the goals of the Union in Article 3, nevertheless Protocol No 27, which is part of EU Treaty law, states the opinion of the High Contracting Parties that ‘the internal market as set out in Article 3 of the Treaty on European Union includes a system ensuring that competition is not distorted’. Hence the economic freedoms which eliminate all obstacles to intra-Union trade in order to merge the national markets into a single market as well as the competition rules which grant undistorted competition on the market, are still the two interrelated pillars of the ‘internal market’ neither of which can be understood without the other. Whereas the economic freedoms are drafted as prohibitions addressed to the Member States of the EU so as to prevent measures restricting cross-border transactions between private parties, the competition rules contain prohibitions addressed directly to market participants (undertakings) in order to prevent strategies that would restrict their freedom to compete and limit consumers’ freedom of choice on the market. The important question that arose immediately after the Treaty of Rome came into force was, of course, who should be competent to enforce these prohibitions. The Treaty provided for a right of the Commission (Article 169 EEC, now Article 258 TFEU) as well as of every Member State (Article 170 EEC, now Article 259 TFEU) to bring an action before the European Court of Justice against any Member State which violated any of the prohibitions contained in the economic freedoms; as far as the competition rules were concerned, the Commission was also under an obligation to ensure the proper functioning and development of the common market and, in particular, to ensure that the provisions of the Treaty are applied (Article 155 EEC, now Article 17 of the Treaty on European Union). Consequently, Regulation 17/62/EEC of the Council2

2 [1962]

OJ 13/204; OJ [1959–62] special edn, 87.

The Internal Market as a Constitution Context  315 authorised the Commission to enforce the competition rules directly against undertakings whereby the uniform application of these rules throughout the common market was guaranteed. In sum, the enforcement of the internal market law was in the beginning primarily in the hands of the Community organs, ie, the Commission and the ECJ. The most remarkable development of the EEC’s enforcement system started, however, as soon as the ECJ was asked to give preliminary rulings concerning the interpretation of the economic freedoms as well as of the competition rules. According to Article 177 EEC Treaty (now Article 267 TFEU) the Court was from the very beginning competent to authoritatively determine the meaning of the economic freedoms and the competition rules whenever a court or tribunal of a Member State considered such determination by the ECJ necessary in order to decide the case before it and to give judgment. Article 177 EEC Treaty clearly implied the possibility that national judges, in order to decide a case, might be compelled to apply the rules on which the internal market rests. In other words: the Treaty implied the possibility that these rules might be self-executing or directly applicable. As is well known, this hypothesis was tested as early as 1962 with regard to the economic freedoms in the case Van Gend & Loos.3 The issue was whether the increase of an import duty by the Netherlands authorities on goods imported from another Member State after the entry into force of the EEC Treaty violated the freedom of movement of goods, in particular Article 12 EEC Treaty which prohibited any increase of customs duties on imports during the transitional period at the end of which all customs duties were to be abolished. The ECJ was requested by a Dutch administrative court to give a preliminary ruling, because a Dutch importer contested before that court the applicability to his imports from Germany of a customs duty that had nevertheless been increased by the Netherlands. The ECJ’s decision did clearly not only require an answer to the substantive law question pertaining to the violation of Article 12 EEC but, more importantly, a response to the question whether an individual may derive from an economic freedom laid down in the EEC Treaty individual rights which Member States’ courts must protect. The famous answer of the ECJ was crystal clear: ‘Article 12 of the Treaty establishing the European Economic Community produces direct effects and creates individual rights which national courts must protect’. Having thereby established the direct effect of the economic freedoms, the ECJ had to determine the relationship between the individual rights derived from these freedoms and Member States’ national laws in case of a conflict between the two. This problem was resolved – as is also well known – in the leading case Costa v ENEL.4 The issue in this case was whether an Italian law that nationalised electricity producers and transferred their assets to a newly established National Electricity Board (ENEL) violated, among others, the freedom of establishment, and if so, what the legal consequence should be. Should Italy merely be obliged to

3 Case 4 Case

26/62 Van Gend & Loos EU:C:1963:1. 6/64 Costa v ENEL EU:C:1964:66.

316  Peter Behrens amend its law in order to make it compatible with the Treaty provisions or should the provisions conflicting with the Treaty become immediately inapplicable? The ECJ emphasised once more that the economic freedoms are, to the extent that the relevant rules are complete and unconditional (ie, not depending on further implementing legislation), capable of creating individual rights which national courts must protect and, to the extent that these rights are not contingent, national courts must protect them even where a Member State’s laws provide otherwise. Hence, Member States, being legally bound to the Treaty provisions, were held to be legally unable to unilaterally nullify such individual rights by means of national legislative measures. In other words: the ECJ established in Costa v ENEL the fundamental principle that Community law takes precedence over conflicting national law which becomes not null and void, but inapplicable to that extent. As far as the competition rules are concerned, the ECJ established their direct effect even before van Gend & Loos in the case De Geus v Bosch and Van Rijn.5 The issue was whether a contractual system of distribution of Bosch products, which allocated the territory of each Member State to an exclusive dealer, violated Article 85 EEC Treaty (now Article 101 TFEU). De Geus, a Dutch dealer who was not part of the system, imported Bosch refrigerators from a German company which was in fact part of the system but violated its contractual obligation not to export Bosch products into other Member States. Van Rijn, the exclusive dealer of Bosch products in the Netherlands, brought an action against de Geus asking for injunctive relief and damages based on the argument that the purchase of Bosch refrigerators by de Geus was outside and in contravention of the exclusive dealership system and hence unlawful. De Geus, however, argued in defence that the exclusive dealership agreements and their export prohibitions were violating Article 85(1) EEC Treaty (now Article 101(1) TFEU) and hence void under Article 85(2) of the EEC Treaty (now Article 101(2) TFEU). The defence, in order to use the prohibition of cartels as a shield, was therefore clearly assuming its direct applicability in a private litigation before the courts of the Netherlands. Both the first instance Court of Rotterdam as well as the Court of Appeal of the Hague, where the case ended up, were in agreement that a defendant could in principle rely on Article 85 EEC, the only question being whether this was possible before the Community had passed an implementing regulation on the basis of Article 87 EEC (now Article 103 TFEU) which would authorise the Commission to grant exemptions under Article 85(3) EEC Treaty (now Article 101(3) TFEU). Upon the appellate court’s request for a preliminary ruling, the ECJ stated in very clear terms, that ‘in principle’ Article 85 EEC Treaty had been ‘applicable’ from the time of entry into force of the Treaty, but that only those agreements would be automatically void which, even though they had to be notified to the Commission for an assessment of their exemptability under Article 85(3) EEC Treaty (now Article 101(3) TFEU), had not been so notified. This holding implied at least to

5 Case

13/61 De Geus v Bosch and Van Rijn EU:C:1962:11.

The Internal Market as a Constitution Context  317 that extent the direct applicability of the competition rules by national courts of the Member States. Whereas in De Geus v Bosch and Van Rijn the direct applicability was merely implicit in the ECJ’s acceptance of the request by the Dutch courts for a preliminary ruling, the ECJ stated in the case of BRT v SABAM6 in crystal clear terms that ‘the prohibitions of Articles 85(1) and 86 tend by their very nature to produce direct effects in relations between individuals’ and that ‘these Articles create direct rights in respect of individuals concerned which the national courts must safeguard’. The relationship between the competition rules of the Treaty and national competition rules was an issue put before the ECJ in the case of Walt Wilhelm7 which concerned the famous price-fixing cartel of dyestuffs producers who were prosecuted and sanctioned by the Commission under the Treaty rules as well as by the German Cartel Office under German competition law. Here the ECJ established the principle of parallel applicability of Community and national competition laws subject to the proviso that, if the ultimate general aim of the Treaty is to be respected, this parallel application of the national system can only be allowed in so far as it does not prejudice the uniform application throughout the Common Market of the Community rules on cartels and of the full effect of the measures adopted in implementation of those rules.

And the Court went on to emphasise that said Article 87 EEC Treaty (now Article 103 TFEU), ‘in conferring on a Community institution (ie the Commission) the power to determine the relationship between national laws and the Community rules on competition, confirm(s) the supremacy of Community law’. This jurisprudence of the ECJ has in fact ‘constitutionalised’ the internal market in the sense that the individual rights derived from the economic freedoms and the competition rules limit the scope of lawful interference of Member States with private commercial transactions by any legislative or administrative means. The flip side of the principle of supremacy of the economic freedoms and the competition rules is the limitation of Member States’ sovereignty. By signing the Rome Treaty the Member States have given up to an important extent their legislative autonomy: they can no longer take legislative action that is violating any of the prohibitions which constitute the economic freedoms, unless such legislation is appropriate, necessary and adequate (ie, proportionate) to protect a ‘mandatory requirement’ (ie, an imperative public policy); and they can no longer unilaterally change the ‘rules of the game’ which protect the competition within the internal market. For exactly these consequences the ECJ’s jurisprudence has recently come under fundamental attack by the former Justice of the German Constitutional Court, Dieter Grimm. In a book that has drawn widespread attention8 he argues

6 Case 127/73 BRT v SABAM EU:C:1974:25, para 16. 7 Case 14/68 Walt Wilhelm EU:C:1969:4. 8 D Grimm, Europa ja, aber welches? – Zur Verfassung der europäischen Demokratie [Europe, Yes, but which one?: On the Constitution of European Democracy] (München, CH Beck, 2016).

318  Peter Behrens that the impact of the ECJ’s constitutionalisation of the internal market on Member States’ legislative powers goes far beyond any comparable impact of Member States’ national constitutions. Indeed, the individual rights derived from the economic freedoms go far beyond the fundamental rights granted by national constitutions; and the concept of supremacy imposes even in the field of competition law a restriction on national legislation that is unknown in Member States’ legal systems. In other words: matters that would in any national legal system be subject to legislative change by a democratically elected parliament according to changing political preferences are withdrawn from the Member States’ political decision-making processes. This is what Grimm characterises as an ‘over-constitutionalisation’ which has given rise to a gap of democratic legitimation in the internal market. The ECJ’s jurisprudence is said to have restricted Member States’ autonomy to an unintended degree and was allegedly not supported by the ‘political will’ of the people concerned. It is rather considered as an act of self-empowerment by the Court at the expense of the political institutions of the Union (including its Member States) based on a disregard of the intentions of the Parties to the Treaty and on a method of interpretation that inappropriately emphasised the objective teleology of the Treaty without an appreciation of the legal nature of the Treaty as an agreement of public international law. This is, according to Grimm, explaining the widespread dissatisfaction with the path European integration has taken. In order to evaluate the merits of this attack on the ECJ’s jurisprudence it is necessary to first distinguish and understand two different implications of the economic freedoms and the competition rules. On the one hand, they are the basis for legislative powers which primary Treaty law provides for the institutions of the Union in order to enact secondary legislation by way of directives or regulations. Such legislative acts of the Union are, as any legislation, always the result of political compromises within the Union’s institutions and among Member States’ governments all of which are democratically elected. Irrespective of whether such powers are used for the harmonisation of national laws in order to facilitate the use of the economic freedoms or for the implementation of the Treaty’s competition rules, no directive or regulation will be enacted without an appropriate balancing of conflicting political preferences and interests. And what is more: secondary legislation and even the primary law of the Treaty are not unchangeable although it must be admitted that any change is more difficult to achieve than a legislative change on the national level of Member States. It is nevertheless not justified to characterise the internal market law to be devoid of democratic legitimation even though it is less subject to day-to-day political pressure than national laws and democratic legitimation is admittedly less direct than on the national level of Member States. The decision-making process within the institutions of the EU requires and allows nevertheless careful attention to political preferences developing in the Member States. Hence it is equally unjustified to argue that the ECJ’s constitutionalisation of the internal market law is an act of self-empowerment which shields the economic freedoms and the competition rules from politically motivated changes by the competent institutions of the Union.

The Internal Market as a Constitution Context  319 The second aspect which characterises the internal market law relates to the direct effect of the economic freedoms and the competitions rules. This feature of the ECJ’s constitutionalisation of the internal market is an even more inappropriate basis for the argument that the Court has engaged in an act of self-empowerment at the expense of the political institutions of the Union. The contrary is true: by deriving from the Treaty individual rights which national courts have to protect, the Court has empowered the citizens of the Union to take care of their economic activities and transactions by themselves. If one considers this aspect of empowerment in hierarchical terms, the concept of direct effect of internal market law implies a ‘downward’ delegation of powers to the market participants rather than an ‘upward’ delegation of powers to the Commission or the Court itself. These market participants are the ‘people concerned’ whose ‘political will’ should, according to Grimm, have supported the ECJ’s constitutionalisation of the internal market in order for it to gain democratic legitimacy. But did the citizens of the Union not make ample use of their individual rights since the Treaty came into force? Is that not sufficiently reflecting peoples’ approval of the constitutionalisation of the internal market? More fundamentally: is it appropriate to ask for democratic legitimation of the individual freedoms by those who enjoy and make use of these freedoms? And does the principle of democracy really require that such individual freedoms should always and without limits be subject to political interference by democratic majorities? The Treaty already allows that democratically elected national parliaments may overrule the economic freedoms in order protect an important public policy objective provided the protection is proportionate. And the competition rules, including the rules on control of state aids, expressly allow for public service exceptions. It is therefore not at all plausible that, according to Grimm, the lack of a democratic foundation of the ECJ’s concept of direct effect should explain the widespread dissatisfaction with the path European integration has taken. The reasons for such dissatisfaction lie elsewhere. Extremely puzzling is the remedy that Grimm offers to cure the mistake the ECJ has made in his eyes by ‘over-constitutionalising’ the internal market. He suggests that the whole body of the internal market law contained in the Treaty should be ‘de-constitutionalised’, ie, scaled down to the status of secondary legislation of the EU which would then be under a more direct control by the political institutions of the EU (and the Member States) rather than by the Court. This idea totally disregards that secondary legislation is also binding upon Member States. What is more: regulations are expressly made directly applicable in the Treaty (Article 288(2) TFEU) and must therefore be applied by national courts! It looks therefore as if nothing would be gained in terms of a de-constitutionalisation suggested by Grimm. A final word is necessary on Grimm’s critique of the method of interpretation applied by the ECJ in order to establish the direct effect and the supremacy of internal market law. The ECJ is said to have mistakenly interpreted the EEC Treaty like a national constitution and therefore relied on its objective goal (teleological interpretation). Instead, the ECJ should have interpreted the Treaty as an international

320  Peter Behrens instrument governed by the principles of public international law according to which the intentions of the parties are said to be controlling (the assumption being that the parties to the Rome Treaty did not intend to create directly applicable provisions that would take precedence over national law). This is clearly wrong. Article 31(1) of the 1969 Vienna Convention on the Law of Treaties provides that a treaty is 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’. This is exactly what the ECJ has done. In Van Gend & Loos the Court has, in order to determine the meaning of the standstill clause of Article 12 EEC Treaty (no introduction of new customs duties), considered ‘the spirit, the general scheme and the wording’ of the provision. The objective of the Treaty was defined in terms of the establishment of a common market which, according to the Court, implied that the Treaty was more than an agreement which merely created mutual obligations between the contracting States. Article 177 EEC Treaty (now Article 267 TFEU) which provided for a preliminary ruling by the ECJ on the interpretation of the Treaty whenever such a ruling would be necessary to enable a national court or tribunal to render a judgment, was taken by the Court as an indication that the contracting States had acknowledged ‘that Community law has an authority which can be invoked by their nationals before those courts and tribunals’. Furthermore, the wording of Article 12 EEC Treaty was said to provide for a negative obligation of Member States (not to introduce new customs duties) the enforcement of which could not be exclusively left to the Commission or the Member States who were entitled to bring (according to their discretion!) the matter before the Court. The functioning of a common market ‘which is of direct concern to interested parties’ would rather require that this negative obligation gives rise to corresponding individual rights of the market participants concerned, rights which they themselves would be able to enforce before the national courts of the Member States. This reasoning was methodically correct, logically stringent and in substance absolutely convincing. It is inconceivable that a common market could have emerged, if the enforcement of the prohibitions constituting the economic freedoms had become dependent upon initiatives of the Commission or Member States’ governments on a case-by case basis! This consideration is even more compelling with regard to the competition rules of the Treaty: restraints of competition are predominantly coached in terms of private law; they are most of the time contained in or reflected by contractual agreements between private parties (‘undertakings’). If the Rome Treaty expressly provided that an agreement which violates the prohibition of cartels is ‘automatically void’ (Article 85(2) EEC Treaty, now Article 101(2) TFEU), how could the ECJ possibly have denied a party to such agreement the right to invoke this sanction in a civil litigation before a national court? It is again inconceivable that the application of this sanction should depend on an initiative of the Commission or a Member State’s government. In light of the Treaty’s ‘ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose’, the ECJ had in fact no other choice than to acknowledge that, ‘the prohibitions of

The Internal Market as a Constitution Context  321 Articles 85(1) and 86 tend by their very nature to produce direct effects in relations between individuals’ and that ‘these Articles create direct rights in respect of individuals concerned which the national courts must safeguard’.9 In sum, Grimm’s critique of what he characterises as the ECJ’s ‘overconstitutionalisation’ of the internal market is based on a fundamental lack of understanding of the indispensable legal underpinnings of an internal market within the framework of the EU. Far from a ‘self-empowerment’ of the Court at the expense of the political institutions of the EU (including its Member States), its jurisprudence has merely unfolded what may have initially been hidden in text of the Treaty by following generally accepted rules of interpretation. The twin concepts of direct effect and supremacy of the economic freedoms and the competition rules admittedly but deliberately reduce the scope of political intervention by Member States’ governments or parliaments with the free flow of goods, services, capital and persons across national borders or with undertakings’ freedom to compete for the benefit of consumers. But: if the High Contracting Parties really meant what they laid down in the Rome Treaty, ie, the establishment of a common (internal) market, they must have also meant to accept the indispensable legal underpinnings thereof as unfolded by the ECJ’s interpretration of the Treaty. Dismantling the constitutionalisation of the internal market and its re-politisation would seriously undermine the exceptionally successful legal and economic development within the EU. Democracy is not the only basic pillar on which European integration is founded; the importance of the rule of law (as reflected by the constitutionalisation of the internal market) cannot be overestimated as the second pillar and the ECJ should rather be credited for its farsighted wisdom in this regard.



9 See

above, n 6.

322

15 Originalism at the European Court of Justice GARETH DAVIES

I. Introduction There are various criticisms made of the European Court of Justice (the Court), of which the most common tend to share a lot of ground: they cluster around the idea that the Court pushes EU law too far. This chapter does not make an argument about whether those criticisms are fair or justified.1 Rather, the chapter tries to unpack the nature of the criticisms, and consider what kind of action or change would disarm them. The core point of the chapter is to challenge the coherence of a common narrative about the Court. That narrative would present the Court as activist, and integrationist, and purposive, and complain that it behaves more like a political organ than a legal one, constantly pushing the text beyond what it can bear.2 The narrative would then go on to suggest that in order to have a more legitimate and limited legal order, one which respects divisions of power between the EU and Member States, the Court needs to become more cautious and disciplined, and behave more like a traditional court, sticking to the letter of the law and leaving politics to politicians.3 On the contrary, it is suggested here that defining characteristics of the Court’s approach to EU law are that it is backward-looking, static and unresponsive to social change. It is based on a vision of the Treaties which was bon ton in rightthinking EU circles 60 years ago – when these were the circles that mattered – but

1 The author’s view, in the interests of transparency, is that sometimes they are and sometimes they are not. 2 eg, G Conway, The Limits of Legal Reasoning and the European Court of Justice (Cambridge, Cambridge University Press, 2012); H Rasmussen, On Law and Policy in the European Court of Justice (Dordrecht, Nijhoff, 1986). For a balanced overview of the discussion see M Dawson, B de Witte and E Muir (eds), Judicial Activism at the European Court of Justice (Cheltenham, Edward Elgar, 2013). 3 See S Sankari, European Court of Justice Legal Reasoning in Context (Europa Law Publishing, 2013) 3–18 for discussion.

324  Gareth Davies resonates rather weakly with the general public now. If one objects to the undoubtedly far-reaching, disruptive and integration-oriented approach to EU law which the Court takes then one is in substance calling for it to come into the present: to abandon its original conception and develop a new one, more in tune with the times. To become, in short, politically responsive and dynamic in its approach to the law. In a neat reversal of the American debate, Eurosceptics and conservatives really want – whether they know it or not – a more contemporary and political court. Europhile integrationists, by contrast, want a court that continues to adhere to the principles and vision which it developed half a century ago and has been applying ever since. The American debate is mentioned because an essentially American constitutional idea will be used here to develop the argument – the idea of originalism. Originalist interpretations of the constitution are often contrasted with those which treat it as a living document, and when those approaches are compared with Europe it is treated as self-evident that the Court of Justice’s teleological interpretative style is almost opposite to originalism. Conservative, often textualist, American originalists are seen as judicial mirror images of the progressive, purposive, European Court.4 Here it will be argued, as the paragraph above suggests, that the Court of Justice is closer to the originalist end of the judicial scale than to that of the living constitution. Its most striking characteristic is a refusal to change. Why does this matter, beyond the obvious and significant pleasure of making a counterintuitive academic argument? It matters partly because the diagnosis affects the treatment. Whether one takes the view that it is important to expand the scope of EU law, or to rein it in, this has institutional consequences. The various tools potentially available to steer the law – textual change, appointments to the Court, political pressure, judicial resistance, amendment of institutional competences – can only be effectively deployed if the causes of EU law’s current state are understood. It matters also because it shines a certain light on the nature of debates about the Court and its law. Institutional and social conservatism are allied with legal dynamism and politicisation, whereas the desires for social and institutional deepening of the EU are allied with legal consistency and conservatism. In a time of populism, these links are not entirely strange – formalism as a way of resisting autocracy’s tendency to instrumentalise and politicise law has a long and distinguished history5 – but they are not fully appreciated in EU law’s internal debate, where the self-understanding of each side is sometimes as misguided as their understanding of the other. The chapter proceeds in steps. The following section sketches the contrast between originalist and other approaches to interpreting constitutions. The next 4 Conway, above n 2; Sankari, ibid 63. 5 M Bobek, ‘Conclusions: Of Form and Substance in Central European Judicial Transitions’ in M Bobek (ed), Central European Judges Under the European Influence: The Transformative Power of the EU Revisited (Oxford, Hart Publishing, 2015); CR Sunstein, ‘Must Formalism Be Defended Empirically?’ (1999) 66 University of Chicago Law Review 636.

Originalism at the European Court of Justice  325 section outlines common criticisms of the Court. The following two show what is meant by the static and backward-looking nature of EU law. A conclusion suggests policy consequences.

II.  Originalism and Living Constitutionalism Originalism, in the context of constitutional interpretation, is, most simply, the idea that the constitution should continue to bear the meaning that it had when it was first adopted unless and until it is amended by the appropriate procedure.6 It is quite often contrasted with the most obviously opposing position, living constitutionalism, which claims that constitutions should be interpreted in their current social and political context, so that their meaning may change as the world around them changes.7 There are of course infinite varieties of originalism and of anti-originalism, addressing matters such as how the original meaning is to be defined or determined, or, on the other side, the extent and manner in which context should be injected into interpretation. There are also rich scholarly debates on the relationship between these interpretative approaches and other legal doctrines, such as precedent, formalism and textualism.8 The variety of possible positions and nuances is considerable, and probably all of them are adhered to by someone, and most American constitutional scholars and judges would not consider themselves either unmitigated originalists or anti-originalists. However, most would agree that there is a central dichotomy around which these debates revolve, and that at the heart of this is the role of the judge. For an originalist, when a text is adopted it bears a certain meaning, and that is the law, and it is for judges to apply that law, but not to change it: change is for the legislature.9 The fear, often voiced, is that if judges can reinterpret to reflect social change then this will in practice amount to giving judges personal power over the law, and turning it into no more than a vehicle for their political views. Originalists often present their standpoint as a democratic one, reflecting the distinction between those who make the law, and those who apply or adjudicate it.10 The alternative view regards the rule of the dead as normatively unacceptable and suggests that a constitution needs to be read in its current social context, 6 A Scalia, A Matter of Interpretation, A Gutmann (ed) (Princeton, NJ, Princeton University Press, 1997). 7 L Tribe, ‘Comment’ in A Scalia, A Matter of Interpretation, A Gutmann (ed) (Princeton, NJ, Princeton University Press, 1997); B Ackerman, ‘The Living Constitution’ (2007) 120 Harvard Law Journal 1738. See discussion in A Scalia, ‘Originalism: The Lesser Evil’ (1989) 57 University of Cincinnati Law Review 849, 852; cf R Dworkin, ‘Comment’ in A Scalia, A Matter of Interpretation, A Gutmann (ed) (Princeton, NJ, Princeton University Press, 1997). 8 See works, above n 7. 9 Scalia, ‘Originalism: The Lesser Evil’, above n 7. 10 ibid.

326  Gareth Davies reflecting current social understandings and attitudes.11 Partly this is because constitutions are hard to change, so originalism amounts to a significant constraint on majoritarian democracy, and sometimes a bizarre one, for example, were a contemporary American judge to take a seventeenth-century American nobleman’s view of privacy or cruel punishment as the basis for current constitutional interpretation of those ideas.12 However, it is also a discussion about the nature of law: for the believer in a living constitution, the meaning of a legal text changes as the meanings of words and principles change in the world. Thus, while the legislature fixes the text, it is the task of the judge in a case to interpret – to give that text meaning, drawing on the context as a whole.13 Law is constructed by legislatures and courts, and perhaps even by society, together. Originalism is generally associated with social conservatism, because the United States in the seventeenth century was broadly more conservative than it is now, and because it protects certain attitudes from social developments which threaten them, by entrenching those attitudes in law and denying the legitimacy of change – unless that change comes via the extremely difficult route of constitutional amendment. As such, it is part of the culture wars, a way of resisting the ascendance and dominance of the values and social understandings of those viewed as a liberal elite. All this is however contingent. It is quite possible that a meaning which at the time of adoption was orthodox and unproblematic comes, as a result of social change, to be radical. Understandings of privacy and individual freedom might, in an age of internet and the security-oriented state, come to threaten business models and surveillance institutions, and it might be the left pleading for originalist interpretations and the right for new understandings to reflect the new age. The principled constitutional scholar may accept that their constitutional philosophy can lead to legal conclusions that they personally regret – but the law is the law. The political activist may, by contrast, see constitutional approaches as tools to be deployed to achieve desirable social change. The ‘right’ approach to interpretation is the one that achieves the ‘right’ result.

