Comparative Federalism: Constitutional Arrangements and Case Law 9781509901494, 9781509901524, 9781509901517

This is the first comprehensive book that explores the subject of federalism from the perspective of comparative constit

222 44 112MB

English Pages [503] Year 2017

Table of contents :
Preface
Contents
Table of Cases
Introduction
0.1 Why This Book and What to Expect from It
0.2 Understanding Federalism-Understanding of Federalism
0.3 Methodology and Terminology
0.4 Structure
Part I: Foundations
1
Concepts
1.1 Federalism
1.2 Regionalism
2
Manifestations
2.1 Confederation
2.2 Federal State
2.3 Regional State and Related Manifestations
2.4 The EU: Federalism Beyond the State?
2.5 Living Without Definitions
3
History
3.1 Ancient Unions
3.2 The Age of Confederations
3.3 Archetypes of the Federal State
4
Debates
4.1 Sovereignty in Federal Systems: Indivisible or Shared?
4.2 Federalism in Diverse Societies: Cure or Curse?
4.3 Federalism and Participatory Democracy in an Era of Pluralism
Part II: Self-Rule and Shared Rule
5
Autonomy of Subnational Entities
5.1 The Constitutional Dimension: Constitutional Autonomy
5.2 The Legislative Dimension: Distribution of Legislative Powers
5.3 The Administrative Dimension: The Relationship between Legislative and Administrative Powers
5.4 The Judicial Dimension: The Court Systems
6
Participation of Subnational Entities at the National Level
6.1 Scope, Institutions and Procedures for Participation
6.2 Institutional Forms of Participation
6.3 Procedural Forms of Participation
7
Financial Relations
7.1 Financial Relations: Significance, Scope and Regulation
7.2 Revenue Powers
7.3 Spending Power
7.4 Equalisation
8
Prevention and Resolution of Conflicts
8.1 Intergovernmental Relations
8.2 Ex ante Scrutiny of Subnational Law
8.3 Constitutional Adjudication
9
Local Government
9.1 Local Government in a Changed Factual and Legal Context
9.2 Legal Recognition of Local Self-Government
9.3 Powers of Local Governments
9.4 Structure of Local Government
9.5 Local Government as a Genuine Third Level?
Part III: Powers and Policies: Between Autonomy and Homogeneity
10
Fundamental Rights
10.1 United States of America
10.2 Switzerland
10.3 Canada
10.4 Germany
10.5 Spain
10.6 Comparative Conclusions
11
Social Welfare and Healthcare
11.1 United States
11.2 Canada
11.3 Belgium
11.4 Spain
11.5 United Kingdom
11.6 Comparative Conclusions
12
Environmental Protection
12.1 United States
12.2 Switzerland
12.3 Germany
12.4 Italy
12.5 United Kingdom
12.6 Comparative Conclusions
13
Immigration and Migrant Integration
13.1 Switzerland
13.2 Canada
13.3 Australia
13.4 Germany
13.5 Belgium
13.6 Comparative Conclusions
14
External Relations
14.1 United States
14.2 Switzerland
14.3 Canada
14.4 Belgium
14.5 Spain
14.6 Comparative Conclusions
15
Concluding Remarks
15.1 Federalism in the Twenty-First Century
15.2 A Federal Research Agenda: The Four ‘Ps’
Index

Recommend Papers

The Global South and Comparative Constitutional Law 0198850409, 9780198850403

This volume makes a timely intervention into a field which is marked by a shift from unipolar to multipolar order and a

590 9 2MB Read more

Comparative constitutional law in Latin America 9781785369209, 1785369202

418 13 1MB Read more

The Oxford Handbook of Comparative Constitutional Law 9780199578610, 0199578613

A comprehensive reference resource on comparative constitutional law, this title examines the history and development of

106 70 14MB Read more

Comparative Constitutional Law in Latin America 9781785369209, 1785369202

This book provides unique insights into the practice of democratic constitutionalism in one of the world’s most legally

526 43 1MB Read more

The Oxford Handbook of Comparative Constitutional Law 9780199578610, 0199578613

A comprehensive reference resource on comparative constitutional law, this title examines the history and development of

105 21 19MB Read more

Federalism and the Charter: Leading Constitutional Decisions 9780773584280

This extensive revision of the landmark Leading Constitutional Decisions brings together recent Charter cases with the c

104 41 50MB Read more

Comparative Company Law: A Case-Based Approach 9781509909360, 9781509909377, 9781509909346

As attention moves rapidly towards comparative approaches, the research and teaching of company law has somehow lagged b

171 29 8MB Read more

Charter versus Federalism: The Dilemmas of Constitutional Reform 9780773563278

In Charter Versus Federalism, Alan Cairns provides an insightful analysis of the consequences -- for citizen and governm

118 37 9MB Read more

Constitutional Law in Theory and Practice 9781442659759

David Beatty draws on more than twenty years' teaching experience to produce a comprehensive introduction to the ba

114 10 11MB Read more

EU Constitutional Law 0198866526, 9780198866527

This title is a comprehensive textbook of EU constitutional law, setting out the structure, values, procedures, and poli

182 67 11MB Read more

Comparative Federalism: Constitutional Arrangements and Case Law
9781509901494, 9781509901524, 9781509901517

Author / Uploaded
Francesco Palermo
Karl Kössler

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

File loading please wait...

Citation preview

COMPARATIVE FEDERALISM This is the first comprehensive book that explores the subject of federalism from the perspective of comparative constitutional law, whilst simultaneously placing a strong emphasis on how federal systems work in practice. This focus is reflected in the book’s two most innovative elements. First, it analyses from a comparative point of view how government levels exercise their powers and interact in several highly topical policy areas like social welfare, environmental protection or migrant integration. Second, the book incorporates case law boxes discussing seminal judgments from federal systems worldwide and thus demonstrates the practical impact of constitutional jurisprudence on policymakers and citizens alike. Volume 19 in the series Hart Studies in Comparative Public Law

Hart Studies in Comparative Public Law Recent titles in this series: The Right to Freedom of Assembly A Comparative Study Orsolya Salát An Inquiry into the Existence of Global Values Through the Lens of Comparative Constitutional Law Edited by Dennis Davis, Alan Richter and Cheryl Saunders The Scope and Intensity of Substantive Review Traversing Taggart’s Rainbow Edited by Hanna Wilberg and Mark Elliott Entick v Carrington 250 Years of the Rule of Law Edited by Adam Tomkins and Paul Scott Administrative Law and Judicial Deference Matthew Lewans Soft Law and Public Authorities Remedies and Reform Greg Weeks Legitimate Expectations in the Common Law World Edited by Matthew Groves and Greg Weeks The Dynamics of Exclusionary Constitutionalism Mazen Masri Constitutional Courts, Gay Rights and Sexual Orientation Equality Angioletta Sperti Principled Reasoning in Human Rights Adjudication Se-Shauna Wheatle Human Rights and Judicial Review in Australia and Canada Janina Boughey The Foundations and Traditions of Constitutional Amendment Edited by Richard Albert, Xenophon Contiades and Alkmene Fotiadou The Federal Idea Public Law Between Governance and Political Life Edited by Amnon Lev

Comparative Federalism Constitutional Arrangements and Case Law

Francesco Palermo and Karl Kössler

OXFORD AND PORTLAND, OREGON 2017

Hart Publishing An imprint of Bloomsbury Publishing Plc Hart Publishing Ltd Kemp House Chawley Park Cumnor Hill Oxford OX2 9PH UK

Bloomsbury Publishing Plc 50 Bedford Square London WC1B 3DP UK

www.hartpub.co.uk www.bloomsbury.com Published in North America (US and Canada) by Hart Publishing c/o International Specialized Book Services 920 NE 58th Avenue, Suite 300 Portland, OR 97213-3786 USA www.isbs.com HART PUBLISHING, the Hart/Stag logo, BLOOMSBURY and the Diana logo are trademarks of Bloomsbury Publishing Plc First published 2017 © Francesco Palermo and Karl Kössler 2017 Francesco Palermo and Karl Kössler 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/, 1998–2017. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. ISBN: HB: 978-1-50990-149-4 ePDF: 978-1-50990-151-7 ePub: 978-1-50990-150-0 Library of Congress Cataloging-in-Publication Data Names: Palermo, Francesco, author. | Kössler, Karl, 1979-, author. Title: Comparative federalism : constitutional arrangements and case law / Francesco Palermo and Karl Kössler. Description: Portland, Oregon : Hart Publishing, 2017. | Series: Hart studies in comparative public law ; volume 19 | Includes bibliographical references and index. Identifiers: LCCN 2017025633 (print) | LCCN 2017027012 (ebook) | ISBN 9781509901500 (Epub) | ISBN 9781509901494 (hardback : alk. paper) Subjects: LCSH: Federal government—Europe. | Constitutional law—Europe. | Regionalism—Europe. | Sovereignty—Europe. | International and municipal law—Europe. Classification: LCC KJC5048 (ebook) | LCC KJC5048 .P35 2017 (print) | DDC 342.24/042—dc23 LC record available at https://lccn.loc.gov/2017025633 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.

Preface This book has been in the making for about a decade. The idea to write a comprehensive manual on comparative federalism emerged from a certain discontent with the gap in the classical literature on the subject. This literature, even if extremely useful, predominantly approaches federalism from the perspectives of political theory and/or comparative politics, while a book that also sheds light on this phenomenon from the perspective of comparative constitutional law has been lacking so far. At the same time, attention to the constitutional dimension of federalism could not be limited to the institutional facets, but had to include the way federal dynamics are working and evolving in political practice. Put differently, and to echo the motto of Hart Publishing, our aim was therefore to write on this subject a ‘good book for lawyers’—but not exclusively for them. Thus, we hope that this publication will equally attract the attention of non-lawyers, as it decidedly does not only focus on ‘law in the books’ but rather on powers and policies as well as on the practical impact of constitutional jurisprudence on policy-makers and citizens alike. In view of this focus, we trust that the book will prove useful for students and scholars from related academic disciplines, above all political science, and practitioners of federalism in governments and international organisations. As writing this book has been a decade-long journey of ‘Odyssean’ proportions, we have accumulated a large number of debts along the way. Some people have encouraged us not to be distracted by the siren songs to make the project less ambitious and comprehensive, while others have kept our ship on track even during the most tempestuous times. A large number of dear colleagues and friends have offered intellectual support on this journey, some perhaps without being aware of it. Several people have been a constant source of inspiration and encouragement, since they have wholeheartedly endorsed our project as members of our institute’s Scientific Committee and/or have read and commented on the various drafts of the manuscript. We are deeply grateful, in particular, to Anna Gamper, Cheryl Saunders, G Alan Tarr, Balveer Arora and Nico Steytler. Many more people from international academic networks on federalism have had a profound influence on our work, above all those brought together by the International Association of Centers for Federal Studies (IACFS) and those belonging to our institute’s ‘Ring of Friends’. Furthermore, this book would have never been conceived without the invaluable legacy of the institute’s first directors, Sergio Ortino and Peter Pernthaler. Beyond academia, we owe a lot to practitioners of federalism in governments and international organisations who have provided us with valuable insights on how federalism works (or does not work) on the ground. At Hart Publishing, we are much obliged to Bill Asquith and his team for excellent editorial guidance and the skillful management of all aspects of the production process. Moreover, the comments by Hart’s anonymous reviewers are greatly

vi Preface appreciated. They provided valuable inputs for improvements and at the same time strong encouragement by confirming that the book is timely and really fills a significant gap in research. Last, but by no means least, we would like to express profound gratitude to our home institution, Eurac Research in Bolzano/Bozen, and to all the researchers of its Institute for Comparative Federalism. The discussions with our colleagues have helped us to structure a book that covers such a broad field and to revise the single chapters. Jens Woelk and Matteo Nicolini have accompanied this project for some time. We are g rateful for their efforts to advance the project and their generous support. Annika Kress, soon after joining our team, emerged as an important driving force behind the process of producing this book. Both her tireless copy-editing support and sharp-witted observations as our first critical reader have been truly invaluable. Francesco Palermo and Karl Kössler Bolzano/Bozen, May 2017

Contents Preface�� v Table of Cases�� xiii Introduction�� 1 0.1 Why This Book and What to Expect from it�� 1 0.2 Understanding Federalism—Understanding of Federalism�� 3 0.3 Methodology and Terminology�� 5 0.3.1 The Comparative Method�� 5 0.3.2 Terminology�� 8 0.4 Structure�� 10 Part I: Foundations 1. Concepts�� 13 1.1 Federalism�� 13 1.1.1 Traditions in Conceptualising Federalism�� 14 1.1.2 Federalism and Federation—Towards Common Ground�� 17 1.2 Regionalism�� 20 1.2.1 Regionalisation vs Regionalism�� 22 1.2.2 New Regionalism, Multilevel Governance and the ‘Europe of the Regions’�� 25 1.2.3 The Proliferation of Regionalisms�� 30 2. Manifestations�� 34 2.1 Confederation�� 35 2.1.1 Definition�� 35 2.1.2 Are Confederations Transitional?�� 36 2.2 Federal State�� 38 2.2.1 The Compact as the Traditional Definition: The ‘Federal Big Bang’�� 38 2.2.2 Form Follows Function: The Role of History in Shaping Federal Manifestations�� 42 2.2.3 Definition by Means of Institutional Elements�� 47 2.3 Regional State and Related Manifestations�� 50 2.3.1 Definition and Relation to the Federal State�� 50 2.3.2 Rationale and Cases�� 51 2.3.3 Devolution�� 55 2.3.4 Autonomy and Similar Manifestations�� 58 2.4 The EU: Federalism Beyond the State?�� 61 2.5 Living Without Definitions�� 64

viii Contents 3. History�� 67 3.1 Ancient Unions�� 67 3.2 The Age of Confederations�� 68 3.2.1 Classical European Confederations�� 68 3.2.2 The United States of America under the Articles of Confederation�� 70 3.3 Archetypes of the Federal State�� 72 3.3.1 The United States of America (1787)�� 72 3.3.2 Switzerland (1848)�� 76 3.3.3 Germany (1867/1871)�� 79 4. Debates�� 83 4.1 Sovereignty in Federal Systems: Indivisible or Shared?�� 83 4.1.1 The Theory of Indivisible Sovereignty and Its Contestation�� 84 4.1.2 Divided Sovereignty in the United States and Beyond�� 88 4.1.3 The Sovereignty Question in the European Union�� 92 4.2 Federalism in Diverse Societies: Cure or Curse?�� 97 4.2.1 Multinational Federalism as a Concept�� 98 4.2.2 The Legal Status of Internal Minorities�� 101 4.2.3 Secession under Federal Constitutions�� 105 4.3 Federalism and Participatory Democracy in an Era of Pluralism�� 111 4.3.1 Federalism and Participatory Democracy: Two Faces of Pluralism�� 111 4.3.2 The Crisis of Democracy: An Opportunity for Federalism?�� 114 Part II: Self-Rule and Shared Rule 5. Autonomy of Subnational Entities�� 125 5.1 The Constitutional Dimension: Constitutional Autonomy�� 126 5.1.1 Definition and Scope�� 126 5.1.2 Limits to Constitutional Autonomy: Supremacy and Homogeneity Clauses�� 130 5.1.3 Practical Use of Constitutional Autonomy�� 136 5.2 The Legislative Dimension: Distribution of Legislative Powers�� 139 5.2.1 Distribution of Powers and the Residual Clause�� 139 5.2.2 Dual vs Cooperative Systems�� 146 5.2.3 Flexibilisation and Centralisation: Implied Powers�� 149 5.3 The Administrative Dimension: The Relationship between Legislative and Administrative Powers�� 156 5.4 The Judicial Dimension: The Court Systems�� 159 6. Participation of Subnational Entities at the National Level�� 164 6.1 Scope, Institutions and Procedures for Participation�� 164 6.2 Institutional Forms of Participation�� 165 6.2.1 Second Chambers�� 165 6.2.1.1 Origins and Constitutional Entrenchment�� 167 6.2.1.2 The Senate Model: Variations, Appointment and Composition�� 169

Contents ix 6.2.1.3 The Council (or Ambassadorial) Model: A Germany Legacy�� 172 6.2.1.4 Hybrid Second Chambers�� 176 6.2.2 Intergovernmental Bodies�� 177 6.3 Procedural Forms of Participation�� 178 6.3.1 Constitutional Amendment�� 178 6.3.1.1 Direct Participation by the Legislatures or Electorates of Subnational Entities�� 179 6.3.1.2 Indirect (or Absence of) Participation Through Second Chambers�� 183 6.3.2 Territorial Changes�� 186 6.3.3 Legislative (and Administrative) Functions�� 190 6.3.4 Appointments�� 193 6.3.5 Other Functions�� 199 7. Financial Relations�� 201 7.1 Financial Relations: Significance, Scope and Regulation�� 201 7.1.1 Significance�� 201 7.1.2 Scope: Local and Supranational Dimensions�� 202 7.1.3 Regulation: Legal Sources and (A)Symmetrical Design�� 205 7.2 Revenue Powers�� 210 7.2.1 Types of Subnational Entities’ Own Revenues�� 210 7.2.2 Taxing Powers�� 211 7.2.3 A Trend Towards Tax Centralisation?�� 222 7.3 Spending Power�� 228 7.3.1 Spending Power and Conditional Grants�� 228 7.3.2 Legal Basis and Limits of the General Spending Power�� 229 7.3.3 Spending Power and Fiscal Discipline�� 235 7.4 Equalisation�� 240 8. Prevention and Resolution of Conflicts�� 246 8.1 Intergovernmental Relations�� 246 8.1.1 Principles: Federal Loyalty and Similar Precepts�� 249 8.1.2 Institutions: The Predominance of the Executive Branch�� 253 8.1.3 Instruments: Delegation and Intergovernmental Agreements�� 257 8.2 Ex ante Scrutiny of Subnational Law�� 261 8.2.1 Judicial Scrutiny of Draft Legislation�� 261 8.2.2 Political or Judicial Scrutiny of Subnational Constitutions�� 263 8.3 Constitutional Adjudication�� 266 8.3.1 Significance, Scope and Impact�� 266 8.3.2 Constitutional Review and Other Functions�� 272 9. Local Government�� 281 9.1 Local Government in a Changed Factual and Legal Context�� 281 9.2 Legal Recognition of Local Self-Government�� 284 9.2.1 Source of Recognition�� 285

x Contents

9.3

9.4

9.5

9.2.2 Scope and Quality of Recognition�� 286 9.2.3 The Problem of (Over)Regulation�� 287 Powers of Local Governments�� 290 9.3.1 Assignment of Local Powers�� 290 9.3.2 Restrictions of the Exercise of Local Powers�� 295 9.3.2.1 Overburdening with Delegated Tasks�� 295 9.3.2.2 Supervision Prevailing over Intergovernmental Collaboration�� 297 9.3.2.3 Limited Financial Resources�� 300 Structure of Local Government�� 305 9.4.1 Amalgamation�� 305 9.4.2 Inter-municipal Cooperation�� 309 9.4.3 Umbrella Entities�� 312 Local Government as a Genuine Third Level?�� 314

Part III: Powers and Policies: Between Autonomy and Homogeneity 10. Fundamental Rights�� 321 10.1 United States of America�� 322 10.2 Switzerland�� 327 10.3 Canada�� 330 10.4 Germany�� 333 10.5 Spain�� 338 10.6 Comparative Conclusions�� 341 11. Social Welfare and Healthcare�� 346 11.1 United States�� 349 11.2 Canada�� 355 11.3 Belgium�� 361 11.4 Spain�� 365 11.5 United Kingdom�� 370 11.6 Comparative Conclusions�� 376 12. Environmental Protection�� 382 12.1 United States�� 383 12.2 Switzerland�� 387 12.3 Germany�� 391 12.4 Italy�� 393 12.5 United Kingdom�� 396 12.6 Comparative Conclusions�� 398 13. Immigration and Migrant Integration�� 403 13.1 Switzerland�� 404 13.2 Canada�� 407 13.3 Australia�� 411 13.4 Germany�� 412 13.5 Belgium�� 415 13.6 Comparative Conclusions�� 418

Contents xi 14. External Relations�� 423 14.1 United States�� 425 14.2 Switzerland�� 428 14.3 Canada�� 431 14.4 Belgium�� 436 14.5 Spain�� 439 14.6 Comparative Conclusions�� 443 15. Concluding Remarks�� 448 15.1 Federalism in the Twenty-First Century�� 448 15.2 A Federal Research Agenda: The Four ‘Ps’�� 449 Index�� 453

xii

Table of Cases Note that judgments presented in greater detail in the case law boxes are in bold. Australia High Court of Australia Amalgamated Society of Engineers v Adelaide Steamship Co Ltd (1920) 28 CLR 129 (Engineers’ case)�� 268 Attorney-General (Cth) v Colonial Sugar Refining Co Ltd (1913) 17 CLR 644�� 127 Austin v Commonwealth (2003) 215 CLR 185�� 127 Australian Communist Party v Commonwealth (1951) 83 CLR 1�� 132 Clayton v Heffron (1960) 105 CLR 214�� 127 Judiciary and Navigation Acts, Re (1921) 29 CLR 257 (Advisory Opinions case)�� 277 New South Wales v Commonwealth (2006) 229 CLR 1 (Work Choices case)�� 232 Pape v Commissioner of Taxation (2009) 238 CLR 1�� 268 Queensland v Commonwealth (1977) 139 CLR 585 (Second Territories Representation case)�� 172 R v Hughes (2000) 202 CLR 535�� 257 South Australia v Commonwealth (1942) 65 CLR 373 (First Uniform Tax case)�� 223 Streets v Queensland Bar Association (1989) 168 CLR 461�� 132 Thomas v Mowbray (2007) 233 CLR 307�� 257 Victoria v Commonwealth (1957) 99 CLR 575 (Second Uniform Income Tax case)�� 223, 232 Wakim, Re (1999) 198 CLR 511�� 257 Western Australia v Commonwealth (1975) 134 CLR 201 (First Territories Representation case)�� 172 Western Australia v Commonwealth (1995) 183 CLR 373 (Native Title Act case)�� 132 Williams v Commonwealth (2012) 248 CLR 156 (School Chaplains case)�� 231 Williams v Commonwealth (2014) 252 CLR 416�� 232 Austria Constitutional Court VfSlg 8/1921�� 158 VfSlg 206/1923�� 158 VfSlg 2455/1952�� 39, 166 VfSlg 3055/1956�� 263 VfSlg 5676/1968�� 134 VfSlg 6944/1972�� 294 VfSlg 8719/1979�� 294 VfSlg 8831/1980�� 250 VfSlg 9886/1983�� 260 VfSlg 10292/1984�� 133, 250

xiv Table of Cases VfSlg 10510/1985�� 158 VfSlg 11403/1987�� 39 VfSlg 11669/1988�� 39 VfSlg 13500/1993�� 288 VfSlg 14534/1996�� 251 VfSlg 15552/1999�� 250 VfSlg 16241/2001�� 134 VfSlg 17173/2004�� 191 VfSlg 19782/2013�� 170 VfSlg 19868/2014�� 297 VfSlg 19894/2014�� 308 Belgium Constitutional Court No 33/1992�� 439 No 49/1994 �� 249 No 42/1997�� 249 No 04/1998�� 214 No 33/2001�� 363 No 100/2003�� 214 No 166/2003�� 146 No 5/2004�� 417 No 51/2006�� 363 No 101/2008�� 417 No 33/2011�� 247 No 67/2014�� 247 Bosnia and Herzegovina Constitutional Court Partial Decision U5/98 III of 1 July 2000 (Constituent Peoples case)�� 99, 104 Partial Decision U5/98 IV of 18 August 2000�� 147 U 25/00 of 23 March 2001�� 105 U 37/01 of 3 November 2001�� 105 Partial Decision U44/01 of 27 February 2004�� 99 Partial Decision U4/04 of 31 March 2006�� 99 U3/13 of 26 November 2015�� 99 Canada Judicial Committee of the Privy Council (JCPC) and Supreme Court of Canada Angers v MNR [1957] Ex CR 83�� 379 Anti-Inflation Act, Reference Re [1976] 2 SCR 373�� 149 Authority of the Parliament in Relation to the Upper House, Reference Re [1980] 1 SCR 54�� 168 Bank of Toronto v Lambe [1887] 12 AC 575 (JCPC)�� 215 BC Motor Vehicle Act, Reference Re [1985] 2 SCR 486�� 145

Table of Cases xv Bell Canada v Quebec (Commission de la santé et de la sécurité du travail) [1988] 1 SCR 749�� 145, 357 Boucher v Stelco Inc [2005] 3 SCR 279�� 259 Bruker v Marcovitz [2007] 3 SCR 607�� 333 Canada Assistance Plan (BC), Reference Re [1991] 2 SCR 525�� 233, 261, 358, 379 Canada (AG) v Ontario (AG) [1937] AC 326 (JCPC)�� 145, 435 Canada (AG) v Ontario (AG) [1937] AC 355 (JCPC)�� 270, 356, 376 Citizens Insurance Co v Parsons [1881] 7 AC 96�� 144 Edwards v Canada (AG) [1930] AC 124�� 145, 272 Eldridge v British Columbia (AG) [1997] 2 SCR 624�� 360 Employment Insurance Act, ss 22 and 23, Reference Re [2005] 2 SCR 669�� 357 Finlay v Canada (Minister of Finance) [1986] 2 SCR 607�� 231 Ford v Quebec (AG) [1988] 2 SCR 712�� 331, 342 Godbout v Longueuil (City of) [1997] 3 SCR 844�� 333 Hodge v The Queen (Ont) [1883] 9 AC 117 (JCPC)�� 145 Initiative and Referendum Act, Reference Re [1919] AC 935 (JCPC)�� 120, 331 Law Society of British Columbia v Mangat [2001] 3 SCR 113�� 410 Lovelace v Ontario [2000] 1 SCR 950�� 235 Manitoba (AG) v Forest [1979] 2 SCR 1032�� 127 Multiple Access Ltd v McCutcheon [1982] 2 SCR 161�� 133 Nova Scotia (AG) v Canada (AG) [1951] SCR 31�� 257 Objection by Quebec to Resolution to Amend the Constitution, Reference Re [1982] 2 SCR 793�� 181 Ontario (AG) v Ontario Public Service Employees’ Union [1987] 2 SCR 2�� 127 Ontario (AG) v Canada (AG) [1912] AC 571 (JCPC)�� 277 Ontario (AG) v Canada (AG) [1896] AC 348 (JCPC)�� 132 Quebec (AG) v Blaikie [1979] 2 SCR 1016�� 127 R v Crown Zellerbach Canada Ltd [1988] 1 SCR 401�� 149 R v Nat Bell Liquors Ltd [1922] 2 AC 128 (JCPC)�� 120, 331 Resolution to Amend the Constitution, Reference Re [1981] 1 SCR 753�� 180, 277 Same-Sex Marriage, Reference Re [2004] 3 SCR 698�� 145 Schneider v British Columbia [1982] 2 SCR 112 �� 360, 378 Secession of Québec, Reference Re [1998] 2 SCR 217�� 38, 109, 277 Securities Act, Reference Re [2011] 3 SCR 837�� 269 Senate Reform, Reference Re [2014] 1 SCR 704�� 168 Solski (Tutor of) v Quebec (Attorney General) [2005] 1 SCR 201�� 342 Supreme Court Act, ss 5 and 6, Reference Re [2014] 1 SCR 433�� 196 Syndicat Northcrest v Anselem [2004] 2 SCR 551�� 333 The Queen (Ont) v Board of Transport Commissioners [1968] SCR 118�� 235 Toronto (City of) Electric Commissioners v Snider [1925] AC 396 (JCPC) 412�� 360 UL Canada Inc v Quebec (AG) [2005] 1 SCR 143�� 259 Ward v Canada (AG) [2002] 1 SCR 569�� 144 European Union Court of Justice of the European Union (CJEU) Case 8/55 Fédération charbonnière de Belgique v ECSC High Authority [1955] ECR 201�� 152 Case 26/62 Van Gend en Loos v Administratie der Belastingen [1963] ECR 3�� 63, 93

xvi Table of Cases Case 6/64 Costa v ENEL [1964] ECR 1194�� 63, 93 Case 29/69 Stauder v City of Ulm [1969] ECR 419�� 63 Case 11/70 Internationale Handelsgesellschaft v Einfuhr- und Vorratsstelle für Getreide und Futtermittel [1970] ECR 1125�� 63 Case 22/70 Commission v Council [1971] ECR 263 (AETR)�� 153 Case 8/73 Hauptzollamt Bremerhaven v Massey Ferguson [1973] ECR 897�� 152 Case 294/83 Les Verts v Parliament [1986] ECR 1339�� 63 Joined Cases 281, 283, 284, 285 and 287/85 Germany & Ors v Commission [1987] ECR 3203�� 152 Case 45/86 Commission v Council [1987] ECR 1493�� 152 Case 103/88 Fratelli Costanzo [1989] ECR 1839�� 425 Case C-376/98 Germany v Parliament [2000] ECR I-8419�� 144, 424 Case C-103/01 Commission v Germany [2003] ECR I-5369�� 424 Case C-212/06 Government of the French Community and Walloon Government v Flemish Government [2008] ECR I-1683�� 363 France Council of State Case Nicolo (1989)�� 63 Germany Federal Constitutional Court 1 BVerfGE 14 Südweststaat (Southwest State)�� 186, 275, 280 1 BVerfGE 117 Finanzausgleichsgesetz (Financial Equalization Act)��243–44 1 BVerfGE 167 Kommunales Selbstverwaltungsrecht (Local Self-Governance)�� 286 1 BVerfGE 299 Wohnungsbauförderung (Housing Funding)�� 250 2 BVerfGE 237 Hypothekensicherungsgesetz (Mortgage Security Act)�� 336 4 BVerfGE 178 Landesgesetz über die Verwaltungsgerichtsbarkeit (State Law on Administrative Jurisdiction)�� 128 5 BVerfGE 34 Baden-Abstimmung (Baden Home Association)�� 187 6 BVerfGE 32 Elfes-Urteil (Elfes case)�� 338 6 BVerfGE 309 Reichskonkordat (Concordat case)�� 250 8 BVerfGE 122 Volksbefragung Hessen (Hesse Referendum case)�� 250 9 BVerfGE 268 Bremer Personalvertretung (Bremen Civil Servant case)�� 135 12 BVerfGE 205 1. Rundfunkentscheidung (First Broadcasting case)�� 280 22 BVerfGE 267 Einheitliches Grundrecht (Uniform Fundamental Right)�� 336 26 BVerfGE 228 Sorsum�� 299 27 BVerfGE 71 Leipziger Volkszeitung (Leipzig Daily Newspaper)�� 336 32 BVerfGE 333 Ergänzungsabgabe (Supplemental Tax)�� 229 33 BVerfGE 303 Numerus Clausus I�� 260 34 BVerfGE 9 Besoldungsvereinheitlichung (Federal–State Salary)�� 91 36 BVerfGE 342 Niedersächsisches Landesbesoldungsgesetz (Remuneration Law of Lower Saxony)�� 91, 259, 336 37 BVerfGE 363 Bundesrat�� 173, 192 49 BVerfGE 89 Kalkar I�� 393

Table of Cases xvii 49 BVerfGE 168 Vertrauensschutz bei Verlängerung der Aufenthaltsgenehmigung (Legitimate Expectation Regarding the Renewal of Residence Permits)�� 413 52 BVerfGE 95 Schleswig-Holsteinische Ämter (Ämter of Schleswig-Holstein)�� 311 53 BVerfGE 185 Hessisches Gesetz über die Neuordnung der gymnasialen Oberstufe�� 336 55 BVerfGE 274 Berufsausbildungsabgabe (Vocational Training Act)�� 192 61 BVerfGE 149 Amtshaftung (State Liability)�� 250 72 BVerfGE 330 Finanzausgleich I (Financial Equalization I)��241, 243–44, 250 73 BVerfGE 339 Solange II�� 63 79 BVerfGE 127 Rastede�� 293 81 BVerfGE 310 Kalkar II�� 250, 393, 400 83 BVerfGE 37 Ausländerwahlrecht I (Foreign Voters I)�� 135, 415 86 BVerfGE 148 Finanzausgleich II (Financial Equalization II)�� 239 89 BVerfGE 155 Maastricht�� 64, 94 92 BVerfGE 203 EG-Fernsehrichtlinie (EC-Television-Directive)�� 250 93 BverfGE 1 Kruzifix im Klassenzimmer II (Classroom Crucifix II)�� 334 101 BVerfGE 158 Finanzausgleich III (Financial Equalization III)�� 243 106 BVerfGE 310 Zuwanderungsgesetz (Immigration Act)�� 174 116 BVerfGE 327 Berliner Haushalt (Budget of Berlin)�� 239 122 BVerfGE 342 Bayerisches Versammlungsgesetz (Bavarian Assembly Act)�� 335 123 BVerfGE 267 Lissabon (Lisbon)�� 64, 94 125 BVerfGE 141 Gewerbesteuer (Trade Tax)�� 303 128 BVerfGE 1 Gentechnikgesetz (Genetic Engineering Act)�� 391 134 BVerfGE 366 OMT-Beschluss (Outright Monetary Transactions)�� 64, 204 135 BVerfGE 317 ESM-Vertrag (European Stability Mechanism)�� 64, 204 BVerfG 21 March 2012, 1 BvR 2492/08�� 335 BVerfG 21 June 2016, 2 BvR 2728/13�� 204 BVerfG 16 December 2016, 2 BvR 349/16�� 91 Staatsgerichtshof für das Deutsche Reich (25 October 1932)�� 81 StGH Hess 30 December 1981, PSt 880�� 336 India Supreme Court of India Ganga Ram Moolchandani v State of Rajasthan, 2001 AIR 2616�� 39 Golaknath v State of Punjab, 1967 AIR 1643�� 41 His Holiness Kesavananda Bharati Sripadagalvaru v State of Kerala, 1973 AIR 1461�� 41, 185, 276, 291 Indira Nehru Gandhi v Raj Narain, 1975 AIR 2299�� 41 International Tourist Corporation v State of Haryana, 1981 AIR 774�� 213 ITC Limited v Agricultural Produce Market, 2002 AIR 852�� 276 Minerva Mills Ltd v Union of India, 1980 AIR 1789�� 185, 276 Rai Sahib Ram Jawaya Kapur v State of Punjab, 1955 AIR 549�� 154 SR Bommai v Union of India, 1994 AIR 1918�� 41, 155, 200 State of Bihar v Bal Mukund Sah, 2000 AIR 1296�� 41 State of Karnataka v Union of India, 1978 AIR 68�� 279 State of Rajasthan v Union of India, 1977 AIR 1361�� 154

xviii Table of Cases State of West Bengal v Union of India, 1963 AIR 1241�� 39, 279 State of West Bengal v Kesoram Industries, 2005 AIR 1646�� 276 Western Coalfields Ltd vs Special Area Development, 1982 AIR 697�� 146 Italy Constitutional Court No 49/1958�� 251 No 170/1984�� 63 No 219/1984�� 251 No 358/1985�� 251 No 359/1985�� 396, 401 No 151/1986�� 396, 401 No 153/1986�� 251 No 273/1998�� 394 No 382/1999�� 394 No 42/2001�� 251 No 106/2002�� 121 No 274/2003�� 275 No 379/2004�� 120 No 365/2007�� 7, 51 No 367/2007�� 394 No 378/2007�� 394 No 61/2009�� 394, 400 No 74/2009�� 210 No 225/2009�� 395, 401 No 282/2009�� 395 No 119/2010�� 395 No 186/2010�� 394 No 234/2010�� 394 No 44/2011�� 394 Malaysia Judicial Committee of the Privy Council (JCPC) and Federal Court of Malaysia Mamat bin Daud v Government of Malaysia [1986] 2 MLJ 192�� 271 Petroliam Nasional Bhd v State Government of Terengganu [2004] 1 MLJ 8�� 211 Stephen Kalong Ningkan v Tun Abang Haji Openg and Tawi Sli [1966] 2 MLJ 187�� 130 Stephen Kalong Ningkan v Government of Malaysia [1968] 2 MLJ 238�� 130 Mexico Supreme Court of Justice Constitutional Controversy 31/1997�� 279 Constitutional Controversy 29/2000�� 147 Constitutional Controversy 5/2004�� 301

Table of Cases xix Nigeria Supreme Court of Nigeria Attorney General of the Federation v Attorney General of Abia State SC 28/2001 of 5 April 2002�� 221 Attorney General of Ondo State v Attorney General of the Federation SC 6/2002 of 7 June 2002�� 275 Russian Federation Constitutional Court Judgment of 18 January 1996, no 2-P, Vestnik Konstitutionnogo Suda RF 1996 no 1�� 130 Judgment of 1 February 1996, no 3-P, Vestnik Konstitutionnogo Suda RF 1996 no 1�� 160 Judgment of 7 June 2000, no 10-P, Vestnik Konstitutionnogo Suda RF 2000 no 5�� 128 Judgment of 21 December 2005, no 13-P, Vestnik Konstitutionnogo Suda RF 2006 no 3�� 130 South Africa Constitutional Court Cape Bar Council v Judicial Services Commission 2012 (4) BCLR 406 (WCC)�� 195 CDA Boerdery (Edms) Bpk v Nelson Mandela Metropolitan Municipality 2007 SCA 1 (RSA)�� 286 Certification of the Amended Text of the Constitution of the Republic of South Africa 1997 (2) SA 97�� 313 Certification of the Constitution of the Republic of South Africa 1996 (4) SA 744 (CC)�� 265, 312 Certification of the Constitution of the Western Cape 1998 (1) SA 655�� 265 Certification of the Kwazulu-Natal Constitution 1996 (4) SA 1098�� 136, 264 City of Cape Town v Robertson 2005 (2) SA 323 (CC)�� 286 Constitutionality of the Liquor Bill 2000 (1) SA 732�� 147 Executive Council of the Province of the Western Cape v Minister for Provincial Affairs and Constitutional Development, Executive Council of KwaZulu-Natal v President of the Republic of South Africa 2000 (1) SA 661�� 147, 306 Judicial Service Commission v Cape Bar Council 2013 (1) SA 170 (SCA)�� 195 Liebenberg NO v Bergrivier Municipality 2013 (5) SA 246 (CC)�� 203 Matatiele Municipality v President of the Republic of South Africa 2006 (5) SA 47 (CC)�� 188 MEC, Mpumalanga v Imata 2002 (2) SA 76 (SCA)�� 304 Premier, Western Cape v President of the Republic of South Africa 1999 (3) SA 657�� 147, 220 Premier: Limpopo Province v Speaker of the Limpopo Provincial Legislature 2012 (4) SA 58 (CC)�� 254 Premier: Limpopo Province v Speaker: the Limpopo Provincial Legislature 2011 (6) SA 396 (CC)�� 254 The National Education Policy Bill 1996 (3) SA 289�� 92, 264 Uthukela District Municipality v President of the Republic of South Africa 2003 (1) SA 678�� 252

xx Table of Cases Spain Constitutional Court STC 4/1981�� 299, 302, 339, 341 STC 32/1981�� 92, 147, 289, 314 STC 18/1982�� 251, 442 STC 26/1982�� 441 STC 44/1982��440–41 STC 84/1982�� 288 STC 76/1983�� 53 STC 89/1984�� 339 STC 66/1985�� 339 STC 11/1986�� 251 STC 146/1986�� 367, 377 STC 37/1987�� 214 STC 179/1987�� 224 STC 252/1988�� 442 STC 124/1989�� 377 STC 137/1989�� 440 STC 153/1989�� 440 STC 214/1989�� 288 STC 46/1990�� 442 STC 96/1990�� 302 STC 13/1992�� 224, 232, 234, 258 STC 135/1992�� 224 STC 80/1993�� 440 STC 165/1994��440–41 STC 109/1998�� 289 STC 289/2000�� 214, 224 STC 62/2001�� 236 STC 159/2001�� 294 STC 48/2004�� 299 STC 58/2004�� 289 STC 13/2007�� 209 STC 247/2007�� 224, 339, 341 STC 49/2008�� 198 STC 31/2010�� 139, 209, 244, 339, 341 STC 134/2011�� 236 STC 42/2014�� 110 STC 31/2015�� 110 STC 32/2015�� 110 STC 259/2015�� 110 Switzerland Federal Court BGE 101 Ia 67�� 328 BGE 104 Ia 215�� 328 BGE 116 Ia 359�� 122

Table of Cases xxi BGE 116 Ib 265�� 389 BGE 118 Ia 124�� 265 BGE 121 I 267�� 329 BGE 121 I 334�� 389, 399 BGE 121 I 367�� 328 BGE 125 I 458�� 227 BGE 129 I 217��406–07 BGE 130 I 140�� 406 BGE 132 I 291�� 307 BGE 139 I 16�� 78 UK House of Lords, High Court of Justice and Supreme Court of the United Kingdom Agricultural Sector (Wales) Bill—Reference by the Attorney General for England and Wales [2014] UKSC 43�� 262 AXA General Insurance Ltd v Lord Advocate (Scotland) [2011] UKSC 46�� 278 Martin & Miller v Lord Advocate [2010] UKSC 10�� 278 Rothwell v Chemical & Insulating Co Ltd [2007] UKHL 2�� 278 R (Miller) v Secretary of State for Exiting the European Union [2016] EWHC 2768 (Admin)�� 57 R (on the application of Miller and Dos Santos) v Secretary of State for Exiting the European Union [2017] UKSC 5�� 57 Thoburn v Sunderland City Council [2002] EWHC 195 (Admin), (Metric Martyrs case)�� 56 USA Supreme Court of the United States ALA Schechter Poultry Corp v United States 295 US 495 (1935)�� 349 Arizona State Legislature v Arizona Independent Redistricting Commission 576 US ___ (2015)�� 266 Atkin v Kansas 191 US 207 (1903)�� 281, 292 Baker v Carr 369 US 186 (1962)�� 136 Baker v Nelson 291 Minn 310, 191 NW 2d 185 (1971)��326–27 Baker v Vermont 744 A 2d 864 (Vt 1999)�� 324 Baldwin v GAF Seelig Inc 294 US 511 (1935)�� 143 Barron v Baltimore 32 US (7 Pet) 243 (1833)�� 322 Bowers v Hardwick 478 US 186 (1986)�� 325, 327 Brown v Board of Education 347 US 483 (1954)�� 131 Brushaber v Union Pacific Railroad Co 240 US 1 (1916)�� 216 Carter v Carter Coal Co 298 US 238 (1936)�� 143, 349 Chicago, Burlington & Quincy Railroad Co v Chicago (City of) 166 US 226 (1897)�� 322 Chisholm v Georgia 2 US (2 Dall) 419 (1793)�� 90, 270 Clark v Des Moines (City of) 19 Iowa 199 (1865)�� 291 Clinton (City of) v Cedar Rapids & Missouri River Railroad Co 24 Iowa 455 (1868)�� 291 Collector v Day 78 US (11 Wall) 113 (1870)�� 216

xxii Table of Cases Comptroller of the Treasury of Maryland v Wynne 575 US ___ (2015)�� 227 Cooley v Board of Wardens 53 US (12 How) 299 (1852)��142–43 Cooper v Aaron 358 US 1 (1958)�� 131 Crosby v National Foreign Trade Council 530 US 363 (2000)�� 426 DaimlerChrysler Corp v Cuno 547 US 332 (2006)�� 426 Di Santo v Pennsylvania 273 US 34 (1927)�� 143 Erie Railroad Co v Tompkins 304 US 64 (1938)��161–62 Ex Parte Young 209 US 123 (1908)�� 91 Field v Clark 143 US 649 (1892)�� 156 Garcia v San Antonio Metropolitan Transit Authority 469 US 528 (1985)�� 166, 273 Gibbons v Odgen 22 US (9 Wheat) 1 (1824)�� 142 Gitlow v New York 268 US 652 (1925)�� 322 Gonzales v Raich 545 US 1 (2005)�� 143, 384 Goodridge v Department of Public Health 798 NE 2d 941 (Mass 2003)�� 324 Graves v New York ex rel O’Keefe 306 US 466 (1939)�� 216 Gregg v Georgia 428 US 153 (1976)�� 323 Hammer v Dagenhart 247 US 251 (1918)�� 143 Helvering v Davis 301 US 619 (1937)�� 226, 230, 234, 349, 376 Home Building & Loan Association v Blaisdell 290 US 398 (1934)�� 349 Jacobellis v Ohio 378 US 184 (1964)�� 3 Lawrence v Texas 539 US 558 (2003)�� 325, 327 Lichter v United States 334 US 742 (1948)�� 156 Lloyd Corp v Tanner 407 US 551 (1972)�� 324 Loving v Virginia 388 US 1 (1967)�� 326 Luther v Borden 48 US (7 How) 1 (1849)�� 136 Marbury v Madison 5 US (1 Cranch) 137 (1803)��75, 272, 274–75 Massachusetts v Environmental Protection Agency 549 US 497 (2007)�� 386, 400 Massachusetts v Mellon 262 US 447 (1923)�� 234, 350 McCulloch v Maryland 17 US (4 Wheat) 316 (1819)�� 150, 272 McDonald v Chicago (City of) 561 US 742 (2010)�� 322 Medellín v Texas 552 US 491 (2008)�� 427 Memorial Hospital v Maricopa County 415 US 250 (1974)��350–51, 379 Michigan v Long 463 US 1032 (1983)�� 325 Missouri v Holland 252 US 416 (1920)�� 272, 428 National Federation of Independent Business v Sebelius 567 US ___ (2012)�� 151, 234, 355 National Labor Relations Board v Canning 573 US ___ (2014)�� 198 National League of Cities v Usery 426 US 833 (1976)�� 273 Nebbia v New York 291 US 502 (1934)�� 349 New State Ice Co v Liebmann 285 US 262 (1932)�� 318 New York v United States 505 US 144 (1992)�� 273 Norman v Baltimore & Ohio Railroad Co 294 US 240 (1935)�� 349 Obergefell v Hodges 576 US ___ (2015)�� 326 Oregon Waste Systems Inc v Department of Environmental Quality of Oregon 511 US 93 (1994)�� 384 Pacific States Telephone & Telegraph Co v Oregon 23 US 118 (1912)�� 136 Pennhurst State School and Hospital v Halderman 451 US 1 (1981)�� 234 Pennsylvania v Nelson 350 US 497 (1956)�� 132 People v Anderson 493 P 2d 880, 6 Cal 3d 628 (Cal 1972)�� 324 Pollock v Farmers’ Loan & Trust Co 158 US 601 (1895)�� 216, 270

Table of Cases xxiii Printz v United States 521 US 898 (1997)�� 151 Pruneyard Shopping Center v Robins 447 US 74 (1980)�� 324 Roe v Wade 410 US 113 (1973)�� 323 Schuette v Coalition to Defend Affirmative Action 572 US ___ (2014)�� 266 Shapiro v Thompson 394 US 618 (1969)�� 350, 379 South Dakota v Dole 483 US 203 (1987)�� 234 Southern Pacific Co v Arizona 325 US 761 (1945)�� 143 Steward Machine Co v Davis 301 US 548 (1937)�� 349, 376 Swift v Tyson 41 US (16 Pet) 1 (1842)�� 161 Texas v New Mexico 482 US 124 (1987)�� 279 Texas v White 74 US (7 Wall) 700 (1869)��40, 90, 105–06 United States v Butler 297 US 1 (1936)�� 230, 349 United States v Leon 468 US 897 (1984)�� 323 United States v Lopez 514 US 549 (1995)�� 143, 273, 384, 398 United States v Morrison 529 US 598 (2000)�� 384 United States v Windsor 570 US ___ (2013)�� 326 West Coast Hotel Co v Parrish 300 US 379 (1937)�� 349 Wilson v Black-Bird Creek Marsh Co 27 US (2 Pet) 245 (1829)�� 142

xxiv

Introduction 0.1 WHY THIS BOOK AND WHAT TO EXPECT FROM IT La multitude qui ne se réduit pas à l’unité est confusion; l’unité qui ne dépend pas de la multitude est tyrannie.1

F

EDERALISM AND RELATED concepts are some of the most studied topics in social, political and legal sciences and beyond. Not only is federalism an essential component of theory and practice of states and power structures, but the federal principle is indeed much older than the modern and even the pre-modern state.2 In fact, federalism is at the heart of one of the most pressing challenges in the history of mankind: how to order public life and how to limit, organise and regulate power in a way that guarantees freedom and efficiency, unity and plurality, autonomy and coordination. For this reason, federalism is not in itself a form of government but rather its ideological underpinning. It is an idea—and an ideal. As such, it is a relative—not an absolute—concept, that might take different institutional manifestations,3 and its essence lies in society itself,4 the institutional structures being a product of societies and of the ideas developed therein. The federal idea and its institutional implications and manifestations have been broadly debated for centuries by scholars from different disciplines, with a clear prevalence of political philosophers and political scientists, but also by constitutional lawyers, macroeconomists and others. Since federalism is an unavoidable element of the theory of the state, it is widely taught in universities, and a bulk of manuals exist on this topic. So why one more handbook on federalism? Because an essential disciplinary approach is still largely missing in the literature: that of comparative constitutional law. Most of the existing books on comparative federalism take a decidedly political-philosophical approach. Furthermore, even those books that deal with federalism from an institutional perspective are mostly written by political scientists rather than by lawyers and, in general, neglect the role of the courts. The courts, however, shape the practical functioning of federal systems if not more at least not less than political actors.5 Finally, even when studies take into account the legal institutional

1 ‘Plurality which is not reduced to unity is confusion; unity which does not depend on plurality is tyranny.’ B Pascal, Pensées de M Pascal sur la religion et sur quelques autres sujets (Paris, Guillaume Desprez, 1670); translation of CW Eliot, The Harvard Classics: The Shelf of Fiction (New York, PF Collier and Son, 1917) Vol XLVIII. 2 See ch 3. 3 See ch 2. 4 See WA Livingston, ‘A Note on the Nature of Federalism’ (1952) 67 Political Science Quarterly 81. 5 See J Kincaid, ‘Comparative Observations’ in J Kincaid and GA Tarr (eds), Constitutional Origins, Structure, and Change in Federal Countries (Montreal, McGill–Queen’s University Press, 2005).

2 Introduction perspective, the role of the courts and even their case law, they usually approach the topic through individual case studies of countries, occasionally comparing two or more specific systems. And even the most brilliant analyses of specific aspects of federalism focus on countries rather than on functions.6 This book has as its ambition to look at the most relevant aspects of federalism from the perspective of functions, which is the typical area of comparative legal investigation. What are the central issues of, and the challenges for, federal governance? What are its main rules, how does it work in practice and why is it the way it is? How have such rules evolved in different countries, which ones became models that inspired other countries and why? Which ones remained isolated and why? What are the main policy fields in which federal governance can be tested, regarding both its existence and its impact? Based on concise explanations (and, where necessary, reinterpretations) of concepts, the comparative and issue-based approach of this book focuses on the practical meaning and utility of federalism in explaining and shaping constitutional realities and in determining the concrete governance of complex phenomena. Such focus, which can be called the concept-inspired practical approach of the book, is reflected, first, in the analysis of the interaction of different government levels, ie their powers and policies, regarding various highly topical areas such as environmental protection, immigration and integration, and social welfare and healthcare; second, it is reflected in the analysis of the case law of constitutional and supreme courts, which illustrates the practical impact of federalism on policymakers, civil servants and citizens. In order to highlight the importance of constitutional jurisprudence on the development of federalism, the book features text boxes with case law supposed to illustrate the arguments made in the main text. Being sometimes simultaneously understood as a concept, an ideal (and an ideology), a system of government and a set of institutions and instruments,7 federalism can, of course, be approached in many ways and from many disciplinary angles. This is precisely the reason why its definition is so controversial and ultimately impossible to agree upon, depending essentially on whether the focus is on the system of government (descriptive theories) or on its advocacy (normative theories).8 This book has rather a descriptive and explanatory purpose. It is not limited, however, to a description of how federations work; rather, it also considers the theoretical and ideological foundations underpinned (sometimes implicitly and even unconsciously) by the legal norms and case law that are at the core of the book. In other words, the book focuses on the toolbox of federalism, including its historical and conceptual embedment.

6 One may think, for example, of the Global Dialogue on Federalism (McGill–Queen’s University Press), a prestigious series edited by distinguished (legal and non-legal) scholars in the area of federalism with the support of the Forum of Federations and the International Association of Centres for Federal Studies (IACFS), dealing with fundamental issues such as division of powers, financial relations, local government, unity and diversity, and the like. See also AL Griffiths (ed), Handbook of Federal Countries, 2005 (Montreal, McGill–Queen’s University Press, 2005). 7 See ch 1. 8 An extremely useful collection of fundamental papers on federalism can be found in J Kincaid, Federalism (Los Angeles, Sage, 2011) Vol 1–4.

Understanding Federalism 3 0.2 UNDERSTANDING FEDERALISM—UNDERSTANDING OF FEDERALISM

Defining federalism and classifying federal states have kept scholars busy for centuries, filling libraries in the process. Nevertheless, there is no universal agreement on what federalism means,9 nor is there agreement on how to classify federal countries. Nor can there be one. In 1994, Daniel Elazar estimated that about half of around 180 sovereign states in the world were federal or had some form of federal arrangement,10 whereas, just a few years later, another leading scholar in the field, Ronald Watts, determined that only 23 fully fledged federal states existed in the world.11 What does then the term ‘federal’ mean? The answer obviously depends on the criteria one chooses, which entail a significant degree of subjectivity. The central question, however, is whether the question itself is meaningful. 0.A Is There a Bold (or a Bald) Definition of Federalism? Towards a Pragmatic Approach The underlying dilemma is reminiscent of the ‘bald man paradox’, as formulated by the Greek philosopher Eubulides in the fourth century BC. A man with a full head of hair is obviously not bald, and the removal of a single hair will not turn a non-bald man into a bald one. Yet it is obvious that a continuation of that process must eventually result in baldness. When can a man thus be considered bald? Can the difference between a bald and a non-bald man consist of just one hair? If yes, of which one? Since criteria are inevitably subjective and based on perception (the same person can be perceived as bald by someone and as not bald by others), the response, however grounded in scientific criteria, will always be open to controversy, although some common features undoubtedly exist.12 The alternative to a controversial theory is pragmatism, which is no less subjective but has the merit of not pretending to be universal or scientific. Pragmatism can be, and sometimes cannot but be, a criterion that can help get one out of an abstract definitional conundrum, as US Supreme Court Justice Potter Stewart bluntly proved when describing his threshold test for obscenity: ‘I shall not today attempt further to define the kinds of material I understand to be embraced within that shorthand description; and perhaps I could never succeed in intelligibly doing so. But I know it when I see it’.13 A more sophisticated form of pragmatism is a rather general and open definition, one that is broad enough to accommodate all different views. It is no surprise that the most quoted and perhaps the most appropriate definition of federalism is the one suggested by Elazar, according to which it is ‘self-rule plus shared rule’.14 Such a definition has the merit

9 See A Gamper, ‘A “Global Theory of Federalism”: The Nature and Challenges of a Federal State’ (2005) 6 German Law Journal 1297. 10 See DL Elazar, Federal Systems of the World: A Handbook of Federal, Confederal and Autonomy Arrangements, 2nd edn (Harlow, Longman Group Limited, 1994) XVI ff. 11 See RL Watts, Comparing Federal Systems, 2nd edn (Montreal, McGill–Queen’s University Press, 1999), and RL Watts, ‘Typologies of Federalism’ in J Loughlin et al (eds), Routledge Handbook of Regionalism and Federalism (Abingdon, Routledge, 2013). 12 See ch 2. 13 Jacobellis v Ohio 378 US 184 (1964). 14 DJ Elazar, Exploring Federalism (Tuscaloosa, University of Alabama Press, 1987) 12.

4 Introduction

of being sufficiently generic that no one can disagree, while at the same time providing a rough idea of what it is all about. In particular, such a generic definition embraces models of government that would not be classified as federal in the classical view put forward by most scholars.

While extremely useful in explaining concepts and narratives around federalism, the debate on classification of federal countries remains partisan and cannot lead to shared conclusions. It is not for this book to delve too much into this debate or into the theoretical (historical, political, ideological, philosophical) foundations of federalism, although one must be aware of them in order to contextualise and understand what all that is about. Therefore, the approach followed in this book is to give an account of the origins of the idea and of the most recurrent definitions of the concepts,15 and to focus on the toolbox of federalism, on how it works and why. For this reason, all experiences that combine self-rule and shared rule have to be considered. Not only is this the most correct comparative approach,16 but in each and every country (including those undoubtedly included in the category of federal countries in every classification), the degree to which self-rule and shared rule are combined is different. Distinctions among them are thus essentially quantitative, not qualitative.17 This book originates from the firm belief in the practical meaning of federalism in today’s world. While based on the universal value of making power better organised and more inclusive by accommodating territorial and, where necessary, cultural and other forms of pluralism, federalism is deeply contextual about how to achieve such an ideal goal. The comparative analysis of different instruments reveals the commonality of functions: how are subnational entities to be represented in the national decisionmaking process? How are financial relations between levels of government to be regulated? How are uniform rules to be combined with the necessity of treating various situations differently in areas such as fundamental rights, environmental protection, social welfare and healthcare and many others? How are conflicts between the levels of government to be prevented and resolved? How are legislative, administrative and judicial powers to be divided? How are specific territories or groups or nations to be accommodated within a common constitutional framework? How are local government and, where appropriate, the challenges of supranational integration to be dealt with? These are the fundamental questions, the responses to which are channelled through legal instruments that can be similar or different depending on the country, the history, the legal tradition, the economic situation and the political choices. Only a comparative and issue-oriented analysis of such instruments (institutions,

15 This is done especially in chs 1 and 2 and, where necessary, throughout the book when introducing concepts, institutions and rules. 16 See section 0.3 below. 17 See also ch 2.

Methodology and Terminology 5 procedures, rules of interpretation) makes it possible to understand the fundamental questions raised above and their possible solutions. 0.3 METHODOLOGY AND TERMINOLOGY

0.3.1 The Comparative Method The added value of a comparative study is to get to the heart of issues by identifying the legal instruments that have been developed in various countries or groups of countries, to test their practical impact (this is what makes special attention to case law and to policy analysis necessary) and, even more importantly, to check whether they have become transplants in other countries, be it due to historical, political, military, cultural or other influences. This way, common issues are more easily identified and the legal responses more effectively tested. Taking into account the historical legacies and the cultural specificities of each case, including in terms of legal tradition,18 a comparative legal approach suggests that while each case is different, problems are similar and, often, so are the instruments with which to tackle them and, more broadly, the functions that a complex polity19 is called upon to perform. Comparative work cannot just consist of describing how one single issue is regulated in a number of pre-selected countries. This is, of course, helpful in terms of providing information that might serve as background for a more profound analysis, but it is not the essence or the goal of comparison. This book takes what we believe is a more useful methodological approach: Instead of pre-determining the countries to be considered from a particular angle (be it some institution like second chambers, some procedure such as constitutional amendment, some substantive issue such as the division of powers or some specific policy field such as social welfare), the opposite is done. First, the item is identified and the issue described. Then, how it has been tackled in model countries is considered, meaning a) where the problem has been dealt with first, or b) in a way that has been exported to other countries, or c) in an original way that is peculiar to that particular case only. Finally, an explanation is given for each case and the rationale of each instrument is identified. This is why, while taking a global approach, some countries are quoted much more often than others. Understandably, the oldest and more established federations (also in terms of general performance of the country) such as the United States,

18 See T Fleiner and C Saunders, ‘Constitutions Embedded in Different Legal Systems’ in M Tushnet et al (eds), Routledge Handbook of Constitutional Law (New York, Routledge, 2013). 19 The term ‘polity’ originates in US political science and indicates the structural, formal and institutional dimension of a political system (as opposed to its procedural—politics—and content dimension— policy). See JG March and JP Olsen, ‘The New Institutionalism: Organizational Factors in Political Life’ (1984) 78 American Political Science Review 734, and G Almond and S Verba, The Civic Culture (Princeton, Princeton University Press, 1963). The terms ‘polity’ and ‘political system’ are used in this book interchangeably and do not necessarily refer, such as in the case of the EU, to states.

6 Introduction Switzerland and Germany, have somewhat more comparative weight than other countries. Also, much more recent federations (at least in a functional sense)20 such as Belgium, Spain and Italy often present model instruments in terms of how to deal with (asymmetric) devolutionary dynamics. In between classic aggregative and more recent devolutionary federations are several federations that emerged out of British colonialism, from Canada to Australia to India, and these are also quoted rather frequently due to their model function. Rules developed in individual countries that had comparative weight are also discussed, such as in the case of Austria especially for federal constitutional adjudication, or Bosnia and Herzegovina and Ethiopia with regard to multinational federalism. And, of course, the book analyses original constitutional or interpretative solutions that have been peculiarly developed in specific contexts, from South Africa to Malaysia, from Latin American countries to Russia, even from the United Kingdom (with its model of devolution) to a supranational polity like the European Union.21 The non-formalistic approach to federalism that characterises the book explains why less attention is given to countries that are nominally considered or consider themselves as federal but in which a) federalism is merely a paper principle (often due to authoritarian regimes or to violent conflicts), or which b) have developed no model or original solution, or where c) neither the management of specific policy fields nor constitutional adjudication presents features that are relevant in a comparative functional analysis. Doing the opposite, ie pre-identifying which countries are federal and which are not and then using only the supposedly federal countries for a transversal, symmetric juxtaposition, would have no comparative added value, as it would result in a sterile compilation and, more importantly, would eventually put theory before reality. Law is (or at least it should be) a practical discipline, which helps us understand phenomena and, based on such understanding, possibly resolve problems. Looking at the phenomenon of federalism from a legal point of view therefore means focusing on a functional and problem-oriented perspective. As mentioned above, the essential function of federalism is to guarantee institutional pluralism within a common constitutional framework, and this is why a broad understanding of federalism and an open approach to countries is followed in this book in order to identify the instruments and solutions in the federal toolkit that can be identified even in countries that would themselves refrain from identifying as federations. Key in a comparative perspective is the identification of issues and instruments that are recurrent in a number of countries or that are peculiar to just one country. Therefore, the identification of such issues is more important than, and might even prevail over, a classification provided not only in the literature but even by courts or by the constitution.

20

See section 0.3.2 below and ch 1. EU is treated in this book as a transversal issue which is, in one way or another, addressed in many chapters. 21 The

Methodology and Terminology 7

0.B How Comparison Can Read beyond Formalism For example, if the goal of the research is to look at how the interaction of a plurality of governments and decision-makers works, or at how cultural or linguistic diversity is accommodated, then this should be done by studying the countries that have developed instruments for that. In so doing, one realises that, for instance, the Spanish case might be extremely rich in this regard despite the fact that most scholars do not consider Spain a federal state and rather prefer other concepts (such as ‘regional state’ or ‘compound state’ or the category a ‘state of autonomies’)22 and that, in Spanish political discourse (like in that of many other countries), the term ‘federalism’ is extremely contentious and sensitive. Of course, there are (or there might be) differences between these concepts, and these need to be explained, but in the end it is essential to understand how the fundamental questions mentioned above are dealt with in that particular context and what each particular case can tell us in relation to the overall trends and dynamics of the governance of federal issues that are common to several other countries. For this reason, in a comparative perspective, definitions provided in judicial rulings or even in normative texts can also be challenged. For instance, the Italian Constitutional Court has expressly denied that Italy is a federal state.23 But this does not prevent the country from offering interesting examples of federal management of issues. The Swiss Constitution calls Switzerland a confederation, reflecting the historical origin of the name, but there is general agreement that, in fact, as confirmed by comparative analysis, the country is indeed a federation and not a confederation.24 Comparative analysis provides a wide range of similar examples. The Soviet Constitutions defined the USSR as federal, but many scholars argue that federalism is not really possible in non-democratic countries, and a similar argument has been made for Latin American federations during times of military dictatorship. Venezuela remains a peculiar case, even after the downfall of its military regime, as Article 4 of the 1999 Constitution defines the country as ‘a decentralised federal state’, a term completely unknown in the literature and actually bringing together two concepts that are usually considered at odds with one another.25 Furthermore, some federal countries call themselves ‘union’ rather than ‘federations’ (India26 and Comoros), some stress the union in their name (United States of America), others the federative element (Federated States of Micronesia) and others only the republican one (Republic of Trinidad and Tobago). Put simply, identification of a country’s federal nature goes far beyond naming conventions and definitions.

The purpose of comparative analysis is to look at phenomena in a critical way and to highlight certain specific cases instead of others because of the contribution of that particular country in providing relevant models from among the instruments that they have developed and experimented with. As Boris Mirkine-Guetzévitch, one of the founders of comparative constitutional law, once put it, ‘comparative constitutional analysis tells us how relative texts, words and dogmas are. Texts

22

See section 0.3.2 below. Italian Constitutional Court No 365/2007. 24 See chs 3.2.1 and 3.3.2. 25 See DJ Elazar, ‘Federalism vs. Decentralization: The Drift from Authenticity’ (1976) 6 Publius 9. 26 On the reasoning behind the use of the term ‘union’ in the Indian Constitution see ch 4.2.3. 23

8 Introduction t hemselves do not create democracy … Texts only create some conditions—for political evolution, transformation and realization.’27

0.3.2 Terminology The same non-formalistic and functional approach is taken in addressing the delicate issue of terminology. Each language, legal culture and country uses specific terms that can hardly be generalised. Identifying common terms to fit a whole range of different experiences is a slippery slope and must be done with extreme care, knowing that sometimes choices can be contested either on ideological grounds or because of the semantic variety of a specific term in a specific context. While difficult and at least in part subjective, there is no alternative to this exercise. It is essential, however, that this be done in a consistent way, whilst bearing in mind the manifold facets of the terms and the different ways they are understood in various contexts. This begins with the term federalism itself, which refers to an idea, an ideal and an ideology. There is general consensus in the literature that this concept is inherently different from its institutional manifestations.28 When only the latter are meant, our preferred term is federal systems. This notion includes, in our view, all those systems in which at least two political tiers of government exist, thereby combining self-rule and shared rule and thus making use (to a greater or lesser extent) of the federal toolkit. This means that the term federal system (or federal country) is used in this book not only to denominate fully fledged federal states (synonymously called federations). It also includes what are often referred to as regional or devolved states.29 The latter are federal systems in legal terms, as they are—like consolidated federal countries—constitutionally unitary states, subordinated to the national constitution and with constitutionally entrenched political autonomy of the subnational units. Notwithstanding the use of federal system and federal country as umbrella terms, the differentiated notions federal or regional or devolved states30 are used when specific countries are meant or when a distinction between these notions needs to be made. For example, even if both Spain and Germany are, using our terminology, federal systems, the former is a regional state and the latter a federal state.

27 B Mirkine-Guetzévitch, ‘Les méthodes d’étude du droit constitutionnel comparé’ (1949) 1 Revue international de droit comparé 396, 417 (author’s translation). 28 See, among others, Elazar (n 10); Watts, Comparing Federal Systems (1999); and M Burgess, Comparative Federalism: Theory and Practice (Abingdon, Routledge, 2006). See also ch 1.1.2. 29 In some contexts, like in Spain, other terms are also used, such as autonomic state (ie a ‘state [made] of autonomies’, especially indicating the peculiar Spanish experience) or compound state (again indicating a territorially plural polity ruled by a superordinate constitution). See, eg, E Aja, Estado autonómico y reforma federal (Alianza, Madrid, 2014). 30 The term ‘state’ is generally used in this book as a synonym for a country as a whole, and it only denominates a subnational entity when we write about a country with this specific terminology (eg states in the US, India, Nigeria, etc).

Methodology and Terminology 9

0.C Same Words, Different Meanings Interestingly enough, the underlying culture deeply influences the very meaning of the words used. For instance, in contexts such as the United States, Canada or Australia, federal generally means ‘central’, ‘national’ (eg federal government), whereas, in the German-speaking context, the adjective föderalistisch (eg föderalistische Gewaltenteilung) usually stresses the plurality of entities that constitute a unity, normally referred to as Bund. And in political terms, a federalist in the United States is normally associated with support for a larger role for the federal government, whereas in most devolutionary federations the same word has rather the opposite meaning, ie advocating more powers for subnational units.

While respecting the peculiar terminology of particular countries when one specific system is dealt with (eg using confederation and cantons in Switzerland, provinces in Canada and South Africa, autonomous communities in Spain, Länder in Germany and Austria, regions and autonomous provinces in Italy, state and entities in Bosnia and Herzegovina, communities and regions in Belgium, etc), we use the adjectives national (government) and subnational (entity, unit or government) as general terms, as is usually done in the Anglo-American context, being aware that sometimes the adjective national is read in different ways especially in the framework of nationalistic claims in some countries or in their units. In light of such claims, for example, the Canadian province of Quebec has, for the exercise of its autonomous law-making powers, an institution called National Assembly, even though this is, of course, a subnational legislature. Occasionally, the terms central and peripheral governments can also be used with the same meaning. Sometimes, units or constituent units indicate subnational entities, ie the levels of government immediately below the national one. While federalism implies a certain explicit or implicit hierarchy between national and subnational governments (this term also often being used in the American sense of ‘level’ rather than in the narrow sense of ‘cabinet’), words like levels, tiers, orders or spheres of government are used interchangeably.31 In the same way as for the general term ‘federalism’, scholars are divided on the potentially countless adjectives that can be attached to the concept. Trying to adhere to the most consolidated terminology (and explaining it whenever necessary), this books refers to ‘aggregative’ and ‘devolutionary’ federations (and sometimes to ‘coming-together’ and ‘holding-together’ federations) to indicate the historical process of the formation of the respective state and to ‘dual’ or ‘cooperative’ systems to reflect the most common ways to divide powers between spheres of government in a federal country. Further conceptual categories such as ‘symmetric’ or ‘asymmetric’, or ‘calm’ or ‘restless’ federal systems are also used and explained where appropriate.32 31 Sec 40(1) of the South African Constitution states that ‘[i]n the Republic, government is constituted as national, provincial and local spheres of government which are distinctive, interdependent and interrelated’. This is just but one further example of how terminology—even when constitutionally entrenched— does not necessarily reflect the reality. Despite the all-but-hierarchical expression ‘spheres of government’ laid down in the Constitution, South Africa is generally considered a rather centralised federation in which the hierarchy between the national and the provincial governments is quite acute. 32 See chs 2.2.2 and 2.5.

10 Introduction One of the essential and uncontested elements of federalism is that power be shared. Reflecting this basic rationale through consistent terminology is, however, a rather complex task. In terms of the theory of the state, especially in continental Europe where, traditionally, more attention has been paid to this aspect, the reference is often to divided sovereignty. The term sovereignty is used rather sparingly in this book and only when necessary to refer to specific theoretical issues.33 Since the focus is on instruments especially, it is preferable to use a terminology that more appropriately reflects the functional approach, although there is some difference in this respect between the Anglo-American tradition and the European terminology. In the former, terms such as powers or jurisdiction are normally used, whereas in the latter it is common to speak about competences, as reflected in the English version of the European Treaties, which use this term extensively. In this book, the terms are generally used as synonyms (especially division of powers or division of competences), unless there is a specific reason for not doing so. Other, less general terminological choices are explained in the individual chapters where necessary. 0.4 STRUCTURE

This book claims to be a modest step towards exploring new approaches to federal studies based on the acknowledgement of the new trends in federalism and on the deep conviction of the paramount importance of federalism in general, and of the federal toolkit in particular, in addressing the contemporary challenges of ever more complex governance. Part I (chapters 1–4) examines the foundations of federalism in terms of concepts (chapter 1), the manifestations through which these are realised (chapter 2), the historical evolution from ancient unions to archetypes of the federal state (chapter 3) and some of federalism’s most crucial debates (chapter 4). Part II explores the institutional and the procedural toolbox of federalism, looking at how powers and structures are divided and jointly exercised in federations (self-rule and shared rule). These include, in particular, the autonomy of subnational entities (chapter 5), the forms of subnational participation in national decision-making (chapter 6), financial relations (chapter 7), instruments for the prevention and resolution of conflicts (chapter 8) and the role and regulation of local government (chapter 9). Part III looks at some of the most relevant and complex powers and policies, fundamental rights (chapter 10), social welfare and healthcare (chapter 11), environment protection (chapter 12), immigration and migrant integration (chapter 13) and external relations (chapter 14), in order to examine how they are managed in federal countries. It is up to the readers to establish whether or not the purposes of the book have been achieved.

33

See especially ch 4.1.2.

Part I

Foundations

12

1 Concepts 1.1 FEDERALISM People are not likely to go to the stake, or the barricades, to defend federalism as such. They may undertake heroic actions for the sake of some value which federalism happens at the minute to favour, and may even then inscribe federalism on their banner—‘Liberty and Federalism’—‘Equality and Federalism’—but never just ‘Federalism’.1

A

S THIS SECTION’S brief overview of conceptual history illustrates, federalism has always been applied as an organisational principle in a pragmatic and flexible manner to serve quite different purposes.2 Precisely because of this mutability as an instrument to realise quite distinct values such as individual freedom, equality or solidarity, its essence is difficult to grasp. The fact that federalism is not easily tangible and lacks a widely accepted definition also results from its long history, which goes back to ancient times.3 Over the course of centuries, it has been seen as a polymorphic organisational principle underlying very different organisational forms, which sometimes go far beyond the legal–political realm. It is selfevident that a book focusing on the perspective of comparative constitutional law has its natural focus on federalism as embodied in state-related manifestations of this concept.4 Nonetheless, it is important to stress that federalism is much broader and is sometimes also seen as a principle of economic and social organisation. 1.A Proudhon’s Federalism beyond the Legal–Political Sphere The relevance of federalism as a principle of economic and social organisation has a long tradition that is usually traced back to the writings of Pierre-Joseph Proudhon in the nineteenth century.5 The French philosopher claimed that federalism would be futile if restricted to state organisation because the economic centralisation inherent to modern capitalism would render it ineffective. According to Proudhon, a federal state was only viable if complemented by an industrial–agricultural federation composed of autonomous enterprises, as well as associations of producers and consumers. Such a form of b ottom-up

1

G Sawer, Modern Federalism (New York, The New American Library, 1961) 153. See RL Watts, ‘Contemporary Views on Federalism’ in B De Villiers (ed), Evaluating Federal Systems (Dordrecht, Martinus Nijhoff Publishers, 1994) 7. 3 See ch 3.1. 4 For an overview, see ch 2. 5 See P-J Proudhon, ‘The Federal Principle’ in D Karmis and W Norman (eds), Theories of Federalism: A Reader (Basingstoke, Palgrave Macmillan, 2005). 2

14 Concepts

organisation would—unlike top-down Jacobinism—in his view, protect individual and collective freedom against a potentially oppressive state. Proudhon’s view of federalism later influenced anarchist movements and the make-up of the short-lived First Spanish Republic (1873–74). After World War II, Alexandre Marc, Denis de Rougemont and other proponents of ‘integral’ federalism and a federal Europe revived Proudhon’s ideas.6

Due to federalism’s polymorphic character, the principle has over time been a major topic of research for different academic disciplines. This holds true even if the focus is exclusively on state-related federalism, which is the subject of studies in constitutional law, political science and economics. A common understanding is difficult to reach not only across disciplines, but also among researchers from different geographical areas with distinct traditions in approaching federalism. Broadly speaking, we may identify, from a historical perspective, a typical Anglo-American and (continental) European approach. This presence of two main theoretical traditions clearly distinguishes federalism from regionalism, which is, as a concept, very much linked to Europe and, above all, to the European Union and its Member States.7 1.1.1 Traditions in Conceptualising Federalism In academic literature on federalism, there are quite different approaches to this concept. Without doubt, much of the literature, above all from the Anglo-American space, is linked directly or indirectly to a series of 85 newspaper articles that were published by Alexander Hamilton, John Jay and James Madison under the pen name Publius in the aftermath of the Philadelphia Constitutional Convention. This collection, called The Federalist, has had—beyond the immediate purpose of defending the constitutional draft of 1787 and, quite importantly, asserting a specific interpretation of it—immense theoretical and practical implications: ‘The Federalist converted federalism from an expedient into an article of faith, from an occasional accident of history into an enduring expression of the principles of constitutionalism.’8 The Federalist has subsequently served as the main point of reference for many scholars, among them Kenneth C Wheare. His seminal study, Federal Government,9 continues to play a key role in the English-language literature and reflects two typical characteristics of the Anglo-American understanding of federalism. These are the reduction of federalism contrary to its above-mentioned variety to its state-related dimension and, closely connected with this, the equation of ‘federalism’ with the state-related terms ‘federation’, ‘federal state’ and ‘federal government’. Such a reductionist view is most obvious in the example of Wheare, who defines the federal principle as the ‘method of dividing powers so that the general and regional governments

6

See L Roehmheld, Integral Federalism: A Model for Europe (Frankfurt, Peter Lang, 1990). See section 1.2 below. 8 C Rossiter (ed), The Federalist Papers (New York, The New American Library, 1961) xii. 9 See KC Wheare, Federal Government (New York, Oxford University Press, 1947). 7

Federalism 15 are each, within a sphere, co-ordinate and independent’.10 In this regard, he is not an exception, but rather stands in a long line of like-minded authors ranging from Albert V Dicey11 to contemporary authors such as Vernon Bogdanor. The latter strongly underlined that devolution must be distinguished from federalism, which, according to his view, ‘would divide supreme power between Parliament and the various provincial bodies’.12 Wheare’s definition has attracted both ardent support and fierce opposition, with the latter focusing primarily on three aspects: his emphasis on independence of national and subnational governments, a static understanding of federalism and an allegedly overly legalistic approach. As to the first point, several scholars have criticised the idea that a notion of strictly separated and independent governments would be unrealistic and instead emphasise their inevitable interdependence.13 Wheare’s perception of federalism as something static was challenged by Carl J Friedrich in particular, who claimed that attention should be paid to the dynamic process of federalising rather than to the final outcome of this process. In short, he introduced an understanding of federalism as a process.14 A third group of critics, spearheaded by William A Livingston and William H Riker, pointed to Wheare’s reputedly excessive legalism. By looking primarily at federal government, he would neglect the importance of federal societies and, as a result, disregard political realities. Livingston urged a reorientation of research from the pure form of federal constitutions to a sociological analysis of their actual operation in the context of federal societies. Such societies are, in his view, characterised by social cleavages like linguistic and religious diversity, communal traditions or territorial identities that have political salience, are territorially distributed throughout the country and crucial to its constitutional and political stability.15 Thus, according to Livingston, ‘the essence of federalism lies not in the institutional or constitutional structure but in the society itself. Federal government is a device by which the federal qualities of the society are articulated and protected.’16 Riker even goes further by considering federalism a ‘constitutional legal fiction which can be given whatever content seems to be appropriate at the moment’ and reduces its significance to being instrumental at the founding of a state. At that stage, the fiction would lure politicians from different parts of a country into accepting a national government: ‘Once the central government is actually in operation, however, what maintains or destroys local autonomy is not the more or less superficial feature of federalism but the more profound characteristics

10

ibid 11. refers to ‘the system of government known as Federalism’. See AV Dicey, Introduction to the Study of the Law of the Constitution (Indianapolis, Liberty Fund, 1982) 73. 12 V Bogdanor, Devolution (Oxford, Oxford University Press, 1979) 2. 13 See AH Birch, Federalism, Finance and Social Legislation in Canada, Australia and the United States (Oxford, Clarendon Press, 1955); DJ Elazar, The American Partnership: Intergovernmental Cooperation in the Nineteenth-Century United States (Chicago, Chicago University Press, 1962); M Grodzins, The American System: A New View of Government in the United States (Chicago, Rand McNally, 1966). 14 See CJ Friedrich, ‘Federal Constitutional Theory and Emergent Proposals’ in AW Macmahon (ed), Federalism: Mature and Emergent (New York, Russell & Russell, 1962) 528f. 15 See M Burgess, Comparative Federalism: Theory and Practice (Abingdon, Routledge, 2006) 140. 16 WA Livingston, ‘A Note on the Nature of Federalism’ (1952) 67 Political Science Quarterly 81, 84. 11 He

16 Concepts of the political culture’.17 Riker essentially construed federalism as a ‘constitutional bargain’ between future national and subnational government leaders to aggregate territory, increase tax revenue and facilitate army recruitment for the purpose of aggression or defence against an external threat.18 In hindsight, Livingston and Riker have been instrumental in increasing awareness of the fact that federal government always has to be seen as operating within a specific social and political context. The approach to defining federalism that has prevailed in German-speaking parts of Europe, essentially in the classical federations of Switzerland, Austria and Germany, differs significantly from the Anglo-American tradition. Most authors from these three countries reject the above-mentioned reductionist view of ‘federalism’ and refrain from equating it with ‘federal state’. At around the turn of the twentieth century, the German philosopher and politician Constantin Frantz had already regarded federalism as a multifaceted principle of organisation not only for the state but, more broadly, for society.19 This view ironically links the conservative Frantz with the above-mentioned anarchist Proudhon. Around the same time, the Austrian-German constitutionalist Georg Jellinek was also arguing the case for a clear distinction between federalism and the federal state. For him, federalism was an idea involving the advocacy of a balance between union and autonomy that would pervade federal states such as the German Empire and the United States.20 After Jellinek, this differentiation came to be widely accepted among constitutional law scholars in German-speaking countries.21 Most notably, in Germany, federalism has also become a major research subject for political scientists, which emphasises—similar to the above-mentioned studies of the US federation by Birch, Elazar and Grodzins—the interdependence of government levels.22 This interdependence, a typical feature of Germany’s cooperative federalism,23 was famously blamed by Fritz W Scharpf for creating a certain inertia in political decision-making. An excessive number of veto players, all caught in a ‘joint-decision trap’,24 would result in decisions being made at the lowest common denominator or being blocked altogether. Another example of the transatlantic flow of ideas concerns the concept of functional federalism. Building on earlier work by the US political economist Vincent Ostrom,25 several European scholars have also

17 WH Riker, ‘Six Books in Search of a Subject or Does Federalism Exist and Does It Matter?’ (1969) 2 Comparative Politics 135, 142 and 146. 18 WH Riker, Federalism: Origin, Operation, Significance (Boston, Little, Brown & Company, 1964) 1. 19 See C Frantz, Der Föderalismus als das leitende Prinzip für die soziale, staatliche und internationale Organisation, unter besonderer Bezugnahme auf Deutschland (Mainz, Kirchheim, 1879) v. 20 See G Jellinek, Allgemeine Staatslehre, 3rd edn (Berlin, O Häring, 1914) 785. 21 See, eg R Zippelius, Allgemeine Staatslehre, 13th edn (München, Beck, 1999) 389; F Ermacora, Allgemeine Staatslehre: Vom Nationalstaat zum Weltstaat (Berlin, Duncker & Humblot, 1970) 625; P Pernthaler, Allgemeine Staatslehre und Verfassungslehre, 2nd edn (Wien, Springer, 1996) 291. 22 See A Benz, ‘Intergovernmental Relations in the 1980s’ (1989) 19 Publius 203, 218f. 23 On dual and cooperative federalism, see chs 2.2.2 and 5.2.2. 24 FW Scharpf, ‘The Joint-Decision Trap: Lessons from German Federalism and European Integration’ (1988) 66 Public Administration 239, 239. 25 See V Ostrom et al, ‘The Organization of Government in Metropolitan Areas: A Theoretical Inquiry’ (1961) 55 American Political Science Review 831.

Federalism 17 started to argue in a similar manner that both the territorial scope and powers of entities at the supranational, national, regional and local level should be subjected to flexible change according to current needs. Varying concepts falling within the broader category of functional federalism have been developed in quite different academic environments, such as among Swiss economists26 and Italian constitutional lawyers.27 The above-mentioned examples of common research foci on the interdependence of government levels, as well as functional federalism, demonstrate a certain level of academic exchange between the Anglo-American and (continental) European spaces. Nonetheless, both areas have, as illustrated, quite distinct traditions in conceptualising federalism. Specific features of the typical Anglo-American view are the reduction of federalism to its state-related dimension and, thus, its frequent equation with ‘federation’, ‘federal state’ and ‘federal government’. However, the past three decades have arguably witnessed a tendency towards a convergence of views that were traditionally much further apart. 1.1.2 Federalism and Federation—Towards Common Ground Federalism has now become one of those good echo words that evoke a positive response but that may mean all things to all men … We see the term applied to almost any successful combination of unity with diversity, to almost any form of pluralism and cooperation within and among nations.28

Even after almost five decades, there is still quite some truth in this classical statement. First, it alludes to the fact that federalism often evokes positive connotations despite, or precisely because of, its conceptual vagueness. One should add here, however, that at times federalism also arouses negative reactions, especially in the context of ethno-culturally diverse societies, where some see it as a recipe for state disintegration.29 This is just the other side of the coin of polarised views on federalism, which do not see it in a pragmatic manner as a toolbox of institutional solutions.30 Second, even though the characteristic vagueness mentioned in Duchacek’s statement is still palpable, some progress has been made in the meantime towards a common understanding of federalism. This is not least the result of enhanced international communication and cooperation among scholars working in this field. In this regard, the publication in 1982 of Preston King’s book with the programmatic title Federalism and Federation was certainly a most significant event.

26 See BS Frey and R Eichenberger, The New Democratic Federalism for Europe: Functional, Overlapping and Competing Jurisdictions (Cheltenham, Edward Elgar, 1999). 27 See S Ortino, ‘Functional Federalism between Geopolitics and Geo-Economics’ in S Ortino et al (eds), The Changing Faces of Federalism: Institutional Reconfiguration in Europe from East to West (Manchester, Manchester University Press, 2005) 289f. 28 ID Duchacek, Comparative Federalism: The Territorial Dimension of Politics (New York, Holt, Rinehart and Winston, 1970) 191. 29 See ch 4.2. 30 See ch 0.2.

18 Concepts While he regarded ‘federalism’ in a normative way as an ideology, he took ‘federation’ to mean this ideology’s institutional manifestation and thus a descriptive term. As federalism is perceived as the driving force behind the establishment of a federation, ‘there may be federalism without federation’ but ‘no federation without some matching variety of federalism’.31 Either literally or with slight modifications, King’s fundamental distinction was received by many authors from the Anglo-American space, thus deviating from the previous equation of ‘federalism’ and ‘federation’. Ronald L Watts’ seminal book Comparing Federal Systems was most likely the most influential publication representing this new tradition. Building on King’s original dualist distinction, he not only differentiated ‘federalism’ from ‘federation’, but also from another, much broader descriptive term, that is, the ‘federal political system’.32 The latter umbrella category was intended, similar to what Elazar had called ‘varieties of federal arrangements’,33 to cover beyond federations sensu strictu also other manifestations of federalism under domestic law or even international law.34 Even though Michael Burgess is sceptical about ‘subordinating federation to a mere species of a genus that itself remains somewhat ambiguous’35 and reverts to a dualist view, he still shares with Elazar and Watts the now common distinction between a descriptive and normative term. Burgess defines ‘federation’ as ‘a particular kind of state’ and ‘federalism’ as ‘the recommendation (and sometimes) the active promotion of support for federation’.36 Among scholars, there seems to be today not only a broad consensus concerning this normative–descriptive distinction, but also a lowest common denominator regarding the meaning of federalism. Most see the essence of this concept best expressed in Elazar’s classical formula of ‘self-rule and shared rule’.37 In other words, federalism is about at least two entities being part of a larger union while at the same time enjoying autonomy. This widely recognised basic characterisation is implied by the origin of the concept in the Latin word foedus, which means compact. This etymological background often also serves as a point of departure for deriving both inherent commitments to loyalty,38 cooperation,39 peaceful conflict resolution40 and mutual aid41 so as to effectively realise the foedus. This stems from an understanding of ‘compact’ as indicating a partnership among co-equals instead of a mere contract following a rationale of quid pro quo.42 Likewise rooted in the etymological origin

31

P King, Federalism and Federation (London, Croom Helm, 1982) 76. Watts, Comparing Federal Systems, 3rd edn (Montreal, McGill–Queen’s University Press, 2008) 8f. 33 DJ Elazar, Federal Systems of the World: A Handbook of Federal, Confederal and Autonomy Arrangements (Harlow, Longman, 1991) xvi. 34 See ch 2. 35 Burgess, Comparative Federalism (2006) 48. 36 ibid 2. 37 DJ Elazar, Exploring Federalism (Tuscaloosa, University of Alabama Press, 1987) 5. 38 On principles of federal loyalty, see ch 8.1.1. 39 On institutions and instruments of intergovernmental cooperation, see ch 8.1.2 and 8.1.3. 40 See ch 8. 41 On financial solidarity, see ch 7.4. On solidarity regarding social welfare and healthcare, see ch 11. 42 See TO Hueglin and A Fenna, Comparative Federalism: A Systematic Inquiry (Peterborough, Broadview Press, 2006) 51; J Kincaid, ‘Introduction’ in AL Griffiths (ed), Handbook of Federal Countries, 2005 (Montreal, McGill–Queen’s University Press, 2005) 8. 32 R

Federalism 19 is the definition of federalism by means of structural characteristics, namely the element of union, coordination rather than subordination of multiple levels, integration of entities at the federal level and their autonomy following the principle of subsidiarity.43 It is more than doubtful, however, whether this ideal of federalism as a concept is ever reached in practice. To be sure, the above-mentioned structural characteristics may not pass for indispensable features that have to be present in all of federalism’s extremely diverse manifestations. For example, the coordinate status of multiple levels is, even in fully fledged federal states to a greater or lesser extent, compromised by supremacy and homogeneity clauses,44 which imply a certain degree of subordination of the subnational level of government. Another example is subsidiarity, which, rather than being a conditio sine qua non, seems to provide merely an additional value to federalism by framing the autonomy of entities along the lines of interest, suitability and efficiency.45 This realistic view is not to deny, of course, that federalism and subsidiarity have been closely intertwined throughout their conceptual history. This is epitomised most obviously by the early confluence of both concepts in Johannes Althusius’ model of a commonwealth comprising four different levels; namely families, cities, provinces and, finally, the commonwealth itself.46 This federal model is grounded in the idea that the lowest possible level should exercise autonomous powers insofar as it has an interest in this exercise and is most suitable for it. This idea is in line with its etymological origin in the Latin word subsidium, meaning reserve (troops), and was much later also reflected in, and spread by, the Catholic social doctrine. Based on Aristotle and Thomas Aquinas, the encyclical Quadragesimo Anno developed a conception of a social and political order based on the subsidiarity principle instead of on the previously advocated centralism.47 Today, this principle is certainly most closely associated with the European Union so that its contemporary meaning is primarily established by the treaties of Maastricht and Lisbon. However, the extent to which subsidiarity truly governs the functioning of the EU and its Member States, some of which have entrenched this principle in their national constitutions, still remains in doubt. 1.B Subsidiarity—Europe’s Organisational Principle? The Maastricht Treaty, concluded in 1992, entrenched the subsidiarity principle in the legal framework of the newly established European Union. Its Article 3b was formulated as follows: ‘In areas which do not fall within its exclusive competence, the Community shall take action, in accordance with the principle of subsidiarity, only if and in so far

43 See TO Hueglin, ‘Johannes Althusius: Medieval Constitutionalist or Modern Federalist?’ (1979) 9 Publius 9. 44 See ch 5.1.2. 45 See also A Gamper, ‘A “Global Theory of Federalism”: The Nature and Challenges of a Federal State’ (2005) 6 German Law Journal 1297, 1300. 46 See ch 4.1.1. 47 See Pius XI, On the Reconstruction of the Social Order (Quadragesimo Anno) (New York, The America Press, 1938).

20 Concepts

as the objectives of the proposed action cannot be sufficiently achieved by the Member States and can therefore, by reason of the scale or effects of the proposed action, be better achieved by the Community.’ Even though this provision was initially praised as effectively realising the subsidiarity principle, Member States have been rather reluctant to invoke it before the European Court of Justice (ECJ), now the Court of Justice of the European Union (CJEU).48 Another factor limiting the provision’s potential is the discretion granted by the ECJ, which was only willing to overturn European measures in the rare case of a ‘manifest error’ of assessment. Furthermore, some proponents of the subsidiarity principle have overlooked that Article 3b is only concerned in an entirely dualist manner with relations between the competences of the EU and its Member States without even mentioning the regions. At least in Germany, whose Länder had campaigned for subsidiarity within the EC throughout the 1980s, Maastricht has had a more far-reaching impact because it fuelled a debate as to whether subsidiarity would be inherent in some of the provisions of the 1949 Basic Law. This debate eventually gave rise to the adoption of Article 23(1) of the Basic Law, which has been interpreted as an obligation to strive for the realisation of subsidiarity throughout Europe, including also regions and municipalities. Article 5(3) of the 2007 Lisbon Treaty on European Union (TEU), which amended the Maastricht Treaty,49 almost literally repeats the latter’s Article 3b on subsidiarity. Nonetheless, it still introduced some changes. First, it makes, regarding the capacity or incapacity of the Member States to achieve the objectives of the proposed action, an explicit reference to this achievement ‘either at central level or at regional and local level’. Second, it outlines in Protocol No 2 a procedure that makes the national parliaments the custodians of subsidiarity. Article 6 of this Protocol foresees that within six weeks from the publication of a draft European legislative act any chamber of a national parliament may issue a reasoned opinion on the alleged non-compliance with the principle of subsidiarity. In this process, it is for the respective chamber to decide whether ‘to consult, where appropriate, regional parliaments with legislative powers’ so that the latter do not have a mandatory advisory role. Moreover, where reasoned opinions represent at least one-third of all votes of the national parliaments, the draft must be reviewed, but it may eventually be maintained, provided that reasons are given for this decision (Article 7). 1.2 REGIONALISM [R]egion is an elusive concept, covering a variety of territorial levels and a range of social contents. A minimal definition would present it as an intermediate territorial level, between the state and the locality.50

While it is clear that ‘region’ is derived from the Latin word regio, meaning line, boundary, as well as the territory between boundaries, the contemporary understanding 48 The Court of Justice of the European Union (CJEU) denominates, since the entry into force of the 2007 Lisbon Treaty, a court system that comprises three individual courts: the Court of Justice established in 1952, the General Court set up in 1988 and the Civil Service Tribunal created in 2004. 49 The Treaty of Lisbon amended the two treaties that together form the constitutional foundations of the EU: the Treaty on European Union (TEU), originally adopted in Maastricht in 1992, and the Treaty on the Functioning of the European Union (TFEU), initially signed in Rome in 1958 as the Treaty establishing the European Economic Community. 50 M Keating, The New Regionalism in Western Europe: Territorial Restructuring and Political Change (Cheltenham, Edward Elgar, 1998) 9.

Regionalism 21 of this term and the concept of regionalism is not easy to grasp. It is evident from the etymological origin that a region implies both territorialism and diversity, as this geographical area between boundaries must be in some way distinct and distinguishable from other regions.51 These criteria of differentiation may then vary substantially so that different academic disciplines arrive at radically different perceptions of these distinct areas as historical, cultural, economic or legal–political regions. Regardless of the determining criterion in each individual case, the ‘region is not a natural entity, but a social construction’.52 Territories that are in some way distinct but not legal–political regions exist, of course, regardless of whether or not they have legal recognition. And vice versa, legal recognition does not necessarily coincide with, or give rise to, regions that are distinguishable according to any of the other above-mentioned criteria. A case in point is regionalisation of Central and Eastern European countries, which has been perceived in many cases as nothing more than a precondition for accession to the EU and, more specifically, to its regional policy.53 In the absence of endogenous initiative and little, if any, recent tradition of regions, it comes as no surprise that this process has largely failed to create strong regions in legal and political terms.54 Interestingly, the understanding of the term ‘region’ in a legal–political sense may be traced back to, of all countries, France, with its Jacobin centralist tradition. During the time of the Third Republic (1870–1940), political movements in Brittany and Provence demanded the establishment of régions to deal with subnational affairs. Whereas these efforts were eventually fruitless, the first constitutional entrenchment of ‘regions’ dates back to the Spanish Constitution of 1931. According to its Article 8, the so-called Estado Integral was to include the municipalities, provinces and regions (regiones) as government levels.55 Today, the term ‘region’ has constitutional relevance in some countries but is used in very diverse ways. For instance, in Belgium, it denominates one of two categories of subnational entities (Article 3); in Iraq the only autonomous subnational unit, Kurdistan (Article 113); and in India autonomous territories in the tribal areas of only a few subnational entities, ie the states of Assam, Meghalaya, Tripura and Mizoram (Article 1(2) of the Sixth

51 See A Gamper, ‘Regions and Regionalism(s): An Introduction’ in GM Grabher and U Mathis-Moser (eds), Regionalism(s): A Variety of Perspectives from Europe and the Americas (Wien, New Academic Press, 2014) 3ff. 52 M Keating, ‘The Invention of Regions: Political Restructuring and Territorial Government in Western Europe’ (1997) 15 Environment and Planning 383, 390. 53 See M Brusis, ‘Between EU Requirements, Competitive Politics, and National Traditions: Recreating Regions in the Accession Countries of Central and Eastern Europe’ (2002) 15 Governance: An International Journal of Policy, Administration, and Institutions 531. On regional policy, see sections 1.2.1 and 1.2.2 below. 54 See F Palermo, ‘Concluding Remarks—New Regionalism in Central, Eastern and South-Eastern Europe: Traditional Models and Beyond’ in F Palermo and S Parolari (eds), Regional Dynamics in Central and Eastern Europe: New Approaches to Decentralization (Leiden, Brill, Martinus Nijhoff, 2013) 246. 55 See T Wiedmann, Idee und Gestalt der Regionen in Europa: Rechtsvergleichende Untersuchung zu Unitarismus und Föderalismus, unter besonderer Berücksichtigung des Vereinigten Königreichs, Frankreichs, Spaniens und Deutschlands (Baden-Baden, Nomos, 1996) 23.

22 Concepts Schedule). From a constitutional viewpoint, the term ‘region’ may thus have manifold meanings and different relations to the subnational entities of federal systems. Even if legal scholars do not use the vocabulary of ‘region’ and ‘regionalism’ so frequently, they have become increasingly interested in regionalism as a concept, developed rather in other disciplines like political science, geography or economics, and particularly in its relation to federalism.56 1.2.1 Regionalisation vs Regionalism Most accounts of regionalism differentiate between a bottom-up and a top-down dimension, with the latter often being referred to as regionalisation. Whereas bottom-up regionalism claims the distinctiveness of a territory and advocates on this basis for a status of territorial autonomy, regionalisation means the granting of such autonomy or territorial restructuring merely for the delivery of regional development policies. This definition does not say anything, of course, about the scope of territorial autonomy. But most scholars tend to exclude a status of merely administrative autonomy and claim that genuine regionalism requires legislative powers exercised by an elected assembly, as only such a status would put a territory in a position to pursue its own policies reflecting its own priorities.57 As to its origins, there is considerable consensus that the concept of regionalism may be traced back to the late nineteenth century,58 which obviously makes it a much more recent concept than federalism. At that time, regionalism, like in the above-mentioned cases of Brittany and Provence, was a response to the consolidation and expansion of the nation state with its project of economic, social and cultural homogenisation. If we leave this first phase aside, the later history of the concept in the post-World War II era is characterised by an overall tendency of topdown regionalisation being first the sole dominant approach, before being complemented in the 1970s by calls for stronger bottom-up regionalism.59 Throughout the 1950s and 1960s, an approach prevailed that regarded regionalisation for the sake of regional development policies, practised extensively, above all in France, as a means of promoting even territorial development across the country in question. These policies, which primarily relied on tax incentives, financial transfers and infrastructure investments, were initially clearly technocratic and top-down in nature. In respect to the latter, they were not dissimilar from the more recent example of decentralised Special Economic Zones in China, which have been established since the 1980s primarily to attract foreign capital with more market-oriented

56 See especially A Gamper, Die Regionen mit Gesetzgebungshoheit: Eine rechtsvergleichende Untersuchung zu Föderalismus und Regionalismus in Europa (Frankfurt, Peter Lang, 2004) 69ff; Palermo and Parolari, Regional Dynamics (2013). 57 See J Loughlin, ‘Introduction: The Transformation of the Democratic State in Western Europe’ in J Loughlin (ed), Subnational Democracy in the European Union: Challenges and Opportunities (Oxford, Oxford University Press, 2001) 14. 58 See Keating, ‘The Invention of Regions’ (1997) 384. 59 See J Loughlin, ‘Reconfiguring the Nation-state: Hybridity vs Uniformity’ in J Loughlin et al (eds), The Routledge Handbook of Regionalism and Federalism (Abingdon, Routledge, 2013) 12f.

Regionalism 23 policies and governmental flexibility.60 In post-war Europe, political conflict concerning policy directions and the need for efficient implementation soon attenuated the technocratic top-down approach and prompted national governments to create specialised development agencies involving both national and regional representatives. This eventually gave rise, in some countries, like France, to regions with directly elected institutions. Around 1970, this regionalisation approach was complemented by growing bottom-up regionalism, which started to take root in Europe through what has been aptly termed the continent’s ‘regional revolution’,61 albeit, of course, only in its Western part. These processes put the position of regions forcefully on the political and, frequently, constitutional agenda and triggered a number of reforms or at least reform attempts. The heyday of regionalism in the 1970s witnessed the following events in particular:62 in Italy, the transfer of powers to the ordinary regions; in the United Kingdom, devolution projects for Scotland and Wales, albeit both repealed after post-legislative referendums in 1979;63 in Belgium, the start of, so far, six consecutive state reforms (1970–2011); in Spain, the adoption in 1978 of the constitutional framework for the state of autonomies (Estado de las autonomías); and, finally, even in France, the incremental process of transforming the above-mentioned regions established for territorial development into fully fledged territorial entities with direct democratic legitimation (collectivités territoriales) that culminated in the 1982 decentralisation reform.64 Beyond these legal reforms in a number of countries, the rise of regionalism in the 1970s was also epitomised by the increasing Europe-wide interconnectedness of subnational entities. An evident example is the replacement in 1975 of the Council of Europe Conference of Local Authorities with the Congress of Local and Regional Authorities. Moreover, the Council of Europe issued in 1975 and 1978, respectively, the declarations of Galway and Bordeaux, which addressed the increasingly topical issue of regionalism. Around 1990, regions came to the fore once again in the context of key events surrounding tighter European integration,65 not least because the above-mentioned legal reforms and forms of interregional cooperation also had repercussions for this process. While back in the 1950s Germany was the only federal country within the European Communities, Austria joined the EU in 1995 as a federal state and several original members like Belgium, Spain and Italy have been transformed during their

60 See C Carter, ‘A Tale of Two Chinese SEZs: From Exogenous to Sustainable Endogenous Growth?’ in C Carter and A Harding (eds), Special Economic Zones in Asian Market Economies (New York, Routledge, 2010). 61 J Hopkins, Devolution in Context: Regional Federal and Devolved Government in the European Union (London, Cavendish, 2002) 39. 62 For an overview, see Y Mény (ed), Dix ans de regionalisation en Europe: Bilan et perspectives 1970–1980 (Paris, Cujas, 1982). 63 In Wales, only 20.3% voted for devolution, while 51.6% did so in Scotland. As the latter figure represented only 32.9% of the registered electorate, it failed to reach the required threshold of 40%. 64 See X Philippe, ‘France: The Amendment of the French Constitution “on the Decentralized Organization of the Republic’’’ (2004) 2 International Journal for Constitutional Law 691, 692ff. 65 See S Bartolini, Restructuring Europe: Centre Formation, System Building, and Political Structuring between the Nation State and the European Union (Oxford, Oxford University Press, 2005) 257ff.

24 Concepts membership to the EU into federal or regional states. Joint lobbying activities resulted in the adoption in 1988 by the European Parliament of the Community Charter for Regionalisation, which demanded sufficient legislative and executive powers for regions to regulate explicitly enumerated matters.66 Beyond that, the same year marked, more famously, a watershed in the history of EC/EU regional policy, which was supposed to be managed from then on jointly by multiple levels of government. 1.C EC/EU Regional Policy and the Partnership Principle The late-1980s reform of the EC/EU regional policy had two main impacts. First, it vastly increased the endowments of the two funds that were the financial levers to realise this policy, namely the European Social Fund (ESF) and the European Regional Development Fund (ERDF). This massive effort contributed to a development that saw the previously domestic regional policies undergo a process of comprehensive Europeanisation.67 Second, the 1988 reform has been considered crucial because it declared the partnership principle to be one of four new cornerstones of this increasingly important policy.68 Both programming and implementation would now become joint tasks of partnerships with obligatory consultation between the representatives of the national, regional and local levels, as well as the Commission.

However, the regions that the joint EC/EU regional policy was to be based on, that is, the three-tier scheme of the Nomenclature des unités territoriales statistiques (NUTS), merely served the administration of funding. Unsurprisingly, they have thus hardly served as a nucleus for their evolution into strong legal–political regions. The fact that the NUTS classification relies in principle on pre-existing subnational entities, is linked to the EU’s ostensible ‘regional blindness’, as expressed in its indifference, according to Article 4(2) TEU, towards the internal territorial and constitutional arrangements of its Member States. The classification does so, however, by mixing various types of regions with very different domestic constitutional statuses.69 Where pre-existing subnational entities are absent, the EU remains only indifferent in legal terms, but uses forceful political incentives, that is, the financial benefits from its regional policy to press Member States to create regions, at least of some formal kind. This is in line with the more general assessment that EU blindness towards internal arrangements is a rather ‘deceptive notion’.70 Another important event for

66

Community Charter for Regionalization [1988] OJ C 326/296. See R Leonardi, ‘Cohesion in the European Union’ (2006) 40 Regional Studies 155. 68 The other three are the principles of programming, concentration and additionality, with the first two aimed at preventing excessive dispersion of funds and the last one requiring Member States to provide matching funds. 69 While, for instance, German Länder are in the NUTS I category, Spanish autonomous communities and Austrian Länder are in NUTS II. Their supposed German equivalents in this category are the government districts (Regierungsbezirke), which have purely administrative functions. 70 S Weatherill, ‘The Challenge of the Regional Dimension in the European Union’ in S Weatherill and U Bernitz (eds), The Role of Regions and Sub-National Actors in Europe (Oxford, Hart Publishing, 2005) 30. 67

Regionalism 25 European regions was without doubt the conclusion of the Maastricht Treaty in 1992, a moment aptly described as ‘the passage from functional to institutional regionalism’.71 The first pillar of institutional reform was the newly introduced possibility for a member of a regional government to head a delegation and to vote in meetings of the Council of the European Union. In this regard, Article 16.2 TEU today foresees that the ‘Council shall consist of a representative of each Member State at ministerial level, who may commit the government of the Member State in question and cast its vote’72 and thus merely requires that certain rules concerning representation be in place in national law.73 Maastricht’s second institutional novelty was the Committee of the Regions, consultation with which was made obligatory regarding certain matters with repercussions on the regional or local level and optional in others. 1.2.2 New Regionalism, Multilevel Governance and the ‘Europe of the Regions’ If it is true that post-World War II developments have been characterised by a trend from top-down regionalisation focused on economic development to more comprehensive bottom-up regionalism, there remains the question of how these have been conceptualised. New Regionalism, a concept popularised since the 1980s, holds that trends towards regionalism are the result of the exigencies of globalisation and integration within organisations such as the North American Free Trade Agreement (NAFTA), the Mercado Común del Sur (Mercosur) and, above all, the EU. The fact that this new economic context has substantially diminished the capacities of national governments to effectively realise top-down regional development policies like in the early post-World War II era, would have forced regions to become more self-sufficient and, indeed, autonomous. They would have been from then on no longer complementary to one another within an essentially national framework of regional development, but, in the context of competitive regionalism, rivals in a global market. Arguing from a normative perspective, ‘new regionalists’ also claim that responses to these changed circumstances are most effective on a regional scale. 1.D New Regionalism—Is it All About the Economy? Many proponents of New Regionalism have seen the above-mentioned changes merely and exclusively as functions of a newly emerged economic context. As against such monist interpretations of this concept, arguably more appropriate pluralist accounts also emphasise the importance of other causes for the emergence of regions.74 Among these are, for

71 B de Witte, ‘Regional Autonomy, Cultural Diversity and European Integration’ in Ortino et al, The Changing Faces (2005) 207. 72 For details regarding this reform, see J Woelk, ‘A Place at the Window: Regional Ministers in the Council’ in R Toniatti et al (eds), An Ever More Complex Union: The Regional Variable as a Missing Link in the EU Constitution? (Baden-Baden, Nomos, 2004). 73 Regarding such rules in Belgium and Spain, see ch 14.4 and 14.5. 74 See especially Keating, The New Regionalism (1998).

26 Concepts

instance, claims to institutional efficiency, political responsiveness to bottom-up mobilisation around regionalist demands, increasingly complex policy challenges requiring the involvement of multiple levels and, in cases like Catalonia, Flanders or Scotland, the confluence of regionalism and nationalism. In short, they reject economic determinism and perceive the ‘regional question as a political question’.75

Whereas the scope of New Regionalism goes beyond Europe,76 other concepts are less global and, indeed, focused on regionalism within the European Union. One example is the concept of multilevel governance (MLG) which was, at its origins, inextricably linked to EU regional policy in the aftermath of the above-mentioned 1988 reform. This field has, despite its subsequent extension to the functioning of EU governance in general, remained its ‘home ground’ and main area of reference. In short, the concept of MLG holds that the functioning of the EU defies the traditional dichotomy between domestic and international politics and is instead characterised by ‘a system of continuous negotiation among nested governments at several territorial tiers’.77 Further reforms of regional policy78 and in other fields went on to also include what Jacques Delors famously called ‘les forces vives’,79 that is, actors from civil society. As a consequence, MLG today has not only a vertical but also a horizontal dimension, with private actors involved at all government levels. Terminologically, ‘multilevel’ thus captures the increasing interdependence of governments, while ‘governance’ reflects the growing interdependence of the latter with private actors. While scholars sometimes make normative arguments about MLG by claiming its superiority over other concepts of EU governance,80 its main uses are that of a political buzzword, especially at the EU level,81 and of an analytical tool. In the latter context, it is quite controversial whether, in regional policy and even more in other fields, there is only involvement of multiple actors, ie multilevel participation in a formal sense, or involvement plus real influence on the part of these actors, ie effective multilevel governance.82 In hindsight, there is much evidence for regional

75 J Agnew, ‘The “New Regionalism” and the Politics of the Regional Question’ in Loughlin et al, The Routledge Handbook (2013) 134. 76 See, eg, AJ Scott, Regions and the World Economy: The Coming Shape of Global Production, Competition, and Political Order (Oxford, Oxford University Press, 1998). 77 G Marks, ‘Structural Policy and Multi-Level Governance in the EC’ in AW Cafruny and GG Rosenthal (eds), The State of the European Community Vol 2: The Maastricht Debates and Beyond (Boulder, Lynne Rienner, 1993) 392. 78 These reforms increased the diversity of actors in these partnerships by including social and economic partners (1993) as well as environmental agencies and other non-governmental actors (1999). 79 J Delors, ‘Speech 8 July 1991’ in Commission of the European Communities (ed), La reconversion des régions industrielles: rencontre des 60 régions éligibles à l’objectif 2 des fonds structurels (Brussels, CEC, 1991). 80 See especially L Hooghe and G Marks, ‘Optimality and Authority: A Critique of Neo-Classical Theory’ (2000) 38 Journal of Common Market Studies 795. 81 The notion soon entered the vocabulary of politicians, as well as of civil servants. See, eg Commission of the European Communities, ‘European Governance: A White Paper’ COM (2001) 428 final. 82 On this important distinction, see I Bache, Europeanization and Multi-Level Governance: Cohesion Policy in the European Union and Britain (Lanham, Rowman & Littlefield, 2008) 162.

Regionalism 27 policy that a strong role for regions has depended decisively on the pre-existing constitutional empowerment of these entities and that the post-1988 functioning of this policy has not per se facilitated a development ‘from NUTS to constitutional regions’.83 As opposed to MLG’s obvious relevance for regions within the EU, the concept of multilevel constitutionalism focuses, of course, from a constitutionalist perspective, on the interaction merely between EU and national sources of law in the postMaastricht framework and is thus essentially dualist.84 It is revealing that the term ‘regional’ matters for this concept only to denominate the tighter integration of Member States within Europe as a region.85 Another EU-related notion, that of the ‘Europe of the regions’,86 refers to the repercussions of tighter European integration on the status of regions in the new institutional framework. Owing to the above-mentioned increasing role of regions in the EU’s functioning and institutional framework, the initial European–national dualism has given way since the late-1980s to a more complex, triangular relationship.87 It is certainly true that the notion ‘Europe of the regions’ appeared in quite different contexts and soon evolved into a political slogan used by regionalists and, for a rather short time, by the European Commission. But what holds the ‘Europe of the regions’ together as a concept, albeit a very malleable one, is its convergence on the claim that regions have evolved into a genuine ‘third level of government’.88 This differs from the original perception of this notion in the 1960s insofar as it was no longer embedded in a comprehensive federal vision of Europe.89 At the beginning of the new Millennium, however, the hype of the 1990s concerning a ‘Europe of the regions’ quickly began to give way to growing disillusionment in many respects. First, effective lobbying in EU affairs came to be seen as limited only to some economically strong and politically well-positioned regions.90 Second, EU 83 F Palermo and A Santini, ‘From NUTS to Constitutional Regions: Addressing EU Regions in the EU Framework’ in Toniatti et al, An Ever More (2004). The crucial role of the pre-existing state structure was already emphasised in J Kelleher et al, The Thematic Evaluation of the Partnership Principle: Final Synthesis Report (London, The Tavistock Institute, 1999). 84 Whereas the German Federal Constitutional Court famously characterised the new supranational framework in state-derived terms as a Staatenverbund (on the seminal ‘Maastricht judgment’, see box 4.B), multilevel constitutionalism drew an explicit contrast to this position by proposing, in the original German-language publications of Ingolf Pernice, the notion of the Verfassungsverbund. Today, this original concept and multilevel constitutionalism sensu lato coexist. See N Walker, ‘Multilevel Constitutionalism: Looking Beyond the German Debate’ (2009) 8 LSE, Europe in Question Discussion Paper Series 1, 1ff. 85 See I Pernice, ‘Constitutional Law Implications for a State Participating in a Process of Regional Integration: German Constitution and Multilevel Constitutionalism’ in E Riedel (ed), German Reports on Public Law Presented to the XV International Congress on Comparative Law (Baden-Baden, Nomos, 1998) 40. 86 M Keating, ‘A Quarter Century of the Europe of the Regions’ (2008) 18 Regional & Federal Studies 629, 630. The expressions ‘Europe and the Regions’ and ‘Europe with the Regions’ are sometimes used with similar meanings. 87 See especially B Jones and M Keating (eds), The European Union and the Regions (Oxford, Oxford University Press, 1995). 88 See, eg, U Bullmann, ‘The Politics of the Third Level’ (1996) 6 Regional & Federal Studies 3. 89 See, eg, D de Rougement, The Idea of Europe (New York, Collier-Macmillan, 1996). 90 See C Jeffery, ‘Sub-national Mobilization and European Integration’ (2000) 38 Journal of Common Market Studies 1.

28 Concepts regional policy was increasingly perceived as failing to empower regions in general. Many had before arguably fallen into a ‘descriptive trap’. As the structural funds have ‘regional’ written all over them, observers have too often jumped to the conclusion that they are about regional devolution and politics, and even represent a strategy on the part of the Commission to bypass the member states and enlist allies at the sub-state level to create a Europe of the Regions.91

In truth, however, the rationale behind the 1988 reform of regional policy, that is, the effective integration of new members Greece, Portugal and Spain and marketcorrecting measures to counterbalance the supposed negative impact of the single market92 on already marginalised regions,93 had hardly anything to do with the political strengthening of regions. Finally, the Committee of the Regions has remained far from the ‘European Senate of Regions’ as a ‘third chamber’ in the legislative process, which subnational entities had envisioned and campaigned for during the 1980s.94 Today still, after the Lisbon Treaty, it only assists the main institutions ‘in an advisory capacity’ (Article 13(4) TEU). Similar to the Council of Europe Congress of Local and Regional Authorities (CLRAE), which had evolved from the above-mentioned Conference of the same name in 1994, there is also a more structural problem. As the Committee of the Regions hosts both regional and local representatives, it tends to become entangled in diverging or even irreconcilable interests. But even the regional representatives alone often have little in common, as the domestic constitutional status of their respective territories varies so enormously, so does their capacity to uphold their interests in Brussels individually.95 More generally, the fact that a ‘Europe of the regions’ has hardly materialised is therefore also caused by the immense diversity of European regions in terms of both legal status and political clout. In fact, the much invoked ‘third level’ is, to the limited extent that it exists, ‘homogenous in neither composition nor degree of influence.’96 This asymmetry and the consequences that it entails are probably best illustrated by two developments. The first one is the formation in 2000 of an alliance beyond the Committee of the Regions, namely the Conference of European Regions with Legislative Power (REGLEG). This alliance primarily aims at increasing the influence of its members in the EU legislative process following the principle of subsidiarity. It was initiated when, in preparation of the Treaties of Amsterdam and Nice, signed in 1997 and 2001 respectively, stronger regions had become increasingly disillusioned with the limited success of the Committee of the Regions in pushing the

91

Keating, ‘A Quarter Century’ (2008) 630. 1986, the Single European Act had set an objective of establishing a single market by 31 December 1992. 93 Commission of the European Communities. ‘Completing the Internal Market: White Paper from the Commission to the Council’ COM (1985) 310 final. 94 de Witte, ‘Regional Autonomy’ (2005) 207. 95 On representation offices of the regions in Brussels, see ch 14.4 and 14.5. 96 Weatherill, ‘The Challenge’ (2005) 15. 92 In

Regionalism 29 regional agenda.97 REGLEG clearly differs from the latter regarding its status, as it is not a formal institution of the EU but an informal political network. Perhaps even more importantly, it is also distinct due to its more exclusive character as a club comprising only subnational entities with law-making powers.98 A second development that reflects the diversity of European regions is the troubled history of the draft European Charter of Regional Self-Government.99 This charter was proposed by the CLRAE in 1997 with the backing of the Committee of the Regions. 1.E Is Europe Heading towards Regional Democracy? As early as 1985, the European Charter of Local Self-Government was adopted.100 It was the first internationally binding treaty to enshrine the principle of subsidiarity and it has been ratified in the meantime by all members of the Council of Europe. To effectively implement this principle by closing the gap between the local and national levels, a complementary draft European Charter of Regional Self-Government was issued in 1997. The idea was thus to go beyond Article 13 of the 1985 Charter, according to which each party may ‘specify the categories of local or regional authorities to which it intends to confine the scope of the Charter or which it intends to exclude from its scope’. Instead of as a second tier of local government, regions were now to be recognised as a distinct level of government in their own right. But unlike its very successful predecessor regarding the local level, the 1997 draft failed from the beginning to gain broad support. Controversial debates soon revolved in particular around the very concept of self-government, as autonomy, including fiscal and legislative powers instead of purely administrative functions, proved to be unacceptable for many states, especially from Central and Eastern Europe. In 2008, the CLRAE responded to the persistent deadlock by changing the Charter’s title and revising the text. The replacement in the title of ‘regional self-government’ with the less specifically autonomy-related ‘regional democracy’ was supposed to cater to the concerns of certain states. Disagreement, however, regarding the concept of self-government could not be resolved on a terminological basis. Consensus was eventually reached only for the adoption in 2009 of a Reference Framework for Regional Democracy101 at a Council of Europe Conference of Ministers responsible for Local and Regional Government. This was, however, merely a compilation of documents that had been passed by the CLRAE and the Committee of Ministers up until then.102 Instead of being legally binding, the Reference Framework was thus limited to functioning as a point of reference for processes of territorial restructuring.

97 J Loughlin, ‘The Regional Question, Subsidiarity and the Future of Europe’ in Weatherill and Bernitz, The Role of Regions (2005) 164ff. 98 Currently, REGLEG is composed of 73 regions from eight Member States: Austria, Belgium, Finland (only Åland Islands), Germany, Italy, Portugal (only Azores and Madeira), Spain and the United Kingdom (only Northern Ireland, Scotland and Wales). 99 Congress of Local and Regional Authorities of Europe, ‘Recommendation 34 (1997) on the Draft European Charter of Regional Self-Government’ (3 June 1997). 100 See ch 9.1. 101 Council of Europe Reference Framework for Regional Democracy, MCL-16(2009)11. 102 The Helsinki principles adopted by the Conference of Ministers responsible for Local and Regional Government on 28 June 2002, the European Charter of Local Self-Government and the CLRAE Recommendation 240 (2008).

30 Concepts 1.2.3 The Proliferation of Regionalisms Unlike the concept of federalism, which has arguably experienced some consolidation over the last three decades,103 various understandings of regionalism have rather proliferated. This proliferation also accounts for the considerable space that this chapter dedicates to the explanation of these different regionalisms. Some of the manifold conceptions even go beyond the above-mentioned definition of a region as a subnational entity, that is, ‘as an intermediate territorial level, between the state and the locality’.104 Rather than subnational regionalism, the focus is in these cases on a sort of cross-border or supranational regionalism.105 The latter notion of regions above states is, of course, anything but new. After all, this is the understanding that has always prevailed in the fields of international law and international relations. Up until the 1960s, ‘regionalism’ was arguably used overwhelmingly to denominate political, economic or military cooperation between states and only then became more common as a term referring to an intermediate level within states.106 But, as is still the case today, primarily economy-focused collaboration within such organisations as NAFTA, Mercosur, APEC or ASEAN107 is said to constitute ‘macro geo-economic regions’.108 While the EU has developed, in contrast to the aforementioned examples, elements of clearly more far-reaching political integration, it has recently created essentially economy-focused ‘macro-regions’ within its territory. These transcend international boundaries, as they are groupings of (subnational) regions from different Member States that originated in the 1980s and that are held together by certain common interests, problems or potentials. Back then, an overarching aim of European integration was, not least due to the establishment of the single market precisely at that time, to counteract the feared marginalisation of peripheral regions and to tackle problems of urbanisation. More recently, efforts towards macro-regions have been undertaken, especially by the entrenchment with the 2007 Lisbon Treaty of territorial cohesion, that is, the balanced development of European regions, as an EU objective (Article 174 TFEU) and, soon thereafter, the adoption of macro-regional strategies.109 As pointed out by the Committee of the Regions, the new macro-regions ‘are not an extra institutional tier within the European Union, but a network, a modus operandi or, rather, a form of joint action that involves various European, national, regional and local players,

103

See ch 1.1.2 above. Keating (n 50) 9. 105 ‘Supranational’ is meant here, of course, not in the legal sense as a form of union that exercises certain powers delegated by the Member States, but, instead, signifies regions above the state level. 106 See Keating (n 50) 9. 107 These are the North American Free Trade Agreement (NAFTA), the Mercado Común del Sur (Mercosur) in Latin America, the Asia-Pacific Economic Cooperation (APEC) and the Association of Southeast Asian Nations (ASEAN). 108 Ortino, ‘Functional Federalism’ (2005) 282. 109 These are the EU Strategies for the Baltic Sea Region (2009), the Danube Region (2010), Adriatic and Ionian Region (2014) and the Alpine Region (2015). See European Commission, ‘Report from the Commission Concerning the Governance of Macroregional Strategies’ COM (2014) 284 final. 104

Regionalism 31 various policies and various funding programmes’.110 The fact that the focus is thus clearly on pragmatic collaboration between existing government levels within the macro-regions by making use of existing policies and funds,111 makes the potential of this relatively new strategy critically dependent on efficient coordination. Another form of regions transcending states is created through cross-border cooperation (CBC), which may be defined in broad terms as any concerted action of subnational entities or local governments involving bi- or multilateral cooperation across international boundaries.112 This definition therefore does not include collaboration between local authorities merely across subnational boundaries, which is sometimes subsumed under CBC in a broader sense.113 But even CBC in the abovementioned sense is arguably an old phenomenon of everyday practice, even if not, in many cases, legally entrenched. Only the paradigm of the modern nation state has strongly reinforced sovereignty and thus erected barriers that are easily surmountable for national governments with their centralised power over external relations but less for the thus newly entrenched peripheries. In Europe, this state of affairs has changed in recent decades through a process of legal institutionalisation that has given rise to ‘new centers in the periphery’.114 In other words, the establishment of cross-border regions has gradually softened international boundaries and transformed former ‘barriers’ into ‘spaces of contact’.115 This process started, as in other parts of the world, with pragmatic functional cooperation in areas like infrastructure or economic development, which was to overcome the inherent disadvantages of peripheral areas. Put differently, what has been aptly termed in the US context as ‘cross-border housekeeping’116 has sparked the establishment of many of Europe’s cross-border regions, often denominated ‘Euroregions’. But the first such arrangement, concluded in 1958 between German and Dutch municipalities, in some way already foreshadowed that there could be more rationales for collaboration than functionalism. More concretely, this cooperation was also regarded as contributing to mutual understanding and, in a post-war context, even reconciliation. In particular, the 1990s then witnessed a trend towards

110 Committee of the Regions Opinion on ‘Territorial Cooperation in the Mediterranean through the Adriatic-Ionian Macro-Region’ (own-initiative opinion) 2012/C 9/03. 111 A Dubois et al, ‘EU Macro-regions and Macro-regional Strategies: A Scoping Study’ (2009) Nordregio Electronic Working Paper 4/2009, 24 www.diva-portal.org/smash/get/diva2:700381/FULLTEXT01.pdf. 112 For CBC as part of external relations and concrete examples from five federal systems, see ch 14. 113 This understanding underlies most chapters, especially those on Australia, Germany and Switzerland, in B De Villiers (ed), Crossing the Line: Dealing with Cross-Border Communities (KonradAdenauer-Stiftung, Johannesburg, 2009). Local cooperation across subnational boundaries, which occurs particularly but not exclusively in metropolitan areas, is in this book treated in ch 9.4.2. 114 R Toniatti, ‘How Soft Is and Ought to be the Law of Interregional Transborder Cooperation?’ in R Kicker et al (eds), Changing Borders: Legal and Economic Aspects of European Enlargement (Frankfurt, Peter Lang, 1998) 43. 115 See R Ratti, ‘Spatial and Economic Effects of Frontiers: Over-view of Traditional and New Approaches and Theories of Border Area Development’ in R Ratti and S Reichman (eds), Theory and Practice of Transborder Cooperation (Basel, Helbing & Lichtenhahn, 1993). 116 J Kincaid, ‘Foreign Relations of Sub-national Units’ in R Blindenbacher and A Koller (eds), Federalism in a Changing World: Learning from Each Other (Montreal, McGill–Queen’s University Press, 2003) 86.

32 Concepts more symbolically loaded cooperation, which was aimed, aside from the functional logic, partially, sometimes even predominantly, at strengthening the ties between people belonging to an ethno-cultural group separated by an international boundary.117 The increasing application of CBC at that time in contexts of ethno-cultural diversity is reflected, for instance, in its entrenchment in the Council of Europe’s 1995 Framework Convention for the Protection of National Minorities (FCNM),118 in new arrangements especially in post-communist Central and Eastern Europe,119 and in the particularly well-institutionalised collaboration between Northern Ireland and the Republic of Ireland as a result of Strand 2 of the 1998 Belfast Agreement. While the functional and/or symbolic rationales of CBC are typical of Europe as well as of other parts of the world, a main difference lies in the legal framework. Beyond its promotion through funding within the EU’s regional policy,120 such cooperation has also taken root, in Europe, in both international and supranational law. More precisely, the Council of Europe initiated the Madrid Outline Convention of 1980,121 which was the first binding international treaty requiring from the contracting states domestic legal recognition of cross-border activities, albeit only pursuant to vague minimum standards. Even if this rather weak framework was strengthened through the adoption of three protocols to the Madrid Convention,122 it is without doubt eclipsed in practice by the more effective legal tools provided by the 2006 EU regulation on a European Grouping of Territorial Cooperation (EGTC).123 As an EU regulation, this legal act is immediately enforceable law in all Member States and provides a binding framework for the so-called EGTCs, which are public bodies for cross-border cooperation with legal personality under European law and domestic public law. With the legal framework for these bodies and the above-mentioned declaration of territorial cohesion as an EU objective that requires ‘particular attention … to cross-border regions’ (Article 174 TFEU), the European Union turned from a mere promotor into a regulator of CBC. Interestingly, this recent activism at the EU level has enhanced, as a side effect, the legal role played by national governments because the Member States are the 117 F Palermo, ‘Trans-border Cooperation and Ethnic Diversity’ in J Kühl and M Weller (eds), Minority Policy in Action: The Bonn-Copenhagen Declaration in a European Context 1955-2005 (European Centre for Minority Issues & Institut for Graensregionsforskning, Syddansk Universitet, 2005) 162. 118 Art 17(1) FCNM: ‘The Parties undertake not to interfere with the right of persons belonging to national minorities to establish and maintain free and peaceful contacts across frontiers with persons lawfully staying in other States, in particular those with whom they share an ethnic, cultural, linguistic or religious identity, or a common cultural heritage.’ Art 18(2) FCNM: ‘Where relevant, the Parties shall take measures to encourage transfrontier co-operation.’ 119 See A Engl and J Woelk, ‘Crossborder Cooperation and Minorities in Eastern Europe: Still Waiting for a Chance? A Summary and Evaluation of the Four Case Studies’ (2008) 6 European Yearbook of Minority Issues (2006/2007) 235. 120 See section 1.2.1 above. Most notable in this regard is the INTERREG initiative, which has provided funding since 1989. 121 European Outline Convention on Transfrontier Co-operation between Territorial Communities or Authorities ETS No 106 (Madrid, 21 May 1980). 122 See, especially Protocol No 3 to the European Outline Convention on Transfrontier Co-operation between Territorial Communities or Authorities concerning Euroregional Co-operation Groupings (ECGs) ETS No 206 (16 November 2009). 123 Regulation (EC) 1082/2006 on a European Grouping of Territorial Cooperation [2006] OJ L210/19, 31.

Regionalism 33 source of implementing legislation and, regarding the EGTCs, are in possession of important means of intervention.124 This seems rather ironic, of course, as CBC was originally conceptualised precisely ‘as a form of emancipation of sub-national entities from the intrusiveness of the state’.125 Anyway, it is certainly a very significant development that cross-border regions are today entrenched in supranational law and thus enjoy even greater legal guarantees than subnational entities, the status of which may be modified by unilateral national acts alone. Of course, CBC is by no means unknown beyond Europe. The new EGTCs actually share with cross-border regions in other parts of the world certain intrinsic problems, namely the democratic accountability and legitimacy of cooperation that is largely managed by bureaucracies and, linked to this, people’s acceptance of decisions. Beyond Europe, however, CBC activities are typically carried out with a much lower degree of legal institutionalisation, either on the basis of private law instruments, such as associations and companies, or even in entirely informal ways.126 As a result, cross-border regions have hardly entered, so far, the sphere of public law.

124 They may influence cooperation (Art 4 of the 2006 EGTC Regulation), interrupt it (Arts 13 and 14) and even themselves become members of an EGTC (Art 4). 125 F Palermo, ‘The “New Nomos” of Cross-Border Cooperation’ in F Palermo et al (eds), Globalization, Technologies and Legal Revolution: The Impact of Global Changes on Territorial and Cultural Diversities, on Supranational Integration and Constitutional Theory—Liber Amicorum in Memory of Sergio Ortino (Baden-Baden, Nomos, 2012) 82f. 126 For North America, see, eg, JK Blatter, ‘Debordering the World of States: Towards a Multi-level System in Europe and a Multi-polity System in North America? Insights from Border Regions’ (2001) 7 European Journal of International Relations 175.

2 Manifestations Federalism is a natural constitution for a body of states which desire union and do not desire unity …1

… but, it could be added, the form such a ‘natural’ constitution takes is deeply influenced by the process that led to its formation. As described in the previous chapter, federalism is a manifold concept which may have several institutional manifestations and often means different things to different people in different contexts. Nearly all states, with very few exceptions (usually the so-called micro-states) encompass some form of vertical division of powers.2 At the same time, countless attempts have been made in literature to define the federal state as something opposed to other forms of vertical division of powers and functions. Daniel Elazar and Ronald Watts have tried to capture all these diverse manifestations by using, respectively, the broad umbrella terms ‘varieties of federal arrangements’ and ‘federal political systems’. Both scholars list within these overarching categories different forms of organisation such as unions, federations, confederations, federacies, associated states, condominiums, leagues and joint functional authorities; plus—needless to say—hybrids, ie, combinations of two or more categories.3 The fact that the manifestations of federalism are extremely diverse and difficult to distinguish was known long before the times of Elazar and Watts. Indeed, some scholars observed as early as 1940 that ‘it is not always possible to draw clear and incontestable distinctions … alliance shades into league, league into confederation, confederation into federal state, federal state into unitary state.’4 This chapter looks at how the toolkit of federalism can be used for very different constitutional purposes and shows that federal and regional states differ essentially in their historical origins, the former being the first manifestation of the federal principle, the latter representing a more recent way in which this principle is used to accommodate territorial pluralism. The historical turning point is represented by the nation state paradigm: when the creation of nation states was the overall trend in order to create bigger, militarily and economically more competitive countries, smaller sovereign

1 AV Dicey, Introduction to the Study of the Law of the Constitution, 6th edn (London, Macmillan & Co, 1902) lxxv. 2 See M Frenkel, Federal Theory (Canberra, Australian National University, 1986) 389. 3 Beyond these manifestations mentioned by both authors, Elazar also lists consociations and Watts constitutionally decentralised unions. See DJ Elazar, Federal Systems of the World: A Handbook of Federal, Confederal and Autonomy Arrangements (Harlow, Essex, Longman, 1991) XVI; R Watts, Comparing Federal Systems, 3rd edn (Montréal, McGill–Queen’s University Press, 2008) 8f. 4 H Greaves, Federal Union in Practice (London, G Allen & Unwin, 1940) 10.

Confederation 35 entities were pushed by historical and political circumstances to pool together in a bigger polity; conversely, when the age of nation states came to an end, the driving forces of federalism shifted from military and economic–industrial supremacy towards a more effective administration and/or accommodation of ethno-cultural or other kinds of differences, thus promoting decentralisation of existing countries. Further, the chapter describes the most recurrent institutional forms of the vertical division of power. It explores the legally and historically most relevant manifestations of what is to be considered a common phenomenon (confederations, federal states, regional states), and it deals with other, less recurrent forms (associations of states, federacies, devolution), as well as with some important conceptual challenges (the legal meaning of autonomy, the nature of the European Union). Finally, it concludes by warning against simplistic or formalistic definitions of federalism and its manifestations. 2.1 CONFEDERATION

2.1.1 Definition In times when federations were established by the coming-together of sovereign states, the first aggregative form was the confederation of states.5 A confederation is a union of independent states that transfer only limited parts of their sovereignty (eg, some defence powers) to a government with an enumerated and rather limited scope of responsibilities. The limitation of scope, however, cannot go as far as to comprise only one goal; thus, it would not be appropriate, for example, to call a defence organisation, such as NATO, a confederation. Three legal elements are needed for a confederation to exist. First, a confederation is a union of independent states ruled by international law rather than by constitutional law. Second, and consequently, fundamental decisions require unanimity, and each participating state thus has a veto power on every single decision of the confederation. Third, the confederation is financially dependent on funds transferred by the participating states, meaning that it lacks its own revenues. These fundamental elements of a confederation are generally uncontested, and for this reason there is no ambiguity in literature and in political practice as to the irrelevance of the nomen iuris: when these conditions are not met, there is no confederation regardless of the name of the polity. Conversely, even if Switzerland, for historical reasons, still calls itself a ‘confederation’ in Article 1 of its 1999 Constitution,6 its status as a prototype of the federal state7 is uncontested.

5 See M Forsyth, Unions of States: The Theory and Practice of Confederation (Leicester, Leicester University Press, 1981). See GJ Ebers, Die Lehre vom Staatenbunde (Breslau, M&H Marcus, 1910). 6 For example, as mentioned in box 0.B, similarly, Australia is still called a Commonwealth (Art 1 of the Constitution of 1901), but it is a federal state, as was the Soviet Union, at least nominally, in spite of being called a union. 7 See ch 3.3.2.

36 Manifestations One of the prototypes of modern confederations was the United States of America between independence in 1776 and the entry into force of its federal Constitution in 1789.8 More precisely, after the independence of the 13 former colonies, the Articles of Confederation and Perpetual Union were drafted in 1777 and entered into force on 1 March 1781. For the eight years they were in force, the Articles of Confederation created a confederation of American states called the United States of America. The system of government was based on a permanent Congress of the Confederation, more formally called ‘the United States in Congress assembled’, in which each state had one vote. The Congress was in charge of a limited number of functions, including common defence, setting weights and measures (including currency) and serving as an arbitration court for disputes between states. When the Congress was not in session, a Committee of the States, based on the same principle of equal representation of each state with a rotating presidency, performed the functions of the Government. 2.A The Articles of Confederation In the Articles of Confederation, the confederal nature clearly comes to the fore: Article 2 stated that ‘[e]ach state retains its sovereignty, freedom, and independence, and every power, jurisdiction, and right, which is not by this confederation expressly delegated to the United States, in Congress assembled’. Moreover, any change in the Articles of Confederation required unanimity (Article 13). Article 3 further clarified the nature of the Union, affirming that the states formed a ‘firm league of friendship with each other, for their common defense, the security of their liberties, and their mutual and general welfare, binding themselves to assist each other, against all force offered to, or attacks made upon them, or any of them, on account of religion, sovereignty, trade, or any other pretense whatever’. The Union was called upon to guarantee to the citizens of each state ‘all privileges and immunities of free citizens in the several States; and the people of each State shall have free ingress and regress to and from any other State, and shall enjoy therein all the privileges of trade and commerce’ (Article 4). The functions attributed to the Congress were limited to war powers, other important issues of foreign policy (Articles 6 and 9) and a few in the domestic arena (Article 4). Article 8 established that the common treasury had to be supplied by the states ‘in proportion to the value of all land within each State’, by taxes to be ‘laid and levied by the authority and direction of the legislatures of the several States’.

2.1.2 Are Confederations Transitional? The second most relevant historical example of a confederation, the Swiss confederation, was also eventually replaced by a federal constitution.9 While its history was much longer as compared to the American case—the origins of the confederation date back to the ‘eternal confederal compact’ of 1291 among the original cantons as a military and trade alliance—it was nevertheless replaced by a federal constitution

8 9

See on this transition, chs 3.2.2 and 3.3.1. See ch 3.2.1 and 3.3.2.

Confederation 37 with a stronger national government in 1848 after the Civil War. The evolution of the two most relevant historical confederations towards a federal state10 raises the question as to whether confederations are doomed to fail and will necessarily end up centralising in a federal setting that represents, using the terminology of the drafters of the US Constitution, ‘a more perfect union’. Such an approach, however, does not entirely explain confederations and their importance for the federal theory. Like most of the traditional readings of federalism, this approach is biased by the excessive attention paid to certain important historical experiences that tend to be generalised.11 Following more careful observation, however, it must be noted that, first, no form of government is eternal (in spite of the terminology normally used by constitution drafters), and second, the ‘inevitable evolution’ from a confederation into a federation tends to be rather common when federal states come together through ever deeper integration of previously sovereign entities, such as in the case of the United States or Switzerland. Furthermore, confederation is not a transitional step from an imperfect to a more perfect union when it is established as the consequence of the dissolution of a previously more unitary state. One could take the examples of the Commonwealth of Nations (previously known as the British Commonwealth) or of the Commonwealth of Independent States, which are loose and rather non-political associations of sovereign states originating from, respectively, the British Empire and the Soviet Union—although not all former members of these two entities participate in these organisations, and some members of the Commonwealth of Nations were never British colonies.12 Another recent example is the evolution of Serbia and Montenegro, first established in 1992 as a federation called the Federal Republic of Yugoslavia following the break-up of Yugoslavia, then reconstituted in 2003 as a confederation. The State Union was a very loose confederation, even with separate currencies and economic policies (Montenegro adopted the euro, while Serbia kept the dinar as its currency). Strongly promoted by the EU in an attempt to stave off further conflict in the Balkans, this transitional stage was decided while waiting for a referendum on Montenegro’s independence that would determine whether to finally establish two separate countries.13 The referendum was held in 2006, with the majority of Montenegrin voters opting for independence, and the State Union officially came to an end with the formal declaration of independence of both Serbia and Montenegro that same year. A counter-example is the European Union, which includes elements of a federation and others of a confederation.14 While the EU has certainly developed the

10 The trajectory of the Republic of the Seven United Netherlands was rather different for historical reasons, see ch 3.2.1. 11 See A Gamper, ‘A “Global Theory of Federalism”: The Nature and Challenges of a Federal State’ (2005) 6 German Law Journal 1297, 1297ff. 12 See KC Wheare, The Constitutional Structure of the Commonwealth (Oxford, Clarendon Press, 1960). 13 See M Suksi, ‘On the Voluntary Re-definition of the Status of a Sub-state Entity: The Historical Example of Finland and the Modern Example of Serbia and Montenegro’ (2004) 4 Faroese Law Review 33. 14 On the nature of the EU, see section 2.4 below.

38 Manifestations former in the course of its history, it has also clearly refused to become a fully-fledged federal state, as epitomised by the rejection of the Constitutional Treaty in 2005. In sum, it would be misleading to see confederations as just a transitional step towards the establishment of federal states. Throughout history, this has been the case for some confederal experiences but clearly not for all of them. Like federal states, confederations also do not all look the same, nor did they originate in the same way, nor is their destiny predetermined. However, they generally carry little political weight vis-a-vis the individual countries and the international community and are often weak in terms of decision-making and effectiveness.15

2.2 FEDERAL STATE

2.2.1 The Compact as the Traditional Definition: The ‘Federal Big Bang’ As mentioned earlier in this book, to define federalism in abstract terms is not only impossible but also futile. Every scholar tends to advocate a specific understanding of what constitutes a federal state, and nearly every constitutional or supreme court in countries that define themselves as federal has provided some definition of the essential elements of a federal state. Quite recently, the Canadian Supreme Court, for instance, provided a very much democracy-focused definition. The judges regarded as the essence of the federal state its function ‘to enable citizens to participate concurrently in different collectivities and to pursue goals at both a provincial and a federal level’.16 What many of the apex court definitions have in common is that they look at the domestic experience and, not unlike academic classifications, naturally tend to generalise the essential elements of one particular experience in order to subsume a general rule. This approach is reflected, for instance, in the definitions provided by the Indian Supreme Court and the Constitutional Court of Austria. 2.B India 1963–2001 and Austria 1952–88: Can Court Definitions of the Federal State be Generalised? Since the adoption of its national Constitution in 1949, India has always been termed a federation and/or a union. For clarification, it is useful, in this regard, to go back to the explanations of terminology and concepts by BR Ambedkar, the Chairman of the Drafting Commission. In his words, ‘India was to be a federation’, but not one that the states had joined through an original act of union and could, therefore, leave in a reciprocal act of disunion. The term union was a deliberate choice to reflect a unitary penchant17 and a specific ratio essendi: ‘Though the country and the people may be divided into different States for convenience of administration the country is one integral whole, its people a

15

See ch 3.2.1. Reference Re Secession of Québec [1998] 2 SCR 217, para 66. See box 4.E. 17 This unitary penchant is reflected, for example, in the Indian Constitution’s emergency clauses to the benefit of the national government, see ch 5.2.3. 16

Federal State 39

s ingle people living under a single imperium derived from a single source.’18 Even though the Supreme Court, in its early jurisprudence, sometimes affirmed, in contrast to Ambedkar’s definition, that India’s Constitution ‘is not truly federal in character’,19 it has in its more recent case law tended to recognise the country’s federal nature more unequivocally. In 2001, it stated that the ‘Indian Constitution is basically federal in form and is marked by the traditional characteristics of a federal system, namely, supremacy of the Constitution, division of power between the Union and States and existence of an independent judiciary. From Kashmir to Kanyakumari, the country is one and there is no intelligible differentia which distinguishes advocates practising within the State of Rajasthan and those practicing outside Rajasthan but within the territory of India.’20 By contrast, the Constitutional Court of Austria has provided a partially divergent definition. According to this court, elements of the federal state are the division of competences and the participation of the Länder in the federal legislative process,21 the constitutional autonomy of the Länder22 and the responsibility of the Länder for the implementation of national law, the so-called indirect federal administration.23 It is evident that the definitions of the two courts emphasise beyond some commonalities certain country-specific elements, ie, in the Indian case, an independent judiciary and identical requirements for the exercise of a legal profession and, in the Austrian case, the indirect federal administration. But both these characteristics are far from typical of all federal states so that these court definitions may not serve as bases for a generally applicable definition.

However defined, the federal state presents some common elements, which can be described as necessary but not sufficient. In other words, they must exist in each federally organised state, but are not enough, alone, to identify a federal state as opposed to other, similar state organisations.24 The ‘skeleton’ of a federal state is represented by: a) the division of state functions between at least two different orders of government both enjoying political autonomy; b) the supremacy of the federal/ national constitution; and c) a system of cooperation among the levels, including the judicial adjudication of disputes between and among the entities over the respective constitutional powers. Put differently, federally organised states find (different) ways to divide public powers among different spheres of government, combine (in different ways) self-rule and shared rule and (in different ways) unite without merging and divide without separating. As with federalism, the literature also often refers to federation in relation to the origins of the term foedus (compact) in order to consider that federal states are only those established by the coming-together of previously independent entities. Such an approach is too narrow and is inherently wrong, because it looks at only one of the origins of federalism. First, some states, whose federal nature is uncontested,

18

Parliament of India, Constituent Assembly Debates, vol VII, 4 November 1948, 43. State of West Bengal v Union of India, 1963 AIR 1241. 20 Ganga Ram Moolchandani v State of Rajasthan, 2001 AIR 2616. 21 VfSlg 2455/1952. 22 VfSlg 11669/1988. 23 VfSlg 11403/1987. On the indirect federal administration, see ch 5.3. 24 On the futility of a definition, see section 2.2.2 below. 19

40 Manifestations have been formed by the transfer of powers from a previously unitary entity, such as in the case of Belgium and to some extent Austria, India, Nigeria and others. Second, and more importantly, the difference between ‘coming-together’, ‘holdingtogether’ and ‘putting- or forced-together’ federations25 is only historical, not legal, in nature. In coming-together federations, the member units were originally both historically and legally sovereign, while the federal government is historically derived. However, when the ‘federal big bang’ takes place, ie the federal constitution enters into force, the federal government becomes legally sovereign and the member units legally derived.26 The federal big bang turns the original sovereignty of member units into autonomy, due to the primacy of the federal constitution. This is confirmed by the so-called homogeneity clauses included, not by chance, in all federal constitutions of coming-together federal states, such as Article 4, Clause 4 of the US Constitution (the Republican Clause), Articles 5 and 6 of the Swiss Constitution, Article 28 of the German Basic Law, etc.27 In some cases, the impact of the federal big bang is even more profound as it entrenches the federal system so firmly that it is even protected against the power of constitutional amendment. While several federal countries have declared the federation, in their constitutional texts28 or jurisprudence,29 as indissoluble, Germany’s so-called ‘eternity clause’ is a particularly notable example. Article 79(3) of the Basic Law stipulates that amendments ‘affecting the division of the federation into Länder’ on principle in the legislative process’ shall be inadmissible. Even if this provision does not guarantee the individual existence of each Land,30 it entrenches both the country’s territorial division as such into (at least two) Länder and their participation. Whereas the firm entrenchment of the federal system has been an integral part of Germany’s constitutional order from its outset in 1949, a similar rule was belatedly introduced in India by the case law of the Supreme Court.

25 This threefold distinction goes back to A Stepan, ‘Federalism and Democracy: Beyond the US Model’ (1999) 10 Journal of Democracy 19. While Alfred Stepan’s term ‘putting-together’ was supposed to capture the unique character of federal states that were formed through coercion, Nancy Bermeo’s similar term ‘forced-together’ aims to place particular emphasis on the influence of external actors and the element of system frailty. See N Bermeo, ‘The Import of Institutions’ (2002) 13 Journal of Democracy 96, 108. 26 The expression ‘federal big bang’ to describe the entry into force of a supra-ordinated federal constitution is sometimes used in the discourse on European integration. See N Walker, ‘The Shifting Foundations of the European Union Constitution’ (2012) University of Edinburgh, School of Law, Europa Working Papers 2012/1. For a more contextual use in the theory of federalism, see R Toniatti, ‘Federalismo e potere costituente’ in Regionalismo e federalismo in Europa (Trento, Giunta della Provincia Autonoma di Trento, 1997) 171. 27 See ch 5.1.2. 28 For instance, Art 1 of the Brazilian Constitution states the federal republic is ‘formed by the indissoluble union of the states and municipalities and of the federal district’. The Preamble to the Australian Constitution underscores explicitly that the people of the colonies ‘have agreed to unite in one indissoluble Federal Commonwealth’. 29 Texas v White 74 US (7 Wall) 700, 725 (1869). On this seminal ruling, see box 4.D. 30 See box 6.F.

Federal State 41

2.C India 1973: The Federal Character as Part of the Constitution’s Basic Structure? In 1970, the religious guru Sri Kesavananda Bharati challenged attempts of the state of Kerala to impose certain restrictions, under land reform acts of that state, on the management of his community’s property. While in its ruling the Supreme Court overruled previous decisions implying that the right to property could not be restricted, the case is today more known for its development of the ‘basic structure doctrine’. In their early jurisprudence after 1949, the judges had asserted that any provision of the Indian Constitution would be amendable, including fundamental rights (Part III) and the procedure of constitutional amendment itself (Article 368). Already in 1967, the Supreme Court had initiated a change of course in the Golaknath case.31 In this decision it ruled that an act of amendment under Article 368 was also a ‘law’ within the meaning of Article 13(2) of the Constitution and thus, according to this provision, subject to constitutional fundamental rights. In the Kesavananda Bharati judgment of 1973, the Court overruled the Golaknath decision, as it affirmed that a constitutional amendment was not a ‘law’ for the purposes of the above-mentioned Article 13(2), but at the same time established certain limits.32 Confirming the constitutionality of the Twenty-Fourth, Twenty-Fifth and Twenty-Ninth Amendment, the Court famously held that an amendment could not damage, abrogate or alter what it called the ‘basic structure’ of the Constitution. The fact that the parliament only had the power to amend would indicate implied limits. Whereas changes would be admissible in order to adapt the Constitution to changing conditions, they would have to leave intact its very foundations and basic institutional pattern. Regarding the concrete elements of the Constitution’s basic structure, there has been a certain ambiguity and flexibility from the beginning, as even several of the judges adopting the majority opinion laid them out separately. However, most of them shared the view that the federal character of the Constitution constitutes such an element which was reiterated in the Bommai ruling of 1993.33 Although the Court introduced the basic structure doctrine in a 7–6 decision and has tended in subsequent judgments to determine elements on a case-by-case basis, the doctrine has gained broad acceptance. To a considerable extent, this is due to its role during the state of emergency declared by Indira Gandhi in 1975 based on Article 352(1) of the Constitution.34 The basic structure doctrine was used by the Supreme Court to strike down the Thirty-Ninth Amendment,35 which was supposed to prevent Indira Gandhi’s prosecution, and is thus widely credited with restoring Indian democracy.36

It appears to be clear that the sovereignty debate, although key, cannot provide definite answers to the phenomenon of federal states. While in legal terms it seems incontestable that the prevalence of the federal constitution turns original subnational sovereignty into legal autonomy, in broader terms it seems that sovereignty

31

Golaknath v State of Punjab, 1967 AIR 1643. His Holiness Kesavananda Bharati Sripadagalvaru v State of Kerala, 1973 AIR 1461. 33 SR Bommai v Union of India, 1994 AIR 1918. 34 On this provision, see ch 5.2.3. 35 Indira Nehru Gandhi v Raj Narain, 1975 AIR 2299. 36 Another seminal ruling that further developed this doctrine is State of Bihar v Bal Mukund Sah, 2000 AIR 1296. 32

42 Manifestations simply does not explain the whole issue and remains the subject of controversial debates.37 As Scottish scholar Neil MacCormick effectively put it, sovereignty is rather like ‘virginity: something that can be lost by one without another’s gaining it’.38 In his view, sovereignty has been lost within the European Union by its Member States without the European level gaining it.39 While the supremacy of the national constitution legally strips sovereignty from the member units, the historical formation of federal states has, however, normative consequences for the institutional setting of each federal system. In coming-together federal countries, the reduction of sovereignty of member units into autonomy is compensated by the units’ participation in the exercise of sovereign powers at the national level: this occurs especially by means of guaranteed representation in powerful second chambers, such as the US Senate, the Swiss Council of States (Ständerat) or the German Federal Council (Bundesrat) and by retaining some control over the constitutional amendment power, up to a veto right of each individual member unit as to some essential elements of the federal compact.40 One may think of Article 5 of the US Constitution, which prescribes, inter alia, that ‘no State, without its consent, shall be deprived of its equal suffrage in the Senate’.41 In other words, before agreeing to the federal big bang, future member units must accept the predominance of the federal government in exchange for maintaining some influence on national policies (both collectively through the second chamber or other mechanisms of participation and, more rarely, individually).42 Conversely, in holding-together federations, such institutional representation of member units is either non-existent or extremely weak: there is usually no strong involvement of member units in the procedure for amending the constitution, no veto right as to essential elements of it and usually the national government retains the power to legislate instead of the member units when legal or economic unity is at stake.43 2.2.2 Form Follows Function: The Role of History in Shaping Federal Manifestations Federal states are created in different ways, and the origins usually influence the institutions of each federation and ultimately the understanding of federalism that is behind any attempt to define a federal state. From a historical perspective, it must be noted that the era of coming-together federations had momentum that lasted up to, roughly, World War I: before, basically all federations were created by

37

See ch 4.1. N MacCormick, Questioning Sovereignty: Law, State and Nation in the European Commonwealth (Oxford, Oxford University Press, 1999) 126. 39 For details regarding sovereignty in the EU context, see ch 4.1.3. 40 See H Wechsler, ‘The Political Safeguards of Federalism: The Role of States in the Composition and Selection of the National Government’ (1954) 54 Columbia Law Review 543. 41 For details, see ch 6.3.1. 42 See ch 6. 43 Examples of this are Art 120 of the Italian Constitution or Art 150(3) of the Spanish Constitution. For further details, see ch 5.2.3. 38

Federal State 43 uniting previously sovereign entities, sometimes experimenting with a looser union (a confederation) during the process. One may think of the classical federations such as the USA, Switzerland and Germany.44 Also other coming-together federations were established before World War I, although with some peculiarity from the legal point of view: formally, Canada and Australia obtained their (federal) constitutions, respectively, in 1867 and 1901 as acts of the British Parliament and acquired de iure sovereignty only with the Statute of Westminster of 1931. Yet, they can be considered coming-together federations in the material sense, since in both countries federations were established, in the Canadian case at least for the most part, on top of existing quasi-sovereign units.45 The process of the establishment of coming-together federations is very much linked to nation state formation: especially during the nineteenth century, nation states represented the winning model, and countries that had not yet achieved this type of statehood were pushing for it. In some cases, for historical, political, ethnic/religious or other reasons, the formation of nation states was carried out by the coming-together of sovereign entities. At the same time, however, in some circumstances the federal state was created after the establishment of the nation state, like in the case of nearly all Latin American federations. After World War I, especially due to the reconfiguration of the geopolitical landscape in Europe after the dissolution of the Habsburg Empire, as well as of the Russian and Ottoman Empires, the process towards the formation of nation states essentially came to an end. Some historically hybrid federations were created by reuniting, under a federal umbrella, entities that were previously parts (with different statuses) of bigger empires, as in the case of Austria, Czechoslovakia or even, mutatis mutandis, the Soviet Union. For all these federal experiences, it is debatable whether they represent coming-together or holding-together types of federations, and it is probably correct to consider them to be both. Certainly, the vast majority of all federal systems, however defined, established after the 1920s are holding-together federal systems that resulted from a transfer of power from a unitary state.46 One may think of Belgium, Spain (both the constitutions of 1931 and 1978), Italy, the United Kingdom, and most recently Nepal, but, for the most part, also the cases of India47 and Bosnia and Herzegovina. In other words, historical origins are indeed 44

See ch 3.3. regard to the historical facts, the situation is of course more complex, as many states/provinces were formed by separation from existing entities. One may think, for example, of the establishment of the states of Victoria and Queensland in Australia in 1851 and 1859, respectively, whose territories were previously part of New South Wales. See ACV Melbourne, Early Constitutional Development in Australia: New South Wales 1788–1886, Queensland 1859–1922, RB Joyce (ed), 2nd edn (St Lucia, University of Queensland, 1963) esp 443f. What was before 1867 the United Province of Canada came to be split up at the founding of the federal state into the two separate provinces Ontario and Quebec. The provinces of Manitoba, Alberta and Saskatchewan and the territories of Yukon and Nunavut were over the years carved out from the vast Northwest Territories. 46 In the few cases when federations were established according to the old system, they usually collapsed. One may think of the confederation (which never became a federation) of Senegambia (1982–89), of the federation of the West Indies (1958–62), of French West Africa (which started prior to World War I and formally ended in 1958) or of the Federation of Rhodesia and Nyasaland (1953–64), etc. 47 India was established as a federal state by the Constitution of 1950 and cannot be considered a coming-together federal system, since the states and territories of the Indian Union were not fully-fledged 45 Having

44 Manifestations very relevant in shaping the constitutional features of federal states. History, however, should not be seen as the only defining element of a federal or regional state. Rather, it is important to identify historical cycles that have determined the ways of forming federations. Such a clear shift in the historical formation of federal states has to do with the changing meaning of the foedus. In both the past and the present, the foedus has indicated nothing other than the will to be together. Such a will evolves over time and takes on different constitutional forms. Sometimes it is even imposed, by force (such as in the American Civil War) or by international pressure (such as in Bosnia and Herzegovina, where it has been largely replaced by the international community),48 and its evolutions are part of the physiology of federalism. One may wonder, for instance, whether such a compact is still present in today’s Belgium, even though, constitutionally, it clearly exists. For all these reasons, the historical perspective seems much more effective than the formalistic sovereignty approach in explaining the legal nature and the political dynamics of federalism in its various manifestations. History and the political–societal functions to be achieved at the time of the establishment of the federal compact (both in terms of coming-together and holdingtogether) also define the intrinsic elements and the institutional character of the various federal states. Put differently, the form each federation takes and even its ideological underpinnings are determined by the historical and political motives behind each case. This is why several dichotomies are highlighted in the literature depending on such intrinsic and largely pre-institutional factors. The first and perhaps most common distinction is the one between mature and emergent federal systems.49 A distinction based essentially on time (mature federal countries are older, emergent ones are younger), but also on institutional features (only mature ones exhibit certain characteristics of many federations such as typically residual powers of subnational entities, a second chamber, etc) and on the

sovereign states but were more or less autonomous under British, French or Portuguese colonial rule, although most subnational entities were modelled according to the existing provinces established by the British Government of India Act of 1935. Moreover, at the time of independence, a huge number of more than 500 formally independent small territories remained out of the British Raj, and most of them subsequently voluntarily joined the Union (in the cases of Hyderabad and Jammu and Kashmir, they were annexed by force), and a few joined Pakistan. In sum, while not established by the coming-together of previously existing sovereign entities, the Indian Union is also partly the outcome of a union of provinces, states and territories. India should thus also be considered a hybrid federal system as to its origins. See G Austin, The Indian Constitution: Cornerstone of a Nation (Oxford, Clarendon Press, 1966) 186ff. Further, see B Ray, Evolution of Federalism in India (Calcutta, Progressive Publishers, 1967); and N Mukarji and B Arora (eds), Federalism in India: Origins and Development (New Delhi, Vikas Pub, 1992). 48 The fact that the High Representative for Bosnia and Herzegovina, appointed by the international community, was granted in 1997 the power to impose constitutional amendments (see box 4.C on the ‘Constituent Peoples Case’) has arguably changed the role of this official ‘from that of a supervisor of the peace implementation process to its main actor’, J Woelk, ‘Bosnia-Herzegovina: Trying to Build a Federal State on Paradoxes’ in M Burgess and GA Tarr (eds), Constitutional Dynamics in Federal Systems: Sub-national Perspectives (Montreal, McGill–Queen’s University Press, 2012) 119. 49 See RL Watts, ‘Typologies of Federalism’ in J Loughlin et al (eds), The Routledge Handbook of Regionalism and Federalism (Abingdon, Routledge, 2013) 25; Watts, Comparing Federal Systems (2008) 24f, 29f.

Federal State 45 ‘adaptability’ over time of mature federal systems when facing changing conditions. For Watts, the United States (1789), Switzerland (1848), Canada (1867), Australia (1901), Austria (1945), Germany (1949) and India (1950) match the criteria and therefore are mature federations.50 By contrast, emergent federal countries were created in the second half of the twentieth century, and are still looking for their equilibrium.51 Further, they do not present all the characteristics of mature federal systems; Watts calls these ‘quasi-federations’. Most emergent federal countries are then considered ‘fragile’,52 because they experienced long periods of non-democratic regimes. Among them, Watts includes Spain (1978), Belgium (1993), Russia (1993), Bosnia and Herzegovina (1994), the Latin American federations (Brazil, Mexico, Venezuela, Argentina), Nigeria (1960), Ethiopia (1995), Pakistan (1973), as well as the ethno-federalisms of the Soviet era and four micro-federations (Comoros, Saint Kitts and Nevis, Micronesia, Belau). However, the category of ‘emergent’ federal systems especially lacks homogeneity: it is difficult to group Spain and Belgium together with Pakistan, Venezuela or the micro-federations since the former are consolidated multilevel democracies, whereas the latter’s federal arrangements have mainly remained on paper. Furthermore, both Germany and Austria—two classic prototypes of mature federations—experienced a long period of authoritarian regime under Nazi rule. Finally, the classification does not take into account that the Latin American federations are vital and long-lasting and that Italy—a regional state that is surely ‘quasi-federal’—has also proved to be long-lasting. Precisely for the historical reasons mentioned above, the further distinction between aggregative (or integrative) and devolutionary federal systems also has a strong historical dimension. Aggregative systems are, as a matter of fact, older than devolutionary ones. The focus in this distinction is on the origin of the (federal) compact. Nicholas Aroney points out that aggregative federal countries are those in which the rationale was, at least predominantly, that ‘previously independent political communities have been integrated into a federal system’ (such as the United States,53 Switzerland, Canada, Germany and Australia), whereas devolutionary ones are ‘those in which a formerly unitary state has devolved governmental powers upon a number of regions within that state’ (such as Spain, Belgium, Italy and South Africa and, for the most part, India).54 In aggregative federal systems, pre-existing units must have been independent states (in the United States, Switzerland, Germany) or must have enjoyed a high degree of autonomy before their aggregation into the new federal structure, such as in Canada, Australia and Malaysia, dominions and 50

See Watts (n 3) 25.

51 ibid.

52 See N Steytler and J de Visser, ‘“Fragile Federations” and the Dynamics of Devolution’ in F Palermo and E Alber (eds), Federalism as Decision-Making: Changes in Structures, Procedures and Policies (Leiden, Brill–Nijhoff, 2015) 80. 53 In fact, even in the United States only the original 13 states came together. The subsequent states were formed from territory originally governed by the federal government, which had to approve their formation into states. 54 N Aroney, ‘Formation, Representation and Amendment in Federal Constitutions’ (2006) 54 The American Journal of Comparative Law 277, 282; N Aroney, The Constitution of a Federal Commonwealth: The Making and Meaning of the Australian Constitution (Cambridge, Cambridge University Press, 2009) 40f.

46 Manifestations colonies enjoying autonomy and self-government within the former British Empire. Again, the distinction is helpful but not all-encompassing. It does not explain, for instance, why some devolutionary federal systems such as Spain and Belgium present structures that are similar to those of integrative federations, nor it can respond to the question as to how to classify borderline cases, such as Argentina and Austria. And some federations, such as Canada and Australia, while materially aggregative, were legally created by an act of the British Parliament. Be that as it may, the aggregative or devolutionary origin of a federal system may have ‘a vital influence on the way in which a federal constitution can be interpreted’.55 When it comes to the institutional features of federal systems, a common distinction is made between dual and cooperative systems.56 Dual federalism is typical of the oldest federations (such as the United States, Canada, Switzerland, Australia), which adopted a distribution of powers, often explicitly based on a divided model of sovereignty:57 the powers allocated at different levels of government were separated as if they were watertight compartments.58 Further, laws enacted by each government level are, according to the model of ‘legislative federalism’,59 as a rule ‘implemented and administered by their own separate civil services and departments of states’.60 Such an approach was more frequent in older federations (with the notable exception of Germany), as it reflects the idea of two separate and parallel legal orders, while history has proven that this is practically impossible, as cooperation and coordination are inherent to federal structures. Even in the United States, initially dual federalism has given way to more entanglement between governments by becoming over time cooperative or, as some put it, rather coercive.61 Dual federalism, famously described due to the distinct and separate layers (of government) also as ‘layer cake’ federalism, has been replaced since the New Deal era by ‘marble cake’ federalism.62 Cooperative systems, on the contrary, are based on the institutional assumption that, in most areas, decision-making and implementation require action by both levels of government and thus their integration to a certain degree, in order to reduce conflicts, to increase efficiency and avoid duplication. At the same time, too much integration is likely to create confusion and reduce accountability, as well as ‘competition and autonomous action and initiative of each level of government’.63

55

See Aroney, The Constitution (2009) 40. For details, see ch 5.2.2. 57 On divided sovereignty in the US and Switzerland, see ch 4.1.2. 58 On the watertight compartments doctrine, see ch 5.2.1. 59 See ch 5.3. 60 TO Hueglin and A Fenna, Comparative Federalism: A Systematic Inquiry (Peterborough, Broadview Press, 2006); P de Vos and W Freedman (eds), South African Constitutional Law in Context (Oxford, Oxford University Press, 2014) 268. 61 See J Kincaid, ‘From Cooperative Federalism to Coercive Federalism’ (1990) 509 The Annals of the American Academy of Political and Social Science 139. 62 On these metaphors, see M Grodzins, ‘The Federal System’ in Goals for Americans: The Report of the President’s Commission on National Goals (New York: Columbia University Press, 1960). 63 Watts (n 3) 122. Critical approaches to cooperative federalism are discussed in RL Watts, ‘Origins of Cooperative and Competitive Federalism’ in SL Greer (ed), Territory, Democracy and Justice: Regionalism and Federalism in Western Democracies (London, Palgrave Macmillan, 2005) 121f. 56

Federal State 47 A further distinction is occasionally made between calm and restless federations.64 The rationale of this distinction is primarily sociological, as it indicates the attitude when it comes to the accommodation of ethnic groups. In general, federal states in the context of diverse societies are likely to be restless, whereas countries that are more homogeneous in terms of population tend to be calm(er). However, history and political considerations do play a fundamental role in this distinction too. It occurs, in fact, that long-lasting (therefore mature and thus older) federal systems are indeed calm even if they operate in the context of diverse societies, such as in the case of Switzerland or Canada, and it is difficult to determine whether this attitude depends on the fact that the federation is long-lasting or if the federation is long-lasting because of a generally calm, cooperative attitude. Furthermore, as most federal systems in the context of diverse societies have been established more recently, ie, after the Cold War,65 and definitely after World War I, there is a certain historical overlap between these (as a rule, restless) systems and those that follow a devolutionary pattern. The same also goes for the last recurrent distinction, ie, between symmetrical and asymmetrical federations which concerns differences between subnational entities in terms of institutional elements such as financial relations and representation in a second chamber. Having regard specifically to the element of the distribution of powers, symmetrical systems are those in which (legislative) powers are distributed uniformly, whereas asymmetry occurs when certain units are provided with more powers than the others.66 In fact, the historical element largely also covers this distinction, as usually older, aggregative federal systems are based on the principle of the same powers for the subnational units, whereas younger, devolutionary federal countries often provide some specific areas with a different status in order to accommodate specific differences, be they ethno-cultural or territorial.67 2.2.3 Definition by Means of Institutional Elements Many valuable attempts have been made in the literature to define the federal state (rather, according to the terminology adopted in this book, federal systems) based on institutional elements. The idea behind such an approach is simple and meaningful. While federalism always entails political and ideological elements and is thus also subjective,68 federal systems are concrete, constitutional manifestations of the federal principle; thus, a comparative observation should make it possible to determine the common elements, those that are essential in order to identify a federal system.

64 See G Poggeschi, Language Rights and Duties in the Evolution of Public Law (Baden-Baden, Nomos, 2013) 152. 65 See ch 4.2.1. 66 See ch 5.1.1. 67 See R Agranoff (ed), Accommodating Diversity: Asymmetry in Federal States (Baden-Baden, Nomos, 1999). 68 See chs 0.2 and 1.1.

48 Manifestations

2.D Institutional Definitions of Federal States (or Systems) Such an approach has indeed inspired several attempts to define federal states (or systems) by means of institutional elements. Among the most convincing institutional definitions are those provided by Ronald Watts and Peter Pernthaler. Watts proceeds from a political perspective and first distinguishes between federalism, federal political systems and federations. Federalism ‘refers to the advocacy of multi-tiered government combining elements of shared rule and regional self-rule’.69 ‘Federal political systems’ and ‘federations’ are instead descriptive terms: the former being broader and referring to a ‘political system in which … there are two (or more) levels of government’, thus encompassing the whole spectrum of non-unitary manifestations (confederations, federations, etc); the latter are a particular species of the genus ‘in which neither the federal nor the constituent units are constitutionally subordinated to the other, ie each has sovereign powers derived from the constitution rather than another level of government, each is empowered to deal directly with its citizens in the exercise of the legislative, executive and taxing power and each is directly elected by its citizens’.70 Based on this assumption, Watts lists six institutional elements that, in his view, are common to all federations: ‘1) two orders of government each acting directly on their citizens; 2) a formal constitutional distribution of legislative and executive authority and allocation of revenue resources between the two orders of government ensuring some areas of genuine autonomy for each order; 3) provision of the designated representation of distinct regional views within the federal policy-making institutions, usually provided by the particular form of the federal second chamber; 4) a supreme written constitution not unilaterally amendable and requiring the consent of a significant proportion of the constituent units; 5) an umpire (in the form of courts or provisions for referendums) to rule on disputes between governments; 6) processes and institutions to facilitate intergovernmental collaboration for those areas where governmental responsibilities are shared or inevitably overlap.’71 Pernthaler takes instead a more legal perspective, which handles concepts like sovereignty, citizenship and constitutional subordination more carefully.72 For him, the institutional minimum common denominator includes four elements: 1) division of statehood (sovereignty) mirrored by a constitutionally guaranteed division of powers that can only be changed with the consent of the member units; within the respective competences, the federation and member units are independent and equally sovereign (dual government); 2) effective participation of subnational entities in the decision-making process at the central level, especially with regard to the procedure to amend the constitution, to adopt national laws and to be represented in common institutions, such as the constitutional court or in the election of the head of the state; 3) constitutional autonomy of the subnational entities, ie, their power to autonomously make their fundamental choices within the limits of the national constitution; 4) a federal financial arrangement, ie, a division of taxation and spending powers that corresponds to the division of competences and includes solidarity among all members of the federation. For Pernthaler, these four elements are of an existential, functional nature: if they are not given in a polity, the latter cannot be considered to be federal.

69

Watts (n 3) 6. Watts (n 3) 7. 71 Watts (n 3) 7. See also RL Watts, ‘Comparative Conclusions’ in A Majeed et al (eds), Distribution of Powers and Responsibilities in Federal Countries (Montreal, McGill–Queen’s University Press, 2006). 72 P Pernthaler, Österreichisches Bundesstaatsrecht (Wien, Verlag Österreich, 2004) 299ff. 70

Federal State 49 All such definitions are extremely valuable and certainly useful from a comparative perspective as an orientation in the complex field of federal studies. From a comparative point of view, however, two aspects need to be considered as caveats against a possible all-encompassing institutional definition of a federal system. First, each definition based on a comparative analysis presupposes the selection of case studies and thus strongly influences the outcome of the research. Put differently, if certain countries are pre-identified as federal and thus only their constitutions are analysed, this leads to a very different outcome than if other cases are included too. For example, Watts considered in 2008 that there were 25 fully fledged federations, and these were considered as a comparative basis for his analysis,73 while Anderson listed 28 federations the year before.74 Such an approach is rather amphibolic, as it is likely to prove that the countries matching the definition are precisely those taken as the basis for the analysis. Second, federal constitutional arrangements are very peculiar to each and every country, while pursuing similar goals. Therefore, upon closer observation, each common element has in fact some exceptions. For example, the constitutional autonomy of subnational units is not given in some uncontested federations like Canada or Belgium and, except for Jammu and Kashmir, India, whose member units do not have their own codified constitutions in the sense of both formally and materially qualified acts of autonomous legislation.75 Typical constitutional matters are then essentially regulated by the national constitution and/or by national special laws.76 Conversely, in some regional states such as Italy, the regions adopt their own constitutions by means of entrenched laws. And in any case, the national constitution prevails over the constitutions of the subnational units. Another example is the effective participation of the subnational units in the decision-making process at the national level, often considered as a main institutional indicator of federal statehood.77 It is often argued, for example, that the adoption of the Seventeenth Amendment to the US Constitution, which mandated the popular election of senators in each state, transformed the representation from federal into democratic, de facto eliminating the genuine representation of the states.78

73 Austria, Belgium, Germany, Russia, Spain and Switzerland in Europe; Canada, the Federation of Saint Kitts and Nevis, Mexico and the United States in North America; Argentina, Brazil and Venezuela in South America; the Union of the Comoros, Ethiopia, Nigeria and South Africa in Africa; India, Pakistan and Malaysia in Asia; the United Arab Emirates in the Middle East; Australia, Palau and the Federated States of Micronesia in Oceania. See Watts (n 3) 9. 74 Argentina, Australia, Austria, Belau (Palau), Belgium, Bosnia and Herzegovina, Brazil, Canada, Comoros, the Democratic Republic of Congo, Ethiopia, Germany, India, Iraq, Malaysia, Mexico, Micronesia, Nigeria, Pakistan, Russia, St Kitts and Nevis, South Africa, Spain, Sudan, Switzerland, the United Arab Emirates, the United States of America and Venezuela. See G Anderson, Federalism: An Introduction (Oxford, Oxford University Press, 2007) 3. 75 See chs 5.1.1 and 5.1.2. 76 See C Saunders, ‘The Relationship between National and Subnational Constitutions’ in KonradAdenauer-Stiftung (ed), Subnational Constitutional Governance (Pretoria, KAS South Africa, 1999) 29. 77 For this assumption, see M Moushkeli, La théorie juridique de l’État fédéral (Paris, Pedone, 1931) 229ff. 78 See B Ackerman, ‘The New Separation of Powers’ (2000) 113 Harvard Law Review 633, and G Doria, ‘The Paradox of Federal Bicameralism’ (2006) 5 European Diversity and Autonomy Papers 1.

50 Manifestations And while some established federations have experienced severe problems with the effectiveness of their second chambers, especially in terms of the representativeness of the subnational entities (Belgium, Canada, India, Mexico, Austria, etc— Venezuela even abolished the Senate in 1999), some regional states have developed more effective forums for the institutional representation of the subnational units via intergovernmental conferences (Spain, Italy, to some extent even France), which are common also to some classical federal states such as Canada.79 This means, in turn, that institutional definitions are also incomplete because they are related to, and determined by, the historical evolution of each country. Not only federalism80 but also the federal system is a process, and any picture of it taken at a given moment cannot reflect the movement or the dynamic of the process. If federalism and the federal system were movies, any definition would be just a screenshot, thus inevitably partial and ultimately doomed to fail, although definitions are no doubt extremely useful for orientation and for knowing what the rule is in general. Definitions, in other words, only matter if they are taken critically, knowing that not only federalism, but also the federal state (and system), are abstract categories that should define very concrete phenomena. Thus, the only possible use of a definition is to provide a very general indication of what the phenomena entail. 2.3 REGIONAL STATE AND RELATED MANIFESTATIONS

2.3.1 Definition and Relation to the Federal State A regional state can thus somehow be defined a contrariis: the easiest and probably most precise definition is that it is neither a unitary nor a federal state. To say precisely what a regional state is, however, is not that simple, and it is even more difficult to draw a clear line between federal and regional states. In this respect, an effective description has been put forward by Peter Häberle. Noting that a difference exists between federal and regional states, although it is hardly possible to find decisive qualitative elements, he wrote that regionalism is the ‘little brother’ of federalism.81 This expression makes clear that both typologies produce the same phenomenon: vertically dividing power according to the rules provided by the national constitution. At the same time, due to the different historical origins (at least between regional states and classical coming-together federations), the different aims and the different traditions of the two categories, they are not entirely the same, while sharing the same genes. ‘Little brother’ is to be understood not only in quantitative terms, but—in addition to Häberle’s intuition—also in a time perspective: as a matter of principle, it is fair to say that subnational entities of regional states enjoy a lower degree of autonomy as compared to those of federal

79

See ch 6.2. CJ Friedrich, Trends of Federalism in Theory and Practice (New York, Praeger, 1968). 81 P Häberle, ‘Föderalismus, Regionalismus und Kleinstaaten in Europa’ in Europäische Rechtskultur: Versuch einer Annäherung in zwölf Schritten (Baden-Baden, Nomos, 1994) 257ff. 80

Regional State and Related Manifestations 51 states, although, as always, there are significant exceptions: there is no doubt, for instance, that Spanish autonomous communities or (some of the) Italian regions enjoy many more powers than, for instance, subnational units in countries such as Mexico, Venezuela, Russia, Austria and others. And certainly, the regional state is a far more recent creation (younger brother) than the federal state. In any event, whatever (if any) the difference between federal and regional states, in recent decades there has undoubtedly been a convergence between them: while federal states usually experience centripetal tendencies, with an increase in the power of the centre, in regional states the trend is towards further decentralisation.82 As a matter of fact, the qualitative difference, if one exists, is also being considerably reduced. 2.E Italy 2007: Autonomy and Sovereignty of the Sardinian People? It would be wrong to infer from the above that categories do not matter. An interesting example is provided by a judgment of the Italian Constitutional Court in 2007. In 2006, the regional Parliament of Sardinia initiated the procedure for the adoption of a new regional constitution, setting up a special commission called the Commission for Drafting the New Basic Law on Autonomy and Sovereignty of the Sardinian People. The name suggested, on the one hand, the existence of a Sardinian people as different from the Italian people. On the other hand, it indicated that the Sardinian people could enjoy ‘sovereignty’, ie, that the autonomy of the region does not derive from the constitution but from a free determination of the sovereign Sardinian people, which could also decide differently if they so wish. The terminology used was clearly aimed at provoking a debate. As expected, the national government immediately challenged the Sardinian law, and the Constitutional Court found it in breach of the constitution.83 The Court offered a formalistic definition of the qualitative difference between a federal and a regional system, pointing out that, in the latter, subnational entities are autonomous and not sovereign. Even the most significant degree of decentralisation does not alter the unity and indivisibility of sovereignty, which cannot but be vested in the state and in its unitary people. Definitions, thus, sometimes matter, even though it might be contended that in federal states there would have been the same outcome (although maybe not the same reasoning).

2.3.2 Rationale and Cases Like the federal states, so-called regional states are also very difficult to define. Precisely like the federal state, the regional state is an umbrella term used to describe the same phenomenon with a different intensity. From a historical perspective, the very idea of a regional state developed much later and to some extent in opposition to that of the federal state. When the age

82 According to L Hooghe et al, The Rise of Regional Authority: A Comparative Study of 42 Democracies (1950–2006) (New York, Routledge, 2010), between 1970 and 2005 only two states became more centralised, while many experienced an increase in what the authors call ‘regional authority’. 83 Italian Constitutional Court No 365/2007.

52 Manifestations of coming-together federations came to an end, after World War I, regionalisation processes started by means of the decentralisation of unitary polities. Such decentralisation, in the mind of its advocates, should precisely not go as far as to create a federation. Instead, it should help establish a polity in between the federation— which was considered too decentralised and was historically maybe no longer feasible—and the unitary state, which was far too centralised and could no longer accommodate the ethno-cultural, geographical, political, social and economic diversities. In most cases, such a middle ground between the unitary and the federal state was an experiment with the main aim of accommodating ethno-cultural differences, such as in the case of Spain and of (pre-federal) Belgium. In other cases, the accommodation of ethno-cultural differences was just one out of many other reasons for the establishment of a regional state, like in Italy. In any case, these countries, however defined, have experimented with ‘innovative federal and quasi federal forms’ that reflect different ‘degrees of centralization or decentralization’.84 Spain is usually considered the main laboratory of the regional state. In fact, during the nineteenth century, several attempts were made to break the centralisation of the Napoleonic system. However, the prototype of a regional state, in between a federation and a unitary state, aimed in particular at guaranteeing ethnocultural differences, was developed by the Republican Constitution of 1931. Then, the idea was that of an ‘integral state’ (this was the terminology of the Constitution), compatible with a substantial autonomy of municipalities and regions. The Constitution of 1931 was trumped by the Franco regime, which repressed all forms of autonomy, but it served as an essential precedent (together with the German and the Italian models) for the democratic Constitution of 1978, adopted after the death of Franco. In response to the strong nationalistic movements in the Basque Country, Catalonia and Galicia, a form of so-called pre-autonomic regime was extended to the whole Spanish territory between 1977 and 1978. The Constitution of 1978 reached a difficult compromise between unitary and nationalistic instances, by defining a process but not a clear end: the process of decentralisation was left open, and only a procedural framework was established in which different forms of autonomy could be realised according to the needs and specificities of each territory.85 While proclaiming the indissoluble unity of the nation, the constitution ‘recognizes and guarantees the right to autonomy of the nationalities and regions’ (Article 2 of the Spanish Constitution). The regionalising process set up by the constitution moves from the bottom up and provides two procedures for achieving regional autonomy: Article 143 outlines the so-called ‘slow track’ to form regions (autonomous communities), by stating that neighbouring provinces with common historical, cultural and economic elements, the islands and the provinces of historical regional relevance can form autonomous communities. The ‘fast track’ outlined in Article 151 was instead

84 85

Watts (n 3) 5. See JJ González Encinar, El Estado unitario–federal (Madrid, Tecnos, 1985).

Regional State and Related Manifestations 53 reserved for the territories where historical nationalities were settled86 and those, in fact only Andalusia, that completed a complex and lengthy procedure, including a referendum. These territories were allowed to proceed immediately to the establishment of autonomous communities. The difference between these autonomous communities and those on the slow track, lies not only in the procedure for their establishment, but also in the attribution of powers: fast-track autonomous communities could immediately seize from the centre all powers foreseen in Article 149 of the Spanish Constitution, while other autonomous communities were for the initial five years limited to the competences listed in Article 148. Such a strongly asymmetrical design of the Spanish regional state was progressively reduced, although the autonomous communities of the historical nationalities still enjoy wider powers than the others in several fields, including local policing and, for the Basque Country, financing.87 2.F Spain 1983: The State of Autonomies and the LOAPA Decision Following the attempted coup in February 1981, the process of regionalisation was put under stricter control by means of a controversial Organic Law on the Harmonisation of the Autonomy Process (LOAPA). This law, based on Article 150(3) of the Spanish Constitution,88 stipulated, inter alia, that the slower way of regionalisation was to be preferred, and it introduced strong elements of oversight by the central authorities. The law was scrutinised by the Constitutional Court in 1983, and the future development of the just-born state of autonomies (Estado de las autonomías) depended on the decision. The ruling struck down significant parts of the LOAPA (14 out of 38 Articles) and affirmed that the law in fact did not achieve the goal of harmonising the autonomy process and that the national government is not allowed to unilaterally reshape the distribution of powers.89 Among the provisions declared unconstitutional were those stipulating that the state’s norms should take automatic precedence over those of the autonomous communities and that civil servants in the regions should be seconded by the state rather than being recruited locally. The ruling had the effect of setting the autonomy process in motion.

While decentralisation in Spain has gone as far as to suggest that it might already be a federal state, although a ‘federation in disguise’,90 Italy is another classical example of a regional state.91 Since the achievement of national unity, completed in

86 According to the constitutional terminology, these were those territories ‘which in the past approved by referendum their autonomy statutes and that at the time of entry into force of this constitution enjoy provisional autonomy regimes’, Spanish Constitution, Transitional Provision No 2. 87 See ch 7.1.3. 88 On this provision, see ch 5.2.3. 89 STC 76/1983. 90 L Moreno, ‘Federalization in Multinational Spain’ in M Burgess and J Pinder (eds), Multinational Federations (London, Routledge, 2007) 86–107. 91 For an overview, see F Palermo, ‘Italy: A Federal Country without Federalism?’ in Burgess and Tarr, Constitutional Dynamics (2012).

54 Manifestations the 1860s, the Italian state has been modelled according to the French blueprint of a centralised and bureaucratic state. It was only with the Republican Constitution of 1948 that an innovative, but at the same time feeble, experiment with regionalisation was carried out. Not unlike Spain, but for different reasons, Italian regionalism is characterised by its asymmetrical design, both as a matter of constitutional law and in terms of the effective use of powers transferred to the regions. Despite the constitutional provisions for a general regional design for the whole country, at first, only five special, or autonomous, regions were established, all situated in the periphery (three in the Alpine arch in the north, with consistent minority groups, Aosta Valley, Trentino-South Tyrol, Friuli-Venezia Giulia, and the two main islands of Sicily and Sardinia). Their differentiated treatment was mainly a reaction to the complex problems of regional diversity: international obligations imposed by the peace treaty and fears regarding the secession of these peripheral areas. Each of them is guaranteed by a special statute, a basic law with constitutional rank. As an innovative experiment, the regionalisation of the whole country, a third way between a federal and a unitary system, was aimed at avoiding too strong an asymmetry between the special regions and the rest of the territory. However, the regional design was only fully developed in the 1970s, when the ordinary regions were established and legislative powers effectively devolved. Since then, a permanent increase in the regional powers gradually narrowed the gap between ordinary and special regions. The path has been anything but straightforward, influenced by shifting political priorities and very much determined by constitutional adjudication. After a series of reforms in the 1980s and 1990s, two constitutional amendments were approved in 1999 and 2001 that considerably increased the powers of the (ordinary) regions. The first reform introduced the direct election of regional presidents in order to enhance political stability in the ordinary regions. It also strengthened their constitutional autonomy, as the regional basic laws are now to be adopted by the ordinary regions themselves in a special procedure (double approval and a possible referendum). The second reform of 2001 completely reshaped the constitutional provisions on relations between the state and the regions. It drastically changed the distribution of legislative and administrative powers between the state and the regions: Article 117 of the Italian Constitution now lists all legislative powers of the state, as well as the fields of concurrent legislation (ie, those in which the regions can legislate only within the framework of general guidelines determined by a national law). By contrast with the situation before, the residual powers lie with the regions according to classic federal schemes. Administrative powers are no longer connected with legislative ones but are distributed in a flexible manner according to the criteria of ‘subsidiarity, differentiation and proportionality’ (Article 118). The new provision on fiscal federalism provides for the partial financial autonomy of subnational entities (Article 119), and preventive state control over regional legislation was abolished, although the state retained the power to replace regional norms with a view to guaranteeing the legal and economic unity of the system (Article 120). There are several other relevant examples of regional states, although, quite interestingly, nearly all of them are in Europe. Among them, some vest subnational units with legislative powers, and others only with (albeit broad) administrative

Regional State and Related Manifestations 55 powers. The former include Belgium with subnational entities having had the power to enact (quasi)laws92 even before 1993—when it officially became, by constitutional denomination, a federal state—and, to a more limited extent, Serbia (the autonomous province of Vojvodina can pass legislation). The latter range from the case of France after the constitutional reform of 2003, when the principle of the ‘unitary decentralized state’ was introduced in the Constitution (Article 1) and the regions were given very broad administrative powers, including in important fields that are usually regulated by law. Further, usually temperate examples of regional states are to be found in some Central, Eastern and South-eastern European countries.93 From a comparative perspective, some elements common to all regional states can be identified. First, power also historically proceeds from the centre and is gradually devolved to the periphery. Second, the distribution of powers and financial prerogatives among the regions is asymmetrical: such asymmetry may vary over the course of time but it remains a structural factor of regional states. Third, the regionalising process is often prompted by the need to accommodate ethno-cultural differences: while this is not the only reason for the regionalisation of a state, it is definitely one of the main ones. Other motives are linked with the complexity of modern governance and the factual impossibility for one single level of government to cope with such complexity. Fourth, being the regional state initially conceived as a compromise between two categories with longer traditions—the unitary and the federal state— it is more dynamic overall, and it undergoes processes of reform of the territorial setting more often. Fifth, from a terminological point of view, while many constitutions define the respective state as ‘federal’ (even when this is all but uncontested), no constitution defines itself as ‘regional’. 2.3.3 Devolution The term ‘devolution’ is becoming more commonly used in contemporary constitutional and political jargon. As a general phenomenon, it indicates the transfer of public functions from the centre to subnational elected authorities. However, it is also used especially in the peculiar context of the United Kingdom, where it has been part of the constitutional discourse since at least the 1970s. In the UK context, ‘devolution’ describes the process of the decentralisation of powers from the central government. According to the probably most cited definition provided by Vernon Bogdanor, devolution means ‘the transfer of powers from a superior to an inferior political authority’94 and it consists of three elements: a transfer to a subordinate but

92 These acts of the subnational legislatures are called decrees and have the same legal force as laws passed by the federal parliament. Only in the case of the Brussels Parliament these acts are referred to as ordinances which have a slightly inferior legal status. The federal government may suspend ordinances if it regards them as compromising the role of Brussels as the capital of Belgium or its international role. 93 See F Palermo and S Parolari (eds), Regional Dynamics in Central and Eastern Europe: New Approaches to Decentralization (Leiden, Brill–Nijhoff, 2013). 94 V Bogdanor, Devolution in the United Kingdom (Oxford, Oxford University Press, 2001) 2—a definition that was provided for the process in the UK but would also suit the phenomenon in general.

56 Manifestations elected body, a territorial/geographical basis and functions previously exercised by ministers and parliament. The parts of the UK covered by such a transfer of powers include Scotland, Northern Ireland and Wales, while attempts to extend devolution to England proper failed after a referendum in the north-east in 2004. This has sparked the still unanswered ‘English question’, which refers to the interrelated issues of England’s place in the Union and of decentralisation within it.95 In the 1970s, the idea of devolution was developed by nationalists as an update of the nineteenth century’s ‘home rule all round’, ie, a fully developed selfgovernment.96 Under the Labour Government of 1974–79, two bills were proposed (the Scotland Bill and the Wales Bill) but were rejected by referendums held in the affected areas. When Labour returned to power in 1997, devolution was an essential part of the electoral agenda, and in 1998 four bills were adopted by Westminster and confirmed in local referendums: the Scotland Act, the Government of Wales Act, the Northern Ireland Act and the Greater London Authority Act. Albeit with successive, significant amendments, these acts are still in force. In terms of the entrenchment of these acts, some scholars even assume that they could be repealed and amended only by a law supported by a referendum.97 While procedurally entrenched, devolution is legally speaking a reversible process: like any national law in the UK, in theory the Westminster Parliament could, at any time, withdraw it by simple majority. As unlikely as that would be, this is an essential element that distinguishes (British) devolution from other forms of the regional state. This aspect should, however, not be overestimated. It has to do mostly with the essential elements of British constitutionalism, which presupposes unlimited sovereignty of the parliament.98 Yet, British courts have affirmed the existence of so-called ‘constitutional statutes’, among them the devolution acts. 2.G United Kingdom 2002: Devolution and Factual Limits on the Sovereignty of Parliament? In Thoburn v Sunderland City Council, the Divisional Court advanced the existence of ‘constitutional statutes’.99 In particular, Lord Justice Laws writing for the Court identified the Magna Carta 1215, the Bill of Rights 1689, the Acts of Union 1707, the Reform Acts, the Human Rights Act 1998, the Scotland Act 1998, the Government of Wales Act 1998 and the European Communities Act 1972 as constitutional acts that are, because of their constitutional importance, to be protected from the implied repeal rule, according to which later acts of legislation take precedence over earlier acts and result in the conflicting

95 See M Burch et al, ‘Devolution, Change and European Union Policy-making in the UK’ (2005) 39 Regional Studies 465; R Hazell (ed), The English Question (Manchester, Manchester University Press, 2005). 96 See K Robbins, Great Britain: Identities, Institutions and the Idea of Britishness since 1500 (The Present and the Past) (London, Routledge, 1998) 206ff. 97 See B Hadfield, ‘Devolution: A National Conversation?’ in J Jowell and D Oliver (eds), The Changing Constitution (Oxford, Oxford University Press, 2011) 218. 98 See AV Dicey, The Law of the Constitution, 10th edn (London, Macmillan, 1965) 39. 99 Thoburn v Sunderland City Council [2002] EWHC 195 (Admin), (Metric Martyrs case).

Regional State and Related Manifestations 57

parts of the earlier act being repealed. In other words, ‘constitutional statutes’ can only be repealed explicitly. What is uncontested, in any case, is the supremacy of the British institutions over the devolved ones. Since the establishment of the UK Supreme Court under the Constitutional Reform Act 2005 and the start of its work in 2009, this judicial body has the power to adjudicate conflicts regarding devolution.