III.  The Criticisms of the Court: Activism and More The Court has been called activist – meaning something like that it reads the Treaty in a way that prioritises certain political goals, rather than the most plausible or clear meaning of the text.14 It is also sometimes accused of having insufficient 11 See above, n 7. 12 A Reed Amar, ‘Introduction, xvii’ in A Scalia, A Matter of Interpretation, A Gutmann (ed) (Princeton, NJ, Princeton University Press, 1997). 13 Ackerman, above n 7. 14 See above, n 2; see also M Poiares Maduro, ‘In Search of a Meaning and not in Search of the Meaning: Judicial Review and the Constitution in Times of Pluralism’ (2013) 54 Wisconsin Law Review 541.

Originalism at the European Court of Justice  327 reasoning – its judgments are concise, to some readers cryptic, and while they invariably refer to other past judgments they do not contain a full analysis of the different possible legal perspectives or all the interests at stake.15 It is also, implicitly, accused of over-constitutionalising.16 That is a currently fashionable critique of the state of EU law, but it is one where the Court must surely share the blame. The essence of the critique is that too many policy choices are fixed in EU law, because they are embedded in Treaties which are very hard to change. Matters which ought to be the subject of political debate are instead constitutionalised, and then no longer meaningfully contestable. To the extent that this is true, it is because the Court has chosen to interpret the Treaties in a certain way. It is hardly controversial to claim that most of the salient Treaty articles are relatively open in their wording, and could plausibly bear different meanings. What is a market? What is free movement? What, indeed, is necessary? The Court might have reacted to this open texture by declaring that it was for the legislature – perhaps in conjunction with itself – to fill in these meanings, and that it would defer to political interpretations, perhaps within limits. Instead, it chose to provide its own definitions, to gradually refine them, and to expect the legislature to be constrained by them.17 It claimed ownership of the text. That the Treaties have taken certain matters out of politics is because of the way the Court has understood the Treaties: if the EU is over-constitutionalised in this sense, it is because the Court has over-constitutionalised it.18 A constitution with an open and contestable meaning would not have been a political or democratic problem. All three criticisms portray the Court as a bull in a political and constitutional china shop. It charges forward, pushing an integrationist agenda that is hardly inevitable from the text, and without considering social or political context, nor legitimising its actions by the traditional judicial discursive route of providing persuasive reasons. It does not seem to be constrained by the literal words of the Treaty, nor by the views of society or the legislature, nor by any sense that it must justify its decisions in a non-self-referential way. It must be noted that the claim discussed here is that the Court pushes an agenda – not that it achieves it. For whatever it may decide, Member States have techniques much discussed in the political science literature for resisting its pressure.19 Compliance and the capacity of law to cause societal change are 15 See Conway, above n 2. 16 D Grimm, ‘The Democratic Costs of Constitutionalisation: The European Case’ (2015) 21 European Law Journal 460; FW Scharpf, ‘De-Constitutionalisation and Majority Rule: A Democratic Vision for Europe’ (2017) 23 European Law Journal 315. 17 G Davies, ‘The European Union Legislature as an Agent of the European Court of Justice’ (2016) 54 Journal of Common Market Studies 846. 18 G Davies, ‘Does the Court own the Treaties? Intepretative pluralism as a solution to overconstitutionalisation’ (2018) 24 European Law Journal 358. 19 See eg D Martinsen, ‘Judicial Influence on Policy Outputs? The Political Constraints of Legal Integration in the European Union’ (2015) 48 Comparative Political Studies 1622; L Conant, Justice Contained: Law and Politics in the European Union (Ithaca, NY, Cornell University Press, 2002).

328  Gareth Davies another issue. However, if we regard the Court as a constructor of doctrine, whatever the extra-legal consequences of that doctrine, then most of the traditional lawyers’ criticisms portray it exceeding, or at least stretching, its mandate. It is then not surprising that studies of the Court’s interpretative style have tended to put it far away from originalism on the scale of possibilities.20 Its teleological interpretation prizes the effectiveness of the law above all else, and in this sense is forward-looking, and unfettered by the practices and constraints of the past.21 As a court which accepts partial responsibility for a political, integrative, mission, subordinating text to policy imperatives in that aim, it seems an originalist’s bogeyman.

IV.  The Static Nature of EU Law And yet, it is also true that while EU law’s effects march on, the law itself is remarkably static, in several ways. Taking the internal market as an example, instances of the Court changing its mind, or adopting a new direction in the law are extremely rare. Keck is the most famous, and it is striking that it is a case almost without consequences: the reason why the Court excluded selling arrangements from the category of obstacles to free movement was because they almost always had a minimal effect on trade while serving a legitimate social purpose which justified their existence.22 To engage in endless judicial review of national measures which would almost always turn out to be compatible with the Treaty was a waste of judicial resources.23 Keck is not really a rule about what kinds of things are allowed or prohibited, but about evidential presumptions. Contrary to the Court’s suggestion that it was changing its mind, it did not really change very much. We see this in the fact that the principles in the major cases which preceded it, Dassonville and Cassis, survived and continued and have never been called into question.24 Another example where change is sometimes claimed is in the law on citizenship, where it has been suggested that Dano and Alimanovic represented a move away from a deeper, more expansionist vision of citizenship expressed in the earlier Martinez-Sala and Grzelczyk.25 This is contested, yet the fact that a shift 20 See Conway, above n 2; Sankari, above n 3, 63. 21 For an overview of the construction of the dynamic image of the Court as a motor of integration, see R Byberg, ‘The History of the Integration Through Law Project: Creating the Academic Expression of a Constitutional Legal Vision for Europe’ (2017) 18 German Law Journal 1531. 22 Joined Cases C-267/91 and C-268/91 Keck and Mithouard EU:C:1993:905; G Davies, ‘The Court’s Jurisprudence on Free Movement of Goods: Pragmatic Presumptions, Not Philosophical Principles’ (2012) 2 European Journal of Consumer Law/Revue Europeenne de Droit de la Consommation 25. 23 ibid. 24 Case 8/74 Procureur du Roi v Benoît and Gustave Dassonville EU:C:1974:82; Case 120/78 Rewe-Zentral AG v Bundesmonopolverwaltung für Branntwein EU:C:1979:42. 25 Case C-333/13 Dano EU:C:2014:2358; Case C-67/14 Alimanovic EU:C:2015:597; Case C-85/96 Martinez-Sala EU:C:1998:217; Case C-184/99 Grzelczyk EU:C:2001:458; N Nic Shuibhne, ‘Limits rising, duties ascending: The changing legal shape of Union citizenship’ (2015) 52 Common Market

Originalism at the European Court of Justice  329 in legal approach was arguable made the cases stand out, and is part of the reason for the intense discussion around them.26 This was fuelled by the fact that they seemed to be in tune with the political mood, and to show a certain responsiveness (perhaps regrettable) by the Court.27 If so, these cases might indeed be an example of a living constitution approach to the law. However, these are isolated examples in a general trend in which most of the law, and most of its impact, comes from repeated application of old-established principles. Where there is controversy, it is usually because the Court is sticking to a rule, not because it is going in a new direction. Dassonville expressed the idea that it is not form, but effects which matter, and that effects should be understood in a commonsensical way: is there, or might there be, an actor whose movement is made more difficult? Laval, Viking, the free movement of medical patients, and the cases on Free Movement of Capital are just applications of this simple approach.28 Cassis expressed the idea that each actor in the market should be regulated by their home state.29 The law on services, the Services Directive and Centros are applications of this.30 In all of these controversial situations arguments were made that the situation was different – trade unions were special, medical care was special, free choice of company location threatened models of incorporation, services deregulation would undermine society, removing all obstacles to investment would amount to dis-embedding capitalism. Whatever one might think about the merits of these arguments, or about the Court’s final decisions, the characteristics of the judgments are that faced with a new context, and a plausible argument for a new approach, the Court instead stuck to its old one. Free movement law is really just footnotes to Dassonville.31 Even in the law of citizenship and resources, where, as discussed above, there is the suggestion of different phases in the law with different approaches, it is Law Review 889; A Iliopoulou-Penot, ‘Deconstructing the Former Edifice of Union Citizenship? The Alimanovic Judgment’ (2016) 53 Common Market Law Review 1007; D Thym, ‘The Elusive Limits of Solidarity: Residence Rights of and Social Benefits for Economically Inactive Union Citizens’ (2015) 52 Common Market Law Review 17. 26 G Davies, ‘Has the Court Changed or Have the Cases? The Deservingness of Litigants as an Element in Court of Justice Citizenship Adjudication’ (2018) 25 Journal of European Public Policy 1442; H Verschueren, ‘Preventing “Benefit Tourism” in the EU: A Narrow or Broad Interpretation of the Possibilities Offered by the ECJ in Dano?’ (2015) 52 Common Market Law Review 363. 27 M Blauberger et al, ‘ECJ judges read the morning papers. Explaining the turnaround of European citizenship jurisprudence’ (2018) 25 Journal of European Public Policy 1422; C O’Brien, ‘The ECJ sacrifices EU citizenship in vain: Commission v United Kingdom’ (2017) 54 Common Market Law Review 209. 28 G Davies, ‘Tough Love in the Internal Market’ in F Amtenbrink, G Davies, D Kochenov and J Lindeboom (eds), The Internal Market and the Future of European Integration: Essays in Honour of Lawrence W Gormley (Cambridge, Cambridge University Press, 2019). 29 Case 120/78 Rewe-Zentral, above n 24. 30 G Davies, ‘The Services Directive: Extending the country of origin principle and reforming public administration’ (2007) 32 European Law Review 232; C-212/97 Centros EU:C:1999:126. 31 LW Gormley, ‘Inconsistencies and misconceptions in the free movement of goods’ (2015) 40 European Law Review 925; JHH Weiler, ‘The Constitution of the Common Market Place: Text and Context in the Evolution of the Free Movement of Goods’ in P Craig and G de Búrca (eds), The Evolution of European Law (Oxford, Oxford University Press, 1999); Davies, ‘Tough Love’, above n 28.

330  Gareth Davies hard to see that any of the twists and turns are not comfortably encompassed by the Court’s first substantive judgment on the issue, Baumbast, where it stated its vision: In any event, the limitations and conditions which are referred to in Article 18 [TEU] and laid down by Directive 90/364 [now copied into 2004/38] are based on the idea that the exercise of the right of residence of citizens of the Union can be subordinated to the legitimate interests of the Member States. In that regard, according to the fourth recital in the preamble to Directive 90/364 beneficiaries of the right of residence must not become an ‘unreasonable’ burden on the public finances of the host Member State. However, those limitations and conditions must be applied in compliance with the limits imposed by Community law and in accordance with the general principles of that law, in particular the principle of proportionality.32

Nothing since could be described as a break with this approach, or even an amendment of it. This is not to say that there are no doctrinal developments. Ruiz Zambrano, U-turns, the cases on use of goods, all brought new situations before the Court and resulted in new rules being added to the law.33 However, they were extensions, not contradictions, of what came before. They do not represent a turning away from earlier law or ideas, but a continuation of them. Criticism of those judgments would not be ‘why is the Court taking this new direction?’ but ‘why is the Court still following the path of ever-more and ever-deeper?’ For underlying the fundamentally static nature of the substantive law is a deeper stasis: that of the underlying vision of the market and of free movement. The Court’s understanding of this field of law is driven by the idea of ongoing and open-ended integration. Interpretation is guided by the idea of effectiveness – that the law should mean whatever is necessary to get its job done. Free movement is understood as imperative, as urgent, and the burden of adaption to its requirements is placed on the Member States, either directly, through negative harmonisation, or through positive harmonisation. There is an absence in the case law of any suggestion that there should be limits to movement – except in de minimis situations, which, by definition, barely matter – or that we should accept that the existence of diverse institutions and regimes is a good in itself which entails certain constraints on ease of movement.34 Sometimes movement gives way to national institutions in the law, but this is contingent and temporary, not a desirable end point. The vision of the Court is one in which States redesign themselves, or are redesigned, to fit the requirement of complete inter-State openness. This may happen slowly and carefully, with respect for important interests, but the goal is clear. 32 Case C-413/99 Baumbast v Secretary of State for the Home Department EU:C:2002:493, paras 90–91. 33 Case C-34/09 Ruiz Zambrano EU:C:2011:124; Case C-256/11 Dereci EU:C:2011:734; Case C-82/16 KA EU:C:2018:308; Case C-370/90 Singh EU:C:1992:296; Case C-127/08 Metock EU:C:2008:449; Case C-110/05 Commission v Italy EU:C:2009:66. 34 G Davies, ‘Internal market adjudication and the quality of life in Europe’ (2015) 21 Columbia Journal of European Law 289.

Originalism at the European Court of Justice  331

V.  The Choice for Europe The argument above can be extrapolated and made more general. It is not just the Court’s vision of the market which is static, but its vision of the nature of EU law, and indeed of the European Union. The narrative would then be something like this: although the Treaties could be read in different ways, early in its career, the Court made a central interpretative choice. It chose to read them as the expression of a policy programme, a legal plan to achieve change, rather than as a mere statement of rules.35 By signing, the Member States wished to commit to self-transformation, away from their nationalistic, self-destructive past and their damaged present. The means of this was to be integration, and the individual articles were there, in the Court’s eyes, to achieve this broader goal. Implicitly, the Treaties were a legal mandate to do what was necessary to make a wounded Europe whole again, and enable it to achieve its destiny, or at the very least to be peaceful, prosperous and free. This teleological understanding, as it is usually described, has the distinctive consequences so familiar to EU lawyers: Treaty articles (serving integration) are read broadly, whereas exceptions (limiting it) are read narrowly; the law is read holistically, so that the interpretation of a provision may be influenced by very different provisions elsewhere, because each part is understood as contributing to a single whole, and must be made coherent with that whole; and where the Treaty fails to provide explicitly for a rule that the Court considers necessary to make the whole function properly, that rule will then be found, as a matter of logic, to be present by implication – for it would be nonsensical to read the Treaty so that it did not work.36 And even while a rejection of constitutional primacy pur sang becomes the orthodoxy in every court and textbook in Europe,37 the Court does not budge an inch: the constitutional principles of EU law are non-negotiable, whatever the constitutional responses – coming from the Europe which must be changed – to these may be.38 All this is quite defensible – indeed, EU lawyers have defended it with great success for decades.39 For one thing, there are textual invitations to this

35 M Poiares Maduro, ‘Interpreting European Law’ in H Koch, K Hagel-Sørensen, U Haltern and JHH Weiler (eds), Europe: The New Legal Realism (Copenhagen, Djøf Publishing, 2011); cf Scalia, A Matter of Interpretation, above n 6, 40, on the US Constitution. 36 C-6/90 Francovich EU:C:1991:428; E Herlin-Karnell, The Constitutional Dimension of European Criminal Law (Oxford, Hart Publishing, 2012). 37 A Bobić, ‘Constitutional Pluralism Is Not Dead: An Analysis of Interactions Between Constitutional Courts of Member States and the European Court of Justice’ (2017) 18 German Law Journal 1395; G Beck, ‘The Lisbon Judgment of the German Constitutional Court, the Primacy of EU Law and the Problem of Kompetenz‐Kompetenz: A Conflict between Right and Right in Which There is No Praetor’ (2011) 17 European Law Journal 470. 38 JHD Stone, ‘Agreeing to Disagree: The Primacy Debate between the German Federal Constitutional Court and the European Court of Justice’ (2016) 25 Minnesota Journal of International Law 127. 39 See generally Poiares Maduro, ‘Interpreting European Law’, above n 35; Sankari, above n 3.

332  Gareth Davies approach in the Treaty, such as the preambles, the commitment to creating a Union of certain values, the broad statements of Union purposes, and the ‘evercloser Union’ provision now in Article 1 TEU. It is one thing to be suspicious of ‘intention’ as discovered in travaux preparatoires, but when the intention of a Treaty is expressed within that Treaty it is hard to blame judges for taking it into account. Teleology and textualism are sometimes contrasted, but here the teleology is founded in the text. However, the point of this chapter is that the Court’s vision of EU law, just as the vision of the market which corresponds to it, is rooted in the past and not the present. When the Court first formulated its understanding, it would have been a reasonably orthodox one among the relevant legal community of the time – not the understanding shared by all, certainly, but a mainstream one held by many reasonable and informed people involved in EU law, and one which quickly became predominant.40 The first judges of the European Court of Justice – just like the first leaders of the Commission – were drawn from a pool of educated, internationally oriented Europeans and among these the view that European integration was a moral imperative, and the Treaties were tools to achieve this vital goal, would not have been strange.41 Certainly there will have been diversity of opinion, and some will have disagreed while others will have been nuanced. However, the Court’s early understanding was rooted in its moment, in the relevant legal community, and in the text. The trouble is, in the meantime more than half a century has passed and the politics of Europe have changed. If the Treaties were adopted today, and a court was formed from internationally minded European lawyers to adjudicate them, it seems very likely that a more restrained, more technical, less messianic and more balanced approach to the text would be taken, in which the policy freedom of Member States, and the precise wording of articles, carried greater weight. Europe would then not be built as it has been, but the prevailing social mood and attitudes towards international law and organisations would probably be reflected in the law. Part of that change, at least among lawyers, may be because of the very success of the Court’s law: it is because integration has gone so far that some feel it should be stopped. A broad reading in the past leads some to conclude that it is time for a narrower reading now. Thus among the many effects of EU law, it has shaped European society to the extent that reactions to that same law are now different, in a sort of constitutional reflexivity. One does not, therefore, have to think that the Court was ‘wrong’ in its early approach to the Treaties to think that a different approach would be better today. Rather we might say, if we were living constitutionalists, that each reading has its time, and that now it is time to revisit some of the constitutional choices which have served so well. 40 See discussion in M Rasmussen, ‘Revolutionizing European law: A history of the Van Gend en Loos judgment’ (2014) 12 International Journal of Constitutional Law 136; A Vauchez, ‘“IntegrationThrough-Law”: Contribution to a Socio-History of EU Political Commonsense’ (2008) EUI Working Paper No RSCAS 2008/10; also W Phelan, ‘The Revolutionary Doctrines of European Law and the Legal Philosophy of Robert Lecourt’ (2017) 28 European Journal of International Law 935. 41 Rasmussen, above n 40; Vauchez, above n 40; and Phelan, above n 40.

Originalism at the European Court of Justice  333

VI. Conclusion When critics complain that EU law seems never to stop invading new fields, they are really complaining that it never changes. Rather than constantly expanding its interpretations, the Court takes fixed and long-standing interpretations and constantly applies them to new situations. It hammers away with its post-war constitutional tools at the rock of Member State law, and with each blow it exposes new substrates and creates new fissures. The cry of the critic is that surely it is time to pick up a new tool, to polish and repair, to smooth over. A sculpture is not made with one technique alone. By contrast, the Court goes deeper and deeper with its beloved constitutional hammer and chisel, as if no other way of working could ever cross its mind. The power of free movement law, for example, is simply that its rules are widely applicable, strict and unchanging. It is not judicial activism that is putting Member States on the back foot – it is dogged judicial consistency. If, by contrast, the Treaties were a living constitution, the Court would be asking whether its ideas should change; whether the constitutional reading which fitted the mood, the consensus and the needs of the post-war period is necessarily the one which best fits society today. If changing circumstances may justify changing constitutional interpretations then the importance of the insight that integration proceeds in stages – which the Court intermittently notes42 – would lie in the corollary that constitutional law may do so too. There would be adaption of the law and of the rules – or there would be at least recognition of the possibility, and discussion of why the old readings still fit the present day. This is not what we see. The Court is path-dependent, loyal to its original readings of the law, and does not recognise any reasons why this might be different. To the frustration of those who want a European Union that is contemporary, adaptive and politically constituted, and the fury of those who above all love national law, it adheres to the original understanding of the Treaties even as the consequences of that understanding become ever harder to bear. It is, in this sense, an originalist court, and therein lies the crisis of European law. That is not to say that the judges would self-identify with this philosophy – although many might acknowledge their loyalty to the original ideas of Europe – but that in practice it behaves in this way. There is a lesson here for those would like to see the Court reined in. They tend to accuse it of being activist, or political, and wish it were more conservative. What they miss is that a reading and a Court which may have been activist in the 1960s have become, through time and the absence of change, conservative. The case law has taken on a nostalgic tone, and the Court has become a backwardlooking organ. If we think it should not be, then we need to inject it with activism once again – appoint more activist judges. It is only when the law becomes political, and the constitution comes to life, that it is likely to respond and adapt to the Europe of today. 42 Case 283/81 Cilfit and Others EU:C:1982:335; Case C-379/15 Association France Nature Environnement EU:C:2016:603.

334

16 The ‘Fundamental Freedoms’ and (Other) Fundamental Rights: Towards an Integrated Democratic Interpretation Framework SACHA GARBEN

I. Introduction For provisions so long-standing and central to the project of European integration as those on the free movement of goods, services, workers, capital and the freedom of establishment, the degree of disagreement on their precise meaning and application – as also evidenced in the various contributions to this book – is somewhat startling. The uncertainty is deep and broad, comprising technical, theoretical and normative issues. This is, of course, because (even) after all this time, we (still)1 don’t know what these provisions, in the bigger scheme of things, are – and should be – for. This translates into endless discussion about, for instance, to what extent ‘rules whose effect is to limit commercial freedom

1 It could be argued, in line with Gareth Davies in ch 15 in this volume, that there once – in the early days of European integration – was a more pronounced ‘vision’ on what EU law and the provisions on goods, establishment, services, workers and capital were to mean, and that the current uncertainty comes from the fact that this vision no longer corresponds to current social and political thinking. ‘It is based on a vision of the Treaties which was bon ton in right-thinking EU circles sixty years ago – when these were the circles that mattered – but resonates rather weakly with the general public now’. While the integrationist approach of the CJEU was clear (as it still is, as Davies also argues in ch 15), and indeed more broadly supported by EU circles, as well as legitimised by a ‘permissive consensus’ on European integration by the general public, I am not sure that there really was more of a uniform, substantive ‘vision’, even back then, about the position of these provisions in relation to the core questions of state intervention and regulation in the market and society, economic freedom, non-market interests and human rights. Integration coûte que coûte, which was and still appears to be driving the Court, is not much of a vision, and cannot carry anything that would be labelled ‘constitutional’. For that latter point, see M Cahill, ‘European Integration and European Constitutionalism: Consonances and Dissonances’ in D Augenstein, ‘Integration Through Law’ Revisited: The Making of the European Polity (Surrey, Ashgate, 2012).

336  Sacha Garben even where such rules are not aimed at products from other Member States’2 should be caught as prima facie restrictions,3 what type of justification test should be applied to ‘restrictive’ public interest rules of all sorts,4 how the free movement provisions should be interpreted in relation to each other,5 as well as fundamental human/social rights,6 and whether their ‘negative reach’ should correspond to the EU’s ‘positive’ competence to adopt harmonising legislation – opening in turn another can of worms about the limits of Article 114 of the Treaty on the Functioning of the European Union (TFEU) and the place of nonmarket interests in internal market law, and the choice between minimum and maximum harmonisation.7 All this could even be said, only half-jokingly, to translate into uncertainty as to how to actually even name these provisions. Are they ‘free movement provisions’ – which would suggest an emphasis on equal treatment rather than general economic freedom; ‘internal market provisions’ – which place them alongside the legal bases for positive integration and suggests an emphasis on their joint contribution to the same project and thus an argument for convergence; ‘four freedoms’ – which, apart from arguably miscounting them,8 suggests an emphasis on economic freedom rather than equal treatment; or ‘fundamental freedoms’ – which suggests that they are fundamental rights and deserve to be interpreted as such9 (whatever that means)?

2 Joined Cases C-267/91 and C-268/91 Criminal proceedings against Bernard Keck and Daniel Mithouard EU:C:1993:905, para 14. 3 See on the issue of discrimination, market access and selling arrangements, ch 2 by Stefan Enchelmaier in this volume. For an authoritative case to ‘restrict restrictions’, see C Barnard, ‘Restricting Restrictions: Lessons for the EU from the US?’ (2009) 68(3) The Cambridge Law Journal 575. 4 See ch 8 by Illetra Antonaki in this volume in the context of the free movement of capital. For a case to adapt proportionality to the nature of the restriction, especially in ethically contentious questions, see F de Witte, ‘Sex, Drugs & EU Law: The Recognition of Moral and Ethical Diversity in EU Law’ (2013) 50 Common Market Law Review 1545, 1566. 5 On the convergence between the free movement provisions, see ch 2 by Stefan Enchelmaier. 6 See S de Vries et al, Balancing Fundamental Rights with the EU Treaty Freedoms: The European Court of Justice as ‘Tightrope’ Walker (The Hague, Eleven International Publishing, 2012). 7 See ch 12 by Stephen Weatherill in this volume. 8 The four freedoms are considered to be those of goods, persons, services and capital, based on Art 26, para 2, TFEU which defines the internal market as ‘comprising an area without internal frontiers in which the free movement of goods, persons, services and capital is ensured’. In this approach, the freedom of establishment should be considered as comprised of the free movement of persons; however, while there is a partial overlap between the free movement of persons and establishment, so is there with services, and the freedom of establishment is moreover not only a right of self-employed individuals but a right for companies and may not always entail the actual movement of persons. It seems more logical to treat freedom of establishment as separate. ‘Persons’ could be redefined as ‘workers’, with the parts of free movement of persons that concerns the self-employed (and posted workers) comprised under establishment (and services). Simply extending the numbering to five in this way (goods, workers, services, establishment and capital) may not be fully satisfactory either, however, if one wants to account for the influence of EU Citizenship and its movement rights, which is part of the free movement of persons but extends (precisely) beyond workers and self-employed. 9 Not to speak of the fact that when they are called, as they often are, ‘the fundamental freedoms’ this seems to reflect an approach that places them at the ‘centre of the EUniverse’, suggesting that they are the most important values or at least the most important fundamental rights for the EU project.