Parliamentary sovereignty and devolution were also, in the context of so-called ‘Brexit’, an issue for the courts. The High Court ruled in November 2016 that the UK Secretary of State does not have power under the Crown’s foreign affairs prerogative to formally notify the Council of the EU about the country’s withdrawal (Article 50 of the Treaty on European Union), but required an authorising act by the sovereign UK Parliament, not by the devolved legislatures.100 The Supreme Court upheld this judgment in January 2017, albeit with a partially different reasoning,101 and after the adoption of the European Union (Notification of Withdrawal) Bill 2017 the United Kingdom eventually invoked Article 50 on 29 March 2017. Another essential element of devolution is its inherent asymmetrical character: each affected area has its own institutional system, its own powers and its own cooperation procedures with the central government.102 This depends on the very different reasons behind each individual devolution process. In the UK, Scottish devolution is the product of autonomous historical constitutional traditions that have never ceased to exist; Welsh devolution is the constitutional answer to ethnocultural claims; the devolution in Northern Ireland is an attempt to provide an institutional and constitutional answer to dramatic social and political cleavages; and even the ‘quasi-devolution’ of the Greater London Authority is aimed at providing solutions to complex governmental issues common to many big cities.103 At the same time, asymmetry is largely compensated by the unity of the civil service: apart from in Northern Ireland, there is a single and unitary civil service in the UK, which guarantees uniformity in interpretation and implementation of the rules and, even more importantly, a unified administrative culture. This contributes significantly to the effectiveness of the system in spite of this being largely regulated by conventional rules, including in sensitive matters such as financial relations. As to the relationship between devolution and other concepts, some authors consider devolution as a functional element and sometimes a prerequisite to

100

R (Miller) v Secretary of State for Exiting the European Union [2016] EWHC 2768 (Admin). R (on the application of Miller and Dos Santos) v Secretary of State for Exiting the European Union [2017] UKSC 5. 102 Bogdanor, Devolution in the United Kingdom (2001); J Bradbury and J Mawson (eds), British Regionalism and Devolution: The Challenges of State Reform and European Integration (London, Routledge, 1997). 103 Greater London is, legally speaking, part of the local government and thus not included in devolution in the strict sense. The phenomenon, however, is similar and the debate as to whether the Greater London Authority (GLA) is to be considered part of devolution or not resembles the one as to whether federalism and regionalism are the same phenomenon or not. Interestingly, in 2011 the Greater Manchester Combined Authority (GMCA) was also established. 101

58 Manifestations federalisation,104 while others believe that devolution and federalisation are two antithetical constitutional processes.105 From a comparative observation, it can be affirmed that devolution very much resembles the process of regionalisation. The ‘devolutionary state’ is in some way the British version of the regional state, especially because devolved powers are delegated powers. Devolution has been effectively described as ‘the delegation of central government powers to subordinate units, these powers being exercised with some degree of autonomy though with ultimate power remaining with the central government’.106 Devolution is in sum an advanced form of territorial decentralisation, which nevertheless does not change the fundamental constitutional nature of the state. Especially in its most advanced Scottish form, devolution goes as far as possible without crossing the border of national sovereignty.107 2.3.4 Autonomy and Similar Manifestations ‘It is fair to claim that no clear account of the concept of autonomy is available.’108 Another slippery issue is the demarcation between regional/devolved states and autonomy. The latter is a very broad concept that appears in different branches of scholarship, not only in law and politics, but also in philosophy and the natural sciences. In law, Ruth Lapidoth identifies four conceptions of autonomy: a) the right to act upon one’s own discretion in certain matters; b) a synonym of independence; c) a synonym of decentralisation; d) the exclusive power of legislation, administration and adjudication in specific areas of an autonomous entity.109 Moreover, autonomy has different meanings if used with regard to constitutional law or international law or if attributed to territorial entities or to groups of people (territorial or personal autonomy).110 For our purposes, (territorial) autonomy (of a territorial entity) can be considered a synonym of decentralisation, ie: [G]ranting internal self-government to a region … thus recognizing a partial independence from the influence of the national or central government. This independence can be determined primarily by the degree of actual as well as formal independence enjoyed by the autonomous entity in its political decision-making process.111

104

See J Kendle, Federal Britain: A History (London, Routledge, 1997). MA Fazal, A Federal Constitution for the United Kingdom: An Alternative to Devolution (Aldershot, Dartmouth Publishing, 1997). 106 P Norton, ‘Devolution: A Threat to the UK? Or a Reinforcement?’ in P Norton (ed), The Constitution in Flux (Oxford, Blackwell Publishers, 1982) 174. 107 See C Himsworth and CR Munro, The Scotland Act 1998 (Edinburgh, Sweet & Maxwell, 1999). 108 M Wiberg, ‘Political Autonomy: Ambiguities and Clarifications’ in M Suksi (ed), Autonomy: Applications and Implications (The Hague, Kluwer Law International, 1998) 43. 109 See R Lapidoth, ‘Autonomy: Potential and Limitations’ (1993) 1 International Journal of Minority and Group Rights 269, esp 277. 110 See TH Malloy and F Palermo (eds), Minority Accommodation through Territorial and NonTerritorial Autonomy (Oxford, Oxford University Press, 2015). 111 H-J Heintze, ‘On the Legal Understanding of Autonomy’ in Suksi, Autonomy (1998) 7. 105 See

Regional State and Related Manifestations 59 For that matter, autonomy is a constitutionally guaranteed form of decentralisation of state functions, precisely like the regional state. As seen above, the same also applies to federations, whereby the powers enjoyed by the constituent units are—in a purely constitutional perspective—nothing other than a form of autonomy. Thus, what makes autonomy different from federations and from regional states? While, broadly speaking, autonomy can be considered an overarching concept comprising all constitutional forms of the vertical division of powers, strictly speaking, the term is usually used to identify the self-government enjoyed by a certain, specific region. In other words, autonomy usually does not refer to all parts of the national territory, but only to some, denoting self-government for one or more particular regions.112 These might be the only effectively self-governing territories in one country, as it is the case in Finland for the Åland Islands, in Denmark for Greenland, in Portugal with respect to the Azores and Madeira, in Moldova with Gagauzia, in the United Republic of Tanzania with Zanzibar, etc, or some regions enjoying a particular status as compared to others in the same country, like the so-called ‘autonomous’ or special regions in Italy, Vojvodina in Serbia, etc. The aforementioned examples illustrate that a key driver behind the establishment of such asymmetrical autonomy arrangements is often ethno-cultural diversity113 or, sometimes, also the remoteness of an area,114 and in certain cases like Greenland a combination of both. Simplifying, and not engaging in complex debates about the multifaceted meanings of autonomy,115 this limitation to only certain areas of a national territory can be considered the distinctive element of regional states and autonomous entities. For this reason, autonomy is considered in this book under the heading of the regional state. The same goes for other concepts related to the phenomenon of autonomy in the broad sense, such as federacies and associations of states. Federacies are defined as political arrangements between a larger unit and one or more smaller unit(s). The latter retain substantial autonomy, participate at least at a minimum level in the government of the larger unit, and the relationship can be dissolved only by mutual agreement.116 The best-known example of a federacy is the relationship linking Puerto Rico and other unincorporated territories to the United States. Puerto

112 For a differentiation between the terms autonomy and self-government (which are usually used as synonyms) see Y Dinstein, ‘Autonomy’ in Y Dinstein (ed), Models of Autonomy (New Brunswick, Transaction Press, 1981). 113 It must be pointed out that autonomy is indeed often seen as autonomy (from the state) for a particular group and is therefore often used in minority-rights discourse, where it is intended to be synonymous with internal self-determination. See, eg A Eide, ‘The Universal Declaration in Space and Time’ in J Berting et al (eds), Human Rights in a Pluralist World: Individuals and Collectivities (Middelburg, Roosevelt Study Center, 1990) esp 25. 114 See E Hepburn, ‘Recrafting Sovereignty: Lessons from Small Island Autonomies?’ in A Gagnon and M Keating (eds), Political Autonomy and Divided Societies: Imagining Democratic Alternatives in Complex Settings (Basingstoke, Palgrave Macmillan, 2012). 115 See, eg, Suksi (n 108). Others speak of ‘autonomism’ as a peculiar ideology in many respects different from, and even opposed to, federalism. See J Lluch, ‘Autonomism and Federalism’ (2012) 42 Publius 134. 116 See RL Watts, The Institutions of a Federal State: Federalism and Democracy as Fundamental Counterweighing Principles (Fribourg, Institut du fédéralisme, 1996) 4.

60 Manifestations Ricans, in particular, share all the rights and obligations of US citizenship, including service in the armed forces. At the same time, they do not pay federal taxes and cannot vote in national elections. The foreign affairs of the islands are handled by the US Government. Also, the unusual constitutional relationship within the Kingdom of the Netherlands between the Netherlands and the former Dutch Antilles (Curacao, Sint Maarten and the already autonomous Aruba) after the reform from 2010 might be considered a form of federacy, although formally all four countries have an equal, constitutive nature.117 An association of states is another form of privileged linkage between sovereign states of different size and weight. The essential elements are very similar to federacies, the only substantial difference being that associations, unlike federacies, ‘can be dissolved by either of the units acting alone on prearranged terms’.118 An example of an association of states is the constitutional link between the Cook Islands and New Zealand based on the Cook Islands Constitution Act 1964 (an act of the Parliament of New Zealand) as amended in 1965, setting up an association agreement. This entails full executive power for the government of the islands, full legislative power unless the New Zealand Parliament is authorised be the islands’ government to pass legislation for the islands and New Zealand citizenship for Cook Islanders. Other typologies are sometimes dealt with in the literature due to their lato sensu linkage with a federal-type relationship. According to the widely used definition developed by Elazar and Watts, the above-mentioned ‘varieties of federal arrangements’ and ‘federal political systems’,119 respectively, include condominiums, leagues, joint functional authorities and other hybrids. Condominiums would be ‘political units which function under the joint rule of two or more external states in such a way that the inhabitants have substantial internal self-rule’, as was the case in Andorra under the joint rule of France and Spain until 1993.120 Leagues are even weaker linkages than confederations, are established for limited or just one purpose only (usually defence) and unilateral withdrawal is always possible. Contemporary examples mostly include intergovernmental organisations such as the Association of Southeast Asian Nations (ASEAN), the Arab League, the African Union, etc. Joint functional authorities would be agencies established by two or more states ‘for joint implementation of a particular task’: these include, in Watts’ opinion, several international organisations, such as the International Labour Organization (ILO), the International Atomic Energy Agency (IAEA), etc, but also many of the forms of cross-border cooperation set up by subnational entities especially in Europe. Needless to say that these classifications serve just the purpose of creating a common interpretative framework for phenomena somewhat linked with the federal idea

117 See HG Hoogers, ‘UPG voor Curacao, Aruba en St. Maarten’ in HE Bröring et al (eds), Schurende rechtsordes: Over de Europese Unie, het Koninkrijk en zijn Caraibische gebieden (Groningen, Europa Law Publishing, 2008) 119ff. 118 See Watts, The Institutions (1996) 4. 119 See Elazar, Federal Systems (1991) XVI; Watts (n 3) 8f. 120 See Loughlin et al, The Routledge Handbook (2013) 23.

The EU: Federalism Beyond the State? 61 of states associating in order to do something together, and are, of course, very subjective. Categorisation is, however, all the more contested when it comes to such residual concepts. In the literature, it is not uncommon to find different classifications for the same cases.121 2.4 THE EU: FEDERALISM BEYOND THE STATE?

There has been much scholarly debate about the legal nature of the European Union, as well as about its federal traits.122 Undoubtedly, the federal idea has often been linked to the process of European integration, even long before the actual process started, and in fact the idea of devising institutional solutions to safeguard peace on the European continent has occupied many great thinkers for centuries. 2.H Ideas of Europe: Kant, Spinelli and Monnet In his ground-breaking Perpetual Peace, Immanuel Kant portrayed in 1795 the state of nature among states as characterised by war. Since peace is nothing to be taken for granted, it is something to be actively established and guaranteed through law. This international legal order, he claimed, ‘shall be based on a federalism of free states’.123 Kant advocated ‘a league of nations, which, however need not be a state of nations’. This foedus pacificum, a league of a special kind, differs in its level of integration from both a world republic and a mere peace treaty (pactum pacis). While the latter ‘seeks to end only one war’ and is therefore rather a ceasefire agreement, the foedus pacificum ‘seeks to end all war forever’.124 What Kant had in mind when calling for ‘a federalism of free states’ is not a world republic organised as a federal state but a multilateral league for peace. That federalism would gain prominence in the context of peace-building after World War II was anything but predictable before it. Then it was widely regarded as a conservative concept and as such highly controversial. Others were downright contemptuous of it and proclaimed the ‘obsolescence of federalism’.125 However, during and after World War II, the idea of federalism as a tool for peace in Europe gained momentum. In 1941, Altiero Spinelli and Ernesto Rossi, exiled on the island of Ventotene, wrote the manifesto Towards a Free and United Europe (Ventotene manifesto), in which they warned that after the War ‘national systems … will

121 For example, Greenland, Åland, Madeira and the Azores are seen by some (mostly Europeans) as autonomies, see Wiberg, ‘Political Autonomy’ (1998), and by others (especially from outside Europe) as federacies, see Watts (n 116) 5, and Lluch considers Puerto Rico a non-federal autonomy rather than a federacy, a concept he considers misplaced in federal studies, J Lluch, ‘Federacies and Conceptual Stretching: A Critique of the Category of ‘Federacy’ (Political Studies Association Annual Conference, London, April 2011). 122 See, eg, K Nicolaidis and R Howse (eds), The Federal Vision: Legitimacy and Levels of Governance in the United States and the European Union (Oxford, Oxford University Press, 2001) and D McKay, Designing Europe: Comparative Lessons from the Federal Experience (Oxford, Oxford University Press, 2001). 123 I Kant, ‘Toward Perpetual Peace’ [1795] in M Gregor (ed), Practical Philosophy (Cambridge, Cambridge University Press, 1996) 325. 124 ibid 327. 125 See H Laski, ‘The Obsolescence of Federalism’ (1939) 98 The New Republic 367, 367f.

62 Manifestations

ostentatiously begin to reconstruct the old State institutions’.126 This resurgence of nation states would inevitably lead to another war. Therefore, they argued, self-sufficiency has to be replaced by economic cooperation and military as well as political integration of all democratic states in a European federation. The manifesto envisaged ‘a steady federal state, that will have at its disposal a European armed service instead of national armies; that will break decisively economic autarkies, the backbone of totalitarian regimes’.127 The revolutionary ideas of the Ventotene manifesto were diametrically opposed to Jean Monnet’s preference for a functional, intergovernmental and incremental approach. As the head of the General Planning Commission, Monnet was in charge of reconstructing the French economy after World War II and became the driving force behind a plan that was eventually presented in the famous Schuman Declaration: ‘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. The coming together of the nations of Europe requires the elimination of the age-old opposition of France and Germany. Any action taken must in the first place concern these two countries … the French Government proposes … that Franco-German production of coal and steel as a whole be placed under a common High Authority, within the framework of an organization open to the participation of the other countries of Europe.’128 The initiative of merging industries so that war in Europe would become materially impossible embodies Monnet’s functionalist approach. Integration in this first sector, accomplished with the founding of the European Coal and Steel Community in 1952, was supposed to trigger integration in other sectors. These functional links, primarily of an economic nature, were expected to produce tangible benefits that would create, in turn, de facto solidarity among Europeans and a positive attitude towards an ever more united Europe on rational grounds, which should be the culmination of a gradual integration process.

While the federal idea has accompanied the evolution of the European integration process since before its start, a European federation has so far not materialised, nor is this likely to happen in the foreseeable future. Even the attempt to apply constitutional discourse and semiotics to the European Union—while largely symbolic— failed when the ratification of the Treaty establishing a Constitution for Europe (so far the closest to a federal big bang) was rejected by the people of France and the Netherlands in referendums of 2005. And while the political and academic debate has for a long time been polarised between ‘federalists’ and ‘intergovernmentalists’, it seems that in some way both have won and both have lost, since both elements are clearly present in today’s EU. Importantly, for our purposes, the issue of the legal nature of the European Union and of its relationship with the legal systems of the Member States has also been

126 See A Spinelli and E Rossi, ‘Towards a Free and United Europe: A Draft Manifesto’ (June 1941) II para 2 (The Ventotene Manifesto). 127 ibid para 21. 128 R Schuman, ‘The Schuman Declaration—9 May 1950’, available at europa.eu/abc/symbols/9-may/ decl_en.htm.

The EU: Federalism Beyond the State? 63 addressed by several courts, notably the Court of Justice of the European Union (CJEU) and the constitutional or supreme courts of most Member States. The CJEU has inferred from the Treaties that the duty to give full effectiveness to the European Treaties implied the precedence of EU law over any domestic source of law129 and that European law should have direct effect in the domestic legal systems of the Member States,130 thus also protecting the fundamental rights of individuals as unwritten general principles of European law.131 Other principles developed by the Court, which are likewise typical of a federal system, are now even expressly codified in the primary law of the European Union (TEU and TFEU) and thus accepted by the Member States. Cases in point are regarding the distribution of powers, the principles of delegated competences, subsidiarity, proportionality (all in Article 5 TEU) and the principle of loyal cooperation (Article 4(3) TEU). National courts, however, have generally taken a different approach, aimed at guaranteeing the predominance of their national constitution or, more precisely, of the fundamental, essential principles thereof. Most courts thus adopt a so-called dualistic view of the relationship between the domestic and the European legal order, according to which the two systems form separate orders. However, should they exceptionally collide, EU law can prevail except when it comes to the essential constitutional principles of the domestic constitution, those that incorporate the constitutional identity of the state. In case such principles are affected, the supremacy of the domestic constitution must be guaranteed by the national constitutional court. This doctrine is normally called the theory of counter-limits, meaning that domestic constitutions accept being limited by EU legal integration, with the only limitation regarding their vital elements.132 Many EU Member States have included specific clauses in their own constitutions reflecting these positions, essentially prescribing that European integration should be based on the same fundamental values and organisational principles laid down in their domestic constitution (the most glaring example of this type of clause being Article 23 of the German Basic Law). 2.I Germany 1993-2014: What Is the European Union? As to the very nature of the European Union, while the CJEU has stated that it represents ‘a Community of Law’ based on the rule of law and on the same foundational principles of the constitutions of the Member States,133 some domestic courts have clearly affirmed that the relationship between European and domestic law is different in nature as compared to that of federal systems. In spite of the precedence (as a rule) of EU law over domestic law, the EU still remains an international organisation, albeit a peculiar one. According

129

Case 6/64 Costa v ENEL [1964] ECR 1194. Case 26/62 Van Gend en Loos v Administratie der Belastingen [1963] ECR 3. 131 Settled case law since Case 29/69 Stauder v City of Ulm [1969] ECR 419 and Case 11/70 Internationale Handelsgesellschaft v Einfuhr- und Vorratsstelle für Getreide und Futtermittel [1970] ECR 1125. 132 73 BVerfGE 339 (Solange II); Italian Constitutional Court No 170/1984; French Council of State, case Nicolo (1989). 133 Case 294/83 Les Verts v Parliament [1986] ECR 1339. 130

64 Manifestations

to the judges it is what they famously called a Staatenverbund (compound of states), and is not a single European (federal) state.134 Moreover, the transformation into a federation would not be possible because this would imply a change of the essential, non-negotiable constitutional principles and values of the domestic constitution and thus be impeded by the counter-limits to European integration mandated by the national constitution.135 More recently, the German Constitutional Court has gone as far as to acknowledge that some measures adopted by the European Central Bank to respond to the financial crisis might be ultra vires (although it referred the question to the CJEU for a preliminary ruling)136 and has reiterated its approach according to which counter-limits represent an insurmountable obstacle to the process of European integration.137

In fact, it is uncontested that in many aspects the European Union works like a federation. Principles such as the supremacy of EU law over domestic law, the direct effect of EU law, the exclusive interpretation by the Court of Justice of the EU regarding conflicts of attribution between the EU and the Member States, the institutional representation of the Member States in the bodies of the EU, and several other elements are examples of this.138 And even more generally, there is clearly a vertical division of powers between the EU, the Member States and in several cases their subnational entities,139 which indicates that the European Union is in the broader sense part of the federal phenomenon. Looking not at federalism but at the federal state and other legal manifestations of the vertical division of power, as this chapter is doing, the debate about the federal traits of the European Union is not extraordinarily relevant. What needs to be stressed, however, is the presence of federal elements in the process of European integration, which are overall reminiscent of a confederation, and the fact that this process does not seem to be leading towards a fully-fledged federation,140 thus confirming the assumption that the time of federations through aggregation is over. 2.5 LIVING WITHOUT DEFINITIONS

In summary, as there is no agreed definition of federalism—nor can there be one— the federal state is also extremely difficult to define. Even more debatable is the legal

134

89 BVerfGE 155 (Maastricht). 123 BVerfGE 267 (Lisbon). 136 134 BVerfGE 366 (Outright Monetary Transactions). 137 135 BVerfGE 317 (European Stability Mechanism). See box 7.B. 138 See also M Burgess, ‘Territoriality and Federalism in the Governance of the European Union’ in M Burgess and H Vollaard (eds), State Territoriality and European Integration (London, Routledge, 2006). 139 There is an immense amount of literature on the role of the subnational entities and the three-tiered vertical division of powers in the EU. See especially S Weatherill and U Bernitz (eds), The Role of Regions and Subnational Actors in Europe (Oxford, Hart Publishing, 2005). 140 See A Lollini and F Palermo, ‘Comparative Law and the ‘Proceduralization’ of ConstitutionBuilding Processes’ in J Raue and P Sutter (eds), Facets and Practices of State-Building (Leiden, Martinus Nijhoff, 2009). 135

Living Without Definitions 65 distinction between federal and regional states and, to some extent, also between other categories, all of which are nonetheless somehow related to the federal idea. Each distinction ultimately depends on the criteria used for comparison: if institutional criteria are employed, the classification of federal, regional or other types of states greatly differs from what results from a classification based on history (old vs recent federal systems), or on trends (aggregative vs devolutionary), or on quantitative or qualitative criteria (the allocation of powers) or on positivistic criteria (are some countries federal just because their constitutions say so, or should federalism exist in reality and not just on paper, and how can the presence of a federal element be measured?). At the same time, it would be fallacious to believe that these different conceptual categories have no meaning simply because they cannot be easily distinguished from one another.141 To be manifestations of the same phenomenon, ie the vertical division of powers, does not mean to be the same thing. As Giandomenico Falcon once put it, nobody knows in abstract terms how much air is needed for a balloon to be recognized and used as such, but if you give an empty balloon to a child, he will definitely refuse to play with it.142 This is to say that while neither a general, all-encompassing theory of federalism nor an abstract, legally all-explaining definition of a federal state as opposed to other related categories is possible, a pragmatic observation is in the end what makes it possible to identify the phenotype and to work with it better than if these never fully definable categories did not exist. Similar to other concepts, both the quantitative and the qualitative elements are important to define the nature of the federal phenomenon. Thus, definitions are necessarily incomplete but nevertheless fundamental to approach a full understanding of a complex phenomenon. While the development of an abstract model is not possible, the long history of the vertical division of power might instead give important indications as to the relevant differences. In the first place, until modern times, such division of power mostly took the form of confederations. The classical, centripetal federal state was the form of government of the modern era, intimately linked with the formation of nation states when the circumstances so provided: for example, in the nineteenth century the processes of the national unification of Germany and Italy were similar in many respects, aimed as they were at creating a nation state by bringing together small kingdoms divided for centuries. But due to historical and political circumstances, Germany was united as a federal state and Italy was not. The contemporary era is dominated by the regional state or by federal ones created (gradually and often asymmetrically) in a devolutionary process. As the time of national unification has passed and the recognition of regional diversity is commonly accepted, decentralisation in all its forms (a regional state sometimes leading to a fully-fledged federation, devolution, various forms of autonomy, etc) better fits contemporary government. In particular, it helps address some of the key challenges

141

See ch 0.2. Falcon, ‘Federalismo–regionalismo: alla ricerca di un sistema in equilibrio’ in L Mariucci et al (eds), Il federalismo preso sul serio: una proposta di riforma per l’Italia (Bologna, Il Mulino, 1996) 115. 142 G

66 Manifestations of the present day, such as accommodation of ethno-cultural, economic and other differences and coping with unprecedentedly complex governance. Thus, differences exist and different concepts do matter. In constitutional terms it should never be forgotten that there are various historical and present manifestations of the federal idea, however achieved. All these are part of the family of federal systems and reflect, in different ways, the balance between self-rule and shared rule, or, in more modern terms, between autonomy and integration.143

143 See J Marko, Autonomie und Integration: Rechtsinstitute des Nationalitätenrechts im funktionalen Vergleich (Wien, Böhlau, 1995).

3 History 3.1 ANCIENT UNIONS

I

‘The origin of the federal idea is wreathed in mist, as indeed is the origin of life itself … there are no secure details of parentage, no reliable paternity tests.’1

NDEED, IT IS a hopeless endeavour to determine with an adequate degree of certainty the historical roots of federalism and its diverse institutional manifestations. Its basic idea was most likely first discovered and practised independently at different times and in different places. One example of an early form of a union was the Amphictyonic League of ancient Greece, through which tribes and later cities were loosely bound together for the purpose of organising the protection of sanctuaries and the practice of common religious rites. Similarly, the federal idea was the driving force behind the integration of the tribes of Israel. 3.A The 12 Tribes of Israel As mentioned above, federalism and its earliest manifestations may not be traced back to a time or place. In this sense, there is no ‘cradle of federalism’. Yet, even with similar unions existing throughout ancient times elsewhere as well, the compact that bound together the 12 tribes of Israel for several centuries is worth a closer look. There are essentially two elements that make this case particular, namely the quite detailed written constitution and the—by the standards of the time—high level of integration.2 From the thirteenth century BC onwards, the Israelites were united for nearly 700 years under the Torah of Moses as a common constitution that was binding for all tribes. The concept of a foedus was thus transferred from being applied in relations between God and his people to those among the 12 tribes.

Even though Greeks and Israelites are thus often seen as having pioneered this idea, similar forms of union existed in other parts of the world. Cases in point are such political traditions among Bedouin tribes throughout their history and among some Native American tribes. Most famous is probably the Iroquois League, which arguably inspired to a certain degree the Founding Fathers of the United States.3

1 SR Davis, The Federal Principle: A Journey through Time in Quest of a Meaning (Berkeley, University of California Press, 1978) 2f. 2 See DJ Elazar, Exploring Federalism (Tuscaloosa, University of Alabama Press, 1987) 117. 3 See DS Lutz, ‘The Iroquois Confederation Constitution: An Analysis’ (1998) 28 Publius 99.

68 History What all of these unions have in common is that their aim—apart from religious motives—was primarily the pooling of military forces for defensive and/or offensive purposes. While it remains uncertain who should be credited with the invention of some forms of federal union, it seems to be quite clear what provoked them. Regardless, the creation of a larger unit that could succeed in the event of war was probably seldom an end in itself, but was seen rather as being linked to the enhancement or maintenance of economic welfare. This ultimately military and economic rationale was particularly forceful not only regarding ancient unions but also in the case of the classical confederations that followed.4 3.2 THE AGE OF CONFEDERATIONS

3.2.1 Classical European Confederations The late Middle Ages, a period when European cities were experiencing an economic boom and, as a corollary, were also gaining political weight, marked the dawn of a new era of federalism: the age of confederations. Increasingly autonomous cities in both southern and northern Europe overcame their isolation by integrating into larger entities for the sake of military and/or economic cooperation. The Lombard League (1167–1250), for example, was established primarily in an attempt to fend off the Holy Roman Emperors on the Apennine Peninsula. The Hanseatic League (1250–1669) had a military element and also served economic expansion. Particularly in its golden age until 1400, the members of the Hanse—cities and their guilds—dominated trade in the coastal areas of the Baltic Sea and the North Sea. While these two examples of alliances among cities are mainly of historical interest, other confederations composed of larger territories occupied a far more important place in the history of federalism.5 This is because some of them, namely the United States of America under the Articles of Confederation (1781–88) and the Old Swiss Confederacy (1291–1798 and 1815–48), eventually took the decisive step from confederation to federation.6 Another confederation, however, and one that was for some time the most promising, finally vanished. The Dutch Republic of the Seven United Provinces (1581–1795) was, in economic terms particularly, a remarkable success story. 3.B The Republic of the Seven United Netherlands The foundations of the Dutch Republic’s long-term leading position in global trade were laid in 1579, when the counties, cities and bishoprics of the northern Netherlands concluded the Union of Utrecht, an agreement of mutual military support against Habsburg Spanish rule, which was followed two years later by a declaration of independence. Besides

4 See especially M Forsyth, Unions of States: The Theory and Practice of Confederation (Leicester, Leicester University Press, 1981) 160. 5 See DJ Elazar, Federalism: an Overview (Pretoria, HSRC Publishers, 1995) 21. 6 See sections 3.3.1 and 3.3.2 below.

The Age of Confederations 69

opposition to the persecution of Protestants, one of the primary causes of this resistance was discontent with Habsburg Spain’s efforts of absolutist centralisation, which were aimed at supplanting the typically medieval network of autonomous territorial, professional and religious entities. Though conceived as a confederation of seven sovereign provinces, the de facto power relations were highly asymmetrical, as the common assembly of provincial representatives, the Staten-Generaal, was clearly dominated by Holland.7 Thus, it is hardly surprising that the policies of this institution such as foreign affairs and supervising the extremely profitable Dutch East India Company came to be largely guided by Holland’s commercial interests.8 However, the most significant reason for the confederation’s eventual demise was arguably the increasing rivalry within provinces between royalists, favouring strong governors (stadhouder) from the House of Orange-Nassau, and republicans, supporting the assemblies of the provincial estates (staten) and the transformation into a true republic. Ultimately, the invasion by Napoleon’s army in 1795 sounded the death knell to the already weakened confederation.

As a product of early modern times, the Republic of the Seven United Netherlands was predated by another notable confederation, the Old Swiss Confederacy (Alte Eidgenossenschaft), which had its roots in the Middle Ages. Already in 1291, the alpine territories of Uri, Schwyz and Unterwalden settled on a letter of alliance (Bundesbrief).9 Starting from this small group of original cantons in remote mountainous areas, the confederation was gradually extended to urban areas. In the Swiss case, integration also came about for both military reasons, that is, common resistance against the Habsburg emperors, and economic motives, especially the safeguarding of important trade routes. Nevertheless, the interests of the cantons diverged in many respects, especially in economic terms between rural–agricultural and urban–commercial areas. This cleavage even deepened during the age of Reformation, as many cities welcomed the religious revolution, while rural areas strongly opposed it. Bearing these severe antagonisms in mind, it should come as no surprise that the confederation’s asymmetrical legal framework put much emphasis on diversity and granted each canton (Ort) a different status. With a standardising single act of union lacking from the beginning, the individual legal position was determined by an extremely complex conglomerate of letters of alliance. The Tagsatzung, the confederation’s only common institution, was composed of representatives bound to instructions from their governments and as generally required to reach unanimity, which often resulted in a deadlock, especially between rural and urban cantons.

7 See J Israel, The Dutch Republic: Its Rise, Greatness, and Fall 1477–1806 (Oxford, Oxford University Press, 1998) 106ff. 8 See JAF de Jongste, ‘Ein Bündnis von sieben souveränen Provinzen: Die Republik der Vereinigten Niederlande’ in T Fröschl (ed), Föderationsmodelle und Unionsstrukturen: Über Staatenverbindungen in der frühen Neuzeit vom 15. zum 18. Jahrhundert (München, R Oldenbourg Verlag, 1994) 138ff. 9 This pact makes reference to a preceding alliance, namely the antiqua confoederatio of 1273, see WE Rappard, Collective Security in Swiss Experience: 1291–1948 (London, George Allen and Unwin, 1948).

70 History Moreover, after the Reformation, the Protestant majority of the cantons in the Tagsatzung no longer coincided with the Catholic majority of the people.10 An invasion by Napoleon’s army seemed to put an end—similar to the Dutch Republic—to the experiment of confederation. In 1803, however, in the aftermath of federalist insurgencies, Napoleon set in motion the return to a confederation, which was eventually confirmed in 1815. Nevertheless, the legacy of French constitutionalism during the brief period of the unitary Helvetic Republic (1798–1803) changed its design, as the re-established cantons were granted equal status.11 Moreover, the French July Revolution of 1830 sparked constitutional change towards liberalism mainly in Protestant urban cantons. Their ideological and, in the end, military clash in 1847 with the Sonderbund, an alliance of seven Catholic–conservative cantons, and the ensuing federal transformation marked the end of the Old Swiss Confederacy.12 In retrospect, there are many similarities between these two most notable European confederations, including the military–economic rationale, the impact of the Protestant Reformation—albeit unifying in one case and divisive in the other—and the royalist–republican and liberal–conservative ideological cleavages within or between the members of the confederations. Although the danger of deadlock was politically tempered in the Dutch case by Holland’s power to impose its will on the others, both the Staten-Generaal and the Tagsatzung clearly proved unfit for efficient decisionmaking and thus exposed the need for change, which was eventually propelled by Napoleon. In the Netherlands, this change meant that the provinces would survive after the reorganisation of Europe in 1815 only in the context of a decentralised unitary monarchy, whereas the Swiss created more integrated joint institutions and eventually a federal state in 1848. 3.2.2 The United States of America under the Articles of Confederation The decisive step from confederation to federation was first taken in another part of the world. The fact that 13 British colonies would eventually found the United States of America was anything but predictable, even shortly before this historic event. Until the 1760s, the colonies had enjoyed a considerable degree of self-rule with their own written charters, with discontent growing mainly through the tightened grip of the British in the aftermath of the French and Indian Wars in 1754–63. In particular, tax policies and the imposition of legislation through a remote parliament with no colonial representation fuelled dissatisfaction and led to the American Revolutionary War (1775–83) and the Declaration of Independence (1776). Advocates of

10 See M Kopp, Die Geltung des Mehrheitsprinzips in eidgenössischen Angelegenheiten vom 13. Jahrhundert bis 1848 in seiner Bedeutung für die alte Eidgenossenschaft (Winterthur, Schellenberg, 1959) 70. 11 See U Häfelin and W Haller, Schweizerisches Bundesstaatsrecht, 5th edn (Zürich, Schulthess, 2001) 13ff. 12 See T Fleiner, ‘Swiss Confederation’ in K Le Roy and C Saunders (eds), Legislative, Executive and Judicial Governance in Federal Countries (Montreal, McGill–Queen’s University Press, 2006) 266f.

The Age of Confederations 71 independence had started to gather in the 1770s in local ‘committees of correspondence’, which increasingly acted as parallel structures to the colonial legislatures.13 Mainly in order to coordinate their revolutionary efforts, the 13 now self-styled ‘states’ brought into being first a loose confederation, which was then formalised through the Articles of Confederation and Perpetual Union.14 These were drafted in 1777 by the Continental Congress, which was composed of representatives from all states with each having one vote, and took effect after the completion of the ratification process in 1781. Nevertheless, the Articles of Confederation did not avoid the central pitfalls of the preceding informal arrangement. The weak Congress of the Confederation had only those powers expressly granted. Moreover, it had to rely on the states to implement the laws that it adopted and could merely make requests to the states to raise an army and to grant funds without the power to enforce such requests. Finally, decision-making rules proved extremely cumbersome because fundamental issues still required unanimous assent.15 Whereas the legal design of the confederation still allowed for success in the Revolutionary War, it turned out to be unfit to solve the problems of the post-war period, especially those of an economic nature, with each state creating trade obstacles by passing protectionist legislation and issuing its own currency. Nonetheless, the inappropriateness of the Articles of Confederation was only in part responsible for the eventual demise of the confederation. James Madison asserted, for instance, that it was more discontent with the state constitutions that made federation inevitable and also impinged on its design.16 The political and economic elite in the states became increasingly convinced that their constitutions did not contain sufficient checks on democracy that could prevent the encroachment of unrestrained state legislatures on individual liberty. Indeed, the degree to which republicanism and sovereignty of the people as guiding principles were translated into people’s effective political participation varied considerably. Both the size of the electorates, usually more restricted in the southern states,17 and the powers of the legislatures elected by them differed from state to state. Early state constitutions tended to repudiate the idea of overly concentrated leadership and thus aimed at empowering the people and legislature at the expense of the executive and the judiciary. Most far-reaching in this regard was Pennsylvania’s Constitution of 1776. Any amendment by a constitutional convention required popular ratification, and the people elected their representatives to the General Assembly on the basis of what was, by the standards of the time, an exceptionally broad franchise (Section 6). The members of this almost omnipotent unicameral legislature (Section 9) acted on an

13 See M Tushnet, The Constitution of the United States of America: A Contextual Analysis (Oxford, Hart Publishing, 2009) 9f. 14 See box 2.A. 15 See RH Fallon, The Dynamic Constitution: An Introduction to American Constitutional Law (Cambridge, Cambridge University Press, 2004) 3. 16 See M Farrand, The Records of the Federal Constitutional Convention of 1787 (New Haven, Yale University Press, 1966) 134ff. 17 See DS Lutz, The Origins of American Constitutionalism (Baton Rouge, Louisiana State University Press, 1988) 58.

72 History imperative mandate, with the role of the executive being limited to carrying out their will (Section 20). Equally weak was the position of the Supreme Court, whose members could be removed ‘for misbehaviour at any time by the General Assembly’ (Section 23). Interestingly, this radical design soon sparked a countertrend towards an emphasis on checks and balances, which were epitomised by the institution of a popularly elected single executive possessing a legislative veto.18 This countertrend not only characterised a second wave of state constitutions from 1777 onwards,19 but it also informed, in the end, the federal Constitution. Critics of Pennsylvania’s model not only set the tone for later constitution-making in the states, but they turned, after futile amendment attempts in some cases, to the superior level and eventually came to dominate the Constitutional Convention of 1787.20 Therefore, it is hardly surprising that the Convention was very much in line with second-wave state constitutions: ‘What is the Constitution of the United States … but that of Massachusetts, New York and Maryland! There is not a feature of it which cannot be found in one or the other.’21 3.3 ARCHETYPES OF THE FEDERAL STATE

3.3.1 The United States of America (1787) From a legal point of view, the transformation of the just described confederation under the Articles of Confederation into a federation was a revolutionary act. Taking the fundamental distinction by Emmanuel-Joseph Sieyès as a starting point, the constituent power (pouvoir constituant) exercised in 1787 was original and not derivative. In other words, the new constitutional order did not evolve from the A rticles of Confederation but resulted in spite of the evident historical continuity from legal discontinuity. Although only charged with amending the Articles of C onfederation, the 55 delegates of the Constitutional Convention in Philadelphia soon decided to replace them with an entirely new constitution.22 They disregarded the amendment procedures, which required the unanimous assent of all state governments and ratification by ‘conventions of people’ in nine states (Article VII). The successful completion of this procedure was, of course, quite unlikely because the interests of large and small states, pro- and anti-slavery states, proponents of states’ rights and of extensive federal Government were diametrically opposed.

18

See GA Tarr, Understanding State Constitutions (Princeton, Princeton University Press, 2000) 87f. On the waves of state constitution-making, see Lutz, The Origins (1988) 103ff. 20 It is quite revealing, for instance, that Pennsylvania’s delegation was composed of people who opposed the state’s constitution of 1776. See RF Williams, The Law of American State Constitutions (Oxford, Oxford University Press, 2009) 66ff. 21 John Adams quoted in C Bowen, Miracle at Philadelphia: The Story of the Constitutional Convention, May–September 1787 (Boston, Little, Brown and Company, 1966) 199. 22 See Fallon, The Dynamic Constitution (2004) 3f. 19

Archetypes of the Federal State 73 Notwithstanding these cleavages, more integration was seen, like later in Switzerland, as the only option. The conundrum of abandoning confederation without establishing an equally unwanted unitary state was resolved pragmatically by identifying a third option, that is, the creation of what then came to be known as a federal state. Similar to the first integrative step towards confederation, both military and economic motives played a role, as the new constitution was to facilitate free trade and the defence of independence against still-present British and Spanish armed forces. The rationale behind the new constitutional design was to reconcile necessary augmentation of central power with the protection of individual freedoms, including property, against violations by the federal Government. Rather than choosing the straightforward solution of entrenching a bill of rights, which only occurred in 1789,23 the Constitutional Convention created a federal state as a structural restriction on the exercise of power. Since human imperfection would make governmental power necessary but, at the same time, a danger to liberty,24 James Madison argued for a federal design as one pillar of a system of checks and balances. The classical horizontal dimension of the separation of powers25 was to be complemented by a vertical one: In a single republic, all the power surrendered by the people is submitted to the administration of a single government. … In the compound republic of America, the power surrendered by the people is first divided between two distinct governments, and then the portion allotted to each subdivided among distinct and separate departments. Hence a double security arises to the rights of the people.26

Apart from preventing abuse of governmental power, but linked to this rationale, federation should protect individual freedoms more generally against a tyranny of the majority. For Madison, liberty would be permanently threatened by factions, that is, groups in society with special interests,27 which he perceived as being rooted in human nature and thus impossible to eliminate. While majority rule already kept factions made up of a minority in check, the federal design was supposed to do the same with regard to majority factions. This is because the latter factions would be 23

See ch 10.1. The Federalist no 51: ‘If men were angels, no government would be necessary. If angels were to govern men, neither external nor internal controls on government would be necessary. In framing a government which is to be administered by men over men, the great difficulty lies in this: you must first enable the government to control the governed; and in the next place oblige it to control itself.’ 25 The Federalist no 10: ‘But the great security against a gradual concentration of the several powers in the same department, consists in giving to those who administer each department the necessary constitutional means and personal motives to resist encroachments of the others. The provision for defense must in this, as in all other cases, be made commensurate to the danger of attack. Ambition must be made to counteract ambition. The interest of the man must be connected with the constitutional rights of the place.’ 26 The Federalist no 51. For a critique of the view that a federation ensures better protection of individual rights, see FL Neumann, ‘Federalism and Freedom: A Critique’ in D Karmis and W Norman (eds), Theories of Federalism: A Reader (Basingstoke, Palgrave Macmillan, 2005) 207ff. 27 The Federalist no 10: ‘By a faction, I understand a number of citizens, whether amounting to a majority or a minority of the whole, who are united and actuated by some common impulse of passion, or of interest, adversed to the rights of other citizens, or to the permanent and aggregate interests of the community.’ 24

74 History fragmented and, overall, less likely within a larger and more diverse federation than within the single states: Different interests necessarily exist in different classes of citizens. If a majority be united by a common interest, the rights of the minority will be insecure. … in the federal republic of the United States … the society itself will be broken into so many parts, interests, and classes of citizens, that the rights of individuals, or of the minority, will be in little danger from interested combinations of the majority.28

Having analysed the legal origin and rationale of the first federal constitution in history, we can now turn to the basic features that set it apart from the Articles of Confederation. First, the 1787 Constitution was recognised as supreme law, which could be changed single-handedly neither by the states nor by Congress. By adopting the principle of constitutional supremacy (Article VI, clause 2 of the Constitution),29 the Convention explicitly denied the British heritage of parliamentary sovereignty, of which the absence of a legal distinction between constitutional and other law is one of three main characteristics.30 In terms of legal evolution, it is interesting that the recognition of this principle also meant that the federal constitution ironically adhered, regarding this aspect, to the otherwise rejected Pennsylvania Constitution of 1776.31 The newly established hierarchy between constitutional law and ordinary law had two profound implications. Constitutional amendments were subjected to a special procedure. As opposed to unanimity under the Articles of Confederation, the new Constitution stipulated the consent of two-thirds of both houses of Congress and three-fourths of the states (Article V). The other implication was that the Constitution would now impose as supreme law certain limits on ordinary legislation and form the standard against which the latter was measured. 3.C United States 1803: Judicial Review Through the Supreme Court? Shortly before the end of his term, US President John Adams, from the more centralist Federalist Party, attempted in 1801 to restrain the newly elected Democratic–Republican administration of Thomas Jefferson. He did so by creating new offices through a reform of the Judiciary Act and then filling those positions with like-minded circuit judges and justices of the peace. One of these infamous ‘midnight judges’ was William Marbury, who was denied his commission documents by the incoming administration despite his appointment by Adams. Marbury therefore petitioned the Supreme Court to order James Madison, the new Secretary of State, to deliver the documents.

28

The Federalist no 51. See ch 5.1.2. 30 According to Dicey, the characteristics of parliamentary sovereignty are the following: ‘first, the power of the legislature to alter any law, fundamental or otherwise, as freely and in the same manner as other laws; secondly, the absence of any legal distinction between constitutional and other laws; thirdly, the non-existence of any judicial or other authority having the right to nullify an Act of Parliament, or to treat it as void or unconstitutional’, AV Dicey, The Law of the Constitution, 10th edn (London, Macmillan, 1965) 39. 31 See Williams, The Law (2009) 69. It was stipulated in sec 9 of the Pennsylvania Constitution that the otherwise extremely powerful state legislature ‘shall have no power to add to, alter, abolish, or infringe any part of this constitution’. 29

Archetypes of the Federal State 75

The Court found, with Chief Justice John Marshall authoring the opinion, that the nominee definitely had a vested legal right to the commission, but added that the Court was not empowered to order Madison, by a writ of mandamus, to deliver the commission.32 Even though the Supreme Court was granted original jurisdiction over such a writ by Section 13 of the 1789 Judiciary Act, this provision went beyond the jurisdiction as defined in Article III, section 2 of the US Constitution. In contrast to Marbury’s argument that this section would only set a floor, the judges held that the above-mentioned Act of Congress could not extend the Court’s jurisdiction and was therefore unconstitutional. In such cases of conflict between the Constitution and Acts of Congress, the latter were, according to Marshall’s reasoning, not law and the courts would be bound only by the Constitution, whereby he affirmed the principle of judicial review. One of his main arguments was that any written constitution would be useless if Congress could ignore it and the courts could not enforce it: ‘To what purpose are powers limited, and to what purpose is that limitation committed to writing, if these limits may, at any time, be passed by those intended to be restrained?’ (p 176). Furthermore, Marshall argued that, in order to fulfil their task of deciding cases, courts would have to be able to decide what law applies to each case: ‘It is emphatically the province and duty of the Judicial Department [the judicial branch] to say what the law is. Those who apply the rule to particular cases must, of necessity, expound and interpret that rule. If two laws conflict with each other, the Courts must decide on the operation of each. … the Court must determine which of these conflicting rules governs the case. This is of the very essence of judicial duty’ (pp 177–78). As to why the Constitution would take precedence, Marshall referred to the oath of judges requiring them to uphold it and to the Supremacy Clause, which lists the ‘Constitution’ before the ‘laws of the United States’.

A second feature of the 1787 Constitution concerns the direct effect and extensive scope of central legislation. The direct effect of laws enacted by the Congress on individuals was a main innovation, as this had still required during the time of confederation the transposition of the legislation concerned by the states. Whereas central legislative powers had been merely derived under the Articles of Confederation, needing explicit delegation by the states, they were now enshrined in a constitution that was, as emphasised above, unalterable without the consent of central institutions. The pivotal Article I, section 8 expressly granted the Congress legislative power in such eminent areas as interstate and foreign commerce, taxation and raising an army. As mentioned above, the failure of the Articles of Confederation was ascribed to a considerable extent precisely to the lack of such central powers in economic and military affairs. Furthermore, Article I, section 8, clause 18 conferred on the Congress the authority to make laws ‘necessary and proper’ to implement its explicit powers. Even though fiercely resisted by the Anti-Federalists, who feared boundless central power, this provision eventually became the foundation of the influential concept of implied powers.33 A third characteristic of the 1787 Constitution was the creation of a second chamber of federal parliament, which was allegedly intended to ensure the participation 32

Marbury v Madison 5 US (1 Cranch) 137 (1803). See ch 5.2.3. For an account of Anti-Federalists’ criticism of the Constitution, see C Kenyon, Men of Little Faith (Amherst, University of Massachusetts Press, 2003). 33

76 History of the states in central legislation.34 Though Madison claimed that the Senate ‘will derive its powers from the states as political and coequal societies’,35 it did not actually evolve as an institution representing states’ interests36 and was probably never supposed to do so.37 The depiction of the Senate as a stronghold of state interests seems inconsistent with several constitutional rules that disfavour such a role. Cases in point are the removal of the pre-existing imperative mandate and recall, which overall weakened the links between the states and their alleged representatives at the federal level. Contrary to what Madison’s ex post justification suggested to mute Anti-Federalist critique, the design of the Senate did not quite result from conceptual considerations as a symbol of the states still being equal. It was rather the result of a pragmatic bargain, the so-called Connecticut Compromise,38 to please the smaller states so that ‘it is rather muddleheaded to romanticize a necessary bargain into a grand principle of democratic politics’.39 Arguably, there was, however, an element of principle insofar as the Senate was to reflect the theory of mixed government, blending aristocratic and democratic elements, as espoused by British Whigs and Montesquieu.40 As an ‘American House of Lords’,41 it had, much like many other cornerstones of the new federation, the function of protecting against the excesses of democracy. This is best illustrated by an anecdote involving George Washington explaining to Thomas Jefferson, who had been absent from the Constitutional Convention, the function of the Senate: ‘Washington asked, “Why do you pour your coffee into your saucer?” Jefferson replied, “To cool it.” “Even so,” Washington responded, “we pour legislation into the senatorial saucer to cool it”.’42 3.3.2 Switzerland (1848) Federal Switzerland, brought into being by the Constitution of 1848, to some extent followed the US example. Nevertheless, it also added some new features that make it a second archetype of the federal state. Similar to the United States, the Swiss Constitution was based on the exercise of original constituent power, as it violated 34

On federal bicameralism, see ch 6.2.1. The Federalist no 10. 36 See S Patterson and A Mughan, ‘Senates and the Theory of Bicameralism’ in S Patterson and A Mughan (eds), Senates: Bicameralism in the Contemporary World (Cleveland, Ohio State University Press, 1999) 11. 37 See G Doria, ‘The Paradox of Federal Bicameralism’ (2006) 5 European Diversity and Autonomy Papers 1, 19ff. 38 Also known as the Sherman’s Compromise or Great Compromise. 39 RA Dahl, A Preface to Democratic Theory (Chicago, University of Chicago Press, 1956) 112. At one point, the Federalists even conceded this bargain nature: ‘But it is superfluous to try, by the standard of theory, a part of the Constitution which is allowed on all hands to be the result, not of theory, but ‘of a spirit of amity, and that mutual deference and concession which the peculiarity of our political situation rendered indispensable. … A government more consonant to the wishes of the larger States is not likely to be obtained from the smaller States’, The Federalist no 62. 40 See G Wood, The Creation of the American Republic: 1776–1787 (Charlotte, University of North Carolina Press, 1998) 553ff. 41 E Swift, The Making of an American Senate: Reconstitutive Change in Congress, 1787–1841 (Ann Arbor, University of Michigan Press, 2002) 9. 42 Patterson and Mughan, ‘Senates’ (1999) 15. 35

Archetypes of the Federal State 77 the confederal requirement of unanimous consent. Instead, the draft elaborated by the Tagsatzung in the aftermath of the 1847 Civil War43 was only approved by a majority of the cantons. Likewise, similar to the United States, the new federal Constitution, which replaced the complex network of confederal letters of alliance, was granted the status of supreme law and could be amended exclusively by a special procedure. The Constitution also modelled some features of the Council of States (Ständerat) on the US Senate, albeit enabling a stronger link between the cantons and their supposed representatives. For instance, it left the regulation of fundamental issues such as the appointment and term of office entirely to the discretion of the cantons and permitted dual membership in the subnational and national legislatures. Another difference is the departure from equal representation by granting some cantons, until 1999 officially called ‘half cantons’,44 only one vote as opposed to the two votes that the others had. However, more than merely deviating from the US archetype regarding certain elements of bicameralism, Switzerland is noteworthy for three major innovations. One such innovation is the combination of federalism with a constitution that foresees the widespread use of direct democracy. The Philadelphia Convention had failed to introduce instruments of direct democracy, already existing in municipalities and some in states, equally at the national level.45 In the Swiss case, by contrast, the Constitution of 1848, and even more so that of 1874, entrenched such instruments.46 They thereby followed the example of cantonal constitutions47 and, more generally, responded to postulates of the Democratic Movement of the 1860s. Today, the long-term effects of the changes initiated at that time are still tangible. The most notable tool of Swiss direct democracy is probably the requirement for amendments of the federal Constitution to be ratified, after simple majorities in both chambers of the federal parliament, in a mandatory referendum (Article 140 of the Swiss Constitution).48 Compared to the crucial role at the federal level, direct democracy is, as a rule, even more developed in the cantons. Many of them also foresee popular initiatives beyond constitutional amendments for ordinary legislation, as well as referendums on financial decisions.49 Two cantons, namely Appenzell Inner Rhodes and Glarus, still practise the old instrument of the Landsgemeinde, that is, an annual assembly in a public space of all people entitled to vote, which makes decisions on the basis of the majority rule and a non-secret ballot

43

See section 3.2.1 above. The new Arts 1 and 142(3) of the 1999 Constitution no longer use the term ‘half cantons’, but leave their legal status untouched. 45 Examples are the town meetings and instruments foreseen in early state constitutions like Pennsylvania’s of 1776. See L Morel, ‘Referendum’ in M Rosenfeld and A Sajó (eds), The Oxford Handbook of Comparative Constitutional Law (Oxford, Oxford University Press, 2012) 512. 46 See ZT Pállinger, ‘Direct Democracy in the Swiss Federation’ in C Fraenkel-Haeberle et al (eds), Citizen Participation in Multi-level Democracies (Leiden, Brill–Nijhoff, 2015) 253ff. 47 Regarding subnational entities as potential laboratories of democratic innovation, see ch 4.3. 48 See ch 6.3.1. 49 See Pállinger, ‘Direct Democracy’ (2015) 256. 44

78 History voting system.50 Overall, these instruments of direct democracy provide decisionmaking at both the national and subnational level with particularly strong legitimacy, which has recently raised questions as to the relationship of such formally qualified acts with Switzerland’s obligations under international law. 3.D Switzerland 2011: International Law as a Limitation on Direct Democracy? In 2010, majorities of both the Swiss electorate and the cantons approved a popular initiative launched by the Swiss People’s Party that foresaw the automatic expulsion of foreigners convicted of certain major crimes (so-called Ausschaffungsinitiative). In the aftermath, the Federal Court declared51 that the newly introduced Article 121(3-6) of the Constitution would not be directly applicable but would instead require ordinary legislation to establish more precise rules about the criminal offences concerned.52 Even more importantly, it held that the new provision would collide with Switzerland’s obligations under international law, most notably the 1999 Bilateral I Agreement with the EU on the free movement of persons and the European Convention on Human Rights (ECHR). The principle of non-refoulement would mandate the examination of every single expulsion against the ECHR, in particular, its Article 8 about the right to respect for private and family life. An automatic expulsion evidently runs contrary to such an examination on a caseby-case basis. As international law, which the Federal Court is under Article 190 of the Constitution obliged to apply, the ECHR would take precedence over conflicting domestic law even if adopted later and with the enhanced popular legitimacy of a referendum.53 The main findings of this seminal judgment may have much broader repercussions on other cases as well. The precedence of international law and a necessary case-by-case assessment could collide, especially with the absolute ban on the construction of minarets (Article 72(3) of the Constitution). This ban was introduced in 2009, likewise a result of a popular initiative by the Swiss People’s Party. According to the ruling’s reasoning, a balancing on an individual basis with the freedom of thought, conscience and religion, guaranteed by Article 9 ECHR, seems mandatory. In any case, the judgment strongly underlined the limits that international law established for direct democracy at both the national and the cantonal level.

‘Together we defend the right to remain different.’54 This statement very well captures another characteristic of Switzerland, which was the first federal system to operate in the context of a society with legally recognised linguistic, religious and cultural diversity. Whereas the federal state had been introduced in the United States primarily as part of a system of checks and balances, it was to serve in Switzerland

50

For an example of the functioning of a cantonal Landsgemeinde, see box 9.F. BGE 139 I 16. 52 This was later done through a reform of the Criminal Code on 20 March 2015. 53 See A Epiney, ‘Das Verhältnis von Völkerrecht und Landesrecht aus der Sicht des Bundesgerichts. Anmerkung zum BGE 2C_828/2011 vom 12. Oktober 2012’ Jusletter (18 March 2013) jusletter.weblaw. ch/juslissues/2017/885.html. 54 D de Rougemont, La Suisse ou l’Histoire d’un peuple heureux (Lausanne, L’Age d’Homme, 1965) 18. 51

Archetypes of the Federal State 79 in 1848, among other things, the purpose of managing diversity.55 As, for instance, Article 72(1) of the 1999 Swiss Constitution makes the regulation of the relationship between state and church a prerogative of the cantons, they may, and indeed do, follow very different models. While there is—in line with the French tradition— no officially recognised cantonal church (Landeskirche) in Geneva and Neuchâtel, other cantons grant this special status to one or more religious communities.56 Similarly, cultural affairs and the related area of education are to a considerable extent regulated at the subnational level (Article 61a–71). The same applies to the issue of official language use. In this regard, each canton may, in principle, decide autonomously. But ‘[i]n order to preserve harmony between linguistic communities, the Cantons shall respect the traditional territorial distribution of languages and take account of indigenous linguistic minorities’ (Article 70(2)).57 Today, there are 22 unilingual cantons (17 German, four French, one Italian), three bilingual ones with German and French and a trilingual one—Graubünden/Grigioni/Grischun— with German, Italian and Romansh. For Switzerland as a whole, German, French, Italian and Romansh are defined as official languages, even though the latter is not a language of federal legislation and, moreover, is restricted to communication with people who speak Romansh (Article 70(1)). A third notable innovation introduced by federal Switzerland concerns the primarily cantonal implementation of national legislation. In the United States, this task had been assigned, as a rule, to the federal administration, a model that came to be known as ‘legislative federalism’. This coincidence between legislative and executive power clearly emphasises the idea of separating, as much as possible, the national and subnational levels of government. The Swiss Constitution of 1848, by contrast, established what is today called ‘administrative federalism’,58 as it largely assigned the implementation of federal laws to the cantons and granted them broad discretion in this regard. Administrative federalism has allowed the cantons to adjust national legislation to their specific circumstances. This rationale is still reflected in Article 46(3) of the 1999 Swiss Constitution, according to which the federal Government ‘shall allow the Cantons all possible discretion to organise their own affairs and shall take account of cantonal particularities.’ 3.3.3 Germany (1867/1871) The early German federation, created by the constitutions of 1867 and 1871, again differs from the United States and Switzerland in several regards and may be regarded as a third archetype. One of its primary characteristics was that it first introduced a

55

See on this, more generally, ch 4.2. are mostly the Roman Catholic Church and the Swiss Reformed Church, as well as the Old Catholic Church and Jewish congregations in some cases. 57 See EM Belser, ‘Accommodating National Minorities in Federal Switzerland’ in A-G Gagnon and M Burgess (eds), Revisiting Unity and Diversity in Federal Countries: Changing Concepts, Reform Proposals and New Institutional Realities (Leiden, Brill–Nijhoff, forthcoming). 58 On ‘legislative federalism’ and ‘administrative federalism’, see ch 5.3. 56 These

80 History federal system in a political and legal context dominated by one hegemonic entity. Unlike in the United States and Switzerland, where some power equilibrium was in place between the constituent units, Germany was from the outset characterised by immense de facto asymmetries. Prussia covered more than half of the entire state territory and was clearly dominant in political and economic terms. Why Germany eventually became a federation, however, may only be understood properly against a specific historical background. As far back as in the Middle Ages, and particularly after the 1648 Peace of Westphalia, self-governing intermediary powers like principalities, cities, bishoprics, and professional corporations had restrained the exercise of central imperial power and created a complex network that led Samuel von Pufendorf to famously describe the Holy Roman Empire as some kind of ‘irregular body resembling a monster’.59 It seems natural that the transformation of this ‘irregular body’ into a modern state appeared only conceivable in a federal form. But when the Empire was dissolved in 1806, there was a rivalry between two leading powers, Prussia and Austria, which rendered it impossible to agree on too much integration. Therefore, rather than establishing a federation, the settlement at the Congress of Vienna in 1815 foresaw the creation merely of a confederation (Deutscher Bund). After Prussia had won the 1866 war with Austria, the crucial nineteenth-century ‘German question’ of how to unify 37 predominantly Germanspeaking states was resolved according to its preference for a state without Austria. This resulted in the short-lived North German Confederation (Norddeutscher Bund) of 1867 and the quite similarly designed German Empire (Deutsches Kaiserreich) of 1871.60 The clear predominance of Prussia characterised the federation not only in the time of the Empire but also in that of the Weimar Republic established after World War I. 3.E Germany 1932: Prussian Hegemony and the Power of the Federal Government Tensions between the national Government and Prussia culminated in 1932 when German President Paul von Hindenburg issued an emergency decree under the infamous Article 48 of the Weimar Constitution. This act foresaw the removal—also for ideological reasons— of the Social-Democratic Prussian Government and its substitution with German Chancellor Franz von Papen as imperial commissioner (so-called Preußenschlag). In response, the Prussian Government, the parliamentary group of the Catholic Centre Party, the SocialDemocrats in the Prussian Parliament and, interestingly, the Länder Bavaria and Baden challenged the constitutionality of the emergency decree. The following legal case, Preußen contra Reich, involved the most prominent legal scholars of the Weimar Republic with the German Government being supported by Hans Kelsen and even represented in court by Carl Schmitt. Hermann Heller, on the other hand, acted

59 The original Latin wording is ‘irregulare aliquod corpus et monstro simile’, S de Monzambano Veronensis (S von Pufendorf), De Statu Imperii Germanici ad Laelium fratrem, Dominum Trezolani, liber unus (Geneva, 1667) VI:9. 60 For a concise overview, see W Heun, The Constitution of Germany: A Contextual Analysis (Oxford, Hart Publishing, 2010) 13ff.

Archetypes of the Federal State 81

as counsel for the Social-Democratic Party of Prussia.61 Without an in-depth examination of the requirements for invoking Article 48, the Constitutional Court held that the emergency decree was in line with the Constitution insofar as it appointed an imperial commissioner to temporarily exercise the powers of the Prussian Government.62 It reaffirmed, on the other hand, the latter’s continued right to be represented in the Reichsrat, which was to ensure subnational participation in the drafting of certain national legislation. This bad compromise effectively did away with federalism in the Weimar period and was, in the aftermath, regarded as a milestone on the path to the Republic’s eventual demise.63 After their abolishment by the National Socialists in 1933, the Länder were to be reestablished after World War II, which again raised the question of Prussia’s hegemonial status. Interestingly, the territorial changes effected during that time only left the boundaries of Bavaria, Hamburg and Bremen unchanged and resulted in a consolidation of Länder in the south. In connection with the creation by the Allied Occupation Powers of new entities in the north by breaking up Prussia,64 this finally gave rise to a more balanced Länder structure.

Besides federalism working in a context of very asymmetrical power relations, another notable feature of the early German federation was its attenuation by being combined with certain anti-democratic and anti-republican elements. This concerned, above all, subnational representation at the federal level and the implementation of federal legislation. As monarchic Prussia had relied, in its abovementioned rivalry with Austria, on a tactical alliance with the otherwise despised liberal national movement,65 the conservative Bismarck subsequently sought to mitigate the consequences of this unusual coalition through a specific federal design that limited the influence of liberals and ran contrary to their agenda of democratisation and republicanism. In concrete terms, this was ensured by the crucial roles of the monarchic governments representing their state at the federal level and dominating the implementation of federal legislation. The latter gave rise to a form of ‘administrative federalism’ that was quite different from the one operating in Switzerland. After all, the extensive legislation of the national parliament (Reichstag), eagerly awaited by the liberals, regarding criminal, civil and social security law66 was to be not only adapted to local contexts but also tempered through a decentralised administration dominated by aristocrats. Quite atypically, the constituent monarchies also

61 See D Dyzenhaus, Legality and Legitimacy: Carl Schmitt, Hans Kelsen and Hermann Heller in Weimar (Oxford, Clarendon Press, 1997). 62 Staatsgerichtshof für das Deutsche Reich (25 October 1932) 1ff. 63 See M Stolleis, Geschichte des öffentlichen Rechts in Deutschland. Band 3: Staats- und Verwaltungsrechtswissenschaft in Republik und Diktatur 1914–1945 (München, CH Beck, 1999) 121ff. 64 Law concerning the dissolution of the State of Prussia (25 February 1947), Law No 46 in the Official Bulletin of the Control Council for Germany (31 March 1947). 65 See S Oeter, Integration und Subsidiarität im deutschen Bundesstaatsrecht (Tübingen, Mohr Siebeck, 1998) 27ff. 66 See ER Huber, Deutsche Verfassungsgeschichte seit 1789. Band 3 (Stuttgart, Kohlhammer, 1963) 861ff.

82 History had authority over external relations and defence, which have been in most federations classical prerogatives of the national level.67 The institution that was supposed to coordinate actions in these two fields and concerning the implementation of federal legislation was the Federal Council (Bundesrat). Instead of serving like the US Senate or the Swiss Council of States (Ständerat) as a real second chamber of the national parliament, this body, composed of delegates of the monarchic state governments instead of elected representatives, was conceived as a bulwark against parliamentarism.68 As its unanimous decisions were de facto put into practice by the powerful administration of Prussia, the latter’s hegemony regarding the executive branch was guaranteed. Indeed, it was even reinforced by the fact that the King and Prime Minister of Prussia were at the same time Emperor and Chancellor of Germany. Whereas both the Federal Council as an institution and administrative federalism are nowadays sources of power for the Länder, they historically had, due to Prussian predominance, strong centralising effects. As we will see throughout the following chapters, the specific traits of the German federal system have had, like those of the United States and Switzerland, enormous influence on several more recently adopted federal constitutions.

67

On this and Belgium as an exception, see chs 14.4 and 14.6. M Kotzur, ‘Federalism and Bicameralism: The German ‘Bundesrat’ (Federal Council) as an Atypical Model’ in J Luther et al (eds), A World of Second Chambers (Milan, Giuffrè Editore, 2006) 266f. 68 See

4 Debates

W

HILE THE THREE preceding chapters have familiarised readers with federalism in terms of concepts, manifestations and history, this one is aimed at deepening the understanding of three debates. All three of these appear to be of crucial importance, either because the debate has accompanied federalism throughout much of its history until today or because it has become highly topical more recently. The first holds true for the question of whether sovereignty may be divided and shared in federal systems, which has shaped controversies from the time of Johannes Althusius until the European Union today (section 4.1). This debate has always been inextricably linked with another one, that of the viability of federalism in diverse societies (section 4.2). Especially in an age of federal systems being increasingly expected to prevent and resolve conflicts in the face of ethno-cultural diversity, this second debate seems to be of vital importance. It is linked to the first one by having, as one of its problems, the issue of secession, that is, a territory’s achievement of sovereignty as an independent state rather than only sovereignty shared with the national level within a federation. Diversity, beyond its ethno-cultural dimension, however, is also central to the third debate, which revolves around pluralism in a broader factual and legal sense. More precisely, this debate focuses on the question of whether federalism may, together with participatory democracy, facilitate governance in an era of pluralism and thus abate the much lamented current crisis of democracy (section 4.3). 4.1 SOVEREIGNTY IN FEDERAL SYSTEMS: INDIVISIBLE OR SHARED?

‘Sovereignty is that absolute and perpetual power vested in a commonwealth which in Latin is termed majestas.’1 The question of whether the nature of sovereignty allows it to be divided and shared between the different levels of government has concerned generations of theorists, as well as framers and interpreters of constitutions. This section starts with an exploration of the conceptual roots of indivisible sovereignty, the impact of this notion in practice and its eventual contestation (section 4.1.1). It then goes on to shed light on the invention of divided sovereignty in the constitutional practice of the United States, as well as on its adaptation or rejection by other federal systems such as Switzerland, Germany and Spain

1 J Bodin, The Six Bookes of a Commonweale [1606] (Cambridge MA, Harvard University Press, 1962) Book 1 VIII para 1.

84 Debates (section 4.1.2). The section then concludes with a glance at the particularly intricate debate on (in)divisible sovereignty that continues to absorb the European Union, constitutional courts and legal scholars alike. At the outset, it needs to be clarified that this section focuses on the legal dimension of sovereignty. Being aware of the classical Diceyan differentiation between legal and political sovereignty,2 it is not principally about the political dimension, defined as the power to ensure people’s ultimate obedience towards one’s own will. Nor does this section delve deeply into the old theoretical debate about whether political sovereignty is a prerequisite for legal sovereignty or vice versa.3 It does, however, recognise that a purely positivist construction of the concept is too narrow and that its understanding has always been ‘an expression of contemporary juridico– political discourse’.4 In this light, sovereignty should be regarded as inherently flexible, well captured by the notion of ‘floating sovereignty’,5 which allows it to be adapted to new realities. As we will see, views on sovereignty and its (in)divisibility in the context of federal systems have, over time, undergone significant changes. 4.1.1 The Theory of Indivisible Sovereignty and Its Contestation The quasi-natural point of departure in debates about sovereignty and federalism are, as surprising as that may seem, the works of two champions of concentrated power, namely Jean Bodin and Thomas Hobbes. Both formulated a conception of sovereignty as something indivisible, which then became the object of contestation by certain theorists of federalism claiming precisely the opposite.6 The seminal works of Bodin and Hobbes must be seen as embedded in, and influenced by, the specific historical situation of sixteenth- and seventeenth-century Europe with its omnipresent dynastic and religious wars. It was not by coincidence that this situation of political turmoil witnessed the contemporaneous birth and lasting linkage of sovereignty, on the one hand, and the modern state, on the other. This era’s understandably pessimistic zeitgeist is vividly illustrated by a phrase from Hobbes’ autobiography: ‘[i]t was my Mother Dear, Did bring forth Twins at once, both me and Fear’.7 In this light, it is hardly surprising that a philosopher in constant fear of civil war and an invasion of the Spanish Armada famously pictured the state of nature without government very negatively as a ‘war of all against all’.8 To escape from this 2 See AV Dicey, Introduction to the Study of the Law of the Constitution (Indianapolis, Liberty Fund, 1982) 27. 3 On this debate, see N MacCormick, Questioning Sovereignty: Law, State and Nation in the European Commonwealth (Oxford, Oxford University Press, 1999) 127ff. 4 M Loughlin, ‘Ten Tenets of Sovereignty’ in N Walker (ed), Sovereignty in Transition (Oxford, Hart Publishing, 2003) 57. 5 D Kostakopoulou, ‘Floating Sovereignty: A Pathology or a Necessary Means of State Evolution?’ (2002) 22 Oxford Journal of Legal Studies 135, 148. 6 SR Davis, The Federal Principle: A Journey Through Time in Quest of a Meaning (Berkeley, University of California Press, 1978) 46f. 7 T Hobbes, Vita Carmine Expressa (London, 1681) 27f. 8 T Hobbes, De Cive [1642] (Oxford, Clarendon Press, 1983) Praefatio para 14.

Sovereignty in Federal Systems 85 situation, people would have to submit to Leviathan, that is, state authority with unlimited and indivisible sovereignty: ‘This is the generation of that great Leviathan, or rather, to speak more reverently, of that mortal god to which we owe, under the immortal God, our peace and defence’.9 More than seven decades earlier, Bodin had already similarly argued the case for ‘absolute and perpetual power’10 as a distinctive characteristic of the modern state. In doing so, he was inspired, like Hobbes, by his own experiences of war and chaos, among them the 1572 St Bartholomew’s Day massacre of French Huguenots by Catholics. Unlike for Hobbes, however, sovereignty was, in Bodin’s view, not the result of a social contract among people ceding their individual right to decide their fate to an omnipotent authority, but simply a divine right of the king. The impact of these two philosophers proved immense in terms of both theory and practice. More importantly for our purposes, however, they also came to be contested. As to practice, it is true that absolute monarchs all over Europe and later in other parts of the world readily used arguments made by Bodin and Hobbes to justify the concentration of power in their hands. Yet, the idea of indivisible sovereignty was in fact not implemented as categorically and comprehensively as the prevailing narrative suggests. In particular, the 1648 Peace of Westphalia is often overestimated as the revolutionary moment that brought into being, like a big bang, indivisible sovereignty all over Europe. To be sure, this event provided a first impulse for the elaboration of a sovereignty concept that features the still-relevant distinction between an external dimension, ie, the lack of a superior power from without, and an internal dimension, ie, the ultimate authority to make decisions regarding a specific territory.11 Yet, the impact of the Peace of Westphalia should be put into perspective because it was, despite its undoubted importance, not as revolutionary as often assumed. In practice, many of the Holy Roman Empire’s typical patchworks of superimposed and entangled entities, enjoying a certain degree of territorial autonomy or, in the case of professional corporations and religious authorities, non-territorial autonomy within a multilevel system, actually continued to be in place well beyond 1648. After all, it was precisely the persisting reality of such nested self-governing entities that led Samuel von Pufendorf to characterise the Holy Roman Empire as ‘some kind of irregular body resembling a monster’.12 Uniform and comprehensive sovereignty was equally an illusion, around the same time, in other parts of the world like the Muscovite state and Ottoman Empire.13 In summary,

9

T Hobbes, Leviathan [1651] (Cambridge, Cambridge University Press, 1992) XVII para 12. The Six Bookes [1606] (1962) Book 1 VIII para 1. 11 On external and internal sovereignty, see especially MacCormick, Questioning Sovereignty (1999) 129f and M Troper, ‘Sovereignty’ in M Rosenfeld and A Sajó (eds), The Oxford Handbook of Comparative Constitutional Law (Oxford, Oxford University Press, 2012) 354ff. 12 S de Monzambano Veronensis (S von Pufendorf), De Statu Imperii Germanici ad Laelium fratrem, Dominum Trezolani, liber unus (Geneva, 1667) VI:9. See also ch 3.3.3. 13 See K Kössler, ‘Conclusions: Beyond the Illusion of Ethno-culturally Homogenous Territory’ in T Malloy and F Palermo (eds), Minority Accommodation through Territorial and Non-Territorial Autonomy (Oxford, Oxford University Press, 2015) 254. 10 Bodin,

86 Debates the Westphalian notion of sovereignty ‘was always an exercise in hyperbolic selfaggrandisement by states’.14 As to theory, the impact of Bodin and Hobbes is epitomised not least by the aforementioned quote of Pufendorf. Only on the basis of a notion of indivisible sovereignty and its position as ‘the regular form’ of legal–political organisation could he perceive the Holy Roman Empire as an ‘irregular body’. The concept of federalism was in this light often relegated, by Immanuel Kant15 and many others, to the realm of international relations, more exactly, to alliances between states, which were themselves imagined as enjoying uniform sovereignty. The contrary idea of sovereignty as something divisible has its roots, to a large extent, in the early federal concept of Johannes Althusius.16 According to his model, associations of smaller autonomous entities should constitute larger entities and thus produce a four-level system ranging from families, cities and provinces to the overarching commonwealth. In stark contrast to Bodin before and Hobbes after him, Althusius regarded the individual, explicitly following Aristotle, in a positive light, as a ‘social, political creature’ (zoon politikon) by nature.17 By applying the biblical idea of a compact with god to relations among people,18 he went on to define as the central process of politics that of ‘association [consociatio], in which the symbiotes [symbiotici] pledge themselves each to the other’.19 Through its bottom-up logic, the formation of such associations differed, of course, fundamentally from the rationale of absolute monarchies and their notion of uniform sovereignty. Explicitly challenging Bodin, Althusius argued the case for sovereignty as something divisible. With reference to the fact that families existed before cities and cities before provinces and the commonwealth, he claimed that, in his model, each constituting body is not only prior but also superior to the body constituted by it: ‘For however great is the power that is conceded to another, it is always less than the power of the one who makes the concession.’20 In this light, sovereignty was for Althusius something that is shared by the great number of associations constituting, at various levels, the commonwealth as an associated body: ‘From this body, after God, every legitimate power flows to those we call kings or optimates. Therefore, the king, prince, and optimates recognize this associated body as their superior, by which they are constituted, removed, exiled, and deprived of authority.’21 Whereas Althusius’s reasoning concerned the relationship between sovereignty and his peculiar commonwealth model inspired by the concept of federalism, later

14 See S Tierney, Constitutional Law and National Pluralism (Oxford, Oxford University Press, 2004) 86. 15 See box 2.H. 16 On the link in this concept between federalism and subsidiarity, see ch 1.1.2. For a comprehensive treatment of Althusius’s legacy, see TO Hueglin, ‘Johannes Althusius: Medieval Constitutionalist or Modern Federalist?’ (1979) 9 Publius 9. 17 ‘Clearly man by nature is a gregarious animal born for cultivating society with other men, not by nature living alone as wild beasts do, nor wandering about as birds’, J Althusius, Politica [1603] (Indianapolis, Liberty Fund, 1995) I, paras 24f. 18 On the influence of the biblical compact idea on the 12 tribes of Israel, see ch 3.1. 19 Althusius, Politica [1603] (1995) I, para 1. 20 Althusius (n 17) I, para 22. 21 Althusius (n 17) I, para 22.

Sovereignty in Federal Systems 87 theoretical debates focused from 1787 onwards on its relationship with federal states as constitutional realities. In this regard, it is important to distinguish, just as concerning scholarship on federalism as concept,22 between the Anglo-American common law and the Continental European civil law traditions. These two traditions differ fundamentally with regard to the weight given to sovereignty and the approach to it as a subject of research. Whereas the term sovereignty is, for example, notably absent in several law dictionaries following the Anglo-American tradition, it is usually an extensively covered centrepiece of similar dictionaries of French, Italian, Spanish or German origin.23 This simple fact reflects divergent historical patterns in terms of a more or less pragmatic approach. For most European theorists, sovereignty has remained a concept of central importance and something indivisible. This is because their imagination has traditionally revolved, also in light of nineteenth-century nationalism, around the ideal of the (indivisibly) sovereign state.24 The idea of indivisibility remained prevalent in spite of Althusius, whose works had sunk into oblivion and were only rediscovered in the late nineteenth century.25 But if sovereignty was indeed seen as characterised by the ‘principle of the state’s omnipotence’,26 this naturally led to the question of how to make this perception compatible with the reality of the federal state.27 Some scholars argued that sovereignty would reside in such cases exclusively with the national level of government. The latter would merely transfer powers by means of its KompetenzKompetenz, ie, the power to define its own powers (and those of others) to the subnational entities and would thus retain undivided sovereignty.28 For rather few European theorists, by contrast, the locus of sovereignty, again interpreted as undivided, was exclusively at the subnational level. One of them was the Bavarian scholar Max von Seydel, who claimed that the German Länder had merely delegated certain functions to the federation but, in doing so, had not lost their sovereign status. It is true that von Seydel’s works built heavily and explicitly on the idea of John C Calhoun that the federal state would be a compact between sovereign states and thus, through the exercise of a right to secession, dissoluble.29 However, the central importance of sovereignty (of the US states) in Calhoun’s constitutional theory is anything but representative for the Anglo-American tradition.

22

See ch 1.1. See O Beaud, ‘Conceptions of the State’ in Rosenfeld and Sajó, The Oxford Handbook (2012) 274. 24 For two among numerous examples, see H Kelsen, Das Problem der Souveränität und die Theorie des Völkerrechts (Tübingen, Mohr Siebeck, 1920) 64; R Carré de Malberg, Contribution à la théorie générale de l’État (Paris, CNRS, 1985) 139. 25 See O von Gierke, Johannes Althusius und die Entwicklung der naturrechtlichen Staatstheorie (Breslau, Marcus, 1880). 26 O Beaud, La Puissance de l’État (Paris, Presses Universitaires de France, 1994) 144. 27 For an overview of the positions of European theorists on sovereignty and the federal state, see A Gamper, ‘A “Global Theory of Federalism”: The Nature and Challenges of a Federal State’ (2005) 6 German Law Journal 1297, 1302ff. 28 See, eg, G Jellinek, Die Lehre von den Staatenverbindungen [1882] (Berlin, Keip, 1996) 291. 29 M von Seydel, ‘Der Bundesstaatsbegriff’ (1872) 28 Zeitschrift für die Gesamte Staatswissenschaft 185, 208ff; JC Calhoun, ‘A Discourse on the Constitution and Government of the United States’ [1854] in CN Wilson and SB Cook (eds), The Papers of John C Calhoun, vol XXVIII (Columbia, University of South Carolina Press, 2003) 82. On federalism, sovereignty and secession, see ch 4.2.3. 23

88 Debates Instead, this tradition typically started, quite pragmatically, from the constitutional reality of the federal state to reflect on sovereignty, whereas most European theorists took, as mentioned above, indivisible sovereignty as a given and attempted to make it somehow compatible with federations as evident phenomena. The AngloAmerican approach is vividly expressed by the following view on (mostly European) debates about the categorisation of federal systems and the locus of sovereignty: All these definitional arguments, however, say more about the conceptual imagination or intellectual agenda of those arguing for one view over another than they do about the entity being discussed. At times, such inquiries focus on the rather elusive (and frequently unhelpful) idea of sovereignty.30

From such a pragmatic perspective rooted in constitutional reality, some scholars from the Anglo-American tradition have simply concluded that Bodinian absolute sovereignty is logically impossible: ‘Sovereignty, in the classic sense, has no meaning: divided as power is, the element of absoluteness which is essential to the concept of sovereignty is not present.’31 Others have not dismissed the notion of sovereignty in a federal state but have opened the door towards its reconceptualisation as something divisible. Unlike Bodin and his followers, sovereignty was not regarded as the Kompetenz-Kompetenz for an entity to itself determine its own powers and to extend them, if it wishes, to every area, but as a divisible bundle of powers. Consequently, it is theoretically conceivable, at least for dual federal systems,32 that each level of government has ultimate authority within distinct areas and is thus sovereign, in a more limited and relative sense, regarding these jurisdictions. This notion of a bundle of powers, which allows one government level to partly relinquish and partly retain sovereignty, is reflected in the compromise reached at the Philadelphia Convention of 1787. This was the site of the first practical debate about sovereignty and the federal state among the framers of a constitution. 4.1.2 Divided Sovereignty in the United States and Beyond Defending and interpreting this compromise of 1787, the Federalists claimed that the Articles of Confederation had ‘inconsiderately endeavored to accomplish impossibilities; to reconcile a partial sovereignty in the Union, with complete sovereignty in the States; to subvert a mathematical axiom, by taking away a part, and letting the whole remain’.33 Moreover, they asserted that the opposition to the settlement of 1787 would still cling to indivisible sovereignty of the states and thus ‘cherish with blind devotion the political monster of an imperium in imperio’.34

30 D Halberstam, ‘Federalism: A Critical Guide’ (2011) 251 University of Michigan Public Law Working Paper 1, 8. 31 RH Leach, American Federalism (New York, WW Norton, 1970) 1 [emphasis added]. 32 See chs 2.2.2 and 5.2.2. 33 The Federalist no 42. 34 The Federalist no 15.

Sovereignty in Federal Systems 89 However, the majority of the constitutional architects present in Philadelphia firmly supported divided sovereignty, even if the text of the Constitution did not explicitly establish where sovereignty resides. In their view, the states had merely relinquished parts of their sovereignty, not all of it, and thereby created an ‘incomplete’ national government. Thus, no level of government would enjoy sovereignty absolutely, but only in relative terms. In the words of de Tocqueville, even more remarkable for a compatriot of Bodin, ‘two sovereignties are necessarily in presence of each other’35 so that the ‘sovereignty of the United States is shared between the Union and the States, while in France it is undivided and compact’.36 Terminologically, the perception of sovereignty as divided and shared is also reflected in the wording used by the framers of the Constitution, who early on deliberately dismissed the notions ‘federal’ and ‘national’: ‘Since the word “federal” was thought to tend towards state sovereignty and the word “national” toward national sovereignty, James Madison contended that what was adopted was “neither a national nor a federal constitution, but a composition of both”.’37 Remarkably, the idea of divided sovereignty was not contested through claims of absolute national sovereignty, not even by the most centralist delegates to the Convention.38 It met, however, with criticism on the part of some of the AntiFederalists, which was then echoed during much of the nineteenth century by politicians and theorists such as John Taylor of Caroline and the above-mentioned John C Calhoun. According to their main argument, closely resembling Althusian thought, only the states possess ‘innate sovereignty’ because only they were ‘selfconstituted’ and afterwards constituted, in a reversible act of sovereignty, the federal government as a derived and therefore subordinated entity.39 As a corollary of the purported indivisible sovereignty of the states, Calhoun claimed a right for them not only to secede, but also to nullify national laws incompatible with vital states’ interests. Both claims rested upon his interpretation of the 1787 Constitution as a ‘compact between the states’ rather than a ‘constitution over them’.40 Eventually, the outcome of the US Civil War dealt a death blow to any claims of indivisible sovereignty of the states. It also paved the way, more generally, for a lasting reinforcement of federal power through the Fourteenth Amendment of 1868 and thus made divided sovereignty lean increasingly towards the national government.41 Still one year later, the Supreme Court also ultimately rejected, with its ruling in the

35

A de Tocqueville, Democracy in America, vol 1 [1835] (New York, Vintage, 1954) 172. ibid 128. 37 CG Haines, The Role of the Supreme Court in American Government and Politics—1789–1835 (Berkeley, University of California Press, 1944) 105. 38 ibid 106. 39 See J Taylor, New Views of the Constitution of the United States (Washington DC, Way and Gideon, 1823) 43. 40 JC Calhoun, ‘A Discourse’ (2003) 82. 41 For some of the many Supreme Court cases involving the 14th Amendment’s Equal Protection and Due Process Clauses, see boxes 9.C and 10.B. For the Due Process Clause as a basis for the Supreme Court to make fundamental rights of the federal Constitution applicable to the states, ie, so-called incorporation, see ch 10.1. 36

90 Debates landmark case Texas v White,42 Calhoun’s theory of a compact between sovereign states. Afterwards, this theory still had, however, some repercussions beyond the United States, most notably in Canada and Bavaria.43 Apart from the framers of the Constitution and their opponents, the question of sovereignty soon also emerged as a subject for the Constitution’s ultimate interpreters. In Chisholm v Georgia, the Supreme Court had to rule on a state’s claim of sovereign immunity and on sovereignty of the states, more generally. 4.A United States 1793: Sovereignty and Sovereign Immunity of the US States During the American Revolutionary War (1775–83), a merchant from South Carolina sold supplies on credit to the state of Georgia. As the merchant was a British loyalist, G eorgia later refused to pay for the goods. In response, the executor of the merchant’s estate, Alexander Chisholm, brought a lawsuit against Georgia. State officials refused to appear before the Supreme Court because, as a sovereign state, Georgia would have immunity. This claim was made, even though the US Constitution expressly affirmed the extension of federal judicial power to cases ‘between a State and citizens of another State’ (Article 3, section 2). The Supreme Court refused to accept the argument of sovereign immunity as a necessary corollary of sovereignty and obliged Georgia to pay for the goods.44 As the above-mentioned provision clearly allows a state to sue another state, the role as such of being a defendant before the Supreme Court would not be a violation of sovereignty. But the ruling went beyond just that. The judges also gave an explanation of the reasons behind the establishment of a federal judiciary. Prior to the 1787 Constitution, each state had to rely, regarding legal disputes involving parties beyond its borders, due to the strictly separate state judiciaries, on justice being ensured by another state. The possible partiality of the latter became a source of animosity so that ‘a common tribunal for the termination of controversies became desirable’ (p 474). Moreover, the Court observed that one of the objectives explicitly emphasised in the Constitution’s Preamble is precisely to ‘establish justice’, which would have to be ensured for all. In this light of equal justice, the judges found it implausible that the above-mentioned Article 3, section 2 of the US Constitution was supposed to cover only lawsuits on the part of a state, more exactly, the collective citizens of the state, against an individual citizen of another state but not vice versa. If the states, which they explicitly called ‘the joint and equal sovereigns of this country’ (p 477), had intended a narrow interpretation of such state–citizen cases to refer only to those with the state as plaintiff, they would have had to foresee this exception with explicit wording.

This judgment, which confirmed the states’ (co-)sovereignty with the national government but negated their sovereign immunity, proved highly controversial. In response, Congress initiated the Eleventh Amendment to overturn the Court’s doctrine regarding sovereign immunity. Even though this constitutional amendment

42

See box 4.D. Canada, see FR Scott, ‘French-Canada and Canadian Federalism’ in ARM Lower et al (eds), Evolving Canadian Federalism (Durham, Duke University Press, 1958); and for Bavaria, von Seydel, ‘Der Bundesstaatsbegriff’ (1872). 44 Chisholm v Georgia 2 US (2 Dall) 419 (1793). 43 For

Sovereignty in Federal Systems 91 dealt with a rather specific aspect, it illustrates very well, more generally, the very practical consequences of divided sovereignty. In short, the reintroduction of states’ immunity against lawsuits filed by citizens of another state (following from state sovereignty) posed questions as to its relationship with the justiciability of the Supremacy Clause (following from national sovereignty).45 The Supreme Court eventually attenuated the effects of state immunity with a legal fiction. In its view, the Eleventh Amendment would only disallow lawsuits against a state itself but not against state officials acting on its behalf.46 Thus, it opened up an avenue for the judicial enforcement of states’ compliance with national law as ‘the supreme Law of the Land’ (Article VI, clause 2).47 While the perception of divided sovereignty was first advanced by the constitutional practice of the United States, it is likewise characteristic of several other countries. In Switzerland, for example, there is a long tradition of such an understanding, which was already reflected in the Constitution of 1874 and recently reaffirmed by its total revision in 1999. The new Article 3 almost literally repeats the previous sovereignty provision. According to Article 3, ‘[t]he Cantons are sovereign except to the extent that their sovereignty is limited by the Federal Constitution’. Even if the cantons are thus explicitly labelled right at the beginning of the Constitution as (co-)sovereign, this declaration is mostly of historical–symbolic importance.48 In Germany, sovereignty has clearly had a more practical legal impact, paradoxically, even without the term being explicitly mentioned in the Basic Law. Instead, it was the Federal Constitutional Court that not only recognised the notion of divided sovereignty,49 but also derived from it concrete prescriptions for the relations between the Länder and the other levels of government. First, their sovereignty is the distinct difference with regard to the merely decentralised local authorities that form part of the respective Land.50 Second, they must be guaranteed, by virtue of their sovereignty, constitutional autonomy, as well as a constitutionally protected core of responsibilities and financial autonomy.51 However, the co-sovereign status of the Länder does not imply a right to secede. In 2016, the Federal Constitutional Court refused to accept a constitutional complaint for adjudication that had been directed against the refusal to hold a referendum on the secession of Bavaria from Germany. The Court merely stated that the federation is based on the constituent power of the German people as a whole so that separatism of single Länder violates the constitutional order.52

45

On the Supremacy Clause, see especially ch 5.1.2. Ex Parte Young 209 US 123 (1908). 47 While several Supreme Court rulings of the 1990s weakened the Supremacy Clause vis-à-vis sovereign immunity, political and normative incentives for compliance with national law are still powerful. See M Tushnet, The Constitution of the United States of America: A Contextual Analysis (Oxford, Hart Publishing, 2009) 173ff. 48 See G Biaggini, ‘Föderalismus im Wandel: Das Beispiel des schweizerischen Bundesstaates’ (2002) Zeitschrift für öffentliches Recht 359, 380. 49 36 BVerfGE 342, 360f (Remuneration Law of Lower Saxony). 50 See ch 9.2.2. 51 34 BVerfGE 9, 20 (Federal–State Salary). 52 BVerfG 16 December 2016, 2 BvR 349/16. 46

92 Debates In Spain and South Africa, it was similarly the constitutional courts that clarified, both in early decisions, the question of sovereignty and its implications for the various levels of government. But unlike in Germany, they were reluctant to endorse the idea of subnational entities as being (co-)sovereign. The South African Constitutional Court already held in 1996 that ‘[u]nlike their counterparts in the United States of America, the provinces in South Africa are not sovereign states. They were created by the Constitution and have only those powers that are specifically conferred on them under the Constitution.’53 In a similarly clear and explicit manner, the Constitutional Court of Spain ruled in 1981 that [t]he Constitution prefigures a vertical distribution of public authority between entities of different levels that are primarily the State, holder of sovereignty; the autonomous communities, characterised by their political autonomy; and the provinces and municipalities, granted a different sphere of administrative autonomy.54

Of course, the fact that both Spain’s Constitution and its Constitutional Court refrain from dividing sovereignty is, in view of the country’s devolutionary pattern, quite understandable. This only demonstrates, once more, the crucial impact of the different paths to becoming a federal system.55 4.1.3 The Sovereignty Question in the European Union A particularly contentious and ongoing debate on sovereignty and federalism has involved the EU and its Member States. In fact, this debate has, like the related but distinct question about the EU’s ‘federal’ nature,56 accompanied European integration from its earliest days. It is worth remembering that, according to Jean Monnet, the ‘indispensable first principle’ of the 1951 Schuman Plan for the establishment of the European Coal and Steel Community (ECSC) ‘is the abnegation of sovereignty in a limited but decisive field’ because, in his view, mere ‘cooperation between nations, while essential, cannot alone meet our problem.’57 In practice, this abnegation was effected by subjecting these two industries, certainly a limited but decisive field, to decisions of a supranational body called the High Authority, that is, the predecessor of today’s European Commission. According to Article 9(5) of the 1951 ECSC Treaty, [t]he members of the High Authority shall exercise their functions in complete independence, in the general interest of the Community. In the fulfilment of their duties, they shall neither solicit nor accept instructions from any government or from any organization. They will abstain from all conduct incompatible with the supranational character of their functions.

53

The National Education Policy Bill 1996 (3) SA 289. 32/1981. On this judgment’s interpretation of political and administrative autonomy, see box 9.B. 55 See ch 2.2.2. 56 See ch 2.4. 57 Quoted and discussed in N Nugent, The Government and Politics of the European Community (London, Macmillan, 1991) 35. 54 STC

Sovereignty in Federal Systems 93 This provision demonstrated the technocratic detachment of the supranational High Authority, even though appointed by the Member States, from national interests and its function as a vehicle to abnegate, within its field of action, the sovereignty of ECSC members. Little more than a decade after the ECSC Treaty, two seminal decisions of the European Court of Justice (ECJ) in the cases Van Gend en Loos and Costa58 further elaborated on the practical legal consequences of this limited self-restriction of sovereignty by the states involved in the, meanwhile intensified, process of European integration.59 First, it famously held that ‘[T]he Community constitutes a new legal order of international law for the benefit of which the states have limited their sovereign rights, albeit within limited fields, and the subjects of which comprise not only Member States but also their nationals.’60 Importantly, this passage from the ruling in Van Gend en Loos not only marked a milestone in the ECJ’s characteristic judicial activism, but also reinforced the Court’s idea of Europe as a community composed not only of Member States, but also of peoples and individuals.61 Two years later, the judges went on to hold explicitly that this ‘new legal order’ takes precedence over the domestic law of the Member States. The Court ruled that it would be impossible for the Member States to give preference to a unilateral and subsequent measure against a legal order accepted by them on a basis of reciprocity. … [L]aw 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.62

Notwithstanding this seminal judgment, the concrete extent of this precedence has remained contested and limited by the constitutional courts of various Member States.63 This became the legal framework for European integration, which was supposed to be, according to Jean Monnet’s functionalist approach, an incremental process of generating, through tangible benefits for citizens, increasingly broad support for a ‘federal’ Europe.64 Altiero Spinelli, by contrast, one of this approach’s primary critics, warned that national interests would in this manner always prevail over the common European interest. Put briefly, Spinelli ‘acknowledged that Monnet had made the first steps easier to achieve, but he had done so by making the later steps more difficult’.65 In fact, while Monnet’s approach was successful in rapidly intensifying

58 See B de Witte, ‘Direct Effect, Supremacy, and the Nature of the Legal Order’ in P Craig and G de Búrca (eds), The Evolution of EU Law (Oxford, Oxford University Press, 1999). 59 The year 1958 witnessed the conclusion of the Treaties of Rome establishing the European Economic Community (EEC) and the European Atomic Energy Community (EAEC). 60 Case 26/62 Van Gend en Loos v Administratie der Belastingen [1963] ECR 3. 61 See P Pescatore, ‘The Doctrine of “Direct Effect”: An Infant Disease of Community Law’ (1983) 8 European Law Review 155, 157f. 62 Case 6/64 Costa v ENEL [1964] ECR 1194. 63 See ch 2.4. 64 See box 2.H. 65 M Burgess, Comparative Federalism: Theory and Practice (Abingdon, Routledge, 2006) 231.

94 Debates material integration through the adoption of substantive law regarding an increasing number of subject matters, Europe’s degree of institutional integration has continued to clearly lag behind. At the centre, the EU still today has, despite the gradual empowerment of other institutions, the ‘Medusa-like Council’,66 which safeguards the predominance of the Member State governments. This combination of strong material and weak institutional integration has been famously termed ‘Europe’s Sonderweg’.67 In spite of the Member States’ continued powerful role in EU decision-making, the incremental functionalist process of extending European legislation to ever more areas has raised among them concerns that Monnet’s proclaimed ‘abnegation of sovereignty’ is no longer confined only to ‘a limited but decisive field’. Indeed, as early as in 1990, when material integration was still weaker than today, a prominent observer stated that there ‘simply is no nucleus of sovereignty that the Member States can invoke, as such, against the Community’.68 Most notably, this backlash against an ostensibly too far-reaching encroachment on sovereignty is epitomised by two landmark judgments of Germany’s Federal Constitutional Court. 4.B Germany 1993–2009: Is Germany Still Sovereign after the Treaties of Maastricht and Lisbon? In its 1993 landmark decision regarding the Maastricht Treaty,69 the Federal Constitutional Court emphasised that Germany had maintained its status as a sovereign state. The reasoning of the judges centred around the principle of democracy. More concretely, the EU would lack certain pre-legal conditions for a genuine European democracy, most notably, common civic values and institutions like parties and media giving rise to a truly European public opinion (paras 87–88 and 41). Democratic legitimacy of EU legislation would still rest on the national peoples because the Maastricht Treaty had created a Staatenverbund and not of a European state based on the people of one European nation (para 89). In its judgment upholding the Lisbon Treaty,70 the Court further clarified the sovereignty question by defining the Staatenverbund explicitly as ‘an association of sovereign states’ (para 229). In its reasoning, it invoked once again the argument of the EU’s democratic legitimacy being based on the national level. Even if the Lisbon Treaty claims the functioning of the EU to be founded on representative democracy (Article 10(1) TEU) and the European Parliament to be composed of representatives of the Union’s citizens (Article 14(2) TEU), the judges espoused a different view. As its members are not elected EU-wide according to the principle of electoral equality, they deemed the European Parliament ‘a representation of peoples in the respective national contingents of Members’ (para 280).

66 JHH Weiler, ‘Federalism without Constitutionalism: Europe’s Sonderweg’ in K Nicolaïdis and R Howse (eds), The Federal Vision: Legitimacy and Levels of Governance in the United States and the European Union (Oxford, Oxford University Press, 2001) 55. 67 ibid 54. 68 K Lenaerts, ‘Constitutionalism and the Many Faces of Federalism’ (1990) 38 American Journal of Comparative Law 205, 220. 69 89 BVerfGE 155 (Maastricht). 70 123 BVerfGE 267 (Lisbon).

Sovereignty in Federal Systems 95

As for Germany, the Court held with reference to the principle of democracy (Article 20 of the Basic Law), which is under the ‘eternity clause’ (Article 79(3) of the Basic Law), an unamendable element of constitutional identity, that the Basic Law ‘not only presumes sovereign statehood for Germany but guarantees it’ (para 216). Moreover, the democracy principle would entail, in conjunction with the subsidiarity principle (Article 23(1) of the Basic Law), that the transfer of sovereign powers to the EU may only be effected ‘in a predictable manner, particularly in central political areas of the space of personal development and the shaping of living conditions by social policy’ (para 251). This means that democracy and subsidiarity make the principle of conferral, in addition to its entrenchment in EU law (Articles 4(1), 5(1) and 5(2) TEU), a necessary requirement under the Basic Law (paras 233–35). Moreover, the two principles prohibit a transfer of the power to decide on its own powers. This crucial distinction between the non-transferable Kompetenz-Kompetenz and single transferable sovereign competences is actually not so specific for Germany, but resembles the differentiation of the French Constitutional Council between the essence and exercise of sovereignty. In its view, France would be prohibited by its Constitution from relinquishing the essence of its sovereignty and actually has never done so because single conferred competences may always be recovered by renegotiations, denouncement of the EU treaties or constitutional amendment.71 According to the German Court, its own assessment of whether the EU really acts only within these limited transferred powers (ultra vires review) is a constitutional obligation unless legal protection can be obtained at the EU level (para 240).