Fundamental Freedoms and Fundamental Rights  337 The Court of Justice of the European Union (CJEU), for all its wisdom, has not given us much structured guidance. For each of the above doctrinal questions, there are judgments pointing in different directions. Certainly, to a significant degree, unsettled issues and multiple competing interpretations of core concepts and doctrines are, for anyone except a die-hard positivist, inherent in the indeterminate nature of the law and part of its very method.10 Yet, a minimum degree of coherence, knowability and predictability of the law, and its limits,11 are cornerstones of the Rule of Law. And the legitimacy of EU law, and of the judiciary’s work within it, is dependent on sound legal methods of interpretation, embedded in a sound legal methodology – which has to comprise a substantive vision of the constitutional order.12 Here we must note that Weiler’s turn-of-the century finding about the vacuity of European constitutionalism still holds true and goes to the core of the problem.13 He considered that European integration has not produced a European legal order of constitutionalism without a formal constitution, but the opposite: ‘a constitutional legal order the constitutional theory of which has not been worked out, its long-term, transcendent values not sufficiently elaborated, its ontological elements misunderstood, its social rootedness and legitimacy highly contingent’.14 More concretely still, Cahill points out the ‘conceptual dissonance’ between European integration and constitutionalism, namely that the latter ‘has a long-established and well developed internal connection to orienting ideals of democracy, self-determination, representation, constituent power, separation of powers and fundamental rights, both as legitimating foundations and as outcomes to be achieved’ while the former is not sufficiently oriented to these ideals (but instead mostly to integration itself).15 The case of the provisions on goods, services, workers, capital and establishment are emblematic thereof. It is common ground that they have been central in the CJEU’s constitution-building, yet they operate in a normative and theoretical vacuum – one could even say they have created

10 D Kennedy, ‘A Left Phenomenological Alternative to the Hart/Kelsen Theory of Legal Interpretation’ (2007) 40 Kritische Justiz 296. 11 In a federal order such as the EU, we should add the constitutional concern of respecting the competence and autonomy of the States, enshrined in EU law through the principles of conferral, subsidiarity and national constitutional identity. 12 See the seminal work of G Conway, The Limits of Legal Reasoning and the European Court of Justice (Cambridge, Cambridge University Press, 2012). Conway puts forward an argument for a normative theory of interpretation for the CJEU based on the rule of law and democracy. The approach and argument of this chapter concur with that, with as a main difference that human/fundamental rights are formulated alongside the rule of law as the principle of constitutionalism, whereas Conway’s approach does not develop a dialectic relationship between constitutionalism and democracy with human rights as an important component, but instead seems to take Waldron’s view about the inherently contestable nature of human rights and therefore does not consider them an integral part of the meta-principles of political morality that should dictate judicial interpretation. 13 JHH Weiler, The Constitution of Europe: Do the Clothes have an Emperor and Other Essays on European Integration (Cambridge, Cambridge University Press, 1999). 14 ibid, 8. 15 See Cahill, above n 1, 28.

338  Sacha Garben at least part of that vacuum due to their blunt force combined with, ironically, their effectiveness to integrate – but their inadequacy to create.16 What does it all matter? Don’t we by now know enough: that the free movement of goods, services, workers, capital and the freedom of establishment have a tremendously broad reach in terms of prima facie scope of application, catching most rules that are liable to affect economic activity, and that Member States simply have to learn how to defend (and adopt) their rules in an EU-proof and EU-savvy manner? As for those Member States, have they not had the chance, over and over again, to amend the Treaties to alter the interpretation of the CJEU if they disagreed with it? And to the extent that the overriding goal of these provisions was to further integration, have they not been a break-out success, despite – or perhaps thanks to – some of the above concerns? And there is no pleasing everyone anyway. Since one will always find a judgment to disagree with, no matter what ‘vision’ is adopted, should we not be careful not to let a handful of imperfect rulings (whichever one considers these to be) inform an all-encompassing condemnation of the state of internal market law as a whole? Even if it is clear that a coherent vision, idea or theory is lacking, do we really miss it, and would we really be better off with one? Would that not stifle innovation and adaptability at the Court, prevent it from achieving the right outcomes in individual cases, and overshoot its legitimacy – as a vision is basically another word for ideology and that it is something to be left to politics? This chapter argues that what we currently have is not enough and that the EU is in dire need of a more coherent theory on the basis of which it could define and improve its legitimacy. It considers that this theory should be based on the two equal principles of democracy and constitutionalism17 – the latter consisting in the Rule of Law and fundamental rights.18 The development of this theory is a responsibility of the entire EU legal community, presumably starting in academic

16 Davies has similarly accused EU internal market law of ‘conceptual poverty’, and considers that ‘establishing’ an internal market entails something more than ‘integrating’ it: ‘For what does it mean to establish a functioning internal market? What is required? One can argue that it requires trust between communities, effective communication, perhaps a certain shared understanding and expectations. Cultural and educational similarities can make trade easier, as can a shared language. One can also certainly argue, and indeed it is fiercely argued by many in these days of permanent economic crisis, that a market without redistribution cannot truly work – that it will lead to stresses and inequalities that will ultimately destroy it or the fabric of the societies in which it is embedded. It may be that an internal market requires a high degree of political and economic integration. To ‘establish’, it may be noted, implies creating a thing in a way that is solid and rooted. An established market is not a disembodied one, but one that is solidly connected with and rooted in its society’. G Davies, ‘The Competence to Create an Internal Market: Conceptual Poverty and Unbalanced Interests’ in S Garben and I Govaere (eds), The Division of Competences between the EU and the Member States: Reflections on the Past, the Present and the Future (Oxford, Hart Publishing, 2017). 17 See J Tully, ‘The Unfreedom of the Moderns in Comparison to Their Ideals of Constitutional Democracy’ (2002) 65 Modern Law Review 204. 18 See also S Garben, ‘The Principle of Legality and the EU’s Legitimacy as a Constitutional Democracy: A Research Agenda’ in S Garben, I Govaere and P Nemitz (eds), Critical Reflections on Constitutional Democracy in the European Union (Oxford, Hart Publishing, 2019).

Fundamental Freedoms and Fundamental Rights  339 writing, but with the aim to persuade the CJEU to adopt it, or for a Treaty change to embrace it explicitly. In Schiek’s words: ‘as the Court must be viewed as a political actor, the preconditions for changes in its case law presuppose the development of a wider discourse on the adequate constitutional frame for the future of EU integration’.19 It is imperative that this frame ensures an appropriate embedding of the internal market,20 and in particular its directly effective ‘negative’ provisions on goods, services, workers, establishment and capital. This ‘thick constitutional context’ should provide the context and telos for the interpretation of these provisions, both as regards their individual substantive meaning and in terms of their relationship to other rights. As will become clear, this does mean that there is a legitimate place for the free movement provisions as part of EU fundamental rights and constitutionalism, but only to the extent that they serve human dignity and to foster a robust and well-functioning transnational democracy. This provides something of a middle way between the argument to de-constitutionalise the internal market out of a concern for democracy on the one hand,21 and the current approach of internal market-exceptionalism that considers ‘the fundamental freedoms’ to be the normative apex and constitutional foundation of the European integration process. In order to make this argument, the chapter proceeds in the following way: section II will first consider the ‘fundamental’ nature of the free movement provisions, asking whether they are currently to be considered as EU fundamental rights, and what is at stake in that qualification and treatment. Section III in turn will set out the integrated democratic interpretation framework within which the free movement provision, as well as other fundamental rights, and the legal order more generally, should operate. Section IV concludes.

II.  How Fundamental are the Free Movement Provisions? The Relationship between ‘the Fundamental Freedoms’ and Fundamental Rights A.  Are the Free Movement Provisions Fundamental Rights? It is a rather axiomatic statement that the free movement provisions are part of the foundations of the EU project and that they are among the fundamental principles of the Treaty. It is much less obvious to state that they are ‘fundamental 19 D Schiek, ‘Towards More Resilience for a Social EU – the Constitutionally Conditioned Internal Market’ (2017) 13 European Constitutional Law Review 611. 20 In this regard, it concurs with the view of Schiek, who has proposed to ‘constitutionally condition’ the internal market. My approach differs to the extent that in determining the hierarchy and interpretation of the various constitutional rights in the Charter and the Treaties, this chapter argues using the principle of democracy rather than infer that hierarchy from a (social) reading of the Charter itself. See Schiek, above n 19. 21 D Grimm, The Constitution of European Democracy (Oxford, Oxford University Press, 2017).

340  Sacha Garben rights’, which would suggest a moral imperative for constitutional protection against majoritarian decision-making that, at least initially, could not be deduced from the text of the Treaties, nor – at face value – seems necessitated by human dignity22 (or entailed by ‘what it means to be a person’)23 or democracy.24 Yet, since the 1980s, it is commonplace in EU circles to speak of the free movement provisions as ‘fundamental freedoms’, a term not found in the Treaties but which had circulated in this context in German legal literature before the CJEU adopted it in 1981,25 stating that ‘the free movement of capital constitutes, alongside that of persons and services, one of the fundamental freedoms of the Community’.26 And indeed, as such, the term ‘fundamental freedom’ does not denote something different from ‘fundamental right’ or, for that matter, ‘human right’.27 The CJEU has furthermore specifically used the term ‘fundamental right’ for the free movement of workers,28 and for the free movement of goods.29 This, together with the privileged status in EU primary law afforded to the free movement provisions and the limited scope accorded to derogations from these provisions, would therefore suggest that indeed, the free movement provisions are – in the EU legal order – treated as fundamental rights. Some would however consider that these are just semantics, and that for the purposes of EU law, the fundamental freedoms are clearly distinct from fundamental rights, their parallels notwithstanding.30 De Cecco considers that ‘nothing 22 Human dignity is considered the source of all human rights. 23 For Hoffman, there are three factors to weigh in the question whether a right is fundamental: (i) is the right specifically identified as fundamental by a controlling text or authority or logically entailed by a right to be recognised; (ii) is the right empirically necessary to the realisation of a recognised fundamental right; (iii) is possession of the right entailed by what it means to be a person, so that no person devoted to human dignity could reasonably prefer to live in a society where the right was not recognised? D Hoffman, ‘What Makes a Right Fundamental’ in D Hoffman, Our Elusive Constitution: Silences, Paradoxes, Priorities (Albany, NY, State University of New York Press, 1942) 191. 24 It is argued that fundamental rights should not be conceptualised to constrain democracy (contra J Waldron) but instead as necessary to ensure the well-functioning of an inclusive and robust democracy in the long term. See also C Gearty, Principles of Human Rights Adjudication (Oxford, Oxford University Press, 2004) 88. 25 The first mention appears to be by AG Mayras in Case 2/74 Reyners, referring to the freedom of establishment as a ‘fundamental freedom for the benefit of nationals of Member States’. See also C Walter, ‘History and Development of European Fundamental Rights and Freedoms’ in D Ehlers (ed), European Fundamental Rights and Freedoms (Berlin, De Gruyter, 2007). 26 Case 203/80 Criminal proceedings against Guerrino Casati EU:C:1981:261, para 8. 27 Up until the moment that the CJEU claimed the free movement provisions to be ‘fundamental freedoms’, that term had been used by the ECtHR in reference to the set of human rights guarantees within its purview. 28 Case 152/82 Forcheri v Belgium EU:C:1983:205, para 11; see also Case 222/86 UNECTEF v Heylens EU:C:1987:442, para 14; Case C-415/93 UEFA v Bosman EU:C:1995:463. 29 Case C-228/98 Dounias v Minister for Economic Affairs EU:C:2000:65, para 64. It could be argued, as de Cecco has, that this ‘is an isolated and dubious reference point’. F de Cecco, ‘Fundamental Freedoms, Fundamental Rights, and the Scope of Free Movement Law’ (2014) 15 German Law Journal 383, 391. 30 T Kingreen, ‘Fundamental Freedoms’ in A von Bogdandy and J Bast (eds), Principles of European Constitutional Law, 2nd rev edn (Oxford, Hart Publishing, 2011) 519; W Frenz, ‘Grundfreiheiten und Grundrechte’ (2002) 37 Europarecht 606. See for a nuanced discussion of the issue: P Oliver and W Roth, ‘The Internal Market and the Four Freedoms’ (2004) 41 Common Market Law Review 407.

Fundamental Freedoms and Fundamental Rights  341 appears to turn on this form of words, except the fundamental importance of free movement in the scheme of the Treaty, and the emphasis on the narrow interpretation to be given to derogations/justifications’.31 The former President of the CJEU has argued that despite their similarities in name, application, interpretation and effect, the fundamental freedoms and fundamental rights are different for three main reasons: (i) the fundamental freedoms are enumerated while fundamental rights are not; (ii) the fundamental freedoms only apply in economic, cross-border circumstances; and (iii) fundamental rights are typically human rights, while the fundamental freedoms are less based on the individual (except the free movement of persons).32 With the comprehensive bill of rights that is the EU Charter becoming part of primary law since Lisbon, the first argument loses most of its force, if it ever constituted an analytically relevant difference to begin with. One may however counterargue that with the Charter, fundamental rights have indeed also become enumerated, but that that enumeration excludes most aspects of the free movement provisions. Only Article 15(2) of the EU Charter provides that ‘every citizen of the Union has the freedom to seek employment, to work, to exercise the right of establishment and to provide services in any Member State’, leaving most notably the free movement of capital and goods outside the Charter’s scope – and thus outside the EU’s body of fundamental rights.33 Yet, apart from the fact that nothing seems to prevent the treatment of the EU Charter as non-exhaustive and thus to allow the recognition of other fundamental rights, it has been argued that various EU Charter provisions – most notably Article 21 on non-discrimination on grounds of nationality, Article 16 on the freedom to conduct a business and Article 17 on the right to property, can be interpreted/reconstructed to encompass most aspects of the free movement provisions.34

31 de Cecco, above n 29, 390. 32 V Skouris, ‘Fundamental Rights and Fundamental Freedoms: The Challenge of Striking a Delicate Balance’ (2006) 17 European Business Law Review 225, 234–35. 33 de Cecco, above n 29, 392. 34 Oliver and Roth argue, as regards the free movement of goods, that the right to import and export goods between Member States could be regarded as part of the freedom to conduct a business or of the right to property. P Oliver and W Roth, ‘The Internal Market and the Four Freedoms’ (2004) 41 Common Market Law Review 409. V Trstenjak and E Beysen note that ‘the substantive guarantees inherent in fundamental freedoms can be formulated in terms of fundamental rights which protect economic activity’: V Trstenjak and E Beysen, ‘The growing overlap of fundamental freedoms and fundamental rights in the case-law of the CJEU’ (2013) 38 European Law Review 293, referring to V Skouris, ‘Das Verhältnis von Grundfreiheiten und Grundrechten im europäischen Gemeinschafts­ recht’ (2006) 59 Die Öffentliche Verwaltung 89, 93–96. For the argument that the principle of equality can be used to anchor a fundamental rights approach to the free movement provisions, see Opinion of Advocate General Stix-Hackl in Case C-36/02 Omega Spielhallen v Oberbürgermeisterin der Bundesstadt Bonn ECLI:EU:C:2004:162. For a general discussion see de Cecco, above n 29, 392 ff. Nota bene an ‘integrated’ approach that treats (aspects of) the free movement provisions as fundamental rights can be used to determine the appropriate scope of application of the free movement provisions, as will be argued in section III below. As will be seen, that does not necessarily have to lead to a further expansion of the free movement provisions and can rather be a method of limiting their constitutional scope and reach.

342  Sacha Garben As to the second argument, it is true that the scope of application of the free movement provisions is determined on the basis of the respective Treaty provisions, which require a (potential) cross-border element. Article 51 of the EU Charter on the other hand, the reference point for EU fundamental rights, provides that the ‘provisions of this Charter are addressed … to the Member States only when they are implementing Union law’, which may not always require a crossborder element, nor does a cross-border element automatically mean the situation is in the ‘scope of EU law’. Yet, this should not distract from the fact that these are basically two different operationalisations of the same criterion: a connecting factor to the EU. For the Charter rights this is determined by the ‘scope of EU law’, and for the fundamental freedoms by a cross-border element. Indeed, the crossborder element could be considered the way to establish that we are in the scope of EU law for the purposes of the fundamental freedoms. One might object that the free movement provisions are the scope of EU law in a very direct sense, while the fundamental freedoms come into play only indirectly. Although there is some truth to that, there are several fundamental rights that also have a legal basis in the Treaties and some, even if not all,35 Charter provisions have been afforded direct effect,36 meaning that fundamental rights can also constitute a very ‘direct’ part of EU law. On balance, all this does not seem to provide a valid reason to categorically differentiate the fundamental freedoms and the fundamental rights. Also the third alleged difference can be mitigated. There is no clear-cut difference between the fundamental freedoms that would concern economic operators, on the one hand, and the Charter rights which would concern individual citizens, on the other. For one, the free movement of persons, which entails not just the free movement of workers but also aspects of the freedom of establishment and the freedom to provide services, places the rights of the individual citizen central. To reiterate, this aspect of the free movement provisions is explicitly included in the EU Charter in Article 15(2). Furthermore, a range of fundamental rights under the EU Charter can be enjoyed by businesses, both substantively (right to property, freedom to conduct a business) and procedurally (right to an effective remedy, presumption of innocence, rights of defence).37 In particular, Article 16 of the EU Charter, concerning the freedom to conduct a business, shows that the Charter is not only concerned with traditional human rights; it potentially overlaps with the fundamental freedoms to some extent.38 This is not to say that the position of the individual is not important in the qualification of a right as fundamental, or that it is uncontroversial to afford fundamental rights protection to companies.

35 eg, Judgment of 15 January 2014, Case C-176/12 Association de médiation sociale EU:C:2014:2. 36 See recently, Case C-414/16 Vera Egenberger v Evangelisches Werk für Diakonie und Entwicklung eV EU:C:2018:257. 37 See P Oliver, ‘Companies and Their Fundamental Rights: A Comparative Perspective (2015) International & Comparative Law Quarterly 661. 38 See generally R Babayev, ‘Private Autonomy at Union level: On Article 16 CFREU and Free Movement Rights’ (2016) 53 Common Market Law Review 979.

Fundamental Freedoms and Fundamental Rights  343 These are arguments that however need to be treated separately from the question whether in the current state of interpretation of EU law the free movement rights are treated as fundamental rights. In subsection B, as well as in section III, we will consider what criteria should be used to determine the fundamental status of a right and its scope and interpretation. Beyond the above arguments of the former CJEU President, a remaining reason to distinguish the fundamental freedoms from fundamental rights could be that they serve different purposes in the EU constitutional settlement. In the words of Kingreen, the ‘functional differences between fundamental freedoms and rights can be condensed into the terms of “transnational integration versus supranational legitimation” … In short, what the fundamental freedoms created the fundamental rights must now seek to legitimate’.39 It is of course true that the free movement provisions have been used in functional ways, serving to propel the project of the internal market – and therewith European integration more generally – onwards and upwards. They have been the oil that has powered the motor of European union. Fundamental rights, by contrast, ostensibly do not have that same integrative aim, and are seen instead as a check on that power, thereby legitimising it. They place limits on the action of the EU institutions, and they are not intended as self-standing obligations that constrain the powers of the Member States outside the scope of EU law. Yet, such a description does not do justice to the ‘unsettling nature’ of EU fundamental rights’ competence.40 Fundamental rights do also empower the EU legislator, as several of them are coupled to a legal basis in the Treaties,41 or can inform the content of EU legislation more indirectly.42 The CJEU has furthermore interpreted the scope of EU law generously,43 and while fundamental rights may sometimes curb the fundamental freedoms, sometimes it is the other way around, and furthermore, sometimes they compound each other in further empowering the EU vis-à-vis the Member States.44 Muir notes in this regard that it is precisely because fundamental rights share with the fundamental freedoms their capacity to seriously limit Member

39 Kingreen, above n 30, 531– 32. 40 E Muir, ‘Fundamental rights: An unsettling EU competence’ (2014) 15 Human Rights Review 25. 41 Beyond the often-cited Arts 17 and 19 TFEU on data protection and non-discrimination, we could consider the legal bases relating to social policy (Art 153 TFEU); the environment (Art 192 TFEU); and health (Art 168 TFEU) and their respective mainstreaming clauses as legal competence to ‘implement’ the fundamental rights contained in the EU Charter, and thus as ‘fundamental rights competences’. 42 Muir, above n 40. 43 As Trstenjak and Beysen note: ‘A common trait of the CJEU case law on fundamental freedoms and rights is the gradual expansion of their respective scope of application. Not only have the fundamental freedoms been applied in the most diverse fields of law, but the CJEU has also widened the circle of the parties to whom they are addressed and it has extended the obligations which flow from those provisions. A similar, though more tentative, evolution can be observed in the CJEU’s case law on EU fundamental rights, especially with regard to the determination of the areas of law in which those rights must be observed’. Trstenjak and Beysen, above n 34. 44 eg, Case C-60/00 Mary Carpenter v Secretary of State for the Home Department EU:C:2002:434.

344  Sacha Garben State regulatory choice and capacity, that they are often contested by the Member States.45 Fundamental rights thus also lead to further integration and are instrumental in the project of EU-building, in ways that again do not seem categorically different from the fundamental freedoms. The functional nature of fundamental rights is underlined by the Court itself, where it says that the reason for pursuing (the objective of protecting fundamental rights in EU law) is the need to avoid a situation in which the level of protection of fundamental rights varies according to the national law involved in such a way as to undermine the unity, primacy and effectiveness of EU law.46

There is ground to worry about this ‘parochial’ view from the CJEU,47 but for the purposes of our discussion it illustrates how EU law sees fundamental rights and fundamental freedoms as agents of integration (surely as well as inherently valuable). At the same time, like fundamental rights, the fundamental freedoms as interpreted by the Court do not only empower the EU, they also circumscribe it – especially the discretion of the EU legislator, which is held by the Court’s interpretation of the free movement provisions by the hierarchy of norms. The fundamental freedoms function as constitutional checks on all public power, both at national and EU level, like fundamental rights do. To the extent that fundamental freedoms primarily bind the Member States and fundamental rights primarily the EU, this seems, in light of the foregoing analysis, a difference of degree rather than a categorical distinction. Finally, Kingreen’s more implicit point seems to be that the fundamental freedoms are ultimately closer to the EU’s raison d’être than fundamental rights are. While true, this is a historical approach that certainly explains why EU legal scholars tend to distinguish the four freedoms and fundamental rights, but does not point at a decisive analytical difference, especially not from a more forwardlooking perspective. It should be noted that recent years have clearly shifted the EU’s self-perception and political commitment from a project of market integration to a project of values,48 and this is likely to become an even more important steer of European integration in the future, presumably with fundamental rights as a spearhead. Of course, as Augenstein argues, it remains to be seen whether the EU will succeed in evolving into a polity ‘with a robust valuebased “institutional ethos” that transcends the narrow confines of economic and technocratic rationality’.49 He warns that the conception of fundamental rights 45 Muir, above n 40. 46 Case C-206/13 Siragusa EU:C:2014:126, para 32. 47 E Spaventa, ‘The interpretation of Article 51 of the EU Charter of Fundamental Rights: the dilemma of stricter or broader application of the Charter to national measures, Study for the European Parliament PETI Committee’ (2016) 21, available at: www.europarl.europa.eu/RegData/ etudes/STUD/2016/556930/IPOL_STU(2016)556930_EN.pdf. 48 See T von Danwitz, ‘Values and the rule of law: Foundations of the European Union – an inside perspective from the ECJ’, available at: www.kcl.ac.uk/law/research/centres/european/KingsCollegeLondon-TVD-k-def-Verteilung.pdf. 49 D Augenstein, ‘Engaging the Fundamentals: on the Autonomous Substance of EU Fundamental Rights Law’ (2013) 14 German Law Journal, referring to A Williams, The Ethos of Europe: Values, Law and Justice in the EU (Cambridge, Cambridge University Press, 2010).