The evident fact that the question of sovereignty within the EU tends to erupt, like in the aftermath of the Maastricht and Lisbon Treaties, especially in times of political and constitutional dispute, has been blamed on the widespread perception of the EU as a legal phenomenon sui generis.72 Such a perception underlies, for example, the just-mentioned notion of the Staatenverbund. Arguably, sui generis thinking has been nurtured by the above-mentioned European tradition of state-centred sovereignty and the resulting incapacity to imagine, unlike the United States, divided sovereignty.73 This deeply entrenched paradigm has also been seen as provoking three denials: that the EU has no people, no constitution and no constitutionalism.74 As the sui generis view only provides ‘a tranquilizing non liquet in times of constitutional peace’,75 it cannot prevent Europe’s inevitable tendency to lapse back, in times

71

See Troper, ‘Sovereignty’ (2012) 358. a systematic critique of the sui generis view, see R Schütze, From Dual to Cooperative Federalism: The Changing Structure of European Law (Oxford, Oxford University Press, 2009) 59f. Scepticism regarding this view had already been expressed much earlier: ‘It is all very well to say that the EU or EC is an entity “sui generis”, “of its own special kind”. But this does not take us far in the way of constructing a reasonable account of what this thing-of-its-own-kind is’, MacCormick (n 3) 142. 73 See ch 4.1.2. 74 See Schütze, From Dual to Cooperative (2009) 63ff: ‘Because sovereignty could not be divided, it had to be in the possession of either the Union or the Member States; that is, either a European people or the national peoples. Depending on the locus of sovereignty, the European Union would be either based on a (national) constitution or an (international) treaty. And even if Europe had a constitutional treaty, the lack of a “constitutional demos” denied it a constitutionalism of its own’. 75 Schütze (n 72) 60. 72 For

96 Debates of dispute, into a situation where indivisible sovereignty is localised at just one level of government. And this level is typically, like in the above-mentioned judgment, that of the Member States. Such a lapse back to national sovereignty in moments of crisis is evidently also illustrated by the ‘take back control’ slogan of the eventually successful ‘Brexit’ campaign for the withdrawal of the United Kingdom from the European Union. One crucial question is then related to the reasons why the remaining sovereign states within the union should comply with EU law. Indeed, the acceptance of the European constitutional discipline has been regarded as having elements of voluntariness and reciprocity because it is based on a principle of ‘constitutional tolerance’.76 This means that the Member States mostly comply not because as a matter of legal doctrine, as is the case in the federal state, they are subordinate to a higher sovereignty and authority attaching to norms validated by the federal people, the constitutional demos. They accept it as an autonomous voluntary act, endlessly renewed on each occasion, of subordination, in the discrete areas governed by Europe to a norm which is the aggregate expression of other wills, other political identities, other political communities … The Quebecois are told: in the name of the people of Canada, you are obliged to obey. The French or the Italians or the Germans are told: in the name of the peoples of Europe, you are invited to obey.77

Given this lack of a clear legal hierarchy, various theories of constitutional p luralism have become an increasingly popular way of conceptualising sovereignty in the EU. Notwithstanding their differences, these theories share the avoidance of the traditional state-centred notion of indivisible sovereignty. Instead, they focus on the coexistence of constitutional orders that are overlapping in terms of people and territory and relate to each other in a heterarchical rather than hierarchical manner.78 It is not by coincidence that constitutional pluralism has been closely intertwined from its outset with the study of the European Union.79 After all, the phenomenon of the EU defies the purported legal monism and etatism of the ‘Westphalian intermezzo’,80 which in Europe actually only separated two eras of pluralism. It is indeed striking how some depictions of the Continent’s present condition of constitutional pluralism, eg, that of ‘late sovereignty’, resemble the above-mentioned constellation of medieval Europe. The state of late sovereignty has been described as being characterised by ‘functionally limited polities within a multidimensional order’,81 which do not claim, unlike states in the Westphalian era, comprehensive

76 Weiler, ‘Federalism without Constitutionalism’ (2001) 62ff. For Weiler, this principle of constitutional tolerance is embodied as a ‘meta-political objective’ in the Preamble to the EC Treaty, which refers, for the sake of preserving the continent’s diversity, to distinct peoples in the plural: ‘Determined to lay the foundations of an ever closer union among the peoples of Europe’, p 67. 77 Weiler (n 66) 68. 78 See N Walker, ‘The Idea of Constitutional Pluralism’ (2002) 65 Modern Law Review 317, 317; M Poiares Maduro, ‘Contrapunctual Law: Europe’s Constitutional Pluralism in Action’ in Walker, Sovereignty in Transition (2003) 521. 79 Constitutional pluralism was introduced, precisely in this EU context, by MacCormick (n 3) 104. 80 Kössler, ‘Conclusions’ (2015) 253. 81 N Walker, ‘Late Sovereignty in the European Union’ in Walker (n 4) 23.

Federalism in Diverse Societies 97 and mutually exclusive jurisdiction over a specific territory. Much of this holds true similarly for Europe which was likewise a patchwork of overlapping territorial and functional jurisdictions. Another variation of constitutional pluralism sees the legal orders of the EU and its Member States as ‘contrapunctual law’,82 with these orders requiring, for the sake of compatibility and coherence, their constant integration and harmonisation with each other. Other theories similarly consider the legal systems of the EU and its members as interacting and having a hierarchical structure internally, but without this ‘adding up to any sort of all-purpose superiority of one system over another’ so that ‘the interpretative power of the highest decision-making authorities of the different systems must be, as to each system, ultimate’.83 Thus, there would be today a ‘post-sovereign’ Europe, characterised by the coexistence of ‘the states of Europe, now not-fully-sovereign states, and the European Union, still a non-sovereign union’.84 While the Member States would still have ‘compendious external legal sovereignty towards the rest of the world’,85 they would have lost their internal legal sovereignty through transfers of powers to the EU, without the latter gaining this sovereignty. This does not mean that (indivisible) sovereignty has disappeared as a political buzzword that is, as indicated above, especially powerful in times of constitutional dispute. Certain observations regarding the controversies after Maastricht seem today just as appropriate as more than two decades ago: There is a widespread, but perhaps misguided, belief that there are a lot of sovereign states in the world, that this is a good thing, that the United Kingdom is one, and that it will be a bad thing if the UK ceases to be so. … Has not the Prime Minister recently assured the House of Commons that the “Sovereignty of this House” is in no way infringed by the Maastricht Treaty? But anyway, sovereignty is thought mightily important, and always somewhere to be found on the legal-political stage.86

In view of present signs of renationalisation and retreat from European integration, the following statement from 1999 deserves to be mentioned. Europe’s ‘successful transcendence of the sovereign state and state sovereignty is greatly to be welcomed. It has been and will be a condition for the security of peace and prosperity among us.’87 4.2 FEDERALISM IN DIVERSE SOCIETIES: CURE OR CURSE?

‘Macedonia’s sovereignty and territorial integrity, and the unitary character of the State are inviolable and must be preserved. There are no territorial solutions to ethnic issues.’ This statement in Article 1(2) of the 2001 Ohrid Framework Agreement, which ended armed conflict between the Macedonian Government and ethnic-Albanian

82

Poiares Maduro, ‘Contrapunctual Law’ (2003). MacCormick (n 3) 118. 84 MacCormick (n 3) 142. 85 MacCormick (n 3) 133. 86 N MacCormick, ‘Beyond the Sovereign State’ (1993) 56 Modern Law Review 1, 1. 87 MacCormick (n 3) 129. 83

98 Debates rebels, reflects the widespread belief that a unitary state structure is preferable for ethno-culturally diverse societies, because any territorial solution would inevitably endanger the respective state’s most basic interest, namely its own survival.88 On the other hand, however, a wealth of literature has argued the case for multinational federalism as a viable solution. This section first tracks and evaluates the lively debate between the proponents and opponents of this concept (section 4.2.1). It points out that multinational federalism as prescription should not be equated with the much broader field of federalism in ethno-culturally diverse societies. The section then draws attention to two particularly salient issues in such societies. The first one is the legal status of internal minorities within subnational entities (section 4.2.2). The second one concerns the highly topical issue of secession—due to a recent rise of independence movements— and its treatment by constitutional law (section 4.2.3). 4.2.1 Multinational Federalism as a Concept Scholars arguing, in line with the above-mentioned quotation, against the viability of federal design in a context of diversity, have pointed since the 1990s to the disintegration of the three communist ethno-federations, ie, the Soviet Union, Yugoslavia and Czechoslovakia.89 Since around the same time, however, others have started to praise autonomy as an instrument of internal self-determination for minorities that may prevent claims for external self-determination, ie, secession. The latter view is advocated, in particular, by academics who limit territorial autonomy more or less exclusively to an instrument of minority protection90 and by proponents of what has been variously conceptualised with some variation as multinational, plurinational, ethnic or post-conflict federalism.91 What all this scholarly work has in common is that it endorses the concept that a country’s cultural diversity should form the basis of its federal territorial structure. From a practical perspective, it is thus recommended that internal boundaries be drawn or even redrawn in such a way that nationwide minorities, at least large ones with a compact area of settlement, are transformed into regional majorities

88 Ironically, even though the Ohrid Agreement refrained from introducing a federal system, one of its core elements, namely the redrawing of municipal boundaries to the benefit of the Albanian minority, was actually a ‘territorial solution’. See J Marko, ‘The Referendum on Decentralization in Macedonia in 2004: A Litmus Test for Macedonia’s Interethnic Relations’ (2004) 4 European Yearbook of Minority Issues 695. 89 See PG Roeder, ‘Soviet Federalism and Ethnic Mobilization’ (1991) 43 World Politics 196; V Bunce, Subversive Institutions: The Design and the Destruction of Socialism and the State (Cambridge, Cambridge University Press, 1999). 90 For instance, according to Yash Ghai’s classic definition, ‘autonomy is a device to allow minorities claiming a distinct identity to exercise control over affairs of special concern to them’. See Y Ghai, ‘Autonomy as a Participatory Right in the Modern Democratic State: Public Participation, Autonomy and Minorities’ in ZA Skubarty (ed), Beyond a One-Dimensional State: An Emerging Right to Autonomy? (Leiden, Martinus Nijhoff, 2005) 38. 91 For an introduction, see J McGarry and B O’Leary, ‘Federation and Managing Nations’ in M Burgess and J Pinder (eds), Multinational Federations (London, Routledge, 2007).

Federalism in Diverse Societies 99 within ‘nationality-based units’.92 In constitutional practice, which has, since the 1990s, often followed this prescription, the perception of subnational entities as benefitting, at least predominantly, one group is reflected particularly in, first, statements of identity and, second, provisions about territorial (re)organisation. As to the first point, for instance, former Article 1 of the 1992 Constitution of the Republika Srpska explicitly defined this constituent unit of Bosnia and Herzegovina as ‘a State of the Serb people and of all its citizens’.93 More recently, this subnational entity held in 2016 a referendum about upholding a National Day of the Republika Srpska which the Constitutional Court had deemed discriminatory against non-Serbs because of the day’s historical connotations and thus unconstitutional.94 As to the second dimension, that of drawing subnational boundaries, multinational federations do not regard territory, in contrast to mononational federations like the United States, as something neutral, as ‘a blank slate to be filled in by whoever lives on the territory’.95 Instead, they perceive geographical areas, historically, as homelands of specific groups. This is epitomised by Article 47 of the 1995 Ethiopian Constitution, which first draws the boundaries of most constituent states to the benefit of a specific titular group and then also facilitates their redrawing along ethnic lines by granting ‘nations, nationalities and peoples within the states … the right to establish, at any time, a state of their own’.96 The two above-mentioned contrasting views regarding the viability of multinational federalism have dominated the academic debate since the 1990s.97 Critics of this concept have argued that it would provide nationality-based subnational units with important incentives and political resources, which they may draw upon in a possible attempt at secession. Existing subnational parliaments, government and administration, which wield power in significant policy fields, could be easily transformed into strong national institutions of an independent state. M oreover, they could use their jurisdiction over crucial issues such as culture, language use, education and media to push a project of minority nation-building beyond the

92 W Kymlicka, ‘Is Federalism a Viable Alternative to Secession?’ in PB Lehning (ed), Theories of Secession (London, Routledge, 1998) 125. Kymlicka contrasts them with ‘regional-based unity’. 93 The wording ‘a State of the Serb people and’ was later declared unconstitutional, Constitutional Court of Bosnia and Herzegovina, Partial Decision U5/98 III of 1 July 2000 (the seminal Constituent Peoples case), see box 4.C. 94 Constitutional Court of Bosnia and Herzegovina, U3/13 of 26 November 2015. Other highly controversial identity-related cases were Partial Decision U 4/04 of 31 March 2006 concerning the insignia of the two entities which had been introduced during the times of war, as well as Partial Decision U 44/01 of 27 February 2004 on the renaming of municipalities by adding the prefix ‘Serbian’. 95 J Kincaid, ‘Territorial Neutrality and Coercive Federalism in the United States’ in S Mangiameli (ed), Federalism, Regionalism and Territory (Milan, Giuffrè, 2013) 133f. 96 The Ethiopian Constitution does not differentiate between these three categories of groups, but only mentions them in a common definition in Art 39(5): ‘A “Nation, Nationality or People” for the purpose of this Constitution, is a group of people who have or share large measure of a common culture or similar customs, mutual intelligibility of language, belief in a common or related identities, a common psychological make-up, and who inhabit an identifiable, predominantly contiguous territory.’ 97 For an introduction into this debate, see S Choudhry and N Hume, ‘Federalism, Devolution and Secession: From Classical to Post-conflict Federalism’ in T Ginsburg and R Dixon (eds), Comparative Constitutional Law (Cheltenham, Edward Elgar, 2011) 363ff.

100 Debates limits of the current state.98 Another critique points at regional and minority parties, which would thrive under the conditions of a multinational federation and sooner or later demand secession, either out of conviction or out of political and economic self-interest.99 Most proponents of multinational federalism do not deny the risk of secession. Their claim is instead that sufficient autonomy for nationality-based units would dissuade them from pursuing independence in the first place. In this sense, they see the concept of multinational federalism as paradoxical: ‘while it provides national minorities with a workable alternative to secession, it also helps to make secession a more realistic alternative’.100 From today’s perspective, this debate about multinational federalism seems to have reached an impasse. The historical evidence is far from being conclusive and does not clearly verify the hypothesis of either side. Such verification even seems impossible for two reasons. First, constitutional designs are just too diverse and extralegal determinants like social and economic factors or political culture too influential to identify an unambiguous causal link between multinational federalism and secession. Second, a general assessment appears impossible because of the grave methodological problem that both groups of scholars tend to be biased and concentrate on those case studies that confirm their view.101 As mentioned above, opposition to multinational federalism largely originated from early analyses of the collapse of communist ethno-federations, and, to a much lesser extent, also of equally failed post-colonial federations propagated above all by the British, in Africa, Asia and the Caribbean.102 Advocates of multinational federalism, on the other hand, tend to ascribe the evident failure in these cases not to the concept itself but to contextual factors. Chief among them are lacking or still-fragile democracy and the character of ‘putting-together federations’,103 which were imposed by an outgoing colonial power or the Russian Soviet Federative Socialist Republic (RSFSR) as the hegemonic entity. Instead, multinational federalists point to more encouraging counterexamples like Canada, Belgium, Spain and sometimes India. The latter cases lead us to a third problem regarding a general assessment of multinational federalism, namely the fluctuation of the historical record when looked at over a longer time frame. While it is still easy, from today’s perspective, to see Canada as a remarkable success story of constitutional design, albeit not one that can simply be transferred to other social, economic and political contexts,104 the long-term

98 With reference to the Soviet Union, see R Brubaker, Nationalism Reframed (Cambridge, Cambridge University Press, 1996) 9. 99 See JL Snyder, From Voting to Violence: Democratization and Nationalist Conflict (New York, Norton, 2000) 327. 100 Kymlicka, ‘Is Federalism’ (1998) 142. 101 See Choudhry and Hume, ‘Federalism, Devolution’ (2011) 368. 102 For an early analysis of post-colonial federations, see RL Watts, New Federations: Experiments in the Commonwealth (Oxford, Clarendon Press, 1966). 103 On the classification of ‘coming-together federations’, ‘holding-together federations’ and ‘puttingtogether federations’, see ch 2.2.1. 104 See S Choudhry, ‘Does the World Need More Canada? The Politics of the Canadian Model in Constitutional Politics and Political Theory’ (2007) 5 International Journal of Constitutional Law 606.

Federalism in Diverse Societies 101 success of multinational federalism in some European cases has recently grown more and more uncertain. Little more than a decade ago, it had been claimed with reference to the forces of globalisation and European integration that in ‘the stateless nations of the United Kingdom, Spain, Belgium and Canada, nationalist leaders have increasingly eschewed separatism in favour of seeking a place within the new complex systems of multi-level government’.105 Since then, however, the secession option has gained support, albeit not necessarily majority support, in all of the above-mentioned European stateless nations: a development that has been polemically stigmatised as a ‘form of irredentist Euro-tribalism’.106 Of course, this wave of secessionist sentiment might wane, like a similar wave in Quebec in the mid-1990s, without actually leading to secession. But it might also prove to be enduring and eventually make some European countries join the ranks of failed communist and post-colonial experiments with multinational federalism. Looking beyond Europe, there are other cases that cannot be said to have eliminated the threat of secession forever. Even in the case of relatively stable Ethiopia, the fact that the much-publicised constitutional right to secede (Article 39) has never been invoked is mostly due to the political hegemony of the ruling-party coalition since 1991.107 The same provision, however, might in the future, in a different political context, easily facilitate secession. After all, the latter’s effective achievement seems to depend in most cases on a combination between independentist sentiments and incentives provided by the (geo)political and economic context. In conclusion, there is just no simple and general answer as to whether the concept of multinational federalism forestalls or facilitates secession.

4.2.2 The Legal Status of Internal Minorities As collateral damage, the sometimes oversimplified and general confrontation between believers in multinational federalism and their adversaries has entailed a tendency to obstruct a much needed, more differentiated debate. In this regard, it is important to underline that multinational federalism should not be equated with federalism in diverse societies. In reality, it is, as we have seen, a specific concept that advocates for such societies that nationwide minorities are, where possible, transformed into regional majorities and therefore only one among different options of federal design. Beyond that, more differentiation is needed even between multinational federations that do fit the above-mentioned basic definition. A closer look reveals that these vary considerably, for example, concerning the degree to which the constitution actually emphasises the link between territory and ethnicity and its

105 M Keating, ‘Plurinational Democracy in a Post-Sovereign Order’ (2002) 1 Queen’s Papers on Europeanisation 1, 1. Similarly, Nancy Bermeo claimed that, ‘In Spain and Belgium, federalization has held the growth of exclusive identities in check and stymied support for separatism’. See N Bermeo, ‘The Import of Institutions’ (2002) 13 Journal of Democracy 96, 105. 106 JHH Weiler, ‘Editorial’ (2012) 23 European Journal of International Law 911. 107 For details, see section 4.2.3 below.

102 Debates position towards the recently highly topical matter of secession, in particular, merit closer analysis. As to this ethnic–territorial link, we may identify three main abstract approaches.108 A legal system may, first, make subnational entities the exclusive framework for the self-government of groups; second, let territorial and ethnic elements coexist and interact; or, third, emphasise the territorial dimension with ethnicity being reduced to merely historical importance. It is obvious that the more law links territory to ethnicity, which usually means to the self-government of the regionally dominant group, the more it compromises adequate recognition of internal diversity within a subnational entity. The single dominant group then tends to be seen as owning the autonomous territory and, as a consequence, territorially based power instead of sharing it with other groups. Such a monistic approach is an inherent problem of multinational federalism. On the one hand, advocates of this concept may be credited for dispelling the deeply rooted belief of traditional liberalism that a homogeneous nation and its ‘own’ unitary state would be prerequisites for democracy109 because they legally recognise cultural diversity on the national scale. On the other hand, however, they typically fail to acknowledge cultural diversity on the subnational scale. After all, the idea of empowering sufficiently large groups with relatively compact areas of settlement by giving them their ‘own’ territories, does not do away with problematic majority– minority relations, but merely reverses them. In doing so, multinational federalism often reproduces the nation state model that it is directed against, albeit at the subnational level, in what may be called ‘nation regions’. This reflects, of course, a very biased approach towards diversity that sees nationwide minorities more worthy of legal protection and empowerment than minorities within subnational entities. The latter groups, which have been variously termed ‘internal minorities’110 or ‘intra-unit minorities’,111 are, in such a context, through the lack of their ‘own’ territory, at risk of being marginalised. This risk will, in most cases, be greater for ‘minorities within minorities’, which are, like indigenous people in Quebec, numerically inferior on both the national and subnational scale. By contrast, ‘majorities within minorities’, which belong, like Anglophones in Quebec, to the nationwide majority, will generally be in a better position. After all, the latter may benefit from a protective role played by the national government, and they are potentially outnumbered only regarding subject matters assigned to the subnational

108 See F Palermo, ‘Owned or Shared? Territorial Autonomy in the Minority Discourse’ in Malloy and Palermo, Minority Accommodation (2015) 14ff. 109 This has been the prevailing view of traditional liberalism ever since John Stuart Mill’s famous words that ‘free institutions are next to impossible in a country made up of different nationalities’, JS Mill, Considerations on Representative Government (London, Parker, Son and Bourn, 1861) 289. In this context, it is not surprising, of course, that many multinational federalists are at the same time leading proponents of the Liberalism II school, which strongly opposes traditional liberalism (for an excellent overview of this school, see W Norman, Negotiating Nationalism: Nation-Building, Federalism, and Secession in the Multinational State (Oxford, Oxford University Press, 2006) 1ff. 110 Choudhry, ‘Does the World’ (2007) 624. 111 RL Watts, ‘Multinational Federations in Comparative Perspective’ in Burgess and Pinder, (eds), Multinational Federations (2007) 232.

Federalism in Diverse Societies 103 government. Minority status thus has both a territorial and functional dimension inasmuch as it depends not only on territorial demarcation of subnational entities, but also on the distribution of powers. While the issue of internal minorities has not passed unnoticed, many instruments proposed to diminish their risk of being marginalised have remained within the interconnected paradigms of the ‘nation-region’ and majoritarian democracy. Thus, they merely establish some external substantive limits to the scope of majoritarian decision-making within a subnational entity. These instruments do not, however, change, for certain subject matters, this form of decision-making as such. In other words, they refrain from introducing internal procedural limits and thus fail to directly address the fundamental problem of internal minorities being outnumbered.112 External substantive limits may be set, for instance, with two instruments suggested by Ronald Watts,113 namely the power of the national government to intervene on behalf of these groups in certain sensitive policy fields (eg Articles 349–51 of the Constitution of India) and a comprehensive catalogue of fundamental rights in the national constitution (eg the 1982 Canadian Charter of Rights and Freedoms). It also applies to the idea of supplementing territorially based autonomy within a federal state regarding certain policy fields with decision-making along the lines of non-territorial autonomy.114 Another instrument, by contrast, does directly address the crucial problem of outnumbering. If elements of power-sharing are introduced at the subnational level, this establishes internal procedural limits to majoritarian decision-making. In concrete terms, instruments for sharing legislative and executive power within a subnational entity, like special parliamentary voting procedures for certain subject matters or mandatory representation in the subnational government and public administration, protect internal minorities at least partially against being outnumbered.115 The period since the 1990s has witnessed a new focus on such regional power-sharing not only in constitutional practice, above all in Asia and Europe, but also in academic work. This is epitomised by research efforts that have gone beyond, since the late-1960s, the traditionally national frame of reference of the classic literature on power-sharing.116 Some more recent studies have placed their emphasis explicitly on ‘regional consociations’117

112 On constitutionalist justifications for external substantive and internal procedural limits, see Kössler (n 13) 269ff. 113 See R Watts, Comparing Federal Systems, 3rd edn (Montreal, McGill–Queen’s University Press, 2008) 165f. 114 For Karl Renner’s comprehensive model of the Nationalitätenbundesstaat, which almost a century ago combined territorial and non-territorial elements, see K Renner, Das Selbstbestimmungsrecht der Nationen (Leipzig, Deuticke, 1918). 115 For a critical analysis of potential and risks, see K Kössler, ‘Beyond Majoritarian Autonomy? Legislative and Executive Power-Sharing in European Regions’ in M Nicolini et al (eds), Law, Territory and Conflict Resolution (Leiden, Brill Nijhoff, 2016). 116 A Lijphart, Democracy in Plural Societies (New Haven, Yale University Press, 1977). 117 See J McGarry and B O’Leary, ‘Introduction: The Macro-Political Regulation of Ethnic Conflict’ in J McGarry and B O’Leary (eds), The Politics of Ethnic Conflict Regulation: Case Studies of Protracted Ethnic Conflicts (London, Routledge, 1993) 35. Opposed to the until-then exclusive focus on classical ‘sovereign consociations’ at the national level, McGarry and O’Leary started to pay increasing attention to these ‘regional consociations’.

104 Debates and ‘complex power-sharing’,118 with the word ‘complex’ referring, among other things, to the consideration of other government levels than merely the national one. Beyond framers of constitutions and scholars, the issue of power-sharing within subnational entities also became relevant for the judges of a constitutional court, that of Bosnia and Herzegovina. 4.C Bosnia and Herzegovina 2000: Equal Status for the Constituent Peoples Also at the Subnational Level? In 1998, Alija Izetbegović, then Chair of the three-person Presidency of Bosnia and Herzegovina, requested that the Constitutional Court evaluate the conformity of the constitutions of the country’s two entities, ie, the Republika Srpska and the Federation of Bosnia and Herzegovina, with the national Constitution. In four partial decisions, the Court then declared significant parts of the two subnational constitutions unconstitutional.119 Before these rulings, both constitutions had reflected the rationale that Bosniacs and Croats would predominate in the Federation of Bosnia and Herzegovina and Serbs in the Republika Srpska. This was obvious, for instance, from constitutional statements of identity120 and, above all, from the institutional design. Power-sharing had thus been limited in the first entity to include Bosniacs and Croats and was completely absent in the second entity so that majoritarian democracy could fully benefit the Serb group there. The national Constitution, adopted in 1995 only after the two entity constitutions, left the latter largely unchanged. The main question for the Court then was whether the definition of the three constituent peoples of Bosnia and Herzegovina in the Preamble to the national Constitution would stipulate the equal status of those groups only at the national level or also at the entity level. The Court supported in a narrow decision, with the three international and two Bosniac judges outvoting the two Serb and two Croat judges, the latter view and emphasised in this regard the constitutional principle of ‘collective equality’ of the constituent peoples.121 This principle would imply that the constitutional distinction of two entities in Article 3 cannot be interpreted as ‘a constitutional legitimation for ethnic domination, national homogenization or a right to uphold the effects of ethnic cleansing’ (p 59) within these entities. In practice, this meant that Serbs had to be included in the previously only Bosniac-Croat power-sharing regime of the Federation of Bosnia and Herzegovina and that power-sharing comprising all three constituent peoples had to be newly introduced in the Republika Srpska. The impact of this ruling on the institutional design of both entities

118 See C Kettley et al, ‘Self-Determination Disputes and Complex Power Sharing Arrangements: A Background Paper for Debate’ (2001) Cambridge Centre of International Studies Working Paper, 4f. 119 For a concise overview of the decisions, see J Woelk, ‘Bosnia-Herzegovina: Trying to Build a Federal State on Paradoxes’ in M Burgess and GA Tarr (eds), Constitutional Dynamics in Federal Systems: Sub-national Perspectives (Montreal, McGill–Queen’s University Press, 2012) 120ff. 120 On the former Art 1 of the Constitution of the Republika Srpska, see section 4.2.1 above. The Constitution of the Federation of Bosnia and Herzegovina did not declare its commitment to one nation, but still defined in former Art 1 only ‘Bosniacs and Croats as constituent peoples’, while Serbs and further minorities would fall into the category of ‘Others’. 121 Constitutional Court of Bosnia and Herzegovina, Partial Decision U5/98 III of 1 July 2000.

Federalism in Diverse Societies 105

proved to be enormous, at least from a legal point of view.122 After the entities had failed for almost two years to bring their constitutions in line with the judgment, this was eventually done through intervention by the High Representative for Bosnia and Herzegovina. By virtue of the authority conferred upon him in 1997 by the Peace Implementation Council (PIC), an international forum of 55 states involved in the peace process (‘Bonn Powers’), the High Representative issued decisions that amended the constitutions of both entities and thus laid the foundation for their institutional design today.123

4.2.3 Secession under Federal Constitutions Another highly topical issue specific to federalism in diverse societies is the constitutional approach to the question of secession.124 From a comparative perspective, it is evident that a clear majority of constitutions either refrain from granting a right to secession or even explicitly prohibit secession.125 Generally, it may hardly come as a surprise that most federal states refrain, in the spirit of Madisonian constitutionalism, from adopting ‘provisions that prevent the defeat of the basic enterprise’,126 ie, constitutional self-preservation: paraphrasing a famous statement regarding international law,127 constitutions are therefore ‘no suicide pacts’. Scepticism towards secession is even less surprising in the case of post-communist and post-colonial countries, which all too often directly associate federalism with secession and are thus characterised, partly due to historical experience, by a sort of ‘autonomy-phobia’.128 Much more interesting than the mere fact that these bans on secession exist is how they are formulated in constitutional texts or jurisprudence. Very often, they use specific terms, such as ‘perpetual’, ‘indissoluble’, ‘indestructible’ union, which go back to the judgment of the US Supreme Court in Texas v White. 122 Some observers have questioned whether power-sharing is in practice really as extensive and effective as the constitutional amendments after the ruling suggest. See F Bieber, Post-War Bosnia (Basingstoke, Palgrave, 2006) 132. 123 Office of the High Representative (OHR), Decision on Constitutional Amendments in the Federation (19 April 2002); OHR, Decision on Constitutional Amendments in Republika Srpska (19 April 2002). See also Constitutional Court of Bosnia and Herzegovina, U 25/00 of 23 March 2001 regarding the authority of the High Representative to both enact laws and change them and U 37/01 of 3 November 2001 on the denial of a Constitutional Court competence to review decisions of the High Representative to remove public officials. 124 On the normative question of whether a federal constitution should entrench a right to secession, see Norman, Negotiating Nationalism (2006) 170ff. 125 See K Kössler, ‘Federal Constitutions and Secession: From “Perpetual Union” to a Right to Separate?’ in A Eppler et al (eds), Qualified Autonomy und Federalism versus Secession in EU Member States (Innsbruck, StudienVerlag, forthcoming). 126 CR Sunstein, ‘Constitutionalism and Secession’ (1991) 58 University of Chicago Law Review 633. 127 ‘International Law is not a suicide club for states’, P Thornberry, ‘The Democratic or Internal Aspect of Self-Determination with Some Remarks on Federalism’ in C Tomuschat (ed), Modern Law of Self-Determination (Dordrecht, Martinus Nijhoff, 1993) 118. 128 F Palermo, ‘Central, Eastern and South-Eastern Europe and Territorial Autonomy: Are They Really Incompatible?’ in A-G Gagnon and M Keating (eds), Political Autonomy and Divided Societies: Comparative Territorial Politics (Basingstoke, Palgrave Macmillan, 2012) 82.

106 Debates

4.D United States 1869: The United States as a Perpetual Union? The Government of Texas claimed that the Confederate state legislature had, during the American Civil War (1861–65), illegally sold United States bonds owned by Texas since 1850. The state instituted proceedings before the US Supreme Court, which, as a consequence, had to make a decision about its original jurisdiction in the case.129 By accepting jurisdiction, the Court affirmed that Texas had never ceased to be a US state, even if it had temporarily joined the Confederate states. It held that the state had merely been under military rule during this time because the US Constitution disallows unilateral secession. Four years after the outcome of the Civil War had provided a military answer to the question of secession, the position of the victorious side was thus legally confirmed by the Supreme Court. In doing so, it echoed Abraham Lincoln, who had already claimed on the eve of the War that ‘perpetuity is implied, if not expressed, in the fundamental law of all national governments’ and that ‘No government proper ever had a provision in its organic law for its own termination’.130 According to the Court’s line of reasoning, it was clear that the ‘constitution, in all its provisions, looks to an indestructible Union, composed of indestructible States’ (p 725). If the title of the 1777 ‘Articles of Confederation and Perpetual Union’ had highlighted the perpetuity even of the Confederation, this would have to apply a fortiori to the federation as well. After all, the federal Constitution of 1787 aims, according to its Preamble, ‘to form a more perfect Union’. In the Court’s words, it was ‘difficult to convey the idea of indissoluble unity more clearly than by these words. What can be indissoluble if a perpetual Union, made more perfect, is not?’ It is important to underline here, however, that the judges did not perceive the United States to be absolutely indissoluble. Even this strongly anti-secessionist judgment acknowledged that separation could be effected either through revolution or agreement: ‘There was no place for reconsideration or revocation, except through revolution or through consent of the States’ (p 700).

Similarities to this ruling regarding terminology and arguments characterise the constitutional approaches to secession in a number of federal states. This appears to suggest that they were, besides being driven by their own specific historical context of existing or alleged separatist tendencies, at least partly influenced by the US precedent established in Texas v White. This holds true for Article 1 of the Brazilian Constitution131 and the Preamble of the Australian Constitution.132 But the link to the United States is particularly obvious in BR Ambedkar’s explanation of the choice of the word ‘Union’ in the Indian Constitution. The use of this term was to underline that India actually is a federation, but not a federation that is the result of

129

Texas v White 74 US (7 Wall) 700 (1869). Inaugural Address on 4 March 1861’, reprinted in TH Williams (ed), Selected Writings and Speeches of Abraham Lincoln 117 (New York, Hendricks, 1943)—‘Lincoln Writings’. 131 Art 1 states the following: ‘The Federative Republic of Brazil, formed by the indissoluble union of the states and municipalities and of the federal district’. 132 The Preamble underscores explicitly that the people of the colonies ‘have agreed to unite in one indissoluble Federal Commonwealth’. In conjunction with the absence of a secession clause in the main text, this wording of the Preamble has been regarded as clearly excluding a right to secession, N Aroney, ‘Australia’ in L Moreno and C Colino, Diversity and Unity in Federal Countries (Montreal, McGill– Queen’s University Press, 2010) 30. 130 ‘First

Federalism in Diverse Societies 107 an agreement.133 Precisely through the lack of such an agreement, ‘no State has the right to secede from it. The federation is a Union because it is indestructible. … The Americans had to wage a civil war to establish that the States have no right of secession and that their Federation was indestructible. The Drafting Committee thought that it was better to make it clear at the outset rather than to leave it to speculation or to dispute.’134 However, not all constitutions remain silent on secession or ban it. A second group of cases even solemnly proclaimed a unilateral right to secede. Typically, however, such provisions, entrenched most notably in all Soviet constitutions from 1924 onwards and in communist China in the 1930s,135 have lacked both a legal procedure to enforce this right and the necessary political context to do so. Thus, they merely amounted to fig leaves and were instruments used to create a fiction of voluntary union and, correspondingly, free exit, which was supposed to lure minorities groups into accepting their integration into the state. A similarly small third group includes those federal constitutions that grant not only a unilateral right to secession but also a procedure to enforce it. Most notably, Ethiopia’s 1995 Constitution first vests the ‘nations, nationalities and peoples of Ethiopia’ with ‘all sovereign power’ (Article 8) and then grants them, on that basis, a right to secession after having completed a procedure described in great detail (Article 39). Remarkably, secession was constitutionalised in Ethiopia, against the conventional wisdom that this would be a recipe for disintegration, to integrate the many different ethnic groups. In particular, the purpose was to dissuade them from following the example of Eritrea, which had gained de facto independence in 1991. In the Constitutional Commission, the opinion prevailed that such a clause would be necessary ‘for the sake of peace and stability’ and as a symbol of ‘voluntary union’.136 Today, Article 39(1) grants Ethiopia’s ‘nations, nationalities and peoples’ an ‘unrestricted right to self-determination up to secession’.137 In contrast to the above-mentioned Soviet and Chinese cases, this proclamation is complemented by a clearly defined legal procedure, which has two stages. The secession clause first requires that the demand be approved ‘by a two-thirds majority of the members of the legislative council of the Nation, Nationality or People concerned’ and then supported in a referendum organised by the federal government within three years by a simple majority of voters (Articles 39(4) a–c). However, in light of the

133

See box 2.B. Parliament of India, Constituent Assembly Debates, vol VII, 4 November 1948, 43. 135 Art 72 of the 1977 Soviet Constitution proclaimed that ‘each Union Republic shall retain the right freely to secede from the USSR’. Similar provisions were Art 4 of the 1924 Constitution and Art 17 of the 1936 Constitution. Art 14 of the 1931 Constitution of the Chinese Soviet Republic (1931–37) still provided a later abolished right to secession: ‘Mongolians, Tibetans, Miao, Yao, Koreans and others … shall enjoy the full right to self-determination, ie they may either join the Union of Chinese Soviets or secede from it and form their own state as they may prefer’. 136 Quoted in A Habtu, ‘Multiethnic Federalism in Ethiopia: A Study of the Secession Clause in the Constitution’ 35 Publius 313, 324f. 137 See YT Fessha, Ethnic Diversity and Federalism: Constitution Making in South Africa and Ethiopia (Farnham, Ashgate, 2010) 207ff. 134

108 Debates political hegemony of the Ethiopian People’s Revolutionary Democratic Front (EPRDF), which has held, since 1991, a monopoly on power at all government levels, it is highly doubtful whether this procedure, if ever invoked, would in fact be implemented. Also in the lesser-known case of St Kitts and Nevis, a tiny Caribbean archipelago, a secession clause still has to stand the test of practice. According to Article 113 of the 1983 Constitution, the legislature of the island of Nevis may decide, with a two-thirds majority of its members, that it will cease to be part of the two-island federation. This decision then has to be confirmed, like in Ethiopia, in a referendum, even by two-thirds of all valid votes cast.138 In spite of the specific context of such a small bipolar federation, this clause is interesting because it even goes beyond—for instance, regarding the right to participate in the vote—the Ethiopian case in constitutionally regulating procedural details. A fourth constitutional approach towards secession is to grant a right not to secession itself but merely to negotiations towards that end. This is the singular and highly exceptional case of Canada. Unlike in most of the above-mentioned cases, this position is not explicitly enshrined in the text of the Constitution. Instead, it goes back to a seminal reference of the Supreme Court, which has been widely credited for carefully balancing different views and being highly innovative.139 In sum, the ruling outlined a middle-way solution that was welcomed by both conflicting parties, albeit later interpreted differently in national and provincial laws.140 This solution essentially rejected the extreme views at both ends, namely that a positive referendum result would directly pave the way to unilateral secession and that such an outcome would have no effect at all. 4.E Canada 1998: A Constitutional Right to Unilateral Secession or to Negotiations? When the Parti Québécois was in power in the province of Quebec, it staged two referendums, first in 1980 and then in 1995. The first referendum was not about the province’s future status itself, but merely about a mandate for the provincial government to negotiate towards such a status, which came to be known as sovereignty-association. Any possible change in political status would then have required popular approval through another referendum. This contrasts strikingly with the question asked in the referendum 15 years later, which asked not merely for a mandate to negotiate but directly whether ‘Québec should become sovereign, after having made a formal offer to Canada for a new economic and

138 On referendums in secession procedures in general, see K Kössler, ‘Territorial Referendums from a Constitutionalist Perspective: Functions, Justifications and Requirements’ in A López Basaguren (ed), Claims for Secession and Federalism: A Comparative Study with a Special Focus on Spain (Heidelberg, Springer, forthcoming). 139 See AC Cairns, ‘The Quebec Secession Reference: The Constitutional Obligation to Negotiate’ (1998) 10 Constitutional Forum 26. 140 See ‘An Act to give effect to the requirement for clarity as set out in the opinion of the Supreme Court of Canada in the Quebec Secession Reference SC 2000, c 26’—the federal Clarity Act and Quebec’s Bill 99—‘An Act respecting the exercise of the fundamental rights and prerogatives of the Québec people and the Québec State RSQ, c E-20.2’.

Federalism in Diverse Societies 109

political partnership’. Thus, sovereignty would have required, unlike in 1980, no second referendum, but only the closure of partnership negotiations, either with an agreement about secession or merely a formal offer (Section 26(2) of provincial Bill 1 ‘An Act respecting the future of Québec’). After a slight majority of 50.6 per cent had voted ‘no’ in the referendum, the Governor in Council, effectively the Canadian Government, asked the Supreme Court for an advisory opinion regarding the following questions: Are Quebec’s institutions entitled to effect unilateral secession under the Canadian Constitution or under international law and which one of these two sources of law would take precedence in the event of conflicting positions? In short, the ruling held that both sources of law would disallow unilateral secession.141 However, it explicitly approved of secession by constitutional amendment, even if the Court refused to determine which one of the five Canadian amendment procedures would be applicable (para 105).142 What is more, the judges declared that if the desire to pursue secession were expressed in a provincial referendum by a ‘clear majority’ on a ‘clear question’, this ‘would give rise to a reciprocal obligation on all parties to Confederation to negotiate constitutional changes to respond to that desire’ (para 88). Instead of sticking, as is their usual practice, to specific provisions of the constitutional text, the judges derived this obligation from certain unwritten fundamental principles that they regarded as the ‘vital assumptions upon which the text is based’ (para 49): federalism, democracy, constitutionalism and the rule of law and the protection of minorities. The Court thus differentiated between the constitutional text and the more far-reaching constitutional law and engaged from a positivist perspective in making norms rather than only applying them.143 Remarkably, the above principles would not only oblige, in case of a clear expression in favour of secession, ‘all parties to Confederation’, ie, the other provinces and the national government, to multilateral negotiations, they would also govern the conduct in such negotiations, which amounts in the view of the Court to the—rather vague—obligation of the parties to act in good faith (paras 88 and 90). Moreover, the unwritten principles would also form the standard that the substantive outcome of any negotiated agreement would have to be measured against. This left two crucial questions unresolved. Which institution, if any, is vested with the authority to judge good faith and what happens if even the good faith of all parties does not lead to an agreement? As to the first question, the Court made clear that it had ‘no supervisory role over the political aspects of constitutional negotiations’ (para 100). The question of good faith was regarded besides the question of what is a ‘clear majority on a clear question’ as such a political aspect, which would be subject only to political evaluation. Regarding possible sanctions, the Court merely noted that a failure to negotiate in good faith ‘may have important ramifications at the international level’ (para 103) because it would undermine, in the eyes of the international community, the legitimacy of the government concerned. As to the second question, a lack of agreement despite the good faith of all parties, the Court explicitly refused to speculate what would happen in such a case (para 97). It simply accepted the possibility of such an impasse, but failed to provide any rule as to how to break it.

141

Reference Re Secession of Québec [1998] 2 SCR 217. an account of the different expert views concerning the relevant amendment procedure, see PW Hogg, Constitutional Law of Canada, 5th edn (Scarborough, Carswell, 2000) 5-39. 143 See S Choudhry and RL Howse, ‘Constitutional Theory and the Quebec Secession Reference’ (2000) 13 Canadian Journal of Law and Jurisprudence 143, 155. 142 For

110 Debates Europe has also recently witnessed secessionist tendencies. In Catalonia, secessionist sentiment has grown significantly over the last few years, not least driven by the Constitutional Court’s decision to strike down parts of the 2006 Statute of Catalonia.144 The Court then attempted in a series of rulings to counter moves towards separation. The judges regarded the Declaration of Sovereignty, adopted by the Catalan Parliament in 2013, as unconstitutional.145 But in the same ruling they also held that the ‘right to decide’, in Catalan dret a decidir, of Catalan citizens was not per se illegal, if exercised in accordance with the constitutional framework. After the Spanish Parliament had refused to authorise a referendum, Catalonia called for a unilateral ‘popular consultation’ on independence, which was, contrary to an injunction of the Constitutional Court, held on 9 November 2014. After the vote, the judges ruled that both the decree and the Catalan law it was based on had violated the Constitution.146 The fact that secessionist parties then won in the regional elections of September 2015 a majority of seats (but only 47 per cent of the popular vote) was interpreted by these parties as democratic legitimation for further moves towards independence. On 9 November 2015, precisely one year after the abovementioned ‘popular consultation’, the Catalan Parliament passed the Declaration of the Initiation of the Process of Independence of Catalonia which stipulated that the government of the autonomous community should disobey national legislation and disregard rulings of the Constitutional Court on the validity of Catalan legislation. The Constitutional Court invalidated this resolution in a judgment147 that had, like the other aforementioned rulings, unanimous support, also from the two Catalan judges. Nonetheless, the Government of Catalonia has announced that it plans to hold a binding referendum on independence in September 2017. In contrast to the case of Catalonia, Scottish moves towards independence were much more an intergovernmental issue than a matter for the courts. This is, of course, due to the highly flexible form of ‘political constitutionalism’ that characterises the United Kingdom. In 2011, the Scottish National Party (SNP) won a majority of seats in the Scottish parliamentary elections with a manifesto that pledged to hold a referendum on independence. In response, the UK Government offered to provide the Scottish Parliament with the powers to hold a referendum, providing that it was ‘fair, legal and decisive’. The Edinburgh Agreement concluded by the UK and Scottish governments in October 2012 paved the way for an Order in Council under Section 30 of the 1998 Scotland Act, formally made in the name of the Queen, which temporarily devolved the power to hold the vote. The referendum eventually took place on 18 September 2014 with a majority of 55 per cent voting against independence. Following the United Kingdom’s decision to leave the European Union, in which a majority of 63 per cent of the Scottish electorate voted to remain in the EU, the Government of Scotland declared that it would aim at a second referendum to be held between Autumn 2018 and Spring 2019. On 31 March 2017, two days after

144

See box 5.F. STC 42/2014. 146 STC 31/2015 and STC 32/2015. 147 STC 259/2015. 145

Federalism and Participatory Democracy 111 the United Kingdom had officially notified the EU about its withdrawal, the Scottish First Minister sent a formal request to the UK Government for another order under the above-mentioned Section 30. 4.3 FEDERALISM AND PARTICIPATORY DEMOCRACY IN AN ERA OF PLURALISM

‘[T]he prestige of the absolute has rather crumbled in our hands.’148 More than a century after this observation by a pioneering theorist of pluralism, the latter is ubiquitous. This is true both as an undeniable empirical reality and in scholarly debates about pluralism, albeit with very different meanings. Indeed, political scientists and theorists in particular have used the term in various contexts to denote, for example, political competition among multiple interest groups149 or to refer to diversity of values or sets of values.150 In studies of law, we have witnessed, since the mid-1980s, a rise of what has come to be known as legal pluralism.151 Several scholars interpret today’s global legal configuration as postnational and pluralist.152 Constitutional pluralism has become a widespread notion to describe, as mentioned above, the European Union.153 In many respects, we therefore seem to be living in an era of pluralism, both in fact and in legal terms. In this section, we argue that pluralism is the missing link that provides for an inherent connection between federalism and participatory democracy. The latter two tools of governance are arguably two faces of pluralism, as they represent, respectively, its institutional and societal dimensions (section 4.3.1). We then explore the interactions, positive or negative, between federalism and participatory democracy in times of a much lamented crisis of democracy (section 4.3.2). More specifically, we examine, for example, the extent to which federalism provides more fertile ground for innovative participatory tools through the introduction of an additional level of relatively small entities and the potential to diffuse successful subnational practices both horizontally and vertically. 4.3.1 Federalism and Participatory Democracy: Two Faces of Pluralism Even if pluralism tends to mean, in different fields of research, quite different things, it is nonetheless possible and, for our purposes, even imperative to distinguish

148

W James, A Pluralistic Universe [1909] (Cambridge, Harvard University Press, 1977). RA Dahl, Who Governs? Democracy and Power in an American City (New Haven, Yale University Press, 1961). 150 See I Berlin, Four Essays on Liberalism (Oxford, Oxford University Press, 1969); R Bellamy, Liberalism and Pluralism: Towards a Politics of Compromise (London, Routledge, 1999). 151 For a thorough overview, see BZ Tamanaha, ‘Understanding Legal Pluralism: Past to Present, Local to Global (2008) 30 Sydney Law Review 375. 152 See, eg, N Krisch, Beyond Constitutionalism: The Pluralist Structure of Postnational Law (Oxford, Oxford University Press, 2012). 153 See section 4.1.3 above. 149 See

112 Debates two basic types of pluralism. Understanding the notion in a normative way is not a necessary corollary of pluralism in the descriptive sense of the term. The latter denotes the natural diversity of society and thus the coexistence within it of conflicting interests and conceptions of the public good. In short, it is about ‘pluralism-in-fact’.154 Of course, the mere presence of such diversity does not provide any indication about how to deal with this reality. Regarding this question then, ‘pluralism-as-a-norm’ supports the view that this factual diversity is desirable and should be maintained or even fostered by the legal–political order through its institutionalisation.155 Since pluralism essentially pits, as illustrated by the initial quotation, diversity against the absolute and uniform, it may hardly be surprising that early pluralists at the beginning of the twentieth century were unified in their opposition to unitary state structures and unlimited sovereignty.156 Even if they were less unified concerning the desirable alternative, this opposition prompted some of them to actively endorse the federal state and thus to directly link ‘pluralism as a norm’ with federalism. Most notable here is Harold Laski, who long fervently endorsed federalism157 before famously declaring, in the face of post-World War II reconstruction needs, its obsolescence.158 Today, conversely, federalism scholars sometimes refer to pluralism and see it as inextricably linked with their subject of research. In fact, pluralism may be regarded as ‘the backbone of federalism since federalism is at odds with any state organization aimed at protecting only one interest’,159 be it the interest of the dominant denomination in theocracies, of the majority group in ethnocracies, of the supreme leader in autocracies and, as pointed out by James Madison, even societal factions in majoritarian democracies.160 As pluralist constitutional democracy, more generally, is precisely predicated on the view that the interest of one person or one group should give way to a pluralism of interests and political competition between these, it applies both substantive rights and institutional arrangements to safeguard this factual pluralism. The latter arrangements include, most notably, the horizontal separation of powers and, in some cases, a federal state structure as a means to additionally diversify power in the vertical dimension. Together with substantive rights, these essential structural provisions are, as part of a precommitment strategy,161 typically protected through entrenchment, although to different degrees, against being amended too easily. Otherwise,

154 M Rosenfeld, Just Interpretations: Law between Ethics and Politics (Berkeley, University of California Press, 1998) 200f. 155 For the argument that ‘pluralism as a norm’, more precisely ‘comprehensive pluralism’, is superior to rival approaches like liberalism, communitarianism and republicanism, see ibid 199ff. 156 See PQ Hirst, ‘Introduction’ in PQ Hirst (ed), The Pluralist Theory of the State: Selected Writings of GDH Cole, JN Figgis and HJ Laski (London, Routledge 1989) 3. 157 See H Laski, ‘The Pluralistic State’ (1919) 28 The Philosophical Review 562. 158 See H Laski, ‘The Obsolescence of Federalism’ (1939) 98 The New Republic 367, 367ff. 159 F Palermo, ‘Regulating Pluralism: Federalism as Decision-Making and New Challenges for Federal Studies’ in F Palermo and E Alber (eds), Federalism as Decision-Making: Changes in Structures, Procedures and Policies (Leiden, Brill–Nijhoff, 2015) 504. 160 On Madison’s factions and federalism, see ch 3.3.1. 161 See S Holmes, ‘Precommitment, and the Paradox of Democracy’ in J Elster and R Slagstad (eds), Constitutionalism and Democracy (Cambridge, Cambridge University Press, 1988).

Federalism and Participatory Democracy 113 simple majority rule might, with little difficulty, privilege the interests of what Madison called the majority faction over those of minorities in a very broad sense of the term. In short, the federal state and other institutional arrangements of pluralist constitutional democracies seek to establish predefined rules and procedures for political competition162 instead of suppressing factual pluralism. Such suppression would be characteristic of bluntly non-democratic constitutions or what has been termed partisan constitutions. The latter result from constitution-making processes that lack consensus-oriented pluralism because of usurpation by one faction, even if represented by a parliamentary majority.163 From the perspective of legal pluralism, it seems important to place federalism in the broader context that it is embedded in today. As the Westphalian era has gradually come to an end with the post-World War II proliferation of international, supranational and transnational law, the global legal configuration has undergone dramatic changes. In the 1980s, this new reality directly inspired legal pluralism, which captures the essence of this existing configuration as a coexistence of multiple and semi-autonomous legal regimes that overlap in terms of people and territory.164 This pluralism has incrementally replaced the dualism of ‘the conventional, and conventionally separate, structures of constitutional law (considered as the law of the Keynesian–Westphalian state) and international law (considered as the law between Keynesian–Westphalian states)’.165 Looking at the whole picture of ever more dense interactions between legal orders, there is not, or not yet, a basic grid to conceive it. In contrast to the above-mentioned dualism of the Westphalian era, between the predominant domestic law and largely facilitative international law, there arguably prevails today a ‘disorder of orders’.166 How then does federalism fit into this overall picture of global legal pluralism? As a matter of fact, it seems fair to state flatly that ‘[f]or federalism, all this talk about pluralism and the unsettled nature of legal authority among different levels of governance is a coming home of sorts’.167 This does not mean, of course, that there are no differences between the logic of federalism and that of legal pluralism. One of the more important ones is certainly that in federations the interplay of diverse legal orders is not unrestrained, but it occurs within the context of a constitution

162 See M Loughlin, Sword and Scales: An Examination of the Relationship between Law and Politics (Oxford, Hart Publishing, 2000) 189ff. 163 In the wake and under the impression of World War II, the making of European constitutions, as an exercise of constituent power, typically aimed at reflecting pluralism by institutionalising social conflicts, whereas partisanship was largely left to the sphere of constituted power. Recently, this dualism has been challenged by the intrusion of partisanship into the sphere of constituent power. As social conflict is no longer institutionalised and thus contained, it is becoming potentially more disruptive for legal and political stability. See M Dani, ‘The “Partisan Constitution” and the Corrosion of European Constitutional Culture’ (2013) LSE LEQS paper 68/2013, www.lse.ac.uk/europeaninstitute/leqs/leqspaper68.pdf. 164 See J Griffiths, ‘What is Legal Pluralism?’ (1986) 24 Journal of Legal Pluralism and Unofficial Law 1, 4. 165 N Walker, ‘Beyond Boundary Disputes and Basic Grids: Mapping the Global Disorder of Normative Orders’ (2008) 6 International Journal of Constitutional Law 374. 166 ibid 373ff. 167 See Halberstam, ‘Federalism’ (2011) 49.

114 Debates that includes, to a greater or lesser extent, certain principles of supremacy and homogeneity,168 as well as a legal hierarchy that is usually guaranteed by an ultimate judicial authority of constitutional interpretation. As outlined so far, there are complex and manifold links between federalism and pluralism. Since both federalism and the above-mentioned separation of powers have their roots in ancient times,169 they appear to be the oldest and plausibly most developed tools of what we may call institutional pluralism. In contemporary societies, however, this mere institutional dimension seems no longer sufficient because the increasing diversity of interests makes it practically impossible to have all of them channelled through the classical institutions of representative democracy, be they at the subnational or other levels of government. There seems to be a need for additional and supplementary forms of decision-making that increase the degree of pluralism and thus enhance the overall legitimacy of contemporary constitutions. This need has given rise in recent decades to participatory democracy in a broad sense.170 Against this backdrop, it seems fair to say that there is a link between federalism and participatory democracy and that it lies in their ‘common innate connection with pluralism … Both are expressions of pluralism: institutional and societal, respectively.’171 4.3.2 The Crisis of Democracy: An Opportunity for Federalism? ‘If the fear of being bypassed by popular initiatives or overridden by popular referendums was enough to force public officials to behave honestly and responsively, well and good; if not, the people themselves would simply take over.’172 This statement about the rationale behind early twentieth century populism in the west of the United States and Canada is remarkable for two reasons. First, it demonstrates that disenchantment with elected representatives, perceived as inherently biased towards promoting special interests, is on no account a new phenomenon. Second, it illustrates the attempt as a remedy to supplement representative democracy with another form of decision-making. This was, at that time, direct democracy, which came to be introduced in both countries precisely at the subnational level.173 Today, the focus of democratic innovation has arguably shifted from direct democracy to the abovementioned participatory democracy.174 A crucial question then is whether federalism

168

See ch 5.1.2. Concerning federalism, see ch 3.1, and regarding the separation of powers, see MJC Vile, Constitutionalism and the Separation of Powers, 2nd edn (Indianapolis, Liberty Fund, 1998) 1ff. 170 For definitional issues, see section 4.3.2 below. 171 F Palermo, ‘Participation, Federalism, and Pluralism: Challenges to Decision Making and Responses by Constitutionalism’ in C Fraenkel-Haeberle et al (eds), Citizen Participation in Multi-Level Democracies (Leiden, Brill Nijhoff, 2015) 44. 172 D Butler and A Ranney, Referendums: A Comparative Study of Practice and Theory (Washington, American Enterprise Institute, 1978) 27. 173 For the western United States, see GA Tarr, Understanding State Constitutions (Princeton, Princeton University Press, 2000) 150ff; for the Canadian Prairie provinces, see D Laycock, Populism and Democratic Thought in the Canadian Prairies, 1910 to 1945 (Toronto, University of Toronto Press, 1990). 174 See L Morel, ‘Referendum’ in Rosenfeld and Sajó (n 11) 507. 169

Federalism and Participatory Democracy 115 is rather conducive to the novel tools proposed under this broad umbrella term or not. In other words, does federalism, as an expression of institutional pluralism, also reinforce instruments of societal pluralism or rather constrain them? At the outset, it is important to make clear that neither direct nor participatory democracy has sought to abolish representative democracy. Both have been merely aimed at complementing it by introducing additional dimensions of popular involvement. That is why, empirically, representative democracy has remained clearly prevalent and has been rightly qualified, from that perspective, as the ‘proper form of democracy’.175 From a normative point of view, the fundamental question about the configuration of democracy boils down to two contrasting views: whether the people should only choose their legislators or be involved in the process of law-making themselves, either by directly making decisions or by, at least, preparing and influencing them. The latter idea, which forms the foundation of both direct democracy and participatory democracy, was espoused most notably by Jean-Jacques Rousseau.176 The first view, however, has certainly prevailed among constitutional theorists.177 Beyond that, representative democracy has been, over time, closely intertwined with the broader concept of liberal democracy because it came to be seen as a guarantee for a rational legislative process and thus ultimately for ‘government of law and not of men’ (Article 30 of the 1780 Constitution of Massachusetts). Yet, it has also been seen, at times, as requiring complementary forms of democracy. As for direct democracy, its introduction was proposed as early as in the late nineteenth century by movements in Switzerland and North America and by prominent constitutional theorists,178 quite notably by AV Dicey even for the United Kingdom of Great Britain and Ireland, the prototype of a parliamentary system.179 In fact, the Diceyan approach of introducing direct democracy as a complement only regarding particularly important issues such as constitutional matters gained much support during the twentieth century on the part of theorists and in practice. Over the last three decades in particular, we have arguably witnessed a ‘new wave of “direct constitutional democracy”’180 that has included, perhaps most controversially, the recent territorial referendums concerning independent statehood in Scotland and other cases.181

175 See E-W Böckenförde, ‘Mittelbare/repräsentative Demokratie als eigentliche Form der Demokratie: Bemerkungen zu Begriff und Verwirklichungsproblemen der Demokratie als Staats- und Regierungsfrom’ in G Müller et al (eds), Staatsorganisation und Staatsfunktion im Wandel: Festschrift Kurt Eichenberger zum 60. Geburtstag (Basel, Helbing & Lichtenhahn, 1982). 176 Famous is Rousseau’s criticism of British representative democracy: ‘The English people thinks it is free. It greatly deceives itself; it is free only during the election of the members of Parliament. As soon as they are elected, it is a slave’. See JJ Rousseau, On the Social Contract (New York, St. Martin’s Press, 1978) 102f. 177 For an overview, see Morel, ‘Referendum’ (2012) 502. 178 See, for example, H Kelsen, Vom Wesen und Wert der Demokratie, 2nd edn (Tübingen, Mohr Siebeck, 1929) chs III and IV; R Carré de Malberg, ‘Considérations théoriques sur la question de la combinaison du référendum avec le parlementarisme’ (1931) 48 Revue du Droit Public 225. 179 See AV Dicey, A Leap in the Dark, 2nd edn (London, John Murray, 1911) 19. 180 S Tierney, Constitutional Referendums: The Theory and Practice of Republican Deliberation (Oxford, Oxford University Press, 2012) 1. 181 See Kössler, ‘Territorial Referendums’ (forthcoming).

116 Debates Nonetheless, the traditional combination of merely two forms of democracy, which individual constitutions mix, of course, quite differently, has been facing increasing criticism since the 1960s for not allowing real participation.182 The inappropriateness of this combination today is also reflected, for example, in the statement that democratic decision making resembles the penny-farthing bicycles of the late nineteenth century: the small rear wheel is direct democracy, the big one is representative democracy, and on it sits the driver (the executive). Such a bike was a revolutionary innovation at the time it was invented, but for contemporary standards it is far too slow.183

The anachronistic character of the traditional blending of representative and some direct democracy is also reflected in current diagnoses of a crisis of democracy or, considering that the ‘history of democracy is also a history of its crises’,184 at least a crisis of its identity. A testament to that search for identity is the emergence of innumerable ‘hyphenated democracies’, such as ‘post-democracy’, ‘audience-democracy’ or ‘unpolitical democracy’ to name just a few.185 What then are the changed circumstances that made the penny-farthing bicycle appear outdated? These certainly include the generally rising expectations regarding the participation of citizens who, having benefitted from broadened access to education and easier access to information, are today increasingly reluctant to leave democratic decision-making solely to elected representatives.186 Another main factor is the naturally and inevitably growing complexity of contemporary societies, which are characterised by an ‘explosion of fact’187 that is impossible to be grasped and managed exclusively by parliamentary representatives. This has led to the involvement in the governance process of an increasing number of actors, both governments and private actors such as interest groups, technical commissions and non-governmental organisations (NGOs),188 which are vested with quite different legitimacies like electoral mandates or expertise. Finally, there has emerged over time an increasing factual pluralism in the above-mentioned sense due to, among other reasons, the extension of suffrage and growing international migration. As a result, ever more diverse interests can hardly be aggregated exclusively by parties in classical institutions of representative democracy. All these changed circumstances arguably require decision-making, especially in view of securing the social acceptance of norms, to move towards more

182 See, eg, B Barber, Strong Democracy: Participatory Politics for a New Age (Berkeley, University of California Press, 1984). 183 Palermo, ‘Participation’ (2015) 33. 184 K-P Sommermann, ‘Citizen Participation in Multi-Level Democracies: An Introduction’ in C Fraenkel-Haeberle et al (eds), Citizen Participation (2015) 5. 185 On these notions, see, respectively, C Crouch, Post-Democracy (Cambridge, Polity Press, 2004) 6; B Manin, The Principles of Representative Government (Cambridge, Cambridge University Press, 1997) 223; N Urbinati, Democracy Disfigured: Opinion, Truth, and the People (Cambridge, Harvard University Press, 2014) 81. 186 On this ‘cognitive mobilisation’, see R Inglehart, The Silent Revolution: Changing Values and Political Styles among Western Publics (Princeton, Princeton University Press, 1977) 295ff. 187 C Geertz, Local Knowledge: Further Essays in Interpretive Anthropology, 3rd edn (New York, Basic Books, 2000) 171. 188 More specifically, on the interplay between governmental and private actors in the context of EU multilevel governance and local governance, see chs 1.2.2 and 9.1, respectively.

Federalism and Participatory Democracy 117 elaborate solutions than just the traditional dualist mix of representative democracy with some direct democracy. Primarily for this reason the focus of democratic innovation has been shifting for well over two decades to participatory democracy.189 It is important to clarify, however, that this notion may have quite different meanings. In the United States, for example, the term participatory democracy typically refers to instruments like referendums, legislative initiatives and recalls. A similarly broad perception of participatory democracy as including direct democracy also characterises Article 1(4) of the Constitution of the Austrian Land of Vorarlberg, according to which ‘[t]he Land is committed to direct democracy in the form of citizens’ initiatives, referendums and popular consultation and also encourages other forms [emphasis added] of participatory democracy’. Nevertheless, the arguably prevailing opinion in democratic theory and practice, which this section follows, denotes participatory democracy as a tertium genus alongside representative and direct democracy. Yet, this differentiation leaves untouched, of course, the enormous internal conceptual diversity within this umbrella term. Participatory democracy is commonly seen as encompassing both deliberative democracy190 and participatory democracy in a more narrow sense,191 sometimes even associative democracy.192 To be sure, the first two notions are distinct in terms of their origins, with participatory democracy being in today’s discourse largely linked to concrete practices implemented in South America and deliberative democracy emanating from theories put forward since the 1980s in Europe and North America. Arguably, there is also a slight difference in emphasis. Whereas participatory democracy is seen as prioritising broad participation in decision-making, deliberative democracy stresses the quality of deliberation as a process, in particular its rational and impartial character.193 Notwithstanding these differences and two largely distinct research communities, both these novel forms of democracy certainly do have much in common. Both aim to attenuate the most negative effects of majority rule, most notably, the marginalisation of minority interests through polarised and polarising yes-or-no decisions. Even if decisions might still be made in the end by the majority of a legislature or by the people in a referendum, popular involvement in the preparation and deliberation beforehand is expected to foster broader social acceptance of the decision and the resulting law or administrative act. How does federalism then relate to the new tools of participatory democracy in the broad sense of the term? A first relevant observation concerns the typical penchant of participatory democracy to small-scale settings. This is illustrated by the

189

See Morel (n 174) 507. See, for example, JS Fishkin, Democracy and Deliberation: New Directions for Democratic Reform (New Haven, Yale University Press, 1991); A Follesdal, ‘The Value Added by Theories of Deliberative Democracy: Where (Not) to Look’ in S Bosson and JL Martì (eds), Deliberative Democracy and its Discontents (Aldershot, Ashgate, 2006). 191 See, for instance, Barber, Strong Democracy (1984); B de Sousa Santos, ‘Participatory Budgeting in Porto Alegre: Toward a Redistributive Democracy’ in B de Sousa Santos (ed), Democratizing Democracy: Beyond the Liberal Democratic Canon (London, Verso, 2005). 192 This form of democracy is aimed at devolving decision-making to voluntary self-governing associations. See PQ Hirst, Associative Democracy (Cambridge, Polity Press, 1994). 193 See J Teorell, ‘Political Participation and Three Theories of Democracy: A Research Inventory and Agenda’ (2006) 45 European Journal of Political Research 787; J Elster, ‘Introduction’ in J Elster (ed), Deliberative Democracy (Cambridge, MA, Harvard University Press, 1998) 8. 190

118 Debates fact, for example, that even ancient Athenian assembly democracy, an early combination of both direct popular participation and a deliberative process, focused on the polis. Only in large modern states, at least those with a unitary structure, was this combination later deemed generally unfeasible and dissolved, with deliberation being transferred from the people to their representatives in parliaments, explicitly characterised as ‘deliberative assemblies’,194 and being thus again applied on a smaller scale. The great advantage of a large state with a federal structure is, of course, the possibility to deviate from the norm of representative democracy at least on smaller territorial scales, namely, in addition to municipalities, also subnational entities. The above-mentioned Swiss Landsgemeinde, a form of local assembly democracy involving all people entitled to vote, epitomises this most clearly.195 Unsurprisingly, this instrument is for practical reasons of scale limited to the local level and two of the smallest cantons. In the United States, by contrast, the much larger size, in terms of population, of subnational entities has prevented the application of a long tradition, particularly in New England, of holding town meetings on local matters to the level of the states. Overall, it therefore seems natural that direct democracy is in these two federations and others like Germany, Austria or Argentina clearly more prevalent at the local and subnational levels.196 While this prevalence is, regarding tools of direct mass democracy, like legislative initiatives or referendums, a matter of choice, it is for more recently invented democratic tools that focus on both broad participation and effective deliberation, a necessary consequence of the problem of scale. The penchant of these tools towards small-scale settings may therefore hardly come as a surprise. It is evident, for instance, in several highly publicised inventions like ‘deliberative polls’197 or neighbourhood meetings on a ‘deliberation day’.198 In terms of applicability on the larger territorial scales of the subnational or even national levels, much certainly depends on the specific instrument. For example, participatory budgeting, ie, collective decision-making about the allocation of a part of public funds, has remained largely confined to the level of municipalities or even boroughs. Exceptions like its application in the Brazilian state of Rio Grande do Sul or the Indian state of Kerala are few. Citizens’ assemblies, by contrast, are deliberative forums that have been implemented even primarily at the subnational level. This holds true, in particular, for the pioneering Canadian examples concerning electoral reform in British Columbia and Ontario.199 After all, these citizens’ assemblies were held in two provinces with a population of a medium-sized European country.

194

E Burke, The Works of the Right Honourable Edmund Burke (London, Henry G Bohn, 1854) 446. See ch 3.3.2. 196 See C Saunders, ‘Comparative Conclusions’ in K Le Roy and C Saunders (eds), Legislative, Executive, and Judicial Governance in Federal Countries (Montreal, McGill–Queen’s University Press, 1996) 373; Morel (n 174) 513. 197 JS Fishkin, When the People Speak: Deliberative Democracy and Public Consultation (Oxford, Oxford University Press, 2009) 25ff. 198 B Ackerman and JS Fishkin, ‘Deliberation Day’ (2002) 10 The Journal of Political Philosophy 129. 199 See J Ferejohn, ‘Conclusion: The Citizens’ Assembly Model’ in ME Warren and H Pearse (eds), Designing Deliberative Democracy: The British Columbia Citizens’ Assembly (New York, Cambridge University Press, 2008). 195

Federalism and Participatory Democracy 119 Yet, for participatory democracy, its self-declared aim of inclusiveness generally still continues to pose a challenge in territorial entities with larger populations. In fact, these large populations aggravate James Fishkin’s famous trilemma asserting that three essential democratic values, namely political equality regarding inclusion and voting, qualified deliberation and mass participation, cannot all be achieved at once.200 To be sure, it seems plausible that participatory democracy’s problem of scale may be attenuated through new communication technologies which have indeed become, over time, important elements of many participatory practices.201 However, this frequent combination with what has come to be known as ‘e-democracy’ or ‘digital democracy’ cannot completely solve the inherent problem. Ambitious proposals, such as linking the above-mentioned tradition of town meetings with the Internet to create a national network of citizens’ assemblies in the US,202 have so far remained unimplemented. To be sure, federal states provide, in view of this problem of scale, more opportunities for participatory democracy because it may create, in the form of subnational entities, smaller territorial frames of reference. It grants these entities a set of autonomous decision-making powers, which may then be exercised with the participation of citizens. It is precisely at this point that we encounter another problem of participatory democracy, that of diluted political accountability. In fact, as Madison has already pointed out, in his plea for representative democracy, a precondition for accountability is a clear distinction between the ‘rulers, men who possess most wisdom to discern, and most virtue to pursue the common good of the society’203 and those they represent. Arguably, the innate complexity of supplementing indirect democracy in a federal state with channels for participation makes accountability and democratic oversight more difficult. After all, this blurs the above-mentioned distinction between the people and their representatives, even at multiple levels, and thus creates opportunities for shifting blame.204 Yet, this risk of intricate decisionmaking seems to be something inevitable because the above-mentioned changed circumstances for democracy in an era of pluralism requires the participation of more governmental and even private actors. In fact, this increased involvement of different institutions and people controlling each other may, provided that their interaction is efficient, actually enhance transparency and accountability.205 Anyway, the concrete relationship in a federal constitutional system between participatory democracy and representative government remains crucial. But while the

200 In Fishkin’s view, all constellations achieving two of these values, described as ‘mass democracy’, ‘mobilised deliberation’ and ‘deliberative microcosms’, will undermine the realisation of the third one. See Fishkin, When the People (2009) 32ff. 201 See BR Barber, ‘Three Scenarios for the Future of Technology and Strong Democracy’ (1998/1999) 113 Political Science Quarterly 573. 202 See K O’Leary, ‘The Citizen Assembly: An Alternative to the Initiative’ (2007) 78 University of Colorado Law Review 1489. 203 The Federalist no 57. For James Madison’s preference of representative democracy over direct democracy, see also The Federalist no 10. 204 See RK Weaver, ‘The Politics of Blame Avoidance’ (1986) Journal of Public Policy 371. 205 See Palermo, ‘Regulating Pluralism’ (2015) 500.

120 Debates latter’s relationship with direct democracy at the subnational level has from the outset sparked numerous constitutional disputes, with those regarding the initiative and referendum procedures of Manitoba and Alberta being early examples,206 participatory democracy has hardly attracted the attention of constitutional courts. This comes as no surprise, of course, as its tools typically do not challenge the predominant role of representative institutions as strongly as direct democracy does. In Italy, at least, the Constitutional Court certainly played a role in the process of complementing indirect democracy at the regional level with participatory practices.207 After several reforms in recent years, Tuscany today stands out with its ‘regional public debate’, as does Emilia-Romagna, drawing partly on Tuscany’s experience, with its ‘organised discussion paths’. Earlier, the latter region had been involved in a constitutional dispute with the national government that had to some extent clarified the potential space for subnational participatory democracy under the Italian Constitution. 4.F Italy 2004: Is Participatory Democracy Inefficient and a Threat to Representative Government? The Statute of the region of Emilia-Romagna introduced two remarkable instruments of participatory democracy. Article 17 provided the possibility of subjecting legislative and administrative acts of general character to a process of public examination, a debate that could be triggered according to the implementing Regional Law 8/2008 by a minimum of 50,000 signatures. Article 19 granted a right to participate in the legislative procedure to all civil society associations that made a formal request. The Italian Government decided to challenge both provisions before the Constitutional Court. In the Government’s view, Article 17 would run contrary to the constitutional requirement that ‘[p]ublic offices are organised according to the provisions of law, so as to ensure the efficiency and impartiality of administration’ (Article 97). The national Government also contested Article 19 of the regional statute because, as it claimed, it would violate Article 121 of the Italian Constitution, according to which ‘[t]he bodies of the Region are: the Regional Council, the Regional Executive and its President’. The Government alleged that the provision in question would alter the system of representative democracy and the role of political parties, which make decisions within the Regional Council by virtue of their electoral legitimacy and through a political contest between the parliamentary majority and minority. The Constitutional Court, however, did not follow this line of argument and upheld both provisions.208 As to Article 17, it held that the process of public examination is not an overburdening procedural requirement that would result in an unconstitutional impairment of efficiency. With regard to Article 19, it ruled that the foreseen participation of associations does not hinder the proper functioning of the regional institutions of representative democracy, but merely facilitates the consultation of interest groups from civil society. To thus sustain the consultative process and make it more transparent negates

206 Reference Re Initiative and Referendum Act [1919] AC 935 (JCPC); R v Nat Bell Liquors Ltd [1922] 2 AC 128 (JCPC). 207 For an overview, see S Piattoni, ‘Reconnecting Representative and Participatory/Deliberative Democracy in Italy’ in J Torfing and P Triantafillou (eds), Interactive Policymaking, Metagovernance and Democracy (Colchester, ECPR Press, 2011). 208 Italian Constitutional Court No 379/2004.

Federalism and Participatory Democracy 121

either the independence of the representative institutions nor the role of the political n parties. Overall, the Constitutional Court’s judgment therefore seems to be in line with an earlier ruling from 2002, in which it had regarded the forms of exercising popular sovereignty as being manifold and not limited exclusively to representative democracy.209

If subnational entities really do provide, as in Italy and other above-mentioned cases, a fertile ground for democratic innovation through participatory practices, there remains the question of which benefits this may entail. Indeed, the development of innovative tools at this level may have positive effects either directly for the subnational entity concerned or indirectly for others. In the first regard, the possibility of creating a distinct individual mix of representative, direct and participatory democracy enables correspondence with a specific subnational political culture that may be inclined more towards one or another form of democracy. For example, this differentiation is evident in the rather diverse configurations of democratic decision-making among Swiss cantons or Italian regions, with the above-mentioned Tuscany and Emilia-Romagna taking a decidedly more participatory approach. Finding a ‘democratic background consensus’ that reflects a common political culture seems easier at the subnational than higher levels of government and virtually impossible at the global level.210 But yet, there have also been, regardless of their success in practice, interesting experiments of participation at the EU level such as the Open Method of Coordination (OMC) and the European Citizens’ Initiative (ECI).211 Moreover, variation at the subnational level may allow not only for more or less direct involvement of the people, but also for different conceptions of the demos. As a matter of fact, subnational tools of participatory democracy are on the whole more open to include non-citizens, albeit not least because they are usually less empowering than the right to vote in elections or referendums. Indirectly, the success of such tools may have an even broader positive effect if they are then, according to the logic of what is known as experimental or laboratory federalism,212 emulated by others. Even though federalism is not per se a sufficient condition for such processes, a spread of innovative practices, both horizontally between subnational entities and vertically to other government levels, has actually occurred in a number of policy fields.213 As to vertical diffusion, there seems to be, on the one hand, at higher levels of government further removed from citizens, a particular need for participatory democracy as a complement that enhances legitimacy. On the other hand, it is precisely at the national or even supranational levels that the design of such instruments is hampered by the above-mentioned problem of scale,

209

Italian Constitutional Court No 106/2002. See Sommermann, ‘Citizen Participation’ (2015) 4. 211 See M Dougan et al (eds), Empowerment and Disempowerment of the European Citizen (Oxford, Hart Publishing, 2012). 212 On this, see the introduction to Part III. 213 See chs 10–14. 210

122 Debates which requires good practices from lower levels to be considerably reconfigured or makes them utterly impossible. It is hardly surprising, therefore, that evidence regarding horizontal emulation and further development of tools of participatory democracy is without doubt more conclusive. Such dynamics have occurred internally within states, for instance, regarding the above-mentioned practices between British Columbia and Ontario, as well as between Tuscany and Emilia-Romagna. Nevertheless, those practices have failed to reach a more substantial number or even a majority of subnational entities in both the Canadian and Italian case. It is remarkable though that several innovative instruments have been adopted and adapted internationally. This holds true for ideas behind the Canadian citizens’ assemblies travelling as far as Australia and, above all, for participatory budgeting, which has spread in a relatively short time after the first experiments in 1989 in Porto Alegre to local governments worldwide.214 In conclusion, two caveats appear to be important when it comes to the function of federalism for democratic innovation. While intermediary entities between the local and national levels may particularly lend themselves in terms of their size to experiments of participatory democracy, one should be wary of overestimating federalism’s impact. This does not mean that one needs to subscribe to the view that a society achieves what it aims for anyway and regardless of the state structure so that ‘the accidents of federalism (i.e., the constitutional and administrative detail) do not make any difference at all’.215 It means, however, that a federal structure is not per se the driver behind democratic innovation but rather a catalyst, reinforcing pre-existing inclinations, if they exist, to introduce participatory practices, and it may serve as a procedural matrix to implement them. A second caveat concerns the instructive experience that subnational entities are, of course, not always pioneers of democratic reform, from where it can then spread to other sites. On some occasions, they have actually been stumbling blocks for such reform. A most glaring example is the extension of female suffrage in Switzerland, whose differentiation at the cantonal level illustrates the above-mentioned distinct subnational political cultures regarding the configuration of democracy. Whereas women’s suffrage was introduced in some predominantly French-speaking cantons in 1959 and at the national level after a popular referendum in 1971, some mostly German-speaking cantons continued to deny the right to vote at the cantonal level. They thus literally functioned as ‘laboratories of reluctance’.216 The last canton to abandon this denial, Appenzell Innerrhoden, did so only because a Federal Court decision in 1990 forced it to do so.217

214 For an overview, see Y Sintomer et al (eds), Participatory Budgeting Worldwide (Bonn, Engagement Global and Service Agency Communities in One World, 2013) 25ff. 215 WH Riker, ‘Federalism’ in FI Greenstein and NW Polsby (eds), Handbook of Political Science, vol 5 (Reading, Addison-Wesley Publishing Company, 1975) 144. 216 N Schmitt, ‘Subnational Institutional Innovation and Participatory Democracy’ in Palermo and Alber (eds), Federalism as Decision-Making (2015) 482. 217 BGE 116 Ia 359.

Part II

Self-Rule and Shared Rule

124

5 Autonomy of Subnational Entities

T

‘The ideal distribution of powers between governments in a federation is the one in which each government is able to act independently within its own watertight sphere of responsibility.’1

HIS STATEMENT REFLECTS only one of the many institutional approaches to federalism, the one that supports the view that separation rather than integration of jurisdictions allows for a more effective government. Comparative analysis suggests that the interaction between the national and subnational levels of government can be based on other principles too, and in fact is a lot more complex than the ideal type of dual federalism and its watertight compartments doctrine suggests.2 Federalism is, in the end, all about the right mix between self-government and shared government, whereby each country has its own formula to combine these two (apparently opposite) principles. Each federal system, whatever its nature, is composed of two or more orders of government, operating under the authority of the national constitutional framework. This framework provides for a certain degree of shared rule through common institutions and powers at the national level, as well as for a certain degree of self-rule by regulating the scope of subnational units’ jurisdiction and powers.3 The quality and quantity of powers and responsibilities attributed to each level, as well as the overall approach to the distribution of powers (symmetrical or asymmetrical, dual or integrated, affecting all areas of ‘sovereignty’—legislative, administrative and judicial powers—or just some, etc) vary in each case, confirming the absence of a pure, ideal model.4 Rather, there are as many variations as there are federal systems. Nonetheless, regardless of how each federal system comes together and is shaped, in constitutional terms the powers of subnational units can be framed in terms of degree of autonomy, as they either derive from or are guaranteed by, and in any case are exercised within, the limits of the national constitution. Looking at the autonomy of subnational units means analysing the degree of their self-rule. Since both classical, aggregative federal countries and younger, devolutionary federal or regional states assume that subnational units emulate states in their organisation, it is not surprising that the common dimensions of self-rule are those affecting the essential components of statehood. These include the power to 1

KC Wheare, Federal Government, 4th edn (London, Oxford University Press, 1963) 14. On the origins and meaning of this doctrine, see box 5.I and section 5.2.2 below. 3 See DJ Elazar, Exploring Federalism (Tuscaloosa, University of Alabama Press, 1987) 5. 4 See A Gamper, ‘A “Global Theory of Federalism”: The Nature and Challenges of a Federal State’ (2005) 6 German Law Journal 1297. 2

126 Autonomy of Subnational Entities adopt their own constitution—or at least to have their constitutions guaranteed by the national constitution (section 5.1), the power to pass their own legislation (section 5.2), the power to implement legislation and to administer public functions (section 5.3), their own judicial power (section 5.4).5 This is not to say that to be a federal or regional state all these powers should be vested in (or retained by, depending on the historical process) the subnational units: many federal countries, such as Austria, Belgium and South Africa, have, for instance, a centralised judiciary. Rather, it means that the division of jurisdiction, powers and responsibilities is aimed at mirroring the ‘state nature’ of both national and subnational governments, although only to the extent made possible by the historical, political and other circumstances of each country. 5.1 THE CONSTITUTIONAL DIMENSION: CONSTITUTIONAL AUTONOMY

5.1.1 Definition and Scope The first and most relevant aspect of subnational autonomy relates to the constitutional sphere. Constitutional autonomy can take either the ‘pure’ form whereby subnational entities have the power to adopt their own constitutions or the functionally equivalent guarantee of subnational autonomy by the national constitution. The former is the case in most aggregative federal countries. The latter means that constituent units do not have their own written constitution, as their powers and functions are disciplined by the national one (like in Belgium or India),6 or they do have basic laws that are, however, adopted by means of a national constitutional or special law (like the statutes of Spain’s autonomous communities enacted as organic laws).7 As usual, intermediate forms also exist (like in Canada). Due to the influence of the historic formation of each constitution, as a rule in aggregative federal systems the pre-existence of the constituent units left these mostly unchanged, and often the federal constitution was drafted based on the model of those of the constituent units, like in the cases of the United States or Australia.8 This is because the constituent units’ constitutions were fully fledged constitutions, drafted as constitutions of (quasi-)sovereign states. Once again, however, while as a rule one can identify a basic difference between aggregative and devolutionary federations in the different historical process of formation of the respective state, it would be wrong to consider that, in aggregative federal systems, there is full constitutional 5 Of course, there is a fifth, essential element of autonomy: financial autonomy. From a purely legal perspective, financial autonomy can be broken down into its legislative and administrative (and judicial) dimensions. However, given its practical relevance, financial autonomy is treated separately in ch 7. 6 In India, only the state of Jammu and Kashmir has its own constitution. In Belgium, Arts 118(2) and 123(2) of the Constitution grant the Flemish and French communities, as well as the Region of Wallonia, only very selectively the power to adopt formally qualified legislation regarding certain aspects of institutional and procedural self-organisation. 7 See on the 2006 Statute of Catalonia box 5.F below and, more generally, on the status of the autonomy statutes as organic laws ch 10.5. 8 See C Saunders, ‘The Relationship between National and Subnational Constitutions’ in KonradAdenauer-Stiftung (ed), Subnational Constitutional Governance (Pretoria, KAS South Africa, 1999) 30.

The Constitutional Dimension 127 autonomy for subnational units while that is not the case in devolutionary ones: regional Italy has constitutional autonomy for its ordinary regions, while in federal Canada, there are no provincial constitutions in the sense of both formally qualified and materially complete legal documents. The fact that codified subnational constitutions, including fundamental provisions regarding issues such as the conduct of government and fundamental rights, do not exist is in part due to the history of Canada as a former British dominion.9 At the same time, state constitutions in Australia were drafted despite the colonial context. As we have seen in other parts of this book, historical formation and legal culture are essential also in determining the reading of the constitutional autonomy of the subnational units. 5.A Australia 1913–2003, Germany 1955 and Russia 2000: Subnational Constitutions as Remnants of Sovereignty The historical formation of a federal system and each country’s legal culture are understandably reflected in how courts perceive the concept of federalism and its essential elements. In Australia, the High Court has stated that a country is ‘really federal’ only when the constituent units, while accepting the irrevocable transfer of part of their original sovereignty to the central government, keep their original constitutions unchanged. As a consequence, Australia is considered a ‘true’ federation, since the constitutions of individual states have remained substantially unchanged since the establishment of the Commonwealth Government.10 The same Court has repeatedly stated that subnational constitutions provide in principle for full legislative power on the part of the states and that federal competences, while prevailing, should be considered exceptional in the systematic reading of the Constitution11 and that the Commonwealth by no means can exercise its powers in a way that jeopardises the existence or the effective functioning of the states.12 In Germany, several subnational constitutions define the respective Land as a ‘state’ or ‘free state’,13 and the Federal Constitutional Court has declared that the constitutional

9 However, in Ontario (AG) v Ontario Public Service Employees’ Union [1987] 2 SCR 2, the Canadian Supreme Court made clear that what can be identified as ‘provincial constitutions’ are composed by a plurality of legal texts: some still originating from the British Parliament, some from the common law, some are unwritten such as constitutional conventions or fundamental principles. Still other sources are ordinary provincial statutes that deal with typical constitutional matters. Moreover, such matters are extensively regulated by Title V ‘Provincial Constitutions’ and s 133 of the 1867 Constitution Act. According to the Supreme Court these provisions have a dual nature and form part of both the national and provincial constitutions, Quebec (AG) v Blaikie [1979] 2 SCR 1016; Manitoba (AG) v Forest [1979] 2 SCR 1032. Ironically, the only subnational document called ‘Constitution Act’ is in reality only a statute (of British Columbia) and can thus be easily amended by the normal legislative process. Besides, this act is not a systematic consolidated document, but ‘a random collection of miscellaneous sections’, C Sharman, ‘The Strange Case of a Provincial Constitution: The British Columbia Constitution Act’ (1984) 17 Canadian Journal of Political Science 87, 91. 10 Attorney-General (Cth) v Colonial Sugar Refining Co Ltd (1913) 17 CLR 644. 11 Clayton v Heffron (1960) 105 CLR 214. 12 Austin v Commonwealth (2003) 215 CLR 185. 13 Bavaria, Saxony and Thuringia are the three contemporary ‘free states’. Also, usage of the expression ‘free state’ has a historical basis: ‘Freistaat’ was the term used in Germany in the 19th century for a state not ruled by a monarch, ie, for a republic. The current use is just linked to that history and has no legal implications.

128 Autonomy of Subnational Entities

spaces of the federal Government and the Länder are independent and parallel to each other,14 even though the constitutional practice shows that Germany is an outstanding example of intertwined government.15 In Russia, the Constitutional Court explicitly denied the sovereignty of subnational units by finding the provision of the Constitution of Altai asserting republican sovereignty and declaring that the Republic of Altai ‘owns its land and territory’ to be in violation of the national Constitution. The Court pointed out that, in the Russian context, no other source of power exists besides the ‘multinational people of Russia’ and thus the sovereignty of the federation precludes the existence of other levels of sovereignty.16

Formally, constitutional autonomy is simply a special legislative power of the subnational unit: the power to adopt the unit’s own constitution as an act of legislation. However, a constitution differs from ordinary legislation in both form and substance. As to the form, the constitution is adopted by means of an entrenched legislative procedure, requiring, inter alia, a qualified majority for adoption by parliament, sometimes a referendum, or even the election of special bodies in charge of the drafting of the constitution. In the United States, most subnational constitutions were approved by special constitutional conventions and ratified by popular vote. The amendment power is normally vested with the state legislature (with qualified majority), but in some cases it can be initiated by a qualified number of voters. In Switzerland, cantonal constitutions (in conformity with the federal one) need to be approved by popular vote, and the same goes for their amendments, which can also be initiated by popular initiative. In Germany, the constitutions of most Länder are adopted and approved by the regional parliament, either with17 or without a successive popular referendum;18 only exceptionally have constitutions been approved by a constitutional assembly and ratified by popular vote.19 Typically, constitutional autonomy implies that subnational constitutions themselves determine the procedure for their own amendment, although within the limits of the homogeneity clause20 or of a functionally equivalent certification by the central authorities:21 this is again the rule in aggregative federal systems such as the United States,22 Germany, Switzerland and others, whereas in devolutionary systems such as Spain, Italy or South Africa the procedure for subnational constitutional amendment is normally determined by the national constitution.

14

4 BVerfGE 178, 189 (State Law on Administrative Jurisdiction). See section 5.2.2 below. 16 Russian Constitutional Court, judgment of 7 June 2000, no 10-P, Vestnik Konstitutionnogo Suda RF 2000 no 5. See JP Goode, The Decline of Regionalism in Putin’s Russia: Boundary Issues (Abingdon, Routledge, 2011) 92. 17 Berlin, Brandenburg, Bremen, Mecklenburg-Vorpommern, North Rhine-Westphalia, Thuringia. 18 Hamburg, Lower Saxony, Saxony, Saxony-Anhalt, Schleswig-Holstein. 19 Bavaria, Hesse, Rhineland-Palatinate. 20 See section 5.1.2 below. 21 On Switzerland and South Africa, see ch 8.2.2. 22 See GA Tarr (ed), Constitutional Politics in the States: Contemporary Controversies and Historical Patterns, (Westport, Greenwood Publishing Group 1996) 40ff. 15

The Constitutional Dimension 129 With regard to the contents, a constitution regulates issues that are constitutional in nature,23 such as the system of government, the composition, election and function of the legislature and the executive, instruments of direct democracy, the organisation of local government24 and often further elements such as (sub-)national identity, the use of language(s) and fundamental rights, at least in addition to those guaranteed by the national level.25 To be considered as such, a (subnational) constitution generally needs to differ from (subnational) ordinary legislation in both form and substance and to express a margin of its own authority, not being a mere specification of the national constitution. If these elements are given, a (subnational) constitution can be considered as such regardless of its formal name like constitution, statute, basic law, etc. As a rule, subnational constitutions that historically preceded the national one (as is usually the case in aggregative federal systems)26 have a broader scope than subnational constitutions adopted after them.27 Asymmetry of the federal system28 is often reflected in asymmetrical constitutional autonomy.29 In Belgium, subnational units do not have their own constitutional documents, and in South Africa only one province has its constitution in force after fulfilling the criteria posed by the national Constitution (Articles 142–45). In Italy, ordinary regions adopt their constitutions, the so-called statutes, as their own entrenched laws (Article 123 of the Italian Constitution), while the statutes of the special regions are adopted as constitutional laws of the state (Article 116), although the seemingly broader constitutional autonomy of the former does not correspond overall to a broader but rather to a more limited subnational autonomy. In the Russian Federation, only republics can adopt their own constitutions, whereas the other federal units (krais, oblasts, federal cities, autonomous oblasts and autonomous okrugs) can only adopt charters or statutes, although the distinction is essentially semantic due to the principle of equal standing of the constituent units.30 In the early years after the enactment of the Russian Constitution of 1993, the Constitutional Court closely scrutinised the establishment by the subnational entities of own systems of government (Article 77 of the Constitution). It held, for example, that the assumption of vast powers by Altai’s Legislative Assembly over the executive

23 See HJ Boehl, Verfassungsgebung im Bundesstaat: Ein Beitrag zur Verfassungslehre des Bundesstaates und der konstitutionellen Demokratie (Berlin, Duncker & Humblot, 1997) 171ff. 24 See ch 9. 25 See ch 10. 26 This depends on the history of each country and can happen regardless of the aggregative or devolutionary formation of federalism. In Germany, for instance, while historically the Länder preceded the federation, the formation of the democratic state after World War II had different dynamics, and only a few subnational constitutions (notably the Bavarian one) were adopted before the Basic Law. Conversely, for instance, Sicily in Italy adopted its own constitution in 1946, before the national Constitution was approved (1948) and had then to adapt to it accordingly. 27 This is particularly evident in the field of fundamental rights. See ch 10. 28 See ch 2.2.2. 29 On the causes and dynamics of constitutional asymmetry, see F Requejo and K-J Nagel (eds), Federalism beyond Federations: Asymmetry and Processes of Resymmetrization in Europe (Ashgate, Farnham, 2011). 30 See GA Tarr, ‘Subnational Constitutional Space: An Agenda for Research’ (World Congress of the International Association of Constitutional Law, Athens, 2007) 5f.

130 Autonomy of Subnational Entities branch was unconstitutional because it violated the principle of the separation of powers enshrined in the federal Constitution.31 Anyway, in light of the establishment in 2000 of seven (now nine) federal districts, which are groupings of the above-mentioned subnational entities and headed by a presidential envoy, the effective scope of both the constitutions of republics and the charters or statutes of the other entities has become unclear.32 The fact that the governors of the constituent units are, since a federal law was enacted in 2004, not elected by the people anymore, but appointed by the Russian President and the subnational legislatures, was not regarded by the Constitutional Court as a violation of electoral rights.33 In India, states are not entitled to adopt their own constitution, except Jammu and Kashmir (Article 370 of the Indian Constitution), although a special procedure is provided in order to allow each state to determine whether its legislature should be unicameral or bicameral.34 From a comparative perspective, in a number of federal countries subnational entities are faced with the problem that their formal constitutional autonomy has been considerably eroded. This holds true for the above-mentioned constituent units of the Russian Federation and Jammu and Kashmir, as well as for the state of Sarawak in Malaysia. There, a protracted legal dispute about the dismissal of Sarawak’s Chief Minister indicated already in the 1960s a lack of legal protection against the national Government’s interference with a state constitution.35 5.1.2 Limits to Constitutional Autonomy: Supremacy and Homogeneity Clauses As subnational constitutions are part of the overall legal system of a country, they must be in compliance with the national constitution. Subnational units are not entirely sovereign,36 and their constitutional autonomy is to a larger or narrower extent always limited by the national constitution. The degree of such limitation is determined primarily by two constitutional tools that exist, in one form or another, explicitly or implicitly, in all constitutions of compound states: supremacy clauses and homogeneity clauses, as well as, of course, their judicial interpretation. Supremacy clauses are simpler in their structure and function, as they establish that the national constitution and federal law in general, takes precedence over any piece of subnational legislation, even over subnational constitutions. The prototype

31 Constitutional Court, judgment of 18 January 1996, no 2-P, Vestnik Konstitutionnogo Suda RF 1996 no 1. 32 See A Heinemann-Grüder, ‘Russian Federalism—Legacies, Reforms, and Prospects’ (2002) 3 Indian Journal of Federal Studies 69. 33 Constitutional Court, judgment of 21 December 2005, no 13-P, Vestnik Konstitutionnogo Suda RF 2006 no 3. 34 Art 169 of the Indian Constitution provides that state legislative assemblies can pass a resolution demanding the creation or abolition of a second chamber, the Legislative Council (Vidhan Parishad), and the national parliament can make it a law through an ordinary bill that does not require any amendment of the Constitution. 35 Stephen Kalong Ningkan v Tun Abang Haji Openg and Tawi Sli [1966] 2 MLJ 187; Stephen Kalong Ningkan v Government of Malaysia [1968] 2 MLJ 238. 36 On the question of divided sovereignty, see ch 4.1.2.

The Constitutional Dimension 131 of federal supremacy clauses is laid down in Article VI, clause 2 of the US Constitution, declaring federal law ‘the supreme law of the land’, notwithstanding any contrary law of the states.37 Accordingly, any piece of federal law, even a regulation of a federal agency, trumps any conflicting state law, including the state constitution. 5.B United States 1958: Supremacy and the Prohibition of Elusion Despite the recurrent efforts of some states to ‘nullify’ federal laws they disagree with, the interpretation of the Supremacy Clause is quite consistent and unequivocal. In 1958, the US Supreme Court, unanimously, spoke clear words in Cooper v Aaron,38 a famous case in which the Court scrutinised the efforts of state authorities to block integration in the Central High School of Little Rock (Arkansas), despite the famous 1954 ruling in Brown v Board of Education39 and the subsequent school desegregation policy. The governor and the legislature of Arkansas contended that they were not bound by Brown and by its holding that ‘the Fourteenth Amendment forbids states to use their governmental powers to bar children on racial grounds from attending schools where there is state participation through any arrangement, management, funds or property’. The state authorities maintained that they were entitled to suspend the school board’s plan to do away with segregated public schools in Little Rock until state laws and efforts to upset and nullify Brown had been further challenged and tested in the courts. In Cooper v Aaron, the Court held that ‘[n]o state legislator or executive or judicial official can war against the constitution without violating his undertaking to support it … If the legislatures of the several states may, at will, annul the judgments of the courts of the United States and destroy the rights acquired under those judgments, the Constitution itself becomes a mockery.’ As the supreme law of the land, federal law cannot be eluded by state laws.

The most immediate consequence of a supremacy clause in terms of federal theory is what Americans call pre-emption, while other federations use other notions of inconsistency. Anyway, pre-emption or inconsistency poses some difficult questions. When does a state law or action that is at least arguably consistent with federal law create sufficient conflict so as to justify finding it ‘pre-empted’, ie, precluded by existing federal legislation? Pre-emption can be either express or implied. When the US Congress chooses to expressly pre-empt state law, the only question for the courts becomes determining whether the challenged state law is one that the federal law is intended to pre-empt. Implied pre-emption is more subtle, at least when the state law in question does not directly conflict with federal law. The Court then looks beyond the express language of federal statutes to determine whether Congress has ‘occupied the field’ in which the state is attempting to regulate, or whether a state law directly conflicts with federal law, or whether enforcement of the state law might frustrate federal purposes.

37 ‘This Constitution, and the Laws of the United States which shall be made in pursuance thereof; and all treaties made, or which shall be made, under the authority of the United States, shall be the supreme law of the land; and the judges in every state shall be bound thereby, anything in the constitution or laws of any state to the contrary notwithstanding.’ 38 Cooper v Aaron 358 US 1 (1958). 39 Brown v Board of Education 347 US 483 (1954).