Fundamental Freedoms and Fundamental Rights  345 in the EU polity to date has been conditioned by the fundamental freedoms, and that this will be difficult to overcome.50 That, however, is an argument to be treated in relation to the question of hierarchy between different fundamental rights (subsection C below) and in the development of an integrated interpretation framework (section III below). It does not deny the argument of the present enquiry that fundamental freedoms are – by virtue of EU law – treated as fundamental rights.51

B.  Should they Be? Considering the foregoing, it thus seems that little stands in the way of considering the free movement provisions as EU fundamental rights, because for all relevant aspects (application, interpretation and effect), they function largely the same, and even if they have a different legacy in the process of European integration to date, there does not seem to be a convincing categorical analytical difference between them. That finding, however, does not pertain to the distinct question of whether the free movement provisions should be considered ‘real’ fundamental rights, in other words, whether such a label and treatment is appropriate. The CJEU, in its judgments constitutionalising52 the free movement provisions according them primacy and direct effect, including over national constitutional law, has not clarified the normative reasons or justifications for treating them as such. Some would argue that a theoretical vision does implicitly underlie the CJEU’s approach, or in any event rationalises and legitimises it post facto, namely that of ordoliberalism.53 The latter promotes the constitutionalisation of an economic order through provisions of competition and economic freedom as a ‘political

50 Augenstein, ibid. 51 Indeed, Augenstein’s assessment confirms that finding. Augenstein, above n 49: ‘With the adoption of the EU Charter, the fundamental market freedoms have become incorporated into the body of EU fundamental rights law’. 52 On the term, see M Loughlin, ‘What is Constitutionalisation?’ in P Dobner and M Loughlin (eds), The Twilight of Constitutionalism: Demise or Transmutation? (Oxford, Oxford University Press, 2010). On EU constitutionalism: M Poiares Maduro, We The Court. The European Court of Justice and the European Economic Constitution (Oxford, Hart Publishing, 1998); A Hatje, ‘The Economic Constitution within the Internal Market’ in A Bogdandy and J Bast (eds), Principles of European Constitutional Law (Oxford, Hart Publishing, 2010); Schiek, above n 19. It should be noted that ‘constitutionalisation’ in itself does not have to equate the free movement provisions to ‘real’ fundamental rights as such. This touches on the related, yet distinct, question of hierarchy or ‘balance’ between the free movement provisions and fundamental rights as protected national constitutions, the EU Charter and the ECHR. It is one thing for the free movement provisions to have primacy and direct effect in relation to national legislation, another for them to circumscribe the EU legislative process, yet another for them to condition fundamental rights protected at national, at EU and/or at the international level. 53 PC Müller-Graff, ‘Die wettbewerbsverfaßte Marktwirtschaft als gemeineuropäisches Verfassung­ sprinzip?’ (1997) 31 Europarecht 422, attempts to establish these principles as a common constitutional core across the European Union. See also PC Müller-Graff and E Riedel (eds), Gemeinsames Verfassungsrecht in der Europäischen Union (Baden-Baden, Nomos, 1998).

346  Sacha Garben theory of society’ providing a ‘“third way” beyond liberalism and socialism as a permanent and liberal order of peace’.54 While this school of thought naturally favours the Court’s treatment of the free movement provisions as fundamental rights, it does not seem to be the ‘grand theory’ behind the CJEU’s case law on the internal market and competition law or European integration more generally, which is instead driven by functionalism more than anything.55 There are too many judgments like Keck,56 Albany,57 Altmark58 and AKT,59 to name but a few, that pull in a different direction and cannot be accommodated in a holistic constitutional vision based on ordoliberalism, or neoliberalism,60 or any coherent vision for that matter61 – as argued in the introductory section. Such an approach would furthermore be highly problematic from the perspective that Article 3(3) of the Treaty on European Union (TEU) provides that the Union shall ‘work for the sustainable development of Europe based on balanced economic growth and price stability, a highly competitive social market economy, aiming at full employment and social progress’, mentioning in the same paragraph the Union’s task to establish an internal market. The EU constitution simply does not allow the judicial application of a constitutional theory of ordo- or neoliberalism to European integration generally, or the internal market specifically. But one does not have to subscribe to ordo- or neoliberalism, nor to vacuous integrationism, to support giving (a degree of) constitutional value to (some 54 R Wiethölter, ‘Wirtschaftsrecht’ in A Görlitz (ed), Handlexikon zur Rechtswissenschaft (München, Ehrenwirth, 1972) 531, 534. On Ordoliberalism in the EU see C Joerges, ‘The Market without a State? States without Markets? Two Essays on the Law of the European Economy’ (1996) EUI Working Paper Law 1/96, San Domenico di Fiesole, available at: eiop.or.at/eiop/texte/1997‐019 and − 020.htm; C Joerges, ‘Economic Law, the Nation‐State and the Maastricht Treaty’ in R Dehousse (ed), Europe after Maastricht: an Ever Closer Union? (Munich, CH Beck, 1994); J Hien and C Joerges, ‘Dead Man Walking: Current European Interest in the Ordoliberal Tradition’ (2018) EUI Working Paper LAW 2018/03. 55 For a comprehensive analysis of the case law of the CJEU on the free movement and competition provisions in relation to the opposing Ordoliberal and Service Public schools of thought, and the conclusion that what drives the case law is functionalism and neither of these theories: W Sauter and H Schepel, ‘“State” and “Market” in the Competition and Free Movement Case Law of the EU Courts’ (2007) TILEC Discussion Paper 24. 56 Joined Cases C-267 and C-268/91 Keck and Mithouard EU:C:1993:905; As Babayev argues, ‘the Keck line of reasoning has been followed in cases such as Case C-441/04 A-Punkt Schmuckhandels EU:C:2006:141; Case C-531/07 LIBRO EU:C:2009:276; Case C-483/12 Pelckmans ECLI:EU:C:2014:304; Case C-198/14 Visnapuu EU:C:2015:751. 57 Case C-67/96 Albany International BV v Stichting Bedrijfspensioenfonds Textielindustrie EU:C:1999:430. 58 Case C-280/00 Altmark Trans GmbH and Regierungspräsidium Magdeburg v Nahverkehrsgesellschaft Altmark GmbH, and Oberbundesanwalt beim Bundesverwaltungsgericht EU:C:2003:415. 59 Case C-533/13 Auto- ja Kuljetusalan Työntekijäliitto AKT ry v Öljytuote ry and Shell Aviation Finland Oy EU:C:2015:173. For a discussion, see S Garben, ‘The Constitutional (Im)balance between “the Market” and “the Social” in the European Union’ (2017) 13 European Constitutional Law Review 23. 60 See, however, M Wilkinson, ‘Authoritarian Liberalism in Europe: A Common Critique of Neoliberalism and Ordoliberalism’ (2019) 45(7–8) Critical Sociology 1023; and M Wilkinson, ‘Authoritarian Liberalism as Authoritarian Constitutionalism’ (2018) LSE Law, Society and Economy Working Papers 18/2018. 61 Babayev, above n 38, 985.

Fundamental Freedoms and Fundamental Rights  347 aspects of) the free movement provisions alongside more traditional fundamental rights. If we consider the common ground of contemporary constitutional democracy to be the commitment to the triptych of democracy, the Rule of Law and fundamental rights,62 and within that context the latter to be concerned with upholding human dignity and guaranteeing a robust and inclusive democracy,63 then we can still identify three justifications for the treatment of the free movement provisions as fundamental rights. As we will come to see, however, that treatment can only be defended to a narrower extent than the current interpretation of the free movement provisions. The first, most persuasive defence of the constitutional treatment of free movement provisions, is the one that Floris de Witte has called ‘the argument from aspirational justice’: that the free movement provisions expand ‘individual agency beyond the parameters of permissive behaviour in the citizens’ own Member State, and as such provides a trampoline for the attainment of the individual’s aspirations’.64 This is an individualistic approach, that considers the right to move freely between Member States as a way to foster individual dignity through enhanced agency and self-determination. It could furthermore be considered an essential part of what it means to be a member of a political order: that one has the possibility to pursue one’s life in every territorial part of it, on an equal basis with other members. That would therefore also provide the normative ground for the right to equal treatment in relation to the economic opportunities offered in different parts of the federal territory, and ties this argument to citizenship and the general principle of equality and the fundamental right to non-discrimination. A less ‘aspirational’ and more basic dimension of the argument, but all the more relevant for justice, is that free movement provides an exit from fundamental rights abuses and oppressive regimes within the federal territory – which in turn may act as a disincentive for rulers to oppress and thereby act as a federal bulwark against autocracies.65 All this provides a strong normative justification for treating free movement provisions as fundamental rights, but only when it comes to the actual movement of individuals in their capacity as citizen, worker or self-employed – which is the dimension of the ‘fundamental freedoms’ that is indeed already captured in the EU Charter in Article 15(2). It may perhaps be stretched (rather thinly) to apply to the individual investor in the context of the free movement of capital in certain circumstances, but it is certainly less convincing in relation to business economic rights to ‘market access’ in relation to goods, services, capital and establishment. A second defence of the constitutionalisation of the free movement provisions is an argument of transnational democracy,66 also referred to as the argument from 62 Garben, Govaere and Nemitz, above n 18. 63 Gearty, above n 24, 88. 64 de Witte, above n 4, 1152. 65 For the general argument that federalism acts as a protection against autocracies, see D Halberstam, ‘Federalism: A Critical Guide’ (2011) University of Michigan Public Law Working Paper No 251. 66 C Joerges, ‘European Law as Conflict of Laws’ in C Joerges and J Neyer, “’Deliberative Supranationalism” Revisited’ (2006) 20 EUI Working Paper Law 22.

348  Sacha Garben ‘transnational effects’.67 Market integration is, to the extent that it de-nationalises decision-making to ensure that affected interests from other Member States are taken into account, ‘desirable on account of democracy itself ’.68 This transnational democracy argument is undoubtedly very strong in relation to positive market integration through the adoption of EU-level rules regulating the internal market. We would argue that this is an important counterbalance to the often-heard argument that the EU legislative process is not as democratically legitimate as national decision-making, and that it mitigates the concerns of ‘competence creep’ in relation to the use of Article 114 TFEU. Here, we are however concerned with the ‘negative’ integration dimension of the internal market, through the application of the free movement provisions, and the constitutionalisation thereof. It is in this context that the transnational democracy argument has been most widely used, becoming ‘a shibboleth for the European pro-attitude’,69 even if its validity is much less obvious than in relation to positive market integration. The idea is that the fact that individual actors can confront the rules of other Member States that disadvantage them but in which they have had no representation, and to have these rules subjected to a judicial balancing test, is democracy-enhancing because it takes account of those excluded from the decision-making process. Poiares Maduro rendered the first comprehensive account of this argument in relation to the free movement provisions, considering that ‘the representation of the interests of nationals of other Member States within national political processes’70 argues in favour of a ‘fundamental rights conception of the free movement rules’.71 The normative basis for the right to have one’s interests to be taken into account that underlies this argument could be that the democratic principle demands that the interests of those who are affected by a decision need to be weighed in that decision somehow. But first, there needs to be some qualification as to what it means to be ‘affected’ to the extent that it triggers the right to be ‘represented’ or have one’s interests weighed. If any theoretical and elusive effect were to be counted, with the inclusion of opportunity costs, everyone should always be represented everywhere, leading to theoretically and practically absurd outcomes. The original democratic idea of ‘no representation without taxation’ would suggest a much higher threshold: an actual, direct cost needs to be incurred. That is very different from the opportunity cost of potentially having made more money if the rules on economic activity were more lenient. Somewhat tongue in cheek, it may be noted that the free movement provisions seem to have been given an interpretation closer to ‘representation without taxation’ rather than the other way around. 67 A Somek, ‘The Argument from Transnational Effects I: Representing Outsiders through Freedom of Movement’ (2010) 16 European Law Journal 315. 68 ibid. 69 ibid. 70 M Poiares Maduro, ‘Article 30 and the European Economic Constitution: Reforming the Market or the State?’ in M Poiares Maduro, We The Court: The European Court of Justice and The European Economic Constitution (Oxford, Hart Publishing, 1998) 167. 71 ibid, 168.

Fundamental Freedoms and Fundamental Rights  349 It would thus have to be determined which foreign ‘affected interests’ are legitimately so important that they need to be judicially enforced in national democracies. Second, for the democratic argument to make sense, it needs to be ensured that all the interests are balanced fairly. But what does fairness mean in this context? It would seem to necessitate some proportional weighing of interest, as is done in democratic procedures. Note bene: that is a very different exercise from a judicial proportionality test, where interests are weighed in the abstract, detached from the actual democratic weight given in numbers of citizens attached to these interests. And whose excluded interests are we to weigh exactly? Are we speaking of the individual interests of the foreign company that has to comply with product composition rules when wanting to sell on the host market? If so, the free movement rules are a rather disproportionate ‘counterbalance’: they put the interests of, say, some foreign alcohol-producers on par with those of potentially millions of citizens who may favour their specific rules on the production of beer.72 Giving priority to this foreign minority interest through judicial imposition of ‘mutual’ recognition amounts to a decisively non-democratic way to ‘enhance transnational democracy’. Moreover, this foreign private interest is very difficult to separate from the more general private interests, shared by all alcohol-producers both within and outside the host State, of having a laxer regulatory regime. That particular interest has already been weighed in the national democratic process and has therefore not been really excluded,73 and therefore does not deserve to be enforced by the constitution. Or are we instead speaking of the ‘economic benefit’ of other Member States more generally and abstractly? But apart from that fact that is very difficult to determine the due weight that should be attached to such an abstract interest, it is also not entirely true that other Member States are without any representation in national decisions of market regulation: to the extent that this falls within the EU’s legislative competence under Article 114 TFEU, common rules can be adopted to supersede the national ones. That means that there is a genuinely democratic avenue to have all relevant federal interests taken into account. Objections that this decision-making process is arduous and prone to deadlock do not deny that what matters most for the transnational effects argument is the existence of a forum of transnational democratic decision-making in which the relevant interests can be brought to the table, regardless of the result. Furthermore, such objections would sit uneasily with the often-heard complaints about EU over-regulation and competence creep on the basis of Article 114 TFEU: apparently the legislative process is very able, too able according to some, to produce output. Finally, it should be noted that to the extent that mutual recognition implies that market actors should be able to export their ‘home’ regulatory regimes to other Member States, it does

72 Not to mention the fact that the free movement provisions in their current interpretation are the rule and all other interests the exception, arguably thereby weighing the former more heavily than the latter, see subsection C below. 73 Somek, above n 67.

350  Sacha Garben not resolve but instead ‘flips’ the democracy failure: the people of the host State are forced to accept the rules decided on in another Member State in which they have had no say. That can hardly be a satisfactory solution from the very democratic perspective that lies at the heart of the justification. All in all, the democratic argument seems to be able to justify the constitutionalisation of the free movement provisions only to the limited extent that it concerns the adoption by a Member State of a rule that specifically harms the economic interests of foreign economic actors, ie, protectionism. It is only that interest that is really excluded from the national democratic process (as general economic interests are not) and may not sufficiently be protected by the EU legislative process. This interest may however still be counterbalanced by other interests: the democratic principle merely authorises/requires that the foreign actor’s interest is weighed in due proportion alongside the (many!) other interests that make up the decision-making process. This suggests a rather broad scope for possible ‘justification’ of even overtly protectionist measures. As to the argument that we are only able to effectively enforce that rule if we adopt an obstacle-approach rather than an intention-approach, and thereby also include indistinctly applicable measures to trigger the judicial weighting exercise (since Member States will simply cover up their protectionism behind a veneer or impartiality), we need to insist again that there is always an even more transnationally democratic remedy available: the adoption of common EU-level rules. Since the negative application of the free movement provisions is the much less refined, less effective and more contentious method of market integration from a democratic perspective, the democratic argument itself requires that precedence is given in principle to positive market integration. This is a crucial point, because the constitutionalisation of the free movement provisions does not only mean that they trump national legislation, but in some ways more importantly also that they trump EU-level legislation. After all, the free movement provisions are primary law and in the EU’s hierarchy of norms they outrank ordinary legislation. Far from sustaining the current sweeping reach and application of the free movement provisions, the democratic argument can thus only support a rather narrow version of their constitutionalisation, limited to protectionist measures. Not respecting these limits will mean that the democratic argument collapses in on itself, with the free movement provisions creating a democratic deficit, not only in relation to national democracy but also in relation specifically to transnational democracy. Finally, one could consider the protection of individual economic autonomy as a normative justification for the treatment of the free movement provisions as fundamental rights. Babayev considers that there is an overlap between the freedom to conduct a business, as laid down in Article 16 of the EU Charter, and the free movement provisions.74 He argues that Article 16 of the EU Charter ‘reflects

74 Babayev,

above n 38.

Fundamental Freedoms and Fundamental Rights  351 the customary liberal meaning of individual economic autonomy and protects it as an end in itself. On the other hand, a more specific form of such autonomy appears to be manifested through free movement rights’.75 Seen as such, the free movement provisions are still ‘not aimed to ensure the freedom to carry out an economic activity per se’, but they would ‘confer the freedom to access the market in a Member State’76 in a way that potentially exceeds the justifications under the transnational democracy and aspirational justice arguments. For our purposes, the relevant idea is that the protection of economic entrepreneurship is already part of the fundamental rights catalogue of the EU, and the free movement provisions can therefore legitimately be seen as specific manifestations of that fundamental right in a federal, cross-border context. But that does not yet provide us with the underlying normative justification, from a perspective of democratic constitutionalism, of such a fundamental right to economic autonomy in the first place, and what that underlying value means for the interpretation of that right in general and the specific manifestations in a free movement context. Article 16 of the EU Charter provides that ‘[t]he freedom to conduct a business in accordance with Union law and national laws and practices is recognized’. The wording indicates that this freedom is inherently relative and can be limited by both regulations and practices, suggesting it is one of the ‘weaker’77 rights in the EU Charter.78 It is widely considered to be ‘one of the less traditional rights’ not generally protected in international human rights instruments.79 The freedom to conduct a business under Article 16 of the EU Charter is drawn from the Court’s case law and the constitutional traditions common to Member States. To what end and to what extent does national constitutional law protect this individual autonomy against majoritarian decision-making? It is not traditionally universally present in the national constitutional law of the Member States, with many having only recognised versions of this right very recently and some not as a(n enforceable) constitutional right at all.80 Where the freedom to conduct a business is recognised in national constitutional law, it tends to allow a relatively wide scope of limitation in the public interest and it is generally conceived as a right of individuals to set up an economic activity or join a profession rather than concerning the general exercise of economic activity.81 As Groussot et al state, ‘[t]hey have in common that they do not constitute unfettered prerogatives and must thus be viewed in the light of their social function’.82 75 ibid. 76 ibid, 981. 77 X Groussot et al, ‘Weak right, strong Court – The freedom to conduct business and the EU charter of fundamental rights’ in S Douglas-Scott and N Hatzis (eds), Research Handbook on EU Law and Human Rights (Cheltenham, Edward Elgar Publishing, 2017). 78 Groussout et al, above n 77. 79 European Union Agency for Fundamental Rights, ‘Freedom to conduct a business: exploring the dimensions of a fundamental right’ (Luxembourg, Publications Office of the European Union, 2015). 80 For an overview, see European Union Agency for Fundamental Rights, above n 79. 81 ibid. 82 Groussot et al, above n 77.

352  Sacha Garben The EU Fundamental Rights Agency considers that the ‘freedom to conduct a business is about enabling individual aspirations and expression to flourish, about encouraging entrepreneurship and innovation, and about social and economic development’.83 Applying the principles of democracy and constitutionalism, the freedom to conduct a business has a certain role in ensuring a robust democracy, to the extent that it helps to empower individuals – especially those who are disadvantaged – through economic activity, to prevent concentrations of (economic) power and equalise societal asymmetries. This would suggest a reading where the right is about fostering the possibility of individual entrepreneurship, in the sense of enabling citizens to set up an economic activity or join a profession, within a broader picture of creating more socio-economic progress and equality in society. This human right would not, upon such a constitutional democratic reading, provide a ‘sword’ for companies in the operation of their business against workers, citizens and general public interest standards – that would be a perverse result. This brings us very close to the reading of the fundamental right’s dimension of the free movement provisions already contained in Article 15(2) of the EU Charter: that of individual empowerment and agency in a federal context. In particular, the idea that the free movement provisions (in their constitutional form) are not only about territorial mobility as an end in itself, but also about creating the possibility for upward social mobility, brings us quite firmly within justified fundamental rights territory. But that would reject a reading of Article 16 and the free movement provisions as about businesses’ right to contractual freedom or the running of a business (across borders) without interference. Thus, taken together, the arguments from aspirational justice, individual economic autonomy and of transnational democracy (or transnational effects) support the constitutionalisation of the free movement provisions up to a degree: the right to actual physical movement, a degree of economic autonomy/entrepreneurship and a right to equal treatment of individuals in their capacity as citizen, worker or self-employed – which is the dimension of the ‘fundamental freedoms’ that is already captured in the EU Charter in Articles 15(2), 16 and 17 – and the right to challenge protectionist restrictions on the cross-border activities of companies. Any constitutionalisation beyond this important, but – compared with the status quo deeply internalised by EU law and its community – rather limited dimension, cannot benefit from the arguments from aspirational justice, individual autonomy and transnational democracy even cumulatively applied84 and seems only justifiable on the basis of a normative commitment to ordo- or neoliberalism. An overly broad approach, such as is currently taken, may not be a conscious choice to support ordo- or neoliberalism however, as the legitimate scope of constitutionalisation of free movement combined with a concern 83 European Union Agency for Fundamental Rights, above n 79. 84 Contra F de Witte, who considers that the arguments from aspirational justice, transnational effects and a more general argument of constraining democracy to combat minority-rights abuses, can justify the current approach in EU free movement law. See de Witte, above n 4.

Fundamental Freedoms and Fundamental Rights  353 for ‘effectiveness’ can easily ‘slip’, but the result remains illegitimate despite the intentions. In the words of Alexander Somek: It takes more intellectual effort to restrict the normative implication of the argument from transnational effects to the choice of anti-protectionist measures. In other words, it is not easy to establish why the adverse effect of national policies on non-citizens should be remedied only where a national interest, narrowly understood, is pursued intentionally. As a consequence, the argument from transnational effects is by default disposed to support the more broadly sweeping obstacle-based approach. This inclination is reinforced by a widespread intuitive awareness of the normative weakness of anti-discrimination laws. It goes back to the fact that they do not avail of any clearly stated distributive objectives. The lack of distributive objectives translates into the impossibility of determining instances of discrimination. Not by accident, hence, judicial tribunals tend to switch into the obstacle mode, which sweeps comfortably broadly. From this mode it is only a small step to enforcing economic due process. When this step is taken the argument from transnational effects loses its mooring in democratic sensibilities. In fact, democracy – even if only virtual democracy – becomes the casualty in the process of its transnational reinforcement. What remains is the reinforcement of economic liberalism.85

As stated above, ordo- or neoliberalism is not an appropriate constitutional theory for the EU. I personally do not subscribe to ordo- or neoliberalism in terms of political economy. Academically, I do not consider them convincing constitutional theories on the legitimate exercise of authority and the enhancement of public welfare.86 Legally, I do not think that the text of the Treaties allows the EU constitutional order to be interpreted in line with these theories. I thus need to concur, to the extent that the free movement provisions have been given the status of fundamental rights beyond their dignity- and democracy-enhancing role as explained above, with those who argue that the free movement provisions have become over-constitutionalised.87 In other work, I have supported the democratic complaint about this aspect of the EU integration process, concerned that the combination of the substantive (as in, the wide market access or even economic freedom approach to what is a restriction), the procedural (as in, the strict proportionality requirements on justifications of restrictions) and the constitutional (as in, their position in the hierarchy of norms) interpretation of the internal

85 Somek, above n 67, 333. 86 See also Wilkinson, ‘Authoritarian Liberalism in Europe’, above n 60. 87 As also described by Peter Behrens in ch 14 in this volume, an important critique has emerged over the past decade from influential scholars such as Fritz Scharpf, Dieter Grimm and Susanne Schmidt, who take issue with the fact that granting constitutional status to the Treaty provisions on the internal market (and competition law) has detached them from the democratic process in the Member States and the EU itself, while the values they pertain to are highly political and would be ordinary law in the Member States. FW Scharpf, ‘The asymmetry of European integration, or why the EU cannot be a “social market economy”’ (2010) 8 Socio-Economic Review 211. D Grimm, The Constitution of European Democracy (Oxford, Oxford University Press, 2017). S Schmidt, The European Court of Justice and the Policy Process: The Shadow of Case Law (Oxford, Oxford University Press, 2018).

354  Sacha Garben market provisions have placed the judiciary in a too powerful role on politically sensitive questions beyond what is legitimate in a constitutional democracy.88 One of the difficulties with the democratic critique against the fundamentalisation of the free movement provisions, is however that it often – as in my own case – operates hand-in-hand with a concern about the position of ‘the social’ in the EU construct. This concern tends to go somewhat beyond the observation by Somek cited above that over-constitutionalisation of the free movement provisions reinforces economic liberalism, at the expense of any other public interest, including ‘the social’. It is also because the specific way in which the EU constitution has developed, has had the consequence of not just fundamentalising the free movement provisions, but also of actively ‘displacing’ the protection of social rights.89 As Poiares Maduro formulates it: [T]he core of the European Constitution has lain in market integration. It is mainly under this legitimacy and through the rules provided in the Treaties to its achievement, that the Court has developed the very notion of a European Constitution. Though the EC Treaty also contains social provisions … at the core of market integration are the free movement provisions promoting access to the different national markets. The Court has defined these free movement provisions as ‘fundamental freedoms’. The fundamental rights character granted to the free movement provisions and the widening of their scope of action in order to extend EU supervision over national regulation, and to support the constitutionalisation of Community law, has led to a spill-over of market integration rules into virtually all areas of national law. Indeed, it has led some to argue that they should be conceived as fundamental economic freedoms limiting public power and safeguarding competition in the free market. This dynamic … has been expressed in the fundamental rights discourse of the Court. Most cases on fundamental rights decided by the Court under the doctrine now enshrined in Article 6(2) TEU addressed economic rights and freedoms such as the right to property and freedom of economic activity. At the same time the status of social rights in the Treaties appeared to be in a subsidiary position.90

The problem from this perspective, which overall seems to find more traction in the EU legal community than the ‘mere’ democratic critique against the constitutionalisation of EU economic law, is not so much that the free movement provisions have become ‘fundamental’, but that they have become too fundamental in relationship to other, mostly social, rights. The EU legal order has upgraded free movement provisions while at the same time downgrading fundamental social rights. This brings us to the question of hierarchy or ‘balance’ between the fundamental freedoms and (other) fundamental rights.

88 Garben, ‘The Constitutional (Im)balance’, above n 59. 89 On ‘social displacement’ in the EU legal order, see C Kilpatrick, ‘The displacement of Social Europe: a productive lens of inquiry’ (2018) 14 European Constitutional Law Review 62. 90 M Poiares Maduro, ‘The Double Constitutional Life of the Charter’ in T Hervey et al, Economic and Social Rights Under the EU Charter of Fundamental Rights (Oxford, Hart Publishing, 2003) 285.

Fundamental Freedoms and Fundamental Rights  355

C.  The Hierarchy or ‘Balance’ between the Fundamental Freedoms and (Other) Fundamental Rights: Some are More Equal than Others Oliver and Roth in a seminal article on the subject remarked that ‘at the end of the day, the essential issue is not whether the four freedoms are to be categorized as fundamental rights, but rather their relative importance in the Treaty’.91 There is truth to that, but at the same time, it would seem that the question of the relative importance in the EU constitutional settlement accorded to the internal market freedoms and (other) fundamental rights respectively, is – or should be – intimately connected to the question whether they do or do not qualify as such. De Boer argues that ‘(w)hether the Treaty freedoms are to be seen as fundamental rights and as hierarchically equal should affect the way in which we think they are to be balanced in case of conflict’.92 It could legitimately be argued that fundamental rights are supposed to be the most fundamental norms in a legal order, thus trumping all other rights in hierarchy, and that they are the only substantive rules that can lead the judiciary to invalidate an act when procedurally validly adopted by the legislator. As to the former point, the Kadi ruling of the CJEU lends some credence to the argument that indeed, also in the EU legal order, fundamental rights take pride of place on the highest hierarchical rung,93 for it is difficult to imagine that the CJEU would have refused to accept the Council’s execution of a UN Security Council act if it had been a question of infringement of one of the free movement provisions (although it may be noted that one of the fundamental rights in play was an economic one: the right to property). But whatever the status of fundamental rights in the EU’s external dimension, internally they do not seem to enjoy any higher status than the free movement provisions. And the latter are, as we know, used rather relentlessly to circumscribe the legislative process at national and EU level. One could argue, on the basis of a wealth of constitutional and democratic theory considerations, that such a treatment of the free movement provisions is only legitimate if and to the extent that they qualify as fundamental rights. The EU law approach seems to have this upside down. Probably due to the particular ways in which the EU legal order has developed, it has not been built towards orienting ideals of democracy, self-determination, representation, constituent power, separation of powers and fundamental rights, both as

91 Oliver and Roth, above n 34. 92 N de Boer, ‘Fundamental Rights and the EU Internal Market: Just how Fundamental are the EU Treaty Freedoms – A Normative Enquire Based on John Rawls’ Political Philosophy’ (2013) 9 Utrecht Law Review 148, 148–49. 93 Cases C-402/05 P and 415/05 P Kadi and Al Barakaat International Foundation v European Council EU:C:2008:461.

356  Sacha Garben legitimating foundations and as outcomes to be achieved,94 but instead these primordial concepts of legitimate authority have been attached to the project somewhat ornamentally and incidentally. Fundamental rights have come into play as a side effect of economic integration on the basis of the free movement provisions, and the latter have been the analytical starting point and driving source of the EU’s authority. In traditional EU law, the fundamental status of the free movement provisions goes unquestioned (who cares if they are ‘fundamental rights’ or not, they are the fundamental freedoms!), and it is rather the authority of fundamental rights that is up for discussion. The issue does not seem to be whether fundamental rights rank higher than the free movement provisions, but rather whether they are on a par with (regardless of the question whether the free movement provisions deserve such a fundamental status) or perhaps even inferior to free movement. In that context, the EU Charter is held up as a celebratory moment as it, post-Lisbon, received formal recognition of having the ‘same legal value as the Treaties’.95 As Nic Shuibhne states, this entails ‘legal equivalence for Treaty freedoms and fundamental rights’.96 She furthermore points out that in accordance with Article 2 TEU, ‘[t]he Union is founded on the values of respect for human dignity, freedom, democracy, equality, the rule of law and respect for human rights’, which could provide an argument for priority of fundamental rights over free movement provisions.97 However, there is more than a single judgment available to support the argument that the free movement provisions are given more forceful protection and higher status in EU law than certain fundamental rights, especially collective rights98 such as the freedom of assembly and the right to strike,99 the right to collective bargaining100 and the right to protection in case of collective dismissal.101 As many fundamental social rights are collective rights, this has led to a certain hierarchy of fundamental rights that generally favours the economic over the social. This applies both to the pre- and post-Charter case law.102

94 See Cahill, above n 1, 28. 95 Art 6(1) TEU. 96 N Nic Shuibhne, ‘Fundamental Rights and the Framework of Internal Market Adjudication: Is the Charter Making a Difference?’ in P Koutrakos and J Snell (eds), Research Handbook on the Law of the EU’s Internal Market, Research Handbooks in European Law (Cheltenham, Edward Elgar Publishing, 2017). 97 ibid. 98 A Lo Faro, ‘Toward a de-fundamentalisation of collective labour rights in European social law?’ in M Moreau (ed), Before and After the Economic Crisis, What implications for the European Social Model? (Cheltenham, Edward Elgar, 2011). 99 Case C-438/05 International Transport Workers’ Federation and Finnish Seamen’s Union v Viking Line EU:C:2007:772; Case C-341/05 Laval un Partneri EU:C:2007:809. 100 Case C-426/11 Mark Alemo-Herron and Others v Parkwood Leisure Ltd EU:C:2013:521. 101 Case C-201/15 Anonymi Geniki Etairia Tsimenton Iraklis (AGET Iraklis) v Ypourgos Ergasias, Koinonikis Asfalisis kai Koinonikis Allilengyis EU:C:2016:972. 102 Nic Shuibhne, above n 96.

Fundamental Freedoms and Fundamental Rights  357 The Court’s official line is to insist that all are equal.103 Some are just, it turns out, more equal than others. Of course, in the CJEU’s analytical framework, indistinctly applicable restrictions of the free movement provisions can be justified by reference to the protection of fundamental rights.104 Former CJEU President Skouris considers, citing Omega105 and Schmidberger,106 that the order in which the Court of Justice looks at fundamental freedoms and fundamental rights is not indicative of a hierarchy between the two. It is no more than the consequence of the fact that the Court has jurisdiction only when fundamental freedoms come in to play. In other words, pertinence to fundamental freedoms is a necessary pre-requisite for Community law to apply. It follows from that that they are to be examined first.107

While it is of course true that a prima facie restriction of the free movement provisions has to be established first, it does not mean that the Court has to follow the analytical framework that it imposes in terms of justification. The Court could, after establishing that there has been a prima facie restriction, consider that we are in the scope of EU law for the purposes of both the free movement provisions and the fundamental right in question as protected by the EU Charter, and conduct a ‘bi-directional’ assessment:108 considering both the justification of the infringement entailed by the fundamental right on the free movement provision, and the

103 Skouris, ‘Fundamental Rights and Fundamental Freedoms’, above n 32; K Lenaerts, ‘Exploring the Limits of the EU Charter of Fundamental Rights’ (2012) 8 European Constitutional Law Review 375, 392–93. Trstenjak and Beysen, above n 34, consider that fundamental rights and free movement provisions should be on a par with an exception with regard to Title I of the EU Charter, but that the CJEU has not respected that with regard to the judgments in Viking and Laval. That view is shared by S Weatherill, who considers the case law to generally show an equality between fundamental rights and the free movement provisions, but states that Viking and Laval are ‘the rulings which are most vulnerable to the allegation that, for all the rhetoric of respect for non-economic values, EU free movement law is contaminated by an economic bias’ and that ‘disturbingly, the Court did not follow the model regularly preferred in the case law considered above: it left wholly out of account any margin of appreciation apt to permit recognition of local circumstances’ in these rulings. S Weatherill, ‘From Economic Rights to Fundamental Rights’ in S de Vries, U Bernitz and S Weatherill (eds), The Protection of Fundamental Rights in the EU After Lisbon (Oxford, Hart Publishing, 2013) 27–28. For further criticism on the approach in Viking and Laval, see the literature cited below, n 109. The approach has also been condemned internationally as in breach of fundamental rights by the ILO and Council of Europe’s European Social Committee. 104 It is not yet clear whether, apart from indistinctly applicable restrictions, directly discriminatory restrictions can also be justified by reference to fundamental rights. On the basis of the Court’s general approach, one had reason to expect that the Court would be highly reluctant to allow the justification of directly discriminatory measures on this basis, but would contradict the assertions that the fundamental freedoms and fundamental rights rank ‘equal’. See Trstenjak and Beysen, above n 34, 312. 105 Case C-36/02 Omega Spielhallen v Oberbürgermeisterin der Bundesstadt Bonn EU:C:2004:614. 106 Case C-112/00 Eugen Schmidberger, Internationale Transporte und Planzüge v Republik Österreich EU:C:2003:333. 107 Skouris, ‘Fundamental Rights and Fundamental Freedoms’, above n 32. 108 Schiek, above n 19, 629.

358  Sacha Garben justification of the (potential) infringement entailed by the free movement provision on the fundamental right. In that bi-directional assessment, it could assume the equivalence of all fundamental rights and freedoms, it could determine the ‘essence’ of each fundamental right and freedom in isolation, or it could come to an integrated and hierarchical understanding of the meaning of fundamental rights and freedoms in the EU legal order which would inform the interpretation of the substance of each right and their relationships. Alternatively, or in conjunction with the above approach, it could consider that in the case of fundamental rights, the Member State should benefit from a wide margin of appreciation, and thus conduct a very deferential proportionality test. The Court does not really do any of these things. While it arguably carried out a deferential proportionality test in Omega, it clearly did not in other cases such as Viking and Laval.109 Schmidberger is an interesting example because it is often held up as a good practice of the Court in relation to the fair balancing of fundamental freedoms and rights.110 While the Court did say that the ‘Member States enjoy a wide margin of discretion’,111 it did so only to continue that ‘nevertheless, it is necessary to determine whether the restrictions placed upon intra-Community trade are proportionate in the light of the legitimate objective pursued’112 and to do the exact opposite of applying a wide margin of discretion with a very detailed assessment of all the extenuating elements that could eventually serve to justify this infringement of what for the Court is clearly a particularly fundamental right: the free movement of goods. Despite appearances, the assessment is unidirectional, and thus inherently not ‘fair’.113 The Court did consider that taking account of the Member States’ wide margin of discretion, in circumstances such as those of the present case, the competent national authorities were entitled to consider that an outright ban on the demonstration would have constituted unacceptable interference with the fundamental rights of the demonstrators.114

That is however very far from the idea to assess with equivalence whether the free movement of goods, if interpreted to preclude the demonstration, would not lead to an infringement of the fundamental rights of the demonstrators. It is even

109 See J Malmberg and T Sigeman, ‘Industrial actions and EU economic freedoms: the autonomous collective bargaining model curtailed by the European Court of Justice’ (2008) 45 Common Market Law Review 1115, 1130; de Vries et al, above n 6, 70; Trstenjak and Beysen, above n 34, 312; A Davies, ‘One step forward, two steps back? Laval and Viking at the ECJ’ (2008) 37 International Law Journal 126; T Novitz, ‘A human rights analysis of the Viking and Laval judgments’ (2007–08) 10 Cambridge Yearbook of European Legal Studies 541; S Sciarra, ‘Viking and Laval: collective labour rights and market freedoms in the enlarged EU’ (2007–08) 10 Cambridge Yearbook of European Legal Studies 563; P Syrpis and T Novitz, ‘Economic and social rights in conflict: political and judicial approaches to their reconciliation’ (2008) 33 European Law Review 411. 110 See Trstenjak and Beysen, above n 34, 312. 111 Para 82 of the judgment. 112 ibid. 113 ibid, para 81. 114 ibid, para 89.

Fundamental Freedoms and Fundamental Rights  359 further from the idea that, since we are in the scope of EU law and Member States are acting as agents of the EU in this case (all the more since the demonstration was at least partially about the environmental impact of EU transport policy),115 the assessment should lead to the conclusion that Member States are obliged (rather than authorised, through gritted teeth) to allow such demonstrations by virtue of EU fundamental rights. The important point is decisively not that Austria in this case gets away with the measure. This outcome in no way shows that there is equivalence nor that there would even be precedence accorded to the freedom of expression and assembly. In the abstract, the normative framework developed by the Court values free movement more. The unidirectional analytical framework means that the stakes are unfair. The demonstration (and thus the expression of the freedom of expression and assembly) risks being excluded by order of EU law. The only thing it can achieve is that the Member State gets permission (but is not obliged) to allow it. On the other hand, free movement (or, the prohibition of the demonstration) will (apparently) never be excluded by EU law and thus can still be permitted by the Member States. Concretely, the conditions outlined in paragraphs 83–92 of the judgment mean that in many situations, free movement will be given precedence over the right to demonstrate, and there may furthermore be a chilling effect116 emanating from the judgment in that Member States may choose to interpret the conditions strictly in future situations to exclude the risk of being condemned by the Court (again, because in the Court’s analytical framework, the Member State does not risk being condemned for prohibiting the demonstration, even if we are clearly in the scope of EU law). For all the lofty rhetoric about fair balancing and commitment to fundamental human values, when it comes down to it, the Court applies the free movement provisions as the rule, and fundamental rights as the exception. In determining whether there should be a margin of discretion for national authorities, the approach is not one that looks at the fundamental nature of the right in question and its purpose, but instead, as Weatherill says, the more sensitive and the more remote from commercial considerations the matters advanced in the context of justification of trade barriers are, the more generous the Court is to the available scope for justification and also to the breadth of the margin of appreciation enjoyed by the regulator.117

Davies has argued against this framework in relation to all sorts of public interest justifications, stating: The starting point for the Court when adjudicating on either ‘mandatory requirements’ or Treaty derogations is that they should be interpreted restrictively, because 115 J Morijn, ‘Balancing Fundamental Rights and Common Market Freedoms in Union Law: Schmidberger and Omega in the Light of the European Constitution’ (2006) 12 European Law Journal 15. 116 L Hayes, T Novitz and H Reed, ‘Applying the Laval Quartet in a UK Context: Chilling, Ripple and Disruptive Effects on Industrial Relations’ in A Bücker and W Warneck (eds), Reconciling Fundamental Social Rights and Economic Freedoms after Viking, Laval and Ruffert (Baden-Baden, Nomos, 2011). 117 Weatherill, ‘From Economic Rights to Fundamental Rights’, above n 103, 25.

360  Sacha Garben they represent limitations on the fundamental freedoms of Europeans. A restriction on free movement involves a Member State pleading special interests in order to deprive a European of the freedoms that are part of their legal ‘heritage’ and so this must be strictly supervised and regarded as exceptional. This has a certain rhetorical grandeur, but is substantively nonsense. When national laws conflict with free movement there is substantive legitimacy on both sides. One can equally phrase the situation as free movement law restricting the choices made by democratic processes within the national constitutional order. Respect for national constitutional orders, for substantive subsidiarity, and for democracy entails that free movement must in such a situation be interpreted restrictively. Certainly, there seems little to be gained from a rhetorical battle to see whether it is national democratic choices or free movement which can be phrased in ways seeming to give them the most legitimacy. However, the Court’s choice to put one side of the conflict on a high moral ground while the other is then treated as implicitly less important does not seem to rest on any objective view of what might give rules or policies legitimacy or moral weight. Rather than interpreting restrictions restrictively, they should simply be interpreted fairly: the interests represented should be respected, albeit not misused. A derogation is something that is regarded as more important than the policy derogated from, not less: that is what makes it a derogation.118

This would apply with all the more force when it concerns fundamental rights that are to be weighed against free movement. The Charter does not seem to have made a fundamental difference in this regard. As Nic Shuibhne states, ‘reviewing the post-Lisbon case law on internal market adjudication, neither the methodology nor priorities engaged by the Court have yet adjusted to reflect or to manage more successfully the inevitable tensions that emerge between economic rights and fundamental freedoms’.119 In fact, it appears that in some respects, the coming into force of the Charter has aggravated the problem. In particular, the Court has given a very forceful interpretation of Article 16 EU Charter on the freedom to conduct a business. In Alemo-Herron,120 a judgment so poor that according to Weatherill it deserves to be ‘consigned to the bottom of an icy lake’,121 the Court used Article 16 to read a minimum harmonisation directive adopted on an internal market basis but intended to protect the interests of workers in the event of a transfer of undertakings in a way that it precluded a Member State from providing, in the event of a transfer of an undertaking, that dynamic clauses referring to collective agreements negotiated and adopted after the date of transfer are enforceable against the transferee. The

118 Davies, above n 16. 119 Nic Shuibhne, above n 96. 120 Case C-426/11 Mark Alemo-Herron and others, above n 100. 121 S Weatherill, ‘Use and Abuse of the EU’s Charter of Fundamental Rights: on the improper veneration of “freedom of contract”’ (2014) 10 European Review of Contract Law 167.

Fundamental Freedoms and Fundamental Rights  361 judgment has been widely criticised on a variety of grounds.122 What is interesting for our purposes, is – first – that Article 16 was given an interpretation that goes very much in the direction advocated by Usai, namely that it should be used to ‘push the throttle in favour of an even more developed economic union’ by allowing the right to conduct business to be used as a ‘safeguard against barriers that the member states may want to put up in the internal market’ even in purely internal situations.123 Second, the Court does not even mention, let alone fairly balance, the various fundamental social rights that would be relevant to the case, and that should at least have been used to balance Article 16 of the EU Charter, such as Article 28 of the EU Charter on collective bargaining and Article 31 of the EU Charter on fair and just working conditions. Similarly, in AGET,124 the Court practically ignores125 the fundamental social right in Article 30 of the EU Charter on protection against unjustified dismissal while rather forcedly drawing Article 16 of the EU Charter into its assessment of a restriction of the freedom of establishment. The Court in that judgment deploys a language that seems more cognisant of the need for public interest restrictions of Article 16 of the EU Charter, but the asymmetry in the treatment of economic fundamental rights and fundamental social rights is striking, as is the conflation of the freedom of establishment and the freedom to conduct a business. Taken together, these developments suggest that the Charter is leading to an even more forceful application of the free movement provisions in conjunction with Article 16 of the EU Charter, moving towards a general ‘economic freedom’ approach, while it is instead a bit of a ‘damp squib’126 for the fundamental social rights contained therein.

122 ibid; X Groussot, GT Pétursson and J Pierce, ‘Weak Right, Strong Court – The Freedom to Conduct Business and the EU Charter of Fundamental Rights’ (2014) Lund University Legal Research Paper 01/201; J Prassl, ‘Freedom of contract as a general principle of EU law? Transfers of undertakings and the protection of employer rights in EU labour law’ (2013) 42 Industrial Law Journal 434; P Syrpis and T Novitz, ‘The EU Internal Market and Domestic Labour Law: Looking Beyond Autonomy’ in A Bogg et al (eds), The Autonomy of Labour Law (Oxford, Hart Publishing, 2015); M Bartl and C Leone, ‘Minimum Harmonisation after Alemo-Herron: The Janus Face of EU Fundamental Rights Review’ (2015) 11 European Constitutional Law Review 140. 123 A Usai, ‘The Freedom to Conduct a Business in the EU, its Limitations and its Role in the European Legal Order’ (2013) 14 German Law Journal 1871. For discussion see E Gill-Pedro, ‘Freedom to conduct a business in EU law: freedom from interference or freedom of domination?’ (2017) 9 European Journal of Legal Studies 103. 124 Case C-201/15 Anonymi Geniki Etairia Tsimenton Iraklis (AGET Iraklis), above n 101. 125 Art 30 is mentioned once, almost as an afterthought, to indicate that Art 16 may be limited, but is not made an integral, let alone equal, part of any ‘balancing’. Para 89 of the judgment. 126 C Barnard, ‘EU Employment Law and the European Social Model: The Past, the Present and the Future’ (2014) 67 Current Legal Problems 199, 207. See also Niall O’Connor, ‘(Re)constructing the employment law hierarchy of norms: The Charter will not, should not and need not apply?’ (EU Law Analysis, 14 December 2017), available at: eulawanalysis.blogspot.com/2017/12/reconstructingemployment-law-hierarchy.html.

362  Sacha Garben

III.  Towards an Integrated Democratic Interpretation Framework for All EU Fundamental Rights A.  The Need for an Integrated Theory and Interpretation of the Different Parts of the EU Legal Order As stated above, the EU is in need of a ‘thick’ constitutional theory to orient and legitimise the exercise of its authority in all areas. This is necessary if the Union is going to successfully carry its own claims of autonomy, both from a normative standpoint and more practically speaking in terms of the continuing/renewed acceptance of these claims especially by Europe’s citizens, civil society and national courts. Based on the constitutional traditions of the EU Member States and the post-Lisbon European Union itself, it should be reasonably common ground that fundamental rights need to take a central place therein, together with the principles of democracy and constitutionalism/Rule of Law.127 As Chalmers and Trotter say in relation to fundamental rights in this respect: EU law has been established as a legal order, which carries with it a claim to possess these qualities of coherence, authority and moral pedigree. To meet this claim, a vision of what EU law is about has been established with fundamental rights a constitutive part of it. Interpretation of EU law in the light of this vision allows these claims to be addressed as it relates individual laws to one another, sets out reasons for obeying EU law and sets out what is good and right about the EU.128

They highlight that an integrated method is currently missing, both concerning the Charter/EU fundamental rights as such, and their role in the wider EU legal order (and vice versa): A series of discrete doctrines govern methods of interpreting the Charter, the standard of protection secured by it, its relationship to other instruments and its scope of review. These are spliced together to form EU fundamental rights law. Such splicing is thin on how different aspects of EU fundamental rights law relate to each other or to EU law as a whole. Without this, however, fundamental rights do not have the coherence to set out an imaginary of what is good and right, something central to both their moral status and their iconographic appeal.129

The current case-by-case, right-by-right assessment, without a sensible hierarchical structure of EU law’s plethora of constitutional values and norms, results in a high measure of unpredictability in the adjudication on the various EU Charter rights, their direct/horizontal applicability, and their interaction or mutual 127 See in more detail the various contributions in Garben, Govaere and Nemitz, above n 18, and in particular S Garben, ‘The Principle of Legality’, above n 18. 128 D Chalmers and S Trotter, ‘Fundamental Rights and Legal Wrongs: The Two Sides of the Same EU Coin’ (2016) 22 European Law Journal 9. 129 ibid.

Fundamental Freedoms and Fundamental Rights  363 ‘balancing’. Furthermore, the absence of a coherent approach to the specific question to what extent the free movement provisions are considered part of EU fundamental rights law, and the equally important related questions of what hierarchical status should be accorded to those (aspects of the CJEU’s current interpretation of the) free movement provisions that do not qualify as fundamental rights, and the interaction between the fundamental and non-fundamental parts as well as the relationship with social fundamental rights, is unsustainable for the Union going forward. This issue has been the EU legal order’s ‘Achilles heel’ towards national courts and legal communities for a long time, and combined with the (related?) fading of the ‘permissive consensus’ and rise of a ‘constraining dissensus’ to European integration,130 and the failure of the EU Charter and general Lisbon reforms to resolve it, it is high time to address it now, with the potential of a fundamental reform for the future of the EU post-Brexit (and post-COVID) looming – even if distantly – on the horizon.131 This is as much a case to establish the normative superiority of fundamental rights in the EU constitutional order, as it is to counter the current approach of internal market exceptionalism, where the free movement provisions float elusively, somehow separate and somehow above, the rest of the legal and political order. As such, it aligns with Schiek’s proposal to constitutionally embed the internal market by reference to the EU Charter132 and – as Nic Suibhne has argued – does justice to Article 26(2) TFEU, which by emphasising that free movement is ensured in accordance with the provisions of the Treaties, ‘suggests that the goal of realising a free movement-driven internal market sits within – and is in fact contained by – the wider structure of the Treaties and the many objectives committed to therein’.133 It also means that any pursuit of coherence between the free movement provisions as such should be abandoned, and that instead we need to embrace the idea that some free movement provisions are to be considered and treated as more fundamental than others. For as the analysis in the previous section already suggests, and as shall be set out in more detail in the following subsection, the aspects of free movement that place the equality and dignity of the individual citizen central, are the ones that deserve to be fully protected as fundamental rights, whereas for the economic interests of businesses in the internal market, the free movement provisions can only legitimately give constitutional protection in relation to protectionist measures, and not those that impede market access to foreign companies in the same way as new national market entrants. Any further protection one may want to accord to the other aspects of the free movement provisions should be decided in the political, legislative, process, and not on the basis of the direct application of primary law. 130 L Hooghe and G Marks, Multi-level Governance and European Integration (Lanham, MD, Rowman & Littlefield, 2001). 131 The Debate on the Future of Europe, scheduled to take place in May 2020, has been postponed due to the COVID-19 crisis to a date yet undetermined at the time of writing. 132 Schiek, above n 19. 133 Nic Shuibhne, above n 96.