132 Autonomy of Subnational Entities

5.C United States 1956: Pre-emption of State Laws Federal pre-emption or ‘occupation of the field’ occurs, according to the US Supreme Court, when there is no room left for state regulation. Courts are to look to the pervasiveness of the federal scheme of regulation, the federal interest at stake and the danger of conflict between federal and state enforcement efforts in making the determination as to whether a challenged state law can stand. These three criteria date back to a seminal ruling in 1956. The State of Pennsylvania had convicted Steve Nelson, a member of the Communist Party, under the Pennsylvania Sedition Act, which made it a state crime to advocate the violent overthrow of the US Government, and sentenced him to 20 years in jail. Nelson claimed that a federal law, the Alien Registration Act of 1940 (‘Smith Act’) pre-empted his conviction. The US Supreme Court held that the Pennsylvania Sedition Act was unenforceable because Congress had indeed ‘occupied the field’ with the just-mentioned law enacted in 1940.40 For the Court, the issue of national security in the broadest sense (it was the time of McCarthyism and of the intelligence concerning ‘Communist seditious activities’ against the government) is ‘a field in which the federal interest is so dominant that the federal system [must] be assumed to preclude enforcement of state laws on the same subject’ (p 503). Interestingly, also the Australian High Court had to deal around the same time with a case involving communism and national government power. The judges invalidated legislation of the federal parliament that had outlawed the Australian Communist Party because this matter did not fall within its powers.41

Supremacy clauses exist in nearly all federal systems. The simplest formulation can be found in Article 31 of the German Basic Law, the shortest article of the federal Constitution, stating, in a crystal-clear way, that ‘federal law trumps Länder law’. Similar provisions are to be found in most constitutions, sometimes even using s imilar terminology as in the United States or in Germany. Further examples of supremacy clauses are the constitutions of India (Article 251), Switzerland (Article 49) and Australia (Section 109). In the latter case, a state law becomes invalid in the event of inconsistency with a federal law, ie, inoperative, so that it may become operative again if the inconsistency ceases to exist.42 Section 109 is therefore an express written guarantee ‘against being subjected to inconsistent demands by contemporaneously valid laws’.43 In Canada, by contrast, the paramountcy of federal legislation does not follow from an explicit provision of the constitutional text but solely from constitutional jurisprudence. Over much of the country’s history, any overlap between federal or provincial laws was regarded as sufficient to render the provincial law inoperative, even where there was no conflict.44 Only more recent case law has

40

Pennsylvania v Nelson 350 US 497 (1956). Australian Communist Party v Commonwealth (1951) 83 CLR 1. 42 Western Australia v Commonwealth (1995) 183 CLR 373 (Native Title Act case). 43 Street v Queensland Bar Association (1989) 168 CLR 461, 522. 44 Ontario (AG) v Canada (AG) [1896] AC 348 (JCPC). 41

The Constitutional Dimension 133 determined that paramountcy can only be invoked in cases of o perational incompatibility, ie, that compliance with one law necessarily entails the violation of the other.45 However, a particular case is Belgium where the ‘decrees’ and ‘ordinances’ adopted by the subnational parliaments are explicitly granted the same legal force as the formal ‘laws’ of the national parliament (Articles 127(2), 128(2), 129(2), 130(2) and 134(2) of the Belgian Constitution). As there is no hierarchy which would give precedence in the case of conflict to the legal acts of one parliament or the other, the only criterion is then the material jurisdiction over the subject matter in question.46 Even if the distribution of powers aimed at creating competences that are as exclusive and clearly defined as possible, both the national and subnational legislatures have attempted, through broad interpretation, to extend their jurisdictions as far as possible. The resulting disputes over material jurisdiction are then solved preventively through consultation in the Council of State or, since 1984, ultimately decided by the Constitutional Court (Articles 141 and 142(2)(1) of the Belgian Constitution). In contrast to Belgium, Austria and Russia are only partial exceptions regarding the supremacy clause. Even though explicit clauses are not contained in the respective federal constitutions, the primacy of federal law, of course, operates there as well. In Austria, the absence of a formal supremacy clause, as well as of a rule on pre-emption, has to do with the Kelsenian understanding of the hierarchy of norms that makes such clauses formally redundant: in theory, separation of competences and hierarchy of norms make a conflict impossible. In practice, however, things are quite different, and the principle operates implicitly. For the Austrian Constitutional Court, it is its duty to strike a balance between the powers of the national government and the Länder, based on the general principle of ‘mutual consideration’ that is functionally equivalent to loyal cooperation.47 Also in Russia a hierarchical relationship is formally excluded (Articles 72 and 76(2) of the Russian Constitution). This is primarily a consequence of the historical context of the formation of the Russian Federation after the dissolution of the Soviet Union. In those first years, Russian federalism was rather chaotic, and secessionist trends were to be found in several subnational units, some of which were against a specific supremacy clause in order to keep the Constitution open ended. According to President Yeltsin’s famous words of 1990, subnational entities were allowed, and even encouraged, to ‘take as much sovereignty as they can stomach’.48 Homogeneity clauses are slightly more subtle and difficult to interpret, as they complement the supremacy principle by establishing that subnational constitutions

45

Multiple Access Ltd v McCutcheon [1982] 2 SCR 161. See F Delpérée, ‘Le Fédéralisme de Confrontation à la Belge’ in J Kramer (ed), Föderalismus zwischen Integration und Sezession (Baden-Baden, Nomos, 1993) 138; F Delmartino, ‘Belgium After the Fourth State Reform: Completed Federalism or Confederalism in the Making?’ in G Färber and M Forsyth (eds), The Regions—Factors of Integration or Disintegration in Europe? (Baden-Baden, Nomos, 1996) 133. 47 VfSlg 10292/1984. 48 Quoted from J Hughes, ‘Managing Secession Potential in the Russian Federation’ in J Hughes and G Sasse (eds), Ethnicity and Territory in the Former Soviet Union: Regions in Conflict (London, Routledge, 2002) 55. This changed very quickly in the following years, especially under Mr Putin’s presidency. 46

134 Autonomy of Subnational Entities (and obviously subnational law more broadly) must comply not only with the text but also with the foundational principles and with the overall ‘spirit’ of the national constitution. This requirement is intrinsically more vague and open to interpretation, and comparative practice shows profound differences among the various experiences. For example, while in the United States the principle has been interpreted as a mere political indication,49 in Austria the federal Constitution contains not only a general homogeneity clause (Article 99 of the Austrian Constitution) but also very detailed provisions regarding the Länder in terms of both their parliaments (Articles 95–98) and their governments (Articles 100–06). 5.D Austria 2001: How Much Can Subnational Units Experiment with Direct Democracy? In Austria, Article 99 of the Constitution provides that the Länder constitutions may not ‘affect’ the federal one. The Constitutional Court constantly interpreted this provision as prohibiting subnational constitutions from contradicting the federal one.50 Thus, the constitutional autonomy of the Länder is limited by both the explicit provisions and the implicit principles of the federal constitution. As of the mid-1980s, a new era of ‘discovering constitutional autonomy of the Länder’51 had begun. In this context, the western Land of Vorarlberg went the furthest in terms of introducing direct democracy in its legislation and constitution. Not only did it create referendums of different kinds, popular initiatives, public consultations and other forms of plebiscites in its legislation, but it amended its constitution by placing ‘popular legislation’ on an equal footing with parliamentary legislation (former Article 33(6) of the Constitution of Vorarlberg). In particular, it was provided that the Land Parliament was obliged to adopt a law reproducing the text of a popular initiative if this initiative was approved in a referendum. The Austrian Constitutional Court found this provision of the V orarlberg Constitution in breach of the representative and the democratic principles of the federal constitution, which impose that legislation shall, as a rule, be adopted by elected a ssemblies and only exceptionally by means of direct democracy.52 For the Court, the federal constitutional order—which cannot be contradicted by subnational provisions according to Article 99 of the federal constitution—is based on the principle of prevalence of parliamentary democracy and this principle prevents the adoption of legislation against the will of an elected assembly (in this case the Land Parliament).

Article 28(1) of the German Basic Law obliges the Länder to conform to the underlying principles of the federal Constitution and thus to be republican, democratic and social states that follow the rule of law. While in its initial case law, the Federal Constitutional Court has pointed out that respect for this homogeneity clause does

49

See box 5.E below. 5676/1968. See A Gamper, ‘The Principle of Homogeneity and Democracy in Austrian Federalism: The Constitutional Court’s Ruling on Direct Democracy in Vorarlberg’ in P Bußjäger and A Gamper (eds), The Homogeneity of Democracy, Rights and the Rule of Law in Federal or Confederal Systems (Vienna, Braumüller, 2003). 51 P Bußjäger, ‘Direct Democracy and Citizens Participation in the Austrian Federal State’ in C FraenkelHaeberle et al (eds), Citizen Participation in Multi-Level Democracies (Leiden, Brill–Nijhoff, 2015). 52 VfSlg 16241/2001. 50 VfSlg

The Constitutional Dimension 135 not automatically imply uniformity in the legal systems of the Länder but only the respect for existential constitutional principles,53 in practice the framework provided by the constitutional clause significantly limits subnational constitutional autonomy.54 The German constitutional structure exemplifies the practical ramifications of the homogeneity principle, which have an impact on the exercise of powers and jurisdictions. A certain uniformity as to the legal and economic framework, as well as of ‘living conditions’ throughout the country, is indeed a condition for adopting federal legislation in the area of concurrent powers (Article 72(2) of the Basic Law).55 In the US case, the national Constitution determines, in addition to the abovementioned Supremacy Clause, that ‘the United States shall guarantee to every state in [the] Union a republican form of government’ and ‘shall protect each of them against invasion and, on application of the legislature (or of the executive when the legislature cannot be convened) against domestic violence’ (Article 4, clause 4, the so-called Republican Clause or Guarantee Clause). The first part of this provision, the guarantee of ‘a republican form of government’, means that the national government is obliged to intervene if a state adopts forms of exercise of power that are non-democratic56 and contrary to the federal Constitution. This clause, however, has not been extensively used in political and judicial practice, except for during the era of reconstruction, ie, after the Civil War. 5.E United States 1849–1963: The Political Question Doctrine At the time of the American Revolution, Rhode Island, unlike other states, did not adopt a new constitution and continued to be governed under the colonial Charter of 1633. Consequently, in the nineteenth century the electoral system was still based on census (and on gender and race), thus excluding from voting not only all women and African Americans, but also many white men. In 1841, a group of ‘suffragists’, advocating the extension of voting rights to all white men, elected a convention that elaborated a new draft constitution. It was subsequently declared that the new text had been adopted by the majority of voters, new elections were called under the new constitution and a suffragist government was elected. The so-called Dorr Rebellion started and the Charter Government introduced martial law. During the rebellion, a suffragist, Martin Luther, was arrested in his home, and he filed a lawsuit for trespass against the soldier who arrested him. The argument for the lawsuit was that the martial law that authorised, inter alia, trespass and home arrests was illegal, as it had been adopted by an illegal government. The Charter government was considered illegal by the suffragists as violating the G uarantee Clause (Article 4, clause 4 of the US Constitution), since it was based on a less democratic

53

9 BVerfGE 268 (Bremen Civil Servant case). 83 BVerfGE 37 (Foreign Voters I). See box 13.C. 55 See section 5.2.3 below and ch 12.3. 56 The understanding of democracy in such a context is that of representative democracy shaped by James Madison: ‘A republic, by which I mean a government in which the scheme of representation takes place, opens a different prospect, and promises the cure for which we are seeking. Let us examine the points in which it varies from pure democracy’ (The Federalist no 10). 54

136 Autonomy of Subnational Entities

constitution.57 When the case reached the US Supreme Court, the Court held that the question of the legitimate government in a state is a political question and only Congress has the power to recognise the legitimate state government.58 Since then, with few exceptions, the Republican Clause has been considered as a political question rather than a legal issue. The Court reiterated this doctrine, for instance, when the clause was invoked against the use of the initiative for law-making in the state of Oregon.59

While analogous in substance, homogeneity clauses can take several different forms. In South Africa, not only does the national Constitution contain numerous detailed provisions in its Chapter VI on issues that are normally regulated by the subnational constitutions,60 but the provinces can adopt and amend their own constitution only after a certification ruling by the Constitutional Court (Section 144 of the South African Constitution) that confirms consistency with the national constitutional requirements regarding both the adoption procedure (Section 142) and content (Section 143).61 As acknowledged in all the above-mentioned cases, it is uncontested that the scrutiny of the homogeneity requirement—to the extent this is carried out by the courts—includes respect not only for explicit constitutional provisions but also for the spirit and the implicit principles of the national Constitution. As the content of subnational constitutions is inherently political, the cooperative principle and even more the cooperative attitude within the overall federal system play a key role.62 5.1.3 Practical Use of Constitutional Autonomy If, on the one hand, the national constitution limits the scope of the constitutional autonomy of subnational units, the same subnational governments, on the other hand, do not always exploit the entire potential of their constitutional autonomy. Many authors share the view that often subnational component units make political, qualitative decisions not to assert their subnational constitution-making competency, not to occupy fully the space legally allotted to them. In fact, it may be that subnational units in federal systems more often underutilize their constitution-making competency than they overutilize it.63

57 See GM Dennison, The Dorr War: Republicanism on Trial 1831–1861 (Lexington, University of Kentucky Press, 1976). 58 Luther v Borden 48 US (7 How) 1 (1849). 59 Pacific States Telephone & Telegraph Co v Oregon 23 US 118 (1912). Another notable example for the political question doctrine is Baker v Carr 369 US 186 (1962). In this judgment the Court held that the redistricting of state legislative districts was not a political question and thus justiciable and it identified six characteristics of cases involving political questions. 60 See Tarr, ‘Subnational Constitutional Space’ (2007) 7. 61 On the declined certification of the KwaZulu-Natal Constitution, see box 8.H. 62 See section 5.2.1 below. 63 Tarr (n 30) 14f. See also C Saunders, ‘Legislative, Executive and Judicial Institutions: A Synthesis’ in K Le Roy and C Saunders (eds), Legislative, Executive and Judicial Governance in Federal Countries (Montreal, McGill–Queen’s University Press, 2006) 372.

The Constitutional Dimension 137 Other scholars, however, see it differently: ‘What we have found is that subnational constitution-makers have made rather significant use of the constitutional space allotted to them, but in some areas more than in others’.64 In fact, the degree to which subnational governments tend to utilise their constitutional autonomy varies depending on the subject field. As far as the governmental system is concerned, there is usually broad similarity between the national constitution and the subnational ones, both because many federal countries have been established by the union of existing states, copying their governmental system, and because of the pervasive influence of the homogeneity principle. The choice for a parliamentary or presidential system is therefore valid at both the national and subnational level, including in countries where, like the United States, this is not explicitly mandated by the federal constitution. Partial exceptions can be found in Malaysia and in South Africa, where subnational units can adopt different forms of government (including peculiar forms of monarchy) while respecting the foundational constitutional principles, and in Italy, where most regional governors are directly elected while the national level has a parliamentary system. As regards other aspects, such as the structure of the parliament, direct democracy and participatory decision-making, subnational units tend to make broader use of their autonomous constitutional choice. In particular, while federal countries normally have a bicameral national parliament, most assemblies are unicameral at the subnational level.65 Exceptions are the subnational parliaments in Australia and the US, which are bicameral except for Queensland and Nebraska, respectively,66 while in Russia subnational legislatures are as a rule unicameral with few exceptions such as Bashkortostan, and in India only some states have bicameral legislatures. Furthermore, subnational constitutions are as a rule more generous than the federal level with regard to instruments of direct democracy, providing for constitutional assemblies (Switzerland, Mexico), referendums (Germany, United States, Switzerland) and popular initiatives (Switzerland, Mexico, United States, Germany/ Bavaria). Similarly, subnational constitutions tend to experiment more than federal ones with instruments of participatory democracy, such as in some subnational units of Austria, Italy or Spain. Examples include various forms of public debate especially on infrastructure, bodies composed of citizens and experts to advise the government or to propose legislation, and other mechanisms aimed at including citizens in deliberative processes in forms more sophisticated than direct democratic tools and complementary rather than alternative to representative institutions.67

64 JJ Dinan, ‘Patterns of Subnational Constitutionalism in Federal Countries’ (2008) 39 Rutgers Law Journal 837, 841. 65 According to L Massicotte, ‘Legislative Unicameralism’ (2001) 7 Journal of Legislative Studies 151, 151 and 166; out of 450 subnational parliaments, only 73 are bicameral. 66 See RV Shumate, ‘The Nebraska Unicameral Legislature’ (1952) 5 Political Research Quarterly 504, 505. Curiously, the unicameral legislature of Nebraska took the name of Senate, which is usually the name of the second chamber. 67 See Fraenkel-Haeberle et al, Citizen Participation (2015), and F Palermo and E Alber (eds), Federalism as Decision-Making: Changes in Structures, Procedures and Policies (Leiden, Brill–Nijhoff, 2015). See also ch 4.3.

138 Autonomy of Subnational Entities Understandably, the practical use of constitutional autonomy also varies within the same country. In diverse societies especially, the subnational constitution of some entities is often used as a means to affirm a distinct identity. In Spain, many statutes of the autonomous communities (especially those of the ‘second wave’, ie, those adopted between 2006 and 2011) stress the peculiar identity of the respective community. They include references to historical kingdoms and traditions, to historical nations and nationalities and to their ‘own languages’ (Catalonia, Andalusia), traditional rights (Basque Country, Andalusia, Valencia), flags, monuments, anthems (Baleares) and more generally what are called ‘differential facts’ (hechos differenciales).68 5.F Spain 2010: The Catalan Statute and the Limits of Identity Affirmation The new autonomy statute of Catalonia was adopted in 2006 after a long and complex process, both legally and politically. Politically, the new statute reached a difficult and delicate compromise between political forces both in Barcelona and in Madrid: it was supported in particular by the socialist party the Partido Socialista Obrero Español (PSOE) which was then in charge of the Spanish Government and by the majority of Catalan parties. Legally, statutes of autonomy are approved as organic laws which require first the adoption by the regional parliament (the Catalan Parliament approved the text by overwhelming majority in September of 2005) and then by the first chamber of the Spanish Parliament by absolute majority. This happened, after a political agreement which amended the text in May 2006 on nationhood and financing of the autonomous community specifically, with support of all groups except the conservative party, the Partido Popular (PP), which lodged a constitutional complaint, and the Catalan separatist party Republican Left of Catalonia. Finally, the text was submitted to a popular referendum in Catalonia in June 2006, and it was approved with 74 per cent of votes, by a turnout of 49.4 per cent (some 10 per cent less than in the approval of the previous statute in 1979). The text significantly differs from the previous one adopted in 1979 in several ways. In particular, it contains a new financial arrangement (more favourable for Catalonia). It defines the competences of the autonomous community (and in part also of the state) in an extremely detailed way (so called ‘armour-plating’)69 and it strikes a compromise on the nationality question by acknowledging in the Preamble that ‘the Parliament of Catalonia has defined Catalonia as a nation by an ample majority, [while] the Spanish Constitution recognises the national reality of Catalonia as a nationality’. In so doing, the statute clearly tried to reach the maximum degree of constitutional autonomy, including with regard to symbols and separate identity, in particular through the long and detailed Preamble. It took the Constitutional Court of Spain four years to decide on the above-mentioned constitutional challenge against the statute lodged by the PP, as well as by some (neighbouring) autonomous communities. In June 2010, the Court declared 14 articles of the statute unconstitutional, while 27 other provisions were interpreted in conformity with the Constitution, which means that they are constitutional only if interpreted in the way indicated by the Court—a further 120 contested provisions were found to be in line with

68 See G Delledonne and G Martinico, ‘Legal Conflict and Subnational Constitutionalism’ (2012) 42 Rutgers Law Journal 881. 69 See section 5.2.1 below.

The Legislative Dimension 139

the Constitution.70 With regard to the more ‘symbolic’ provisions of the statute, laid down in the first title in connection with the Preamble, the Court rendered an interpretation in conformity with the Constitution, ie, it did not strike them down, although it remarkably reduced their legal impact. As to issues such as the national symbols (Article 8(1)) or the historical rights (that Article 5 of the statute invokes as the foundation of self-government, together with civil law, language and culture), the Court affirmed that they have to be read as attributed to the (Catalan) nationality as part of the indivisible Spanish nation. A community can represent itself as a national reality ‘in an ideological, historical or cultural sense’, and it is obviously possible to request this to be legally acknowledged through a constitutional reform, but as far as the current constitutional framework is not amended, the only nation in legal-constitutional terms is the Spanish one (FJ 12). The same goes for the historical rights: the only legal foundation of self-government is the Constitution (FJ 10). Furthermore, a number of provisions regarding the Catalan language are either declared unconstitutional or interpreted in conformity with the Constitution. Among others, the ‘preferential’ use of Catalan in the public administration and in the public media (Article 6(1)) was found in breach of the linguistic provisions of the Spanish Constitution, and the provision on the ‘duty to know Catalan’ (Article 6(2)) was interpreted as not meaning the same as the general duty to know Castilian laid down in the Spanish Constitution, being rather restricted to some areas only. Furthermore, the right of pupils to be taught in Catalan (Article 35(1–2)) is interpreted as a right to be taught this language alongside Castilian and by no means as a right to receive all teaching in only one of the two official languages (FJ 24). 5.2 THE LEGISLATIVE DIMENSION: DISTRIBUTION OF LEGISLATIVE POWERS

5.2.1 Distribution of Powers and the Residual Clause In most federal states, the distribution of powers is based on two connected principles: the principle of enumeration of legislative powers and the residual clause. The former means that legislative powers of the federation are only those expressly listed in the constitution. According to the latter, all legislative (but, in principle, also administrative and judicial)71 powers that are not explicitly reserved to the national level are vested with the subnational units (eg the Tenth Amendment of the US Constitution,72 Article 3 of the Swiss Constitution, Article 30 of the German Basic Law, Article 117 of the Italian Constitution, Article 73 of the Russian Constitution, etc). The reasons are historical and practical. Historically, as ‘model’ federations have been established by the coming together of previously sovereign states, they found it natural to spell out the powers attributed to the national level while retaining all o thers. From a practical perspective, to attribute the general residual 70 STC 31/2010. For a thorough analysis of the seminal judgment, see the special volume ‘Especial Sentència 31/2010 del Tribunal Constitucional, sobre l’Estatut d’autonomia de Catalunya de 2006’ of the Revista catalana de dret public (2010). 71 See sections 5.3 and 5.4 below. 72 Interestingly, the 10th Amendment does not contain the ‘expressly’ found in the Articles of Confederation. Some members of Congress favoured the inclusion of this word, but James Madison successfully opposed that.

140 Autonomy of Subnational Entities power to one level of government reduces the risk of intergovernmental conflicts because it guarantees that the distribution of competences is complete and covers all unforeseen issues. However, as usual, there are, of course, several exceptions to this ideal type. First, in some federal systems, such as India, South Africa, Canada and Spain, the residual legislative powers belong to the national level and not to the subnational units. Conversely, the residual legislative powers can be vested with the subnational units even in countries that do not consider themselves as entirely federal, like in Italy. Third, in some federal countries the judicial power is not divided between the national and subnational level, but is retained by the federal level; thus, the distribution of powers is limited to legislative and administrative powers only (like in Austria or Belgium).73 Finally, the amount and the broadness of the powers of the national level makes, in some cases, the residual clause a principle more relevant for systematic purposes than in practice (Austria, Italy, Australia). In India, the relationship between enumerated and residual powers is the other way around: residual powers are vested with the union, but the detailed list of state powers is intended to limit the scope of the national government.74 Following a more careful analysis, the real issue is the technique for the distribution of powers. Depending also on the legal history and culture of each country, competences can either be framed quite broadly or they can be extremely detailed. To distribute powers based on broad areas (such as the environment, energy, commerce, trade, foreign relations and so on) is easier, but as a matter of fact it makes it almost impossible to clearly determine in advance whether a specific action falls within one or another area. For instance, does the construction of a nuclear power plant fall within the competence area of energy, the environment, economic development, infrastructure or something else? Do the opening hours of shops affect more commerce, competition, property or the labour market? When such a general division of powers is chosen, areas are more homogeneous, but there is potential for conflicts that are normally adjudicated by the courts (unless they are resolved beforehand through intergovernmental relations).75 Conversely, a detailed definition and enumeration of activities may reduce the room for intergovernmental interpretative conflicts, but this creates fragmentation, bothers the constitutions with a lot of detail and does not entirely eliminate the risk of interpretative conflicts, as the practical reality is always beyond what can be imagined in abstract terms when drafting a constitution. Examples of broadly defined competences are to be found in the US Constitution (and in the federal constitutions inspired by it, such as in Mexico, Argentina, Brazil and, more recently, Bosnia and Herzegovina), as well as in the Australian Constitution. As a matter of fact, the trend is clear: more recently adopted constitutions tend to be more detailed as far as the distribution of powers is concerned, whereas older texts preferred a very broad definition of competences.76 Detailed catalogues can be

73

On the judicial dimension, see section 5.4 below. See A Majeed, ‘Republic of India’ in A Majeed et al (eds), Distribution of Powers and Responsibilities in Federal Countries (Montreal, McGill–Queen’s University Press, 2006). 75 See ch 8. 76 See T Hueglin and A Fenna, Comparative Federalism: A Systematic Inquiry, 2nd edn (Peterborough, Broadview Press, 2015) 43. 74

The Legislative Dimension 141 found, for instance, in the Indian Constitution of 1949, the revised Swiss Constitution of 1999, the Italian Constitution after the revision of 2001, and in the German Basic Law even more after the amendments of 2006. This trend can also be clearly observed in recent subnational constitutions, the most glaring example being the Autonomy Statute of Catalonia of 2006, which introduced an incredibly long and detailed list of powers (Articles 110–73). However, a difference in the rationale and purpose of the overall tendency to provide a much longer and more detailed list of powers in recent constitutions is to be noted between national and subnational constitutions. National constitutions tend to become more detailed in order to try to include the complexity and nuances of competence fields, while subnational ones, if they at all deal with the distribution of powers, usually do so in order to try to protect their competences (in Catalonia, the term blindaje, ie, armour plating was used). The effectiveness of each model, however, does not depend exclusively on the system of the division of powers adopted, but on a number of other factors, including the political culture (whether cooperative or rather adversarial), the factual balance of power, the role and acceptance of constitutional adjudication, the presence of ethno-cultural cleavages. In Germany, for example, the political culture is rather cooperative, and most intergovernmental issues are settled politically, but it is also generally accepted that the ultimate authority in resolving intergovernmental conflicts belongs to the Federal Constitutional Court. In Switzerland, political settlement of disputes is constitutionally mandated as a rule, while litigation must be the exception. According to Article 44(3) of the Swiss Constitution, conflicts between the cantons and between cantons and the confederation must be resolved ‘as far as possible through negotiation and mediation’. To the contrary, in more conflictive systems political negotiation and mediation are more likely to fail, and disagreements are much more often taken to the courts. In Italy, for instance, after the amendment to the Constitution that established that residual powers are vested with the regions and that specified competences to a much more detailed degree, constitutional litigation between the state and the regions over competence conflicts increased significantly (almost 10 times in the first few years after the amendment).77 Similarly, in Spain, the statutes of autonomy of the ‘second wave’ adopted between 2006 and 2011 (inaugurated by the Valencian and Catalonian statutes) and a consistent part of the most significant subnational legislation passed in its aftermath, have been widely challenged in the Constitutional Court, as it was not possible to reach any political agreement—and this was one of the main reasons that triggered the Catalonian question up until the call for independence.78 While a cooperative or adversarial attitude plays a significant role in the p ractical definition of the division of powers, the role of the courts in this respect is of particular significance. This is due to two main reasons. First, regardless of a detailed or rather loose division of powers between the national and the subnational level, constitutional practice constantly presents borderline cases that give rise to divergent

77 While in 1998 litigation between the national government and the regions made up only 2.76% of the Court’s workload, by 2006 it had become 29.16%: see Italian Constitutional Court, ‘Activity Report 2006’ www.giurcost.org/cronache/Relazione_8_2_2007.pdf. 78 See box 5.F above.

142 Autonomy of Subnational Entities interpretations and ultimately need to be settled in court. Second, every constitution contains clauses that can have (and in practice always have had) an expansive effect: their rather open phrasing is often used by the courts to make the division of powers rather flexible and adaptable to the overall circumstances, in particular to the centralising or decentralising trend that marks a particular moment of the constitutional history of a given country. The best-known example of such provisions is the Commerce Clause of the US Constitution. According to Article 1, section 8, clause 3: ‘[t]he Congress shall have power … to regulate commerce with foreign Nations, and among the several States, and with the Indian Tribes’. Despite the deceptively simple language, this provision has been and still is the direct source of the most important powers that the federal Government exercises. From the early years, the clause has been interpreted broadly by the Supreme Court in order to make the federal power effectively work and thus expand its scope. In practice, the Commerce Clause has been read to allow Congress to regulate any activity that could have any impact at all on interstate commerce, leaving Congress free to regulate any economic activity it wishes and displacing any separate sphere for state legislative control of the economy. In several cases, it was assumed that Congress did not even have to intervene to justify a federal power, as many matters were considered to be implicitly falling within the ‘dormant’ Commerce Clause.79 5.G United States 1824–2005: The Commerce Clause and its Impact on the Distribution of Power The first comprehensive expression of such an expansive reading of the Commerce Clause was expressed by the Supreme Court in 1824 in the case Gibbons v Ogden:80 Gibbons was operating steamboats in violation of the New York monopoly, but he was acting pursuant to a federal licence. The Court considered Gibbons’ commercial ship to be included in interstate commerce because it was not possible to limit the area of navigation in advance. According to Chief Justice Marshall, ‘commerce is more than mere traffic’. In Marshall’s time, the Court recognised that the Commerce Clause had a negative implication that restricted state laws that burdened interstate commerce. When the Court struck down a state or local regulatory act as being inconsistent with the dormant Commerce Clause, it was interpreting the silence of Congress to hold that, in the absence of federal legislation, a state or local law could not create a trade barrier or impose a discriminatory burden on interstate commerce.81 This doctrine was for the first time limited in 1852 in Cooley v Board of Wardens,82 when it was held that a Pennsylvania law requiring all ships entering or leaving Philadelphia to hire a local pilot did not violate the Commerce Clause since the crucial factor in

79 ‘The purpose of the dormant commerce clause is to protect interstate commerce from unconstitutional discrimination by the states’: JE Nowak and RD Rotunda, Constitutional Law, 5th edn (St Paul, West Publishing, 1995) 282. 80 Gibbons v Odgen 22 US (9 Wheat) 1 (1824). 81 See also Wilson v Black-Bird Creek Marsh Co 27 US (2 Pet) 245 (1829). 82 Cooley v Board of Wardens 53 US (12 How) 299 (1852).

The Legislative Dimension 143

determining the federal or state jurisdiction is the nature and the goal of the challenged regulation.83 After Cooley, the Court recognised the states’ concurrent power to regulate commerce in appropriate circumstances. In his dissent in 1927 in Di Santo v Pennsylvania,84 Justice Stone advocated a new test, abandoning reliance on labels such as ‘direct’ or ‘indirect’. Instead of rigid verbal formulas, Stone suggested a more pragmatic test. This approach was sometimes adopted during the confrontation between the Court and President Roosevelt: on some occasions, the Court, based on the pragmatic approach, adopted a narrower interpretation of ‘commerce’.85 The pragmatic approach was definitively adopted by the majority in 1945 in Southern Pacific Co v Arizona.86 What matters is the effect of the law, much more than it being labelled as a law on commerce or not. The law from Arizona was struck down because it in fact interfered with interstate commerce and its safety motivations were recessive. If the effect of a state regulation was any form of protectionism, the state law was in breach of the Commerce Clause.87 Such pragmatic rationale was extended and in fact implied a further expansion of federal power. For instance, the Court upheld Congress’s authority under the Commerce Clause to prohibit transportation in interstate commerce of goods made in factories that employed children.88 After the enactment of the Civil Rights Act 1964 in particular, the question became whether there were still activities that were so local that they remained beyond the reach of the commerce power.89 As a matter of fact, from 1937 to 1995, the Court did not strike down any single act of Congress for exceeding power under the Commerce Clause. It was only due to the influence of the so-called New Federalism that the Court reached its current, more balanced approach. In the 1995 ruling United States v Lopez,90 the Court acknowledged that ‘congressional power under the commerce clause … is subject to outer limits’. At stake was a law that made it a federal offence to possess a gun in or near a school. A student carrying a handgun at his school was found guilty of violating this statute, but the claim was made that, as the gun was unloaded and Lopez was to deliver it to another person and was being paid for doing so, it should be covered by the Commerce Clause. The conviction was reversed on the grounds that the law was beyond the reach of the commerce power. In Gonzales v Raich,91 however, the Court held in 2005 that the federal Government could outlaw the use of marijuana for medical purposes under the Commerce Clause, even if the marijuana was never bought or sold and never crossed the state lines of California, where it was used for therapeutic purposes.

83 In the majority opinion, Justice Curtis wrote: ‘It is the opinion of a majority of the court that the mere grant to Congress of the power to regulate commerce, did not deprive the States of power to regulate pilots, and that although Congress had legislated on this subject, its legislation manifests an intention, with a single exception, not to regulate this subject, but to leave its regulation to the several states’. This was called the ‘selective exclusiveness’ doctrine. 84 Di Santo v Pennsylvania 273 US 34 (1927). 85 For example, Carter v Carter Coal Co 298 US 238 (1936). For the Court, the effect of the labour provisions in a law on mining fall upon production, not commerce, and thus within the scope of the states’ legislation. 86 Southern Pacific Co v Arizona 325 US 761 (1945). 87 See also Baldwin v GAF Seelig Inc 294 US 511 (1935). 88 See Nowak and Rotunda, Constitutional Law (1995) 288. See, eg, the Child Labour Case— Hammer v Dagenhart 247 US 251 (1918). 89 See N Redlich et al, Understanding Constitutional Law (Toronto, Irwin Law, 1995) 105. 90 United States v Lopez 514 US 549 (1995). 91 Gonzales v Raich 545 US 1 (2005).

144 Autonomy of Subnational Entities Interestingly enough, a not dissimilar path has been followed by the Court of Justice of the European Union: for decades it interpreted the prohibition to establish measures having an equivalent effect to quantitative restriction (then Articles 28 and 29 of the EC Treaty) in an extremely broad way and by doing so it substantially contributed to the establishment of a common European market. Only when this market was consolidated did the Court for the first time declare, in the so-called Tobacco case in 2000,92 that a European regulation exceeded the limits of the treaties. In general, a generous interpretation of the national (enumerated) powers is more likely in coming-together federal systems, as it is functional to the consolidation of the national jurisdiction vis-a-vis pre-existing subnational regulation. Dynamics, however, can vary considerably depending on circumstances. In Canada, for example, interpretation of the commerce clause has been rather different as compared to the United States. 5.H Canada 1881–2002: The (Lesser-known) Commerce Clause The 1867 Constitution Act granted the federal Government, with the aim to facilitate a common market, exclusive jurisdiction over ‘trade and commerce’ (Section 91(2)) that was, unlike the above-mentioned US provision, not subject to further textual limitations. However, the Judicial Committee of the Privy Council (JCPC), which was entrusted until 1949 with the ultimate interpretation of the Canadian Constitution, already started in its very first case to construe this provision in narrow terms.93 Conversely, it interpreted the provincial competence for ‘property and civil rights in the province’ (Section 92(13)) very broadly because much legislation on trade and commerce would inevitably impact on civil claims and thus fall under the latter provision. The bottom line of the Supreme Court’s interrelated construction of these two provisions has over the years become a reliance on the criterion of territorial relevance. The judges have restricted the federal Government to the regulation of international and interprovincial trade, whereas intra-provincial trade has been regarded as a provincial domain.94 To be sure, the Supreme Court still also favours, in most instances, a broad interpretation of provincial jurisdiction over ‘property and civil rights in the province’, which gives them, for example, vast powers regarding the environment, protection of workers, consumer protection, social welfare and health.95 Nevertheless, its construction of Section 92(13) is, compared to the JCPC, less uniform and sometimes also limits the scope of this provision in favour of federal jurisdiction.96

In view of the US and Canadian examples, even more important than the historic origin, is thus the dynamic of inter-governmental relations in each federal system. The division of legislative powers, detailed as it may be, can never avoid conflicts between the levels of government over the attribution of a specific function. But such a risk is obviously higher when the division of powers is framed in a very general way. Each national constitution contains provisions like the above-mentioned

92

Case C-376/98 Germany v Parliament [2000] ECR I-8419. Citizens Insurance Co v Parsons [1881] 7 AC 96. See Hogg, PW, Constitutional Law of Canada, 5th edn (Scarborough, Carswell, 2000) 20-1ff. 95 See ch 11.2. 96 See, eg, Ward v Canada (AG) [2002] 1 SCR 569. 93 94

The Legislative Dimension 145 commerce clauses and broad transversal matters such as social welfare that may be used as centralising or decentralising factors depending on their political and judicial interpretation.97 In any event, a perfect and uncontested division of powers does not and cannot exist and any federal system is based to some extent on interdependence. A minimum of overlap is explicitly (formal concurrency) or implicitly acknowledged (de facto concurrency).98 An interesting example is provided by Canada, a federation based on a dualist system, even if its Constitution has from the very beginning provided for two areas of concurrent jurisdiction between the federal Government and the provinces (immigration99 and agriculture in Section 95) with other ones added after World War II (old-age pensions and supplementary benefits in Section 94a).100 However, the dualism that the framers of the Constitution had intended has been, over time, considerably attenuated not only by constitutional amendments but even more so by the jurisprudence of the Supreme Court. 5.I Canada 1883–1988: Concurrency as a Fruit of the ‘Living Tree’ Whereas the JCPC had quite strictly upheld the doctrine that the competences of each government level are ‘watertight compartments’101 and thus protected provincial prerogatives even during the economic crisis of the 1930s, the Supreme Court later tended to place more emphasis on other, in part antagonistic, doctrines, namely that of the Constitution as a ‘living tree’ and the ‘double aspect’ doctrine. Even if the JCPC had invented the ‘living tree’ metaphor,102 the doctrine of a dynamic and flexible interpretation has only become prevalent under the Supreme Court, especially but not exclusively in its view of the Canadian Charter of Rights and Freedoms.103 Similarly, the JCPC had already acknowledged that in certain cases one aspect of a subject may fall within the powers of the provinces and another aspect within those of the national government,104 but only the Supreme Court has created a powerful doctrine on this basis. In such conflicts between national and provincial legislation referring to different aspects but regulating the same subject, it insists on the precedence of federal law and thus opens the door for intervention in previously ‘watertight compartments’. Especially in light of the broad wording of national and provincial powers, there seems to be a ‘risk that these two fields of exclusive powers will be combined into a single more or less concurrent field of powers governed solely by the rule of paramountcy of federal legislation.’105

97 Examples are the Necessary and Proper Clause of the US Constitution (Art I sec 8 cl 18) or the so-called POGG clause of Art 91 of the Canadian Constitution (see sections 5.2.2 and 5.2.3 below), but also the prohibition of measures equivalent to a quantitative restriction to the free movement of goods laid down in Art 34 TFEU (former Art 28 of the EC Treaty). 98 See N Steytler (ed), Concurrent Powers in Federal Systems: Meaning, Making, and Managing (Leiden, Brill–Nijhoff, 2017). 99 See ch 13.2. 100 See ch 11.2. 101 Canada (AG) v Ontario (AG) [1937] AC 326, 354 (JCPC). See box 14.B. 102 ‘The British North America Act planted in Canada a living tree capable of growth and expansion within its natural limits’. Edwards v Canada (AG) [1930] AC 124, 136 (Persons Case). 103 Reference Re BC Motor Vehicle Act [1985] 2 SCR 486 para 53. See, more recently, also Reference Re Same-Sex Marriage [2004] 3 SCR 698. 104 Hodge v The Queen (Ont) [1883] 9 AC 117 (JCPC). 105 Bell Canada v Quebec (Commission de la santé et de la sécurité du travail) [1988] 1 SCR 749, 766, dissenting opinion of J Beetz.

146 Autonomy of Subnational Entities From a more general perspective, competence fields are increasingly complex following the complexity of contemporary societies.106 Thus, the predisposition towards cooperation or conflict and the existence and effectiveness of intergovernmental links and forms of cooperation are decisive in order to determine the distribution of powers in practice.107 5.2.2 Dual vs Cooperative Systems Legislative powers can be exclusively attributed to each level of government as if they were watertight compartments (so called dual federalism) or they can be shared between them (so-called cooperative federalism). As a rule, Anglo-Saxon federal countries tend to prefer the first model, whereby each layer of government has exclusive jurisdiction on the matters reserved to it, whereas European federal systems are generally based on the second, ie, the levels cooperate in legislation and, consequently, in administration. Within this general framework, four types of legislative powers can be identified: 1) Exclusive powers are those attributed to one level of government only, excluding the possibility for other levels to intervene. These are typically contained in all federal systems with regard to the powers of the national government and, in principle, with regard to the residual powers of the subnational entities.108 In practice, however, the exclusiveness of competences is mitigated by the realities of overlapping jurisdictions. This is particularly obvious in the case of Belgium’s tripolar distribution of powers between the federal Government, communities and regions. To be sure, the principle of exclusive powers was regarded by the framers of the Constitution as a recipe to prevent intergovernmental conflicts and it inspired constitutional jurisprudence in some cases to interpret subnational competences extensively, ie to give them plenary powers (‘la plénitude de compétence’).109 On the other hand, however, case law also acknowledged in other instances that the complexity of Belgian’s distribution of powers makes de facto overlaps of jurisdictions inevitable.110 2) Concurrent powers can be exercised by either one of the levels, and the legislation issued by one level pre-empts the others from adopting laws. This system is used, inter alia, in Switzerland (Articles 74, 123, 62 and 63 of the Constitution), Australia (Articles 51 and 109 of the Constitution), India (Article 246 and List III of Schedule 7 of the Constitution),111 as well as Germany. According to

106

See Part III. See Majeed et al (eds), Distribution (2006) 5; R Watts, ‘Comparative Conclusions’ in Majeed et al (n 74) 323 and 343. 108 As mentioned in section 5.2.1 above, in some federal systems residual powers are vested with the national level. 109 See Delpérée, ‘Le Fédéralisme’ (1993) 136. 110 Belgian Constitutional Court No 166/2003. 111 On state occupancy of a field until the enactment of federal legislation, see Western Coalfields Ltd vs Special Area Development, 1982 AIR 697. 107

The Legislative Dimension 147 Articles 72 and 74 of the German Basic Law, the national Parliament can legislate only concurrently in key fields such as civil law, residence of foreigners, taxation, criminal law, social security and territorial planning. To do so, in most cases it must prove that unitary legislation is necessary to achieve the unity of the legal and economic system in the national interest, or equivalent living conditions throughout the national territory, and in other cases it can do so without even providing necessity, which makes these areas, in practice, subject to federal legislation only.112 When comparing distributions of powers, it is important, however, to be cautious with terminology. What is called ‘concurrent powers’ in Mexico, by contrast, means something entirely different. It denominates areas regarding which the federal Parliament may pass legislation that distributes responsibilities between the two or even levels of government.113 3) Shared powers are those in which the national level legislates as to the principles (also called ‘framework legislation’), and the subnational units adopt detailed legislation: examples can be found, inter alia, in Italy (Article 117(3) of the Italian Constitution) and Austria (Article 12 of the Austrian Constitution). Article 149(1) of the Spanish Constitution lists a number of subject matters as exclusive national powers, but sometimes restricts the legislation of the Spanish Parliament to regulating the ‘bases’ or ‘basic norms’ of these matters. The Constitutional Court has, however, interpreted the scope of these framework laws rather extensively.114 The same holds true for South Africa, where national framework legislation has been regarded by the judges as a characteristic of the constitutionally entrenched cooperative government approach115 and as, therefore, required regarding a number of subject matters.116 In the case of Bosnia and Herzegovina, the Constitutional Court used a different reasoning to recognise a national competence of framework legislation, even in some areas that are, according to the text of the Constitution, exclusive subnational powers.117 Such legislation would not violate the Constitution, if it is required for economic integration, for the establishment of minimum conditions concerning the functioning of the state as a whole and for equal levels of fundamental rights protection throughout the country, especially regarding the official use of languages.

112 In fact, the constitutional reform in 2006 introduced a ‘compensation’ for the Länder in response to the increase in the federal legislative powers: according to Art 72(3) of the Basic Law, the Länder have the right to deviate, in total or in part, from federal legislation by adopting their own laws in a few areas such as banning smoking or opening hours for shops. See F Palermo, ‘What’s Wrong with Concurrency? Comparative Reflections based on Constitutional Reforms in Germany and Italy’ in Steytler, Concurrent Powers (2017). See also ch 12.3. 113 Mexican Supreme Court of Justice, Constitutional Controversy 29/2000. 114 STC 32/1981. 115 On cooperative government in South Africa, see chs 8.1.1 and 8.1.2. 116 Premier, Western Cape v President of the Republic of South Africa 1999 (3) SA 657, para 43. In two immediately following cases, the Constitutional Court established some limits to the legislative powers of the national government. See Executive Council of the Province of the Western Cape v Minister for Provincial Affairs and Constitutional Development, Executive Council of KwaZulu-Natal v President of the Republic of South Africa 2000 (1) SA 661 and Constitutionality of the Liquor Bill 2000 (1) SA 732. 117 Constitutional Court of Bosnia and Herzegovina, Partial Decision U5/98 IV of 18 August 2000.

148 Autonomy of Subnational Entities 4) Delegated powers are those that one level of government (usually the national one) devolves to another level by retaining the function and thus always being able to withdraw them; one—albeit contested—example is the devolution in the United Kingdom,118 whereby the principle of the sovereignty of the Westminster Parliament implies that powers transferred to the devolved administrations could be taken back anytime.119 These are, of course, general ideal types with several nuances in between. In Canada, for example, the federal Constitution contains two separate catalogues of exclusive competences, one for the federation and one for the provinces. The Constitution Act 1867 was drafted in the wake of the Civil War in the United States and aims to avoid some of the elements of that Constitution that were considered to be responsible for the malfunctioning of that system: Article 91 provides for a list of 29 areas of exclusive jurisdiction of the federal Parliament and Article 92 does the same for the 16 areas of provincial exclusive legislative powers. Moreover, Article 91 of the Constitution Act 1867 contains a residual clause by attributing to the federal Parliament the power to legislate in all areas that are not exclusively reserved to the provinces, in order to guarantee ‘peace, order and good governance’ (the so-called POGG clause). In India, the powers are divided according to separate lists: the union list (97 items), the state list (61 items) and the concurrent list (47 items). The same approach is followed in the Islamic Republic of Pakistan (Articles 141–44 of the Constitution). More generally, modern times are marked by a clear predominance of cooperative elements, including in dual systems. This often happens to a degree that makes it extremely difficult to clearly attribute the responsibility for legislative action to one or another level of government, despite the provisions and certainly the intention of the constitution: a phenomenon that is critically labelled in the Anglo-American area as ‘interlocking federalism’120 and in the German area as the ‘joint-decision trap’ (Politikverflechtungsfalle).121 Quite often, such developments take a turn towards ‘coercive federalism’, above all due to powerful steering instruments by the national level, especially in the form of financial grants.122 Cooperation and mutual consultation are often provided in constitutions as underlying principles of the federal system (eg, Belgium, Switzerland and South Africa) or are declared by the constitutional courts as being inherent to the constitutional system (eg, by the constitutional courts of Germany, Austria, Spain and Italy).123 In conclusion, the increasing importance of cooperation and coordination, on the one hand, and of transversal matters and interpretative criteria, on the other,124 makes it inevitable that the powers of different levels of government will overlap. 118

See ch 2.3.3. On delegation in the context of intergovernmental cooperation, see ch 8.1.3. 120 Watts, ‘Comparative Conclusions’ (2006) 343. 121 FW Scharpf, ‘Die Politikverflechtungsfalle: Europäische Integration und deutscher Föderalismus im Vergleich’ (1985) 26 Politische Vierteljahresschrift 323. 122 J Kincaid, ‘From Cooperative Federalism to Coercive Federalism’ (1990) 509 The Annals of the American Academy of Political and Social Science 139. 123 See ch 8.1.1. 124 See A Gamper, Staat und Verfassung, 3rd edn (Wien, Facultas, 2014) 93f. 119

The Legislative Dimension 149 It is undeniable that, in general, such overlapping has had an overall centralising effect since the claims for coordination and uniformity put forward by the national level have generally been backed by the political process and upheld by the courts. Overall, even when concurrency is just implicit (ie, primarily, in dual systems such as the United States, Australia, Ethiopia and to some degree Brazil), the courts have set some hurdles to the expansive exercise of listed federal powers in what originally have been conceived as subnational matters.125 5.2.3 Flexibilisation and Centralisation: Implied Powers Constitutional developments over time and the judicial interpretation of the division of powers clearly show, especially in the older federal systems, a centralising dynamic in the allocation of legislative powers. The reasons are manifold and vary from country to country: wars, economic cycles, the spreading of fundamental rights,126 the welfare state.127 All this has contributed to centralising the legislative competences (either by means of constitutional amendments or, more frequently, by judicial interpretation) and to making all federal systems de facto cooperative rather than dual ones. As mentioned above, all constitutions include explicit or implicit clauses making it possible to make the distribution of powers flexible and adaptable to changing circumstances. More specifically, each compound constitutional system requires instruments that confer the required flexibility in terms of the allocation of powers. The division of powers, as detailed as it might be, needs to adapt to changing and evolving circumstances. Therefore, it is common that national constitutions contain opening clauses in this regard and that judicial interpretation develops additional tools that are instrumental in achieving this goal. Paradigmatic examples of such clauses and of their judicial interpretation can be found in the United States and Canada. In the latter case, the above-mentioned Article 91 of the Constitution Act 1867 empowers the federal Parliament to legislate in all areas that are not exclusively reserved to the provinces to guarantee ‘peace, order and good governance’ (POGG clause). However, the provision has been interpreted rather restrictively,128 above all compared to the so-called Necessary and Proper Clause of the US Constitution (Article I, section 8, clause 18). This clause grants Congress the power ‘to make all Laws which shall be necessary and proper for carrying into execution the foregoing powers, and all other powers vested by this constitution in the government of the United States, or any department or officer thereof’.129 The Necessary and Proper Clause enables Congress to pass the laws 125

See N Steytler, ‘Concurrency of Powers: The Zebra in the Room’, in Steytler (n 98). See ch 10. 127 See ch 11. 128 In spite of certain cases of extensive interpretation, eg, Reference Re Anti-Inflation Act [1976] 2 SCR 373, the Supreme Court has largely maintained the JCPC’s restrictive view of the POGG clause, eg, R v Crown Zellerbach Canada Ltd [1988] 1 SCR 401. 129 See G Lawson et al, The Origins of the Necessary and Proper Clause (Cambridge, Cambridge University Press, 2010). 126

150 Autonomy of Subnational Entities required for the exercise of its other powers established by the Constitution, and it is not by chance that it is also known as the Elastic Clause. This clause means essentially two things: first, that the national level can make any law it needs to in order to more effectively carry out its enumerated powers; and, second, that the division of powers can never be considered as carved in stone but needs to be allowed the necessary flexibility to adapt to changing (political, societal, economic, historic) circumstances. The logical counterpart of this clause is the flexibilising role conferred upon constitutional adjudication in case of interpretative conflicts. In this context, the Supreme Court developed the so-called Implied Powers Theory. What the Necessary and Proper Clause meant was gradually clarified by the Supreme Court, starting from a leading case dealt with in 1819. 5.J United States 1819: The Origins of the Implied Powers Theory In 1791, the federal Government created a national bank, the Bank of the United States. There is nothing in the Constitution that attributes such power to Congress and actually the general rule is that residual powers belong to the states. In 1816, a second national bank was chartered by Congress. Since several states did not agree politically with such a key decision of the federal Government, they tried to challenge it by, in particular, taxing the new bank, as Maryland did. James McCulloch, a federal cashier at the Baltimore branch of the US bank, refused to pay the taxes imposed by the state. Maryland filed a suit against McCulloch in an effort to collect the taxes. The question ended up in the Supreme Court and went straight to the core of federalism. By asking whether the Constitution gave Congress the power to create a bank and whether individual states could ban or tax such a bank, the issue was about the interpretation of the division of powers and ultimately about the very nature of the federal state.130 The Supreme Court, led by Chief Justice John Marshall, held that Article I, section 8, clause 18 of the Constitution does not empower Congress ‘to make all laws which may have relation to the powers conferred on the Government, but such only as may be “necessary and proper” for carrying them into execution. The word “necessary” is considered as controlling the whole sentence, and as limiting the right to pass laws for the execution of the granted powers to such as are indispensable, and without which the power would be nugatory’ (p 413). Regarding the case concerned the judges decided that the chartering of a bank was covered by the Necessary and Proper Clause. Accordingly, the federal Government had the right and power to set up a federal bank, and states did not have the power to tax the federal Government.

Through a series of subsequent rulings, the Elastic Clause and the Implied Powers Theory developed into a system making ‘the exercise by Congress of power ancillary to an enumerated source of national authority constitutionally valid, so long as the ancillary power does not conflict with external limitations such as those of the Bill of Rights and of federalism’.131 The interpretation of the specific meaning of

130 131

McCulloch v Maryland 17 US (4 Wheat) 316 (1819). LH Tribe, American Constitutional Law, 2nd edn (New York, Foundation Press, 1988) 301.

The Legislative Dimension 151 federalism as a limitation to the expanding role of the national government is also ultimately contextual and does not follow codified rules. It is again for the actors of the constitutional process and most notably for the courts to determine in specific cases whether and to what extent such limitations operate. After an almost uninterrupted series of rulings upholding federal interventions under the Necessary and Proper Clause, the US Supreme Court established an important doctrine supporting states’ rights and limiting federal power. 5.K United States 1997–2012: Not Everything Is Necessary and Proper In Printz v United States,132 the Supreme Court held that Congress could not require state agencies to perform federal duties under the Necessary and Proper Clause. Federal legislation governing the distribution of firearms required the Attorney-General to establish a national system for instantly checking prospective handgun purchasers’ backgrounds but, until the national system became operative, local officers had the duty to conduct such checks. Mr Printz, a law enforcement officer from Arizona, challenged the constitutionality of the federal provision that required him and other local chief law enforcement officials to conduct background checks on prospective gun purchasers. In a five–four ruling, the Supreme Court held that the federal act was unconstitutional. For the majority of the Court, the overall structure of the Constitution prevents Congress from directing state officials: ‘The framers explicitly chose a constitution that confers upon Congress the power to regulate individuals, not states’. Rather, the Constitution established a system of ‘dual sovereignty’, where states surrendered many of their powers to the new federal Government, but retained ‘a residuary and inviolable sovereignty’. It is on this dual structure that American federalism is built. A recent case of denial of congressional authority under the Necessary and Proper Clause was the Supreme Court’s judgment in 2012 in National Federation of Independent Business v Sebelius.133

Despite profound differences from one country to another, similar dynamics are typical of federal systems. To explain the existence of a functional need to make the division of powers flexible and to grant the central level some leverage in intervening in order to achieve the objectives posed by the national constitution, regardless of the deepest differences in history, constitutional nature and political system, the very similar dynamic that developed in the European Union can be helpful. Article 352 of the Treaty on the Functioning of the European Union (TFEU) contains a provision that is rather similar in scope to the US Necessary and Proper Clause. This article affirms: (1) If action by the Union should prove necessary, within the framework of the policies defined in the treaties, to attain one of the objectives set out in the treaties, and the treaties have not provided the necessary powers, the Council, acting unanimously on a proposal from the Commission and after obtaining the consent of the European Parliament,

132 133

Printz v United States 521 US 898 (1997). National Federation of Independent Business v Sebelius 567 US ___ (2012). See box 11.A.

152 Autonomy of Subnational Entities shall adopt the appropriate measures. Where the measures in question are adopted by the Council in accordance with a special legislative procedure, it shall also act unanimously on a proposal from the Commission and after obtaining the consent of the European Parliament … (3) Measures based on this Article shall not entail harmonisation of member states’ laws or regulations in cases where the treaties exclude such harmonisation. (4) This article cannot serve as a basis for attaining objectives pertaining to the common foreign and security policy.

Article 352 TFEU is constructed as an exceptional provision to be used only when the Treaty has not provided the powers that prove necessary to effectively pursue one of the objectives set out in the treaties. Analogy with the Necessary and Proper Clause of Article I, section 8, clause 18 of the US Constitution is striking, although the formulation is slightly ‘more prudent’. 5.L European Union 1956–96: Implied Powers and the Role of the Court of Justice The Court of Justice of the European Union initially made clear that this provision could only be activated under the strict condition of the absence of a legal base in the treaties.134 It then acknowledged that it could also be used when powers exist elsewhere in the treaties but are insufficient to effectively achieve the objective.135 However, after the Council began to use that provision much more frequently from the mid-1970s, based on the assumption that the objectives of the then-European Economic Community (EEC) were indeed quite broad and extensive, this article allowed for the creation of new competence fields of the Community, notably in the environmental sector, which was at that time not mentioned in the Treaty.136 It was only in 1996 that the Court put a limit on the expansive interpretation of the provision of then-Article 235 (now Article 352 TFEU) by ruling that this legal basis would not be sufficient to permit the accession of the Community to the European Convention on Human Rights. Such a modification ‘would be of constitutional significance and would therefore go beyond the scope’ of that provision.137 With regard to implied powers, the Court first held that the EEC had jurisdiction not only when expressly granted in the Treaty, but also that it had implied competence to carry out tasks expressly allocated to it.138 In the 1970s in particular, the question arose as to whether the powers conferred to the EEC regarding its internal competence might also be used to regulate external relations in the same fields, although there was no indication in the treaties. The Court gradually affirmed the existence of an implicit parallel between the powers explicitly conferred to the EEC by the treaties and their external projection (so-called in foro interno, in foro externo). Where the EEC has been given powers to achieve a certain objective in its internal dimension, it must logically have the corresponding

134

See especially Case 8/73 Hauptzollamt Bremerhaven v Massey Ferguson [1973] ECR 897. Case 45/86 Commission v Council [1987] ECR 1493 (General Tariff Preferences). 136 For a provocative line of argumentation, see JHH Weiler, ‘The Transformation of Europe’ (1991) 100 Yale Law Journal 2403, who affirmed that based on such an interpretation one could infer that the Implied Powers Clause could be used to introduce a common European defence policy, since it is evident that the common market would not be able to function if one of the Member States were militarily occupied. 137 Opinion 2/94 Accession by the Community to the ECHR [1996] I-1759. 138 Case 8/55 Fédération charbonnière de Belgique v ECSC High Authority [1955] ECR 201; Joined Cases 281, 283, 284, 285 and 287/85 Germany & Ors v Commission [1987] ECR 3203. 135

The Legislative Dimension 153

external powers that are ‘necessary’ (thus repeating the wording of the US Constitution and its Necessary and Proper Clause) for the achievement of the scope laid down in the Treaty. The leading case in this regard was adjudicated in 1971.139 At stake in the AETR case was the implementation of a common transport policy by the EEC, through the drafting of internal common rules (a regulation) excluding the concurrent powers on the part of the Member States throughout the sphere of transport. The Court of Justice, while aware that the system of internal EEC measures may not be separated from that of external relations, concluded that Member States may no longer enter into agreements with third countries in this field. The ruling thus acknowledged that powers that, at the outset, have not been conferred exclusively upon the EEC may become so progressively through the exercise of those powers.

Flexibilisation of the division of competences is in any event an issue in each and every federal system, although instruments to achieve this can vary based on the specific circumstances of each country. More recent, devolutionary federal systems, in particular, often use criteria such as the ‘national interest’, meaning that if the national government identifies a compelling national interest, this can justify usurpation of subnational competences. In Italy, national interest was used as a limit on regional legislation up until 2001, when this formulation was deleted from Article 117 of the Constitution but de facto re-established by the Constitutional Court. In Spain, national interest allows in certain cases the establishment of subnational entities in a top-down process, upon the central parliament’s own initiative and through an organic law (Article 144 of the Spanish Constitution). In this manner, it created the autonomous community of Madrid and the autonomous cities Ceuta and Melilla. Moreover, the national Parliament can adopt harmonising legislation in matters belonging to the realm of the autonomous communities if necessary for the general interest, as expressed by the majority of the members of each house of parliament (Article 150(3)).140 Some older federal systems also feature similar instruments. Article 249(1) of the Indian Constitution provides, for example, that if the Council of States, ie, the second chamber of the national Parliament, has declared by resolution supported by not less than two-thirds of the members present and voting that it is necessary or expedient in national interest that parliament should make laws with respect to any matter enumerated in the state list … it shall be lawful for parliament to make laws for the whole or any part of the territory of India.

A parallel centralising instrument for the executive branch is provided by Article 257(1) of the Indian Constitution which extends federal executive power to the giving of directions to a state. In reaction to misuse, the Supreme Court stated that a direction is only lawful, if a state action compromises the exercise of federal

139

Case 22/70 Commission v Council [1971] ECR 263 (AETR). particularly important example of an organic law adopted on the basis of Art 150(3) is the Organic Law on the Harmonisation of the Autonomy Process (LOAPA). See box 2.F. 140 A

154 Autonomy of Subnational Entities executive power.141 As mentioned above, Article 72(2) of the German Basic Law stipulates that the criterion for allowing the national Parliament to legislate on matters of concurrent competence is ‘necessity in the national interest’, which is given in order to ‘establish equal living conditions throughout the federal territory or the maintenance of legal or economic unity’. In some cases, national interest and similar formulas can be used rather as a limitation on the expansion of national legislation. This is the case when such interest is identified with that of one ethno-cultural group in a diverse society, such as in Bosnia and Herzegovina. There, a proposed decision of the national Parliament can be declared ‘destructive of a vital interest of the Bosniac, Croat, or Serb people’ by a majority of votes from the delegates of each of these communities in the House of Peoples, ie, one of the Parliament’s two chambers.142 In such a case, a joint commission is convened consisting of three delegates, each elected among Bosniac, Croat, and Serb delegates, in order to resolve the issue (Article 4(3)(e–f) of the Bosnian Constitution). If the Commission fails to resolve the issue within five days, the case is transferred to the Constitutional Court. In Belgium, in fields reserved to the national Government that affect the vital interest of one of the communities, a special ‘alarm bell’ procedure exists: if 75 per cent of the members of a linguistic group in the federal Parliament143 sign a motivated motion claiming that the provisions of a draft bill (or of a motion) may gravely damage relations between communities, the matter is referred to the federal Government (Article 54 of the Belgian Constitution). The Government includes an equal number of French- and Dutch-speaking ministers and by consensus makes a recommendation to the Parliament within 30 days. Examples of particular (centralising) ‘flexibilisation’ of the distribution of powers are three provisions of the Indian Constitution. These clauses empower the President of India to impose emergency rule in the event of a national emergency, ie ‘a threat to the security of India or any part of its territory caused by war, external aggression or armed rebellion’ (Article 352), a state emergency due to ‘the failure of the constitutional machinery in a state’ (Article 356) or a financial emergency (Article 360). At the founding of India’s federal system, these extraordinary unitary elements found broad support because of the strong impetus of the independence movement for national unity and the widespread belief in the Constituent Assembly that only a strong Union government could promote economic development and prevent secessionism. In fact, BR Ambedkar, one of the chief architects of the Constitution, emphasised the advantages of these three provisions because with them India ‘can be both unitary as well as federal according to the requirements of time and circumstances’.144 From the perspective of the states, the imposition of ‘President’s

141 State of Rajasthan v Union of India, 1977 AIR 1361. An early landmark ruling that generally established broad federal executive powers was Rai Sahib Ram Jawaya Kapur v State of Punjab, 1955 AIR 549. 142 The House of Peoples consists of 15 members: five Bosniacs, five Croats and five Serbs (Art 4(1) of the Bosnian Constitution). 143 The elected members of both chambers of the national Parliament, ie, the House of Representatives and the Senate, are divided into a Dutch and a French linguistic group (Art 43 of the Belgian Constitution). 144 BR Ambedkar, as cited in B Shiva Rao, The Framing of India’s Constitution, vol 5 (New Delhi, Indian Institute of Public Administration, 1968) 810.

The Legislative Dimension 155 rule’ under Article 356 is particularly critical. As the power to proclaim the abovementioned ‘failure of the constitutional machinery in a state’ falls to the President of India, this provision has provided ample opportunity for abuse. Above all in the 1970s and 1980s it became a widely used tool for the national Government to dismiss state governments controlled by opposition parties. This misuse of Article 356 as a pretext was only countered by a seminal judgment of the Supreme Court in 1994. 5.M India 1994: Judicial Review of ‘President’s Rule’? In 1988, the Janata Party won the majority in the state legislature of Karnataka and subsequently formed a government under the leadership of Chief Minister SR Bommai. Soon thereafter, this state’s Governor, a centrally appointed official, sent a report to the President of India claiming that defections from the Janata Party had led to a situation, in which Bommai would no longer command a majority in the state legislature. Even though a part of the alleged defectors officially declared their support of Karnataka’s government and the latter announced to prove the confidence through a vote in the state’s Legislative Assembly, the President of India moved on, upon advice of the Council of Ministers (Article 74), to proclaim President’s rule under Article 356. As required by the latter provision, this move was then approved by the national Parliament. Through such a proclamation, the President of India could ‘assume to himself all or any of the functions of the government of the state and all or any of the powers vested in or exercisable by the Governor or any body or authority in the state other than the legislature of the state’. Beyond that, the President of India was allowed under Article 356 to ‘declare that the powers of the legislature of the state shall be exercisable by or under the authority of parliament.’ Upon a challenge by the Government of Karnataka, the Supreme Court made two important findings.145 First, the judges held that a proclamation of President’s rule was justiciable, albeit to a limited extent. In concrete terms, three issues would be subject to judicial review: whether there was any reason behind the proclamation, whether these reasons were relevant and whether there was no mala fide exercise of Article 356. S econd, the Court established certain procedural principles for invoking this provision. If the majority support for the state Government in the state legislature is in doubt, this shall be tested through a vote of confidence. Moreover, the judges found that the approval of President’s rule after two months by the national Parliament (Article 356(3)) is meant to be a mechanism of control so that no irreversible measures shall be taken before it. Finally, they affirmed that the Supreme Court would provide, in case of an unconstitutional proclamation, effective remedy and thus have the power to re-establish a state legislature and to restore a state government to office.

In the end, each federal system chooses its own tools to make the flexibility requirement operational based on each country’s historical, political and other circumstances. What is common to all instruments, however, is the function. This function consists of providing explicit or even implicit tools to resolve conflicts of competence and to introduce some degree of flexibility in the constitutional division of powers that cannot be seen as perfect and immutable over time. As a rule—although with

145

SR Bommai v Union of India, 1994 AIR 1918.

156 Autonomy of Subnational Entities exceptions, as just mentioned—such tools produce a rather centripetal effect, both because of their being normally drafted as opening doors for national intervention and because of the interpretation of the courts, which often tend to favour a uniform regulation of policy areas. Needless to say, the political dynamics in each country also play a decisive role in orienting the interpretation by the relevant branches of government (including the courts), and such dynamics have more frequently been pushing for centralisation of powers rather than the opposite. 5.3 THE ADMINISTRATIVE DIMENSION: THE RELATIONSHIP BETWEEN LEGISLATIVE AND ADMINISTRATIVE POWERS

The division of administrative power can either follow that of legislative authority or be separated from it. When the administrative powers follow legislation, the same level of government is vested with the power to legislate in one particular matter and also retains the power to implement its own legislation. When legislation and administration are separate, laws can be implemented by a different level than that which adopted the legislation, and in practice legislation is more centralised and administration rather decentralised. The first model (administration follows legislation) is known as legislative or dual federalism, the latter (administration is largely disconnected from legislation) as administrative or executive federalism.146 Again, the dual approach is typical of the Anglo-Saxon and Latin American federal systems such as in the United States, Canada, Australia, Brazil and Mexico, while administrative federalism is the rule in continental Europe (Germany, Switzerland, Austria, Spain and Italy). As usual, however, lines cannot be drawn this clearly: Belgium largely divides legislative and administrative powers according to the dual model, whereas India and South Africa mostly follow the approach of administrative federalism. The underlying idea of legislative/dual federalism is that, in principle, there exist two separate, independent sets of political institutions, each of them as a rule vested with fully fledged state functions.147 This way, each level of government is entirely responsible for the areas assigned to it, adopts the relevant legislation and implements it, making the attribution of responsibility clearer and improving the accountability of each level of government: executives are responsible vis-a-vis the legislatures of the same level, and if additional administrative functions are delegated to them, this is done for specific reasons and within clear limits.148 Conversely, in systems of administrative federalism, subnational governments are in charge of executing most national laws, as this allows for a common framework and t ailor-made

146 The term ‘executive federalism’ may also refer, especially in the Canadian context, to the crucial role of intergovernmental relations that are dominated by the executives of the national and subnational governments at the expense of the respective legislatures. See ch 8.1.2. In this book, we thus prefer to use the term ‘administrative federalism’. On the historical origins of such federalism in Switzerland and Germany, see chs 3.3.2 and 3.3.3. 147 See TO Hueglin, ‘Canada’ in Le Roy and Saunders (eds), Legislative, Executive (2006) 122; and Watts (n 107) 329. 148 See, eg, in the United States the ruling of the Supreme Court in Lichter v United States 334 US 742 (1948), abandoning the original stricter approach of the judgment in Field v Clark 143 US 649 (1892).

The Administrative Dimension 157 implementation at the same time, taking into account local peculiarities. The consequence is overlap of responsibilities and, if such implementation is done within the framework of strict directives, even the annihilation of subnational autonomy, as subnational governments could become mere executive arms of the national one.149 As mentioned above, however, a rigid, watertight-compartment separation between the layers of government, with parallel and non-overlapping structures, is in practice impossible, even more so at present, a time marked by an increasing number of actors and growing complexity of policy areas, which require forms of cooperation among the levels of government. Of course, administrative federalism is more effective, the more efficient the instruments for cooperation. For instance, in Germany most legislation is adopted by the federal Parliament and is executed by the subnational governments, but the subnational governments play a significant role in the drafting of the federal laws they are in charge of implementing due to their direct representation in the Federal Council (Bundesrat).150 In other systems based on a similar approach, such as Austria or Spain, the institutional representation of subnational governments is weaker and it is thus channelled through less formalised and less powerful bodies of intergovernmental cooperation such as the Conference of the Länder Governors (Landeshauptleutekonferenz) in Austria and the sectoral conferences of (prime) ministers in Spain.151 The specific functioning of each system also depends on the structure of the administration. For historical and cultural reasons (including the late establishment of administrative law as an autonomous branch of law), Anglo-Saxon constitutions are quite sparing as to administration, with the consequence being that this is almost entirely regulated by ordinary legislation, which is as such more flexible and easier to adapt. One could think, for instance, of the role of independent and regulatory agencies, which are much less used in continental European systems and are, as a matter of fact, a further driver for centralisation. Finally, when legislation and administration are disconnected and thus a full responsibility of the executives vis-à-vis the legislatives cannot operate, it is natural that the constitution has to include mechanisms that guarantee the effective implementation of national legislation by subnational executives. Such instruments of oversight and enforcement are widely present in national constitutions in the form of enforcement and substitutive powers, such as those under Article 37 of the German Basic Law, Article 120 of the Italian Constitution, Article 52 of the Swiss Constitution and Article 142 of the Austrian Constitution. At the same time, the supremacy clause may be considered in dual systems, due to the precedence even of federal administrative acts, as a functional equivalent of such enforcement and substitutive powers. But even where the latter powers exist, they are typically used very rarely, their function being rather that of emergency breaks in cases of a non-functioning of cooperation mechanisms. 149 See J Kincaid, ‘Comparative Observations’ in J Kincaid and GA Tarr (eds), Constitutional Origins, Structure and Change in Federal Countries (Montreal, McGill–Queen’s University Press, 2005) 422; RL Watts, Comparing Federal Systems, 3rd edn (Montreal, McGill–Queen’s University Press, 2008) 36 and 41. 150 See chs 6.2.1 and 6.3.3. 151 See ch 6.2.2.

158 Autonomy of Subnational Entities

5.N Austria 1921–85: Enforcing a Land Governor’s Compliance with Federal Instructions? In Austria, a special power of enforcement was used several times in cases of ideological conflict between a conservative national and social democratic subnational government or vice versa. According to Article 142 of the Austrian Constitution, the Constitutional Court is authorised to rule on culpable breaches of the law in office by the highest federal and Länder authorities in the performance of their duties. In case of a condemnatory sentence, the Court can pronounce a forfeiture of office, even a temporary forfeiture of political rights or merely state that the law has been violated. Two very early examples of ideological intergovernmental conflicts before the Constitutional Court concerned cases where the social democratic Government of Vienna had refused to follow instructions from the conservative federal Government: one instruction had demanded the review of the performance permission granted regarding a play considered by some as obscene (Arthur Schnitzler’s ‘Reigen’) and the other the closing down of Vienna’s first crematorium.152 A more recent case revolved around the conservative Governor of the Land Salzburg who had permitted in 1984, contrary to an instruction by the federal Minister for Social Affairs, the opening of shops on a national holiday. The Court held that the Law on Rest Periods at Work would fall within the realm of indirect federal administration under Article 102 of the Austrian Constitution.153 Thus, the Land Governor was in charge of executing the law, but at the same time was obliged to follow the instructions of the federal minister responsible. The judges ruled that the reasons allowing to disregard the instruction are listed exhaustively in Article 20(1) of the Austrian Constitution, according to which a ‘subordinate officer can refuse compliance with an instruction if the instruction was given by an authority not competent in the matter or compliance would infringe the criminal code.’ This was obviously not the case. The Land Governor was therefore not allowed, as he did, to autonomously weigh up different legal interests and leave shops open with reference to a public interest. As to its sanction, the Court refrained from the above-mentioned stricter measures and merely declared the unlawfulness of the non-compliance with the federal instruction.

By way of example, the two most significant manifestations of the two models, ie, legislative federalism on the one hand and administrative federalism on the other, should be briefly explained and illustrated. These are the United States and Germany. In the United States, the federal Constitution says very little as to federal administration, besides the fact that executive power is vested with the President (Article II, section 1, clause 1). Administration has thus been developed based on ordinary statutes, and agencies (executive or independent) have played a significant role.154 In principle, administration follows legislation and thus the same level responsible for legislation also carries out administration. Forms of cooperation do exist, however, in several respects, from the informal coordination of governors to steering powers of the national Government by means of the grants system.155 As an 152

VfSlg 8/1921; VfSlg 206/1923. VfSlg 10510/1985. 154 On the important example of the Environmental Protection Agency (EPA), see ch 12.1 and especially box 12.A. 155 See Kincaid, ‘From Cooperative Federalism’ (1990) and E Ryan, Federalism and the Tug of War within (Oxford, Oxford University Press, 2012). 153

The Judicial Dimension 159 ultimate g uarantee, the Supremacy Clause of Article VI, clause 2 of the US Constitution expressly binds all judges to adhere to that principle. In Germany, Article 83 of the Basic Law states that, as a rule, the Länder ‘shall execute federal laws in their own right’.156 Exceptions are laid down in Articles 87 and 87a–f, according to which the federation can implement laws either through its own administration or through federal public law institutions. These exceptions include the diplomatic and consular service, federal finance administration, the postal service, armed forces, air traffic, railways, highways, border control, etc. A federal law can also establish new federal administrative authorities (Article 87(3) of the Basic Law). If the Länder administration executes national laws, this occurs either on the basis of a genuine right to do so (Article 84) or only on federal commission (Article 85) (Bundesauftragsverwaltung). In the former case, supervision by the national Government is restricted to possible violations of laws through implementation measures of the Länder, while in the latter case it also pertains to their appropriateness. Moreover, in case of commissioned execution, the Länder are subject to instructions from the national Government (Article 85(3)). In its case law, the F ederal Constitutional Court has interpreted this power to instruct quite extensively.157 Of course, there are also several nuances between these two models in the area of administration. In Austria, in contrast to Germany, the system is much more coercive than cooperative. In particular, the federal Constitution establishes that the Länder can be tasked with the implementation of federal laws (so-called indirect federal administration, Article 102 of the Austrian Constitution). In this case, unlike in Germany, the Austrian Länder do not execute federal laws as their own prerogative, but as delegates of the federal Government, ie, they are subject to instructions from the federal Government and operate on its behalf, with no substantial autonomy. Another interesting case in point that testifies to the many possible nuances that may exist in this field is Switzerland: while federal law binds the cantons in terms of how they implement it, the confederation must leave them as much leeway as possible and must shoulder the financial burden, as well as take into account the particularities of each canton (Article 46 of the Swiss Constitution). This way, room for differentiation in exercising administrative autonomy is granted. 5.4 THE JUDICIAL DIMENSION: THE COURT SYSTEMS

As noted by Saunders, ‘there tends to be greater interdependence of courts in federations than there is of any other institutions of government’.158 In fact, due to f actors such as, among others, uniformity in interpretation, centripetal tendencies, legal and political culture, judicial power in federal countries is strongly interconnected

156 See AB Gunlicks, The Länder and German Federalism (Manchester, Manchester University Press, 2003) 60ff. 157 See box 12.C. 158 Saunders, ‘Legislative, Executive’ (2006) 365, referring to the chapter by E Osieke, ‘The Federal Republic of Nigeria’ in Le Roy and Saunders (n 63).

160 Autonomy of Subnational Entities even when it is divided between two levels of government. To put it simply: among the functions analysed in this chapter (constitutional, legislative, administrative and judicial), the latter is, overall, the least autonomous, and in several federal systems the judiciary is not even divided between the national and the subnational level: this is the case, for instance, in Austria, Belgium, Russia,159 South Africa and Spain. Normally, the historical formation of the state plays a decisive role in determining the division of judicial powers: aggregative federations, where a judiciary existed in the subnational units prior to the establishment of the federal compact, have normally kept that judiciary, while devolutionary federal systems, departing from a unitary judiciary, as a rule did not transfer judicial powers to the subnational level, except for some organisational aspects. The historical pattern also explains why in the United Kingdom, for instance, Scotland kept its own judiciary and even a partly separate legal system: these were in place at the time of the Act of Union 1707, by which Scotland and England were unified in the Kingdom of Great Britain. Like for administrative competences, the division of judicial powers also follows in principle two main models: dual or integrated. Dual systems involve parallel and basically separate judiciaries for each level of government, exercising the jurisdiction assigned to the respective level independently from one another in general (eg United States and Switzerland). Integrated systems provide for a single judiciary and a single jurisdiction. The authority over the judicial system may either be assigned to the national government only (eg Austria and Belgium) or divided between the national and the subnational level (eg Germany). Canada has a dual administration but an integrated judiciary. In the US case, the federal Constitution only affirms that ‘[t]he judicial power of the United States shall be vested in one Supreme Court, and in such inferior courts as the Congress may from time to time ordain and establish’ (Article III, section 1 of the US Constitution). Congress has established 94 district tribunals (at least one in each state) and 13 courts of appeal. All federal judges are appointed by the US President upon the advice and consent of the Senate. Parallel to the federal judiciary, each state has its own judicial system. Most states have general courts and specialised courts, while a handful of them only have a single court of first instance. Most states have also established state courts of appeal, and all states have their own supreme court (with Texas and Oklahoma also providing for a separate supreme court for criminal issues). In some states, all appellate judges (and sometimes also those of the district courts) are appointed by the governor with the consent of the state senate.160 In 30 states, judges are elected by popular vote, either independently from political parties or being obliged to declare affiliation to one party when submitting their candidature. In other states, the governor appoints judges by choosing from a list submitted by the state bar association and the judges then have to stand periodically for re-election.

159 Constitutional Court, judgment of 1 February 1996, no 3-P, Vestnik Konstitutionnogo Suda RF 1996 no 1. 160 See GA Tarr, Without Fear or Favor: Judicial Independence and Judicial Accountability in the States (Stanford, Stanford University Press, 2012).

The Judicial Dimension 161 The division of jurisdiction between federal and subnational courts in the US is primarily based on the division of legislative power: state courts’ jurisdiction includes all issues not assigned to the exclusive jurisdiction of federal courts, and thus state courts are typically in charge of a much higher number of cases. However, the two judiciaries are not entirely separate. The rulings of state courts can be appealed to the US Supreme Court if the matter deals with an interpretation of the federal constitution or of a law or a treaty concluded by the United States. Furthermore, any issue based on habeas corpus (Article I, section 9 of the US Constitution) can be taken to the US Supreme Court even if lodged in a state court. 5.O United States 1842–1938: Dual Judiciary and the Question of Federal Common Law The US Judiciary Act of 1789 affirms that federal courts have to apply state laws when the relevant field is not regulated by federal law.161 The question arose as to whether this also includes judicial decisions, which, in a common law system, are among the sources of law. The US Supreme Court was first faced with the issue in the case Swift v Tyson adjudicated in 1842.162 The case originated in a circuit court of New York and involved a bill of exchange accepted in the state of New York. The question was whether the applicable law was that of a state (where only judge-made common law and no statutory law existed, as it is still the case in most contract and tort law) or whether, in the absence of a federal law as well, the applicable law had to be derived from federal common law. For the Supreme Court, the provision of the Judiciary Act does not bind federal courts to state commercial jurisprudence and, as common law is not merely local, the federal jurisprudence is free to derive its own common law. Thus, the federal courts, when deciding matters not specifically addressed by the state legislature, had the authority to develop a federal common law. The consequence, however, was not only a widespread use of ‘forum shopping’ by plaintiffs, seeking to sue in federal courts instead of a state court in order to have a different law applied, but also, and more importantly in federal terms, that federal (case) law could be developed in areas where Congress had no power to intervene. This doctrine was overruled almost a century later in 1938. In Erie Railroad Co v Tompkins,163 the US Supreme Court held that federal courts did not have the power to create general federal common law when hearing state law claims. Mr Tompkins, a citizen of Pennsylvania, was walking next to the Erie Railroad’s tracks in Pennsylvania, and was hit by a train. The train company was based in New York and Tomkins sued it in a federal district court of New York, thus avoiding Pennsylvania jurisdiction that was in this case less favourable to the plaintiff. The federal court applied not the state law but the federal common law based on the above-mentioned old doctrine of the case Swift v Tyson, and found in favour of Mr Tompkins, awarding him damages. The US Supreme Court, however, overruled this precedent, acknowledging that it led to two paradoxical consequences: first, it created an unconstitutional extension of the powers of the federal judiciary, as nothing in the Constitution permits Congress to empower federal courts to

161 Para 34 of the 1789 Judiciary Act: ‘the laws of the several states, except where the constitution, treaties or statutes of the United States shall otherwise recognise or provide, shall be regarded as rules of decision in trials at common law, in the courts of the United States, in cases where they apply’. 162 Swift v Tyson 41 US (16 Pet) 1 (1842). 163 Erie Railroad Co v Tompkins 304 US 64 (1938).

162 Autonomy of Subnational Entities

create their own common law for cases not involving issues of federal law; second, the practice of forum shopping that the old doctrine originated made state (judge-made) law increasingly diverge instead of converge. Thus, since Erie Railroad Co v Tompkins, there cannot be a federal common law, and cases not involving federal law are decided in state courts. The consequence was also that, since then, state common law has become much more uniform than it was during the time of the Swift doctrine.

Many countries to some extent follow the US pattern (eg Argentina, Australia and Nigeria), although with certain nuances. In the Australian case, for instance, a final court of appeal deals with matters in both federal and state jurisdictions. However, dual judicial systems also exist in other countries like Switzerland. This is yet another example of a dual system where the judicial powers of the national level are constantly expanding. The federal constitution affirms that cantons are responsible for judicial organisation and for administering civil and criminal jurisdiction (Articles 122(2) and 123(2) of the Swiss Constitution), as well as, implicitly, in all issues for which the cantons are responsible with regard to administration. Over the last decade, however, a number of subsequent reforms have changed the traditional picture of Swiss judicial federalism. Not only has judicial control over administrative acts been expanded (Article 29a of the Swiss Constitution), but the traditional cantonal power to adopt their own codes of civil and criminal procedure has been gradually transferred to the national level, which is now able to adopt unified codes (Articles 122(1) and 123(1) of the Swiss Constitution). As to the court system, the federal jurisdiction includes three high tribunals: the Federal Court, the Federal Criminal Court and the Federal Administrative Court. The respective judges are elected by the Federal Assembly, mirroring the political, as well as the linguistic and regional diversities of the country. The cantons are now responsible for first-instance rulings in civil, criminal and public law matters (Article 191b of the Swiss Constitution), and they can (and do) establish inter-cantonal judicial authorities to rule on these matters. All this shows a clear trend from a more decentralised to a more uniform judicial and court system. Canada and Germany are examples of integrated jurisdictions between the levels of government, as cooperation in the field is mandated by the federal constitution in both countries. In Canada, provinces are responsible for administering justice in their respective territory (Article 92(14) of the Constitution Act 1867), including the establishment of their own civil procedure. The national level has exclusive jurisdiction as to criminal law and criminal procedure (Article 91(27) of the Constitution) and can establish a federal court of appeal and other tribunals ‘for the better administration of the laws of Canada’ (Article 101 of the Constitution Act 1867). It did so by establishing the Supreme Court (1875),164 the Federal Court (1971), the Canadian Tax Court (1983) and the military courts. Provincial courts are regulated by provincial law,

164 Appeal to the JCPC was abolished for criminal matters and for constitutional matters in 1949, which finally made the Court really ‘supreme’.

The Judicial Dimension 163 including regarding the appointment of judges, and must apply both provincial and federal laws.165 The unifying element is represented by the Supreme Court, which rules as the last instance on controversies regarding both provincial and federal law. Continuing on the scale from more dual to more integrated judicial systems in federal countries, the most glaring example of an integrated judiciary is represented by Germany. According to Article 92 of the Basic Law, the judicial power is attributed to the Federal Constitutional Court, to the federal courts established by the Basic Law and to the courts of the Länder. Federal courts are courts of last instance (or, in the case of the Federal Constitutional Court, of exclusive jurisdiction), and subnational courts are in charge of first-instance and appellate jurisdiction. There are only six federal courts: besides the Federal Constitutional Court (Bundesverfassungsgericht), which retains a monopoly on constitutional interpretation at the federal level (Articles 93 and 94 of the Basic Law), there is the Federal Court of Justice for civil and criminal matters (Bundesgerichtshof); the Federal Administrative Court (Bundesverwaltungsgericht); the Federal Fiscal Court (Bundesfinanzhof); the Federal Labour Court (Bundesarbeitsgericht) and the Federal Social Court (Bundessozialgericht) (Article 95 of the Basic Law). In addition, there is also the Federal Patent Court (Bundespatentgericht), as well as military courts (Article 96 of the Basic Law). The Länder are responsible for the lower courts, including the recruitment and payment of judges. While important matters such as the civil, criminal and procedural codes are a matter of concurring legislative competence and could therefore in principle be regulated by each Land, the national level has exercised such power, and thus these codes are uniform for the whole country. This means that judicial litigation is regulated by the same law throughout the entire country, always starts at the subnational level (as first-instance courts are only in the Länder and of the Länder) and always ends up in a federal court for last-instance decision, as supreme courts are only federal. In general, integrated systems require, also and even more so in the judicial field, a high degree of good faith by the subnational units in order to function properly, and this might be a reason why some countries do not divide judicial power at all. Furthermore, the judicial field is an area in which the distinction between common law and civil law traditions (including in federalism) might play some role in designing the different judicial structures.166 What remains common to all systems, however, is the guarantee, although through different tools, of the uniform interpretation of the law. This guarantee is performed everywhere by the national judiciary, following a logic that reflects the same rationale of supremacy and homogeneity of federal law more generally.

165 See PW Hogg, ‘Federalism and the Jurisdiction of Canadian Courts’ (1981) 30 University of Brunswick Law Journal 9, 9ff. 166 See T Fleiner and C Saunders, ‘Constitutions Embedded in Different Legal Systems’ in M Tushnet et al (eds), Routledge Handbook of Constitutional Law (New York, Routledge, 2013).

6 Participation of Subnational Entities at the National Level 6.1 SCOPE, INSTITUTIONS AND PROCEDURES FOR PARTICIPATION

T

HE PARTICIPATION OF subnational units in decision-making at the national level is one of the most distinctive features of federal systems.1 The main reason for this is simple: participation compensates for the loss of sovereignty— ie of legislative and other powers—by subnational units. As a logical consequence, institutions and procedures that account for such participation are generally stronger and more visible in aggregative federal systems because the initial influence of constitutive units was stronger there than in devolutionary countries. Digging under the surface, however, comparative analysis shows that there is neither a unique concept of participation at the federal level nor a single form of it. Rather, each country provides for different mechanisms through which subnational units are represented in institutions or take part in decision-making at the national level, and these depend on several variables peculiar to each country’s history, political system or institutional tradition. Against the background of a common rationale and a broad variety of forms, the mechanisms of participation can be grouped according to certain criteria. The first and most obvious criterion distinguishes between institutional forms of participation (section 6.2), such as representation of subnational units at the national level through second chambers or bodies of intergovernmental cooperation—and procedural forms, ie the functions performed by such institutions (section 6.3). These include, inter alia, their participation in the amendment of the national constitution, territorial changes, as well as legislation and appointment procedures at the national level. One institutional avenue for subnational participation in decision-making at the national level are second chambers. Their variety2 is deeply influenced by the historical formation of each federal system,3 and consequently their federative role is more apparent in aggregative federations. Even within coming-together federations, 1 See, eg, WH Riker, Federalism: Origin, Operation, Significance (Boston, Little, Brown & Company, 1964) 2 and 11; M Burgess, Comparative Federalism: Theory and Practice (Abingdon, Routledge, 2006) 195f. 2 See J Luther et al (eds), A World of Second Chambers: Handbook for Constitutional Studies on Bicameralism (Milan, Giuffrè, 2006). 3 See D Shell, ‘The History’ in NDJ Baldwin and D Shell (eds), Second Chambers (London, Frank Cass & Co Ltd, 2011) 12.

Institutional Forms of Participation 165 however, the models differ from one another. A comparison reveals two structurally different approaches to subnational representation through second chambers, the senate model and the council (or ambassadorial) model.4 The US Constitutions offers the archetype of the senate model. Accordingly, the Senate is part of the federal legislative branch, and senators are not bound by instructions from constituent units, regardless of whether members are directly elected by the citizens (as in the United States, Switzerland, Australia, Argentina, Brazil, Mexico, Nigeria) or are elected by state legislatures (eg India) or appointed by the federal government (eg Canada). Senators can thus vote freely, and they ‘tend to vote along party lines rather than strictly for the regional interests they represent’.5 The German Federal Council (Bundesrat), instead, represents the council or ambassadorial model of institutional representation. Its structure differs from that of the US Senate in that it is historically based on the representation of the executives of the once independent political communities that were to become part of the German federation in 1867/1871.6 According to this model, members of the Federal Council are delegates of the subnational executives so that ‘it is primarily the views of … [subnational] governments that they represent and only indirectly those of the electorate’.7 This influences the system of voting: in the council model, each subnational delegation casts its vote en bloc according to the instructions of the respective government. As is usually the case, intermediate solutions also exist, although the senate model prevails overwhelmingly. The South African National Council of Provinces and the Russian Council of the Federation, for instance, have traits in common with both the senate and the council models.8 Institutional participation, however, is not limited to second chambers only. The declining role of second chambers in effectively representing subnational entities, the growing influence of executives as opposed to parliaments in decision-making processes at all levels and the necessity of permanent coordination, especially in highly technical and complex matters, among others, have led to the creation of institutions that bring together subnational and national executives nearly everywhere. The practical impact of such specialised bodies of intergovernmental cooperation with regard to subnational participation has increased tremendously. 6.2 INSTITUTIONAL FORMS OF PARTICIPATION

6.2.1 Second Chambers The main form of institutional representation of subnational units at the national level are second chambers. For a long time, they were the only representation of 4 See G Doria, ‘The Paradox of Federal Bicameralism’ (2006) 5 European Diversity and Autonomy Papers 1. 5 RL Watts, Comparing Federal Systems, 3rd edn (Montreal, McGill–Queen’s University Press, 2008) 141. 6 See ch 3.3.3. 7 Watts, Comparing Federal Systems (2008) 141. 8 See section 6.2.1.4 below.

166 Participation of Subnational Entities such kind and indeed regarded as an indispensable hallmark of federal design by numerous theorists.9 Only a few scholars have opposed this view and claimed that the doctrine of federalism requiring a second chamber ‘has no self-evidence, and it is assumed, but not proved’.10 But what is a second chamber, from a comparative perspective, in light of very diverse examples? In a material sense, second chambers ‘may also be bodies not incorporated in a parliament, but enjoying a political power linked to the functions of the parliament’.11 Unlike the ‘democratic’ chamber, elected by the people and aimed at representing the national population, second chambers are in federal systems supposed to represent subnational units, although in different ways. The traditional assumption used to be that subnational interests at the national level could be protected by participating in political decision-making only, and especially through representative institutions like second chambers.12 6.A United States 1985 and Austria 1952: How Important Is Institutional Representation? Even in relatively recent times, the US Supreme Court stated that to ensure the role of the states in the federal system the framers of the Constitution ‘gave the states a role in the selection both of the executive and the legislative branches of the federal government’, and, in particular, states ‘were given more direct influence in the Senate’.13 In subsequent judgments, the Supreme Court has not referred concerning the protection of state interests exclusively to such political safeguards of federalism.14 Yet, they still remain important. By the same token, the Austrian Constitutional Court, as early as 1952, defined Länder participation in the federal legislative process as a specific trait of the federal system.15 The case was brought by the regional government of Vorarlberg against a federal constitutional provision that suspended regional citizenship (Landesbürgerschaft) after World War II. Vorarlberg argued that this provision violated the federal principle of the Constitution. The Court then elaborated on this principle and held that the participation of the Länder in national legislation through the Federal Council, or possibly through equivalent channels,16 is one of its essential elements. The abandonment of the federal principle or one of its key elements would qualify as a total revision of the Constitution. Thus, it would only be possible through the procedure laid down in Article 44(3) of the Constitution that requires a popular referendum.

9 See, eg, SR Davis, The Federal Principle: A Journey through Time in Quest of a Meaning (Berkeley, University of California Press, 1978) 142. 10 W Bagehot, The English Constitution (Boston, Little Brown and Co, 1877) 162. 11 J Luther, ‘The Search for a Constitutional Geography and Historiography of Second Chambers’ in Luther et al (eds), A World (2006) 4. 12 Although this assumption has been challenged from the beginning by several scholars. See RE Hero, ‘The US Congress and American Federalism: Are ‘Subnational’ Governments Protected?’ (1989) 42 The Western Political Quarterly 93. 13 Garcia v San Antonio Transit Authority 469 US 528, 550f (1985). 14 See box 8.J. 15 VfSlg 2455/1952. See A Gamper, ‘“Arithmetische” und “geometrische” Gleichheit im Bundesstaat’ in K Weber and N Wimmer (eds), Vom Verfassungsstaat am Scheideweg: Festschrift für Peter Pernthaler (Wien, Springer, 2005) 143ff. 16 On this specific point, see H Schäffer, ‘Alternative Modelle zur Wahrnehmung von Länderinteressen an der Bundesgesetzgebung’, in P Bussjäger and J Weiss (eds), Die Zukunft der Mitwirkung der Länder an der Bundesgesetzgebung (Vienna, Braumuller, 2004) 54.

Institutional Forms of Participation 167 6.2.1.1 Origins and Constitutional Entrenchment The idea of a senate as a representative institution was developed in the Connecticut Compromise reached at the Philadelphia Convention in 1787,17 which reconciled the interests of the largest state (Virginia) with those of the smallest one (New Jersey). In this respect, ‘the essence of the compromise was that while the lower House would be elected on a popular basis, proportionate to the electorate as a whole, the Senate would provide an equal representation for the states irrespective to their size’.18 In addition, a second chamber would permit ‘breathing space in the legislative process to allow for second thought’.19 By contrast, the origins of the council (or ambassadorial) model are deeply rooted in the history of German federalism. For several centuries, long before the establishment of the German nation state, several (mostly small) entities existed, without a dominant centre such as, for instance, Paris or London. Relations among such entities naturally developed as interstate relations among governments and ministerial bureaucracies. In the nineteenth century, this type of relationship was formalised in the German Confederation (1815–66), which remained, however, ‘a union of sovereign principalities and the representative organ of the member states set the stage for princes, not for the people’.20 The drafters of Germany’s subsequent constitutions were deeply influenced by this form of participation in central decisionmaking. Under the Basic Law, in particular, the Federal Council’s nature is that of a representative organ comprising delegates appointed and recalled by regional executives and bound by the instructions of the respective subnational government. In the context of Germany’s administrative federalism, the Länder thus participate through the Federal Council in the drafting of the same federal laws that they are later tasked with implementing. The capacity of both senatorial and ambassadorial second chambers to grant effective representation of subnational interests depends on their institutional design, ie on their composition, on the way their members are appointed, on whether constituent units can impose mandatory instructions on their delegates. Moreover, the role of second chambers is determined by several other variables: the historical environment, the parliamentary or presidential system of government, etc.21 In particular, the alignment of senators/councillors along party lines reduces the influence of subnational entities on second chambers and turns them into national political bodies rather than territorial representations. While different in their origins and, consequently, in their very nature, both senatorial and ambassadorial representative institutions are typically entrenched in their

17 See M Hail and S Lange, ‘Federalism and Representation in the Theory of the Founding Fathers: A Comparative Study of US and Canadian Constitutional Thought’ (2010) 40 Publius 366. 18 Shell, ‘The History’ (2011) 12. 19 Shell (n 3). See also ch 3.3.1. 20 M Kotzur, ‘Federalism and Bicameralism: the German ‘Bundesrat’ (Federal Council) as an Atypical Model’ in Luther et al (n 2) 266. 21 See W Swenden, Federalism and Second Chambers: Regional Representation in Parliamentary Federations. The Australian Senate and German Bundesrat Compared (Brussels, Peter Lang, 2004) 29f.

168 Participation of Subnational Entities respective constitutions. As a consequence, their composition and functions can be altered only by an amendment enacted pursuant to the prescribed amendment procedure. For example, the US Constitution expressly establishes that the composition of the Senate can be altered only with the consent of the states (Article V of the US Constitution). The composition of Germany’s Federal Council can also be modified. However, Article 79(3) of the Basic Law expressly states that a constitutional amendment abolishing the participation of Länder in the federal legislative process is inadmissible.22 In Canada, the question of the constitutional limits on changes regarding the Senate resulted in two seminal decisions of the Supreme Court. 6.B Canada 1980–2014: Can the Senate Be Suppressed or Altered? In 1980, the Supreme Court had to deal with the question of whether Section 91(1) of the British North America Act 1867 allowed for federal legislation that would change or replace the Senate. The answer was negative.23 First, the federal Parliament alone cannot change the name of the second chamber. Second, it cannot alter the number and proportion of Senate members who represent provinces and territories, as the system of regional representation is one of the essential features of that body. Third, changes in the necessary qualifications to become senator cannot refer to the requirement that a senator must be a resident of the province for which he or she is appointed. Fourth, the question of whether senators can be directly elected was answered in the negative because ‘the preamble to the Act referred to “a constitution similar in principle to that of the United Kingdom”, where the Upper House is not elected. In creating the Senate in the manner provided in the Act, it is clear that the intention was to make the Senate a thoroughly independent body which could canvass dispassionately the measures of the House of Commons. This was accomplished by providing for the appointment of members of the Senate with tenure for life’ (p 77). In 2014, the Supreme Court was asked to answer another reference question concerning the Senate, ie whether it was possible to introduce changes relating to the appointment of senators or to abolish the second chamber.24 The Court highlighted the relationship between the Senate and the amendment procedure: ‘The review of constitutional amendments by an upper house is an essential component of the Part V amending procedures … except for the unilateral provincial procedure’ (para 110). Regarding the appointment of senators, the Court held that the changes at hand—senatorial term limits and the consultation of the population of each province and territory about their preferences for potential nominees—could not be introduced by the federal Parliament alone. They would require the lower constitutional amendment threshold of parliamentary approval, plus the consent of seven provinces representing more than 50 per cent of the population (Sections 38 and 42(1)(b) of the 1982 Constitution Act). According to the judges, the abolition of the Senate would even require the higher threshold of parliamentary approval, plus the consent of all 10 provinces (Section 41(e)).

22

See section 6.3.1.2 below. Reference Re Authority of the Parliament in Relation to the Upper House [1980] 1 SCR 54. 24 Reference Re Senate Reform [2014] 1 SCR 704. 23

Institutional Forms of Participation 169 6.2.1.2 The Senate Model: Variations, Appointment and Composition The Senate model originated in the United States and has from there spread in different variations to other federal systems. Whereas, in the US case, the representation of the people of the whole country is entrusted to the House of Representatives, that of the peoples of the states is vested in the Senate. According to Article I, section 3, clause 1 of the US Constitution, the Senate is composed of 100 members, ie ‘two senators from each State, chosen … for six years; and each senator shall have one vote’. Historically, drawing from the US example, the essential features of the senate model were: first, the participation of the constituent units in the composition of the Senate on an equal basis; second, the appointment of the senators by the state legislatures; third, senators are not bound by instructions on the part of the states. The US Constitution eventually abandoned part of this scheme when the popular election of senators was introduced with the Seventeenth Amendment, ratified in 1913. Accordingly, ‘[t]he Senate of the United States shall be composed of two senators from each state, elected by the people thereof, for six years’. This amendment removed the subnational legislatures’ ability to control the voting of their senators—even if this practice had already been abandoned since the middle of the nineteenth century.25 Direct election is a consequence of the prevalence of the principle of popular sovereignty over that of subnational representation in federal systems, following the increasingly important role of the democratic principle from the beginning of the twentieth century. Consequently, the direct election of senators has been introduced in several federal countries, including Mexico (1917), Brazil (1946), Argentina (1946 and 1994) and, in some cases, even prior to the ratification of the US Seventeenth Amendment, as in Australia (1900). Equal representation of subnational units irrespective of their population is another factor that privileges the federal principle over the democratic one. As we will see, the rules about the composition of senatorial second chambers in many cases follow this rule of equal representation, even if there are several exceptions (eg Switzerland, Canada, Austria). As to rules of appointment, there are some senatorial second chambers whose link with subnational entities is extremely weak if it exists at all. In Canada, the model for the Senate was the British House of Lords. During the Quebec Conference of 1864, the delegates of the colonies rejected the idea of adopting the US-style federal and bicameral model. This rejection was a direct result of the US Civil War (1861–65): it was doubtful that the US framework—which was then under major strain—would be suitable for the challenging task of establishing a federation in which Francophone and Anglophone Canadians could live together peacefully. Following instead the British constitutional tradition, the Senate would thus be composed of non-elected members.26 Section 24 of the Constitution Act 1867 stipulates that senators are appointed by the Governor General, ie in practice by the federal government.

25 See WH Riker, ‘The Senate and American Federalism’ (1995) 49 The American Political Science Review 452, 455f. 26 See DC Docherty, ‘The Canadian Senate: Chamber of Sober Thought or Loony Cousin Best Not Talked About’ (2002) 8 The Journal of Legislative Studies 27, 28.

170 Participation of Subnational Entities The existence of ethno-cultural cleavages also played a role in shaping the senates in Belgium and Spain. The Spanish Constitution defines the Senate as ‘the House of territorial representation’ (Article 69(1)). However, the territorial basis for the election of senators is primarily the provinces rather than the autonomous communities, so that the Senate fails to properly represent the latter.27 Article 69 provides that four senators are directly elected in each province and that regional legislatures only appoint one senator and one further senator for every one million inhabitants in their respective territories. Created in 1831 as a chamber fully equal to the Chamber of Representatives, the Belgian Senate has undergone several reforms, most notably in 1993 and 2014. The constitutional amendments passed in 2014 cancelled the previous direct election of 40 senators and introduced the complex system of appointment that is in force today. According to the new Article 67 of the Constitution, 29 senators are appointed by the Flemish Parliament, 10 by the Parliament of the French community, eight by the Walloon Assembly, two by the French-speaking institutions of the Brussels-Capital Region and one by the Parliament of the German Community. Finally, 10 are co-opted by the previously appointed senators: six by the Dutch-speaking and four by the French-speaking senators.28 Although there is an overall trend, as mentioned above, towards the direct election of senatorial chambers, indirect election is not only a recent feature of Belgium. In Austria, indirect election has characterised the Federal Council (Bundesrat) for almost a century. According to Article 35(1) of the Austrian Constitution, the members of the Federal Council are elected by the legislatures of the Länder for the duration of their respective legislative periods in accordance with the principle of proportional representation. At least one seat must fall to the party with the secondlargest number of seats in a Land legislature or, should several parties have the same number of seats, the second-highest number of votes during the last election. That the votes are only relevant, if two parties have the same number of seats, was recently confirmed by the Constitutional Court.29 Importantly, the party with the right to nominate acts, in practice, independently and without any prior debate in a plenary session of the Land Parliament. Article 35(2) stipulates that for the passive voting right it is only mandatory to be eligible for this Land’s legislature, but not to be a member of it. A double mandate is thus allowed but not encouraged. In several federal systems, the selection of the members of second chambers combines different criteria, blending election and federal appointment. Section 45(1) of the Malaysian Constitution prescribes that the Senate (Dewan Rakyat) ‘shall consist of elected and appointed members’. Of the 70 senators, 26 are elected by subnational legislatures, each of which elect two members. With regard to the 44 appointed senators, Section 45(1) states that ‘two members for the Federal Territory of Kuala Lumpur and one member for the Federal Territory of Labuan and one

27 See JM Castellá Andreu, ‘The Spanish Senate after 28 Years of Constitutional Experience: Proposals for Reform’ in Luther et al (eds), A World (2006). 28 See A Feyt, ‘Sénat’ in M Uyttendaele and M Verdussen (eds), Dictionnaire de la Sixième Réforme de l’Etat (Bruxelles, Larcier, 2015) 780ff. 29 VfSlg 19782/2013.

Institutional Forms of Participation 171 member for the Federal Territory of Putrajaya shall be appointed by the Yang di-Pertuan Agong’, ie the monarch and head of state of Malaysia, on the advice of the federal Government. The monarch appoints the remaining 40 members from among individuals who, in his opinion, have rendered distinguished public service or have achieved distinction in their professions or who are representative of ethnic or indigenous minorities.30 India also blends indirect election with federal appointment, even if it does so to a much lesser degree. Section 80(4) of the Constitution establishes that the representatives of each state in the Council of States (Rajya Sabha) ‘shall be elected by the elected members of the Legislative Assembly of the state in accordance with the system of proportional representation by means of the single transferable vote’. This rule applies to currently 204 members of the Council of States, while 12 additional members are nominated by the President of India (Article 80(3)). As always in comparative federal studies, exceptions and peculiarities exist in every possible classification attempt. So, for example, Saint Kitts and Nevis manages to have a unicameral parliament, ie the National Assembly, with 11 members being elected by the population (called ‘representatives’) and four being appointed by the Governor General (called ‘senators’). Regarding the composition of the second chamber, the senate model mostly follows the principle of equal representation of subnational units. This idea is based on the assumption that senators represent the units rather than their people; thus, the equality of the units regardless of the size of their respective population. Such a composition is typical of aggregative federal systems, where equal representation succeeded in accommodating the different positions and interests of bigger and smaller units at the time of the founding of the federation. It also characterises those federal systems that were deeply influenced by the United States of America (eg Article 54 of the Argentinian Constitution, Article 56(1) of the Mexican Constitution and Article 46(1) of the Brazilian Constitution). As for Australia, Section 7(3) of the Constitution likewise establishes that each state has the same number of senators, even though the second chamber today also includes representatives from the territories. The Parliament is, however, allowed to alter the number of senators, ‘but so that equal representation of the several original states shall be maintained and that no original state shall have less than six senators’. At present, the Australian Senate comprises 76 members: 12 from each state and four from the federal territories; two from the Australian Capital Territory and two from the Northern Territory.31 The fact that the Senate no longer reflects its initial conception as a states’ chamber is not least due to jurisprudence of the High Court. When federal legislation provided for the election of senators from the two mainland territories, the Court upheld this law with reference to the federal power to allow the representation of territories in either chamber (Section 122). This provision was

30 See A Harding, The Constitution of Malaysia: A Contextual Analysis (Oxford, Hart Publishing, 2013) 108ff. 31 See T Blackshield and G Williams, Australian Constitutional Law and Theory: Commentary and Materials, 5th edn (Annandale, The Federation Press, 2010) 365.

172 Participation of Subnational Entities given preference over Section 7(1) which characterises the second chamber as being ‘composed of senators for each state’.32 The Swiss Council of States (Ständerat) slightly deviates from the principle of equal representation insofar, as six out of the 26 cantons are, according to Article 150(2) of the Constitution, represented by just one, instead of two, members. These cantons were traditionally called ‘half-cantons’, a term that was only abandoned in the constitutional reform of 1999. While otherwise enjoying the very same legal status as all other cantons, their ‘half’ representation in the Council of States reflects the historic association of the half-cantons to each other and the historic reasons for their partition.33 The Canadian Senate is formed according to the so-called principle of regional equality instead of that of equal representation of the provinces. The British North America Act 1867 aggregated the provinces on a regional basis and granted each senatorial region but not each single province an equal number of representatives, even though both Ontario and Quebec are, at the same time, a province and a senatorial region.34 This composition is entrenched in Section 22 of the Constitution Act 1867. Austria deviates from the principle of equal representation as well, as it rather relies on weighted representation. The aim is to reconcile the federal and the democratic principles by taking into account the demographic size of the subnational units while at the same time over-representing smaller units as compared to larger ones. Article 34(1–2) of the Austrian Constitution states that the Länder are basically represented in the Federal Council in proportion to their respective populations, but that none of them shall elect more than 12 or less than three members. Changes in the representation of the Länder within the Federal Council require the amending procedure as set out in Article 35(4) of the Constitution, ie beyond the absolute majority of the votes cast also the approval by the majority of representatives of at least four Länder. 6.2.1.3 The Council (or Ambassadorial) Model: A Germany Legacy The basic idea of the council (or ambassadorial) model of second chambers has never changed: the Länder participate in decision-making at the national level primarily through their governments. This way, subnational governments participate in the drafting of federal laws that they are, in the context of administrative federalism, responsible for implementing. While the Federal Council participates to a significant extent in the legislative process, its members are appointed by and bound by the

32 Western Australia v Commonwealth (1975) 134 CLR 201 (First Territories Representation case); Queensland v Commonwealth (1977) 139 CLR 585 (Second Territories Representation case). 33 The six cantons are Obwalden and Nidwalden, Basel-Stadt and Basel-Landschaft, Appenzell Ausserrhoden and Appenzell Innerrhoden. 34 The four senatorial regions are Ontario, Quebec, the Maritime Provinces (Nova Scotia, New Brunswick and Prince Edward Island) and the Western Provinces (Manitoba, British Columbia, Saskatchewan, and Alberta).

Institutional Forms of Participation 173 instructions of their respective governments (Article 51(1) of the Basic Law). Thus, they are, in contrast to representatives in the first chamber, not elected members of Parliament with a free mandate. This approach is called the ‘attenuated Federal Council system’ (abgeschwächte Bundesratslösung). The attenuated character lies in the fact that the Federal Council is not a classical second chamber of the legislative branch and does not have absolute veto power through the procedure of approval over all federal legislation (Article 78). These specific issues were at stake in a seminal case before the Federal Constitutional Court. 6.C Germany 1974: Is the Federal Council a Second Chamber? The governments of Rhineland-Palatinate and Bavaria brought a lawsuit against a federal law of 1973 that had modified the previous law on pension insurance. The reason was procedural in nature: the two Länder claimed that the law should be declared null and void, as it had been adopted without the approval of the Federal Council, whereas the previous law was adopted following the approval procedure. For the Court, however, the mere fact that a law was adopted following this procedure is not per se sufficient to require that the approval procedure also be followed to amend the law.35 Rather, the Court stated, the applicability of the approval procedure depends on the substance of the parts of the law that are being amended. These do not require the consent of the Federal Council, if powers of the Länder are not affected. Importantly, the Court also made clear that ‘[t]he Federal Council is not the second chamber of a unitary legislative body, in which the two chambers are on equal footing’ (p 380). Rather, it is a specific body for participation of the Länder in a number of functions at the federal level, including legislation, where this is mandated by the Basic Law. This is self-evident from several provisions of the Basic Law, stating, for example, that ‘laws are adopted by the Parliament (Bundestag)’ (Article 77(1)) and that ‘the Länder shall participate through the Federal Council in the legislation and administration of the Federation’ (Article 50). Such participation might be very extensive, but it nevertheless remains the exception.

As to the composition of the Federal Council, there are changes any time Länder governments change and the presidency rotates every year.36 As to the weight of the Länder in the Federal Council, there is also an element of attenuation, as federal and democratic criteria are combined: not all Länder have the same number of members, but smaller Länder are over-represented as compared to larger ones. According to Article 51(2) of the Basic Law, ‘each Land has at least three votes, Länder with more than two million inhabitants have four, Länder with more than six million have five, and Länder with more than seven million have six’. The system of voting is consistent with this approach: Article 51(3) states that ‘each Land can appoint as many representatives as it has votes’ and that ‘the votes of a Land can be cast only as a unit and only by members present or their alternates’. The underlying idea is

35

37 BVerfGE 363 (Bundesrat case). RL Borthwick, ‘Methods of Composition of Second Chambers’ in Baldwin and Shell (eds), Second Chambers (2011) 22. 36 See

174 Participation of Subnational Entities that Federal Council members shall represent the interests of their respective Land government with one voice, without regard to the partisan composition of this government. But as coalition governments are the norm, a lack of agreement concerning the block vote entails that ‘the Federal Council members from that Land will most likely abstain; this has the effect of a negative vote, since only positive votes are counted.’37 Therefore, coalition partners usually conclude detailed agreements on how they would like to proceed regarding the vote in the Federal Council. No such agreement was in place in 2002 however, when the parties forming the government of the Land Brandenburg were in conflict over how to cast its vote. 6.D Germany 2002: The Block Vote in the Federal Council In 2002, the federal Government, consisting of the Social Democratic Party of Germany (SPD) and the Green Party, enacted a controversial immigration bill. At that time, however, recent Land elections had strengthened the opposition parties in the Federal Council to such an extent that for the adoption of the bill, the federal Government had to rely on Brandenburg casting its block of four votes in favour of it. Yet, a coalition of the SPD and the conservative Christian Democratic Union of Germany (CDU), ie a party in opposition to the federal Government, formed Brandenburg’s Government. After an SPD member of the Brandenburg delegation had cast all four votes in favour of the bill, a CDU member explicitly contradicted this and cast his vote as ‘no’. That Brandenburg’s block of four votes was nonetheless accepted as a ‘yes’ was then challenged by six Länder before the Federal Constitutional Court. The Court ruled that the Federal Council had not validly given its consent which made the Immigration Act incompatible with Article 78 of the Basic Law.38 The judges held that, on the one hand, the Basic Law expects a block vote, normally cast by a delegate determined by the delegation itself or by the Land government before the session. On the other hand, however, it regards each Land as being represented by individual delegates with the right to an individual vote. As Brandenburg split its votes and thus failed to cast them as a block according to Article 51(3), the Court considered simply the vote of this Land as invalid and ineffective.

Although the council model is thus specific in several respects, some of the differences between it and the senate model tend to fade, when a functional rather than formal approach is taken. While the German Federal Council is, as the Federal Constitutional Court ruled in its above-mentioned judgment of 1974, not a (second) chamber in strictly legal terms, it participates in the legislative process in a manner that does not substantially differ from how senatorial chambers take part in legislative processes. It may therefore be regarded, in a functional sense, as an (ambassadorial) second chamber. Moreover, in both cases decisions are in practice often made based on party politics rather than territorial considerations. In fact, the

37 See AB Gunlicks, The Länder and German Federalism (Manchester, Manchester University Press, 2003) 344. 38 106 BVerfGE 310 (Immigration Act).

Institutional Forms of Participation 175 Federal Council has often been used as a tool to oppose political initiatives on the part of the federal Government, when its majority did not match the one supporting the federal Government in the Parliament (Bundestag).39 Despite certain limits, such as the possible political (mis)use of the Federal Council for purposes other than territorial representation, this body has proven to be much more effective overall than any other senatorial second chamber in terms of representing subnational entities as such in national decision-making. And, in fact, all attempts to change its fundamental structure and to transform it into a ‘democratic’ second chamber have failed. In light of the undeniable effectiveness, it is even more striking that the council model has essentially remained limited to Germany. Only some traits of it are to be found in a few hybrid second chambers.40 Although the German Federal Council is widely cited in political and academic debates as the only effective way to guarantee the representation and participation of subnational entities, the model is not followed. Why is that? Part of the explanation lies in the evolution of constitutionalism which has gradually but unequivocally favoured the democratic over the federal principle. Since the beginning of the twentieth century, as mentioned above, the popular election of the members of senatorial second chambers has been widely introduced, for precisely the same reason. Furthermore, as frequently mentioned in this book,41 over the last century, federal systems have been formed rather following the devolutionary pattern than by the coming-together of sovereign states. In most such cases, second chambers already existed in the unitary state preceding the devolutionary process and were based on political/democratic rather than territorial criteria (eg Belgium, Spain and Italy). In such contexts, the introduction of a mechanism of participation that is deeply rooted in a history of intergovernmental, international (ambassadorial) relations among pre-modern entities would appear quite eccentric. Further evidence of this comes from the only institution that is structurally comparable to the German Federal Council, ie the Council of the European Union (Article 16 of the Treaty on European Union (TEU), Articles 237–243 of the Treaty on the Functioning of the European Union (TFEU)). The Council is in fact the body in which the Member States’ governments meet and make decisions, based on a qualified majority procedure or, in some cases, unanimously. It has a fundamental role in EU legislation, but is by no means part of the Parliament. The affinity to the German model very much depends on some historical similarity of progressive intergovernmental integration among sovereign countries sharing common interests. In the absence of such conditions, a mere transplant of the Federal Council model would appear rather difficult. It is thus no surprise that this model has not been used in other systems despite its prestige and effectiveness.

39 See A Benz, ‘From Unitary to Asymmetric Federalism in Germany: Taking Stock after 50 Years’ (1999) 29 Publius 55. 40 See section 6.2.1.4 below. 41 See ch 2.2.1.

176 Participation of Subnational Entities 6.2.1.4 Hybrid Second Chambers Based on their composition, the casting of votes and powers, some federal second chambers resemble certain elements of the ambassadorial model. This is the case, in particular, of the Russian and South African second chambers. These are variants of the senate model, as both chambers are part of the federal legislative branch and their members are not bound by instructions from subnational executives. They show traits of the council model, however, especially when it comes to their composition and voting system. The Constitution of South Africa opts for a mixed composition in order to include both executives and legislatures from the provinces. The National Council of Provinces (NCOP) comprises ex officio members (eg the presidents of the provinces) as well as members appointed by the provincial executives and legislatures (Sections 60(2) and 61(2) of the Constitution). According to the principle of equal representation, each delegation in the National Council of Provinces consists of 10 members (Section 60(1)). With regard to voting, Section 65(1)(a) of the Constitution states that, ‘[e]xcept where the Constitution provides otherwise … each province has one vote, which is cast on behalf of the province by the head of its delegation’. This is particularly true regarding bills amending the national Constitution (Section 74) which require the supporting vote of the province and not of its delegates. When it comes to ordinary bills that do not affect the provinces (Section 75), each delegate in the National Council of Provinces has one vote. Whereas the block vote is an essential feature of Germany’s Federal Council, it therefore only partially applies in South Africa’s second chamber. Due to this specific hybrid design, it seems fair to state that the NCOP is not a classical senatorial second chamber but rather the ‘institutional embodiment of the principles of co-operative government and intergovernmental relations’.42 In Russia, Article 95(2) of the Constitution establishes that the Council of the Federation includes two representatives from each subnational entity: one from the legislature and one from the executive. Each subnational legislature is represented by its speaker. As for the representative of the subnational executives, a constitutional law that entered into force in 2000 foresees that the appointment of this representative must be approved by the legislature of the subnational entity.43 Thus, the composition of the Russian second chamber does neither fully resemble the senate model nor the council model. Hybrid designs also characterise the second chambers of Ethiopia as well as of Bosnia and Herzegovina.44 In the latter case, 15 delegates make up the second chamber of the federal legislative branch (the House of Peoples): the legislative assembly of the Federation of Bosnia and Herzegovina elects 10 delegates (five Croats and five

42 H Klug, The Constitution of South Africa: A Contextual Analysis (Oxford, Hart Publishing, 2010) 155. On these principles, enshrined in Sec 41 of the South African Constitution, see also ch 8.1.1. 43 See J Henderson, The Constitution of the Russian Federation: A Contextual Analysis (Oxford, Hart Publishing, 2011) 157, 166f. 44 See A Morawiec Mansfield, ‘Ethnic but Equal: The Quest for a New Democratic Order in Bosnia and Herzegovina’ (2003) 103 Columbia Law Review 2052, 2052f;

Institutional Forms of Participation 177 Bosniacs); the five Serbian delegates are nominated by the National Assembly of the Republika Srpska (Article IV(1) of the Constitution). In Ethiopia, the state legislatures themselves may appoint the members of the second chamber, ie the House of the Federation, or they may opt for direct popular elections of these representatives (Article 61(3) of the Constitution). In the House of the Federation, each nation, nationality and people, as defined in Article 39(5) of the Constitution, is represented by at least one member and by one additional representative for each one million of its population (Article 61(1–2)). 6.2.2 Intergovernmental Bodies While second chambers have historically been the only institution entrusted with the task of ensuring subnational participation at the national level, today bodies of intergovernmental cooperation also fulfil this function in many federal systems. These bodies do not substitute second chambers but they supplement them to an increasing extent. In part, this trend results from the real or perceived inadequacy of many of the second chambers examined in the preceding section to effectively represent subnational entities. Another major reason behind the growing relevance of intergovernmental bodies is the increasing entanglement of many competences and complexity of political decisions, including more and more technical aspects, which favours executive co-decision-making rather than parliamentary debate. In this new context, such bodies—and intergovernmental relations more generally—are vital, especially for the management of conflicts between different government levels. They are thus primarily explored later in this book.45 However, intergovernmental bodies must be partly anticipated here for systematic reasons because of their role as channels for subnational participation at the national level. In particular, multilateral institutions that include representatives from all governments play an essential role in facilitating cooperative participation where second chambers are ineffective. Even though such institutions are also relevant as supplements, for example, in Germany where a lack of effectiveness of the Federal Council is not an issue, they are in relative terms clearly more significant in federal systems with a weak form of bicameralism (eg Austria, Italy and Canada). In Austria, the real channel for subnational participation in national decision-making is clearly the powerful Conference of the Länder Governors (Landeshauptleutekonferenz) rather than the second chamber. This body can be traced back to informal gatherings soon after World War I and was eventually established as a regular semi-annual conference in 1970. Its influence in practice contrasts with its lack of constitutional entrenchment, while it is just the other way around for the Federal Council. A comparatively more formalised role is played by a multilateral body in Italy, ie the Standing Conference for the Relations between the State and Regions (Conferenza permanente per i rapporti tra lo Stato e le Regioni). The institution includes the Italian Prime Minister and the presidents of the regional executives and must be consulted

45

See ch 8.1.

178 Participation of Subnational Entities regarding measures that could affect regional competences. In Canada, a long tradition of multilateral forums such as the First Ministers’ Conferences bodies, even though somewhat weakened since the early 2000s,46 have compensated for the weakness of the Senate in representing provincial interests over most of the country’s history. It is, therefore, not a coincidence that the role of intergovernmental bodies as functional equivalents of, and compensation for, a weak second chamber has first been studied in Canada. There, the relationship between these two different institutional forms of participation has been captured by the differentiation between ‘intrastate federalism’, involving the representation of subnational interests within national institutions, and ‘interstate federalism’, with these interests being promoted through intergovernmental relations.47 But the decreasing importance of ‘intrastate federalism’ equally characterises other countries with largely ineffective second chambers. As the latter often fail to fulfil the function of ensuring subnational participation that they are, in (federal) theory, supposed to perform,48 the search for more effective functionally equivalent mechanisms is likely to continue. 6.3 PROCEDURAL FORMS OF PARTICIPATION

The areas of decision-making at the national level in which subnational units play a role are numerous and so are the procedures according to which such participation is granted. The main areas—which the following pages will focus on—are constitutional amendment, territorial changes, legislative (and administrative) functions, as well as appointments. Others include specific cases such as states of emergency or impeachment powers. These functions are performed primarily by second chambers, whatever their nature, composition and voting system. In some cases, however, subnational participation is not channelled through a federal second chamber, either due to its absence or because, for certain procedures, different participatory instruments have been chosen by the respective constitution. 6.3.1 Constitutional Amendment The most significant participatory function of subnational units at the federal level is their participation in the constitutional amendment procedure. This power adds up to another fundamental guarantee, ie the rigidity of the national constitution as an essential requirement of federal systems.49 Through participation in the amending power, the very existence of subnational entities, their legislative powers, etc,

46

See ch 8.1.2. D Smiley and RL Watts, Intrastate Federalism in Canada (Toronto, Toronto University Press, 1985). 48 See Doria, ‘The Paradox’ (2006). 49 See AV Dicey, Introduction to the Study of the Law of the Constitution (London, Macmillan, 1885) 142ff. 47

Procedural Forms of Participation 179 in short, all constitutional elements of the federal system, are protected against unilateral modifications. Questions arise, however, when it comes to the different amendment procedures, as there is no unique formula that is common to all federal systems. Furthermore, participation in the amending power takes place according to different criteria: ‘The concern, therefore, is not only with rigidity, but also with the representation of the peoples and/or governing institutions of the constituent states in the amendment process’.50 In this respect, some constitutions provide for direct participation of subnational entities in the amending power, either through their legislatures or their electorate. In other federal systems, subnational participation occurs indirectly through the second chamber. 6.3.1.1 Direct Participation by the Legislatures or Electorates of Subnational Entities The archetype of direct participation in the amending power can be traced to Article V of the US Constitution:51 The Congress, whenever two thirds of both Houses shall deem it necessary, shall propose amendments to this constitution, or, on the application of the legislatures of two thirds of the several states, shall call a convention for proposing amendments, which, in either case, shall be valid to all intents and purposes, as part of this constitution, when ratified by the legislatures of three fourths of the several states, or by conventions in three fourths thereof, as the one or the other mode of ratification may be proposed by the Congress.

State legislatures thus participate in different stages of the amendment procedure. Article V entitles them to both propose amendments—by taking the initiative—and to approve them through ratification. So far, however, only Congress has taken the initiative, as a national convention has never been convened. As for ratification, Congress has typically opted for the process of approval by three fourths of the state legislatures and has convened state conventions only once, ie for the approval of the Twenty-First Amendment, which repealed prohibition in 1933. The Constitution of the United States contains different amendment procedures— and different degrees of rigidity. Beyond the above-quoted general amendment procedure, there is a special procedure, for instance, regarding territorial changes which only requires a joint resolution of both Houses of Congress and the consent of the states affected. Another particular case are amendments concerning the representation of the states in the Senate, as ‘no state, without its consent, shall be deprived of its equal suffrage in the Senate’ (Article V). The practice of ratification by the state legislatures under Article V of the US Constitution has deeply influenced the way in which other national constitutions regulate the participation of subnational units in amendment processes. This holds true, even if some federal constitutions require only a simple majority of state legislatures.

50 N Aroney, ‘Formation, Representation and Amendment in Federal Constitutions’ (2006) 54 The American Journal of Comparative Law 277, 326. 51 See W Dellinger, ‘The Legitimacy of Constitutional Change: Rethinking the Amendment Process’ (1983) 97 Harvard Law Review 386.

180 Participation of Subnational Entities According to Article 135 of the Mexican Constitution, for example, the proposed amendments must be approved by the Congress of the Union and be ratified by a majority of the legislatures of the states. Other constitutions follow the US model more closely and require qualified majorities. According to Article 9 of the Nigerian Constitution, a bill amending the Constitution that has passed both chambers of the National Assembly needs to be approved by no less than two-thirds of all the states. The US archetype to some extent also influenced the Constitution of the Russian Federation, which includes different amendment procedures: first, for territorial changes (Article 137); second, for altering the provisions regarding basic principles of the Constitution, the rights and freedoms and the amending formulas (Article 135); third, a general amendment procedure (Article 136). According to the general formula, the proposed amendments have to be adopted by the federal Parliament through a resolution of the Duma approved by a vote of two-thirds of its members and through a resolution of the Council of the Federation by a vote of three-fourths of its members. The amendment will come into force provided that it is ratified by no less than two-thirds of the subnational legislatures. Canada also has different amendment procedures with graduated subnational participation. This variety of procedures is linked to the historical evolution of constitutional law, as Canada did not have its own domestic amending formula until the so-called patriation of the Constitution in 1982. It was precisely on that occasion that the issue of the participation of the provinces raised fundamental issues of constitutional theory and practice. 6.E Canada 1981–82: Provincial Consent to the Patriation of the Constitution? Before 1982, constitutional amendments had to be formally adopted by the UK Parliament. In this respect, there had been no consistent practice by the federal Government of obtaining the consent of the provinces before requesting an amendment. When Prime Minister Pierre Trudeau proposed what would eventually become the Constitution Act 1982, including the Canadian Charter of Rights and Freedoms and new amendment procedures,52 he made it clear that, if provincial consent could not be obtained, he would proceed unilaterally to request the enactment by the UK Parliament. In reaction, Newfoundland, Quebec and Manitoba requested a reference from their respective provincial courts of appeal and then from the Supreme Court. They asked whether there was a requirement, either by constitutional law or by constitutional convention, to obtain the consent of provincial governments on constitutional amendments. The Supreme Court ruled that as a matter of law the consent of the provinces was not necessary and the federal Government was free to request amendments from the UK Parliament.53 However, the judges went on to recognise the existence of an unwritten dimension of Canada’s Constitution and held that a constitutional convention would require ‘a substantial degree of provincial consent’ (p 905). After the Court’s reference, an agreement was reached between Prime Minister Trudeau and all the provincial premiers

52

See ch 10.3. Reference Re Resolution to Amend the Constitution [1981] 1 SCR 753. See AM Dodek, ‘Courting Constitutional Danger: Constitutional Conventions and the Legacy of the Patriation Reference’ (2011) 54 Supreme Court Law Review 117. 53

Procedural Forms of Participation 181

except for Quebec. This raised a further question as to whether the conventional requirement of a substantial degree of provincial consent was thus fulfilled or whether Quebec had to be included. The province indeed raised the question, and the Supreme Court in another reference held that this was not the case.54

When the Constitution was eventually repatriated in 1982, the UK Parliament granted Canada plenary power of amendment. While Quebec was not granted a veto prerogative in constitutional amendments, the peculiar status of that province enjoys a high level of protection in the new amendment procedures. Indeed, the federal Government partially ensured Quebec a veto, at least on the basis of ordinary law, by enacting the 1996 Act Respecting Constitutional Amendments. This piece of legislation ‘provides that no minister shall present a constitutional resolution to Parliament unless the amendment has first been consented to by a majority of provinces, including Ontario, Quebec, British Columbia, two Western Provinces, and two Atlantic provinces’.55 From a constitutional perspective, however, there are five amendment procedures. Section 38 of the Constitution Act 1982 sets forth the general procedure, which applies when none of the special ones are applicable. This procedure is also known as the ‘seven-fifty’ formula, since it requires a resolution of both chambers of the federal Parliament, as well as the ratification of at least two-thirds of the provincial legislatures, ie seven out of 10, representing at least 50 per cent of Canada’s population (Section 38(1)). Concerning amendments under this provision, which would curtail the legislative powers or rights of a province, there is a possibility of provincial optouts (Section 38(3)). The second amending formula requires unanimous provincial consent. With respect to certain matters that reflect the constitutional foundations of the federal system,56 amendments can be made ‘only where authorized by resolutions of the Senate and House of Commons and of the legislative assembly of each province’ (Section 41). The third amending formula applies to provisions relating to some but not all provinces and requires resolutions of the federal Parliament as well as of the legislature of each province to which the amendment applies (Section 43). Finally, Sections 44 and 45 foresee the fourth and fifth amendment procedures. These pertain to matters regarded as being of interest for only one level of government so that amendments can be effected, respectively, by the federal or the provincial legislatures alone. Under Section 44, the federal Parliament may, for example, amend the constitutional provisions concerning the ‘executive government of Canada’ without any provincial participation. Under Section 45, ‘the legislature of each

54 Reference Re Objection by Quebec to Resolution to Amend the Constitution [1982] 2 SCR 793. See PC Oliver, ‘Canada, Québec and Constitutional Amendments’ (1999) 49 University of Toronto Law Journal 546. 55 See Oliver, ‘Canada, Québec’ (1999) 602f. 56 These include, among other things, the right of a province to have a number of members in the House of Commons not inferior to the number of the senators, the use of English and French, amendments to the amending procedures themselves and the composition of the Supreme Court. On the latter issue, see box 6.H below.

182 Participation of Subnational Entities province may exclusively make laws amending the constitution of the province’.57 In sum, Canadian amending formulas accommodate the need for the direct subnational participation to a varying extent according to the specific matter concerned. Other federal countries also feature graduated subnational participation, even if there is usually less differentiation compared to Canada. Special procedures, in particular concerning amendments that affect the pillars of the federal system or merely one or some of the subnational entities, then have the character of rare exceptions to a general amending formula that otherwise only foresees the indirect involvement of the subnational entities through the second chamber.58 An example is Section 74(8) of the South African Constitution, which prescribes that amendments concerning a specific province or provinces have to be approved by the legislature or legislatures of the province or provinces concerned. Furthermore, Article 368(2) of the Constitution of India provides that amendments affecting the most relevant constitutional issues (eg presidential elections, the distribution of legislative powers, the representation of states in parliament) are to be approved by the national Parliament and ratified by the legislatures of not less than one-half of the states. Besides the legislatures of subnational entities, their direct involvement in amendment procedures can also take place through the participation of their electorate. This holds true for Switzerland and Australia as a consequence of the emphasis which these federations place on the principle of popular sovereignty (Article 1 of the Swiss Constitutions of 1848, 1874 and 1999; the Preamble to the 1900 Constitution of Australia). To put it differently, ‘the idea that the federation consists of both one “people” and a plurality of “peoples” is entrenched within the Constitution’.59 In both countries, a constitutional amendment needs both the consent of the majority of voters in the entire federation and the consent of the majority of voters in a majority of the subnational entities. In Switzerland, the people of the cantons take part in both the total and partial revision of the federal Constitution through a mandatory referendum (Article 140 of the Constitution).60 After simple majorities in both chambers of the federal Parliament, ratification requires simple majorities of both all votes cast by the people throughout the country (Volksmehr) and of the cantons (Ständemehr). The Ständemehr, as part of this double majority rule, signifies the equality, in principle, of the cantons. But this federal corrective to the Volksmehr also means that, with optimal dispersion in the smallest cantons, only 10 per cent of the Swiss electorate could theoretically block a constitutional amendment.61 While the method of determining the vote of a canton was originally left to its own autonomous decision, Article 142(3) now mandates that the majority of the cantonal population is decisive. In doing so,

57

On the peculiar case of provincial constitutionalism in Canada, see ch 5.1.1. See section 6.3.1.2 below. 59 Aroney, ‘Formation, Representation’ (2006) 302. See also MD Kirby, ‘Popular Sovereignty and the True Foundation of the Australian Constitution’ (1996) 3 Deakin Law Review 129. 60 See J Reich, ‘An Interactional Model of Direct Democracy: Lessons from the Swiss Experience’ (5 June 2008) ssrn.com/abstract=1154019. 61 See RE Germann, ‘Die Europatauglichkeit der direkt-demokratischen Institutionen in der Schweiz’ (1991) 31 Schweizerisches Jahrbuch für Politische Wissenschaft 257, 262. 58

Procedural Forms of Participation 183 this provision limited the cantons’ constitutional autonomy62 but further expanded the clout of direct democracy. Beyond the approval of federal constitutional changes in a referendum, amendments may also be triggered by the people themselves through the instrument of a popular initiative. If 100,000 signatures are collected within 18 months in favour of a proposed amendment, there has to be a popular vote on it that again requires the above-mentioned double majority of votes cast and of cantons (Articles 138 and 139). In Australia, a proposed amendment has to be approved by an absolute majority of each chamber of the Parliament and then ratified by a majority of the voters nationwide and a majority of the voters in a majority of the states (Section 128 of the Constitution). Even if this provision was largely modelled on the abovementioned Swiss example, the right to initiate amendments is left, exclusively with the national Parliament. This difference and the generally broader use of referendums and other forms of direct democracy in Switzerland seem to be among the reasons that account for the fact that constitutional amendments in Australia have been much less frequent.63 6.3.1.2 Indirect (or Absence of) Participation Through Second Chambers Indirect participation grants subnational units the possibility to be involved in constitutional amendment procedures only through second chambers. This form of participation does not formally differ from the legislative process, except that it requires special majorities. As a consequence, subnational entities depend on the effectiveness of their representation in the federal second chamber. The German Federal Council participates on an equal footing with the Parliament (Bundestag) in the process of amending the national Constitution, as the same special majorities are required in both organs. Article 79(2) of the Basic Law prescribes that constitutional amendments be passed by a majority of two-thirds of the members of the Parliament and of the votes of the Federal Council. Article 79(3) of the Basic Law also provides for explicit limits on the amending power, as any constitutional changes affecting the division of the federation into Länder, their participation on principle in the legislative process, the fundamental rights catalogue (Articles 1–19) and the crucial constitutional principles (Article 20) are inadmissible. This ‘eternity clause’ entails that constitutional amendments eliminating subnational participation in national legislation or the federal territorial organisation are prohibited, even if all Länder were willing to consent, through the Federal Council, to such amendments. As we will see, however, this clause does not guarantee the existence of each single Land, but merely requires a division of the federation into (at least two) Länder.64 Similar to the above-mentioned direct involvement of subnational entities, also their indirect participation may be graduated according to different amendment

62

See ch 5.1.1. C Saunders, The Constitution of Australia: A Contextual Analysis (Oxford, Hart Publishing, 2010) 48ff. 64 See box 6.F below. 63 See

184 Participation of Subnational Entities procedures. For example, Article 44(1) of the Austrian Constitution only lays down the general amendment procedure: both constitutional laws and constitutional provisions contained in federal laws have to be approved by the first chamber of Parliament (Nationalrat) with the presence of at least half of its members and a two-thirds majority of the votes cast. In addition to this general procedure, however, the Austrian Constitution provides several special procedures. First, Article 35(4) establishes that changes in the composition and selection of members of the Federal Council, regulated in Articles 34 and 35, require in the second chamber beyond the above-mentioned normal quorums a majority of the representatives from at least four Länder. Second, Article 44(2) of the Constitution requires the approval of the Federal Council with the presence of at least half of the members and a two-thirds majority of the votes in case of constitutional amendments that limit the legislative or administrative powers of the Länder. This procedure provides the second chamber with an absolute veto65 which it may use in consultation with the Länder governments to protect subnational interests. In practice, however, the Federal Council, making decisions mainly based on party politics rather than territorial considerations, has been reluctant to use this veto power. Between the entry into force of Article 44(2) in 1985 and the end of 2014 it has given its consent to amendments curtailing Länder powers in as many as 257 cases, among them, the rubberstamping of all amendments necessary for Austria’s accession to the EU even without securing, in return, the promised reform of the federal system.66 A third special procedure is set forth in Article 44(3) which requires a referendum in case of a total revision of constitution, affecting the latter’s basic principles, or in case of a partial revision upon request by one-third of the members of the first or second chamber. As only the nationwide result counts, this referendum has, unlike in the above-mentioned popular votes in Switzerland and Australia, no specific subnational dimension. Federalism, however, is one of the basic constitutional principles whose infringement is regarded as a total revision and thus triggers a mandatory referendum.67 Notable examples of extra-European federal systems where subnational participation in amendments occurs through second chambers include South Africa and India. Section 74(3a) of the South African Constitution sets forth the general amendment formula, according to which two-thirds of the members of the National Assembly must approve proposed amendments. However, amendments regarding several matters require, beyond a vote in the first chamber, the adoption by the National Council of Provinces with a supporting vote of at least six provinces. Examples are amendments affecting the bill of rights (Section 74(2)), the second chamber or certain specifically provincial matters (Section 74(3b)). Whereas, in India, the direct participation of state legislatures is limited to the above-mentioned cases of amendments to fundamental principles, indirect involvement through the Council of States

65

See P Bußjäger, Die Zustimmungsrechte des Bundesrates (Wien, Braumüller, 2001). K Kössler, ‘Reform of the Second Chamber or Its Perpetuation? The Austrian Dilemma and Its Implications for the Italian Senate’ (2016) 37 Istituzioni del Federalismo 339, 354f. 67 The other basic constitutional principles are democracy, republicanism, the rule of law, the separation of powers and fundamental rights. 66 See

Procedural Forms of Participation 185 (Rajya Sabha) is the rule. The general amending formula set out in Article 368(1) prescribes that each chamber of the federal Parliament must pass proposed constitutional amendments by a majority of its members and by a majority of no less than two-thirds of the members present and voting. Similar to Germany, there are some parts of the Indian Constitution, including federalism, that are beyond the power of amendment, even if both the national and state governments were willing to change these parts. Unlike Article 79(3) of the Basic Law, this results in the case of India not from a textual provision but from the basic structure doctrine developed by the Supreme Court.68 As mentioned above, the effectiveness of indirect subnational participation in constitutional amendments depends on whether a second chamber really represents territorial interests. This is not the case, especially in devolutionary and relatively recent federal systems. In Italy, for example, there is no effective representative institution of the regions within the national Parliament, as the Senate is simply ‘elected on a regional basis’ (Article 57 of the Constitution). This provision establishes an extremely weak link between the Senate and the regions, which in fact only affects the electoral system and the minimum numbers of senators that each regional electoral district has. According to the procedure for constitutional amendment, an absolute majority of the members of both chambers in two separate readings is sufficient (Article 138 of the Constitution). If a qualified majority of two-thirds of the members of each chamber is reached in the second reading, no nationwide referendum takes place. If, however, this special majority requirement is not met, a referendum may be requested within three months of the publication of the amendment by one-fifth of the members of one of the chambers, 500,000 voters or five regional legislatures. Apart from this possibility, under certain circumstances, to demand a popular vote, the only possible form of (direct) regional involvement concerns the autonomy statutes of the special regions which have the rank of a national constitutional law (Article 116(1)). All these statutes stipulate regarding their amendment that without a qualified majority of two-thirds of the members in the two chambers of Parliament, a referendum can be held in the affected region. In Spain, qualified majorities in both chambers are required, unlike in Italy, not to exclude the possibility of an optional referendum but for the approval of all constitutional amendments. The Spanish Constitution sets forth a general procedure involving special parliamentary majorities and an optional referendum (Article 167) and a special procedure for changes to certain parts of the Constitution that demands even larger majorities plus a mandatory referendum (Article 168). None of these two procedures foresees any direct participation of the autonomous communities. Since the latter only appoint a small group of senators (Article 69), the participation of the second chamber in the process of constitutional amendment is likewise not an effective mechanism to ensure subnational influence.69

68 His Holiness Kesavananda Bharati Sripadagalvaru v State of Kerala, 1973 AIR 1461. See box 2.C; Minerva Mills Ltd v Union of India, 1980 AIR 1789. See box 8.K. 69 See F Balaguer Callejón, ‘La Constitución’ in F Balaguer Callejón (ed), Manual de derecho constitucional (Madrid, Tecnos, 2007) 121f.

186 Participation of Subnational Entities 6.3.2 Territorial Changes A specific form of subnational participation concerns the key issue of territorial reorganisation. Although such restructuring mostly requires constitutional amendments, we will see that this is not necessarily so in all federal systems and for all types of changes to subnational boundaries. As territory is a constitutive feature of subnational units, altering its boundaries normally requires the consent of the units affected. Such procedures are quite telling about the essence of the federal compact, which would suggest in principle a veto right of subnational entities as to their very existence and territorial integrity. On the other hand, federal constitutions still provide for mechanisms for territorial changes. These issues were at stake in a highly controversial case before Germany’s Federal Constitutional Court soon after World War II. 6.F Germany 1951–56: Abolishment of a Land Without its Consent? Soon after the adoption of the Basic Law, the federal Government was involved in the restructuring of the southwestern part of Germany which then consisted of the Länder of Baden, Württemberg-Baden and Württemberg-Hohenzollern, as established by the Allied Powers. The federal Government attended to this matter after the failure of negotiations between the three Länder under Article 118 of the Basic Law. This provision stipulated that territorial reorganisation could take place either by mutual agreement of the three Länder involved or through a federal law, in which case a referendum had to take place. In 1951, federal legislation was enacted containing a procedure for a referendum on the creation of a single Land Baden-Württemberg, according to which a majority in three of four newly established electoral districts would suffice. The Land Baden, which had been split up for the purposes of the referendum into two electoral districts, challenged the constitutionality of the federal legislation on the ground that it had violated the principles of democracy and federalism. The Court did not agree with this view.70 It held that, in principle, self-determination of the people does indeed follow from a constitutional principle of democracy. But it also emphasised that Baden was, as a member of a federation, not independent but restricted with regard to its sovereign power. The judges acknowledged a certain tension between the constitutional principles of democracy and federalism, but refrained from construing the latter as prohibiting the abolishment of a state against the will of its people. It thus famously characterised Germany as a ‘malleable federal state’ (p 48). Even though federal constitutions would typically guarantee the existence and territory of the subnational entities, the Basic Law was seen as explicitly deviating from this rule. First, the ‘eternity clause’ under Article 79(3) would only ordain, in principle, that the federation must be divided ‘into Länder’ but not protect the territorial existence of an individual Land. Second, Articles 29 and 118 expressly set forth procedures for territorial reorganisation. According to the Court, Article 118 did not require the same procedure as the general provision for territorial reorganisation under Article 29, which foresaw both a local and national referendum. Indeed, Article 118 as a lex specialis concerning southwest Germany even explicitly stated that reorganisation may in this case ‘deviate from the provisions of Article 29’.

70

1 BVerfGE 14 (Southwest State).

Procedural Forms of Participation 187

Two months after the ruling, the referendum took place in December 1951 with indeed three districts voting, as required, in favour of the merger and one, Südbaden, against it. Consequently, the unified Land of Baden-Württemberg, the so-called Southwest State (Südweststaat), was established in 1952. When a few years later the federal Ministry of the Interior rejected the demand of a group from Baden for a popular vote under Article 29 (only applicable after the end of Allied occupation in 1955), the Federal Constitutional Court allowed for this vote to take place.71 Although it reiterated that there is no constitutional guarantee for the existence of an individual Land, the referendum in 1951 could not be considered as a vote within the meaning of Article 29, in particular because it made it possible for the more populous Württemberg and Hohenzollern to outvote the less populous Baden. After this second vote had been constantly postponed, it eventually took place after a court order in 1970. The outcome was that 82 per cent of Baden’s population voted to maintain the unified Land of Baden-Württemberg.

The procedures for altering the territories of subnational entities vary significantly, in particular regarding the degree of constitutional entrenchment. In the United States, territorial alteration requires the consent of the legislatures of the states affected, as well as of Congress (Article IV, section 3, clause 1 of the Constitution). In Canada and Australia, territorial alteration requires a constitutional amendment.72 In the Canadian case, simple changes to boundaries need the consent of the legislatures of the provinces affected (Section 43 of the Constitution Act 1982), while the establishment of new provinces is subject to the above-mentioned ‘seven-fifty’ formula (Section 42(1f)). Australia goes beyond that and foresees an additional requirement. Besides the parliament of a state concerned, a majority of its voters also has to approve an initiative to ‘increase, diminish, or otherwise alter the limits of the state’ (Section 123 of the Australian Constitution). This consent of the electorate is not required in case of the formation of new states (Section 124). In a number of countries, constitutional amendments are required only if the proposed changes affect the existence and the number of subnational entities. In Austria, ‘changes in the composition of the Länder’, ie a merger or breakup of existing subnational entities or the admission of new ones, need amendments to both the federal and Länder constitutions (Article 2(3) of the Constitution). Mere changes to subnational boundaries, by contrast, have been deregulated with a reform in 2008.73 They only require concurrent ordinary laws of the federal Government and the Länder concerned, even if a two-thirds majority is still necessary in the first chamber of the Austrian Parliament. Minor boundary adjustment may be even effected by concurrent legislation of the Länder involved without any federal participation (Article 3(3–4) of the Constitution). Also in Italy, procedures and the role of subnational entities vary

71

5 BVerfGE 34 (Baden Home Association). K Wiltshire, ‘Reforming Australian Governance: Old States, No States or New States?’ in AJ Brown and JA Bellamy (eds), Federalism and Regionalism in Australia: New Approaches, New Institutions? (Canberra, The Australian National University, 2006). 73 See N Sonntag, ‘Federalism and Territory: The Austrian Experience’ in S Mangiameli (ed), Federalism, Regionalism and Territory (Milan, Giuffrè, 2013) 34f. 72 See

188 Participation of Subnational Entities according to the type of territorial change. Article 132(1) of the Constitution requires a constitutional law in order to create new or merge existing regions and the consent of the electorate of the regions affected. Less dramatic territorial changes, ie requests from provinces and municipalities to be aggregated to another region, merely require popular votes in the territories concerned, consultation with the regional legislature and an ordinary law adopted by the national parliament (Article 132(2)). In South Africa, territorial (re)organisation has been a highly disputed issue from the outset.74 In light of this, it is not surprising that Section 74(3b)(ii) of the Constitution requires a constitutional amendment for a change to provincial boundaries. This provision stipulates that a two-thirds majority of the members of the National Assembly plus the approval by the National Council of Provinces with the supporting vote of at least six provinces is needed. If, however, territorial alteration ‘concerns only a specific province or provinces’, there is an additional requirement: a territorial change cannot be passed by ‘the National Council of Provinces … unless it has been approved by the legislature or legislatures of the province or provinces concerned’ (Section 74(8)). 6.G South Africa 2006: Provincial Consent Procedure Regarding Territorial Changes In 1993, South Africa appointed the Commission for Demarcation/Delimitation of States/ Regions/Provinces. The final outcome of the demarcation process was the establishment of nine provinces whose territorial boundaries were then entrenched in the 1996 Constitution (Section 103(2)). When several territorial disputes arose, the national Parliament sought to resolve these controversies by passing the Twelfth Amendment to the South African Constitution in 2005. Under Section 74(8), the KwaZulu-Natal legislature approved the part of the Twelfth Amendment that would transfer the area of the municipality of Matatiele from KwaZulu-Natal to the Eastern Cape province. Matatiele then challenged the validity of the Twelfth Amendment before the Constitutional Court. The Court stated that, although the territorial alterations were consistent with the constitutional amendment procedure set forth in Section 74(8), the national Parliament did not meet the levels of public participation that South Africa’s Constitution requires when changing provincial boundaries.75 Indeed, Section 118(1)(a) stipulates that legislative organs ‘facilitate public involvement in the legislative and other processes of the legislature’. According to the Constitutional Court, ‘[i]n considering whether or not to approve that part of the Twelfth Amendment, the KwaZulu-Natal legislature was required by Section 118(1)(a) to facilitate public involvement by holding public hearings in the area of Matatiele’ (para 85). The failure by the provincial legislature to meet this procedural requirement was thus regarded as a clear violation of the national Constitution.

As a matter of fact, many processes of drawing and redrawing subnational boundaries are related to ethno-cultural differences; they are then an attempt to provide a

74 See J de Visser et al, ‘The Role of Ethnicity in the Demarcation of Internal Boundaries in South Africa and Ethiopia’ in Mangiameli (ed), Federalism, Regionalism (2013) 276ff. 75 Matatiele Municipality v President of the Republic of South Africa 2006 (5) SA 47 (CC).

Procedural Forms of Participation 189 procedural answer to cleavages and group claims in diverse societies.76 Interesting examples can be found, for example, in India and Switzerland. When India and Pakistan gained independence in 1947, the ‘geographical’ legacies of the British Raj were 14 provinces under direct colonial rule and 565 princely states. Unsurprisingly, the establishment of India as a federal country composed of states therefore required large-scale territorial reorganisation. Prime Minister Nehru’s preference was clearly for heterogeneous states which he regarded as a safeguard against separatist tendencies. In 1956, however, the Indian Government responded to pressures from various parts of the country and followed the recommendation of the States Reorganisation Commission to restructure the then-existing 27 states along linguistic lines into 14 new states. This led to a continuous process of state creation and territorial change,77 which first continued until the 1970s and then again gained momentum in the 1990s. The youngest of India’s currently 29 states is Telangana, established in 2014. This process was facilitated by the enormous flexibility (to the benefit of the national Government) under Section 3 of the Constitution. According to this provision, a federal ordinary law is sufficient to form a new state or to alter the boundaries of an existing state and the bill is only referred to the legislatures of the states concerned ‘for expressing its views thereon’. The federal Parliament is, however, on no account bound by these views. The sole exception applies to Jammu and Kashmir. Appendix 2 to the Constitution provides that ‘no bill providing for increasing or diminishing the area of the state of Jammu and Kashmir or altering the name or boundary of that state shall be introduced in parliament without the consent of the legislature of that state’. In Switzerland, territorial reorganisation of the cantons in response to diversityrelated claims has been much less comprehensive and entirely different in terms of the broad procedural involvement of subnational entities (and even local governments). In 1979, the new canton of Jura was carved out of the canton of Bern along linguistic and religious lines. At the Congress of Vienna in 1815, the territories of Jura with their predominantly Catholic, French-speaking population had been integrated into the largely Protestant, German-dominated canton of Berne to compensate the latter for other territorial losses. Tensions between the different groups eventually hit their peak in the 1970s and resulted in a long series of popular votes.78 First, a cantonal referendum was held in 1970 that approved an amendment of the Constitution of Berne to provide for the possibility of a referendum in the Jurassic parts of the canton about their separation. This vote took place in 1974 and resulted in the decision to form a canton of Jura. Then, referendums have been organised since 1975 (until the present day) in a number of predominantly Frenchspeaking districts and municipalities about whether to join the new canton or remain with Berne. In 1977, the population of what was going to be the canton of Jura

76 See M Nicolini, ‘Regional Demarcation, Territorial Alteration, and Accommodation of Divided Societies’ (2015) 94 Revista de Derecho Político 53. 77 See A Majeed, Clouds over Federalism: The Real Working of the Indian Polity (New Delhi, Manak, 2010) 23ff. 78 See EM Belser and R Setz, ‘The Territorial Structure of Federal Switzerland Revised: About the Difficulty of Being Small,’ in Mangiameli (n 73) 5.

190 Participation of Subnational Entities r atified the cantonal constitution.79 Finally, in a national referendum, the people of Switzerland accepted the accession of the new canton to the Swiss federation with a large majority so that Jura was officially created in 1979. Since the total revision of 1999, the Swiss Constitution explicitly states that the procedure for any creation of a new canton must be initiated by the cantons concerned and cannot be prescribed from above, as in India. The federal Government is obliged to ‘protect the existence and territory of the cantons’ (Article 53(1)) and any change in the number of cantons requires the consent of the citizens and the cantons concerned, as well as the consent of the people and all other cantons (Article 53(2)). 6.3.3 Legislative (and Administrative) Functions The second chamber of the national parliament takes part in the legislative procedures either with regard to (nearly) all bills or only concerning specific bills that are of particular relevance for the subnational entities. This participation may then occur on an equal footing with the other chamber of parliament or through a merely suspensive veto which can be overruled by the first chamber.80 At the same time, the second chamber is normally excluded from the relationship between the legislative and executive branches in the countries that follow a parliamentary system. A notable exception is Italy, where the Government can only operate if it enjoys the confidence of both the Chamber of Deputies and the Senate (Article 94 of the Constitution). Legally, withdrawal of confidence and ministerial responsibility are normally in the hands of the first chamber, although politically the more significant the powers of the second chamber, the more they can exercise de facto control over (or at least block) the national executive. This may hold true even for presidential systems, the US Senate being a case in point. The latter’s strong position follows from certain particular functions regarding appointments81 and, above all, from equal participation of both chambers in the legislative process. The House of Representatives and the Senate have the same powers in relation to both legislative initiative and the approval of bills (Article I, section 1 of the Constitution). The only exception is established by Article I, section 7, clause 1: ‘All bills for raising revenue shall originate in the House of Representatives’. However, ‘the Senate may propose or concur with amendments as on other bills’. Even if the House thus takes the initiative, this amendment power gives the Senate a great deal of influence. If both chambers enact different revenue legislation, they must meet, like for any other legislation, in a special conference committee to reach an agreement.82 79

On the controversial ‘reunification clause’ in Art 138 of the 1977 Constitution of Jura, see ch 8.2.2. an overview, see M Russell, ‘Elected Second Chambers and Their Powers: An International Survey’ (2012) 83 Political Quarterly 117. For a classification of second chambers on a continuum between perfect and imperfect bicameralism by means of five categories, see SC Patterson and A Mughan, ‘Fundamentals of Institutional Design: The Functions and Powers of Parliamentary Second Chambers’ (2001) 7 The Journal of Legislative Studies 39, 42. 81 See section 6.3.4 below. 82 See JW Leckrone, ‘The United States Senate’ in Luther et al (n 2) 119. 80 For

Procedural Forms of Participation 191 Almost equal participation of both chambers in the legislative process has not remained a feature limited to the United States. Subsequently, it was also introduced by several federal systems which are otherwise quite diverse (eg Article 148(2) of Swiss Federation, Section 17 of Canada’s Constitution Act 1867, Section 53(5) of the Australian Constitution, Section 107 of the Indian Constitution and Article 44 of the Argentinian Constitution). In practice, however, the second chambers of these countries are clearly weaker than the US Senate which is mostly due to the rules about appointment and composition83 and the predominance of party allegiance. In parliamentary federations influenced by Westminster traditions, there is typically a minor exception to equal participation, as money bills or bills imposing taxation must originate in the first chamber (Section 53(1) of the Australian Constitution, Section 53 of the Canadian Constitution Act 1867 and Section 109 of the Indian Constitution). In several other federal systems, the degree of participation by second chambers in national legislation varies. In South Africa, for example, NCOP is involved in the adoption of most ordinary legislation, but its powers depend on the matter concerned (Sections 75–77 of the Constitution). Money bills and bills not affecting the provinces may be passed by a majority of the votes cast by the individual delegates to the NCOP. Ordinary bills that affect the provinces, by contrast, require a Germanstyle block vote, as a single vote is cast on behalf of each delegation as a whole. Disagreements between the two chambers are referred to a mediation committee (Section 78). In the case of Austria’s Federal Council, participation is similarly foreseen for almost all ordinary legislation with bills regarding federal assets and the federal budget being notable exceptions (Article 42(5) of the Constitution). Even if the second chamber’s decision (within eight weeks) about whether to object to a bill is thus required in almost all cases, its impact is weaker than in the South African case. With respect to most laws, the first chamber of the Austrian Parliament may overrule any objection with a majority decision in the presence of at least half of its members (Article 42(1–4)). Disputes within the Federal Council about whether to impose such a merely suspensive veto or not, appear to have weakened it further. In 2003, for example, the first chamber had passed a controversial bill which in the second chamber found neither sufficient support for a vote of approval nor a sufficient number of representatives to lodge an objection. The Constitutional Court eventually ruled that the bill had become law without any decision taken by the Federal Council.84 Absolute vetoes of the Federal Council are only foreseen regarding few specific types of legislation (eg certain federal framework laws requiring implementation by the Länder (Article 15(6)) and therefore have rather little practical relevance. Compared to Austria, absolute veto power has played a much more crucial role regarding the adoption of federal legislation in Germany (Articles 76–78 of the Basic Law). Basically, the Federal Council only has a suspensive veto which the Parliament

83 84

See section 6.2.1.2 above. VfSlg 17173/2004.

192 Participation of Subnational Entities (Bundestag) may override by a simple majority vote or, if the veto was passed by a two-thirds majority, an equivalent majority of two-thirds. However, in a number of scattered provisions, the Basic Law expressly requires the approval of the Federal Council regarding legislation that affects issues such as finances and Länder implementation of national laws. The jurisprudence of the Federal Constitutional Court extended the scope of this consent legislation enormously by giving a very broad interpretation to the approval of federal legislation regarding Länder administrative procedures and agency organisation (former Article 84(1) of the Basic Law).85 The amendment of this provision in the wake of the 2006 reform of federalism clearly reduced the share of consent legislation, which had amounted over the decades before the reform to 55–60 per cent of all federal law-making.86 On the other hand, the newly introduced Article 104a(4) extended the approval requirement to other cases so that the Federal Council is still destined to play a crucial role. A crucial question concerning the power relations between the first and second chamber of a federal parliament is how disagreements between them are settled in areas where they are on an equal footing and the first chamber therefore does not prevail automatically. In order to facilitate dispute resolution, mediation committees of different kinds exist, as mentioned above, in several countries (eg the United States, Switzerland, India and South Africa). These committees vary significantly in terms of their composition, procedures and significance. For example, the President of India may in the case of a parliamentary deadlock convene joint sittings of the two chambers of Parliament (Article 108 of the Constitution). Yet, these are dominated by the House of the People (Lok Sabha), which has more than twice as many representatives as the Council of States (Rajya Sabha) and may thus easily reach the majority needed in the joint session to pass the controversial bill by itself. Moreover, the President convenes a joint sitting on the advice of the national Government which commands a majority in the House of the People so that this mechanism usually serves the purpose of making the latter prevail over the Council of States. In any event, joint sittings have only been held three times in nearly seven decades. This contrasts starkly with the mediation committee that Article 77(2) of the German Basic Law foresees for disagreements concerning bills under the suspensive veto procedure and, more importantly, bills requiring the consent of the Federal Council. For a number of reasons, this body is clearly more significant and less biased against the (ambassadorial) second chamber than India’s joint sittings. First, it may also be convened by the Federal Council (besides the federal Government and Parliament). Second, the mediation committee is composed on an equal basis of 16 representatives from the Parliament and 16 from the Federal Council (one from each Land), with the latter being formally unbound by instructions. Third, even if a

85

37 BVerfGE 363, 385ff (Bundesrat); 55 BVerfGE 274, 320ff (Vocational Training Act). Gunlicks, ‘Legislative Competences, Budgetary Constraints, and the Reform of Federalism in Germany from the Top Down and the Bottom Up’ in M Burgess and GA Tarr (eds), Constitutional Dynamics in Federal Systems: Subnational Perspectives (Montreal, McGill–Queen’s University Press, 2012) 73. 86 AB

Procedural Forms of Participation 193 simple majority would be sufficient in theory, much broader support is necessary in practice, as majorities in both the Parliament and the Federal Council have to confirm the compromise decision of the mediation committee afterwards. While in the exercise of legislative functions there are thus no fundamental differences between the senate and the council model, a substantial distinction can be noted with regard to administrative functions. In fact, with regard to administrative functions, the role of senatorial second chambers is understandably limited, as they are chambers elected by either the population or the legislatures of the subnational units and their main role is to include such units in the legislative process. To the contrary, Germany’s Federal Council has, as an ambassadorial second chamber in the context of administrative federalism, a quite different role regarding the implementation of national legislation, as this is largely the responsibility of the Länder which it is supposed to represent. Regardless of whether the Länder execute national laws in their own right (Article 84) or merely on federal commission (Article 85), a particular form of participation through the (ambassadorial) second chamber is required. It is only with the consent of the Federal Council that the federal Government may issue general administrative rules (Articles 84(2) and 85(2)). 6.3.4 Appointments Subnational units are granted the right to participate in a number of appointment processes at the national level. Such participation normally takes place through second chambers (only rarely through other participatory bodies), for the simple reason that the most significant appointments are usually reserved for parliamentary bodies and their functional equivalents. Appointments may regard key political figures such as heads of state, but also civil servants and judges. Subnational entities often participate in the election of the head of state in parliamentary federal systems, whereas in monarchic countries the appointment of the head of state normally follows the constitutional provisions regarding succession to the throne. In this regard, Malaysia represents a notable exception, as the monarch and head of state (Yang di-Pertuan Agong) is elected for a term of five years by the Conference of Rulers (Section 32(3) of the Constitution).87 This Conference comprises the different Royal Highnesses governing the states that constitute the federation (Article 38). As to the President of Germany, Article 54(1) of the Basic Law establishes that he or she is elected by the Federal Assembly, which consists of the members of the Parliament (Bundestag) and an equal number of members elected by the Länder parliaments on the basis of proportional representation (Article 54(3)). The same occurs in Italy, where the President of the Republic is elected in a joint session of the Chamber of Deputies and the Senate which is complemented with three delegates from each region elected by the respective assembly (Article 83 of the Constitution). In India, Article 54 of the Constitution provides that ‘[t]he President shall be elected

87

See Harding, The Constitution (2013) 116ff.

194 Participation of Subnational Entities by the members of an electoral college consisting of the elected members of both houses of parliament and the elected members of the legislative assemblies of the states’. The heads of states in parliamentary systems are thus elected by quite different joint assemblies that feature direct participation of subnational entities or merely indirect involvement through the second chamber. As for presidential regimes, the Constitution of the Unites States designs a system in which the states are, on the one hand, construed as districts for nominating electors, and on the other, charged with determining the respective electoral system. Article II, section 1, clause 2 states that ‘[e]ach State shall appoint, in such manner as the legislature thereof may direct, a number of electors, equal to the whole number of Senators and Representatives to which the state may be entitled in the Congress’. In contrast to the just-mentioned examples of presidents and an elected monarch, there is no single person acting as the head of state in Switzerland. The typical functions of this position are fulfilled by the federal executive, ie the seven members of the Federal Council, as a collective body, in particular by the President of the Confederation who is elected from among these seven members for one year. Both the Federal Council and the President of the Confederation are elected by the Federal Assembly, ie the joint sitting of both chambers of the national Parliament (Article 168(1) of the Constitution).88 Article 175(4) calls for ensuring that the various geographical and linguistic regions of the country are appropriately represented within the Federal Council, thus entrenching the representation of French- and Italian-speaking minorities within the federal executive. Moreover, a constitutional convention does not allow a canton to have more than one representative within the Federal Council. A very significant area of participation by subnational units in appointment processes affects the judiciary. While the appointment of (federal) judges by the national government is widespread in common law federal systems, in civil law regimes this normally only affects the constitutional courts. Furthermore, unlike for high political office—where participation can be channelled through other bodies, including through popular vote—judicial appointments normally take place exclusively through the second chamber. Article II, section 2, clause 2 of the US Constitution establishes that the President shall nominate, and by and with the advice and consent of the Senate, shall appoint ambassadors, other public ministers and consuls, judges of the supreme Court, and all other officers of the United States, whose appointments are not herein otherwise provided for, and which shall be established by law.

This so-called Advice and Consent Clause thus requires the involvement of the Senate for the appointment of federal judges, although the divisions over judicial appointments (as well as over other appointments) have gradually become partisan, rather than regional, divisions, and the process now primarily works out to be political rather than federal.89 A legacy of the original federalism-related intent of this

88 See N Schmitt, ‘Swiss Confederation’ in J Kincaid and GA Tarr (eds), Constitutional Origins, Structure, and Change in Federal Countries (Montreal, McGill–Queen’s University Press, 2005) 347–80. 89 See DA Strauss and CR Sunstein, ‘The Senate, the Constitution, and the Confirmation Process’ (1992) 101 The Yale Law Journal 1491, 1495 and 1491ff.

Procedural Forms of Participation 195 appointment system is, however, so-called senatorial courtesy: the Senate can refuse to ratify a presidential appointment of a judge or an official in or from a state when senators from the President’s party from that state oppose the appointment.90 Several federal countries follow, with certain variations, the ‘advice and consent’ model stemming from the US Constitution. In Mexico, Articles 76(2) and 76(8) of the Constitution confer upon the Senate the power to ratify the presidential nominations of the Attorney-General and of the Supreme Court justices. The justices are appointed from a list of three candidates indicated by the President. According to Article 96(1), after hearing all candidates, the Senate selects one of them by a two-thirds majority. The same majority is required by Article 99(4) of the Constitution of Argentina, while Article 101 of the Brazilian Constitution only requires an absolute majority of the Senate and thus less compromise between the main political parties. The Constitution of Nigeria foresees a similar appointment process. Under Section 231(1–2), the Chief Justice and the other members of the Supreme Court are appointed by the President but upon a recommendation by the National Judicial Council (Schedule 3, Part 1(I)) and subject to confirmation by the Senate.91 Similarly to Nigeria, other federal systems also feature consultative bodies, usually composed of leading members of the legal profession, that provide advice to the federal executive in appointment processes. Sometimes this effort of ‘depoliticisation’ goes at the expense of subnational participation. This holds true, especially for Australia where Section 6 of the 1979 High Court of Australia Act requires the federal Government to consult with state attorney-generals. But this often only amounts to a non-binding exchange of views and there is no practice of guaranteeing balanced representation of the states in the High Court. The leeway of the federal executive in South Africa is clearly more limited. Indeed, the Judicial Service Commission enshrined in Section 178 of the Constitution is regarded as having the crucial function,92 together with the President, to nominate judges, including the members of the Constitutional Court (Section 174). In the appointment process, ‘[t]he need for the judiciary to reflect broadly the racial and gender composition of South Africa must be considered’ (Section 174(5)). Moreover, the composition of the Judicial Service Commission itself has been deliberately designed in such a manner that people from diverse backgrounds participate in its deliberations.93 Like in the South African case, there is in Canada no formal role of subnational entities in appointing the members of the highest court, but instead a requirement that the justices have diverse backgrounds. Canada stands out, however, among federal systems insofar as the procedure for the selection of the Supreme Court judges is not at all regulated by constitutional law. Section 96 of the Constitution Act 1867 only provides for the federal appointment of judges to the provincial superior, district and county courts which is simply because the Supreme Court was only established

90 For further details on Senate voting, see L Epstein et al, ‘The Changing Dynamics of Senate Voting on Supreme Court Nominees’ (2006) 68 The Journal of Politics 296. 91 See J Akande, ‘The Legal Order and the Administration of Federal and State Courts’ (1991) 21 Publius 61, 64f. 92 Cape Bar Council v Judicial Services Commission 2012 (4) BCLR 406 (WCC), para 21. 93 Judicial Service Commission v Cape Bar Council 2013 (1) SA 170 (SCA), para 35.

196 Participation of Subnational Entities in 1875.94 It is therefore Section 4(2) of the 1985 Supreme Court Act which states that the nine justices have to be appointed by the Governor General, ie in practice by the federal Government. Section 6 foresees, however, that at least three of the judges shall be appointed from a superior court of Quebec or from among the advocates of that province. Moreover, according to a constitutional convention, three other judges are from Ontario, two are from the Atlantic provinces and one is from the Western provinces, so that the Supreme Court essentially features a regional composition. The guarantee of the three ‘Quebec seats’ serves the dual purpose of recognising the special position of this predominantly French-speaking province and of ensuring representation of Quebec’s civil law jurisdiction in the Court. The controversial question of who may be appointed to these ‘Quebec seats’ was recently clarified in a judgment of the Supreme Court. 6.H Canada 2014: Who is Eligible for the ‘Quebec Seats’ in the Supreme Court? In 2013, the federal Government appointed Justice Marc Nadon to one of the ‘Quebec seats’. According to Section 5 of the Supreme Court Act, in general a potential justice must be either a member of a provincial superior court or a barrister or advocate with at least 10 years’ standing at a provincial bar. Regarding Quebec, the above-mentioned Section 6 additionally provides that ‘its’ three judges must be either from a superior court of Quebec or ‘from among the advocates of that province’. Nadon did not meet the first requirement under Section 6, as he was at the time of the appointment a supernumerary judge of the Federal Court of Appeal. The fulfillment of the second requirement was in doubt because he had only been a member of the Quebec Bar before his appointment. While the province of Quebec and a lawyer from Ontario challenged the appointment, the federal Parliament amended Section 6 of the Supreme Court Act by stating ‘for greater certainty’ that, similar to the wording of Section 5, ‘a judge is from among the advocates of the province of Quebec if, at any time, they were an advocate of at least 10 years standing at the bar of that province’. In short, former members of the Quebec bar, like Nadon, were also declared eligible. The federal Government then referred the issue to the Supreme Court for ultimate clarification. The judges ruled that only current members of Quebec’s superior courts or current members of the Quebec bar are eligible to be appointed to that province’s seats in the Supreme Court.95 Only one judge dissented, arguing that the two provisions in question had to be read together and that, consequently, ‘from among the advocates of that province’ (Section 6) would include, like the general provision of Section 5, former bar members. The majority, however, did not follow this view and held that the above-mentioned amendment of the Supreme Court Act was not merely declaratory of the existing law but in fact changed the composition of the Court. In doing so, the amendment was beyond the jurisdiction of the federal Parliament because any change to the Court’s composition requires the consent of all provincial legislatures under Section 41 of the Constitution Act 1982.96 Thus, the judges ruled that the three Quebec seats had been implicitly entrenched in 1982.

94

See ch 5.4. Reference Re Supreme Court Act, ss 5 and 6 [2014] 1 SCR 433. 96 See section 6.3.1.1 above. 95

Procedural Forms of Participation 197 Notwithstanding the Canadian and Australian cases, many parliamentary federal systems do assign second chambers a role in judicial appointment procedures. Article 94(1)(2) of the German Basic Law states that the Parliament and Federal Council each elect half of the 16 members of the Federal Constitutional Court by a two-thirds majority. In practice, this requirement of qualified majorities has entailed that the two major parties, the conservative CDU/CSU and the social democratic SPD, typically negotiate behind closed doors about potential nominees and compromise on moderate candidates.97 A committee composed of representatives of the Länder governments collaborates with the Federal Council in the selection of the candidates. Although the committee cannot adopt binding resolutions, the Federal Council usually follows its recommendations. In Austria, the Constitutional Court consists of a president, a vice-president, 12 additional members and six substitute members, all appointed by the Austrian President upon the proposal of various institutions. The Federal Council is entitled to nominate three full members and one substitute member (Article 147(1–2) of the Constitution). As decision-making in Austria’s second chamber is dominated by party politics rather than territorial considerations, the influence of the Länder in the appointment procedure is relatively weak. Regarding the 12 members of its Constitutional Court, the Spanish Constitution similarly relies on a procedure of appointment by the head of state, ie in this case the King, upon proposals by various institutions (Article 159(1)).98 Like Austria, Spain has been faced with the problem that its second chamber, authorised to nominate four judges, does not properly represent subnational interests. A reform in 2007 attempted to create a procedural link between the autonomous communities and the Senate that was eventually challenged before the Constitutional Court. 6.I Spain 2008: Senate Nominees to the Constitutional Court In 2007, the Spanish Parliament amended the 1979 Organic Law on the Constitutional Court. According to its new Article 16(2), the legislatures of the autonomous communities would propose candidates and the Senate would then appoint ‘its’ judges to the Constitutional Court from this pool of potential nominees. Yet, there remained the requirement under Article 159(1) of the Constitution that the candidates of each chamber of the national Parliament have to be nominated by a three-fifths majority of its members. This qualified majority requirement has entailed in the past that candidates have been agreed upon by the two major parties, the conservative Partido Popular (PP) and the social democratic Partido Socialista Obrero Español (PSOE). The question was now what would happen if candidates proposed by the regional legislatures failed to obtain the required supermajority.

97 See W Heun, The Constitution of Germany: A Contextual Analysis (Oxford, Hart Publishing, 2010) 169. 98 Four are nominated by the first chamber of the national Parliament, four by the Senate, two by the General Council of the Judicial Power (Art 122(2–3) of the Constitution) and two by the federal Government.

198 Participation of Subnational Entities

The Constitutional Court ruled that in any case the Spanish Constitution stipulates that the Senate must have the final say.99 If candidates are unable to secure the support of a three-fifths majority, the Senate is free to select judges other than those proposed by the legislatures of the autonomous communities.

Apart from the appointment of the head of state and certain categories of judges, second chambers may also be involved in other selection processes, especially those regarding civil servants at the national level, including ambassadorial, cabinet and secondary executive branch appointments. This holds true, especially for federal systems with a presidential system of government. For example, the possibility to participate in the appointments of various types of federal public officials is foreseen by the constitutions of Mexico (Article 76(2)) and Brazil (Article 53(3–4)). In both cases, the idea of such participation has been taken from the United States. There, the above-mentioned Advisory and Consent Clause is crucial (Article II, section 2, clause 2).100 Yet, the US Constitution also allows, under certain circumstances, appointments without the approval of the Senate: ‘The President shall have power to fill up all vacancies that may happen during the recess of the Senate, by granting commissions which shall expire at the end of their next session’ (Article II, section 2, clause 3). To remain in effect, therefore, such recess appointments must be approved by the Senate by the end of the next session of Congress. The scope of such recess appointments was clarified by the Supreme Court in 2014. 6.J United States 2014: Presidential Appointments without ‘Advice and Consent of the Senate’ The National Labor Relations Board (NLRB) found that Noel Canning, a Pepsi distributor, had refused to implement a collective bargaining agreement with a labour union. In response, Canning challenged the order of the NLRB, arguing that the appointment of three of the five board members had been invalid. As a matter of fact, President Obama had invoked the Recess Appointments Clause and had appointed the three members in question when the Senate held merely ‘pro forma’ sessions, with ‘no business transacted’, every Tuesday and Friday between 17 December 2011 and 20 January 2012. The Supreme Court ruled that recess appointments under Article II, section 2, clause 3 are valid for any existing vacancy during any recess of sufficient length.101 Even if the framers of the Constitution had intended this provision as an exception to the regular procedure requiring ‘advice and consent of the Senate’ (Article II, section 2, clause 2), this exception was necessary for the President to ensure the continued functioning of the Government when the Senate was absent. As the Senate is equally unavailable during inter- and

99

STC 49/2008. N McCarty and R Razaghian, ‘Advice and Consent: Senate Responses to Executive Branch Nominations 1885–1996’ (1999) 43 American Journal of Political Science 1122. 101 National Labor Relations Board v Canning 573 US ___ (2014). 100 See

Procedural Forms of Participation 199

intra-session recesses (in the midst of a formal session), this differentiation was deemed irrelevant. The Court gave a similarly broad interpretation of the types of vacancy. Both vacancies that come into existence during a recess and those that initially occur before but continue to exist during a recess may be filled without Senate approval. It is important to note, however, that the judges established one quite restrictive criterion. They held that a three-day recess, as in the case concerned, is too short a period of time for the President to validly invoke the Recess Appointments Clause. Moreover, the Senate must be regarded as being in session when it declares that it is, unless the chamber obviously does not have the capacity to act. The Court thus ruled that President Obama’s appointments were unconstitutional and granted the Senate considerable leeway to block recess appointments by means of ‘pro forma’ sessions.

6.3.5 Other Functions In addition to the main functions just described, representative institutions of subnational entities—and notably second chambers—may participate in a number of other activities at the national level. The German Länder, for example, participate in decision-making processes related to European integration, as Article 23 of the Basic Law requires the consent of the Federal Council for the transfer of sovereign powers to the European Union. Furthermore, the ratification of European treaties requires a two-thirds majority of the votes within the Federal Council. In addition, the Länder take part in the European legislative process: the 1992 constitutional reforms provided for various degrees and forms of participation of the Federal Council, according to the relative competence in comparable domestic matters.102 Similar reforms in Austria in 1992 and 1994 granted the Federal Council the right to both be informed about the nominations of Austrian members within European institutions (Article 23c(5) of the Constitution) and to be consulted and involved in ‘all projects within the framework of the European Union’ (Article 23e). The involvement of representative institutions in the European Union legislative process derives from a more comprehensive right of participation in the national treaty-making power. In both the senate and the council model, second chambers take part in the ratification and implementation of international treaties. In Germany, the Federal Council takes part in the ratification of international treaties, whereby it has, as for ordinary federal legislation, a suspensive or absolute veto depending on the matter affected. Article 59(2) of the Basic Law states that federal law incorporates international treaties into the German legal system, and that such law requires the consent or participation of the bodies that are responsible in such a case for the enactment of federal legislation. In the United States, the Senate is generally responsible for the ratification of international treaties. According to Article II,

102 See J Woelk, ‘A Place at the Window: Regional Ministers in the Council’ in R Toniatti et al (eds), An Ever More Complex Union: The Regional Variable as a Missing Link in the EU Constitution? (BadenBaden, Nomos, 2004) 122f, and—for the situation after the reform of 2006—J Woelk, ‘Chapter 7: Germany’ in D Oliver and C Fusaro (eds), How Constitutions Change: A Comparative Study (Oxford, Hart Publishing, 2011) 161f.

200 Participation of Subnational Entities section 2, clause 2 of the Constitution, the President has the ‘power, by and with the advice and consent of the Senate, to make treaties, provided two thirds of the Senators present concur’. Further functions attributed to second chambers may regard participation in the declaration of a state of emergency, like in Russia (Article 88 of the Constitution) and in India (Article 356 of the Constitution). But while in the Russian case, the President merely has to immediately inform the Council of the Federation about the declaration of a state of emergency, the role of the Indian Council of States goes beyond just that. If the President of India proclaims direct national government control over a state whose constitutional machinery has failed, this later needs the confirmation of both chambers of Parliament (Article 356(3)). This parliamentary approval of so-called President’s rule within two months is, according to the Supreme Court, an essential mechanism of control so that no irreversible measures shall be taken before it.103 In Germany, in the case of a so-called legislative emergency (ie when a confidence vote is rejected, but the Parliament is not immediately dissolved), the Government may continue for up to six months in order to bring about the laws that were not accepted by Parliament if the Federal Council supports them (Article 81 of the Basic Law). However, this procedure has never been used. Second chambers also take part in the supervision over executive branch activities. In the United States, this kind of function is implied in Congress’s enumerated powers (Article I, section 8 of the Constitution). Supervision is conferred upon the Governmental Affairs Committee, which examines ‘the efficiency, economy, and effectiveness of all agencies and departments of the government’ (Senate Rule XXV(k)(2)). Article 173(1)(i) of the Swiss Constitution grants the Federal Assembly—the joint sitting of both chambers of the federal Parliament—the power of deciding over conflicts of jurisdiction between the highest federal authorities. In Germany, the Federal Council participates in supervision over the execution of federal legislation (Bundesaufsicht) under Article 84 of the Basic Law. Among other things, it must express its consent in order to allow the federal Government to send federal commissioners to a Land. Finally, under Article 37 of the Basic Law, the federal executive, with the consent of the Federal Council, may compel a Land to comply with its obligations under federal law (the so-called Bundeszwang). Finally, second chambers may participate in impeachment procedures. In the United States, the House of Representatives impeaches the President—and the Senate acts as a court for adjudicating charges of impeachment (Article I, section 3, clauses 6–7 and Article II, section 4 of the Constitution). The Chief Justice of the Supreme Court in this case presides over the Senate. Under Article 93 of the Russian Constitution, the President may be impeached by the Council of the Federation on the basis of charges of high treason or another grave crime, advanced by the Duma and confirmed by the Supreme Court of the Russian Federation.

103

SR Bommai v Union of India, 1994 AIR 1918. See box 5.M.

7 Financial Relations 7.1 FINANCIAL RELATIONS: SIGNIFICANCE, SCOPE AND REGULATION

7.1.1 Significance

I

N FEDERAL SYSTEMS, financial relations between levels of government are of crucial importance because only sufficient resources enable subnational entities to achieve ‘their policy objectives within their constitutionally assigned legislative and executive responsibilities’.1 In other words, it is the regulation of revenue and spending powers (sections 7.2 and 7.3) that permits subnational units to finance and carry out these responsibilities or not. Rules regarding financial relations are, therefore, decisive as to whether legislative and administrative autonomy2 may be put into practice. In order to ensure this, ‘finance should follow function’.3 In constitutional terms, this principle is epitomised, for example, by Article 47(2) of the Swiss Constitution, according to which the federal Government ‘shall leave the cantons with sufficient sources of finance and contribute towards ensuring that they have the financial resources required to fulfil their tasks’. Beyond that, the same Constitution provides, among the principles for the allocation and fulfilment of tasks, that the ‘collective body that benefits from a public service bears the costs thereof’ and the ‘collective body that bears the costs of a public service may decide on the nature of that service’ (Article 43a). We will see, however, that subnational entities are not usually able to finance all expenditures resulting from their legislative and administrative functions through their own revenues. In the case of national governments, on the other hand, their own revenues typically exceed their spending needs. This mismatch in the vertical dimension of financial relations, the so-called vertical fiscal gap, is one type of disparity in federal systems. Another one follows from horizontal differences between subnational entities in terms of fiscal capacity or expenditure needs. The correction, or at least attenuation, of these differences through equalisation is a third major topic of financial relations in federal systems (section 7.4). What we call in this chapter ‘financial relations’ has been referred to by other authors and with some conceptual variations as ‘distribution of finance’,4 ‘financial

1 RL Watts, Comparing Federal Systems, 3rd edn (Montreal, McGill–Queen’s University Press, 2008) 95. 2 See chs 5.2 and 5.3. 3 A Shah, ‘Introduction: Principles of Fiscal Federalism’ in A Shah and J Kincaid (eds), The Practice of Fiscal Federalism: Comparative Perspectives (Montreal, McGill–Queen’s University Press, 2007) 3ff, 9. 4 See, eg, Watts, Comparing Federal Systems (2008) 95.

202 Financial Relations arrangements’5 or ‘fiscal federalism’.6 This variety of terms has occasionally caused problems and misunderstandings due to divergent perceptions in different academic disciplines. Fiscal federalism, for instance, is a very common expression among political scientists and lawyers, but it is not always used with the specific meaning that economists attribute to it. The latter regard fiscal federalism as designating a particular ‘subfield of public finance’ that ‘lays out a general normative framework for the assignment of functions to different levels of governments and the appropriate fiscal instruments for carrying out these functions’.7 This decidedly prescriptive approach, with the overall aim of maximising the efficiency of public policies, contrasts with the focus of comparative legal studies regarding financial relations. Such studies instead concentrate on analysing the constitutional foundations of financial relations, with the above-mentioned revenue and spending powers, as well as equalisation mechanisms, being the essential areas of interest. Beyond the constitutional framework of financial relations, comparative legal studies also explore how these relations actually work in practice through the interplay of government levels. This is indeed indispensable because practice often differs considerably from what is written in constitutional texts, primarily as a result of change through judicial interpretation (eg United States) or intergovernmental agreements (eg Canada). 7.1.2 Scope: Local and Supranational Dimensions While research often focuses exclusively on the national and subnational levels of government, the scope of financial relations may include, according to a broader understanding, local governments as well.8 From a comparative perspective, the dynamics of this three-level constellation vary significantly. In the United States, for example, the federal Government frequently uses instruments like grants to local governments to bypass the state level and reinforce direct state–local relations.9 In Germany, on the other hand, the municipalities receive significant funds as a percentage of the Land share of national–subnational joint taxes. This percentage is to be determined by the legislation of the respective Land (Article 106(7) of the Basic Law). As to powers of raising one’s own revenues, local governments today often play a rather marginal role, even if, historically, that was not always the case. In Canada, for example, ‘municipalities levied taxes before any provincial tax was enacted’.10 In any event, local taxes typically require a specific basis in legislation, to

5

See, eg, PW Hogg, Constitutional Law of Canada, 5th edn (Scarborough, Carswell, 2000) 6. eg, R Boadway and A Shah (eds), Fiscal Federalism: Principles and Practice of Multiorder Governance (Cambridge, Cambridge University Press, 2009). 7 WE Oates, ‘An Essay on Fiscal Federalism’ (1999) 37 Journal of Economic Literature 1120, 1120f (emphasis added). The expression was first used in RA Musgrave, The Theory of Public Finance (New York, McGraw-Hill, 1959). See also G Färber, ‘Was ist Finanzföderalismus?’ in P Bußjäger (ed), Perspektiven des Finanzföderalismus in Österreich (Innsbruck, Studienverlag, 2013). 8 On the finances of local governments, see more generally ch 9.3.2.3. 9 See WF Fox, ‘‘The United States of America’ in Shah and Kincaid (eds), The Practice (2007) 364. 10 Hogg, Constitutional Law (2000) 6-2. 6 See,

Financial Relations 203 be adopted either by the national or respective subnational parliament, and are thus dependent on legislative action on the part of other government levels. Article 265 of the Indian Constitution mentions this need for enabling legislation explicitly, as ‘[n]o tax shall be levied or collected except by authority of law’. In the South African case, local taxation is entrenched in the Constitution itself, in the long Section 229, and this was recently affirmed by the jurisprudence of the country’s Constitutional Court. 7.A South Africa 2013: Constitutional Basis for Local Taxing Powers Section 40(1) of the South African Constitution describes government as consisting of the ‘national, provincial and local spheres of government which are distinctive, interdependent and interrelated’. In practice, the strength of South African municipalities as autonomous spheres of government depends critically on their ability to raise adequate revenues, which was precisely the subject of a Constitutional Court judgment in 2013.11 For a long time, certain groups of rural landowners had been exempted from the duty to pay property taxes, a crucial source of income for local governments. When the Bergrivier Municipality first obliged them in 2001 to pay property taxes, landowning farmers in the rich wheat farming district of the Swartland refused to do so by referring to the fact that the municipality was already performing well in providing public services. In response to this challenge, the Constitutional Court reinforced the taxing powers of local governments and held that Bergrivier had validly imposed the property taxes.12 The judges stated explicitly that: ‘[a] municipality’s authority to impose rates and levies is derived from Section 229 of the Constitution. The purpose of a municipality’s revenueraising powers is to finance a municipality’s performance of its constitutional and statutory objects and duties as set out in Sections 152(1) and 153 of the Constitution. These include the provision of services to communities in a sustainable manner, promoting social and economic development and providing for the basic needs of the community’ (para 40).

While local government extends the scope of financial relations below the national– subnational dualism, there is sometimes also a relevant level above it. This concerns, of course, particularly federal systems that are Member States of the European Union, even more so for the members of the eurozone in the wake of the global financial crisis. In response to this crisis, European leaders added a new third paragraph to Article 136 of the Treaty on the Functioning of the European Union (TFEU) that authorised the creation of a stability mechanism, which then became a reality with the establishment of a new intergovernmental organisation in 2012. This so-called European Stability Mechanism (ESM) provided a permanent eurozone architecture for bailouts on the basis of conditionality.13 The explicit aim was to achieve financial stability through conditions meant to ensure fiscal discipline.

11 See J de Visser and N Steytler, ‘Confronting the State of Local Government: The 2013 Constitutional Court Decisions’ (2016) 6 Constitutional Court Review 1, 9ff. 12 Liebenberg NO v Bergrivier Municipality 2013 (5) SA 246 (CC). 13 See D Kelemen, ‘Law, Fiscal Federalism and Austerity’ (2015) 22 Indiana Journal of Global Legal Studies 379, 388.

204 Financial Relations As the creation of the ESM could not entirely calm financial markets, the then-EU 27, except for the United Kingdom and the Czech Republic, signed the Treaty on Stability, Coordination and Governance in the Economic and Monetary Union (TSCG) in 2012, commonly known as the European Fiscal Compact. The most important feature of this agreement outside the scope of EU law is the ‘golden rule’ that the budgets of the Member States must be balanced or in surplus and that this rule needs to be entrenched in national constitutions (or equivalent norms as to rigidity). As non-compliance with this duty entails for states, in case of a fiscal emergency, the exclusion from financial assistance and possible sanctions by the Court of Justice of the European Union, the effect of the TSCG amounted to a judicial enforcement of austerity.14 From the perspective of subnational entities, the result was a ‘double upward shifting of authority’:15 from them to the respective national government and from that to the European level. The constitutional changes towards fiscal discipline, agreed to by the national governments, had the effect of limiting the leeway of subnational entities concerning fiscal policies. Beyond that, the TSCG raised, more generally, significant questions concerning the ‘discretion left to member states to make fundamental decisions about taxation and spending’.16 In any event, these very recent developments certainly demonstrate that the scope of financial relations may also include a relevant level above the national–subnational dualism.17 Besides the Outright Monetary Transactions (OMT) programme of the European Central Bank,18 both the ESM and the TSCG proved to be very controversial and provoked legal disputes before Germany’s Federal Constitutional Court. 7.B Germany 2014: The Place of the European Level in Financial Relations In reaction to the European crisis response, legal proceedings were lodged against many key elements of the emergency measures. Most notably, these were Germany’s assent to the above-mentioned adoption of Article 136(3) TFEU and the subsequent approval of the ESM Treaty as well as of the TSCG. The judges ruled that none of these actions had violated the Constitution.19 They emphasised that the principle of democracy, as constituted by the right to vote (Article 38(1) of the Basic Law) and the commitment to a democratic state and popular sovereignty (Article 20(1–2)) forms the basis of the budget autonomy of the German Parliament

14 V Ferreres Comella, ‘Amending the National Constitutions to Save the Euro: Is it the Right Strategy?’ (2013) 48 Texas International Law Journal 223. 15 C Colino and E del Pino, ‘Spanish Federalism in Times of Crisis’ in P Peterson and D Nadler (eds), Fiscal Crisis and the Centralization of Political Power (Washington, Brookings Institution Press, 2014) 161. 16 S Peers, ‘The Stability Treaty: Permanent Austerity or Gesture Politics?’ (2012) 8 European Constitutional Law Review 404, 404. 17 The response to the sovereign debt crisis in the eurozone has been compared to the American response to a similar crisis in the 1780s that was one among many reasons for the creation of the federal Constitution in 1787. See A Loubert, ‘Sovereign Debt Threatens the Union: The Genesis of a Federation’ (2012) 8 European Constitutional Law Review 442. 18 134 BVerfGE 366 (Outright Monetary Transactions); BVerfG 21 June 2016, 2 BvR 2728/13. 19 135 BVerfGE 317 (European Stability Mechanism).

Financial Relations 205

(Bundestag) and requires it to remain the place where autonomous decisions on revenues and expenditures are made. Regarding the TSCG, the Court was very concise and made clear that this treaty by no means granted European institutions powers affecting the budget autonomy of the Bundestag, nor did it coerce Germany into making a permanent and irreversible commitment concerning its fiscal and economic policy. With respect to the ESM, the judges held that, despite the liabilities assumed, budgetary autonomy would be sufficiently safeguarded. First, Article 8(5) of the ESM Treaty, as well as joint declarations of the ESM members and an additional unilateral declaration by Germany on 27 September 2012, ruled out an interpretation according to which the payment obligation created would be unlimited. Second, the admission of new members to the ESM requires a unanimous decision (Articles 44 and 5(6)) and may thus be vetoed by the German representatives so that the chain of legitimation between the Bundestag and the ESM is not interrupted. In order to maintain this veto and thus the constitutionally required budget autonomy, the judges stipulated that budgetary legislation must ensure that Germany meets capital calls fully and in time because it would otherwise face a suspension of its voting rights within the ESM (Article 4(8)).

7.1.3 Regulation: Legal Sources and (A)Symmetrical Design Regardless of whether regulatory frameworks governing financial relations merely focus on two government levels or more, they rely on quite different legal sources.20 Some federal systems are characterised by detailed ‘financial constitutions’21 and thus firmly entrench many aspects of these relations. Cases in point are Switzerland (Article 126–35 of the Swiss Constitution) and Germany (Articles 104a–15 of the Basic Law). In the latter case, the Constitution even goes as far as to determine the allocation of revenue from single taxes, such as that on beer, which it reserves to the Länder (Article 106(2)(5)). If a financial relations matter is not found in the Constitution but addressed by ordinary law, then the federal legislator is clearly predominant. The distribution of competences expressly stipulates that the federal level has concurrent power to legislate on all taxes whose revenue accrues to it wholly or in part (Article 105(2)). When the Länder or municipalities receive a share of the revenue, however, a federal law needs the consent of the Federal Council (Bundesrat) (Article 105(3)). In contrast to the more or less comprehensive financial constitutions in Switzerland and Germany, others are (almost) silent on this issue. A good example is the United States, where finance was not regarded in 1787, with few exceptions like customs tariffs (Article I, section 10, clause 2),22 as an issue to be regulated in the

20 For an overview, see S Parolari, ‘From a Formal to a Substantial Approach: Sources of Law and Fiscal Federalism’ in F Palermo and A Valdesalici, Comparing Fiscal Federalism (Leiden, Brill–Nijhoff, forthcoming). 21 This term is used, in particular, in German-speaking countries and in these cases is derived from the German word Finanzverfassung. See, eg, G Färber, ‘Fiscal Equalization in Germany: Facts, Conflicts and Perspectives’ in G Pola (ed), Principles and Practices of Fiscal Autonomy: Experiences, Debates and Prospects (Farnham, Ashgate, 2015) 118. 22 See section 7.2.2 below.