364  Sacha Garben

B.  An Integrated Interpretation Framework Rooted in Democratic Constitutionalism In the highly sensitive process of judicial review of democratically enacted norms, the judiciary should be guided not so much by an approach of deference per se, but by a clear normative framework within which the fundamental values that it is mandated to uphold are to be accommodated and, importantly, which states how these fundamental values relate to one another, especially in cases of conflict, and how deeply the judiciary is warranted to intervene to uphold them. This would make the adjudication of these values and rights more predictable (a cornerstone of the rule of law) and would provide a more legitimate framework for balancing than one which pursues the effectiveness EU law and a maximisation of the integration of national legal orders. In other words, an ideological meta-framework is needed for constitutional adjudication in the EU legal order.134 This should not be an ideology of political economy, in reference to either an idea of a liberal market economy or a social market economy (even so for the latter, there is an explicit constitutional commitment), as this can ultimately never result in anything but a subjective and highly political view of socio-economic justice that is not suitable for the judiciary to actively enforce. Instead, as already indicated further above, based on the constitutional traditions of the Member States and the EU’s own core values we can consider the common ground of contemporary constitutional democracy to be the commitment to the triptych of democracy, the Rule of Law and fundamental rights,135 and within that context the latter to be concerned with upholding human dignity and guaranteeing a robust and inclusive democracy.136 That provides us with the necessary elements of a legitimate interpretation framework. Of course, that does not provide absolute certainty or conclusive answers to each and every case, and it should not. It provides the appropriate terms within which the arguments can reasonably be made, and within which conclusions can reasonably be drawn. Using that framework, we have already in section II established the extent to which it is legitimate to constitutionalise/fundamentalise the free movement provisions. The same framework should be used to interpret the other fundamental rights as contained in the EU Charter, their substantive interpretation and their interaction, as well as the EU legal order more generally. A central idea underlying the framework is that for a legal order to be legitimate in its exercise of public authority, it needs to serve the principles of democracy and constitutionalism to an equal extent.137 This implies that the judiciary is allowed to overturn democratically decided norms only to the extent that it is necessary to uphold



134 Conway,

above n 12. Govaere and Nemitz, above n 18. 136 Gearty, above n 24 88. 137 Tully, above n 17, by reference to the work of Habermas and Rawls. 135 Garben,

Fundamental Freedoms and Fundamental Rights  365 the key elements of constitutionalism, namely procedurally the decision-maker’s respect of the prescribed decision-making rules, and substantively the respect for fundamental rights.138 The substance of these fundamental rights should in turn be informed by (i) human dignity as the source of all human rights, and (ii) the principle of democracy (which is both an important element of the individual and collective agency required by human dignity and a self-standing concern alongside constitutionalism).139 The fundamental rights that the judiciary should enforce against the legislator are therefore those that guarantee the necessary conditions for respect of and commitment to human dignity, and those that are necessary for an inclusive, robust and long-term democracy. The latter entails, apart from the central democratic rights such as freedom of expression and assembly, ensuring a minimum degree of socio-economic equality (through social mobility, inclusion and emancipation, as well as non-discrimination and the correction of power asymmetries). Every fundamental right should be interpreted in light of this, and where several rights appear to conflict, the ‘balancing’ needs to be performed on the basis of these yardsticks, of course in combination with a textual reading of the provisions in question (in particular, for instance, where the EU Charter indicates that the right is only recognised ‘in accordance with Union law and national laws and practices’).140 It is beyond the scope of this chapter to exhaustively apply the framework to all the fundamental rights in the Charter and their interrelationship. One particular issue deserves to be treated in a little more detail, however, namely the thorny relationship between the free movement provisions and fundamental social rights. The foregoing analysis has set out the reduced scope of fundamentalisation that the current approach allows in relation to free movement, as well as the freedom to conduct a business in Article 16 of the EU Charter. It is important to emphasise that fundamental social rights also need to be interpreted in light of the above normative yardsticks and can be constitutionally enforced only to the extent that they are necessary for human dignity or a well-functioning, inclusive, long-term democracy. This approach of democratic constitutionalism to the purpose of fundamental social rights could help overcome the ‘currently inadequate philosophical underpinning of labour rights (that) deprives their advocates of depth of argument’141 and provide a more robust way to deal with conflicts between the internal market and Social Europe than approaches that argue – even if understandably so – for a more social approach as such and for that reason. 138 Garben, ‘The Principle of Legality’, above n 18. 139 See P Gilabert, Human Dignity and Human Rights (Oxford, Oxford University Press, 2018) ch 10. 140 It may be that on this basis, a general hierarchy of fundamental rights emerges, in which those that are more directly essential for democracy and dignity are placed in a first tier, while others that contribute thereto more indirectly, to be ranked second-tier. See for instance N de Boer, who has proposed to use Rawls’ theory of justice to determine the interpretation and hierarchy of fundamental rights including the free movement provisions to the extent that they foster equality of opportunity: de Boer, above n 92. 141 R Croucher et al, ‘A Rawlsian philosophical basis for core labour rights’ (2011) 33 Comparative Labor Law & Policy Journal 297.

366  Sacha Garben The following points are a very rough first, tentative attempt to apply these insights to the interpretation of the economic and social rights that have been crucial in the narrative so far.142 This is intended to stimulate further debate along these lines of thinking: • The right to strike and bargain collectively have an important role in creating and maintaining the conditions for a robust democracy. They are ‘procedural’ rights rather than establishing certain socio-economic outcomes. As the UN Special Rapporteur put it: ‘protecting the right to strike is … about creating democratic and equitable societies that are sustainable in the long run. The concentration of power in one sector – whether in the hands of government or business – inevitably leads to the erosion of democracy, and an increase in inequalities and marginalization with all their attendant consequences. The right to strike is a check on this concentration of power’. It is also important for the sense of individual and collective empowerment and agency necessitated by human dignity, of being able to take a minimum degree of control over the conditions of one’s life and take a stand alongside others in solidarity. This right should therefore, in principle, be forcefully protected by the judiciary, both against democratic decision-making (at national and European level) as well as in relation to other fundamental rights, especially where they clash. • Substantive workers’ rights or interests such as to be protected against collective dismissal and negative repercussions of restructuring/company transfer, are political decisions in a well-functioning democracy, and the extent of protection should therefore primarily be for the democratic process to flesh out, as it would seem that there is significant scope for counterbalancing with other rights and interests. Strong protection would need to be provided against unjust dismissal generally, as the lack of such protection would undermine the enforcement of any other right in an employment context and would entrench an unacceptable power asymmetry between the employer and the individual worker which cannot be justified in a democratic society. But the scope for collective dismissal for economic reasons is not something that should a priori be determined by the judiciary or the constitution, and instead is a main battleground for the political process. • The fundamental internal market rights serve the purpose, as discussed, of (i) providing individual citizens with the right to seek life/economic opportunities across the entire reach of their federal territory and to be treated equally as other citizens, and (ii) of compensating for the transnational ‘democracyfailure’ where a Member State takes a decision specifically negatively affecting (citizens of) other Member States. They should thus be interpreted as focused specifically on reviewing direct and clear indirect discrimination – and not on

142 See also S Garben, ‘Towards an Adjudication Framework for Economic and Social Fundamental Rights in the EU’ (2020) European Labour Law Journal (forthcoming).

Fundamental Freedoms and Fundamental Rights  367 the general economic freedom of companies, and there is significant scope for counterbalancing. The internal market rights, on this democratic logic, should not be interpreted as forcing national democracies to take decisions applicable on their own territory that are detrimental to their own interest for the benefit of other jurisdictions’ interests, unless this is necessary for the dignity of EU citizens from other Member States. • The freedom to conduct a business has a certain, albeit limited role, in ensuring a robust democracy, to the extent that it helps to empower individuals – especially those who are disadvantaged – through economic activity, to prevent concentrations of (economic) power and equalise societal asymmetries. This supports a reading where the right is about fostering the possibility of individual entrepreneurship, in the sense of enabling citizens to set up an economic activity or join a profession, within a broader picture of creating more socioeconomic progress and equality in society. As a human right it should, upon such a constitutional democratic reading, not provide a ‘sword’ for companies in the operation of their business against workers, citizens and general public interest standards: that would be counterproductive. The current approach under EU law does not seem to follow the above framework, which could be argued raises profound legitimacy questions. This transpires clearly from the various controversial judgments of the CJEU where it has ‘balanced’ fundamental social rights and fundamental economic rights. For one thing, as many have argued, in the seminal and controversial Viking and Laval rulings, economic fundamental rights were taken too far. The above principles would support that criticism, but from a perspective that transcends the ‘economic versus social’ battle, pointing out that on the basis of the meta-values of democratic constitutionalism, the right to strike should receive a forceful interpretation while the internal market provisions are more limited in scope. It is questionable whether there should have been a prima facie restriction of the fundamental aspects of the free movement rights at all, as they should be about combating unjustified (indirect) discrimination against foreign actors. Arguably, this was not the case in Viking and Laval, as in the former case the strike opposed a national company seeking to ‘reflag’ as a foreign actor to escape the national democratic process, and in the latter case the strike was about procuring equal treatment of foreign workers with national workers. As a second example, we can apply the above interpretation principles to Alemo-Herron and AGET. While the social interest of protecting workers against the negative repercussions of company restructuring that were at stake in these cases should, be open to significant (economic) counterbalancing, the counter-interest of freedom of establishment or conducting a business was not even legitimately in play. In AGET, the national rules applied equally to all companies. In Alemo-Herron, no prima facie breach of the internal market was established, and Article 16 of the EU Charter therefore should not have been in discussion at all. And even if it were, this should only be a right for individuals to set up a business or access a

368  Sacha Garben profession (in limited circumstances), and not a right for businesses to conduct economic activities (hiring and firing, contracting) without restriction. And even if the freedom to conduct a business had been considered relevant to the facts of the cases, our above adjudication and interpretation framework suggests that it is the political process that should have decided on the balance. The EU legislator had, in the directives at stake, not decided on that balance as regards the cases at hand – they fell outside the scope of the directives. Thus, it should have been left to the national democratic process: the UK and Greek rules in question should have been upheld until the European legislator had acted to establish a different balance at EU level on these specific questions. Instead, the EU judiciary gave a(n overly) forceful interpretation to the freedom to conduct a business, thereby constitutionalising a certain view on the right socio-economic balance that was not necessary to establish for a democratic society, and was in fact counterproductive to it. These judgments further empower the powerful and entrench societal inequalities at national and European level, disguised as constitutionalism but instead offending the true values of human rights and their meaning for constitutional democracy and democratic constitutionalism.

IV. Conclusion The free movement provisions currently occupy a status in the EU legal order that in most constitutional democracies would only be conferred on fundamental human rights, and they are – maybe not always, but definitively sometimes – treated with a degree of priority over other fundamental rights. Yet, they are not integrated into EU fundamental rights law, nor embedded in any sound constitutional theory of the EU legal order. Such internal market exceptionalism and myopic integrationism is practically and normatively unsustainable. This chapter has made the argument that to a limited extent, a constitutionalisation of the free movement provisions can be defended by transnational democratic and individual justice (dignity) arguments: namely to the degree already covered by Article 15 of the EU Charter, and in the way that they give economic actors a (rebuttable) right not to be subjected to protectionist measures. To the extent that the direct application of the free movement provisions against national measures and in relation to the EU legislative process goes beyond this legitimate degree of ‘fundamentalisation’, they have – as some influential scholars have critically agued – indeed been over-constitutionalised. As this poses deep legitimacy concerns, the CJEU needs to fundamentally reconfigure the free movement provisions, launching an ‘Internal Market 2.0’ that is anchored in a robust theory of democratic constitutionalism. I concur with Schiek that it is necessary and possible to derive a different normative frame of constitutionalisation at EU level from the ‘ambiguous character of EU constitutional law’, and that thinking in such holistic terms is necessary

Fundamental Freedoms and Fundamental Rights  369 to ‘develop a wider discourse on the adequate constitutional frame for the future of EU integration’ on which any changes in the case law of the Court will depend.143 This chapter has accordingly proposed an integrated adjudication framework, within which all fundamental rights, including to a limited extent the free movement provisions, can be interpreted and balanced in a way that does justice to human dignity and transnational democracy. Other aspects of the internal market should be left to the legislative process, as a matter of secondary law. While this framework does not, and should not, provide conclusive answers in itself, it aims to provide ‘the terms of the argument’ within which there can be reasonable disagreement about precise interpretations and resolution of individual cases of conflict between different fundamental rights or between a legislative norm and one or several fundamental rights. Let it be clear that the point of this framework is not to reign in the Court. It is instead, precisely as Davies advocates in chapter fifteen, a call for an activist, forward-looking Court, that takes on itself the responsibility to mature and legitimise the very autonomous constitutional order that it itself created.

143 Schiek, above n 19. While I am furthermore wholly sympathetic to her idea of a socially embedded internal market on the basis of a reading of the Charter that accords a slight precedence to social rights over economic ones, the framework suggested in this paper arrives at a place perhaps not all that different but takes a different road, where democracy is a primary concern and guide, alongside the protection of human dignity.

370

INDEX advertising Hünermund and restriction placed on pharmacists 67, 68, 99, 100 Keck (Joined Cases C-267/91 and Case C-268/91) 99 reach of Art 35 102 rules banning, justifying 102 AGCM (Autorita Guarante della Concorrenza e del Mercato), Italy 230 AGET (Case C-201/15) 8, 367 AI see Artificial Intelligence (AI) Airbnb 137, 152, 213, 218, 220 anti-discrimination policy 226 Akerlof, G 227, 305 Åklagaren v Percy Mickelsson and Joakim Roos see Mickelsson and Roos (Case C-142/05) Åklagarenv Hans Åkerberg Fransson (Case C-617/10) 303 AKT (Case C-533/13) 346 Albany International (Case C-67/96) 346 Alemo-Herron (Case C-426/11) 8, 360, 367 algorithms and labour law 223, 224 Alimanovic (Case C-67/14) 328 Alpine Investments (Case C-384/93) 55–8, 61–2 Anomar (Case C-6/01) 28 Anonymi Geniki Etairia Tsimenton Iraklis see AGET (Case C-201/15) Antonaki, Ilektra 6, 12, 17 Application Programming Interfaces (APIs) 195, 248 open API 188, 207, 208 Article 30 of EEC Treaty see under European Economic Community (EEC) Treaty Artificial Intelligence (AI) 20, 187, 195, 251 ad hoc High Level Expert Group (AI HLEG) 210 ‘AI for good’ (ITU) 211 AI revolution 198 Assessment List on Trustworthy AI (AL TAI) 210–11 embedded AI 197

G20 human-centred AI Principles 211 ‘Trustworthy Artificial Intelligence’ 210 Work Programme of the European Commission 211 Aslam, Farrar et al v Uber (2018) 225 Atlmark (Case C-280/00) 346 Audiovisual Media Services Directive (AVMSD), (Directive 2007/65/EC) 150–1, 243, 244–5 cultural policy provisions 151 obligations on video-sharing platforms 236–7 video-sharing platforms 250 audiovisual services 150–1, 159 Directive see Audiovisual Media Services Directive (AVMSD), (Directive 2007/65/EC) Augenstein, D 344–5 Australia, ACCC Digital Platforms Inquiry 204n36 Austria, indexation of family benefits 121 AVMSD see Audiovisual Media Services Directive (AVMSD), (Directive 2007/65/EC) B&Q plc see Torfaen Borough Council (Case C-145/88) B2B see Business-to-Business (B2B) relationships B2C see Business-to-Consumer (B2C) relationships B2G see Business-to-Government (B2G0 model Babayev, R 167n36, 169n45, 342n38, 346n56, 350–1 banking services 138–9 Banking Union 150 bans, universal 46–7, 50, 62 BAT (Case C-491/01) 273 Baumbast (Case C-413/99) 330 behaviour science, increasing use of findings from 310 Behrens, Peter 10, 16, 353n87 Benedi Lahuerta, S 134

372  Index benefits, indexation of see indexation of family benefits BEREC Regulation (Regulation 2018/1971) 242 Bernard, N 102 Berners Lee, Tim 198 big data 226 bilateral preferential trade agreements 159 Biondi, Andrea 169 Biuro podróży ‘Partner’ (Case C-119/15) 274–5 blockchain applications 212 Bosman (Case C-415/93) 55 Breton, Thierry 4, 188, 205 Brexit 1, 109, 112, 113, 119, 122, 135 fundamental freedoms and rights following 363 referendum result of June 2016 120–1 and single market 187–8 broadcasting, Communication of 2015 on the DSM Strategy 240–1 BRT v SABAM (Case 127/73) 317 Buet (Case 382/87) 274, 275, 277, 278 Bury, Claire 21 Business-to-Business (B2B) relationships 201, 219, 248 securing 205–7 Business-to-Consumer (B2C) relationships competition and value allocation 202–5 consumer protection 219 internet 198, 201 single market 203–5 unfair commercial practices 267 Business-to-Government (B2G) relationships 207, 248 Cahill, M 19, 337 Capital Markets Union/Banking Union 150 capital restrictions derogation from ordinary company law 172–3 discrimination on grounds of nationality 172 golden shares 6, 163, 164, 170–4 non-discriminatory measures derogating from ordinary company law 173 recalibration of test 183 refined test 173–4 refining definition of 6 selling arrangements and Keck 170 see also free movement of capital Carmen Media Group Ltd (Case C-46/08) 36n34

Carpenter (Case C-60/00) 26, 59–61, 62 Cassis de Dijon (Case 120/78) 25 barriers to trade arising out of national legislation differences 45–6, 68 and convergence of the freedoms 56, 57 discrimination and justification of indistinctly applicable restrictions 33–6 landmark judgment 61 mandatory requirements for justification grounds 34, 36, 43, 45 and market access 58 mutual recognition principle 34, 35–6, 95 obstacle to trade resulting from disparities between German and French rules 45–6 potential of 44 proportionality principle 45 scope of application 53 and selling arrangements 49 Chalmers, D 362 Charter of Fundamental Rights, EU 12, 14, 134 data protection 222 fundamental freedoms and rights 341, 342, 352, 360, 364 Article 16 350–1, 361, 365 horizontal applicability between private parties to 60, 62 overriding values 17 social rights protected 113 and Treaty provisions 87 citizens of the EU see free movement of persons Citizens’ Rights Directive (Directive 2004/38) 76, 115, 116 Code of Practice on Disinformation 245 coherence 19, 107 collaborative economy 214, 216, 223, 229 see also Communication of 2016 on the Collaborative Economy; online platform economy; platformisation of the economy Colomer, Ruiz-Jarabo 177 Commission v Belgium (Belgian case) 175–6 Commission v Belgium (Walloon waste) see Walloon Waste (Case C-2/90) Commission v Czech Republic (Case C-525/14) 71–3 Commission v France (riots) (Case C-265/95) 104 Commission v Germany (German insurances) see German insurances (Case 205/84)

Index  373 Commission v Germany (reusable packaging) 106 Commission v Germany (Volkswagen) see Volkswagen (Case C-112/05) Commission v Greece see Infant Formula Milk (C-391/92) Commission v Italy (Italian Trailers) see Italian Trailers (Case C-110/95) Commission v Italy (Lawyers’ fees) see Lawyers’ fees (Case C-565/08) Commission v Italy (Moped trailers) see Moped Trailers (Case C-110/05) Commission v Italy (Museum admissions) see Museum admissions (Case C-388/01) Commission v Italy (Third-party motor liability insurance) see Thirdparty motor liability insurance (Case C-518/06) Commission v Italy (Vinegar) see Vinegar (Case C-193/80) Common Agricultural Policy (CAP) 288, 299 common commercial policy concept 89 common market 46–7, 75 Common Market Law Review, ‘In Search of the Limits of Article 30 of the EEC Treaty’ 65–74 case law following 66–73 Commission v Czech Republic (Case C-525/14) 71–3 Hünermund (Case C-292/92) 67, 68, 99, 100 Infant Formula Milk (C-391/92) 68–9 Keck (Joined Cases C-267/91 and Case C-268/91) 67–8 selected subsequent ‘unfair competition’ cases 69–70 since ‘In Search of the Limits’ (Common Market Law Review) 66–73 Torfaen Borough Council (Case C-145/88) 66–7 use restrictions 70–1 infringement regarding Czech Republic 71 inspiration for 66 product characteristics and selling arrangements 68 see also Common Market Law Review, ‘In Search of the Limits of Article 30 of the EEC Treaty’ Common Market Organisations, within the CAP 299

Communication of 2015 on the DSM Strategy Pillar 1 (better access for business and consumers) 238–41 better consumer protection 239 broadcasting transmissions 240–1 copyright 241 geo-blocking 239–40 portability 240 sales of goods and digital content 238–9 Pillar II (digital networks and services) 241–6 Audiovisual Media Services Directive (AVMSD) 244–5 Electric Communications Code (EECC) 242 GDPR and e-privacy 246 online platforms 243–4 trust and security in digital services 245 Pillar III (maximising the growth potential of the digital economy) 246–9 free flow of non-personal data 20, 247 Public Sector Information Directive (PSI) 247–8 Communication on Shaping Europe’s digital future 80 competences asymmetries of 123 competence creep 10, 11, 348 conferral to the EU 84 EU’s positive competence to adopt harmonising legislation 18 express conferral of shared competence for internal market in TFEU 10 limited, of the EU 123 national, versus free movement paradigm 8 shared 124–5, 261, 263, 264 competition business-to-consumer (B2C) domain 203 ‘competition for eyeballs’ 192 importance of Article 3(3) TEU 87–8 unfair, selected cases 69–70 and value allocation in B2C domain 202–5 Confédération paysanne and others (Case C-528/16) 292, 302–3 conferral principle 5, 308–9 Connected Continent 242 Consten and Grundig (Joined Cases 56/64 and 58/64) 29 constitutionalism 14 conceptual dissonance between European integration and constitutionalism 19, 357

374  Index de-constitutionalisation of internal market 13, 17, 319, 339 and democracy 12, 20, 338, 347, 352, 368 integrated interpretation framework 364–8 and free movement provisions 11, 13, 350 living constitutionalism and originalism 325–6 over-constitutionalisation critique 16–17, 319, 327, 353, 354 thick constitutional context 13, 339, 362 vacuity of European constitutionalism 19, 337 whether constitutional objection to minimum harmonisation 271–8 consumer protection Communication of 2015 on the DSM Strategy 239 Consumer Rights Directive (Directive 2011/83/EU) 281 Consumer Sales and Guarantees Directive (Directive 99/44/EEC) 268, 279n63, 280, 281 Food Information to Consumers Regulation (FIR), Regulation (EU) 1169/2011 296, 297 harmonisation 268, 269 platform disruption 219–21 Unfair Terms in Consumer Contracts Directive (Directive 93/13/EEC) 13, 239n36, 268, 269, 274–7, 280, 281 Consumer Rights Directive (Directive 2011/83/EU) 281 Consumer Sales and Guarantees Directive (Directive 99/44/EEC) 268, 279n63, 280, 281 contextualising the internal market and embedding 9–13 EU-specific contextual factors 78–9 global contextual factors 79–81 post-Lisbon 9 ‘road blockades’ case 78, 79 ‘strawberries’ case 78 convergence of the freedoms 51–61 case law Alpine Investments (Case C-384/93) 55–8 Bosman (Case C-415/93) 55 Carpenter (Case C-60/00) 59–61 Keck (Joined Cases C-267/91 and Case C-268/91) see below Moped Trailers (Case C-110/05) 58–9

defining ‘convergence’ 51 Gebhard (Case C-55/94) 51–2 and human rights, role of 59–61 Keck (Joined Cases C-267/91 and Case C-268/91) and free movement of workers 54 and freedom to provide services 53–4 searching limits to 51–3 market access as the new universal formula 58–9 copyright Communication of 2015 on the DSM Strategy 241 modernisation of framework 233 Corona-crisis see COVID-19 crisis Costa v ENEL (Case 6/64) 301, 302, 315 Court of Justice of the European Union (CJEU) 215, 237 activist, seen as 326–8 and citizenship 76 criticisms of 326–8 deference to choices made by national governments/courts 145 and development of free movement principles over time 25 ‘engine of liberalisation’ 6, 145 freedom to provide services 144, 145, 146 external services law 158 on fundamental freedoms and rights 337, 344 Grand Chamber judgments 145 and healthcare sector 140 and indistinctly applicable restrictions 38, 39 and integration 337 and internal market 313–21 interpretation of Treaty provisions 25 judges 291 Keon Lenaerts as President 293, 294 normative evaluation 17 online sales 220–1 originalism at 323–33 referral to 4 refinement of free movement provisions 25 Uber cases see Uber; Uber France (Case C-320/16); Uber Spain (Case C-434/15) see also European Court of Justice (ECJ) COVID-19 crisis 1–5 balance between common standards and national regulatory autonomy 6

Index  375 borders, keeping open 2, 4, 5 challenges posed by 109 containment measures 3–4 and digital transformation 188, 189 Report of the High Level Expert Group on the Impact of the Digital Transformation on EU Labour Markets 203 fundamental freedoms and rights following 363 kneejerk reactions to 4 and need for connectedness 1, 3 and Single Market 2.0 188 and von der Leyen Commission 202 Criminal proceedings against Bernard Keck and Daniel Mithouard see Keck (Joined Cases C-267/91 and Case C-268/91) crisis, times of Community Method, benefits over executive dominance 4 need for EU-wide solidarity 1, 5 see also COVID-19 crisis cross-border services provision 141–3 see also digital services; freedom to provide services; services Cybersecurity Act 233 Daiichi Sankyo (Case C-547/14) 89 Dano (Case C-333/13) 328 Dassonville (Case 8/74) 25, 58 and capital restrictions 170 direct and indirect hindrances 29–31 formula of trading rules capable of hindering intra-Community trade 65, 66, 67, 95, 99, 101 originalism at the Court 328, 329 special treatment for foreign products 29 data big data 226 cross-border access to 241 data-driven industrial policy 206 Directive on ‘Open Data’ 247 end of open data 205–7 Free Flow of Data Regulation (Regulation 2018/1807) 235 non-personal, free flow of see free flow of data (FFD) ownership of 248 sharing between public authorities 207–8 transfer of 198 data economy 248–9

Data Economy Communication 248–9 data protection General Data Protection Regulation (GDPR) 194, 213, 222, 246 platform disruption 222–3 data spaces 206 Single European Data Space 210 Davies, Gareth 6, 7–8, 11, 14, 16, 19n80, 147, 335n1, 359–60 De Agostini (Case C-34/95) 69n19 De Cecco, F 340–1 De Geus v Bosch (Case 13/61) 316–17 de Witte, Bruno 6, 15, 17 de Witte, Floris 347 de-constitutionalisation of internal market 13, 17, 319, 339 deferential approach, freedom to provide services 6 Deliège (Joined Cases C-51/96 and C-191/97) 53–4, 55, 61 democracy 21, 301, 305, 319, 340n24 complexity 108–9 and constitutionalism 12, 20, 338, 347, 352, 368 integrated interpretation framework 364–8 enhancing 348, 353 failure 350 inclusive 347 majoritarian 326 national 350 risks to 245 robust/well-functioning 340n24, 352, 366, 367 transnational 11, 13, 339, 347–52, 369 forum of transnational democratic decision-making 349 transnational effects 352 virtual 353 devolution, and national regulatory autonomy 7 dIDAS (EU framework for authentication for simple electronic signature) 210 Digital Agenda for Europe (2010) 233 digital ecosystem 192–3, 194, 198 and Commission 204 evolving 199–200 multilayered 212 Single Market 2.0 as a layered ecosystem 200–11 Digital Republic Act (2016), France 230

376  Index digital services defining 139 digital identify (eID) 210 E-commerce Directive (Directive 2000/31) 152 electronic intermediation services 215 electronic trust services (eTS) 210 Internal Market for Digital Services 80 Regulation 1128/2017 on cross-border portability of online services 139 trust and security in 245 see also internet; Single Market 2.0 Digital Single Market (DSM) 189, 201, 233–58 Artificial Intelligence (AI) 20 and citizens of the EU 209, 234 Commission work programmes 139 Communication of 2015 on the DSM Strategy 238–9 compared with four traditional freedoms 81 compared with previous initiatives 233–4 differences in regulating the online environment 236 e-commerce rules 231 embedding the digital dimension 20–1 evolution or revolution 234–5 formulated by Juncker Commission 188 future of 249–51 ‘GAFA’ specific legislation, need for 236 global contextual factors 79–80 Internet of Things (IoT) 20 level playing field at international level 237–8 and offline in the EU 236–7 and Lisbon Treaty 77 as a moving target 237, 249, 252 reasons for DSM strategy 234 SMEs 234, 236 see also internet; online platform economy; Single Market 2.0 direct discrimination 28, 31–2, 46, 96 Dassonville (Case 8/74) 29–31 Directive 85/577/EEC see Doorstep Selling Directive (Directive 85/577/EEC) Directive 89/552 (Television without Frontiers Directive) see Television without Frontiers Directive (Directive 89/552) Directive 93/13/EEC see Unfair Terms in Consumer Contracts Directive (Directive 93/13/EEC)

Directive 93/83/EEC (Satellite and Cable Directive) see Satellite and Cable Directive (Directive 2019/789/EU) Directive 96/71 concerning the posting of workers see Posted Workers’ Directive (Directive 96/71) Directive 97/81/EC see Part-Time Work Directive (Directive 97/81/EC) Directive 99/44/EC on sale of consumer goods 268 Directive 1990/70/EC see Fixed-Term Work Directive (Directive 1990/70/EC) Directive 2000/31 see E-commerce Directive (Directive 2000/31) Directive 2002/65/EC see Distant Selling Directive (Directive 2002/65/EC) Directive 2003/88/EC see Working Time Directive (WTD) (Directive 2003/88/EC) Directive 2004/38/EC see Citizens’ Rights Directive (Directive 2004/38) Directive 2005/29/EC see Unfair Commercial Practices Directive (UCPD) (Directive 2005/29/EC) Directive 2005/60/EC see Money Laundering Directive (Directive 2005/60) Directive 2006/123/EC see Services Directive (Directive 2006/123) Directive 2007/65/EC see Audiovisual Media Services Directive (AVMSD), (Directive 2007/65/EC) Directive 2011/83/EU see Consumer Rights Directive (Directive 2011/83/EU) Directive 2015/2302/EU see Package Travel Directive (Directive 2015/2302) Directive 2017/1128/EU see Geo-blocking Regulation (Regulation 2017/1128) Directive 2018/957/EU see Posted Workers’ Directive (revised) (Directive 2018/957) Directive 2019/1024/EU see Public Sector Information Directive (PSI), (Directive 2019/1024/EU) discrimination 29–43 ambiguity of concept 28–9 arbitrary, prohibition of 27 case law Cassis de Dijon (Case 120/78) 33–6 Dassonville (Case 8/74) 29–31 Lawyers’ fees (Case C-565/08) 40–3 Third-party motor liability insurance (Case C-518/06) 40 van Binsbergen (Case 33/74) 31–2

Index  377 van Wesemael (Joined Cases 110/78 and 111/78) 32 Walloon Waste (Case C-2/90) 37–40 development of concept 26, 29–43 direct and indirect 28, 29–31, 46, 96 early concentration on in the development of free movement principles 61 finding of 26 free movement of capital 172 free movement of persons 121 and Keck (Joined Cases C-267/91 and Case C-268/91) 50 limits of concept and ‘Sunday trading’ case law 44–7 market access 100 nationality grounds 27, 28, 121, 172, 341 and non-discrimination as mutually exclusive 48 prohibition as negative flipside of equality principle 39–40 unequal treatment 27, 28, 39, 40–3 workers, early case law 29 see also free movement of goods; free movement provisions; freedom to provide services; non-discrimination Distant Selling Directive (Directive 2002/65/EC) 279 distinctly applicable restrictive measures (special treatment) distinguished from indistinctly applicable restrictions 27–8, 33 inequality intentionally created 34 see also discrimination; free movement of goods; indistinctly applicable restrictions Distributed Ledger Technologies (DLTs) 210, 212 diversity consequences of establishment of the internal market for management of 285, 304–5 improving 6 managed diversity of internal market 5, 17 versus uniformity 5–9 over-emphasis on uniformity 17 preference for uniformity 7–8 DocMorris (Case C-322/01) 275, 276, 277 Doorstep Selling Directive (Directive 85/577/EEC) 274 Drei Glocken (Case 407/85) 305 DSM see Digital Single Market (DSM) duty to act and free movement of goods 104–8

E-Commerce Directive (Directive 2000/31/EC) 152, 216, 217, 231, 243, 250 economic protectionism 162 edge/cloud model, internet 197, 198, 205, 207 EEC Treaty see Rome Treaty (EEC Treaty) Egenberger (Case C-414/16), horizontal applicability between private parties to EU Charter 60n110, 62 Electronic Communications Code (EECC) 233, 237, 242 electronic intermediation services 215, 216 electronic trust services (eTS) 210 Elite Taxi v Uber Systems Spain see Uber Spain (Case C-434/15) embedding and contextualising 9–13 digital dimension 20–1 multidirectional process 13 of values in constitutional framework of the EU 12 Enchelmaier, Stefan 15 ENISA (European Union Agency for Cybersecurity) 245 e-Privacy Directive (Directive 2002/58/EC) 246 EPSR see European Pillar of Social Rights (EPSR) equality principle equality before the Treaties obligation 7 equals treated equally 46 factual inequality 33–4 inequality intentionally created in special treatment 34 listing all freedoms equally under Article 26(2) TFEF 52 prohibition of discrimination as negative flipside of 39–40 Torfaen Borough Council (Case C-145/88) 46 equivalence principle fundamental freedoms and rights 358–9 legal equivalence for Treaty freedoms and fundamental rights 12 measures having equivalent effect, abstaining from 106 and quantitative restrictions 65 selling arrangements and Keck 49 Ericsson 212 Essent (Joined Cases C-105/12 to C-107/12) 178, 179, 182

378  Index establishment of the internal market, new paradigm for 285–311 adjudication Åklagarenv Hans Åkerberg Fransson (Case C-617/10) 303 Confédération paysanne and others (Case C-528/16) 302–3 Wightman and Others (Case C-621/18) 299–302 consequences for diversity management 285, 304–5 consequences for EU law 305–10 conferral principle 308–9 increasing use of findings from behaviour science 310 proportionality principle 307–8 solidarity principle 309–10 subsidiarity principle 306–7 and harmonisation 285, 286, 294, 295 as an objective in itself 84 requirement to establish internal market 286–94 judges 291–4 legislator 290–1 preliminary conclusions 294 risk regulation 298–9 secondary legislation 295–9 adjudication 299–303 gambling 15, 144–5 preliminary conclusions 303–4 unfair commercial practices law 298–9 supranationality versus managed diversity 5, 285, 295–304 unfair commercial practices law 295–8 Estonia, X-Road ecosystem 198 EU Charter of Fundamental Rights see Charter of Fundamental Rights, EU EU internal market law consequences of the establishment of the internal market for 305–10 diversity management 3 legitimacy dependent on sound legal interpretation 19 and non-market interests 18 purposive or teleological interpretation method 85 static nature 328–30 ‘working well,’ determining whether 6, 16 see also European Union (EU) Europe as a Platform (EaaP) 188 European Agenda for the Collaborative Economy (2016) 214

European Banking Authority 150 European Commission burden of proof for presence of a restriction 49 Communication of 2007 on ‘The Single Market for 21st Century Europe’ 80 Communication of 2015 on the DSM Strategy see Communication on the DSM Strategy (2015) Communication of 2016 on the Collaborative Economy 215–16, 220 Communication of 2018 on Data Economy 248 Communication of 2018 on Subsidiarity and Proportionality 307 Communication of 2018 on ‘The Single Market in a Changing World’ 113, 122, 133 Consumer Policy Programme (2002–06) 279 and digital ecosystem 204 Directorates General 139 draft Directive on Services in the Internal Market 142 European Agenda for the Collaborative Economy (2016) 214 Green Paper on Review of the Consumer Acquis (2007) 279 indistinctly applicable restrictions and Walloon Waste 38–40 information society services, obligation of Member States to notify regarding 152 and internet 195 ‘new’ 201 policy-making, recommended third way 21 Recommendation on the EPSR 126 Reflection Paper of 2017 on the Social Dimension of Europe 126 and Single Market 2.0 202, 203 supervisory role 4 ‘Trustworthy Artificial Intelligence’ 210 White Paper on Completing the Internal Market 140–1 White Paper on the Future of Europe 126 Work Programme of the European Commission 139, 211 European Court of Justice (ECJ) ‘1992’ legislative programme 143, 144 and gender equality 287–8

Index  379 jurisprudence 318 see also Court of Justice of the European Union (CJEU) European Digital Economy 238 European Economic Constitution 179 European Electronic Code 233 European Free Trade Area (EFTA) 209 European High-Performance Computing Joint Undertaking 249 European Labour Authority, creation of 129 European Parliament 158 European Pillar of Social Rights (EPSR) 112, 113, 123, 125, 127, 130–2 20 principles 128 and Lisbon Treaty 134 work–life balance 130 European Union (EU) embedding of values in constitutional framework 12 ‘EU Method’ of executive dominance 4 free movement principles as a hallmark 3 interdependence and interconnectedness 1, 3 international trade in services, seeking to facilitate 159 and internet 194–6 legislative powers 263 market integration process 187 membership of the WTO 73 positive competence to adopt harmonising legislation 18 privatised undertakings, use of golden shares in 163 social deficit 183 as a social market economy 12 withdrawal of the UK from 1, 109, 112, 113, 119, 122, 135, 363 fundamental freedoms and rights following 363 referendum result of June 2016 120–1 and Single Market 2.0 187–8 see also EU internal market law external services law, freedom to provide services in 158–60 free trade agreements 158, 159 international agreements dealing with specific services 159–60 and rule of law 158–9 Familiapress v Heinrich Bauer Verlag (Case C-368/95) 48n73

Federal Republic of Germany v European Parliament and Council of the European Union see Tobacco Advertisment (Case C-376/98); Tobacco Advertisment (Case C-380/03) File Transfer Protocol 189–90 5G connectivity 212 Fixed-Term Work Directive (Directive 1990/70/EC) 224 Food Information to Consumers Regulation (FIR), Regulation (EU) 1169/2011 296, 297, 298 four freedoms (capital, goods, persons and services) 19, 20, 177, 187, 336 development of free movement principles over time 25, 29, 59, 60, 62 internal market dynamics 75, 76, 78 see also free movement of capital; free movement of goods; free movement of persons; free movement provisions; freedom to provide services France, Digital Republic Act (2016) 230 see also Commission v France (riots) (Case C-265/95); Uber France (Case C-320/16) free flow of data (FFD) 20, 188, 231, 247 Free Flow of Data Regulation (FFD) (Regulation 2018/1807) 235, 247, 249 free movement of capital capital restrictions 6, 170–4, 183 clash between economic and social objectives 12 classical liberalism 166 coordinated market economies 165, 171 corporate governance 171–2 derogation from ordinary company law 164, 172–3 and freedom of establishment 181 further purpose doctrine 181 and golden shares 163–6 and human rights 340 liberal market economies 165, 171 liberalisation 161, 163 Maastricht Treaty rewording 76 national regulatory autonomy, space for 6 neutrality principle 176–82 and privatisations 12, 176–82 procedural proportionality 6, 174–6 proportionality principle 165, 171

380  Index and protection of social objectives of the EU 161–84 public interest objectives 182–4 social democratic norms 167 varieties of capitalism 165 free movement of goods 95–109 advertising 99, 102 boundaries 95 clarity in recent case law, determining 15–16 containment measures due to COVID-19 3 duty to act and good administration 104–8 early case law 25–6 enacting quantitative restrictions, duty to abstain from 106 EU-specific contextual factors 78 frontiers of 104–8 and indistinctly applicable restrictions 33 between Keck and market access 97–104 measures having equivalent effect, abstaining from 106 and obstacle to trade 45–6, 66, 89 potential conflict between national rights and free movement provisions 105 provisions framed as constitutional fundamental rights 107 public health requirements 102 reach of Art 35 TFEU 102–4 Regulation 2679/98 (Obstacles to the Free Movement of Goods) 105 removal of physical barriers to trade 141 restrictions, notion of 16, 26 trade and scope of Article 34 TFEU 101 and universal bans 46–7, 50, 62 ‘use’ of a product 70, 71, 101, 102 White Paper on Completing the Internal Market relevant to 141 see also discrimination; distinctly applicable restrictive measures (special treatment); indistinctly applicable restrictions; sale of goods free movement of persons and citizens of the EU 3n5, 4, 21, 76, 119, 122, 130, 133, 195, 198, 209, 242, 251 Digital Single Market (DSM) 209, 234 Directive 2004/38 (Citizens’ Rights Directive) 76, 115, 116 diversity management 304 economically active 116, 118 economically autonomous 116 EECC Code 242 empowerment of 202

general or citizen-based rights and economic freedoms 114–15, 117–18 law 328–9 non-economically active 3n9, 76 prior lawful residence for family members 115 status as 52, 76, 119 see also free movement of persons containment measures due to COVID-19 3 defining scope of ‘persons’ for internal market law purposes 112 effectiveness of EU social policy 112–13 in EU 27 119 and EU internal market 112, 113–19 European Pillar of Social Rights (EPSR) see European Pillar of Social Rights (EPSR) and free movement of capital 167, 234 and free movement of goods 52n84 fused concepts and interpretation principles 117–18 fused legal foundations 115, 118 fused legislative frameworks 116, 118 fused nature of economic and general free movement rights 112 ‘genuine link’ analysis 117, 118 hybrid concepts as examples of economic and general free movement rights 117–18 indexation of family benefits 112, 119–22 migrant workers 116 mobility rights (to move and reside) 60, 114–18 non-discrimination on nationality grounds requirement 121, 274, 341 and political freedom 119–23 privileging of activity over autonomy 116–17, 119 and social freedom 123–34 Treaty definitions 115 ‘use’ of a product, non-discriminatory restrictions on 65–6 workers and self-employed persons 116, 129 free movement of workers 116, 129 development of free movement provisions 25, 26 direct and indirect discrimination 29 early secondary law 25, 61 Free Movement of Workers Regulation (Regulation 492/2011) 121

Index  381 and Keck (Joined Cases C-267/91 and Case C-268/91) 54 see also free movement of persons free movement provisions advertising 99, 102 all EU citizens, rights of 114 certain selling arrangements 47–50 and constitutionalism 11, 13, 350 convergence of the freedoms 26, 51–61 COVID-19 distancing response 3 development over time 25–63 different wordings 52 discrimination see discrimination; unequal treatment distinctly applicable restrictive measures (special treatment) 26–7 early concentration on discrimination 61 four freedoms see four freedoms as fundamental freedoms and rights 107, 340, 342 freedom of establishment 336n8, 338, 342, 361, 367 whether free movement provisions are 339–45 whether free movement provisions should be 345–54 grafting social protection onto 124 hallmark of the EU 3 and human rights, role in unification 59–61 indistinctly applicable restrictions concept 37–40 distinguished from special treatment 26–7 justification 33–6 Weigel (Case C-387/01) 28 justification of grounds beyond the wording of the TFEU 31–2 justification of indistinctly applicable restrictions 33–6 versus national regulatory competence 8 no de minimis threshold applying to 51 normative justifications for constitutionalisation 11 over-constitutionalisation critique 16–17, 353, 354 prima facie restriction 357–8 refinement by the CJEU 25 social objectives 112 special treatment 26–7 structure of free movement law 111 unequal treatment 27, 40–3

wording of TFEU and early secondary law 27–8 see also discrimination; free movement of capital; free movement of goods; free movement of workers; freedom to provide services freedom of establishment 18, 44 and free movement of capital 181 and fundamental rights 335, 336n8, 338, 342, 361, 367 and horizontal regulation 151 negative 57–8 unjustified restriction on 34–5 violation of in Costa v ENEL 315 freedom to provide services 137–60 audiovisual services 150 banking 138–9 Central and Eastern Europe 154 compatibility with TFEU 31 Conclusions of the Lisbon European Council (2000) 142 containment measures due to COVID-19 3 controversial nature 137–60 cross-border provision 141–3 deferential approach 6 Deliège (Joined Cases C-51/96 and C-191/97) 53–4 deregulation and re-regulation 160 and deterrence criterion 60 digital services see digital services Directive see Services Directive (Directive 2006/123) distributed single market for services 207–9 ‘double burdens,’ unnecessary 143–4 early case law 25–6 EU services law as a fragmented legal field 137–40 in external relations law 158–60 fragmented legal field, EU services law as 137–40 fundamental freedom 143–5 ‘general good,’ rules justified by 31–2 General Services Directive 146–8 health law 138 horizontal regulation 138, 141, 147, 151–4 and Keck (Joined Cases C-267/91 and Case C-268/91) 53–4 liberalisation promising to bring economic benefits, specific sectors 148, 155 mandatory requirements test 144 public procurement regime 151–2 sectoral regulation 148–51

382  Index service sectors where national regulation intrusive 148 services of general interest 154–8 legal regime 138, 140 Protocol No 26 added to the Treaties 157 public utilities 154–5 welfare and social services 155–7 specific legislation 138 State control combined with substantive harmonisation 149–50 state of EU services law at the end of 2009 140–3 van Binsbergen (Case 33/74) 31 see also Services Directive (Directive 2006/123) functionalism 346 fundamental freedoms and rights 21, 335–69 differences between 341–5 free movement provisions as 107, 340, 342 freedom of establishment 336n8, 338, 342, 361, 367 whether free movement provisions are fundamental freedoms 339–45 whether free movement provisions should be fundamental freedoms 345–54 hierarchy or ‘balance’ between fundamental freedoms and other fundamental rights 355–61 interpretation of different parts of EU legal order 362–3 laissez-faire approach to policy-making 21 legal equivalence for Treaty freedoms and fundamental rights 12 need for integrated theory 12, 362–3 integrated democratic interpretation framework for all EU fundamental rights 362–8 integrated interpretation framework rooted in democratic institutionalism 364–8 normative basis for 348–9 obstacles test 10 protection against collective dismissal 366 public interest justifications 359–60, 361 and Rule of Law 12 scope of application 342 strike, right to 366 top-down controlled approach to policy-making 21 workers’ rights 366 see also Charter of Fundamental Rights, EU; human rights

Fundamental Rights Agency, EU further purpose doctrine 181

352

G2B see Government-to-Business (G2B) relationships GAIA-X project 204 White Paper 205 gambling and gaming 15, 144–5 procedural proportionality 174–5 Garben, Sacha 8, 10–11, 12, 17, 133, 134 Gaston Schul (Case 15/81) 313 Gebhard (Case C-55/94), convergence of the freedoms 51–2 General Agreement on Tariffs and Trade (GATT) 72 General Agreement on Trade in Services (GATS) 158, 159 General Data Protection Regulation (GDPR) 194, 213, 222, 246 ‘general good,’ freedom to provide services rules justified by 31–2, 34 genetically modified organisms (GMOs) 302 geo-blocking 251 Communication of 2015 on the DSM Strategy 239–40 Geo-blocking Regulation (Regulation 2017/1128) 240 German insurances (Case 205/84), and convergence of the freedoms 57–8 Germany draft Digitalisation Bill 204 harmonisation measures 275 indirect discrimination 29 Gerner-Beuerle, Carsten 169 gig-workers 223–5 Gilli and Andres (Case 788/79), indistinctly applicable restrictions 36n36 golden shares capital restrictions 6, 163, 164, 170–4 clash between economic and social objectives 12 and company law 164, 172–3 and free movement of capital 163–6 growth of use in Europe 164 procedural proportionality 6 and protectionism 162, 164 use of in privatised EU undertakings 163–6 González Sánchez (María Victoria) v Medicina Asturiana (Case C-183/00) 275

Index  383 good administration and free movement of goods 104–8 goods see free movement of goods; sale of goods Google 195 Govaere, Inge 9, 10, 14 Government-to-Business (G2B) relationships 207, 211 ‘Green Deal data space’ 206 Grimm, Dieter 16–17, 317–18, 319–20, 321 Groussot, X 8n35, 351, 361n122 Grzelczyk (Case C-184/99) 328 harmonisation 261–84 diversity management 304 and establishment of the internal market 285, 286, 294, 295 EU’s positive competence to adopt harmonising legislation 18 free choice between models 18, 271 in the internal market 264–5 legislative, practice of 265–71 maximum 9, 261, 266–8, 283, 290 minimum see minimum harmonisation partial 275 policy and priorities 282–4 by stealth 10 terminology and description 265–6 tobacco-related measures 9–10, 272, 273 Treaty superstructure 261–4 unfair commercial practices law 296–7 welfare services 157 Hatzopoulos, Vassilis 20–1 Hauspie, Valérie 22 hazard analysis and critical control points (HACCP) 298 healthcare 3–4, 140, 155 public health requirements 102 High Value Data sets 247–8 historical context of the internal market 313–21 choice for Europe 331–2 European Court of Justice (ECJ) 313–21 criticisms of 326–8 originalism at 323–33 static nature of EU law 328–30 Hoffman, D 340n23 horizontal regulation and Art 63 TFEU 166–70 Charter of Fundamental Rights, horizontal applicability between private parties to 60, 62

freedom to provide services 138, 141, 147, 151–4 post-Lisbon case law on Art 3(3) TEU 88 human rights centrality in EU legal order 26 and dignity 340 and free movement of capital 340 and unification of the freedoms 59–61 see also fundamental freedoms and rights Hünermund (Case C-292/92) 100 Attorney General Opinion 46n65, 67, 68, 99 IHAN project 198, 207 indexation of family benefits 112, 119–22 indirect discrimination 28, 46, 96 broad interpretation 100 Dassonville (Case 8/74) 29–31 indistinctly applicable restrictions 36 indistinctly applicable restrictions additional costs for foreign providers 41 characteristics of products, rules relating to 65 distinguished from special treatment (distinctly applicable measures) 27–8, 33 factual inequality 33–4 as independent of nationality 28, 29 justification of and Cassis de Dijon 33–6 as national treatment 29 proportionality principle 34 reliability of contractual relationships 34 taking market in another Member State as found 41 Walloon Waste (Case C-2/90) 37–40 Weigel (Case C-387/01) 28 see also discrimination; distinctly applicable restrictive measures (special treatment); free movement of goods industrial action, right to 366 Infant Formula Milk (C-391/92), scope of Article 34 TFEU 68–9 integration, European conceptual dissonance between European integration and constitutionalism 19, 337 integrated democratic interpretation framework for all EU fundamental rights 335–69 and internal market 75 market integration process 187, 348 and minimum harmonisation 269–70

384  Index need for integrated theory 362–3 integrated democratic interpretation framework for all EU fundamental rights 362–8 integrated interpretation framework rooted in democratic institutionalism 364–8 negative 11 interconnectedness and the COVID crisis 1, 3 Inter-Institutional Agreement 233 internal market as ‘area without internal frontiers’ 26, 46 avoidance of a free market interpretation 12 broadened interpretation of legal basis 10 change, clarity and continuity in EU law 14–20 and common market 75 competition law 288 ‘conceptual poverty’ 11 contextualising and embedding 9–13 de-constitutionalisation 13, 17, 319, 339 defined in the Treaties 75, 76 demarcation principle 304–5 digital dimension, embedding 20–1 ‘e-Internal Market’ 80 EU internal market disrupted by platforms 215–25 and European integration 75 exceptionalism 13 external dynamics 77, 78–81 four freedoms see four freedoms harmonisation see harmonisation internal dynamics 77, 81–91 ‘Internal Market 2.0’ 20, 21, 367 conference 52 managed diversity of 5 misunderstood concept 75–91 modernisation of concept 81 moving targets and shifting contextual factors 78–81 objective 9, 83 proactive and purposeful instead of the former-reactive and instrumental 81–5, 86 and public interest 10 uniformity versus diversity 5–9 International Monetary Fund (IMF) 162 International Transport Workers’ Federation and Finnish Seamen’s Union v Viking Line see Viking (Case C-438/05)

internet conventional cloud models 197–8 digital ecosystem 192–3, 194, 198, 199–200 Single Market 2.0 as a layered ecosystem 200–11 edge/cloud model 197, 198, 205, 207 end-to-end nature 191, 192, 194 EU Internet Forum 243 evolving ecosystem 199–200 layered architecture 189–90, 198 platformisation of 193 waves of evolution 189–200 first transformation wave 190–6 second transformation wave 196–9 see also digital services; Digital Single Market (DSM); platformisation of the economy; servitisation of the economy; virtualisation of the economy Internet of Things (IoT) 20, 187, 196–7, 199, 205 Internet Protocol 189 ISA2 (Programme on Interoperability solutions for public administrations, businesses and citizens) programme 195, 209 Italian Trailers (Case C-110/95) 71, 74 Italy, Competition Authority 230 ius imperii 169 Jana Pereničová, Vladislav Perenič v SOS financ spol, s.r.o. (Case C-453/10) 275 Jean Thieffry v Conseil De L’Ordre Des Avocats A La Cour De Paris see Thieffry (Case 71/76) Juncker Commission 188, 195, 237, 238 Jyske Bank (Case C-212/11) 276, 277 Kadi (Cases C-402/05P and 415/05P) 355 Kamenova (Case C-105/17) 220n28 Karner (Herbert) (Case C-71/02) 276 Keck (Joined Cases C-267/91 and Case C-268/91) and advertising 99 controversies 26 and convergence of the freedoms 26, 54 and discrimination test 50, 61 and EEC Treaty 67–8 facts of case 67–8 and free movement of workers 54 and freedom to provide services 53–4