206 Financial Relations federal Constitution. While this has not changed until today, also due to the cumbersome constitutional amendment procedures under Article V of the Constitution,23 financial relations matters are regulated, even in great detail, by state constitutions. As a general rule, the latter thereby give ‘priority to taxpayers over service recipients’ and ‘make it more difficult for states and localities to raise funds to finance public services’.24 In Austria, financial relations matters are, in the main, regulated through ordinary legislation and not in the federal or, as in the US case, subnational constitutions. The framers of the codified Austrian Constitution of 1920 had failed to reach a consensus on financial matters and only stated that the ‘competences of the federation and the Länder in the field of taxation will be prescribed in a special [emphasis added] federal constitutional law’ (Article 13 of the Constitution). This was eventually done by the separate Financial Constitutional Law of 1948, which forms the basis for the crucially important Financial Equalisation Law, a sunset law that is adopted by the federal Parliament for four years and only subject to a suspensive veto by the second chamber. Importantly, the Constitutional Law of 1948 only provides general principles for tax allocation and abstract types of taxes, ie exclusive or various kinds of shared taxes (§ 7), but leaves the concrete assignment of taxes to these types and thus to the different government levels to the Financial Equalisation Law. Similar to Austria, the more recent federal systems of Belgium, Italy and Spain rely on their national constitutions only regarding the basic framework for financial relations, which is then specified and complemented by regulations from various other sources of law. The Belgian Constitution addresses financial relations in its Articles 170–81, but Articles 175–77 explicitly delegate much of financial regulation to a special law, ie the 1989 Special Law on the Financing of the Communities and Regions, and in the case of the German-speaking community even an ordinary law. Also, the Spanish Constitution contains only a few general provisions on financial relations, with Articles 156–58 being the most relevant. Having exclusive power over ‘[g]eneral financial affairs and state debt’ (Article 149(1)(14a) of the Constitution), more detailed regulations are made in an organic law adopted by the national Parliament (Article 157(3)), ie the Organic Law on the Financing of the Autonomous Communities (Ley Orgánica de Financiación de las Comunidades Autónomas, LOFCA). Other aspects of financial relations are determined in the bilaterally negotiated statutes of the autonomous communities, which also have the rank of an organic law, and still others, like the ceded taxes, in ordinary legislation.25 As to the relationship between the two relevant organic laws, the Constitutional Court has repeatedly ruled that the autonomy statutes have to be interpreted in

23

See ch 6.3.1.1. R Briffault, ‘The Disfavored Constitution: State Fiscal Limits and State Constitutional Law’ (2003) 34 Rutgers Law Journal 907, 909. 25 On the main characteristics of the Spanish financing system, see A Herrero Alcalde, ‘The Spanish Way of Fiscal Federalism’ in SA Lütgenau (ed), Fiscal Federalism and Fiscal Decentralization in Europe: Comparative Case Studies on Spain, Austria, the United Kingdom and Italy (Innsbruck, Studien Verlag, 2014). 24

Financial Relations 207 conformity with the LOFCA, and not vice versa.26 However, autonomy statutes, especially those reformed in the 2000s, have often prompted amendments to the LOFCA in order to align the regulations of these two legal sources. Beyond this framework set out in special legislation, it is up to ordinary laws to then introduce concrete levies, with the ‘primary power to raise taxes’ being vested exclusively in the national parliament (Article 133(1) of the Constitution). Yet, this does not preclude the imposition of taxes by other levels of government. Autonomous communities may do so, ‘albeit in accordance with the constitution and the laws’ (Articles 133(2)) and 157(1) in fact count autonomous communities’ own taxes among their financial resources). While the Italian Constitution limits itself, like that of Spain, to outlining a few general principles (Article 119), the further regulation of financial relations is, compared to the above-mentioned organic laws, much less entrenched in legal terms. Whereas the bilaterally negotiated autonomy statutes are the main source with regard to the country’s special regions, legislation adopted by the national Parliament is of central importance for ordinary regions. Interestingly, Law 42/2009 delegated to the national Government the power to enact by-laws for the creation of a new financial regime.27 These by-laws, nine of which have been adopted so far, are not self-executing, but often require further executive rules or leave the regulation of certain crucial aspects to a later stage. The fiscal Constitution of South Africa contains, unlike that of Italy, more than only a few provisions and in fact dedicates a long chapter to the matter of ‘Finance’ (Sections 213–230A), but it still does not achieve much more in terms of guarantees for subnational financial autonomy. This is because the South African Constitution itself provides, for many key issues, nothing more than basic principles and delegates the regulation beyond that to national laws. These pieces of legislation, like the annual Division of Revenue Act concerning revenue-sharing28 (on the basis of Section 214) and the 1999 Public Finance Management Act regarding provincial budgeting (on the basis of Section 215), epitomise this partial deconstitutionalisation and the dominance of the federal legislator that came along with it.29 In the United Kingdom, the regulation of financial relations is, in the absence of a written constitution, even entirely deconstitutionalised. Consequently, these relations are essentially defined by the devolution acts for Scotland, Wales and Northern Ireland, on the one hand, and by mere constitutional conventions, on the other.30 The latter holds true, in particular, for a core element of the UK financial arrangement, ie the so-called Barnett formula, whose principle is that any increase or

26 See F Pérez Royo, ‘Cuestiones sobre el poder tributario de las CCAA: la relación entre las disposiciones sobre financiación en los Estatutos y la Ley Orgánica del artículo 157.3’ in AM Pita Grandal (ed), La financiación autonómica en los Estatutos de Autonomía (Madrid, Marcial Pons, 2008). 27 On this law, see F Scuto, ‘The Italian Parliament Paves the Way to “Fiscal Federalism”’ (2010) 2 Perspectives on Federalism 67. 28 See section 7.2.2 below. 29 See P de Vos and W Freedman (eds), South African Constitutional Law in Context (Oxford, Oxford University Press, 2014) 305ff. 30 On constitutional conventions as, in a strict sense, non-legal rules, see P Leyland, The Constitution of the United Kingdom: A Contextual Analysis (Oxford, Hart Publishing, 2012) 32ff.

208 Financial Relations reduction in expenditure in England automatically entails a proportionate increase or reduction in funding for the three devolved administrations. As this formula, which determines the block grants for the devolved administrations, is only based on a conventional rule and the fruit of political negotiations, it cannot be challenged in court and could be renegotiated at any time, if the UK Government decided to do so. On the other hand, it has been argued that the referendums held on devolution, including its financial arrangements, have provided the Barnett formula with particular political legitimacy and protection against its unilateral amendment.31 In sum, this broad variety of financial relations being regulated in very different sources of law also reflects the advantages and disadvantages of a more rigid or more flexible approach. Entrenchment in detailed financial constitutions obviously provides stronger legal guarantees and thus more predictability for all levels of government. By contrast, (partial) deconstitutionalisation of the matter makes the regulation of financial relations more adaptable to changing circumstances. For such adaptation to be profound, federal systems with comprehensive financial constitutions must resort to the often difficult process of formal constitutional amendments. In some cases, such amendments have altered the initial financial arrangement significantly. A case in point is Germany with its repeated reforms of the Basic Law’s Title X ‘Finance’ in 1955, 1969 and 2009. Where financial relations are much less entrenched in constitutional law and not even in some sort of special legislation, similarly far-reaching changes have been effected primarily through two avenues, ie constitutional jurisprudence and intergovernmental agreements. The first holds true, especially for the United States,32 while the latter is characteristic of Canada. Although political accords between the different levels of government play a key role in many federal systems, Canada’s tax rental agreements (1941–62) and tax collection agreements (1962–present) appear to be outstanding in terms of their overall impact on financial relations.33 These broad patterns of adapting financial relations through formal constitutional amendment, constitutional jurisprudence and intergovernmental agreements do not mean, of course, that federal countries in practice rely exclusively on one of these three channels. As we will see in the following sections, financial relations in Germany, for instance, have been shaped beyond the above-mentioned amendments by both intergovernmental agreements and the case law of the Federal Constitutional Court, even to an extent that the latter has been criticised for its judicial activism.34 Apart from rigidity and flexibility, another categorisation to differentiate legal frameworks from financial relations is their asymmetrical or symmetrical design, ie whether a single subnational entity or several of them are treated differently or not. Such differentiation in terms of financial powers is not the case with certain opt-outs

31 See S Eden, ‘United Kingdom’ in G Bizioli and C Sacchetto (eds), Tax Aspects of Fiscal Federalism: A Comparative Analysis (Amsterdam, IBFD, 2011) 551. 32 KW Dam, ‘The American Fiscal Constitution’ (1977) 44 The University of Chicago Law Review 271, 273. 33 See section 7.2.2 below. 34 See DP Kommers, The Constitutional Jurisprudence of the Federal Republic of Germany, 2nd edn (Durham, Duke University Press, 1997) 56.

Financial Relations 209 on the part of Quebec, for example, regarding the just-mentioned tax collection agreements in order to have its own tax collection structures. Other provinces have the same power and might opt out as well. Such cases thus need to be distinguished from differential treatment by law, which is typical especially for more recent devolutionary systems in Europe. This legal differentiation may follow directly from the (financial) constitution, from special laws that are negotiated bilaterally and, in some cases, from ordinary legislation. Asymmetries in Spain are, for example, based on the first two sources of law, ie constitutional provisions and organic laws. The fact that Navarra and the Basque Country, more precisely the latter’s three historical provinces (Álava, Biscay and Gipuzkoa) have a distinct foral regime in terms of finances (régimen foral), which deviates from the common regime (régimen común) for the remaining autonomous communities, is anchored directly in the Spanish Constitution. The latter’s Additional Provision no 1 expressly protects the historic foral rights (fueros) of these territories, which also amounts to special treatment regarding financial affairs. This provision and two organic laws adopted in 1981 and 1982, respectively, entail that the raising and collection of all taxes are decentralised under the foral regime. In return, there is only the obligation to make a payment to the national government for its services provided in the Basque Country and Navarra according to a formula that is renegotiated bilaterally every five years and adopted in an organic law passed by the Spanish Parliament. The exceptional impact, also from a comparative perspective, of this asymmetry is evident from the fact that the autonomous communities under the common regime mostly rely on so-called ceded taxes,35 over which they have limited powers, while those under the foral regime are practically selfsufficient through their own taxes.36 Moreover, the Basque Country and Navarra are not part of Spain’s system of equalisation.37 Beyond the constitutionally guaranteed foral regime, the statutes of the autonomous communities are, as bilaterally negotiated organic laws, a second source of asymmetries because their regulations regarding financial matters may vary. Catalonia’s Statute of 2006, for example, entrusted manifold responsibilities to a specific bilateral body, ie a State-Generalitat Joint Economic and Fiscal Affairs Commission (Article 210). This contrasted with the multilateral approach of the above-mentioned LOFCA, which ascribes key importance to the Fiscal and Financial Policy Council (Consejo de Política Fiscal y Financiera, CPFF): a formally advisory, but practically decisive body that includes the national and regional ministers of finance and is in fact dominated by the national Government. The Constitutional Court deemed the above-mentioned bilateral commission consistent with the national Constitution as long as it neither excludes nor restricts the multilateral mechanisms, above all the CPFF.38 To some degree, asymmetries exist in several other countries as well. In Italy, the five special regions also have special financial regimes that are negotiated bilaterally and

35

See section 7.2.2 below. Herrero Alcalde, ‘The Spanish Way’ (2014) 29. 37 See section 7.4 below. 38 STC 31/2010, FJ 135. The reasoning behind this ruling is based on the earlier judgment STC 13/2007. 36

210 Financial Relations entrenched in their statutes of autonomy.39 This distinguishes them, concerning both procedure and substance, from the regime applicable to the ordinary regions.40 Among the most notable contents of a statute is the determination of which national tax revenues referable to the region are shared and in what percentage (25 to 90 per cent). Moreover, the special regions do not participate in the general equalisation system established on the basis of Article 119(4) of the Italian Constitution. Instead, bilateral agreements with the national Government determine the contribution of each single special region. In the United Kingdom, there is certainly less asymmetry, not least because the above-mentioned Barnett formula was intentionally foreseen to account for the bulk of finances of the devolved administrations.41 Yet, the Scottish Government has been treated differently in some respects. Primary examples are the power to raise or lower the income tax rate by 10 per cent (Section 25 of the Scotland Act 2012), as well as, in the wake of the independence referendum in 2014, several provisions of the Scotland Act 2016: the further extension of autonomy over income taxation (Sections 13–15), the right to receive half of the VAT raised on Scottish territory (Section 16) and new devolved taxes (Sections 17–19). None of these provisions are applicable to Wales or Northern Ireland. 7.2 REVENUE POWERS

7.2.1 Types of Subnational Entities’ Own Revenues The allocation of revenue powers surely concerns first and foremost the problem of tax assignment.42 But beyond this most lucrative source, subnational entities’ own revenues may also stem from fees (or charges) for the provision of public services, profits from public enterprises, royalties related to natural resources and public borrowing. The latter two sources, in particular, sometimes have a profound impact on financial relations and are at the same time often extremely controversial. Revenues from natural resources43 are important and contentious for several reasons. First, they account in some federal countries like Nigeria, Iraq and Russia for a very high share of the combined revenue of all government levels. Second, natural resources are typically distributed unevenly across a country, which leads to the critical question of whether the proceeds from them should primarily accrue to the subnational entities where they are located or to the national government. This may give rise to protracted intergovernmental conflicts, as demonstrated, for instance, by the

39 See A Valdesalici, ‘Features and Trajectories of Fiscal Federalism in Italy’ in Lütgenau (ed), Fiscal Federalism (2014) 91ff. 40 Italian Constitutional Court No 74/2009. 41 A Trench, ‘The UK’s Devolution Finance Debates’ in Lütgenau (n 25) 325. 42 See C McLure, Tax Assignment in Federal Countries (Canberra, Australian National University, 1983); R Bird, ‘Tax Assignment Revisited’ in JG Head and RE Krever (eds), The 21st Century: A Volume in Memory of Richard Musgrave (The Hague, Kluwer Law International, 2009). 43 See, specifically on this, G Brosio, ‘The Assignment of Revenue from Natural Resources’ in E Ahmad and G Brosio (eds), Handbook of Fiscal Federalism (Cheltenham, Edward Elgar, 2006); G Anderson (ed), Oil and Gas in Federal Systems (Oxford, Oxford University Press, 2012).

Revenue Powers 211 legal dispute between Malaysia’s federal Government and the state of Terengganu.44 As illustrated by the case of Alberta within Canada, this question is relevant not only for the distribution of revenue powers but also for other subfields of financial relations such as the equalisation of uneven fiscal capacities.45 Third, disputes over profits from natural resources have amplified conflicts in diverse societies involving subnational entities with separatist potential. The Kurdistan region in Iraq and Scotland in the United Kingdom are cases in point. Public borrowing has become, especially in the recent past in light of the global financial crisis, a highly controversial revenue source. As mentioned above, the European Fiscal Compact obliged its signatories to introduce constitutional provisions regarding fiscal discipline. This also involved restrictions on borrowing for all government levels. Article 119(6) of the Italian Constitution, for instance, now stipulates with regard to the regions (and municipalities, provinces and metropolitan cities) that they ‘may resort to indebtedness only as a means of funding investments. State guarantees on loans contracted for this purpose are not admissible.’ Other federal countries in Europe have adopted limitations with similar wording. In any event, subnational borrowing had been restricted in some cases long before the current crisis. The reasoning behind this was mostly to ensure the national government’s control over the credit rating of the entire country, which is also positively or negatively affected by the creditworthiness of the subnational entities. Thus, Article 293 of the Indian Constitution reserves all foreign borrowing to the national Government. Schedule II, Part I, Item 7 of the Nigerian Constitution does the same even for loans within the country, even if this borrowing restriction for the states has hardly been effective in practice.46 Interestingly, in the US case it is not the national constitution but many state constitutions that contain (self-)limitations such as cumbersome borrowing procedures or a strict ceiling for loans.47 7.2.2 Taxing Powers Although certain other sources of subnational entities’ own revenues may thus be quite relevant, taxes are without doubt the mainstay of subnational receipts. This section will therefore focus on taxing powers. Even if the focus is narrowed down from all types of subnational entities’ own revenues to taxes, there remains, of course, an extremely broad variety ranging from taxes on ‘bads’ like motor fuels to those on ‘sins’ like alcohol and gambling.48 In terms of tax yield, however, there are basically three sources that are clearly the most important throughout federal systems, ie direct taxes on personal and corporate income and indirect taxes on sales or added value (VAT). Concerning all taxes, whether broad-based or narrow-based, revenue powers may be perceived as being composed of different elements: on the

44

Petroliam Nasional Bhd v State Government of Terengganu [2004] 1 MLJ 8. See section 7.4 below. 46 See AH Ekpo, ‘Federal Republic of Nigeria’ in Shah and Kincaid (n 3) 210. 47 See GA Tarr, Understanding State Constitutions (Princeton, Princeton University Press, 2000) 109ff. 48 For an overview, see Shah, ‘Introduction’ (2007) 22–23. 45

212 Financial Relations one hand, powers regarding the tax itself, ie the decision to impose it and to define its base, rate and, if applicable, exemptions and deductions; on the other hand, powers concerning tax collection.49 Broadly speaking, the allocation of taxing powers may follow three different approaches, ie the assignment of exclusive, concurrent and shared powers on taxes. As mentioned above, rather little attention was paid to financial matters at the Philadelphia Convention of 1787. But among the few references in the US Constitution to this issue, there is the express reservation of customs tariffs on imports and exports to the national Government (Article I, section 10, clause 2), which is the first example of an exclusive taxing power. This means that the power to yield revenue from the tax base is reserved to one level of government to the exclusion of others. Similar to the US Constitution, Article 122 of Canada’s Constitution Act 1867 stated explicitly that customs tariffs belong to the competences of the national Government. The 1901 Australian Constitution stipulated the same, while additionally requiring the transfer of a part of customs revenues to the states for 10 years and the imposition of uniform tariffs within two years (Articles 86–88 and 90). In hindsight, it is by no means a coincidence that these older federations had such a focus on the assignment of customs tariffs exclusively to the national government. First, their constitutions simply aimed to ensure that the newly established federal level would have plenary power over a type of revenue source that was at that time extremely profitable and soon accounted for the bulk of federal receipts.50 Second, a national regime of customs tariffs was a precondition for the establishment of an internal market within the boundaries of the newly formed federation. This motive of economic integration was equally strong in Canada and in the United States, in the latter case especially against the background of protectionist interstate trade obstacles under the Articles of Confederation.51 In light of today’s economic integration at the supranational level, the (federal) power to levy customs tariffs has decreased in importance, particularly in relation to the three above-mentioned broad-based taxes of our times. In contrast to older constitutional texts, some of those adopted after World War II, like the constitutions of Germany and India in 1949, have explicitly reserved certain exclusive taxation powers not only to their national governments but also to the subnational entities. However, this has not necessarily entailed much greater tax autonomy in the long run. In the German case, this was the result of radical changes to the original system of tax assignment soon after 1949. Originally, personal and corporate income taxes were exclusive Länder taxes, while that on turnover was an exclusive federal tax. An amendment to the financial Constitution in 1955 stipulated that the first two taxes be shared and another reform in 1969 did the same for the latter tax. As a result, all three of these broad-based taxes are today joint taxes under Article 106(3) of the Basic Law. Like the original German Basic Law of 1949, the Indian Constitution of the same year allocated taxes primarily following the principle of exclusivity. In accordance

49

Similarly, A Shah, ‘Comparative Conclusions on Fiscal Federalism’ in Shah and Kincaid (n 3) 380. See Hogg (n 5) 6-1. 51 See ch 3.2.2. 50

Revenue Powers 213 with Article 246 of the Constitution, taxing powers are included in the union list (List I) or the state list (List II) of Schedule 7. In some cases, however, exclusive powers of one government level are so closely intertwined with the equally exclusive powers of the other level that overlaps regarding taxation have become serious problems in practice.52 Primary examples are state taxes on agricultural income and federal ones on all other income. Another uncertainty had prevailed for some time regarding the relationship between exclusive taxing powers guaranteed in the state list and the residual clause that explicitly reserves to the national government ‘any tax not mentioned in either of those lists’ (Schedule 7, List I, Item 97). This relationship specifically was the issue of a seminal Supreme Court judgment in 1980. 7.C India 1981: Exclusive State Taxes and the Residual Clause The state of Haryana imposed a tax on transport companies running between Delhi and the state of Jammu and Kashmir. It did so even if the carriers used the national highway and only passed through Haryana without picking up or dropping off passengers or goods in this state. As the construction and maintenance of national highways was an exclusive competence of the national Government (Schedule 7, List I, Item 23), the companies claimed in Court that the tax was invalid. The Supreme Court disagreed with this view and held that the state of Haryana had acted in accordance with its power to raise ‘[t]axes on goods and passengers carried by road or on inland waterways’ (Schedule 7, List II, Item 56).53 For the exercise of this exclusive competence to be constitutional, it would be required that there be a sufficient nexus between the state concerned and the object of taxation. According to the Court, a state incurs significant costs even with regard to national highways, not directly for their construction and maintenance, but for the facilitation of transport, ie for lighting, traffic control or motorway service areas. These expenditures would imply the sufficient nexus required. As to the above-mentioned federal residuary power (Schedule 7, List I, Item 97), the judges demanded a restrictive interpretation. This would follow from the wording of this clause itself, which explicitly refers only to taxes not enumerated in List II and III. But the Court also invoked, more generally, the federal character of the national Constitution. In its view, the ‘residuary power cannot be so expansively interpreted as to whittle down the power of the state legislature. That might affect and jeopardize the very federal principle. The federal nature of the constitution demands that an interpretation which would … destroy or belittle state autonomy must be rejected’ (p 327).

Compared to India’s assignment of exclusive taxing powers to both government levels, it is, from a subnational perspective, much more alarming when powers of this kind are allotted only to the federal level, typically as a result of a prohibition of double taxation to the benefit of the national government. Such bans are characteristic, in particular, for several more recently established federal systems. In Belgium, the communities and regions have been allowed to raise their own taxes

52 53

See G Rao, ‘Republic of India’ in Shah and Kincaid (n 3)161. International Tourist Corporation v State of Haryana, 1981 AIR 774.

214 Financial Relations since the 1989 Special Law on the Financing of the Communities and Regions.54 Subsequent constitutional jurisprudence reinforced the fact that subnational entities may thereby, according to the principle of non bis in idem, only create revenue from tax bases that are not occupied by the national Government.55 The prohibition of double taxation in the Spanish LOFCA is even more far-reaching, as the autonomous communities are excluded from tax bases occupied not only by the national Government (Article 6(2)), but also from those covered by local authorities (Article 6(3)). Early case law of the Constitutional Court had still interpreted this ban in a less strict way.56 More recently, however, legislation of the autonomous communities that aimed to circumvent the restriction of double taxation, primarily by giving regional taxes a different (environmentalist) purpose (‘green taxes’), has often been invalidated.57 In both the Spanish and Belgian cases, bans on double taxation have reserved to the respective national government significant exclusive tax room, while subnational entities’ own levies have been reduced to marginal narrow-based taxes with scarce tax yields, often also coupled with high costs of tax administration. In contrast to exclusivity, concurrent taxing powers entail that taxes may be raised, in principle, by both the national and subnational governments at the same time so that they compete for the same revenue source. Notwithstanding the abovementioned exclusive federal power over customs tariffs, the United States and Canada are prime examples of concurrency, even though this approach prevails in the two countries for different reasons. In the Canadian case, it followed not only from the constitutional text but in part also from constitutional jurisprudence. Whereas the provinces were assigned, in addition to certain specific revenue sources,58 the competence concerning ‘direct taxation within the province’ (Section 92(2) of the Constitution Act 1867), the federal Government was granted a general power over ‘the raising of money by any mode or system of taxation’ (Section 91(3)). Thus, the wording clearly implied federal–provincial concurrency regarding direct taxation, above all that on income, which did not even exist in 1867,59 but is today of crucial importance for both levels of government. 7.D Canada 1887: The Broad Scope of Provincial Direct Taxes In search of revenue sources to cover their increasing expenditure requirements, the provinces became more and more interested in raising direct taxes: besides income and inheritance taxes, also corporation and sales taxes. In an early landmark case, the Judicial Committee of the Privy Council (JCPC) declared a Quebec tax on corporations to be in

54 M Verdonck and K Deschouwer, ‘Patterns and Principles of Fiscal Federalism in Belgium’ (2003) 13 Regional & Federal Studies 91, 97ff. 55 Belgian Constitutional Court Nos 04/1998 and 100/2003. 56 See, eg, STC 37/1987. 57 See, eg, STC 289/2000. 58 Examples are ‘licenses’ (s 92(9) Constitution Act 1867) and ‘taxation of natural resources’ (s 92A (4)). 59 See R Knopff and A Sayers, ‘Canada’ in J Kincaid and GA Tarr (eds), Constitutional Origins, Structure, and Change in Federal Countries (Montreal, McGill–Queen’s University Press, 2005) 124.

Revenue Powers 215

line with the Constitution and introduced as a guideline John Stuart Mill’s classic distinction between direct and indirect taxes from his Principles of Political Economy.60 Thus, the former are taxes imposed on the person intended to pay for it, while the latter is demanded from a person expected to get reimbursed at the cost of another. According to this ruling’s narrow interpretation, a tax could only be regarded as indirect if there was a general tendency, not merely the possibility, to shift the tax burden to someone else. As the provinces began in the following decades to introduce more and more taxes of various designs, the distinction of taxes that were direct and thus constitutional became increasingly difficult, even more so during the Great Depression of the 1930s. At that time, the decreasing yield from traditional provincial taxes sparked interest in raising retail sales taxes. New Brunswick, for instance, imposed a tax on anyone purchasing tobacco or importing it into the province, either in person or through an agent. In order to avoid a situation whereby the retailer shifted the cost of the tax to the purchaser, which would have made it a clearly unconstitutional indirect tax, the provincial law named the retailer as the tax collector and the purchaser as the person liable for paying the tax. The Supreme Court—back then not yet the court of last resort—held that this construction would make the tax direct only if the purchaser was the actual consumer, but indirect if it was the consumer’s agent. The JCPC, even in both these constellations, regarded New Brunswick’s tax, in effect a retail sales tax and thus one of the most classical indirect taxes, as direct and therefore consistent with the Constitution. This formalistic approach has since then made it easy for the provinces to circumvent their limitation to direct taxes under Section 92(2) of the Constitution Act 1867.

Also, in the United States, both government levels are bound to compete regarding the same sources of revenue. Similar to the Canadian case, Article I, section 8, clause 1 of the US Constitution vests the federal Government with a general power of taxation. The latter has the right to ‘lay and collect taxes, duties, imposts and excises’, subject to the requirement that ‘all duties, imposts and excises shall be uniform throughout the United States’. Taxation of the states, by contrast, is not based on explicit authorisation but on their residual powers. There are no express constraints on it from the constitutional text itself but rather from the Supreme Court’s dormant Commerce Clause doctrine.61 After all, this doctrine bans the states from passing any laws, including tax legislation, which would discriminate against interstate commerce.62 This makes any state taxation that does not demonstrate a sufficient link between an economic activity and the state territory through the presence of a seller or corporation unconstitutional. Obviously, this is a considerable obstacle, for example, for state taxation of online sales or the income of multinational corporations. In the early days of the United States, the above-mentioned general taxation power of the federal Government was also subject by the Supreme Court to certain limits. First, the Court ruled in 1870 that Congress could not impose a tax upon the salary 60 Bank of Toronto v Lambe [1887] 12 AC 575 (JCPC). JS Mill, Principles of Political Economy [1848] (Indianapolis, Hackett Publishing Company, 2004). 61 See box 5.G. 62 See W Hellerstein et al, ‘Commerce Clause Restraints on State Taxation after Jefferson Lines’ (1995) 51 Tax Law Review 47, 50.

216 Financial Relations of a state judicial officer, as this would contradict the doctrine of dual federalism defined in that very judgment.63 This prohibition was abolished, however, in a ruling nearly seven decades later.64 A second, much more incisive, limit on the federal Government’s plenary taxation power was derived in a ruling in 1895 from the constitutional obligation to apportion direct taxes among the states in proportion to population (Article I, section 2, clause 3).65 7.E United States 1895: Restrictive Conditions for the Federal Income Tax? Income taxation had been introduced with the Revenue Act of 1861 in order to raise funds for the American Civil War. Tax legislation adopted in 1894 then imposed for five years a levy on any gains, profits and incomes exceeding certain thresholds. In response, the Farmers’ Loan & Trust Company notified its shareholders that it would pay the tax and provide to the Treasury the names of all its clients in view of their liability to be taxed under the new legislation. A minority stockholder from Massachusetts, Charles Pollock, brought an action against the company and, after losing in the lower courts, appealed to the US Supreme Court. In a surprising judgment, the Court ruled in Pollock’s favour and qualified the income taxes under the new legislation on interest, dividends and rents as direct taxes so that they violated, unapportioned as they were, the above-mentioned Article I, section 2, clause 3 of the US Constitution.66 According to the judges, ‘the founders anticipated that the expenditures of the states, their counties, cities, and towns, would chiefly be met by direct taxation on accumulated property, while they expected that those of the federal government would be, for the most part, met by indirect taxes’ (p 620–21).

Even if constitutional jurisprudence thus invalidated the federal income tax, it was later superseded by the Sixteenth Amendment in 1913, which expressly allowed Congress ‘to lay and collect taxes on incomes, from whatever source derived, without apportionment among the several states’. The Supreme Court followed suit and upheld unapportioned income taxes laid down in the Revenue Act of 1913.67 The repercussions of this turnaround were enormous, as the Sixteenth Amendment laid the constitutional basis for federal fiscal predominance in the short and long run. First, Congress made extensive use of its newly acquired power during World War I by broadening the coverage of the income tax and by increasing the very low pre-war rates of less than 6 per cent up to 65 per cent.68 Since then, receipts from personal and corporate income taxes have been the mainstay of national

63

Collector v Day 78 US (11 Wall) 113 (1870). Graves v New York ex rel O’Keefe 306 US 466 (1939). 65 On the rationale of this apportionment as a concession to the slave states, see B Ackerman, ‘Taxation and the Constitution’ (1999) 99 Columbia Law Review 1. 66 Pollock v Farmers’ Loan & Trust Co 158 US 601 (1895). 67 Brushaber v Union Pacific Railroad Co 240 US 1 (1916). 68 See BA Wallin, ‘Forces behind Centralization and Decentralization in the United States’ in Commission on Fiscal Imbalance (ed), Texts Submitted for the International Symposium on Fiscal Imbalance (Quebec, Commission on Fiscal Imbalance, 2002) 26. 64

Revenue Powers 217 government revenues, even more so, as there are, unlike in most countries, no federal indirect taxes on sales or added value. The crucial problem of concurrency lies in the fact that it entails, in contrast to exclusive taxing powers, vertical tax competition between the national and subnational governments. This issue does not exist, of course, if one level decides, like the US Government with the sales tax, not to occupy a tax base and, instead, leaves it to the other level. But as government levels in most cases of concurrency indeed do compete for the same tax bases, there have to be some conflict rules or, if a cooperative spirit prevails, mechanisms of tax harmonisation. As to concurrent taxing powers on income in the United States, for example, the ultimate conflict rule is provided by the Supremacy Clause (Article VI, clause 2 of the Constitution), which gives precedence to national (tax) legislation.69 Canada, by contrast, has largely addressed the issue of concurrency by relying on tax harmonisation through intergovernmental agreements. While the ground for vertical tax competition had already been prepared by the Constitution Act 1867, the problem only became salient when the federal Government started, during World War I, to raise taxes on personal and corporate income. After mediation through a report of the Rowell–Sirois Commission had failed, the first solution was a series of tax rental agreements (1941–62) under which the provinces would not relinquish their constitutional power to raise income taxes, but ‘rent’ them to the national Government in return for federal transfers. However, this centralisation turned out to be only temporary, as the provinces, especially Quebec and Ontario, were successful in their struggle for a more cooperative solution that also allowed for opt-outs, and not only for Quebec.70 According to the tax collection agreements (1962– present), the federal Government defines the tax base and collects the revenue at its own expense, but it leaves the determination of tax rates and possible exemptions and deductions to the provinces. Within a uniform regulatory framework, there are bilateral federal–provincial agreements regarding the taxes on personal income (except with Quebec) and corporate income (except with Quebec and Alberta). Such tax harmonisation has been achieved to a much lesser degree, however, concerning the various sales taxes. In this regard, only two partial agreements exist,71 so that uncoordinated concurrency to some extent continues to exist. Another instructive example of tax harmonisation in addressing the problem of vertical tax competition is Switzerland. The particularly important personal and corporate income taxes may be levied concurrently by the federal and cantonal governments and, through surcharges within the limits of cantonal legislation, by local governments as well (Article 128 of the Swiss Constitution). In order to coordinate taxing powers in such a situation of potential overlaps and disputes, Switzerland has applied two instruments. One is a regime of tax harmonisation that is, 69

See ch 5.1.2. DM Brown, ‘Fiscal Federalism: Searching for Balance’ in H Bakvis and G Skogstad (eds), Canadian Federalism: Performance, Effectiveness and Legitimacy, 2nd edn (Oxford, Oxford University Press, 2008). 71 One agreement is with New Brunswick, Nova Scotia and Newfoundland and Labrador, the other one with Quebec. The remaining provinces raise different types of sales taxes or, in the case of Alberta, no such tax at all. See R Boadway, ‘Canada’ in Shah and Kincaid (n 3) 99–124. 70 See

218 Financial Relations however, unlike harmonisation in Canada, outlined in the Swiss Constitution itself (Article 129). After specification through ordinary law, such a regime has been binding since 2001, and it introduced, concerning direct taxation, uniform definitions of tax bases and lists of deductions, while leaving to the cantons the rates and the amounts of deductions.72 A second tool is that the maximum rates for the federal tax are entrenched explicitly in the Swiss Constitution, currently at 11.5 per cent for personal income and 8.5 per cent for corporate income (Article 128(1)), and that the federal Government, in setting its rates, ‘shall take account of the burden of direct taxation imposed by the cantons and communes’ (Article 128(3)). Together with another constitutional guarantee, that of allocating a minimum of 17 per cent of the revenue from the federal tax to the canton where it was collected (Article 128(4)), this has provided the cantons, compared to other subnational entities, with a position of well-ensured tax autonomy. Besides exclusive and concurrent competences, a third and very broad category is that of shared taxing powers. As we shall see in the following analysis, what is actually shared between government levels varies significantly across federal systems. While some of them merely share, in different proportions, the revenue from specific taxes, others also confer certain autonomous regulatory powers to subnational entities. In the first case, that of pure revenue-sharing, such powers are absent, as government levels simply have the (sometimes constitutionally entrenched) right to receive a certain portion of revenue from specified taxes. That is why several scholars subsume revenue-sharing under the category of unconditional federal transfers.73 In contrast to unconditional transfers in a strict sense, however, there is the significant advantage that an increase of the subnational portion typically does not require an explicit decision by the national government but follows directly from a growing tax yield (if it does grow). Our understanding of shared taxes is broader and includes, beyond such pure revenue-sharing, also a second set of cases in which this simple distribution of funds is accompanied by subnational regulatory powers regarding certain aspects of the taxes concerned. As we will see, these may be exercised either autonomously (eg Spain) or jointly within the federal legislative process (eg Germany). From a comparative perspective, shared taxes are more prevalent among more recent federal systems which typically appear to distribute funds based on a combination of various factors and not merely following the territoriality principle. The latter refers to where the revenue was generated and thus tends to reinforce economic disparities between subnational entities. An example of an older federation with extensive revenue-sharing is Austria. In fact, the above-mentioned Financial Equalisation Law,74 adopted every four years, focuses mainly on such distribution, as ‘the main part of revenue of all levels is generated within the so-called joint (shared) taxation system’.75 Upon a mandate

72 See B Dafflon, ‘Swiss Fiscal Federalism: New Roads after the Reform of the Constitution’ in Pola (ed), Principles and Practices (2015) 100f. 73 See Watts (n 1) 104f; Shah (n 3) 28f. 74 See section 7.1.3 above. 75 P Bußjäger, ‘Reforms on Fiscal Federalism in Austria’ in G Robbers (ed), Reforming Federalism: Foreign Experiences for a Reform in Germany (Frankfurt am Main, Peter Lang, 2005) 60.

Revenue Powers 219 established by § 7 of the Financial Constitutional Law of 1948, this law determines, regarding each tax, the revenue portion for the national Government, the Länder and municipalities. Before the federal Government submits the draft of the Financial Equalisation Law to Parliament, tripartite negotiations usually take place between the Federal Minister of Finance and representatives of the Länder and municipalities. Thereby, the federal Government must attempt to reach an agreement and is loosely bound to this accord when the law is finally enacted.76 However, the latter do not participate in these talks on an equal footing, as in reality they ‘really have no legal alternative but to accept the determination of fiscal relations by the federal government’.77 This is entirely different in Germany, where the Länder are in a clearly better position. First, federal laws regarding joint taxes require the consent of the Federal Council (Bundesrat) (Article 105(3) of the Basic Law). Second, the Constitution itself guarantees them and the federal Government equal shares of the revenues from personal and corporate income taxes (Article 106(3)). Regarding the turnover tax, the same provision stipulates that the respective shares are to be determined periodically by a federal law requiring again the consent of the Federal Council.78 It is evident that this constitutionally entrenched framework leaves the Länder and the federal Government no other choice but to cooperate. The significance of this cooperation is epitomised by the fact that the portions from the joint taxes constitute 66 per cent of the total federal tax revenue and even 85 per cent of that of the Länder.79 In Spain, more than half of the financial resources of the autonomous communities stem from ceded taxes (tributos cedidos), which represent another form of shared taxes. With respect to these levies, which were originally assigned completely to the national Government, since 1997 the latter has ceded to the autonomous communities a portion or all of the revenues generated on their territory, as well as certain regulatory powers. Most levies, with the corporate income tax as a notable exception, are today ceded taxes with subnational entities’ regulatory powers being, as a rule, more extensive regarding taxes whose revenue accrues to them fully.80 But they are also highly comprehensive with respect to the personal income tax, whose revenues are shared at 50 per cent. Most often, the autonomous communities have, regarding ceded taxes, the power to raise the rates, which is difficult in the case of high national base rates, and to determine exemptions and deductions with only a few limits. Yet, even fully ceded taxes still seem to be distinct from autonomous communities’ own taxes. Unlike the latter, ceded taxes may only be regulated subject to more or less stringent limits imposed by the national Government, and

76

VfSlg 12784/1991. ibid 61. 78 Before revenues from the personal income tax and turnover tax are shared between the Länder and the federal Government, certain portions are transferred to the local governments: currently 15% and 2%, respectively. 79 See LP Feld and J von Hagen, ‘Federal Republic of Germany’ in Shah and Kincaid (n 3) 140ff. 80 See JA Gimeno Ullastres, ‘Tax Assignment and Regional Co-Responsibility in Spain’ in N Bosch and JM Durán (eds), Fiscal Federalism and Political Decentralization: Lessons from Spain, Germany and Canada (Cheltenham, Edward Elgar, 2008) 82. 77

220 Financial Relations a utonomous communities do not have the power to decide whether to raise them or not. If a region fails to use its regulatory competences regarding a ceded tax, subsidiary national regulations enter into force. For these reasons, most scholars do not regard ceded taxes as subnational entities’ own taxes.81 Apart from the three above-mentioned and quite diverse European cases, revenuesharing is typical, in particular, of many federal systems outside Europe. That subnational entities’ own levies are limited to narrow-based taxes and thus dependent on the sharing of revenues from the broad-based federal taxes is indeed a quite common phenomenon. The provinces of South Africa, for instance, are even explicitly prohibited by Article 228(1a) of the Constitution to impose taxes on any of the major sources, namely ‘income tax, value-added tax, general sales tax, rates on property or customs duties’. Revenue-sharing is then a kind of compensation for this lack of provincial tax autonomy, as most federal receipts have to be transferred to a National Revenue Fund (Article 213). A national law must then safeguard that each province receives an ‘equitable share’ (Article 214(1b)). That the adoption of this law, the annual Division of Revenue Act, needs the prior consultation of the provinces (Article 214(2)) is,82 of course, far from the above-mentioned consent required from the German Länder. Article 162(1) of the Nigerian Constitution quite similarly stipulates that a fund, ie the Federation Account, be established and that the distribution of shared revenues be regulated by a national law, in this case even without consulting the states.83 It is up to the national Parliament to determine, upon a proposal of the President of Nigeria, the formula for revenue allocation from the Federation Account, and it must thereby only ‘take into account, the allocation principles especially those of population, equality of states, internal revenue generation, land mass, terrain as well as population density’ (Article 162(2)). A proviso to this provision puts some constraints on the national Parliament, as any formula must ensure that no less than 13 per cent of the revenue accruing to the Federation Account from a natural resource is payable to the state from which the resource is derived. Unsurprisingly for an oil-rich country like Nigeria, this sparked some controversy and finally litigation before the Supreme Court. 7.F Nigeria 2002: Revenue-sharing and Natural Resources The principle of derivation concerning the sharing of revenues from natural resources obviously implied that the issue of the conditions under which these resources were located within a state’s boundaries became crucial. In this regard, the Attorney-General of Nigeria issued a Writ of Summons to the Supreme Court against the 36 states seeking a declaration

81 See, eg, J Pérez Royo, Curso de derecho constitucional (Madrid, Marcial Pons, 2010) 839. Some observers, however, do qualify them because of the differentiated subnational regulatory powers as subnational entities’ own taxes. See J López-Laborda et al, ‘Kingdom of Spain’ in Shah and Kincaid (n 3) 297f. 82 Premier, Western Cape v President of the Republic of South Africa 1999 (3) SA 657. 83 See RT Suberu, ‘Federalism in Africa: The Nigerian Experience in Comparative Perspective’ (2009) 8 Ethnopolitics 67, 80f.

Revenue Powers 221

from the judges on the federal Government’s definition of the seaward boundaries of the country’s maritime states on offshore oil revenue derivation. The Court agreed with the Attorney-General as plaintiff and determined the low water mark of the land surface to be a state’s seaward boundary.84 On the other hand, however, it also ruled that a number of practices of the national Government concerning the Federation Account had been unconstitutional. These comprised, for example, the exclusion of natural gas from the scope of the proviso to Section 162(2) and thus from the principle of derivation, the inclusion of certain expenditures in the calculation of the Federation Account (eg funding of the judiciary or servicing of external debts) and unilateral allocation of 1 per cent of the revenue in the account to the Federal Capital Territory.

Compared to the aforementioned cases, Brazil stands out through its very comprehensive regulation of revenue-sharing in constitutional law. With its characteristic penchant for detail, the 1988 Constitution does not merely state that the receipts from certain federal taxes are allocated to the states and municipalities (Article 157–58). It then goes on to declare that ‘forty-nine percent of the proceeds from the collection of taxes on income and earnings of any nature and on industrialized products’ is to be turned over to the states and municipalities, even stipulating in detail how this 49 per cent is to be distributed (Article 159). While one part of taxing powers concerns the elements of a tax, ie its base, rate and possible exemptions and deductions, and may be exclusive, concurrent or shared, another part is related to tax collection. As a general rule, collection powers follow the authority to regulate the aforementioned crucial aspects of the tax itself. This link is very explicit in Article I, section 8, clause 1 of the US Constitution, according to which the federal Government may ‘lay and [emphasis added] collect taxes’. In the specific above-mentioned cases of shared taxes, the competence of collecting revenue is typically vested in the national government, which then distributes these funds. As to the legal sources of this power, it sometimes derives directly from the constitution. For instance, Article 128(4) of the Swiss Constitution states explicitly that direct taxes are ‘assessed and collected by the cantons’. The Constitution of India introduces a more complex differentiated regime, as some taxes are levied by the national Government but collected by the states (Article 268), while most of them are both levied and collected by the national Government (Article 270). Other federal systems leave this issue rather to intergovernmental agreements than to regulation by constitutional law. A case in point is, quite obviously, Canada with its above-mentioned tax collection agreements. If a province chooses to conclude such a bilateral accord based on a multilateral framework, the task of tax collection, often perceived as burdensome anyway, is assumed at its own cost by the federal Government, more precisely by the Canada Revenue Agency as an autonomous agency under joint federal–provincial oversight. As mentioned above, tax collection is, at least regarding taxes on personal and corporate income, largely centralised,

84

Attorney General of the Federation v Attorney General of Abia State SC 28/2001 of 5 April 2002.

222 Financial Relations with the exception of Quebec in the first case and of Quebec, Ontario and Alberta in the latter. Autonomous tax collection, however, also has its downsides. First, subnational entities, especially less affluent and smaller ones, are sometimes faced with bigger problems regarding administrative efficiency and thus unable to maximise revenues. Several South African provinces are cases in point.85 Second, tax collection may also incur significant costs for subnational entities.86 Therefore, it does not come as a surprise that some of them, again particularly less affluent and smaller ones, prefer to leave this responsibility to the national Government, even without a constitutional obligation to do so. Yet, this motivation may be outweighed by significant arguments in favour of autonomous tax collection, especially if this competence is viewed as an expression of quasi-sovereignty in the context of an ethno-culturally diverse country. This is epitomised, in the Canadian case, by the fact that the tax collection ambitions of Quebec, in particular, clearly exceed those of any other province. Other cases in point are Bosnia and Herzegovina, where the tax collection power of the two entities follows from Article VIII(3) of the Constitution, as well as Spain. In the latter case, tax collection is only decentralised in the Basque Country and Navarra as a result of their special financial regime.87 With regard to the remaining autonomous communities, taxes have traditionally been collected by the national Government through the State Agency of Tax Administration (Agencia Estatal de Administración Tributaria). Interestingly, however, it was precisely Catalonia that foresaw in its 2006 Statute the establishment of its own Tax Agency (agencia tributaria) intended to collect not only scarce regional taxes but also the ceded taxes levied on the territory (Article 204). Again, subnational entities’ own institutions for tax collection were thus regarded as important vehicles to affirm a distinct identity and create a semblance of quasi-sovereignty. 7.2.3 A Trend Towards Tax Centralisation? In 1986, it was claimed that ‘[t]he period from 1950 to 1980 saw most federations and unions undergo a pattern of decentralization of both government revenues and expenditures’.88 It is not so clear whether the ensuing period of the last three to four decades has been characterised overall, regarding revenues, by a similar trend. There seems to be little doubt, at least, that the ‘control of revenues has been much more concentrated in the federal governments of more recent emergent federations than in mature ones’.89 This is also quite evident from the scope of taxing powers, as outlined in the preceding section, of mature federal systems such as the United States, Canada and Switzerland, on the one hand, and more recent ones like South

85

See B Khumalo and R Mokate, ‘Republic of South Africa’ in Shah and Kincaid (n 3) 273. See Shah (n 3) 20f. 87 See section 7.1.3 above. 88 R Bird, Federal Finance in Comparative Perspective (Toronto, Canadian Tax Foundation, 1986) 17f. 89 See Watts (n 1) 97. 86

Revenue Powers 223 Africa, Spain, Belgium or Italy, on the other. Of course, exceptions to this general rule exist in both groups, with mature Australia, for instance, being very centralised and young Bosnia and Herzegovina being extremely decentralised. As a matter of fact, the latter’s national Government is largely dependent on financial transfers from the two entities. Article VIII(3) of Bosnia and Herzegovina’s Constitution stipulates that, apart from its rather marginal own receipts, such as a value-added tax introduced in 2006, it is the duty of the Federation of Bosnia and Herzegovina to provide two-thirds and of the Republika Srpska to provide one-third of the revenues required for the national budget. Another important observation regarding the trend towards tax centralisation concerns reinforcement in the context of specific events, which is then often maintained. In other words, national predominance is typically justified as inevitable in response to emergencies like war or economic crises90 and not reversed in the aftermath of such events. Glaring examples are income taxes in both Canada and Australia during World War II. The initially temporary tax rental agreements with Canadian provinces were left in place until 1962.91 In Australia, the federal income tax, originally intended to be raised until one year after the War, has been kept ever since. As a matter of fact, the federal income tax legislation enacted in 1942 created in this field a monopoly for the national Government, as an immediate challenge before the High Court failed and a subsequent one, albeit partially successful, left too many political difficulties regarding the reintroduction of a subnational tax.92 To some extent, the relative centralisation of taxation in more recent federal systems is understandable through their typically devolutionary pattern, according to which they had a completely different point of departure. While in mature federations all taxing powers had typically been in the hands of the sovereign entities concluding the original federal compact, the above-mentioned recent federal systems started from the fiscal regime of unitary states. Even though subnational tax autonomy is without doubt less extensive in more recent devolutionary systems, they have still undergone, from the perspective of their unitary legacy, a process of decentralisation, albeit often incremental and centrally controlled. This is best illustrated by long-term developments in Belgium and Spain. Both countries were initially characterised by lump-sum transfers to the subnational entities: regarding the Belgian regions and communities until 1980 and 1988, respectively, and in Spain until the reform of the LOFCA in 1996. In the Spanish case, this move, over time, towards more tax autonomy, even if still comparatively limited, occurred in correspondence with a constitutionalised principle of financial autonomy (Article 156(1) of the Spanish Constitution). The Constitutional Court played a significant role on this journey.

90 For an overview of different reasons for tax centralisation in various countries, see Shah, ‘Comparative Conclusions’ (2007) 378f. 91 Hogg (n 5) 6-3. 92 South Australia v Commonwealth (1942) 65 CLR 373 (First Uniform Tax case); Victoria v Commonwealth (1957) 99 CLR 575 (Second Uniform Tax case). See C Saunders, ‘The Uniform Income Tax Cases’ in HP Lee and G Winterton (eds), Australian Constitutional Landmarks (Cambridge, Cambridge University Press, 2003).

224 Financial Relations

7.G Spain 1987–2007: Financial Autonomy and Its Limits Article 156(1) of the Spanish Constitution specifies the general principle of autonomy enshrined in Article 293 and stipulates that the autonomous communities ‘shall enjoy financial autonomy for the development and exercise of their powers, in conformity with the principles of coordination with the state treasury and solidarity among all Spaniards’. This means, according to the early case law of the Constitutional Court, that the autonomous communities must have sufficient means for the exercise of their powers and the right to dispose of such resources without unreasonable conditions.94 While the Constitutional Court has largely followed the LOFCA and thus political definitions in its interpretation of financial autonomy, the judges have introduced certain outer limits of autonomy. A first limit can be seen in the principle of coordination of autonomy with the financial powers of the national Government, as outlined in the LOFCA.95 A second one concerns the principle of solidarity, which is explicitly mentioned in the above-mentioned Article 156(1) itself and also epitomised by the specific reference to the guarantee of ‘a minimum level of basic public services throughout Spanish territory’ (Article 158(1)). Moreover, the Constitutional Court has ruled that the autonomous communities are obliged, for the sake of solidarity and despite their financial autonomy, to economically contribute to efforts against horizontal imbalances between them.96 More recently, the judges emphasised, however, that financial autonomy has to be understood in a broad sense as also including, beyond the spending dimension, revenue autonomy. Even if this new understanding was first developed in a seminal judgment with regard to local governments,97 it has since been applied to the autonomous communities as well.

The apparent trend towards tax centralisation may also be looked at against the background of economic theory, according to which ‘the case for decentralizing taxing powers is not as compelling as is that for decentralizing public service delivery’, mainly because extended subnational taxing powers entail a ‘trade-off between increased accountability and increased economic costs’.98 Moreover, a certain degree of centralisation is essential in order to guarantee the national government sufficient funds for two tasks that it is commonly expected to fulfil. These are redistribution of financial resources across the country, albeit to very different degrees,99 and macroeconomic stabilisation, which requires enough leverage to undertake costly countercyclical measures.100 On the other hand, a major argument in favour of extensive

93

See ch 2.3.2. STC 13/1992. 95 STC 179/1987. 96 STC 135/1992 and STC 247/2007. See section 7.4 below. 97 STC 289/2000. 98 Shah (n 3) 20f. 99 See section 7.4 below. 100 See W Oates, ‘Assignment of Responsibilities and Fiscal Federalism’ in R Blindenbacher and A Koller (eds), Federalism in a Changing World: Learning from Each Other (Montreal, McGill–Queen’s University Press, 2003) 40. 94

Revenue Powers 225 tax autonomy is that only the raising of subnational governments’ own revenues will ensure their accountability and fiscal discipline. As the claim goes, the more that the politicians who enjoy the electoral pleasure of spending tax money must first experience the electoral pain of extracting it from the taxpayers, the more their fiscal behaviour is likely to be responsible and subject to voter accountability.101

In some federal systems, subnational entities have even found a highly centralised fiscal regime quite convenient. For Spain, it has been claimed, for example, that certain autonomous communities enjoy acting as the ‘fairy godmother’, which offers public services, without assuming the role of the ‘wicked stepmother’,102 which raises taxes to finance them. A rather obvious impact of tax centralisation is that it constrains horizontal tax competition. Such competition between subnational entities presupposes, after all, a certain degree of autonomy. It is emblematic, therefore, that the fear of excessive rivalry between subnational entities was one of the main reasons for the fact that the drafters of the South African Constitution largely centralised taxing powers. In their view, ‘allowing provinces to choose applicable tax rates and tax bases could result in tax competition … thus reinforcing economic disparities’.103 From a historical perspective, tax competition has been much more common as a rationale for financial relations in older federations like the United States and Switzerland, and some classical definitions even regarded it as an essential characteristic of fiscal federalism. The latter’s purpose would be precisely ‘to permit different groups living in various states to express different preferences for public services; and this, inevitably, leads to differences in the levels of taxation and public services’.104 Supposedly mobile citizens were expected to ‘vote with their feet’105 and thus force subnational entities to compete for a mix of public services and low tax burden in order to attract residents. Proponents of such competition at the subnational level endorse a similar argument about offering the best conditions at minimum costs with respect to businesses. Economists also concede, however, that this competition may create inefficiencies regarding the allocation of resources as taxed units (or owners of taxed items) seek out jurisdictions where they can obtain relatively favourable tax treatments. High excise taxes in one jurisdiction, for example, may lead purchasers to bear unproductive travel costs in order to purchase the taxed items in jurisdictions with lower tax rates.106

Interestingly, the US Supreme Court has also, in some instances, pointed to the potential downsides of tax competition. In a landmark ruling of 1937, the judges noticed 101 J

426.

Kincaid, ‘Comparative Observations’ in Kincaid and Tarr (eds), Constitutional Origins (2005)

102 V Ruiz Almendral, ‘Autonomous Communities Taking Advantage of the Mechanism to Ensure the Neutrality of VAT’ (2003) 14 International VAT Monitor (IBFD) 373, 379. 103 Khumalo and Mokate, ‘Republic of South Africa’ (2007) 265. 104 Musgrave, The Theory (1959) 179. 105 CM Tiebout, ‘A Pure Theory of Local Expenditures’ (1956) 65 Journal of Political Economy 416. See also ch 11.6. 106 Oates, ‘An Essay’ (1999) 1120.

226 Financial Relations that ‘states and local governments are at times reluctant to increase so heavily the burden of taxation to be borne by their residents for fear of placing themselves in a position of economic disadvantage as compared with neighbours or competitors’.107 In any event, comparative evidence seems to suggest that horizontal tax competition is not, in reality, a major issue. Some observers even go as far as to claim that ‘the competitive federalism model is a theoretical construct advanced by the fiscal federalism literature and not yet practised anywhere in its pure form’.108 The closest to this pure form is certainly Switzerland, where tax competition is exceptionally strong, not only between cantons but also between municipalities. Uneven tax levels have been normal, with some cantons being able to offer better public services at a lower fiscal burden and income tax rates influencing, especially for high-income earners and self-employed people, their choice of residence. In 2013, for instance, tax rates varied regarding personal income from 3.0 per cent (Zug) to 13.7 per cent (Neuchâtel) and concerning corporate income from 5.0 per cent (Nidwalden) to 18.1 per cent (Geneva).109 While competition between the Swiss cantons has been kept in check, similar dynamics in Brazil took a rather destructive turn and even resulted in what has been called ‘tax wars’ that eventually provoked, as a countermeasure, fiscal recentralisation. In Spain, competition between the autonomous communities has not exactly amounted to ‘tax wars’ but has still sparked some controversy‚ even more so in light of the country’s asymmetrical fiscal regime. This specific situation has led to some sort of one-sided competition, as autonomous communities under the common regime could not even react when the Basque Country decided in 1996 to lower its corporate income tax.110 Apart from these singular cases, however, horizontal tax competition in federal countries is not a serious issue overall,111 not least due to the high degree of tax centralisation in many of them. From a practical point of view, such competition between subnational entities raises certain problems concerning the clear delineation of jurisdictions that resemble those of the above-mentioned vertical tax competition as a result of concurrent powers.112 Precisely such problems caused a dispute between the Swiss cantons of Geneva and Vaud that, in the end, had to be settled by the Federal Court.

7.H Switzerland 1999: Taxing Residents of Another Canton? In 1998, the canton of Geneva informed people who were working on its territory but living in the neighbouring canton of Vaud that they would receive forms to pay income tax to the canton of their workplace. This was the culmination of a longer controversy related to the fact that these commuters were using Geneva’s infrastructure and services but were

107 Helvering v Davis 301 US 619, 644 (1937). In the case cited, the Court used this as an argument for a system whereby old-age pensions were covered by the federal power to spend for the general welfare. See sections 7.3.2 below and 11.1. 108 Shah (n 3) 6. 109 See Dafflon, ‘Swiss Fiscal Federalism’ (2015) 102. 110 See López-Laborda et al, ‘Kingdom of Spain’ (2007) 304. 111 Shah (n 49) 383. 112 See section 7.2.2 above.

Revenue Powers 227

not contributing with their tax payments to financing the related expenditures. Vaud’s minister of finance advised those who had received this notice to disregard the official communication of the canton of Geneva and referred the matter to the Federal Court. The ensuing case was a rare example of this Court being called on to resolve a conflict between two cantons that had proved to be impossible through instruments of intergovernmental relations. In their ruling, the judges held that Geneva’s attempts to raise taxes from commuters were unconstitutional and violated Vaud’s fiscal sovereignty.113 They thereby referred to the jurisprudence that the Court had developed over the years on the basis of former Article 46 of the Swiss Constitution. This provision stipulated that people were subject, concerning their civil law relations, to the legislation of the canton of their residence and prohibited inter-cantonal double taxation. Regarding both principles, former Article 46 referred to specifying ordinary legislation of the federal Parliament. Thus, the judges found that, as long as this legal framework remained unchanged, single taxation by one canton and the fiscal domicile determined by one’s place of residence would be the only applicable principles, even in the face of external effects.

After its total revision in 1999, the new Swiss Constitution now states explicitly that ‘[i]ntercantonal double taxation is prohibited. The Confederation shall take the measures required’ (Article 127(3)). Complex problems with the delineation of tax jurisdiction are common in other countries as well and call for efficient solutions. A very recent case of great importance was decided by the US Supreme Court in 2015. 7.I United States 2015: Full Credit for Income Tax Paid to Another State? Similar to most other US states, Maryland taxes the income that residents earn both within the state and beyond it. More precisely, the income tax of Maryland is composed of a state income tax and a county income tax, which are both collected by Maryland’s comptroller of the treasury. For income earned in other states, taxpayers could get a credit concerning the state tax, but not regarding the county tax. A married couple resident in Maryland and with income generated in other states claimed a tax credit in 2006 concerning both the state and county tax, with the latter being rejected by the comptroller. The US Supreme Court held that the failure to allow this tax credit was clearly unconstitutional as a result of the long-established dormant Commerce Clause doctrine.114 In particular, the judges did not agree with the comptroller’s argument that the state’s tax legislation was constitutional because there was no intention to discriminate against interstate commerce. They made it clear that the Commerce Clause ‘regulates effects, not motives, and it does not require courts to inquire into voters’ or legislators’ reasons for enacting a law that has a discriminatory effect’. As to the determination of such effects, the Court referred to the ‘internal consistency test’, according to which the tax system in

113

BGE 125 I 458. Comptroller of the Treasury of Maryland v Wynne 575 US ___ (2015). On the dormant Commerce Clause, see box 5.G. 114

228 Financial Relations

question is hypothetically applied to every other state as well. It is then asked whether such a situation disfavours interstate commerce in comparison with intrastate commerce. As Maryland imposes a tax on the income of residents in other states and the income of non-residents earned in Maryland, the test results in interstate income being subject to double taxation by both states concerned. The Court declared that the tax legislation in question ‘has the same economic effect as a state tariff, the quintessential evil targeted by the dormant Commerce Clause’.

7.3 SPENDING POWER

7.3.1 Spending Power and Conditional Grants To a large extent, the distribution of competences regarding spending follows the distribution of legislative and administrative powers:115 the more far-reaching the autonomy of subnational entities to make and implement laws, the greater the expenditures that they make. In countries (such as Switzerland and Germany) following the model of administrative federalism, where subnational implementation of national legislation is the rule, the (limited) scope of subnational legislative autonomy obviously does not provide a reliable indication of expenditure needs. The central administrative role of the cantons and Länder indeed entails very significant spending. Quite evidently, one must also take into account that the responsibility for some subject matters is more cost-intensive than for others. This holds true, in particular, for areas like education or social welfare and health, which require not only regulation but also the extensive provision of services. Thus, they clearly account in modern welfare states, whether federal or not, for the bulk of public expenditures. From a theoretical perspective, economists uphold the principle of fiscal equivalence, according to which the distribution of functions in federal systems leverages economies of scale regarding the provision of public goods and avoids external effects in the exercise of competences.116 While the combined scope of legislative and administrative competences thus determines, to some degree, the spending powers of subnational entities, their autonomy may be effectively eroded by expenditures on the part of the federal government beyond its own jurisdictions. To be sure, a general spending power that a government may also use outside of its own legislative and administrative competences is often constitutionally admissible for subnational entities as well. In practice, however, the trend outlined in the preceding section towards tax centralisation117 usually makes extra-jurisdictional expenditures a domain of the national government. From the perspective of subnational entities, such federal spending in their areas of competence often poses a significant dilemma. On the one hand, the

115

See chs 5.2 and 5.3. See M Olson, ‘The Principle of ‘Fiscal Equivalence’: The Division of Responsibilities among Different Levels of Government’ (1969) 59 American Economic Review 479, 482. 117 See section 7.2.3 above. 116

Spending Power 229 above-mentioned trend towards centralised taxation entails that they may welcome this financial assistance in order to bolster their often scarce resources. To compensate for this scarcity of their own revenues, the setting of national minimum standards for services is frequently one among many possible aims declared for the use of federal grants.118 On the other hand, such grants often come with significant strings attached in the form of stricter conditions and may thus constitute an intrusion into subnational autonomy that thwarts the distribution of legislative and administrative powers. They are not always block grants that are only earmarked for a broad area such as social welfare or even entirely unconditional general-purpose transfers for budget support that strengthen subnational spending autonomy (not their revenue autonomy, of course). Whatever the aims that they are supposed to realise, financial transfers must, in any case, be properly designed and systematic in order to be effective: ‘Manna-from-heaven transfers or bilaterally negotiated transfers can build transfers dependencies that cause the slow economic strangulation of fiscally disadvantaged regions.’119 7.3.2 Legal Basis and Limits of the General Spending Power While most constitutions permit extra-jurisdictional spending either explicitly or implicitly, only a few expressly disallow it. A notable example is Germany, where Article 104a(1) of the Basic Law states that the ‘[f]ederation and the Länder shall separately finance the expenditures resulting from the discharge of their respective responsibilities insofar as this Basic Law does not otherwise provide’. Notwithstanding this ban, in principle, on mixed financing, which is, according to the Federal Constitutional Court, one of the cornerstones of the federal structure,120 the Basic Law indeed does provide otherwise in several instances. Such constitutional exceptions from the prohibition of extra-jurisdictional spending concern the highly important joint tasks of improving regional economic structures, the agrarian structure and of coastal preservation (Article 91a), as well as cooperation with the Länder regarding education programmes and research promotion, provided that they are of supraregional importance (Article 91b). Another exception is the explicit authorisation of the federal Government to provide financial assistance to the Länder in certain enumerated cases (Article 104b). Importantly, however, the latter cannot be subject against their (collective) will to unilateral federal spending in their jurisdictions because the Basic Law always prescribes in all the abovementioned cases the need either of the approval of the Federal Council (Bundesrat) or a compulsory intergovernmental agreement. In federal countries where extra-jurisdictional spending is allowed more broadly, this power is often not explicitly entrenched in the constitution but derived from

118

For an overview of these aims, see Shah (n 3) 28ff. Shah (n 49) 392. 120 32 BVerfGE 333, 338 (Supplemental Tax). 119

230 Financial Relations constitutional jurisprudence. This holds true, in particular, for older federations.121 After the US Supreme Court, for instance, had already significantly augmented the national Government’s revenue powers,122 it also granted it more leeway on the expenditure side. In the interwar period, at a time when the Court often still used to outlaw many national initiatives by invoking the residuary power of the states under the Tenth Amendment, the judges started, in several rulings, to regard the General Welfare Clause as a source for a federal power of extra-jurisdictional spending. According to the US Constitution, the national Government is expressly authorised to tax ‘to pay the debts and provide for the common defence and general welfare of the United States’ (Article I, section 8, clause 1). 7.J United States 1936–37: The Roots of the Federal General Spending Power In the early days of the United States, James Madison and Alexander Hamilton had been engaged in a debate about the scope of the General Welfare Clause. The former claimed that federal tax revenue may only be spent for the purpose of enumerated federal powers, whereas the latter regarded federal expenditures for ‘the general welfare of the United States’ as requiring nothing other than congressional judgment.123 Only with the significant augmentation of federal revenues through the above-mentioned Sixteenth Amendment of 1913124 did this theoretical debate start to assume practical significance because the national Government now had considerable money for extra-jurisdictional spending. After a narrow interpretation of the General Welfare Clause had long prevailed, the Supreme Court initiated in 1936 a change of course towards Hamilton’s interpretation, as it invalidated payments to farmers introduced by the 1933 Agricultural Adjustment Act.125 But they did so due to a violation of the Tenth Amendment because the regulation of agricultural production was not among the enumerated federal powers and not, on the basis of such federal spending being per se, unconstitutional. On the contrary, the Court ruled that the General Welfare Clause conferred a separate and distinct power to spend that was independent of federal legislative powers and only subject to the requirement of the expenditures supporting general national welfare. Only one year later, the Supreme Court further reinforced this Hamiltonian view by largely rejecting a role for itself in providing judicial review of the spending policies of Congress.126 Regarding the crucial distinction between what is for general welfare and what is only for particular welfare, there would be ample discretion belonging to Congress. The Court could only step in if the latter’s view was ‘clearly wrong, a display of arbitrary power, not an exercise of judgment’ (p 640).

In the Canadian case, it was similarly constitutional jurisprudence that developed the spending power of the federal Government. This competence is inferred from

121 See RL Watts, The Spending Power in Federal Systems: A Comparative Study (Kingston, Institute of Intergovernmental Relations, Queen’s University, 1999) 56ff. 122 See box 7.E above. 123 See M Tushnet, The Constitution of the United States of America: A Contextual Analysis (Oxford, Hart Publishing, 2009) 179f. 124 See section 7.2.2 above. 125 United States v Butler 297 US 1 (1936). 126 Helvering (n 107).

Spending Power 231 its above-mentioned general taxation power (Section 91(3) of the Constitution Act 1867), its competence to make laws regarding ‘public property’ (Section 91(1a)) and to appropriate federal funds (Section 106). According to the Supreme Court, the scope of this power transcends that of national legislative competences to regulate certain subject matters.127 By contrast, some scholars, including, interestingly enough, Pierre Trudeau before his political career,128 have argued that the federal spending power would be confined to its legislative competences. The Supreme Court has not followed this view. While constitutional jurisprudence in Canada has thus been quite consistent over time, the Australian High Court recently issued several rather surprising judgments whose long-term effects on the use of the federal spending power are still difficult to predict, but are potentially far-reaching. In concrete terms, the Court limited, in one ruling, the long-standing practice of invoking as a basis for the spending power of the national Government its executive power under Section 61 of the Australian Constitution, which ‘extends to the execution and maintenance of this Constitution, and of the laws of the Commonwealth’. 7.K Australia 2012–14: Limits on the Federal Spending Power? In 2006, the Australian Government established a funding programme, called the National School Chaplaincy Programme, under which schools could apply for federal financial assistance concerning chaplaincy services. In creating this programme, the Government relied exclusively on the above-mentioned executive power under Section 61 of the Australian Constitution without additionally adopting ordinary legislation. After the state of Queensland had successfully applied for financial support, Ronald Williams sought to ensure secular education for his children by challenging the constitutionality of such federal funding of chaplaincy services. The High Court opposed the view that the executive power under Section 61, as a basis for the federal spending power, would be coextensive with the national Government’s enumerated legislative competences (Sections 51, 52 and 122) and would allow expenditures by the federal executive without express authorisation through ordinary law.129 According to the judges, Section 61 could be directly invoked only in certain exceptional cases. This flatly contradicted widely held assumptions regarding the scope of the federal executive power that were not even challenged, initially, by the states during the court proceedings. The judges, however, opposed the hitherto shared view of both government levels that the exercise of the federal spending power would be either unlimited or at least constitutional as long as it concerned a subject matter on which the national Parliament could validly enact legislation, even if it was not enacted. In the immediate aftermath of the ruling, the Australian Government reacted by adopting the 2012 Financial Framework Legislation Amendment Act (FFLA Act), which granted the federal executive general permission to spend funds for any purpose specified in regulations made by itself and not by Parliament, with chaplaincy services in schools being listed

127

Finlay v Canada (Minister of Finance) [1986] 2 SCR 607. Trudeau, ‘Federal Grants to Universities’ in PE Trudeau (ed), Federalism and the French Canadians (Toronto, Macmillan of Canada, 1968). 129 Williams v Commonwealth (2012) 248 CLR 156 (School Chaplains case). 128 PE

232 Financial Relations

among these purposes. This workaround was intended to quickly remedy the constitutional problems concerning the National School Chaplaincy Programme and numerous similar funding schemes, but it was again successfully challenged by Ronald Williams. The High Court ruled that the FFLA Act and regulations purporting to authorise the funding programme were invalid because they would not fall within any of the federal legislative powers.130

The fact that the High Court limited Section 61 as a basis of the federal spending power was unusual insofar as the judges based their ruling, in otherwise quite centralised Australia, explicitly on federal grounds, which they still had refused to do in a case only a few years before.131 The restriction of the federal executive power, however, does not affect the alternative mechanism of making expenditures via the states under Section 96. This provision makes it possible to provide ‘financial assistance to any state on such terms and conditions as the parliament thinks fit’ so that grants may be subject to nearly any condition and are not limited by subject matter.132 To be sure, the states may reject funds under Section 96 without any legal means to enforce their acceptance. But from an economic perspective, these grants are, like in the above-mentioned US and Canadian cases, very far from being non-coercive. In practice, states have rarely been provided with influence on federal spending decisions under this provision, even where it concerned their own jurisdictions.133 As the case law of the Spanish Constitutional Court illustrates, an extraordinary role of constitutional jurisprudence in developing the federal spending power is a feature not only of older federal systems. Since a leading case in 1992, the national Government has been deemed to possess the power to make expenditures for any purpose and any matter, even outside its legislative competences, provided that it does not thereby seize competences from the autonomous communities.134 The crucial question therefore is when and under which conditions extra-jurisdictional spending actually amounts to such a (clearly unconstitutional) seizure of subnational competences. This leads us to the second main issue addressed in this section. Like the justexplored legal basis of the spending power, its limits in most federal systems are also defined by constitutional jurisprudence. The Canadian Supreme Court has, in

130 Williams v Commonwealth (2014) 252 CLR 416. See S Chordia et al, ‘Williams v Commonwealth: Commonwealth Executive Power and Australian Federalism (No 2)’ (2015) 39 Melbourne University Law Review 306. 131 New South Wales v Commonwealth (2006) 229 CLR 1 (Work Choices case). See D Hume et al, ‘Heresy in the High Court? Federalism as a Constraint on Commonwealth Power’ (2013) 41 Federal Law Review 71. 132 Victoria v Commonwealth (1957) 99 CLR 575, 605 and 609f (Dixon J) (Second Uniform Income Tax case). See Chordia et al, ‘Williams v Commonwealth’ (2015) 324. 133 See B Galligan, ‘Fiscal Federalism: Then and Now’ in G Appleby et al (eds), The Future of Australian Federalism: Comparative and Interdisciplinary Perspectives (Cambridge, Cambridge University Press, 2012) 320ff. 134 STC 13/1992.

Spending Power 233 this regard, a quite clear approach, as it provides the national Government with a spending power that is virtually unlimited. In its view, the latter may tie to federal grants any condition and ‘[t]he simple withholding of federal money which had previously been granted to fund a matter within provincial jurisdiction does not amount to the regulation of that matter’.135 Setting conditions and the threat to hold back funds in case of non-compliance on the part of provinces would therefore not violate provincial legislative powers. Notwithstanding the obvious financial pressure, there would still be an element of voluntariness as long as the provinces had the option to reject the funds. Also, certain more recent federal systems provide for an essentially unlimited federal spending power. In Belgium, for instance, this is derived from the budgetary power of the House of Representatives for which Article 174 of the Constitution does not set any explicit limits regarding extra-jurisdictional expenditures. Even though this considerable leeway of the national Government might suggest, for both Canada and Belgium, a particularly extensive intrusion into subnational powers, this is not the case in practice. In fact, the assessment that ‘the Belgian fiscal system combines a limited soliciting autonomy with a large spending autonomy’136 still holds true. As for Canada, it is true that the federal Government had initially used, in the aftermath of World War II, shared-cost programmes to gain influence on provincial policies. But today the main grants, the Canada Health Transfer (CHT) and Canada Social Transfer (CST), do not contain, beyond being earmarked for specific purposes, very onerous conditions.137 Quite the contrary, Canada ‘is notable among federations in the percentage (about 80 per cent) of its intergovernmental transfers that take the form of block payments’.138 The opposite is true for the United States, where conditional grants clearly prevail and unconditional block payments, ie the General Revenue Sharing from 1972 to 1982, remained a singular historical episode. Indeed, conditional grants-in-aid by the national Government are considered one of the key characteristics of what has been called ‘coercive federalism’ because they make it possible ‘to achieve federal objectives that lie outside Congress’s constitutionally enumerated powers and to extract more state-local spending on federal objectives’.139 All this suggests that the issue of the limits on the federal spending power is of continued importance. Even before the above-mentioned Butler case, the Supreme Court had regarded the practice of conditional grants in 1923 as not violating the Tenth Amendment, but it did so, instead of the General Welfare Clause, on the basis that the plaintiffs in the case lacked the

135

Reference Re Canada Assistance Plan (BC) [1991] 2 SCR 525, 567. Delmartino, ‘Belgium after the Fourth State Reform: Completed Federalism or Confederalism in the Making?’ in G Färber and M Forsyth (eds), The Regions—Factors of Integration or Disintegration in Europe? (Baden-Baden, Nomos, 1996) 128. See, more recently, also W Swenden, ‘The Belgian Regions and the European Union: Unintended Partners in Unravelling the Belgian State?’ in RW Jones and R Scully (eds), Regionalism and Constitutionalism in an Enlarged European Union (Basingstoke, Palgrave Macmillan, 2009) 21. 137 On the shared-cost programmes as well as CHT and CST, see ch 11.2. 138 DM Brown, ‘Fiscal Federalism: Maintaining a Balance’ in H Bakvis and G Skogstad (eds), Canadian Federalism: Performance, Effectiveness and Legitimacy, 3rd edn (Toronto, Oxford University Press, 2012) 124. 139 J Kincaid, ‘The Rise and Characteristics of Coercive Federalism’ in Pola (n 21) 65. 136 F

234 Financial Relations suffering of particularised harm and thus legal standing.140 The subsequent case law regarding the General Welfare Clause, in particular the above-mentioned judgment in the Helvering case,141 further expanded the leeway of the national Government. Essentially, such grants have been perceived as having the nature of a contract and as constitutional if a state ‘voluntarily and knowingly accepts the terms of the contract’.142 By refusing to qualify certain conditional transfers as coercive, the US Supreme Court has traditionally been reluctant to establish limits on the federal spending power. It upheld, for instance, the withholding of five per cent of federal funds for highway construction if states failed to comply with the federal policy of a minimum age of 21 for purchasing alcohol.143 It was only in 2012 that the Court held, for the first time in its history, a condition of federal aid to be coercive and thus unconstitutional. The part of the 2010 Patient Protection and Affordable Care Act (PPACA) that threatened states in case of non-compliance with the loss of existing Medicaid funding, amounting to more than 10 per cent of a state’s total budget, was struck down as ‘economic dragooning that leaves the states with no real option but to acquiesce’.144 Beyond the ideologically charged Sebelius case, however, the use of the federal spending power has not provoked general discontent in the United States, as it has in other cases, even if in those other cases conditional grants are in fact less prevalent. This holds true especially in comparison with federal systems like Canada that operate in the context of an ethno-culturally diverse society. In such countries, more fundamental opposition against extra-jurisdictional expenditures of the national government is often linked to competing projects of nation-building and the crucial role of spending programmes in this regard, above all for social welfare and healthcare.145 Context and perception therefore matter crucially in the degree to which the federal general spending power is viewed as problematic. Compared to the extensive leeway for the exercise of this power in the United States, constitutional jurisprudence in some other federal systems establishes, at least at first glance, stricter limits. A notable example is the case law of the Spanish Constitutional Court. Even if the Court confirmed that the national Government had the power, as mentioned above, to make expenditures for any purpose and any matter, it also introduced in the very same judgment certain constraints.146 First, it derived from the above-mentioned principle of financial autonomy (Article 156(1)) that the autonomous communities must not only have sufficient funds to exercise their powers, but must also be able to dispose of those funds without unreasonable

140 Massachusetts v Mellon 262 US 447 (1923). The case was consolidated with Frothingham v Mellon, which was about the question of whether a single federal taxpayer had legal standing to sue the national Government to prevent expenditures if his/her only harm was an anticipated increase in taxes. 141 See box 7.J. 142 Pennhurst State School and Hospital v Halderman 451 US 1 (1981). 143 South Dakota v Dole 483 US 203 (1987). 144 National Federation of Independent Business v Sebelius 567 US ___ (2012). See box 11.A. See also Kincaid, ‘The Rise’ (2015) 65f. 145 See H Telford, ‘The Federal Spending Power in Canada: Nation-Building or Nation-Destroying’ (2013) 33 Publius 23. See also ch 11.2. 146 STC 13/1992.

Spending Power 235 conditions. Second, the judges developed a model of a differentiated spending power. Put simply, the more extensive the legislative powers of the national Government, the more comprehensively it may regulate its grants. Consequently, the conditions tied to grants may be more detailed and specific, for example, regarding those matters under Article 149(1) of the Constitution that foresee national legislation on ‘bases’ or ‘basic norms’147 than they may be for matters of exclusive regional competence. In practice, however, there are several exceptions to this rule, and the national Government has often failed to act in compliance with this framework of the differentiated spending power so that constitutional complaints from the autonomous communities against financial intrusions have usually been successful.148 While extra-jurisdictional expenditures as a tool to influence another level of government is typically an issue linked to the federal spending power, this is not as much a problem concerning the equivalent power of subnational entities. Even though they may also have a general power to spend beyond their legislative competences, they usually do not have sufficient revenues or political impetus to do so, at least not on a larger scale. From a legal point of view, the Canadian Supreme Court did not interpret the above-mentioned provincial power to levy direct taxes for ‘the raising of a revenue for provincial purposes [emphasis added]’ (Section 92(2) of the Constitution Act 1867) as excluding the spending of these revenues for other purposes. The Court, for example, upheld the constitutionality of a commuter train service across provincial boundaries and the appropriation of casino profits for First Nations communities.149 But as there have been only a few such cases, the provincial spending power has never become a controversial issue.150 7.3.3 Spending Power and Fiscal Discipline As far as subnational entities are concerned, the issue of the spending power and its limits does not emerge so much regarding intrusive extra-jurisdictional expenditures but mostly in the context of rules about fiscal discipline. Increasingly widespread balanced-budget rules affect, of course, both revenue and spending powers, but the latter in particular because the implementation of such rules typically focuses in large part on cuts of discretionary spending.151 In European federal systems, balancedbudget rules were introduced in many cases due to the fiscal consolidation efforts to meet the euro convergence criteria (‘Maastricht criteria’). The obligation to balance their budgets was imposed, for example, on the subnational entities of Belgium in 147

See ch 5.2.2. C Viver, ‘Centralisation and Decentralization Trends in Spain: An Assessment of the Present Allocation of Competences Between the State and the Autonomous Communities’ in Institut d’Estudis Autonòmics (ed), Decentralizing and Re-centralizing Trends in the Distribution of Powers within Federal Countries (Barcelona, Institut d’Estudis Autonòmics, 2010) 169f; Ruiz Almendral, ‘Autonomous Communities’ (2003) 16. 149 The Queen (Ont) v Board of Transport Commissioners [1968] SCR 118; Lovelace v Ontario [2000] 1 SCR 950. 150 See, eg, Hogg (n 5) 6-21. 151 See J Jonas, ‘Great Recession and Fiscal Squeeze at US Subnational Government Level’ (2012) IMF Working Papers WP/12/184, 17f www.imf.org/external/pubs/ft/wp/2012/wp12184.pdf. 148 See

236 Financial Relations a 1996 intergovernmental cooperation agreement and on those of Spain in a 2001 ‘internal stability pact’, even if in the Belgian case most of the fiscal consolidation burden was actually carried by the national Government.152 In the Spanish case, the dynamics were of a different kind, first with widespread disregard and lack of enforcement of the quite toothless internal stability pact, at least until the onset of the global financial crisis,153 and then increasing tensions that eventually resulted in a dispute before the Constitutional Court. 7.L Spain 2011: Debt Ceilings as a Violation of Financial Autonomy? The ‘internal stability pact’ in response to the ‘Maastricht criteria’ had its legal bases in the Organic Law 2001 Complementing the General Law on Budget Stability and in the 2007 Royal Decree on Budget Stability. According to this new legal framework, the national Government was authorised to set a debt ceiling for all autonomous communities after having received a report from the CPFF154 and following consultations with each autonomous community. In a second step, it then entered into bilateral negotiations with each region regarding its individual maximum share of the overall debt, and in case of disagreement decided unilaterally. In any case, the ‘internal stability pact’ remained a source of significant intergovernmental tensions so that parts of it were finally challenged by Catalonia. The Court held that some of the contested provisions were indeed violations of the financial autonomy under Article 156(1) of the Spanish Constitution. Importantly, however, the judges affirmed at the same time the existence of an implicit balanced-budget principle in the Spanish legal system even without an explicit entrenchment in the Constitution.155 With reference to previous case law,156 the Court emphasised that rules aimed at fiscal discipline were, in principle, admissible. It did not agree with the submission of the plaintiff that the national Parliament had overstepped its powers because the ‘internal stability pact’ would have imposed stricter measures than required by EU law. The Court found that the above-mentioned procedure of setting debt ceilings was clearly supported by Article 149(1)(13) of the Spanish Constitution, which grants the national Government exclusive competence over ‘[b]asic rules and coordination of general economic planning’. The establishment of debt ceilings would not violate financial autonomy because this is necessary to ensure some degree of uniformity within the whole system.

In Spain and other federal systems in Europe, the situation changed significantly with the 2012 European Fiscal Compact.157 In order to comply with this treaty, Spain explicitly entrenched, only two months after the above-mentioned judgment,

152 See W Leibfritz, ‘Fiscal Federalism in Belgium: Main Challenges and Considerations for Reform’ (2009) OECD Economics Department Working Paper 743, 5ff www.oecd-ilibrary.org/content/ workingpaper/218562648706. 153 See V Ruiz Almendral, ‘Curbing the Deficit in Spain and its Autonomous Communities: A Constitutional Conundrum’ (2013) 1 Tijdschrift voor Constitutioneel Recht 68, 68ff. 154 See section 7.1.3 above. 155 STC 134/2011. 156 STC 62/2001. 157 See section 7.1.2 above.

Spending Power 237 a balanced-budget principle in Article 135 of the Constitution. Other European federal systems also made similar amendments aimed at fiscal discipline (eg Article 13(2) of the Austrian Constitution, Articles 109 and 110 of the German Basic Law and Articles 81, 117 and 119 of the Italian Constitution). Beyond the mere obligation to balance budgets, these recent changes sometimes created a broader framework aimed at guaranteeing fiscal discipline, albeit with quite different impacts on the power relations between the levels of government. In Italy, for instance, the 2012 constitutional amendments involved the granting of a new exclusive power to the national Government regarding the ‘harmonisation of public budgets’ (Article 117(2)(e)).158 By contrast, the recent amendments to the German fiscal Constitution, some of which were already passed as part of the Federalism Reform II in 2009 and thus not in response to the European Fiscal Compact, included some seemingly less centralising measures. The new Article 109a of the Basic Law assigned the task of continuously overseeing the budget management of both the federal Government and the Länder to a joint intergovernmental body, ie the Stability Council.159 Together with what has come to be known as the ‘debt brake’ (‘Schuldenbremse’), this council is at the heart of the new institutional framework to ensure budget stability. Among several amendments to introduce the ‘debt brake’, the long Article 109(3) is the key provision. It stipulates that the federal and Länder budgets are to be balanced, in principle, (almost) without borrowing. Apart from exceptional circumstances, such as severe recessions and natural catastrophes, the Länder may not resort to net borrowing to keep a balanced budget, and the federal Government may do so only for up to 0.35 per cent of the GDP. Even though these limits only have to be observed after 2016 by the national Government and 2020 by the Länder (Article 143d(1)), some of the latter have already entrenched them in their own subnational constitutions. Beyond federal systems in Europe, Brazil is an often-cited example of a legal framework to ensure fiscal discipline. As the country’s 1988 Constitution introduced a rather decentralised financial regime, this reinforced existing problems with excessive spending and borrowing by states. These issues were eventually addressed by the Fiscal Responsibility Law of 2000, which featured a set of binding rules for all levels of government.160 These included, among other things, debt limits, restrictions on expenditures for personnel and a ban on extraordinary spending in election years to attract potential voters. With a similar purpose, Argentina enacted a Fiscal Responsibility Law in 2004 that followed the country’s economic crisis from 1998 to 2002.161

158 See G Martinico and L Pierdominici, ‘Crisis, Emergency and Subnational Constitutionalism in the Italian Context’ (2014) 6 Perspectives on Federalism 116, 123. 159 See AB Gunlicks, ‘Legislative Competences, Budgetary Constraints, and the Reform of Federalism in Germany from the Top Down and the Bottom Up’ in M Burgess and GA Tarr (eds), Constitutional Dynamics in Federal Systems: Subnational Perspectives (Montreal, McGill–Queen’s University Press, 2012) 77. 160 See F Rezende, ‘Federal Republic of Brazil’ in Shah and Kincaid (n 3) 82ff. 161 See MA Asensio, ‘Fiscal Federalism and Federal Decisions’ in F Palermo and E Alber (eds), Federalism as Decision-Making: Changes in Structures, Procedures and Policies (Leiden, Brill–Nijhoff, 2015) 240f.

238 Financial Relations Even if these more recent responses to crises in Europe and Latin America have entailed an increase in balanced-budget rules, they are, of course, not a new invention. Indeed, some federal systems have a long and deeply rooted tradition of regulations regarding fiscal discipline. In the US case, this even goes back to the 1840s, when a state debt crisis prompted several states to establish balanced-budget requirements. Today, all states apart from Vermont have such requirements either in their constitutions and/or in ordinary statutes, and some of them, additionally, even restrict the rate of spending increase to the rate of growth of the state economy.162 Even if state balanced-budget rules vary considerably in terms of their stringency and are, like in California in the late-1980s, not always observed, they certainly have had profound effects on subnational spending. During the current crisis, however, they have also had the effect of counteracting and partially offsetting federal efforts of countercyclical economic stimulation through the 2009 American Recovery and Reinvestment Act.163 In other cases, regulations regarding fiscal discipline are characterised by different patterns that are, more generally, typical of these federal systems (eg direct democracy in Switzerland and intergovernmental coordination in Canada). Unlike in the United States, balanced-budget rules exist in Switzerland not only at the subnational level but in the federal Constitution as well (Article 126). Only cantonal constitutions, however, feature an additional instrument, that of the fiscal referendum. Such popular votes on planned spending are either mandatory or optional and, in the latter case, dependent on the collection of a certain number of signatures calling for it. As even the combined effect of these two instruments has in some instances failed to prevent an increase in public debt, several cantons have adopted, starting with St Gallen in 1994, additional provisions. Similar to what is known in the United States as ‘rainy day funds’,164 some of them now require the building up of savings during prosperous years for worse times. In effect, they institutionalised in this manner anti-cyclical fiscal policy without an accumulation of public debt.165 In contrast to the above-mentioned examples, fiscal discipline in Canada is not a constitutional matter. Instead, it is left to mechanisms of federal–provincial coordination during the budgetary process. The relative success of these mechanisms throughout the current crisis has arguably been facilitated by a broad ideological consensus since the country’s fiscal crisis in the 1990s, as both government levels have since then attached great importance to balanced budgets.166 In this context, it is important to note that rules and processes to ensure fiscal discipline are in all three countries, the United States, Switzerland and Canada alike, sustained by an underlying political culture of prudence in subnational spending. Such support certainly plays a critical role more generally because comparative evidence demonstrates that

162 See R Briffault, Balancing Acts: The Reality Behind State Balanced Budget Requirements (New York, 20th Century Fund, 1996). 163 See DA Super, ‘Rethinking Fiscal Federalism’ (2005) 118 Harvard Law Review 2576, 2609f; Jonas, ‘Great Recession’ (2012). 164 See Fox, ‘The United States’ (2007) 366. 165 G Kirchgässner, ‘Swiss Confederation’ in Shah and Kincaid (n 3) 326ff. 166 See T Lewis, In the Long Run We’re All Dead: The Canadian Turn to Fiscal Restraint (Vancouver, University of British Columbia Press, 2003).

Spending Power 239 fiscal discipline is in some instances achieved without rules to induce it and in others absent in spite of them.167 Apart from cultural factors, fiscal prudence on the part of subnational entities is typically seen as being heavily influenced, in particular, by opportunities for a federal bailout, by obscured accountability of subnational entities through large transfers in relation to their scarce own revenues and the degree of their access to borrowing.168 Historically, it was, for example, in the United States the refusal of a bailout during the above-mentioned state debt crisis of the 1840s that led, in turn, to state constitutional balanced-budget rules. It is indeed, more generally, in older federal systems that no-bailout rules for subnational entities are in place. This contrasts sharply with Germany, where financial rescue measures for the Länder had been regarded for a long time not only as a political possibility, but even as a constitutional obligation. 7.M Germany 1992–2006: Bailouts for the Länder and the Limits of Solidarity Beginning in the 1970s, the Länder of Bremen and Saarland amassed considerable debts. Yet, despite increasing reliance on equalisation payments, both continued to spend and incur deficits that were, in the context of the German Länder’s limited taxing powers but unlimited borrowing autonomy, primarily financed through loans. As it became increasingly clear that the debt levels of Bremen and Saarland had become unsustainable, the two Länder eventually requested a bailout and attempted to enforce this before the Federal Constitutional Court. In a judgment in 1992, the Court indeed confirmed that the two Länder suffer from a situation of extreme fiscal stress that they would be incapable of remedying alone.169 For such cases, the judges considered the definition of Germany as a ‘social federal state’ (Article 20(1) of the Basic Law) to require the different parts of the federation to demonstrate solidarity and ensure financial support unless such support would make the other parts themselves unable to fulfil their constitutional obligations. In response, the federal Government provided for a bailout through conditional grants between 1994 and 2004, which, understandably, created certain expectations among the Länder. When the Land of Berlin similarly requested a bailout, with explicit reference to the 1992 judgment, the Federal Constitutional Court ruled in 2006 once again on the conditions of such a financial emergency measure.170 The judges held that a bailout was only consistent with the Constitution and mandated by it if fiscal stress were extreme in comparison with other Länder and in relation to the responsibilities assigned by the Constitution (relative criterion) and if it threatened the very existence of the Land concerned (absolute criterion). In the case of Berlin, the Court denied the existence of such an extreme situation that would pose an existential threat to the Land and thus make a bailout inevitable. Germany’s equalisation system would ultimately aim to enable subnational entities to autonomously fulfil their constitutional tasks, and within this system the requested supplementary grants under Article 107(2) of the Basic Law must be interpreted narrowly as a means of last resort.

167

See Shah (n 49) 385f. J Rodden, ‘The Dilemma of Fiscal Federalism: Grants and Fiscal Performance around the World’ (2002) 46 American Journal of Political Science 670. 169 86 BVerfGE 148 (Financial Equalization II). 170 116 BVerfGE 327 (Budget of Berlin). 168 See

240 Financial Relations 7.4 EQUALISATION

As outlined in the two preceding sections, there has been, in many federal systems, a trend towards tax centralisation, while subnational entities often have significant spending requirements. The result is a vertical fiscal gap, ie a mismatch between disposable revenues and expenditure needs. Conversely, the national government is typically in a situation in which its revenues exceed its own expenditure requirements. This gap may be prevented by extensive subnational taxing powers and addressed ex post by intergovernmental transfers. If this disparity still exists after transfers, the gap turns into what has been called a vertical fiscal imbalance.171 For obvious reasons, the first of the above-mentioned options, that of extensive subnational taxing powers, is typically preferred by subnational entities with greater fiscal capacity (eg Alberta in Canada and the Flemish region in Belgium), while transfers are more popular among poorer ones. This is because the decentralisation of taxing power only permits entities with high fiscal capacity to avoid their individual vertical fiscal gap in relations with the national government. But at the same time, it also allows this differential capacity to play out fully and thus increases horizontal fiscal imbalances. This fiscal divide between subnational entities is remedied in nearly all federal systems, at least to some extent, through arrangements of equalisation. If such an arrangement mainly relies, as it does in many cases, on a transfer of federal funds, then the remedy for horizontal fiscal imbalances requires a certain vertical fiscal gap. Otherwise the national government would simply lack sufficient funds for effective redistribution. The horizontal imbalances that equalisation addresses concern, however, not only uneven fiscal capacity per capita, but sometimes also mismatches stemming from diverging per capita costs of public services. The latter may follow, for instance, from demographic differences between subnational entities like the age structure or from challenging physical environments like mountains. Spending needs are then assessed on the basis of average costs for service provision, historical expenditures or indirect expense indicators like the age or number of inhabitants. This figure may be real or modified as in Germany, where the city states of Berlin, Bremen and Hamburg are in a privileged position (so-called Einwohnerveredelung) due to the multiplication of their actual number by a factor of 1.35 (§ 9(2) of the Financial Equalisation Law). From a comparative perspective, most federal countries rely on some combination of fiscal capacity and fiscal need equalisation, even though Canada, for instance, concentrates on the former and Australia on the latter dimension. Regardless of which dimension of horizontal imbalance a concrete arrangement (primarily) addresses, it has to be noted that the relative position regarding both capacity and needs may change so that equalisation has an element of risk-pooling over time. Today’s net payers may be tomorrow’s net recipients and vice versa. Cases in point are the recent transformation of Ontario due to industrial decline into a beneficiary and the simultaneous turn of Newfoundland and Labrador into a payer province because of offshore oil discoveries. Another example is the complete

171

On the difference between vertical fiscal gap and vertical fiscal imbalance, see Shah (n 3) 28.

Equalisation 241 reversal in the economic fortunes of the Flemish and Wallonian parts of Belgium since the 1960s. Another crucial question in designing an equalisation arrangement is whether to rely primarily on horizontal transfers between subnational entities or vertical transfers from the national government.172 Even if Canada is, in view of its exclusively vertical flow of funds, rather an exception, equalisation through the federal level is in most cases at least predominant. This is due to the fact that the above-mentioned trend towards tax centralisation typically makes a purely self-financing system of subnational entities an unrealistic option. There are, however, federal systems that feature quite extensive horizontal transfers. Traditionally, this has been the case in Germany. Article 107(2) of the Basic Law stipulates that an ordinary law requiring the approval of the Federal Council (Bundesrat) must establish an equalisation system by specifying rules for ‘the claims of Länder entitled to equalisation payments and the liabilities of Länder required to make them’ and may also foresee, as a supplement, ‘grants to be made by the Federation to financially weak Länder from its own funds’. The required ordinary law, the Financial Equalisation Law, the current version of which was enacted in 2001 and is in force from 2005 to 2019, indeed specifies a procedure with first horizontal transfers and then vertical grants. The latter are to be understood, according to the Federal Constitutional Court, as supplementary to the main transfers between the Länder and not as their substitutes.173 More recently than horizontal equalisation in Germany, a new Article 135 was introduced in 2004 into the Swiss Constitution and with it a strong element of inter-cantonal transfers. While fiscal need equalisation is a matter for the national Government, the funds for the equalisation of fiscal capacity are provided by the latter plus the wealthier cantons. It is even expressly stipulated in detail that the contribution of these cantons must be between 66.7 and 80 per cent of the amount put in by the national Government (Article 135(3)). Apart from the question of whether equalisation focuses on the fiscal divide or different costs and whether transfers are primarily vertical or horizontal, the fundamental underlying question is, for any such system, the extent of redistribution. This is linked to the more fundamental issue of how much emphasis a federal country places on certain opposing values: on the one hand, autonomy of subnational entities and competitive service provision and, on the other hand, solidarity and the idea, as in Canada, to have in all parts of the federation ‘reasonably comparable levels of public services at reasonably comparable levels of taxation’ (Section 36(2) of the Constitution Act 1982).174 These more general orientations have a decisive impact on the degree of equalisation and on whether they exist at all. In the Canadian case, for instance, the commitment to redress imbalances between the 172 See R Boadway, ‘Intergovernmental Redistributive Transfers’ in Ahmad and Brosio (eds), Handbook (2006) 363f. 173 72 BVerfGE 330 (Financial Equalization I). 174 This provision merely declares that the national Parliament and Government ‘are committed to the principle of making equalisation payments’ for this purpose. Its justiciability, however, has remained controversial. In favour of enforceability in court, see A Nader, ‘Providing Essential Services: Canada’s Constitutional Commitment Under Section 36’ (1996) 19 Dalhousie Law Journal 306, 349 and rather against it, Hogg (n 5) 6-10 and 33-3.