Index  385 and fundamental freedoms and rights 346 and market access 58, 97–104 neutrality requirement and non-discrimination requirement 69 searching limits to 51–3 and selling arrangements 26, 47–50, 96 capital restrictions 170 direct and indirect discrimination 99 and free movement of capital 166 and Infant Formula Milk (Case C-391/92) 68, 69 ‘Keck objection’ 57, 61 relevant traders and affect in the same manner 49 and static nature of EU law 328 and ‘Sunday Trading’ cases 26 system of presumptions created by 99 uncertainty 15 and unpredictability 70 use restrictions 70 Kingreen, T 340n30, 343, 344 labour law 153, 223–5 Laval un Partneri (Case C-341/05) free movement of goods and duty to act 105 fundamental freedoms and rights 358, 367 originalism at the Court 329 and Posted Workers’ Directive 153 structure of free movement law 111 Lawyers’ fees (Case C-565/08), discrimination case 40–3 legal certainty, and procedural proportionality 174–6 legislative harmonisation see harmonisation Lehtonen (Case C-176/96), convergence of the freedoms 54, 55, 61 Lenaerts, Koen 293, 294 Lessig, Lawrence 189–90 Lisbon Treaty (2007) see Treaty of Lisbon (2007) living constitutionalism, and originalism 325–6 Maastricht Treaty (1993) 76, 268 managed diversity of internal market 5 mandatory requirements test Cassis de Dijon (Case 120/78) 34, 36, 43, 45 freedom to provide services 144 margin of appreciation 175, 357n103, 358, 359

market access change of emphasis from classification of a measure to restricting access 71 collaborative economy 216 discrimination 100 facilitated by Article 34 TFEU 73–4 as free movement between Member States 15 free movement of goods and Keck 97–104 lawyers deprived of opportunity of gaining access to market of host Member State 41–2 as new universal formula in Moped Trailers (Case C-110/05) 15, 26, 58–9, 62, 96 platform disruption 216–19 market liberalisation, versus protection of non-market values 17 Martinez-Sala (Case C-85/96) 328 Mattera, Alfonso 22, 251 maximalist shield interpretation, free movement of capital 177 maximum harmonisation 9, 261, 266–8, 283, 290 Mayras, AG 340n25 Member States burden of proof for justification of a restriction 49 closure of borders in COVID crisis 2 and constitutionalism 338 court proceedings and sale of goods 49 equality before the Treaties 7 and indistinctly applicable restrictions 33 market access as free movement between 15 national competence division, proportionality assessment 2 not absolved from observing obligations if others fail to abide by 35 obligation to notify Commission of new rules affecting information society services 152 recognition of evidence of professional qualifications from other Member States 35 restriction of autonomy 318 rights in strategically sensitive private companies see golden shares see also national regulatory autonomy Metock and Others (Case C-127/08) 115 Mickelsson and Roos (Case C-142/05) 70, 71, 101 migrant workers, rights of children 116

386  Index minimum harmonisation 261, 266, 268–71 benefits claimed for 270, 283, 284 case law AGET (Case C-201/15) 8 Alemo-Herron (Case C-426/11) 8 BAT (Case C-491/01) 273 Biuro podróży ‘Partner’ (Case C-119/15) 274–5 Buet (Case 382/87) 274 DocMorris (Case C-322/01) 275, 276, 277 González Sánchez (María Victoria) v Medicina Asturiana (Case C-183/00) 275 Jana Pereničová, Vladislav Perenič v SOS financ spol, s.r.o. (Case C-453/10) 275 Jyske Bank (Case C-212/11) 276, 277 Karner (Herbert) (Case C-71/02) 276 Philip Morris (Case C-547/14) 273, 278 Rina Services (Case C-593/13) 276, 277 Safe Interenvios (Case C-235/14) 276 Swedish Match (Case C-151/17) 278 Tobacco Advertisment (Case C-376/98) see Tobacco Advertisment (Case C-376/98); Tobacco Advertisment (Case C-380/03) character of 280 collective redundancies 8 constitutional objection to, whether 271–8 evaluation of concept 282–3 and integration 269–70 narrow version 270 policy objection to, whether 278–82 unaccommodating approach to 8–9 see also harmonisation Ministère public v Gauchard (Case 20/87), and ‘Sunday Trading’ cases 44n63 Ministerial Declaration on e-Government, Tallinn (2017) 209 Money Laundering Directive (Directive 2005/60) 276 Monti Report (2010) 156 Moped Trailers (Case C-110/05) 26 market access as new universal formula in 15, 26, 58–9, 62, 96 and use of product 101 most favoured nation (MFN) principle 72 Muir, Elise 118n31, 123n49, 132, 343–4 Multi-Annual Financial Framework (MFF) 251 multilingualism 87 multisided platforms 193

Museum admissions (Case C-388/01), and indistinctly applicable restrictions 39n50 mutagenesis 302 mutual recognition principle and Article 34 TFEU 72 Cassis de Dijon (Case 120/78) 34, 35–6, 95 MyData movement 198 national regulatory autonomy adoption of regulatory frameworks by national authorities 107 and devolution 7 free movement of capital 6 versus free movement paradigm 8 ‘Sunday trading’ cases seen as undue interference with 99 and unequal treatment 42–3 see also Member States nationalism, return to 75 nationality discrimination on grounds of 27, 28, 121, 172, 341 indistinctly applicable restrictions as independent of 28, 29 supranationality versus managed diversity 5, 295–304 negative integration 11 neutrality principle free movement of capital 176–82 and non-discrimination 69 sword interpretation 179 New Valmar BVBA (Case C-15/15), free movement of goods 103 Nic Shuibhne, Niamh 13, 17, 56n94, 356, 360 NIS Directive (network and information systems) 245 Nokia 212 non-discrimination 51n80, 69, 217, 232 data protection 343n41 and discrimination, as mutually exclusive 48 as a fundamental right 347, 365 nationality grounds 121, 274, 341 payment methods 240 public procurement regime 314 use of products 65–6 see also discrimination ‘notice-and-action’ procedures, online platforms 243 Nutrition and Health Claims Regulation (Regulation (EC) 1924/2006) 297

Index  387 Obstacles to the Free Movement of Goods Regulation (Regulation 2679/98) 105 see also free movement of goods; sale of goods oil crisis (1970s) 75 Omega (Case C-36/02) 357, 358 online platform economy defining 213–14 disruption of the internal market 20–1 online platforms Communication of 2015 on the DSM Strategy 243–4 consumer protection 219–21 data protection 222–3 EU internal market disrupted by platforms 20–1, 215–25 regulation requirement for platforms 225–32 EU internal market in need of regulation see below impact on public policy 193 and internal market 213–32 labour law 223–5 legal qualifications of the parties 215–16 market access 216–19 multiple platforms 228 multisided platforms 193 ‘notice-and-action’ procedures 243 obligations on video-sharing platforms 236–7 platform services 137 platform-to-business (P2B) practices 195, 204 proposals for a ‘Platform Directive’ 221, 231 regulation requirement for EU internal market 20–1, 225–32 external regulation, need for 228–32 peer regulation and reputation rating 227–8 sectoral approach to 250 self-regulation 225 examples 226 limits of 226–8 see also Digital Single Market (DSM); online platform economy; platformisation of the economy; portability of online content services open API (open Application Programming Interface) 188, 207, 208 ordoliberalism 345

Organisation for Economic Co-operation and Development (OECD) 162, 195, 230 originalism at the ECJ 323–33 and living constitutionalism 325–6 rule of the dead 325–6 OTT (‘over-the-top’) services 237, 242, 246 P2P (Peer-to-Peer) relationships 219 Package Travel Directive (Directive 2015/2302) 215 pacta sunt servanda principle 34 Part-Time Work Directive (Directive 97/81/EC) 224 Peer-to-Peer (P2P) relationships 219 personally identifiable information (PII) 198 persons, free movement of see free movement of persons in EU 27 Philip Morris (Case C-547/14) and harmonisation 273, 278 importance of Article 3(3) TEU 89, 90 platform economy defining 213–14 scope 214 see also online platform economy platformisation of the economy 20, 187, 193, 194 political freedom, and free movement of persons 119–23 polycentric networks in internal market 305–6 portability of online content services 139, 204n37, 205, 233, 235, 237, 240, 247 Posted Workers’ Directive (Directive 96/71) 141 revising 153 Posted Workers’ Directive (revised) (Directive 2018/957) 154 precautionary principle 3 privatisations and free movement of capital 12, 176–82 neutrality principle 176 post-privatisation phase 176 pre-privatisation phase 176 maximalist shield interpretation 177 prohibition of, in energy distribution system operators 178 reductionist shield interpretation 178, 179–80 sword interpretation 179–80

388  Index procedural proportionality and golden shares 6 gambling 174–5 good governance model 175, 176 judicial review 175–6 legal certainty 174–6 Product Liability Directive 208 products foreign, special treatment for 29 indistinctly applicable restrictions 65 and selling arrangements see sale of goods ‘use’ of 70, 71, 101, 102 non-discriminatory restrictions on 65–6 Programme on Interoperability solutions for public administrations, businesses and citizens (ISA2) programme 195, 209 proportionality principle 5 Cassis de Dijon (Case 120/78) 45 consequences of establishment of the internal market for 307–8 and COVID-19 crisis 4 deferential test 358 free movement of capital 165, 171 indistinctly applicable restrictions 34 procedural proportionality 6, 174–6 Services Directive (Directive 2006/123) 147 strict test 97 protectionism, return to 75 PSI Directive (Public Sector Information Directive) 247–8 public administrations 208 public health requirements, free movement of goods 102 public ownership, incompatibility with internal market 6 Public Sector Information Directive (PSI), (Directive 2019/1024/EU) 247–8 public utilities 154–5, 156 Purnhagen, Kai 5, 17 Quinn v Leathem (1901), convergence of the freedoms 53n85 R v Secretary of State ex parte BAT see BAT (Case C-491/01) reductionist shield interpretation, free movement of capital 178, 179–80 Regulation 2679/98 see Obstacles to the Free Movement of Goods Regulation (Regulation 2679/98)

Regulation 883/2004 see Social Security Co-ordination Regulation (Regulation 883/2004) Regulation 1924/2006 (Nutrition and Health Claims) see Nutrition and Health Claims Regulation (Regulation (EC) 1924/2006) Regulation 1223/2009 267 Regulation 492/2011 (Free Movement of Workers) 121 Regulation 1169/2011 (Food Information to Consumers) see Food Information to Consumers Regulation (FIR), Regulation (EU) 1169/2011 Regulation 1128/2017 (Cross-border Portability of Online Services) 139 Regulation 2017/1128 (Geo-blocking Regulation) 240 Regulation 2018/1807 (Free Flow of Data Regulation) see Free Flow of Data Regulation (Regulation 2018/1807) Regulation 2018/848 see Food Information to Consumers Regulation (FIR), Regulation (EU) 1169/2011 Regulation of Organic Production and Labelling of Organic Products (ROPL) (Reg 2018/848/EU) 297 Reich, Norbert 263 Reinhard Gebhard v Consiglio dell’Ordine degli Avvocati e Procuratori di Milano see Gebhard (Case C-55/94) Renda, Andrea 20 Report of the High Level Expert Group on the Impact of the Digital Transformation on EU Labour Markets 203 reputation aggregators 228 reputation rating, online platforms 227–8 reusable packaging (Case C-463/01) 106 Review of the Public Sector Information Directive (PSI Directive) 247 Rewe-Zentral AG see Cassis de Dijon (Case 120/78) Rickford, Jonathan 168 Rina Services (Case C-593/13) 276, 277 Ringe, Wolf-Georg 168 risk regulation 298–9 Robin-Olivier, S 113n3, 134 Roche-Laguna, Irene 21 Rome Treaty (EEC Treaty) on cartels 320

Index  389 and Commission v Czech Republic (Case C-525/14) 71–3 enforcement system 315 misinterpretation by the ECJ 319–20 wording 82, 313 see also Common Market Law Review, ‘In Search of the Limits of Article 30 of the EEC Treaty’ Ruiz Zambrano (Case C-34/09) 330 and right of citizenship 60 rule of law coherence, knowability and predictability 19, 337 and constitutionalism 338 deterioration in some Member States 188 external services law 158–9 and fundamental rights 12 Safe Interenvios (Case C-235/14) 276 Saint Prix (Case C-507/12), social freedom, and free movement of  persons 130–1 sale of goods ban on internet sales 104 certain selling arrangements and Keck 26, 47–50, 96 and advertising 100 direct and indirect discrimination 99 and Infant Formula Milk (Case C-391/92) 68, 69 ‘Keck objection’ 57, 61 Consumer Rights Directive (Directive 2011/83/EU) 281 Consumer Sales and Guarantees Directive (Directive 99/44/EEC) 268, 279n63, 280, 281 digital content 238–9 doorstep selling 274 equivalent effect and selling arrangements 49 long-distance 100, 102 product requirements and selling arrangements 47, 48, 49, 61, 98 distinction between 99 and TFEU Article 34 102 selling arrangements and Keck 26, 47–50, 96 capital restrictions 170 direct and indirect discrimination 99 and free movement of capital 166 and Infant Formula Milk (Case C-391/92) 68, 69

‘Keck objection’ 57, 61 and product requirements see above relevant traders and affect in the same manner 49 Unfair Terms in Consumer Contracts Directive (Directive 93/13/EEC) 13, 239n36, 268, 269, 274–7, 280, 281 see also free movement of goods; Keck (Joined Cases C-267/91 and Case C-268/91) Sánchez-Bordona, Campos 181, 300, 306 Satellite and Cable Directive (Directive 2019/789/EU) 240 Schepel, Harm 169 Schiek, D 13n58, 132, 339, 345n52, 363, 368–9 Schmidberger (Case C-112/00) 105, 357 Schutzverband gegen unlauteren Wettbewerb v TK-Heimdienst Sass (Case C-254/98) 48n72 Scotch Whisky (Case C-333/14) adoption of regulatory frameworks by national authorities 107 and justification of minimum price per alcohol unit 7 Scotland, devolution issue and Scotch Whisky 7 Second Payment Services Directive (PSD2) 195, 207 secondary legislation 295–9 adjudication 299–303 free movement of workers 25, 61 gambling 15, 144–5 preliminary conclusions 303–4 unfair commercial practices law 298–9 selling arrangements see sale of goods services see digital services; freedom to provide services; Services Directive (Directive 2006/123) Services Directive (Directive 2006/123) 146–8, 160, 230 adoption 140, 147 ‘Bolkestein draft’ 271 coverage 146 and E-Commerce Directive 217 establishment chapter 148 implementation 138–9 lack of clarity of key provisions 147 overriding reasons related to the public interest 146 preamble 146

390  Index substantive rule in Article 16 146, 148 text 146–7 see also freedom to provide services servitisation of the economy 20, 187, 192, 194, 207 Single European Act 75 Single European Data Space 210 Single Market 2.0 187–212 B2C domain see Business-to-Consumer (B2C) relationships and COVID-19 crisis 188 digital government 209–10 digital technology, leading the way on 210–11 distributed single market for services 207–9 ecosystem see digital ecosystem in a layered ecosystem 200–11 new architecture in 211–12 new pillars 188 waves of internet evolution 189–200 first transformation wave 190–6 platformisation see platformisation of the economy second transformation wave 196–9 servitisation see servitisation of the economy virtualisation see virtualisation of the economy see also digital services; Digital Single Market (DSM); internet; platformisation of the economy Single Resolution Board 150 Single Supervisory Mechanism (SSM) 139, 150 SITRA (Finnish innovation agency) 198, 207 small and medium-sized enterprises (SMEs) 234, 236 Smith v Meade (Case C-122/17) 60n110 social freedom, and free movement of persons 123–34 see also free movement of persons Social Security Co-ordination Regulation (Regulation 883/2004) 119–20 revising 120, 121, 128 Solid project 198 solidarity principle 1, 5 consequences of establishment of the internal market for 309–10 Somek, Alexander 353, 354 Sotgiu (Case 152/73), indirect discrimination 29

Souvenirs (Case 113/80), indistinctly applicable restrictions 36n38 Spaventa, Eleanor 7, 15–16, 17, 344n47 special treatment see distinctly applicable restrictive measures (special treatment) SSM see Single Supervisory Mechanism (SSM) Staat der Nederlanden v Essent see Essent (Joined Cases C-105/12 to C-107/12) strategically sensitive privatised companies, rights of Member States in see golden shares subsidiarity calculus 3 subsidiarity principle 5, 309 breach of 154 consequences of establishment of the internal market for 306–7 ‘Sunday trading’ case law and Cassis de Dijon (Case 120/78) 45–6 common market 46–7 free movement of goods between Keck and market access 98 and internal market as ‘area without internal frontiers’ 26, 46 and limits of discrimination concept 44–7 and need for revision of free movement principles 25 seen as undue interference with national regulatory autonomy 99 Torfaen Borough Council (Case C-145/88) 66–7 supranationality versus managed diversity 5, 285, 295–304 see also establishment of the internal market, new paradigm for Swedish Match (Case C-151/17) 278 sword interpretation, free movement of capital 179–80 taxation digital 194 double taxation treaties 86n47 free movement of capital 168 free movement of goods 107, 108 internal 313 ‘no representation without’ 348 Scotch Whisky (Case C-333/14) 7 Taylorism 224 Telecoms Single Market (TSM) 242 Television without Frontiers Directive (Directive 89/552) 150

Index  391 Tesauro, Advocate General Giuseppe, on Hünermund (Case C-292/92) 46n65, 67, 68, 99 Test-Achats (Case C-236/09) 88, 287 TEU see Maastricht Treaty (1993); Treaty on European Union (TEU), 1992 TFEU see Treaty on the Functioning of the European Union (TFEU), 2009 Thieffry (Case 71/76) 25, 34–5 Third-party motor liability insurance (Case C-518/06) 40 TK Heimdienst (Case C-254/98), unfair competition cases 69–70 Tobacco Advertisment (Case C-376/98) and harmonisation 9–10, 272, 274, 277 post-Lisbon case law on Art 3(3) TEU 85–6 Tobacco Advertisment (Case C-380/03) 272, 273 Tobacco Products Directive (Directive 2014/40/EU) 273–4 pre-Lisbon test 77, 81–5, 89, 90, 91 token economy 212 Tom Kabinet (Case C-263/18), global contextual factors 80 Torfaen Borough Council (Case C-145/88) all products, legislation applicable to 46 and EEC Treaty 66–7 and ‘Sunday Trading’ cases 44–5 transnational democracy see democracy transparency principle 217 travaux preparatoires 332 Treaty of Amsterdam (1997) 75–6 Treaty of Lisbon (2007) and Article 3(3) TEU 14, 77 common commercial policy concept 89 entry into force (December 2009) 138 and EPSR 134 global contextual factors 79 post-Lisbon case law on Art 3(3) TEU 85–91 importance for competition and horizontal provisions 87–8 importance for internal market provisions 89–91 pre-Lisbon Tobacco Directive test 77, 81–5, 89, 90, 91 subsidiarity principle 306, 307 textual and contextual interpretation of Treaties following 11 on wording of internal market objective 9 Treaty on European Union (TEU), 1992 Article 3

former, repeal of 10 social dimensions of internal market policy 127 Article 3(3) establishment of the internal market as objective in itself 84, 86 importance for competition and horizontal provisions 87–8 importance for internal market provisions 89–91 post-Lisbon case law 85–91 potential of 77 reformulation of the internal market objective 83 rewording of 14 social dimensions of internal market policy 127 equality of Member States before the Treaties obligation 7 establishment of the internal market 286, 287 on EU as a social market economy 12 multilingualism 87 objective of the internal market 9 see also Maastricht Treaty (1993) Treaty on the Functioning of the European Union (TFEU), 2009 Article 34 as ‘classic freedom’ 108 compatibility of rules with 108 limits of 73 market access facilitated by 73–4 product requirements 102 rebuttable and non-rebuttable presumptions, system of 101–2 scope of application 96–7, 101, 104 Article 35 expansion of scope by Court 103 reach of 102–4 scope of application 96–7, 101, 104 Article 56 144, 145 Article 63 horizontal application 166–70 proportionality principle 171 Article 114 broad legal basis/‘catch-up’ scope 139, 150 competence creep 11 digital services 139 and establishment of the internal market 289, 291 harmonisation 264, 265, 267, 268, 269, 271

392  Index legislative competence 349 limits of 18, 336 subsidiarity principle 306 wording of 10 capital restrictions 6 and Common Agricultural Policy (CAP) 288 compatibility with 31 containment measures, legitimate justification 3 data protection 222 express conferral of shared competence for internal market 10 free movement rights of all Union citizens 114 justification of discrimination grounds beyond wording of (Van Binsbergen and Van Wesemael) 31–2 listing all freedoms equally under Article 26(2) 52 non-discriminatory restrictions on the use of products 65–6 Protocol No. 27 87n40, 289, 314 quantitative restrictions and measures having equivalent effect 65 and Treaty superstructure 262–3 and universal bans 46 wording and early secondary law 27–8 existence of justification grounds beyond 31–2 Trotter, S 362 ‘Trustworthy Artificial Intelligence’ Assessment List on Trustworthy AI (AL TAI) 210–11 see also Artificial Intelligence (AI) Turkey–Textiles (WTO case), and customs union 72 Uber 137, 152, 193, 213, 220, 224, 226 proposed role for 231–2 ‘Rides of Glory’ scandal 222 status of drivers 225 tertius genus 230 Uber France (Case C-320/16) 216, 218, 225, 229 Uber Spain (Case C-434/15) 215, 216, 218, 221, 225, 229 UberPop 237 UCPD see Unfair Terms in Consumer Contracts Directive (Directive 93/13/EEC)

Ugliola (Case 15/69), indirect discrimination 29 unequal treatment 27, 28, 39, 40–3 assessment of national rules on merits 42 Lawyers’ fees (Case C-565/08) 4, 40, 43 Third-party motor liability insurance (Case C-518/06) 40, 41, 43 Unfair Commercial Practices Directive (UCPD) (Directive 2005/29/EC) 267, 268, 279–80, 296 unfair commercial practices law 295–8 waves of harmonisation 296–7 unfair competition, subsequent 69–70 Unfair Terms in Consumer Contracts Directive (Directive 93/13/EEC) 13, 239n36, 268, 269, 274–7, 280, 281 uniformity versus diversity in EU internal market law 5–9 over-emphasis on uniformity 17 preference for uniformity 7–8 Union royale belge des societes de football association and others v Bosman and others see Bosman (Case C-415/93) United Kingdom new settlement for within the EU 120 referendum result of June 2016 120–1 withdrawal from the EU (Brexit) 1, 109, 112, 113, 119, 122, 135, 187–8, 363 fundamental freedoms and rights following 363 referendum result of June 2016 120–1 and Single Market 2.0 187–8 UsedSoft (Case C-128/11) 80 van Binsbergen (Case 33/74) 25 ‘general good,’ freedom to provide services rules justified by 32, 34 justification of discrimination grounds beyond wording of the TFEU 31–2 Van Gend & Loos (Case 26/62) 315, 316, 320 van Wesemael (Joined Cases 110/78 and 111/78) 25, 32, 34 varieties of capitalism 165 veto right, double restriction of national sovereignty in terms of 75 video-on-demand services 245 Vienna Convention on the Law of Treaties (VCLT) 300, 320 Viking (Case C-438/05) free movement of goods and duty to act 105

Index  393 fundamental freedoms and rights 358, 367 originalism at the Court 329 post-Lisbon case law on Art 3(3) TEU 85, 88 structure of free movement law 111 Vinegar (Case C-193/80) 36n37 virtualisation of the economy 20, 187, 192, 194 Visser (Joined Cases C-360/15 and C-31/16) 148 Volker Graf v Filzmoser Maschinenbau (Case C-190/98) 48n71 Volkswagen (Case C-112/05) 162 von der Leyen Commission 80, 188, 196, 202, 205, 211 Walloon Waste (Case C-2/90) 37–40 Walt Wilhelm (Case 14/68) 317 Weatherill, Stephen 8–9, 12, 107, 290, 359, 360 Weigel (Case C-387/01) 28 Weiler, JHH 19, 337 welfare and social services 155–7 White, Eric 15, 52 Wightman and Others (Case C-621/18) 292, 299–302, 306

Work Programme of the European Commission 139, 211 workers in collaborative economy 223 fundamental freedoms and rights 366 gig-workers 223–5 labour law systems 153, 223–5 migrant workers 116 Regulation 492/2011 121 secondary legislation 25, 61 see also free movement of workers; Posted Workers’ Directive (Directive 96/71); Posted Workers’ Directive (revised) (Directive 2018/957) Working Time Directive (WTD) (Directive 2003/88/EC) 223 World Trade Organization (WTO) e-commerce rules 237 external relations law 158 external services law 159 and most favoured nation (MFN) principle 72 setting up 79 Zbyszewska, A 134

394