242 Financial Relations provinces was there long before the introduction of formal equalisation in 1957 and actually already formed part of the initial federal compromise in 1867.175 In Germany, this commitment is even more explicit, as the Basic Law’s proclamation of a ‘social federal state’ (Article 20(1)) is even protected under Article 79(3) against constitutional amendment and forms the context for far-reaching redistribution. Values in both these federations contrast sharply with the United States, where such an equalisation arrangement is conspicuously absent. This is not least because ‘Americans tolerate, even support, large variations in living standards within the country’.176 One should caution, however, against overlooking certain implicit equalising dynamics that operate there just as in other federal systems. These dynamics stem from the joint financing or provision of public services in areas like education or healthcare, which may be regarded as ‘transfers in kind’.177 In the US case, equalising effects are achieved, in particular, by sharing the costs for the Medicaid programme, which demands a higher share of co-funding from richer states than from poorer ones.178 Where formal equalisation arrangements do exist, the above-mentioned general orientations towards solidarity or autonomy make them nonetheless vary significantly concerning the extent of redistribution. Australia, for instance, places much more emphasis on solidarity (also benefitting from not very wide fiscal disparities) than competition-focused Switzerland with an accordingly lower degree of equalisation. But beyond these orientations, there are also inherent systemic restraints on the extent to which such correction is possible that follow from the connection between equalisation and tax (de)centralisation. If subnational entities have, like in Canada, relatively extensive revenue powers, this situation, first, increases the horizontal fiscal divide and, second, leaves to the national government less funds to address this divide. Evidently, the latter is particularly problematic if the transfers are largely vertical, which is, as mentioned above, again characteristic of the Canadian case. Against the background of considerable tax decentralisation and some provinces benefitting immensely from natural resources, the horizontal fiscal divide has over the years reached such proportions that certain limits on the extent of equalisation became practically inevitable. Back in 1957, the initial compromise had not been to bring down the wealthier provinces through horizontal transfers, but to bring the poorer ones up through vertical transfers to the average per capita fiscal capacity of the two highest-income provinces. Ten years later, the objective was lowered to the average of all provinces and in 1982 to that of five representative provinces, excluding both the richest and poorest. After 2004, when the calculation was again based

175 See PE Bryden, ‘The Obligations of Federalism: Ontario and the Origins of Equalization’ in D Anastakis and PE Bryden (eds), Framing Canadian Federalism (Toronto, University of Toronto Press, 2009) 76. 176 DA Kenyon and J Kincaid, ‘Fiscal Federalism in the United States: The Reluctance to Equalize Jurisdictions’ in WW Pommerehne and G Ress (eds), Finanzverfassung im Spannungsfeld zwischen Zentralstaat und Gliedstaaten (Baden-Baden, Nomos, 1996) 38. 177 See R Boadway, The Constitutional Division of Powers: An Economic Perspective (Ottawa, Economic Council of Canada, 1992) 8ff. 178 See ch 11.1.

Equalisation 243 on the average of all provinces, other limits were introduced, like a cap on payments and the inclusion of only half of the provincial revenues from natural resources. Similar limits regarding the extent of redistribution were recently established in Spain. A reform of the LOFCA in 2009 replaced the previous single equalisation payment with a broader system of transfers from four different funds.179 The Essential Public Services Fund, intended for fiscal need equalisation, is not aimed at a full correction of disparities, but only covers spending of the autonomous communities in the fields of education, healthcare and essential social services (Article 15(1–2) of the LOFCA 2009 and before Article 206(3) of the 2006 Statute of Catalonia). This is in line with Article 158(1) of the Spanish Constitution, which expressly states that transfers may be made ‘to guarantee a minimum level of basic public services throughout Spanish territory’. It is assumed that the financial means from this fund and from the ceded taxes are essentially sufficient to cover expenditures and that the Global Sufficiency Fund (Article 13(2) LOFCA) is merely a subsidiary safety net.180 In Germany, the extent of correcting horizontal imbalances, ie the interpretation of the ‘reasonable equalisation of the disparate financial capacities of the Länder’ (Article 107(2)), has been equally controversial, but it was not addressed, unlike in Spain, through a comprehensive reform. As a matter of fact, the Federalism Reform II in 2009 was anything but a profound overhaul of financial relations and concentrated on the above-mentioned ‘debt brake’. Instead of a comprehensive solution, disputes between the Länder have been ongoing for decades and even led to litigation before the Federal Constitutional Court. 7.N Germany 1986–99: The Balance between Länder Autonomy and Solidarity Soon after the adoption of the Basic Law, equalisation turned out to be one of the most contentious topics of intergovernmental relations, with the first judgment in this matter of the Federal Constitutional Court handed down as early as in 1952.181 Divergent positions between the Länder further intensified during the 1980s and then, even more, following German reunification. One notable dispute led in 1986 to a judgment182 that defined equalisation as a balancing act that is about finding ‘the appropriate middle ground between autonomy, individual responsibility and the preservation of the individuality of the Länder on the one hand and the common solidarity and joint responsibility for their existence and autonomy on the other hand’ (p 398). Beyond this general statement, the Court set out, based on expertise from public finance, in a quite detailed manner the requirements for equalisation and its limits, ie a specific minimum and maximum level of equalisation. In 1999, the Court dealt with another legal challenge to equalisation,183 lodged by net payers Bavaria, Baden-Württemberg and Hesse. They had become increasingly disenchanted with the redistribution arrangement because the substantial transfer of

179 These are the Essential Public Services Fund, the Global Sufficiency Fund, the Competitiveness Fund and the Cooperation Fund. For an overview of the new system, see Herrero Alcalde (n 25) 31ff. 180 See Pérez Royo, Curso (2010) 840. 181 1 BVerfGE 117 (Financial Equalization Act). 182 72 BVerfGE 330 (Financial Equalization I). 183 101 BVerfGE 158 (Financial Equalization III).

244 Financial Relations

additional revenue generated by them to poorer Länder would provide disincentives for their own economic development. So they claimed that transfers would be excessive and no longer in compliance with the constitutional standard of ‘reasonable equalisation’ (Article 107(2)). The judges held that the necessary balance between the Länder’s ‘own statehood’ (Eigenstaatlichkeit) and the ‘mutually supportive federal community’ (Solidargemeinschaft) would not be maintained if disparities were not merely reduced, but levelled (Nivellierungsverbot) (p 222).184 The ruling left open where exactly to locate this boundary between such reductions and levelling, but it did not consider bringing up the poorer Länder to at least 95 per cent of the average fiscal capacity per capita excessive. While the judges thus indicated in this regard a certain leeway, they also required the national Parliament to pass two pieces of legislation, ie a Standards Law (Maßstäbegesetz) regarding the criteria for both revenue-sharing and equalisation, as well as, on its basis, the Financial Equalisation Law to be periodically amended. The current such law was adopted in 2001 and introduced some competitive elements into a system that is otherwise still marked by a high degree of solidarity. While it excludes 12 per cent of the above-average tax revenue increase from redistribution, it still goes to great lengths to bring each Land up to at least 95 per cent of the average fiscal capacity per capita through horizontal transfers and then even to 97.5 per cent through the supplementary vertical grants.

As in Germany, equalisation has been, in many federal systems, a subject of continued debate, with these often being reinforced by certain exogenous and endogenous factors. In the latter regard, the faulty design of the system of equalisation seems to have additionally fuelled controversies in several cases. Regions in Belgium, for example, have been faced with a disincentive for their own economic development. The increase through an economic upswing of the income tax transfer from the national government would be outweighed by a simultaneous decrease of the equalisation grant for the then less needy region.185 In the Spanish case, horizontal imbalances regarding fiscal capacity were, under the old system (before the abovementioned LOFCA reform of 2009), not only equalised but even reversed.186 It is evident that such an excessive redistribution nurtured discontent in wealthier autonomous communities and particularly in Catalonia. As a countermeasure, Article 206(5) of the 2006 Statute of Catalonia stipulated that ‘[t]he State shall guarantee that application of the levelling mechanisms shall in no case alter the position of Catalonia in the prelevelling ranking of per capita receipts’. This provision was upheld by the Spanish Constitutional Court, even if subject to its own authoritative interpretation.187

184 The Court thereby referred to 72 BVerfGE 330, 398 (Financial Equalization I) and 1 BVerfGE 117, 131 (Financial Equalization Act). 185 See W Leibfritz, ‘Fiscal Federalism’ (2009) 15. 186 For an analysis relating to the year 2004, see N Bosch and JM Durán, ‘The Financing System of Spanish Regions: Main Features, Weak Points and Possible Reforms’ in Bosch and Durán (eds), Fiscal Federalism (2008) 20. 187 STC 31/2010, FJ 134.

Equalisation 245 In other cases, exogenous factors have contributed to a deepening of controversies about equalisation. In Germany, reunification has, of course, increased fiscal disparities between the wealthiest and poorest Länder. Under the new constellation, all five new Länder are net receivers with the consequence of former net receivers becoming net payers and the traditional net payers, discontent already in the 1980s, being faced with an increasing financial burden. Unsurprisingly, the global financial crisis has, in a number of federal systems, hardened conflicts about the redistribution of increasingly scarce financial resources and in some cases, like that of Catalonia, has even fuelled secessionist tendencies.188 In Canada, it contributed to the resurfacing of an older debate about equalisation, especially in light of Alberta’s superior fiscal capacity and Ontario’s transformation in 2009 from a net payer into a net receiver. The crisis response, however, also tempered this debate because the federal bailout of the automobile industry benefitted Ontario primarily and thus in part countered its claim of being disadvantaged under the equalisation system. Overall, the image of Canada as a ‘giant mutual insurance company’189 still seems to enjoy relatively broad support in both richer and poorer parts of the country.190 Whether such an image holds equally true for all other federal systems discussed in this chapter is less clear. In any case, equalisation arrangements will always be bound to spark controversial debates about their sustainability and efficiency because their very rationale requires a positioning between two hallmarks of federalism: autonomy, on the one hand, and solidarity, on the other.

188

See ch 4.2.3. Saskatchewan premier Allan Blakeney, quoted in R Simeon et al, ‘The Resilience of Canadian Federalism’ in PE Peterson and DJ Nadler (eds), The Global Debt Crisis: Haunting US and European Federalism (Washington DC, Brookings Institution Press, 2014) 211. 190 See ibid 210. 189 Former

8 Prevention and Resolution of Conflicts

T

HIS CHAPTER DEALS with intergovernmental conflicts in federal systems and with how they are prevented or resolved. From a comparative perspective, conflicts may arise specifically in such systems for several reasons. These include, first, uncertainties regarding the scope of powers, especially in case of concurrent jurisdictions.1 Second, conflicts may result from disagreement regarding the interpretation of the distribution of powers or the constitution more generally. Third, they are often linked to subnational resistance against centralising trends, most notably those induced by flexibility clauses concerning the exercise of legislative powers and federal predominance in terms of taxing powers.2 The ways in which federal systems deal with intergovernmental conflicts can be classified as ‘prevention’ and ‘resolution’. The former encompasses a wide range of procedures and institutions of intergovernmental relations through which potential conflicts are prevented from erupting and leading to litigation in court (section 8.1). Prevention of disputes in this sense is also the purpose of mechanisms in some federal systems that seek to ensure a non-conflictual accommodation of subnational law in the overall legal framework. Such mechanisms include the ex ante scrutiny of ordinary laws and constitutions of subnational entities, which may take, however, a form of restrictive federal oversight (section 8.2). Despite such tools to prevent intergovernmental legal conflicts, they may not be avoided, of course, in any single case. This is where the judicial resolution of disputes comes into play, which centres around the power of constitutional review on federal grounds but also includes other court functions (section 8.3). 8.1 INTERGOVERNMENTAL RELATIONS

‘[I]n spite of their ubiquitous character and the impact they have on the lived reality of any federation, intergovernmental relations remain largely opaque to the public, scholars, and even sometimes to public authorities.’3 The fact that intergovernmental relations are of increasing importance overall, albeit with the usual variation in

1

See ch 5.2.2. See chs 5.2.3 and 7.2.3. 3 J Poirier and C Saunders, ‘Conclusion: Comparative Experience of Intergovernmental Relations in Federal Systems’ in J Poirier et al (eds), Intergovernmental Relations in Federal Systems: Comparative Structures and Dynamics (Oxford, Oxford University Press, 2015) 442. 2

Intergovernmental Relations 247 degree, hardly comes as a surprise. After all, this fact is due to changes in the real-life context of federal systems. Unlike more recent federal constitutions, those of the classical mature federations do not make references to intergovernmental relations simply because coordination between governments was considered, under the paradigm of dual federalism,4 as hardly necessary. A clear distribution of powers with ‘watertight compartments’ was still deemed feasible at that time and the (ostensibly few) cases of intergovernmental conflict would be resolved by constitutional adjudication. Yet, the context of federal systems changed dramatically with the arrival of the twentieth-century welfare state and regulatory state, which urgently required legislative and executive coordination in complex areas such as social welfare and healthcare or environmental protection.5 The increasing entanglement of governments and the resulting necessity of cooperation was recently recognised by the Belgian Constitutional Court even regarding a country that otherwise emphasises exclusive powers. 8.A Belgium 2014: Regional Competences and the Need for Intergovernmental Cooperation A decree of the Flemish region from 2009 on the definition of general provisions regarding energy policy included in its Article 9(1)(3) detailed regulations concerning the allocation of emissions certificates to aircraft operators. This aroused concerns on the part of the Council of State, which questioned the constitutionality of this provision and asked the Constitutional Court for a preliminary ruling. According to the Council of State, the Flemish region had exercised a competence alone that was so much entangled with federal powers that its exercise would have required cooperation with the federal Government. The Constitutional Court agreed with the Council of State.6 With reference to the reasoning in an earlier ruling that had invalidated a Flemish decree from 2004,7 the judges acknowledged that Article 6(1)(2) of the 1980 Special Law on Institutional Reforms assigned to the regions the power to prevent and fight against different forms of environmental pollution, among them air pollution. It also pointed out that the Belgian Constitution (Articles 5, 39 and 134) in conjunction with the just-mentioned Special Law of 1980 and the 1989 Special Law on the Institutions of Brussels had introduced a distribution of powers that relied on exclusive competences. In the Court’s view, this presupposes, however, that the matter addressed by a regional decree must be suited to being circumscribed in such a manner that it is fully within this exclusive competence. This requirement is not met in the case of emissions certificates for aircraft operators because a large part of emissions will inevitably concern the airspace of other regions, of Belgian sea areas under the control of the federal Government, of other EU Member States or even beyond the EU. As claimed by the Council of State, the questioned provision of the Flemish decree was unconstitutional, since the exercise of the competence would have required intergovernmental cooperation.

4

See chs 2.2.2, 5.2.2 and 5.3. See Part III. 6 Belgian Constitutional Court No 67/2014. 7 Belgian Constitutional Court No 33/2011. 5

248 Prevention and Resolution of Conflicts While the complexities of governance in the modern welfare state and regulatory state are key to the increase in intergovernmental relations, they are by no means the only driving force. Importantly, the need to fund governmental expansion into new policy fields has inevitably entailed an increase in mechanisms regarding financial issues. Another impetus for this rise has resulted from globalisation and, in some cases, from membership in the European Union, as coordination has become common regarding both the creation and implementation of international and supranational law.8 For these reasons, intergovernmental relations are ubiquitous in federal systems (and in this book). This section focuses more specifically on these relations and on how they contribute to the management of conflicts, which is undoubtedly one of their main functions. The inevitable interdependence nowadays between national and subnational governments and among the latter have brought into being both vertical and horizontal mechanisms.9 As we will see, these mechanisms may concern cooperation regarding different stages of the legislative cycle, ie either the making of law or its implementation. In the former case, they typically involve senior civil servants, ministers or even heads of government and are political in character with the inherent risk of posturing towards the electorate. In the latter case, cooperation is essentially technical because it is driven by specialised civil servants from various ministries. As a rule, vertical cooperation regarding the implementation phase is obviously more common in countries that adhere to the model of administrative federalism where subnational entities execute the bulk of national legislation. It is not by chance that the legal frameworks of Germany, Switzerland and South Africa foresee a number of collaborative mechanisms, even if they are effectively used to different degrees. In many regards, however, intergovernmental relations have been found to work in quite similar ways across most federal systems.10 They certainly differ due to the impact of context factors (eg ethno-cultural diversity) and federal design (eg the number of subnational entities), but nonetheless seem to have common traits. From a legal perspective, it is important to note that there is undoubtedly a trend towards the formalisation of intergovernmental relations. While civil law systems have, quite naturally, a greater propensity for formal, written rules than common law systems, other factors have been deemed relevant as well:11 first, firm entrenchment is sometimes sparked, especially in cases where authoritarianism has been recently experienced (eg Spain and South Africa), by the aim of setting up a legal bulwark against central government dominance; second, it may be favoured in the context of ethno-cultural diversity in order to compensate for, in such cases, a frequent lack of trust in informal mechanisms; third, formalisation may be a response 8

See ch 14.4 and 14.5. section focuses on vertical mechanisms between national and subnational governments. In some federal countries, local governments also play an important role in intergovernmental relations. See ch 9.3.2.2. 10 See A Trench, ‘Intergovernmental Relations: In Search of a Theory’ in SL Greer (ed), Territory, Democracy and Justice: Regionalism and Federalism in Western Democracies (London, Palgrave Macmillan, 2005) 254. 11 See Poirier and Saunders (eds), ‘Conclusion’ (2015) 488f. 9 This

Intergovernmental Relations 249 to a perception of intergovernmental relations as lacking democratic accountability and transparency, which is expressed in the Canadian context through the label of ‘federalism behind (almost) closed doors’;12 finally, the factor of history is, as mentioned at the outset of this section, another important variable. Indeed, many more recently established federal systems espouse, in contrast to mature federations, a greater penchant for formalising intergovernmental relations in ordinary or even constitutional law, with Chapter 3 of the South African Constitution being a particularly notable example. 8.1.1 Principles: Federal Loyalty and Similar Precepts The practice of intergovernmental relations is often based on underlying principles that are called, with certain variations, principles of mutual consideration, federal loyalty or comity and the like. These are very common, especially in countries following the model of administrative federalism. This is because the coordination of national legislation and subnational implementation quite evidently requires a high degree of mutual understanding and cooperation. An exceptional case is Belgium, where the principle of comity, introduced in 1993 and justiciable since 2012, operates in a dual federal system. In this specific case, the need for such a principle largely follows from the complexity of the tripartite distribution of powers between the federal Government, communities and regions. As Article 143(1) of the Belgian Constitution stipulates, all governments must ‘[i]n the exercise of their respective responsibilities … act with respect for federal loyalty, in order to prevent conflicts of interest’.13 Whereas this provision clearly relates only to the exercise of each government’s own competences, the federal loyalty enshrined in the Swiss Constitution is more far-reaching. The latter first states that the federal Government and the cantons ‘shall support each other in the fulfilment of their duties and shall generally cooperate with each other’ (Article 44(1)) and then, more concretely, that they ‘shall provide each other with administrative assistance and mutual judicial assistance’ (Article 44(2)). Moreover, the cantons are also obliged to accept, in nine policy areas of common interest, that the federal Government may declare inter-cantonal agreements, at the request of the interested cantons, to be generally binding or require cantons to participate, also in financial terms (Article 48a). Similarly extensive is the entrenchment of certain principles of cooperative government and intergovernmental relations in South Africa. These entail a number of very specific duties for all spheres of government, ie the national, provincial and local spheres (Article 41(1)), and require national legislation to further regulate institutions and procedures of intergovernmental relations (Article 41(2)).

12 TO Hueglin, ‘Canada: Federalism behind (Almost) Closed Doors’ in R Blindenbacher and A Ostien (eds), Dialogues on Legislative, Executive and Judicial Governance in Federal Countries (Montreal, McGill–Queen’s University Press, 2006) 13ff. 13 See also Belgian Constitutional Court Nos 49/1994 and 42/1997.

250 Prevention and Resolution of Conflicts In contrast to the above-mentioned cases of express entrenchment in the constitutional text, principles of federal loyalty are often developed by constitutional jurisprudence. A paradigmatic example is Germany, where federal loyalty (Bundestreue) is usually traced back to an article published more than a century ago about such an unwritten constitutional principle regarding the relations between the German Empire and its constituent units.14 In its interpretation of the Basic Law, the Federal Constitutional Court has further developed this principle in a set of important judgments. 8.B Germany 1952–95: The Principle of Federal Loyalty (Bundestreue) From an early decision in 1952 onwards, constitutional jurisprudence has developed the principle of federal loyalty (Bundestreue), according to which both the federal Government and the Länder are subject to a ‘legal duty … to federal-friendly behaviour’.15 As explained in more detail in a subsequent judgment, this principle was derived from the very nature of the federal compact (foedus), which would imply for all parties involved a duty of mutual consideration in order to consolidate the compact.16 It is important to note that the unwritten principle of federal loyalty is regarded as being subordinate to written constitutional law and accessory in nature so that it is alone not a sufficient basis for creating rights and obligations. Even though the Federal Constitutional Court has recognised that some provisions of the Basic Law are expressions of federal loyalty (eg Articles 20, 35, 36, 72(2), 106(3–4) and 109(1)), the principle is thus more abstract in its origins and inherent in the federal constitutional order itself. Yet, the Court has over the years distilled from the principle several rather concrete duties. These include certain limits in the exercise of powers, the obligation of active cooperation (also in financial terms), procedural requirements and duties of consideration in the exercise of EU-related competences.17

In Austria, there is similarly a judicially developed principle of federal loyalty. Even if some traces of the so-called principle of mutual consideration (Berücksichtigungsprinzip) may be identified in certain ordinary laws (eg regarding spatial planning), it is, as a general precept, nonetheless based on a set of Constitutional Court decisions.18 It entails that in exercising their legislative powers, the federal Government and the Länder are barred from negating the interests of another government and undermining laws adopted by another parliament. For instance, in a case involving a railway tunnel between two Länder, the Constitutional Court in 1999 invalidated part of Lower Austria’s Law on Nature Protection.19 In enacting this law, the Parliament of

14 See R Smend, ‘Ungeschriebenes Verfassungsrecht im monarchischen Bundesstaat’ in Festgabe für Otto Mayer: Zum 70. Geburtstag dargebracht von Freunden, Verehrern und Schülern (Tübingen, Mohr, 1916). 15 1 BVerfGE 299, 315 (Housing Funding). 16 6 BVerfGE 309 (Concordat case); 8 BVerfGE 122, 138 (Hesse Referendum case). 17 On these four categories of duties see, respectively: 61 BVerfGE 149, 205 (State Liability); 72 BVerfGE 330, 386 (Financial Equalization I) and box 7.N; 81 BVerfGE 310, 334 (Kalkar II) and box 12.C; 92 BVerfGE 203, 239 (EC-Television-Directive). 18 Eg VfSlg 8831/1980; VfSlg 10292/1984. 19 VfSlg 15552/1999.

Intergovernmental Relations 251 that Land had, according to the judges, failed to consider the public interest of the country as a whole and attempted to make the realisation of an infrastructure project that was clearly within federal competence impossible. Even though the principle of mutual consideration primarily concerns the legislative sphere, it is also relevant for implementation. It mandates that laws be applied in a way that is consistent, as much as possible, with regulations made by other governments.20 Spain is another country where the national Constitution is silent on principles of loyalty. It was the Constitutional Court that recognised duties of cooperation and loyalty to the Constitution (fidelidad a la constitución).21 This means that, in conjunction with the principle of legal certainty (Article 9(3) of the Spanish Constitution), governments must not obstruct each other in exercising their competences so as to avoid confusion among those who are subject to their laws. Spain is not the only regional state where constitutional jurisprudence has developed a principle of cooperation. In Italy, it was likewise a series of rulings that gave rise to the recognition of an inherent principle of loyalty.22 8.C Italy 1985–2001: The Principle of Loyal Cooperation (leale collaborazione) After several decisions that had emphasised the necessity of cooperation and integration between the levels of government,23 in 1985 the Italian Constitutional Court explicitly formulated a principle of loyal cooperation. The national Government had issued an order, according to which all public works projects had to be preliminarily submitted to the then-Ministry for Cultural and Natural Heritage. Several regions had claimed at the time that this order violated their constitutional powers, especially in the field of environmental protection, as well as land and urban planning. The Constitutional Court rejected the claim of the regions because it considered the governmental order not to be a violation of regional powers, but rather the expression of the principle of ‘loyal cooperation’.24 This principle would be implicit and indispensable in a constitutional order with two levels of government. It would be expressed primarily, as in the case concerned, in the duty to mutual information. The principle of loyal cooperation was further developed by the Constitutional Court in a number of subsequent rulings, specifying its procedural facets and defining it as ‘one of the fundamental values of the constitution’.25 The principle became so entrenched in the Italian constitutional order that it was also partly constitutionalised in the reform that changed the relations between the national Government and the regions in 2001. Accordingly, Article 120(2) of the Italian Constitution provides that the national Government may act for the regions in certain particular circumstances. However, it must be ensured that such subsidiary powers ‘are exercised in compliance with the principles of subsidiarity and loyal cooperation’.

20

VfSlg 14534/1996. STC 18/1982; STC 11/1986. 22 For a comparison with Germany, see J Woelk, Konfliktregelung und Kooperation im italienischen und deutschen Verfassungsrecht: ‘Leale collaborazione’ und Bundestreue im Vergleich (Baden-Baden, Nomos, 1999). 23 See, eg, Italian Constitutional Court Nos 49/1958 and 219/1984. 24 Italian Constitutional Court No 358/1985. 25 See, eg, Italian Constitutional Court judgments Nos 153/1986 and 42/2001. 21

252 Prevention and Resolution of Conflicts Irrespective of whether principles of federal loyalty are entrenched in the constitutional text or judge-made, they entail a general obligation for governments to cooperate. In some cases, a constitution’s emphasis on such cooperation is further reflected in the fact that consensual solutions to conflicts are explicitly preferred to litigation in court. In Switzerland, for instance, disputes between the cantons or between them and the federal Government ‘shall wherever possible be resolved by negotiation or mediation’ (Article 44(3) of the Constitution). In South Africa, legal provisions for pre-judicial conflict resolution through negotiations are, as for intergovernmental relations in general, particularly extensive. Parties to a conflict are obliged to make ‘every reasonable effort to settle the dispute’ through intergovernmental mechanisms and ‘must exhaust all other remedies before it approaches a court to resolve the dispute’ (Article 41(3) of the Constitution). Such mechanisms are provided in various specific sectoral laws, such as the Division of Revenue Act,26 and they are extensively regulated in Chapter 4 of the 2005 Intergovernmental Relations Framework Act, which applies by default in all other cases. Importantly, if parties fail to comply with intergovernmental mechanisms and do not satisfy the abovementioned criterion of ‘every reasonable effort to settle the dispute’, the Constitutional Court may refer the dispute back to the institutions involved (Article 41(4)). As the Court itself put it, the constitutional principles of cooperative government entail a duty, where possible, to ‘avoid legal proceedings against one another’27 and that this duty is judicially enforced. In contrast to the Swiss and South African cases, there is no principle of intergovernmental cooperation in the United Kingdom that is laid down in written law. Yet, such cooperation occurs de facto very extensively instead of immediately involving the Supreme Court. It is one of the reasons why the latter has so far only played a fairly marginal role regarding devolution issues.28 The fact that collaboration is, like in the above-mentioned cases, prioritised over litigation, even without a codified duty is the result, first, of the implementation of devolution being governed by so-called concordats. These are informal and mostly bilateral agreements between the UK Government and the devolved administration. Such concordats operate within the framework of a general Memorandum of Understanding that outlines certain principles29 and represent a form of non-justiciable bureaucratic law. Second, the Joint Ministerial Committee (JMC), an institution composed of the UK Prime Minister and the First Ministers of the devolved administrations, has acted as a ‘dispute resolution machinery of a political kind’30 and reduced the inclination for litigation in court. The UK example thus clearly demonstrates that both instruments and institutions of intergovernmental relations are, sometimes even without a

26

See ch 7.1.3. Uthukela District Municipality v President of the Republic of South Africa 2003 (1) SA 678, para 13. 28 See P Leyland, ‘The Multifaceted Dynamics of UK Devolution’ (2011) 9 International Journal of Constitutional Law 251, 255. 29 These include good communication and information sharing, early warning of policy initiatives and cooperation on matters of mutual interest and confidentiality. 30 R Hazell, ‘Out of Court: Why Have the Courts Played No Role in Resolving Devolution Disputes in the United Kingdom?’ (2007) 37 Publius 578, 581. 27

Intergovernmental Relations 253 principle, crucial elements of conflict management and merit further analysis in the two following sections. 8.1.2 Institutions: The Predominance of the Executive Branch In federal systems, institutions of intergovernmental relations are typically composed of members of the executive branch rather than legislators. It is not surprising, therefore, that the workings of such institutions have sometimes become almost equated with what is known, in Canadian and Australian terminology, as ‘executive federalism’.31 It is not by coincidence, of course, that this equation has become commonplace precisely in these two federations with parliamentary systems of government. In such systems, where the executive typically commands a majority in the legislature, which it is accountable to, it is certainly easier to secure parliamentary implementation of executive bargains. Such legislative compliance is far from certain in presidential systems where intergovernmental relations thus tend to be dualistic. Some mechanisms then involve executives and others legislators.32 As far as the vertical dimension is concerned, cooperation between legislatures is, from a comparative perspective, without doubt quite a rare phenomenon. Only in some federal countries, such as India and South Africa, do conferences of speakers exist for the sake of sharing information and experiences. The South African Speakers’ Forum, an association composed of representatives from the National Assembly, the National Council of Provinces and from all the provincial legislatures, is a particularly interesting institution, especially in view of its function in coordinating legislative action. In a recent case before the Constitutional Court, the limits of this function were at stake. 8.D South Africa 2011–12: Coordination of Legislative Action and Its (Narrow) Constitutional Limits The 1999 Financial Management of Parliament Act (FMPA) provided that provincial legislation would regulate the financial management of their legislatures. It was the Speakers’ Forum which oversaw the elaboration of a generic bill for the provinces that each of them could then adapt to its specific circumstances. After the Speakers’ Forum had sanctioned the draft bill, laws were enacted in 2009 by the Free State, Eastern Cape and Gauteng, as well as in 2010 by Mpumalanga and Limpopo. When the premier of the province of Limpopo had doubts regarding the legislature’s competence to enact this law, he followed the procedure under Section 121 of the Constitution and referred the bill back to the legislature and, upon the latter’s insistence on the draft law, to the Constitutional Court for a review of its constitutionality.

31 D Smiley, The Federal Condition in Canada (Toronto, McGraw-Hill Ryerson, 1987) 83. Confusingly, the term executive federalism is sometimes also used to describe what is, in this book, called administrative federalism. On this, see ch 5.3. 32 See N Bolleyer, Intergovernmental Cooperation: Rational Choices in Federal Systems and Beyond (Oxford, Oxford University Press, 2009) 111ff.

254 Prevention and Resolution of Conflicts

In 2011, the Court held that Limpopo’s law was indeed beyond the competence of the provincial legislature33 and one year later invalidated the laws of five other provinces for the same reason.34 Section 104(1)(b) of the South African Constitution stipulates that the legislative authority of the provinces pertains to the functional areas listed in (i) Schedule 4 or (ii) Schedule 5, (iii) to any matter ‘expressly assigned … by national legislation’ and (iv) ‘any matter for which a provision of the constitution envisages the enactment of provincial legislation’. There were no doubts among the judges that Limpopo’s provincial power could not be based on one of the first three clauses. As to the fourth clause, the Court interpreted what ‘the constitution envisages’ very narrowly and held that this requires, much like the term ‘expressly assigned … by national legislation’ under the third clause, that provincial legislation be provided ‘in clear terms’ (para 51). This interpretation excluded implied provincial powers and flexible coordination of legislative action (by the Speakers’ Forum) on the fringes of competence areas. The collaboration regarding the enactment of the FMPA and the various provincial laws had actually seemed to be a good example of such legislative coordination, which is even constitutionally required (Section 41(1)(h)(iv)). However, the rigid interpretation of the Court pursued, concerning the assignment of provincial powers, ‘the goal of complete clarity’, which is ‘not only “chimera”, but also contradicts the basic framework of multi-level government’.35

Cooperation between legislatures appears to be somewhat more common and institutionalised in the horizontal dimension. Examples extend from the US National Conference of State Legislatures (NCSL) which involves all state lawmakers, to the Conferences of the Speakers of the Länder parliaments in Austria and Germany. Even if these institutions fulfil, to some extent, the function of settling disputes between subnational entities and of coordinating positions vis-à-vis the national government, they are regarding neither of these two issues as crucial as executive forums. Indeed, horizontal cooperation, driven primarily by executives, has recently been identified as one of the major emerging trends in intergovernmental relations.36 Certainly, collaboration between subnational entities is not necessarily a new phenomenon in all federal countries. In Germany, for instance, it can be traced as far back as the alliances of princes in the nineteenth century.37 In several mature aggregative federations, horizontal cooperation naturally preceded vertical forums, not least because they played, like the 1864 Charlottetown Conference in the Canadian case, a decisive role in the very process of establishing the federal system. Nevertheless, the last two decades have witnessed a sharp increase in institutionalised forums of subnational executives. Cases in point are Canada, with the formalisation of earlier interprovincial meetings, by establishing in 2003, at the initiative of Quebec, 33

Premier: Limpopo Province v Speaker: the Limpopo Provincial Legislature 2011 (6) SA 396 (CC). Premier: Limpopo Province v Speaker of the Limpopo Provincial Legislature 2012 (4) SA 58 (CC). 35 RF Williams and N Steytler, ‘Squeezing out Provinces’ Legislative Competence in Premier: Limpopo Province v Speaker: Limpopo Provincial Legislature & Others I and II’ (2012) 129 South African Law Journal 621, 635ff. 36 See Poirier and Saunders (n 3) 484. 37 See ch 3.3.3. See especially G Lehmbruch, ‘Der unitarische Bundesstaat in Deutschland: Pfadabhängigkeit und Wandel’ in A Benz and G Lehmbruch (eds), Föderalismus (Wiesbaden, Westdeutscher Verlag, 2002) 86ff. 34

Intergovernmental Relations 255 the Council of the Federation and, modelled on it, the Council for the Australian Federation created in 2006. Moreover, recent years have also seen enhanced horizontal collaboration that is restricted to only some subnational entities and specific policy areas. Such doubly limited cooperation finds fertile ground especially where ideological cleavages between subnational entities are so wide that they make broader collaboration impossible. The California-driven Western Climate Initiative established in 2007 is, in this regard, a prime example.38 This quantitative trend regarding the institutionalisation of horizontal councils should not obscure the fact, however, that many of them have had a rather modest impact in practice. An exception seems to be Switzerland, where the small size of the cantons and their common impulse to counteract centralisation has been found to account for an extraordinary degree of horizontal cooperation.39 In fact, there has been a significant increase in such collaboration over the last three decades, especially with the establishment in 1993 of a forum involving the leaders of the country’s executives, ie the Conference of Cantonal Governments.40 This trend was formalised in 1999 with the new Article 48(1) of the Constitution, according to which the cantons may ‘establish common organisations and institutions’. All existing horizontal conferences have recently undergone a process of institutionalisation with permanent structures and have been accommodated since 2008 in a newly established House of the Cantons in Bern. While Switzerland seems to be particularly effective concerning horizontal cooperation between executives, the latter are generally far more engaged in collaboration in the vertical dimension. It has been found that this is much less the case in presidential systems, such as the United States, Argentina, Brazil and Nigeria, where top-level summits of government leaders occur only on an ad hoc basis.41 In the numerous parliamentary federal systems, however, such forums typically play a rather significant role, even if this is subject to certain fluctuations. In certain mature parliamentary federations, meetings between national and subnational government leaders have a tradition lasting many decades. In Canada, they date back to the early twentieth century and have given rise to what has been famously described, due to a certain resemblance to international relations, as ‘federal-provincial diplomacy’.42 Since the early 2000s, however, the previously more regular First Ministers’ Conferences, the country’s main intergovernmental forum, have been convened only occasionally. In the Australian case, the early days of the federation soon after 1901 already witnessed so-called premiers’ conferences. Since their institutionalisation in 1992 as the Council of Australian Governments (COAG) had marked a culmination

38

See ch 12.1. See D Bochsler, ‘Neighbours or Friends? When Swiss Cantonal Governments Co-operate with Each Other’ (2009) 19 Regional and Federal Studies 349. 40 See P Hänni et al (eds), 20 Jahre KdK—Les 20 ans de la CdC—20 anni della CdC—20 onns CdC (Bern, Stämpfli, 2013). 41 See Poirier and Saunders (n 3) 460. 42 R Simeon, Federal–Provincial Diplomacy (Toronto, University of Toronto Press, 1972). 39

256 Prevention and Resolution of Conflicts of ‘collaborative federalism’, this council has lost much of its importance and has become increasingly dominated by the national Prime Minister.43 Both the Canadian and Australian cases thus reveal that vertical cooperation has had significant ups and downs over time even in countries with a long tradition. Both horizontal and vertical executive forums are faced with the challenge of reconciling democratic accountability and transparency with what essentially makes them relevant, ie effectiveness in coordinating governmental action and thus preventing intergovernmental conflicts. Accountability, it has been pointed out, is a particularly complex issue in parliamentary federal systems, where collaboration between executives is most prevalent. On the one hand, there is the principle of responsible government that binds executives to their respective legislatures. On the other hand, there is for executives involved in intergovernmental relations also a certain responsibility vis-à-vis their counterparts according to the principle of pacta sunt servanda.44 The effectiveness of intergovernmental bodies is likewise a complex issue, especially from a legal perspective. Even though it seems intuitive that formalisation facilitates cooperation and thus contributes to success by enhancing compliance and predictability, this does not hold true in all cases. The above-mentioned informal First Ministers’ Conferences and other forums in Canada could be considered, at least until the early 2000s, as highly effective. In India, the only intergovernmental body that is mentioned by the national Constitution (Article 263), ie the Inter-State Council (which also includes federal ministers), has been found clearly less relevant than the informal chief ministers’ conferences.45 Sometimes, vertical executive forums are of considerable relevance, even if the constitution places an emphasis on the role of the second chamber as well. The latter is then, as in the case of Germany’s Federal Council and South Africa’s National Council of Provinces, complemented by such mechanisms. Examples are meetings of the German Chancellor with the leaders of the Länder governments (§ 31 of the Standing Orders of the Federal Government) that take place at least twice a year, often subsequent to the horizontal conference of the Länder government leaders (Ministerpräsidentenkonferenz). In South Africa, there are ‘extended cabinet’ meetings, involving the national and provincial executives as well as organised local government, and additional bodies foreseen by the Intergovernmental Relations Framework Act of 2005.46

43 See M Painter, Collaborative Federalism: Economic Reform in Australia in the 1990s (Melbourne, Cambridge University Press, 1998); G Anderson, ‘The Council of Australian Governments: A New Institution of Governance for Australia’s Conditional Federalism’ (2008) 31 University of New South Wales Law Journal 493. 44 See R Simeon and A Nugent, ‘Parliamentary Canada and Intergovernmental Canada: Exploring the Tensions’ in H Bakvis and G Skogstad (eds), Canadian Federalism: Performance, Effectiveness, and Legitimacy (Oxford, Oxford University Press, 2001) 59. 45 See MP Singh and R Saxena, ‘Intergovernmental Relations in India: From Centralization to Decentralization’ in Poirier et al (eds), Intergovernmental Relations (2015) 252ff. 46 See D Powell and P Ntliziywana, ‘“South Africa Inc”: The Rise of the Developmental State and the Corporatization of Intergovernmental Relations’ in F Palermo and E Alber (eds), Federalism as DecisionMaking: Changes in Structures, Procedures and Policies (Leiden, Brill–Nijhoff, 2015) 299ff.

Intergovernmental Relations 257 8.1.3 Instruments: Delegation and Intergovernmental Agreements Similar to the institutions of intergovernmental relations, their instruments may also be categorised according to whether they pertain primarily to the legislative or executive branch of government. As with the institutions explored in the preceding section, the executive instruments are those more prevalent and important. Yet, certain mechanisms between legislatures merit attention. In some countries, for instance, there is the possibility of delegating legislative competences that may serve purposes of collaboration. An example is Section 51 (xxxvii) of the Australian Constitution, which empowers the national Parliament to make laws on matters referred to it by the legislatures of one or more states. Such laws then ‘extend only to states by whose parliaments the matter is referred, or which afterwards adopt the law’. While the Australian states had usually refrained from referring legislative powers, they became more inclined to do so in the late-1990s.47 This occurred after the High Court had restricted, in two decisions, the use of other cooperative mechanisms by invalidating, respectively, the conferral of judicial and administrative state powers to federal institutions48 and thus jeopardised a joint federal–state regime regarding corporation affairs. Since then, significant references have been made under Section 51 (xxxvii) in such areas as corporation affairs, anti-terrorism49 and industrial relations. The legal situation in Australia contrasts with that in Canada, which can be traced back to two Supreme Court rulings from the early-1950s. First, the judges held that (even revocable) inter-delegation of legislative powers between the federal and provincial legislatures would change the distribution of powers without a formal amendment and was thus unconstitutional.50 Then, however, the Court upheld the transfer of federal powers to a provincial marketing board, ie a non-legislative provincial body. Administrative inter-delegation to the benefit of such bodies has since then been widely used. For such transfers of competences, the Supreme Court does not even require the federal Government to commit the provinces to a basic regulatory framework.51 While Australia and Canada thus treat the issue of delegation very differently, the United Kingdom represents, due to the mutual effects between parliamentary sovereignty and devolution, a case of its own. With the UK Parliament being regarded as sovereign, it remains fully competent to enact laws even concerning devolved matters, and these prevail over acts adopted by the devolved legislatures. But in order to prevent such legislative intervention from becoming common practice, which would undermine devolution, an agreement was made in the course of Scottish devolution that came to be known as the ‘Sewel convention’. According to this convention, which was extended to Welsh and Northern Irish devolution, the UK Parliament may

47 See C Saunders, The Constitution of Australia: A Contextual Analysis (Oxford, Hart Publishing, 2010) 252f. 48 Re Wakim; Ex parte McNally (1999) 198 CLR 511; R v Hughes (2000) 202 CLR 535. 49 Thomas v Mowbray (2007) 233 CLR 307. 50 Nova Scotia (AG) v Canada (AG) [1951] SCR 31. 51 See PW Hogg, Constitutional Law of Canada, 5th edn (Scarborough, Carsweel, 2000) 14-16ff.

258 Prevention and Resolution of Conflicts make laws only upon prior consent of the respective devolved legislature through a legislative consent motion (or ‘Sewel motion’). Such (consented) legislation from Westminster has occurred to an unexpectedly large extent in certain areas.52 Adding a new provision to the original Scotland Act of 1998, the 2016 Scotland Act gave the ‘Sewel convention’ a statutory basis. While the latter’s Article 28(7) reinforces the UK Parliament’s powers regarding Scotland, the new Article 28(8) says explicitly that ‘it is recognised that the Parliament of the United Kingdom will not normally legislate with regard to devolved matters without the consent of the Scottish Parliament’. In contrast to federal states like Australia and Canada, the issue at stake is therefore not the delegation of subnational entities’ constitutionally entrenched own powers, but the concession of a veto right against the exercise of central legislative powers by virtue of parliamentary sovereignty. Much more common than legislative instruments are intergovernmental mechanisms driven by executives, above all intergovernmental agreements. Although such agreements are particularly prevalent in some federal countries, such as Switzerland, Brazil and Spain, they are of increasing importance overall53 and may have innumerable functions. As for vertical accords, they typically have the following purposes: to regulate, in case of administrative federalism, the subnational implementation of national law, to coordinate governmental action in fields of concurrent power (eg immigration and integration agreements in Canada)54 and, most problematically, to create a framework for the exercise of the federal spending power in areas where subnational entities depend on such transfers (eg the growing practice of memoranda of understanding in India).55 These sorts of interventionist agreements sometimes face constitutional limits policed by the courts. In Spain, for instance, vertical accords are particularly common regarding social welfare, but they may not be devised in such a way that the autonomous communities are prevented from determining their own policies.56 Horizontal agreements, by contrast, have quite different functions, such as coordinating legislative initiatives and joint service provision or staving off centralisation with subnational entities voluntarily relinquishing part of their individual autonomy in a certain area for the sake of joint policymaking. Whether vertical agreements are predominantly bilateral or multilateral largely reflects a federal system’s penchant for either symmetry or asymmetry. Whereas accords between the federal Government and a single Land are quite exceptional in Germany,57 bilateral agreements are essential to intergovernmental relations in countries like Canada and Spain. Irrespective of the function, the horizontal or vertical dimension or the number of parties involved, a fundamental issue from a lawyer’s perspective is that of the

52 See A Page and A Batey, ‘Scotland’s Other Parliament: Westminster Legislation about Devolved Matters in Scotland since Devolution’ (2002) Public Law 501. See ch 11.5. 53 See Poirier and Saunders (n 3) 469. 54 See ch 13.2. 55 See Singh and Saxena, ‘Intergovernmental Relations’ (2015) 259. 56 STC 13/1992. See ch 7.3.2 and box 7.G. 57 See C Vedder, Intraföderale Staatsverträge: Instrumente der Rechtsetzung im Bundesstaat (BadenBaden, Nomos, 1996).

Intergovernmental Relations 259 legal basis for concluding intergovernmental agreements. While some federal constitutions remain completely silent on such accords (eg Canada), others include only rudimentary references to certain specific types of agreements and not an extensive general legal regime. Examples of such scarce provisions are Article 1, section 10, clause 3 of the US Constitution on interstate compacts requiring the consent of Congress, Article 105A of the Australian Constitution regarding federal–state agreements on state debts and Article 145(2) of the Spanish Constitution on legally binding accords (convenios) between the autonomous communities. In this context, the regulation of both vertical and horizontal agreements in Article 15a of the Austrian Constitution seems comparatively comprehensive. As with the distribution of powers, Belgium follows its very specific pattern of deconstitutionalisation and regulates intergovernmental agreements in the 1980 Special Law on Institutional Reforms, which it does very extensively in Article 92. Given the widespread scarcity of textual sources, there remains the question of the role of the courts in providing a legal basis. In Canada, besides this lack of a textual constitutional provision, the case law of the Supreme Court has also not provided a stable and reliable basis. In fact, it has sometimes been criticised as being quite inconsistent.58 Agreements are thus mostly left to the political sphere. This is, however, by no means an indication that they lack importance. In Canada specifically, intergovernmental accords are in practice an essential feature in many policy fields.59 Like many other federal constitutions, Germany’s Basic Law contains only a few references to intergovernmental agreements, and they all concern the specific issue of territorial restructuring (Articles 29, 118 and 118a). Yet, constitutional jurisprudence has proved more relevant than, for example, in the Canadian case. First, the power of the Länder to conclude agreements is widely accepted on the basis of their status as co-sovereign ‘states’ (Staatsqualität), accorded to them by the Federal Constitutional Court.60 Second, the Court has in some cases even actively pushed for the conclusion of an intergovernmental treaty. 8.E Germany 1972: Intergovernmental Cooperation Regarding University Admission In the German system of higher education, a sharp increase beginning in the 1950s in applications for admission to very popular degree programmes produced, in combination with a variety of admission restrictions, considerable legal uncertainty. In particular, this involved the question of whether the different Länder rules regarding admission, adopted on the basis of federal framework legislation (former Article 75(1) of the Basic Law), were consistent with the Constitution. Two unsuccessful applicants for medical studies in Hamburg and Munich eventually took this matter to the courts. When legal proceedings arrived at the Federal Constitutional Court, the judges held that the individual right to occupational freedom (Article 12(1) of the Basic Law) and

58 With reference to UL Canada Inc v Quebec (AG) [2005] 1 SCR 143 and Boucher v Stelco Inc [2005] 3 SCR 279, see M-A Adam et al, ‘Intergovernmental Relations in Canada: Competing Visions and Diverse Dynamics’ in Poirier et al (n 3) 164. 59 See, eg, chs 11.2 and 13.2. 60 36 BVerfGE 342, 360f (Remuneration Law of Lower Saxony). See ch 4.1.2.

260 Prevention and Resolution of Conflicts

the equality rule (Article 3) would guarantee, within certain limits, a right to university admission.61 More importantly for intergovernmental relations, they also ruled on how to achieve harmonised admission standards necessary for conformity with the abovementioned fundamental rights. In quite explicit terms, the Court said that, rather than federal legislation, a treaty between the Länder would be the preferable option to remedy the unconstitutional status quo (p 357). Consequently, the ruling resulted in the conclusion, in 1972, of the Treaty on the Admission to University Education, which has since been amended several times. This treaty established criteria and procedures regarding many degree programmes and set up a central agency for university admission. The legal status of this agency was altered only recently, and its responsibilities are now only related to a few highly popular degree programmes.

In Germany, there are also simple administrative agreements in addition to the justmentioned formal treaties. Likewise, other federal countries use different notions for intergovernmental agreements that designate distinct types with several legal implications. In the United States, for example, there are administrative agreements, on the one hand, and compacts on the other, which differ from each other in many ways. Only compacts are subject to the approval of the state legislatures concerned, are legally binding, may establish special agencies and prevail over conflicting state law.62 Similar differences exist in Spain, where it is of crucial importance whether governments conclude acuerdos or convenios. While the former contain mere political commitments, only the latter are legally binding for the parties. As to how intergovernmental agreements concluded by executives may be accorded legal effect more broadly, there appear to be two main patterns: legislative ratification or legislative implementation.63 In the first case, the mere act of ratification provides an accord with the force of law. This applies to treaties in Germany and compacts in the United States. In the second case, an agreement itself never achieves the status of law capable of creating rights and obligations for third parties. This, however, does not prevent the legislatures of all governments involved from implementing its content by passing parallel mirror legislation. At the end of this process of joint law-making, it is this legislation that has legal effect. This holds true for most common law systems, except US compacts, as well as for Austria. In the latter case, it was the Constitutional Court that ultimately declared agreements concluded under the above-mentioned Article 15a of the Constitution as not directly applicable and, just like international treaties, subject to transposition.64 A very specific case that partly defies both patterns described is Switzerland. According to Article 48(4) of the Constitution, the cantons may, under certain conditions, ‘by inter-cantonal agreement authorise inter-cantonal bodies to issue

61

33 BVerfGE 303 (Numerus Clausus I). JF Zimmerman, Interstate Cooperation: Compacts and Administrative Agreements, 2nd edn (Albany, SUNY Press, 2012). 63 See Poirier and Saunders (n 3) 472f. 64 VfSlg 9886/1983. 62 See

Ex ante Scrutiny of Subnational Law 261 legislative provisions that implement an inter-cantonal agreement’. This creates a new legal source, ie that of inter-cantonal law, with which the cantons are constitutionally obliged to comply (Article 48(5)). Although there have so far been only a few examples of bodies, such as the Swiss Conference of Cantonal Directors of Education, that actually enact inter-cantonal legislation,65 the idea seems highly innovative. Regardless of whether an intergovernmental agreement has legal effect, subsequent legislation may in some cases unilaterally derogate from or abrogate such accords. This follows in Canada from parliamentary sovereignty, which trumps the principle of pacta sunt servanda, and loyalty towards an agreement concluded with partners in the federation. As the executive typically commands a parliamentary majority, any party may thus easily alter an agreement or even withdraw from it. 8.F Canada 1991: Parliamentary Sovereignty and Intergovernmental Agreements Under the Canada Assistance Plan, a federal law initially enacted in 1966, the federal Government covered half of the costs of social welfare programmes in the provinces.66 Section 8 of this law foresaw bilateral cost-sharing agreements that could only be amended, according to both the federal law and each bilateral agreement, with the consent of the federal Government and the province concerned. When the federal Government in 1990 introduced, as a cost-cutting measure, a cap on financial transfers with regard to the three wealthiest provinces—Ontario, Alberta and British Columbia—the latter province filed legal proceedings raising two questions: first, whether the federal Government had the authority to limit its (financial) obligations and, second, whether the terms of the bilateral agreement and the conduct of the federal Government in compliance with it had entailed ‘legitimate expectations’ that no bill on such limitations would be introduced in the federal Parliament without British Columbia’s consent. The Supreme Court held that the federal Government had acted in full compliance with the Constitution.67 Notwithstanding the intergovernmental agreements with the provinces, it was free to amend the Canada Assistance Plan. According to the judges, the doctrine of ‘legitimate expectations’ was trumped by the doctrine of parliamentary sovereignty, implying full power to enact, repeal or amend a legislative act. It pointed out explicitly that ‘[a] restraint on the executive in the introduction of legislation is a fetter on the sovereignty of parliament itself’ (p 560). 8.2 EX ANTE SCRUTINY OF SUBNATIONAL LAW

8.2.1 Judicial Scrutiny of Draft Legislation Most constitutional adjudication in federal systems is related to enacted laws.68 In some cases, however, it may concern draft legislation as well. Such pre-enactment

65 See T Pfisterer, ‘Intergovernmental Relations in Switzerland: An Unfamiliar Term for a Necessary Concept’ in Poirier et al (n 3) 408. 66 See ch 11.2. 67 Reference Re Canada Assistance Plan (BC) [1991] 2 SCR 525. 68 See section 8.3 below.

262 Prevention and Resolution of Conflicts scrutiny is or was a feature of some regional states such as Italy and in the specific case of devolution in the United Kingdom. In Italy, the national Government could, until the constitutional reform of 2001, challenge regional bills prior to their adoption. Only since then has it been limited to requesting constitutional review of already-adopted laws under Article 127 of the Constitution.69 In the United Kingdom, there are explicit procedural safeguards to ensure that devolved legislation is not in breach of the distribution of powers as outlined in the devolution acts. Section 35 of the 1998 Scotland Act and Section 112(1) of the Government of Wales Act allow the national Government to refer to the Supreme Court the question of whether a bill or any provision of a bill would be within the legislative competence of the devolved legislatures. 8.G United Kingdom 2014: A Bill from the Welsh Assembly Beyond Its Competence? The Welsh Assembly passed a bill aimed at retaining a regime for the regulation of agricultural wages in Wales, which had just been abolished by the UK Parliament’s Enterprise and Regulatory Reform Act 2013. Under Section 108 of, and Schedule 7 to, the Government of Wales Act 2006, the Assembly is expressly allowed to pass legislation that relates to agriculture. According to the Attorney-General for England and Wales, however, the bill was inconsistent with the distribution of legislative powers, since it was not related to agriculture but to the non-devolved matters of employment and industrial relations. Consequently, the Attorney-General referred to the Supreme Court the question of whether the Agricultural Sector (Wales) Bill 2013 was within the legislative competence of the Welsh Assembly. The Court stated that ‘agriculture cannot be intended to refer solely to the cultivation of the soil or the rearing of livestock, but should be understood in a broader sense as designating the industry or economic activity of agriculture in all its aspects’ (para 49).70 The question was then whether the bill was related to agriculture, for which the judges demanded, in line with earlier jurisprudence, ‘more than a loose or consequential connection’ (para 50). With reference to the consultation process before the adoption of the bill, the judges confirmed that it was indeed related to this subject matter, as its purpose was clearly to regulate agricultural wages in support of the agricultural industry in Wales and its legal and practical effects were consistent with that purpose. If a bill is, after such consideration of its purpose and effects, within the competence of the Welsh Assembly, it does not matter whether it might also be classified as relating to a subject matter that is not devolved, such as employment and industrial relations.

Beyond some regional states, judicial scrutiny of draft legislation also occurs in mature federations. The Austrian Constitutional Court is empowered to decide, upon a request from the federal executive or a Land executive, whether a legislative act or the execution of a law falls within the competence of the national or

69

See section 8.3.2 below. Agricultural Sector (Wales) Bill—Reference by the Attorney General for England and Wales [2014] UKSC 43. 70

Ex ante Scrutiny of Subnational Law 263 subnational level (Article 138(2) of the Constitution). The Court’s ruling in such cases is, according to settled case law, an authoritative interpretation of the distribution of powers and has the same effect as a constitutional provision.71 Hence, it may only be overruled by a constitutional amendment requiring two-thirds majorities in both chambers of the Austrian Parliament and not by an ordinary law. Importantly, this function relates to draft laws and thus needs to be distinguished from abstract constitutional review of a law that has already been enacted, which is foreseen under Article 140(1)(2–3) of the Austrian Constitution and also exists in similar forms in Germany and Spain.72 8.2.2 Political or Judicial Scrutiny of Subnational Constitutions In some federal systems, the constitutions of subnational entities are required to undergo, before their enactment, a process of scrutiny that is aimed at preventing conflicts between two basic laws of different levels of government. These mechanisms may be political or judicial in nature, depending on whether the organ entrusted with the approval of the draft constitution belongs to the national government or to the judiciary. A prime example of political scrutiny concerns state constitutions in the United States. As already mentioned, the so-called Guarantee Clause, or Republican Clause, stipulates that the ‘United States shall guarantee to every state in this union a republican form of government’ (Article IV, clause 4 of the Constitution). It has also been pointed out that the assessment of whether a state has ‘a republican form of government’ is a political question and up to Congress.73 In practice, however, the latter has refrained, with few exceptions such as the Southern state constitutions of the early Reconstruction era until 1867, from actually making this assessment. As a result, the Guarantee Clause has remained a ‘sleeping giant’.74 Apart from the United States, political scrutiny characterises another archetype of the federal state. In Switzerland, it is the Federal Assembly, ie both chambers of the national Parliament together, that is entrusted with approving cantonal constitutions. Article 51(1) of the Swiss Constitution requires that these be democratic through approval by the people and that there be the possibility of amendment ‘if the majority of those eligible to vote so request’. If a cantonal constitution is beyond this requirement in line with federal law, more generally, the Federal Assembly must grant the necessary federal guarantee (Articles 51(2) and 172(2) of the Constitution).75 Although this scrutiny procedure might appear at first glance

71

VfSlg 3055/1956. See section 8.3.2 below. 73 See box 5.E. 74 Charles Sumner quoted in GA Tarr, Understanding State Constitutions (Princeton, Princeton University Press, 2000) 42f. 75 See N Schmitt, ‘New Constitutions for All Swiss Cantons: A Contemporary Challenge’ in M Burgess and GA Tarr (eds), Constitutional Dynamics in Federal Systems: Subnational Perspectives (Montreal, McGill–Queen’s University Press, 2012) 151ff. 72

264 Prevention and Resolution of Conflicts to be a significant restriction, it is usually a mere formality. There are only very few cases in which the federal guarantee has been refused in recent times so that the above-mentioned characterisation of the US Guarantee Clause as a ‘sleeping giant’ holds equally true for the Swiss case. The only more recent case in which the Federal Assembly refused its guarantee concerned the ‘reunification clause’ in Article 138 of the 1977 Constitution of Jura. According to this provision, ‘the Republic of Jura could welcome any part of the Jurassic territory … provided that this part has regularly seceded according to pertinent federal and cantonal law’. The federal guarantee was denied because this provision was interpreted as openly encouraging districts of the neighbouring canton of Bern to secede and thus violated both territorial integrity (then-Article 5 of the 1874 Swiss Constitution) and the above-mentioned principle of federal loyalty.76 Whereas political scrutiny of draft constitutions characterises certain mature federations and reflects, in particular, concerns regarding the maintenance of a democratic order, judicial certification is rather typical of more recently formed federal systems. A paradigmatic case in this regard is South Africa. After the adoption or amendment of a provincial constitution by the legislature, the speaker must submit the text to the Constitutional Court for certification (Article 144(1) of the Constitution). The text then only becomes law once the Constitutional Court has certified its consistency with the national constitutional requirements regarding both the procedure (Article 142) and content (Article 143). On this legal basis, the Court refused in 1996 to certify the provincial constitution of KwaZulu-Natal. 8.H South Africa 1996: The Declined Certification of the KwaZulu-Natal Constitution While the African National Congress (ANC) advocated a predominant national government, it was above all the Inkatha Freedom Party (IFP), with a strong power base in KwaZulu-Natal, which demanded significant autonomy for subnational entities. Even though the IFP had left the negotiations regarding the 1993 Interim Constitution as a sign of protest and did not participate in talks concerning South Africa’s Constitution of 1996, the IFP-dominated legislature of KwaZulu-Natal sought to achieve more autonomy by adopting its own provincial constitution that same year. The Constitutional Court, however, refused to certify this basic law and voiced particularly harsh criticism.77 The judges stated that there were ‘fundamental respects in which the provincial constitution is fatally flawed’ (para 13). In procedural terms, they held that the legislature of KwaZulu-Natal had attempted to avoid the Constitutional Court’s scrutiny of the provincial basic law against the Interim Constitution (paras 36–46). As to the content, the Court ruled that there was an unconstitutional usurpation of national government powers. It cited the distinction in an earlier decision78 between South Africa’s Constitution and that of the United States, which brought several sovereign states together in a federation. In contrast to such an aggregative federation, ‘the provinces in South Africa

76

See section 8.1.1 above. Certification of the Kwazulu-Natal Constitution 1996 (4) SA 1098. 78 The National Education Policy Bill 1996 (3) SA 289. 77

Ex ante Scrutiny of Subnational Law 265 are not sovereign states. They were created by the constitution and have only those powers that are specifically conferred on them under the constitution’ (para 23). Against this backdrop, it held that the provincial constitution had been adopted ‘under a misapprehension that it enjoyed a relationship of co-supremacy with the national legislature and even the Constitutional Assembly’ (para 15). Specifically, the lack of sovereignty of KwaZulu-Natal would imply that it lacks the authority to ‘confer’ legislative and executive powers on the province in some areas and to ‘recognise’ the competences of the national Government in others (paras 32–34). Although this judgment dealt a blow to KwaZulu-Natal’s aspirations for more autonomy, the Court also declined on the same day to certify the national constitution,79 not least because the limited scope of autonomy granted to the provinces was deemed inconsistent with the Constitutional Principles laid down in the 1993 Interim Constitution.

One year later, the Constitutional Court certified the first and, so far, only provincial constitution, that of the Western Cape.80 Yet, this decision was likewise characterised by a restrictive interpretation of provincial constitutionalism. Even if Article 143(1a) of the Constitution allows ‘provincial legislative or executive structures and procedures’ that differ from those provided in the national constitution, a distinct provincial electoral system was not regarded as falling within this power to deviate.81 Judicial scrutiny of subnational basic laws before their enactment is also a feature in Spain, where this occurs in addition to political scrutiny through the necessary adoption of statutes by the national Parliament as organic laws.82 The original 1979 Organic Law on the Constitutional Court had foreseen a pre-enactment review of draft statutes and draft statute amendments that was then abolished in 1985. However, an amendment to this organic law in 2015 reintroduced the possibility of ex ante scrutiny. Importantly, neither political nor judicial certification ex ante necessarily secures subnational constitutions from constitutional review after their enactment. As regards political scrutiny, the guarantee issued by the Federal Assembly would not per se preclude the possibility of abstract review. But the Federal Court has so far refused to accept complaints against cantonal constitutions,83 which has entailed certain problems and arguably a gap in legal protection.84 First, the considerations of the Federal Assembly for issuing its guarantee are often of a political rather than legal nature. Second, this institution and its political scrutiny seem more generally suited to safeguarding the constitutional homogeneity of the two government levels regarding certain principles like democracy rather than a comprehensive review of cantonal constitutions. In the United States, the successfully completed process of

79

Certification of the Constitution of the Republic of South Africa 1996 (4) SA 744 (CC). See D Brand, ‘The Western Cape Constitution’ (2000) 31 Rutgers Law Journal 961. 81 Certification of the Constitution of the Western Cape 1998 (1) SA 655. 82 See box 5.F and ch 10.5. 83 BGE 118Ia 124. 84 See EM Belser and N Massüger, ‘Art 51 BV (Kantonsverfassungen)’ in B Waldmann et al (eds), Basler Kommentar zur Schweizerischen Bundesverfassung (Basel, Helbing Lichtenhahn, 2015) 987f. 80

266 Prevention and Resolution of Conflicts political certification, ie the lack of intervention on the part of Congress on the basis of the Guarantee Clause, does not save state constitutions from review. Based on the above-mentioned Article III, section 2, clause 1 of the Constitution, the latter may be challenged and end up before the Supreme Court. Prime examples are two recent cases in which the apex court scrutinised (and eventually upheld) highly controversial amendments to the constitutions of Arizona and Michigan.85 If scrutiny of draft subnational constitutions is judicial in nature, this is likewise not automatically an obstacle for review in court after their enactment. As to the South African case, it is true that safeguarding the compatibility of provincial constitutions with the national Constitution is left to the above-mentioned certification process. Yet, the Constitutional Court is still responsible for the ex post review of provincial constitutions regarding conflicts with national legislation, for which Article 147(1) of the Constitution provides explicit rules. In other countries, there is a double scrutiny of subnational (quasi-)constitutions, both ex ante and ex post. In Spain, Article 79(9) of the 1979 Organic Law on the Constitutional Court provides that the above-mentioned check on the draft statute of an autonomous community does not prejudice its decision in a possible case concerning the constitutionality of the statute after its enactment. In conclusion, there is in federal systems a broad variety regarding the stringency of procedural control over subnational constitutions. While scrutiny in Switzerland is extremely limited, above all in judicial terms, Spain is an example of very tight political and judicial control both before and after enactment. 8.3 CONSTITUTIONAL ADJUDICATION

8.3.1 Significance, Scope and Impact ‘Federalism … means legalism—the predominance of the judiciary in the constitution.’86 In this classic statement, Albert V Dicey declared the authority of courts as the guardians of the federal constitution to be a defining characteristic (and weakness) of federations. Much more recently, institutional definitions of the federal state have identified the function of courts as umpires in disputes between the levels of government as a main feature.87 Indeed, a constitutional court, defined from a functional perspective, as a constitutionally entrenched independent body whose principal purpose is to protect the supremacy of the (federal) constitution within the legal order, is a feature of nearly all federal systems. Ethiopia is in this respect a clear outlier because Article 83(1) of the Constitution declares that ‘[a]ll constitutional disputes shall be decided by the House of the Federation’, ie the 85 Schuette v Coalition to Defend Affirmative Action 572 US ___ (2014); Arizona State Legislature v Arizona Independent Redistricting Commission 576 US ___ (2015). 86 AV Dicey, An Introduction to the Study of the Law of the Constitution, 10th edn (London, Macmillan, 1959) 175. 87 See, eg, RL Watts, Comparing Federal Systems, 3rd edn (Montreal, McGill–Queen’s University Press, 2008) 7. Watts recognised, however, that this umpire function may also be performed, in obvious allusion to Switzerland, by referendums. But as we shall see, the Swiss case is very exceptional.

Constitutional Adjudication 267 second chamber of the national Parliament. A Council of Constitutional Inquiry, whose composition is regulated in Article 82 of the Constitution, merely investigates constitutional disputes and then submits the matter with recommendations to the House of the Federation for a final decision (Article 84(2)). From an organisational point of view, courts performing the function of the ultimate guardian of the federal constitution, which exist in nearly all federal countries, may be entrusted to quite different bodies. There are Constitutional Courts in the widely used strict sense of the term88 or Supreme Courts at the apex of the ordinary court system. The former are specialist courts focused exclusively on constitutional jurisdiction, while the latter are generalist courts that are also vested with ordinary jurisdiction as the highest appeals court.89 Unlike Supreme Courts, specialist Constitutional Courts are thus located ‘outside of’ the regular court system rather than ‘on top of’ it.90 While constitutional courts of some kind are in theory and practice a typical feature of federal systems, there seems to be less clarity on their relative power in comparison to other institutions. On the one hand, there are those who claim, like Alexander Hamilton did with reference to Montesquieu, that the judiciary is the ‘least dangerous branch’ of government because it has ‘no influence over either the sword or the purse’91 and is thus easily overpowered and influenced by the two political branches. In his view, therefore, Anti-Federalists had no reason to fear an overly strong judiciary, in general, or a Supreme Court with the power of judicial review, in particular. On the other hand, there are those who assert that constitutional courts are, through their role of constitutional adjudication in federal systems, actually very powerful actors in their own right.92 This issue requires some differentiation. It is certainly true that federalism and constitutional adjudication are, from a historical perspective, closely intertwined, not least because both models of judicial review were invented in federal systems.93 Yet, the extent to which constitutional courts have played a role in transforming the federal system, either as an active driver of change (mostly towards centralisation) or by acquiescing to change induced by the political branches, varies significantly.94 There are cases like the United States, Canada and Australia where constitutional adjudication has driven or at least accepted significant alterations of the distribution

88 See V Ferreres Comella, ‘The Rise of Specialized Constitutional Courts’ in T Ginsburg and R Dixon (eds), Comparative Constitutional Law (Cheltenham, Edward Elgar, 2011) 265. Capitals are used hereinafter to differentiate these Constitutional Courts from constitutional courts in a functional sense. 89 On the tension between specialist Constitutional Courts and these two forms of jurisdiction in light of transformative constitutionalism, see FI Michelman, ‘The Interplay of Constitutional and Ordinary Jurisdiction’ in Ginsburg and Dixon (eds), Comparative Constitutional Law (2011). 90 L Favoreu, ‘American and European Models of Constitutional Justice’ in D Clark (ed), Comparative and Private International Law: Essays in Honor of John Merryman (Berlin, Duncker & Humblot, 1990) 111f. 91 The Federalist no 78. 92 For an overview of the debates on the impact of courts in various federal systems, see P Popelier, ‘Federalism Disputes and the Behavior of Courts: Explaining Variation in Federal Courts’ Support for Centralization’ (2017) 47 Publius 27, 29ff. 93 See section 8.3.2 below. 94 On this distinction regarding the US case, see M Tushnet, The Constitution of the United States of America: A Contextual Analysis (Oxford, Hart Publishing, 2009) 39f.

268 Prevention and Resolution of Conflicts of powers, as it had been initially conceived in the constitutional text. The federal constitutions in both the United States and Australia were less centralised than that of Canada, with the framers of the Australian Constitution closely following the US model precisely with the aim of avoiding an overly strong national government. Ironically, however, Australia has ended up today with a clearly more centralised brand of federation than Canada, which is, in both cases, in no small part due to the divergent trajectories of constitutional adjudication. In the Australian case, the High Court has been blamed for having undermined federalism by interpreting the Constitution ‘in a manner widely at variance with the intentions and expectations of its founders’,95 with the Engineers’ case of 1920 marking a watershed.96 In a more recent seminal ruling, the Court reinforced a national view of the Constitution, according to which Australia is no ‘mere aggregation of the federating colonies’ and the federal Government enjoys ‘comparative superiority’.97 Jurisprudence in Canada has likewise turned the federal Constitution on its head, but in the opposite direction. Many doctrines with decentralising effects, which the Supreme Court has only partly countered since 1949, date back to early judgments of the Judicial Committee of the Privy Council (JCPC) whose members were thus quite aptly called the ‘wicked stepfathers of confederation’.98 Even though the JCPC certainly undermined the (centralist) plan of the fathers of the Constitution, it seems quite likely that they thereby also secured the survival of the Canadian federation in the long run: ‘[I]f the law lords had not leaned in that [provincial] direction, Quebec separation might not be a threat today; it might be an accomplished fact.’99 Yet, the transformative role played by the courts in the above-mentioned cases is by no means a general pattern. For various reasons, the impact of judicial decisions on federal systems has been much less significant in other countries. Among these reasons are the quantity and standing of constitutional jurisprudence, the probability of constitutional court judgments being prevented or overturned by mechanisms involving the two political branches of government, the degree of scepticism towards judicial power and resulting limitations of its scope and, finally, the approach to constitutional interpretation. As to the first point, constitutional adjudication may be limited in quantitative terms because courts adhere to the idea of judicial restraint. This holds true, for example, in the case of the Brazilian Supreme Federal Court.100 In other cases, there are simply fewer opportunities for the constitutional court to rule on intergovernmental disputes. Even if constitutional jurisprudence in South Africa has demonstrated

95 N Aroney and J Allan, ‘An Uncommon Court: How the High Court of Australia Has Undermined Australian Federalism’ (2008) 30 Sydney Law Review 245. 96 Amalgamated Society of Engineers v Adelaide Steamship Co Ltd (1920) 28 CLR 129 (Engineers’ case). 97 Pape v Commissioner of Taxation (2009) 238 CLR 1, 85. 98 Eugene Forsey quoted in Hogg, Constitutional Law (2000) 5-18. 99 PE Trudeau, ‘The New Treason of the Intellectuals’ in PE Trudeau (ed), Federalism and the French Canadians (Toronto, Macmillan of Canada, 1968) 198. 100 See T Bustamante and E de Godoi Bustamante, ‘Constitutional Courts as Negative Legislators: The Brazilian Case’ in AR Brewer-Carías (ed), Constitutional Courts as Positive Legislators (Cambridge, Cambridge University Press, 2011).

Constitutional Adjudication 269 leanings towards the national Government, it has had less centralising effects than in other countries because the ANC-dominated political context has entailed that not many cases have been brought to it.101 The volume of case law alone is not, however, necessarily a reliable indication of its impact. In comparison, for example, to the German and Italian Constitutional Courts, the Supreme Courts of both Canada and the United States issue far fewer judgments per year because they only decide in select cases and thus leave considerable room for conflicting interpretations in the lower courts (in US terminology known as ‘intercircuit conflicts’). Still, there is no doubt that both these Supreme Courts have had a significant bearing on their respective federal system. Second, the impact of the courts depends on how their rulings relate to mechanisms involving the two political branches of government, which may either precede or follow judicial decision-making. If intergovernmental mechanisms involving the executives are tasked with the prevention of legal conflicts, as mentioned above either de facto (eg United Kingdom) or upon an express constitutional mandate (eg Switzerland and South Africa),102 many disputes do not even appear on the radar of the judiciary. These cases are far from being exceptional. Also in Canada, the Supreme Court’s ‘job as umpire has been overshadowed by the dominance of intergovernmental negotiation and compromise’.103 It is even more interesting that this was recently acknowledged by the apex court judges themselves. 8.I Canada 2011: The Principle of Federalism and the Need for Intergovernmental Cooperation Regarding legal frameworks for financial markets, Canada is exceptional in that securities have always been regulated by the provinces. Even though the provinces have sought to harmonise many of the policies in this field, a debate has been going on for decades about whether to create a single securities regulator for the whole country. This debate culminated in the publication of a draft Canadian Securities Act in 2010. After Alberta and Quebec had posed reference questions to their respective provincial Courts of Appeal, the matter of whether this act of legislation falls within federal jurisdiction was eventually referred to the Supreme Court. The judges held that the draft act is beyond the federal power concerning the ‘regulation of trade and commerce’ (Section 91(2) of the 1867 Constitution Act).104 They emphasised that both levels of government have jurisdiction over some aspects of securities regulation, the federal level under the just-mentioned ‘trade and commerce’ power and provincial authorities due to their competences regarding ‘property and civil rights in the province’ (Section 92(13)). Nevertheless, the proposed act was regarded as too far-reaching in its

101 See N Steytler, ‘The Constitutional Court of South Africa: Reinforcing South Africa’s Hourglass System of Multilevel Government’ in N Aroney and J Kincaid (eds), Courts in Federal Countries: Federalists or Unitarists? (Toronto, University of Toronto Press, 2017). 102 See section 8.1.1 above. 103 G Baier, ‘The Courts, the Constitution and Dispute Resolution’ in H Bakvis and G Skogstad (eds), Canadian Federalism: Performance, Effectiveness, and Legitimacy, 3rd edn (Oxford, Oxford University Press, 2012) 79. 104 Reference Re Securities Act [2011] 3 SCR 837.

270 Prevention and Resolution of Conflicts

scope because it did not limit itself to aspects of national interest and concern but regulated all aspects of securities: ‘[A]s important as the preservation of capital markets and the maintenance of Canada’s financial stability are, they do not justify a wholesale takeover of the regulation of the securities industry’ (para 128). Importantly, the judges added that ‘[i]t is not for the court to suggest to the governments of Canada and the provinces the way forward by, in effect, conferring in advance an opinion on the constitutionality on this or that alternative scheme. Yet we may appropriately note the growing practice of resolving the complex governance problems that arise in federations, not by the bare logic of either/or, but by seeking cooperative solutions that meet the needs of the country as a whole as well as its constituent parts’ (para 132). In fact, a solution that recognises the essentially provincial nature of securities regulation and at the same time allows the federal parliament to address genuinely national concerns, was considered to be supported by Canadian constitutional principles and existing cooperation schemes in other policy fields as well: ‘The backbone of these schemes is the respect that each level of government has for each other’s own sphere of jurisdiction. Cooperation is the animating force. The federalism principle upon which Canada’s constitutional framework rests demands nothing less’ (para 133).

With respect to mechanisms after a constitutional court’s judgment, a key question is to what extent this decision is final, not only de iure in the sense of its ultimate judicial authority but also de facto vis-à-vis overriding actions by the legislature or through constitutional amendment. As for the latter, the power of the courts is evidently greater in the case of constitutional rigidity, which places onerous hurdles on constitutional amendment. Differences regarding the flexibility of the Constitution have contributed, for instance, to divergent trends in the US states and at the national level. While, at the state level, amendments overturning court decisions occur quite frequently,105 this has only rarely been the case at the national level (eg the Eleventh Amendment in response to the Chisholm case and the Sixteenth Amendment in reaction to the Pollock case).106 In its relation to the (national or subnational) legislature that enacted the scrutinised law, a constitutional court’s position is influenced by whether it has the power of traditional ‘strong-form’ constitutional review or merely ‘weak-form’ review.107 In the latter case, the enacting parliament can respond to the judgment in some way. A prime example is the power of legislative override granted by the Notwithstanding Clause under Section 33 of Canada’s 1982 Constitution Act,108 which was vilified by many, but also interpreted as a feature that facilitates dialogue between courts and legislatures.109

105

See ch 10.1. Chisholm v Georgia 2 US (2 Dall) 419 (1793). See box 4.A. Pollock v Farmers’ Loan & Trust Co 158 US 601 (1895). See box 7.E. In Canada, a famous case of overturning a Supreme Court judgment was the addition in 1940 of an exclusive national power over ‘unemployment insurance’ (Sec 91(2)(A)) which responded to the decision in Canada (AG) v Ontario (AG) [1937] AC 355 (JCPC). See ch 11.2. 107 On this distinction, see M Tushnet, Weak Courts, Strong Rights: Judicial Review and Social Welfare Rights in Comparative Constitutional Law (Princeton, Princeton University Press, 2007). 108 See chs 10.3 and 10.6. 109 See PW Hogg and AA Bushell, ‘The Charter Dialogue between Courts and Legislatures (or Perhaps the Charter of Rights Isn’t Such a Bad Thing after All)’ (1997) 35 Osgoode Hall Law Journal 75, 83. 106

Constitutional Adjudication 271 Third, the impact of a constitutional court is influenced by the degree of scepticism towards judicial power in the setting of a democracy, which may have implications for the scope of its powers. Such scepticism towards judicial power may not only follow, as it does in the UK, from the principle of parliamentary sovereignty. It is, in fact, also deeply rooted in the political traditions of other countries. This holds true for Belgium, where the Constitutional Court’s power of judicial review is limited because it is only allowed to measure national and subnational laws against specific provisions of the Constitution that are explicitly listed in the Constitution itself or in special legislation (Article 142(2)(2–3) of the Constitution). Switzerland is similarly marked by scepticism towards the power of judges, which, in this case, is nurtured by the strong tradition of (direct) democracy. While the Federal Court has been authorised since 1874 to adjudicate intergovernmental disputes (Article 189(2) of the Constitution), its scope for judicial review is restricted. Interestingly, it is limited in a very asymmetrical way, as the Court may review cantonal law (Article 189(1)), but not federal law. Article 189(4) states explicitly that ‘[a]cts of the Federal Assembly or the Federal Council may not be challenged in the Federal Court. Exceptions may be provided for by law’. In fact, the apex court is, like all other judicial authorities, obliged to merely apply federal acts alongside international law (Article 190) because the fate of federal acts and international treaties is supposed to be decided by the people through mandatory and optional referendums (Article 140–41).110 Fourth, the significance of a constitutional court is influenced by its approach to constitutional interpretation. Overall, constitutions or other acts of written law rarely contain explicit rules for interpretation, which, of course, require interpretation themselves.111 The rules laid down in the UK devolution acts are a notable exception (Sections 29(3) and 101 of the Scotland Act 1998, Section 83 of the Northern Ireland Act 1998, Sections 94(7) and 154(2) of the Government of Wales Act 2006). Therefore, it mostly depends on the courts themselves where to draw the outer limits of their interpretive scope. They may do so either leaning towards the idea of judicial activism or that of judicial restraint, with invalidation being (self-)restricted to evidently unconstitutional acts. Restraint is typical, for instance, of the Malaysian Federal Court which has only rarely struck down legislation of the federal Parliament for being beyond its powers according to Article 74 of the Constitution.112 The Court held that for reviewing the consistency of a law with the distribution of powers, its substance, ie the object, purpose and design, are decisive rather than its form.113 This matter of basic judicial self-conception is also linked to the question of whether the (federal) constitution has a meaning that is fixed at a specific point in time or is open to interpretive evolution. In this respect, the long experience and varied patterns of the classical mature federations are illustrative. While the Canadian Constitution has been predominantly regarded as ‘a living tree capable of growth 110

See ch 3.3.2. A Gamper, Regeln der Verfassungsinterpretation (Wien, Springer, 2012) 31ff. 112 See A Harding, The Constitution of Malaysia: A Contextual Analysis (Oxford, Hart Publishing, 2013) 137ff. 113 Mamat bin Daud v Government of Malaysia [1986] 2 MLJ 192. 111

272 Prevention and Resolution of Conflicts and expansion within its natural limits’,114 the view of a constitution’s meaning being fixed at the time of its adoption has been strong in both the United States and Australia. This is reflected in the continued significance of legalism in Australia115 and US originalism, with a main proponent of the latter approach famously declaring that the purpose of a constitution is to ‘obstruct modernity’.116 This contrasts in the United States with several older judgments of the Supreme Court that had epitomised an evolutionary approach117 and thus given judges greater leeway to adapt the federal constitution to changing circumstances. On the other hand, such an approach evidently carries the risk of the Court usurping the role of a positive legislator, rather than that only of a negative legislator, defined by Hans Kelsen as an institution that merely strikes down unconstitutional laws. 8.3.2 Constitutional Review and Other Functions There is no doubt that the establishment of the first modern federation in the United States, and thus federalism more generally, is closely intertwined with the idea and practice of constitutional review. While some point out that the review function may be traced back to the colonial courts under British colonial constitutional law,118 others emphasise that ‘at the time the constitution was adopted nearly all knowledgeable participants in the process assumed that the courts did have that power’.119 Yet, as explained above, the power of a court to review the constitutionality of infra-constitutional legal norms, including laws enacted by national and subnational parliaments, and invalidate them was first asserted and effectively gained by the US Supreme Court with its ruling in Marbury v Madison.120 This marked both a victory for the Federalists advocating a strong national government, among them the author of the judgment, John Marshall, and the birth of judicial control of constitutionality. The latter has since then developed into a mainstay of legal orders worldwide and is to be distinguished from political control of constitutionality.121 For federalism in the United States, constitutional review has entailed that the supremacy of the national Constitution, in particular of the distribution of powers, would henceforth be judicially safeguarded and restored. Both federal and state acts

114

Edwards v Canada (AG) [1930] AC 124, 136 (Persons Case). See box 5.I. However, even if legalism is widely recognised as the orthodox approach to constitutional interpretation, it needs to be stressed that, since Chief Justice Dixon’s commitment in 1952 to ‘strict and complete legalism’, the expression has come to be used as an umbrella term for a variety of sometimes quite different approaches, see Saunders, The Constitution (2010) 90f. 116 A Scalia, ‘Modernity and the Constitution’ in E Smith (ed), Constitutional Justice under Old Constitutions (The Hague, Kluwer Law International, 1995) 315. 117 McCulloch v Maryland 17 US (4 Wheat) 316, 407 (1819) and 415. See box 5.J. Missouri v Holland 252 US 416, 433 (1920). See box 14.A. 118 See AHY Chen and M Poiares Maduro, ‘The Judiciary and Constitutional Review’ in Mark Tushnet et al (eds), Routledge Handbook of Constitutional Law (New York, Routledge, 2013) 98. 119 See Tushnet, The Constitution (2009) 134. 120 Marbury v Madison 5 US (1 Cranch) 137 (1803). See box 3.C. 121 See R Bellamy, Political Constitutionalism: A Republican Defence of the Constitutionality of Democracy (Cambridge, Cambridge University Press, 2007). 115

Constitutional Adjudication 273 could be reviewed and struck down in case they were found ultra vires. In some of its judgments, however, the Supreme Court has refrained from using its review power regarding federal laws and thereby explicitly referred to what is known as the theory of political safeguards of federalism.122 Although this theory, which posits that the two political branches and not the courts should enforce constitutional limits on the federal powers, has recently lost influence, it had still marked a seminal ruling of 1985. 8.J United States 1985: Policing Constitutional Limits on Federal Powers? When the Fair Labor Standards Act (FLSA) was enacted in 1938, it did not apply to employees of state and local governments or private transit companies. After the Supreme Court had ruled in 1976 that the power to regulate interstate commerce123 would not grant Congress the authority to regulate the wages and hours of governmental employees exercising traditional governmental functions,124 the San Antonio Metropolitan Transit Authority informed its employees that it was no longer under a federal obligation to provide them with overtime payment. When the matter was taken to the courts and finally arrived at the US Supreme Court, the latter overruled its above-mentioned decision from only nine years before and held that Congress actually had the authority under the Interstate Commerce Clause to apply the FLSA to a local mass transit system operated by a governmental entity.125 Specifically, the Court now contradicted its own view in National League of Cities v Usery that the recognition of state sovereignty in the US Constitution126 would inevitably imply legal limits on the federal power to regulate employment relations. According to the new ruling, ‘the principal means chosen by the framers to ensure the role of the states in the federal system lies in the structure of the federal government … The framers thus gave the states a role in the selection both of the executive and the legislative branches of the federal government’ (p 550–51). In effect, the Court thus ‘announced that it no longer would examine the constitutionality of federal legislation that threatened to violate the sovereignty of the states’.127 It did so by adopting the rationale of the theory of political safeguards of federalism. As a leading proponent of this theory claimed that ‘the constitutional issue of whether federal action is beyond the authority of the central government and thus violates “states’ rights” should be treated as nonjusticiable, final resolution being relegated to the political branches-i.e., Congress and the President.’128 In several more recent judgments, however, the Supreme Court has reintroduced judicial review concerning federal powers in the light of state sovereignty by policing Congress’s use of its enumerated powers.129

122 See H Wechsler, ‘The Political Safeguards of Federalism: The Role of States in the Composition and Selection of the National Government’ (1954) 54 Columbia Law Review 543; JH Choper, Judicial Review and the National Political Process: A Functional Reconsideration of the Role of the Supreme Court (New Orleans, Quid Pro Books, 1980). 123 See box 5.G. 124 National League of Cities v Usery 426 US 833 (1976). 125 Garcia v San Antonio Metropolitan Transit Authority 469 US 528 (1985). 126 See ch 4.1.2. 127 JC Yoo, ‘The Judicial Safeguards of Federalism’ (1996) 70 Southern California Law Review 1311, 1311. 128 Choper, Judicial Review (1980) 175. 129 New York v United States 505 US 144 (1992); United States v Lopez 514 US 549 (1995). See box 5.G.

274 Prevention and Resolution of Conflicts Whereas, in the US case, it was the Supreme Court that laid claim to the review power and eventually prevailed, more recent federal constitutions have granted such power explicitly. This holds true, in particular, for the 1920 Austrian Constitution—drafted mainly by Hans Kelsen—which is notable for having established a second model of constitutional review.130 First, in contrast to decentralised (or diffused) review by ordinary courts in the United States, which results from the review authority being seen as inherent to judicial power, the Austrian or Continental European model advocates centralised control of constitutionality through a specialist Constitutional Court. Second, the new model allowed for abstract review of a legal norm regardless of its relevance in a specific court case and thus before it has even been applied. In the United States, by contrast, the Case or Controversy Clause (Article III, section 2, clause 1 of the US Constitution) is generally recognised as requiring judicial review to be concrete and relating to enacted law. This is also reflected in the Marbury judgment, declaring that the Court must determine which of ‘conflicting rules governs the case’.131 Third, the Austrian model foresees a review principaliter, where constitutionality is the only matter at stake and not, as with the US review incidenter, merely incidental to the decision of which party wins the case concerned.132 It is important to note, however, that the Austrian Constitution has over time also adopted features of concrete and incidental review. In fact, since 1929, the Administrative Court and the Supreme Court in civil and criminal matters and, since 1985, the appeals courts, are obliged to halt, even in case of mere reservations regarding the constitutionality of a law they would have to apply, their proceedings and refer the law to the Constitutional Court (Article 140(1)(a) in conjunction with Article 89(2–3)).133 This today complements abstract review, which allows oversight over constitutionality to be initiated without a pending court case. A decision may be requested in a reciprocal way by a Land government regarding a federal law or by the federal Government concerning a Land law (Articles 140(1) (2–3)). As we will see, such a combination of abstract and concrete review is foreseen in several federal systems, following, in principle, the Austrian model. Conversely, the US model, as mentioned above, rules out abstract review. For the diffusion of the two archetypes of (federal) constitutional review that have been adopted and often variegated in other federal countries, the legal system has been a determining factor, but not the only one. Overall, countries with civil law systems (eg Germany, Spain and Italy) have tended to follow the Austrian model. The German case is particularly interesting insofar as the courts had asserted

130 See H Kelsen, ‘Judicial Review of Legislation: A Comparative Study of the Austrian and the merican Constitution’ (1942) 4 The Journal of Politics 183. Kelsen thereby drew on earlier ideas in A G Jellinek, Ein Verfassungsgerichtshof für Österreich (Wien, Hölder, 1885). Moreover, the 1920 Constitution of the Czechoslovak Republic had established a comparable system of judicial review several months before, but it never exercised this power. 131 Marbury v Madison 5 US (1 Cranch) 137, 178 (1803). 132 See M Cappelletti, Judicial Review in the Contemporary World (Indianapolis, Bobbs-Merrill, 1971) 69. 133 Regarding the review of the legality of administrative ordinances, all courts are empowered to bring matters to the Constitutional Court (Art 139(1)).

Constitutional Adjudication 275 uring the Weimar Republic, in the absence of an explicit constitutional provision, d a US-style review power, eventually without success.134 Only the Basic Law of 1949 entrenched a strong regime of constitutional review that was seen, like the catalogue of fundamental rights,135 as an essential part of the legal bulwark against lapsing back into dictatorship. Like in the Austrian case, it established a specialist Constitutional Court with the power of both concrete review upon referral by a lower court involved with a supposedly unconstitutional law (Article 100(1)) and the clearly less frequent abstract review. Compared to Austria, the right to referral, granted to the federal Government, Land governments and one-fourth of the members of the Parliament (Bundestag) (Article 93(1)(2)), is even broader. It not only concerns, in a reciprocal manner, acts of the respective other government level but also, for all three potential claimants, both federal and Land law. The Federal Constitutional Court effectively asserted its review power and for the first time invalidated a federal law in the Southwest State case of 1951.136 Similarly, the Spanish Constitutional Court has the power of both concrete review (Article 163 of the Constitution) and abstract review (Articles 161(1)(a) and 162(1)(a)), while Italy only provides for the latter (Article 127 of the Constitution). However, these younger regional states have in common certain asymmetries to the benefit of the national government, even if they do not go as far as in the abovementioned Swiss case to altogether exclude the review of national laws. In Italy, the central Government may challenge the constitutionality of regional laws within 60 days from their publication and a region may do the same regarding both national and regional laws. The former, however, may file a case when regional laws exceed ‘regional competence in the broadest sense of the word: any breach of the constitution could justify [a] challenge’.137 A region, conversely, may only challenge laws that interfere with its own competences. Whereas in Spain a challenge by the national Government of an act of an autonomous community brings about the latter’s suspension (Article 161(2) of the Constitution), this rule does not apply to the reverse case of a region questioning a national law. In contrast to civil law countries, federations with a common law system have been generally inclined to stick closer to the US model (eg Canada, Australia, India and Nigeria).138 Like in the US case, the Canadian Constitution does not expressly recognise judicial review. With arguments, similar to those in the Marbury ruling, it was the JCPC that successfully claimed the review power from the outset, even though this was ‘based more on imperialism than on constitutionalism’.139

134 See DP Kommers and RA Miller, The Constitutional Jurisprudence of the Federal Republic of Germany, 3rd edn (Durham, Duke University Press, 2012) 6. 135 See ch 10.6. 136 1 BVerfGE 14 (Southwest State). See box 6.F. 137 M D’Amico, ‘The Constitutional Court’ in A Helen and D Haljan (eds), Constitutional Law in Italy (Alphen aan den Rijn, Kluwer Law International, 2013) 212. See also Italian Constitutional Court No 274/2003. 138 Sec 4(8) of the Nigerian Constitution. See also Attorney General of Ondo State v Attorney General of the Federation SC 6/2002 of 7 June 2002. 139 PH Russell, The Judiciary in Canada: The Third Branch of Government (Toronto, McGraw-Hill Ryerson, 1987) 93.

276 Prevention and Resolution of Conflicts The precedence of the 1867 Constitution Act, called the British North America Act until 1982, would need to be judicially enforced because it was an imperial statute. Regarding the Australian Constitution, by contrast, the review power was assumed, even if not entrenched, by its framers as a corollary of the status of the Constitution or as inherent to a federation or both.140 Of all common law federations, the constitutional history of judicial review in India is quite likely the most controversial. As with the other above-mentioned cases, the Indian Constitution does not include an explicit general provision granting review power. However, this power has usually been inferred from various provisions imposing specific constitutional limitations on all government institutions141 and it has often been exercised in a way giving federal legislation much leeway.142 Yet, over a certain period of India’s history, the political branches of government have attempted to suppress constitutional review. 8.K India 1980: Judicial Review as a Basic Feature of the Indian Constitution During the state of emergency (1975–77), the Constitution (Forty-Second Amendment) Act 1976 (hereinafter the Forty-Second Amendment) was adopted under the Government of Indira Gandhi. This act contained fifty-nine provisions, inter alia, Section 55 which added a new clause (4) to Article 368 of the Indian Constitution, according to which ‘[n]o amendment of this Constitution … shall be called in question in any court on any ground’. Another additional clause (5) reasserted that the power of amendment would be unlimited and not subject to judicial review. The Supreme Court of India struck down Section 55 of the Forty-Second Amendment and emphasised that the national Parliament cannot use its power to amend the Constitution, which had already been held to be limited in the Kesavananda Bharati case,143 to grant itself unlimited power.144 With reference to the basic structure doctrine developed in the above-mentioned case, the Court declared the power of judicial review to be a basic feature of India’s constitutional system that cannot be eliminated by amendment and, in turn, implies the review not only of ordinary laws but of amendments as well. The importance of this power is epitomised by the fact that the judges used it to strike down another provision of the Forty-Second Amendment besides Section 55 in the same ruling. Its Section 4 had altered Article 31C of the Constitution in order to give precedence to the Directive Principles of State Policy (Part IV of the Constitution) over Fundamental Rights (Part III). The Court, however, strongly defended the rights catalogue, especially the articles guaranteeing equality and civil liberties, against such encroachment: ‘Three Articles of our Constitution, and only three, stand between the heaven of freedom into which Tagore wanted his country to awake and the abyss of unrestrained power. They are Articles 14, 19 and 21. Article 31C has removed two sides of that golden triangle’ (p 660).

140 141

289.

See Saunders (n 47) 75. See DD Basu, Introduction to the Constitution of India, 17th edn (New Delhi, Prentice-Hall, 1995)

142 But see for two examples of considering state powers: State of West Bengal v Kesoram Industries, 2005 AIR 1646; and ITC Limited v Agricultural Produce Market, 2002 AIR 852. 143 His Holiness Kesavananda Bharati Sripadagalvaru v State of Kerala, 1973 AIR 1461. See box 2.C. 144 Minerva Mills Ltd v Union of India, 1980 AIR 1789.

Constitutional Adjudication 277 Although courts in common law federations are mainly focused on deciding cases of an adversarial nature under the power of concrete constitutional review, they may in some cases be asked to issue advisory opinions without a case being litigated. The Indian Supreme Court, for example, has given several such opinions after consultation by the President under Article 143 of the Constitution. In the absence of an explicit provision, advisory opinions are not rendered in Australia or, based again on the Case or Controversy Clause, in the United States.145 This contrasts with Canada, where the so-called reference jurisdiction was recognised by the JCPC more than a century ago,146 even if Article 101 of the 1867 Constitution Act, the legal basis for the creation of the Supreme Court, speaks of ‘a general court of appeal [emphasis added] for Canada’. There have been several significant references since then, especially in terms of whether they settled intergovernmental conflicts in a fair way. Some of these cases were brought to the Court by the federal government (eg the 1998 Secession Reference)147 or, after proceedings at the respective provincial court of appeal, a provincial government (eg the 1981 Patriation Reference upon request of Newfoundland, Quebec and Manitoba).148 Even if both civil and common law federal countries thus demonstrate penchants, respectively, for the Austrian and US type of review, there is no necessary link between the two models and the legal system.149 There are indeed exceptions in both directions. For example, Argentina, a civil law country, has largely emulated, as in many other regards, the United States. Article 116 of the Constitution stipulates that lower courts also ‘have jurisdiction over and decide all cases that deal with matters governed by the constitution’. The Supreme Court of Justice then has, through its appellate jurisdiction, the final say (Article 117). On the other hand, South Africa, a common law country, has adopted the Austrian idea of a specialist Constitutional Court, but in many regards represents a hybrid category of its own. In contrast to the above-mentioned silence of the constitutions of mature common law federations about judicial review, that of South Africa regulates it explicitly. Lower courts are empowered, with few exceptions, to review the constitutionality of ‘an act of parliament, a provincial act or any conduct of the President’, but their orders of invalidity have no force unless confirmed by the Constitutional Court (Articles 167(5) and 172 of the Constitution). A still more exceptional case that defies any clear categorisation is judicial review in the United Kingdom. This is due, of course, to the absence of a codified written constitution. Consequently, political and legal controversies have centred around the question of common law judicial review concerning acts of the devolved legislatures. This question was at stake in a 2011 case before the UK Supreme Court.

145 Re Judiciary and Navigation Acts (1921) 29 CLR 257. On the US case, see LH Tribe, American Constitutional Law, 3rd edn (New York, Foundation Press, 2000) 73ff. 146 Ontario (AG) v Canada (AG) [1912] AC 571 (JCPC). 147 Reference Re Secession of Québec [1998] 2 SCR 217. See box 4.E. 148 Reference Re Resolution to Amend the Constitution [1981] 1 SCR 753 (Patriation Reference). See box 6.E. 149 See AR Brewer-Carías, Judicial Review in Comparative Law (Cambridge, Cambridge University Press, 1989) 128ff.

278 Prevention and Resolution of Conflicts

8.L United Kingdom 2011: Devolved Legislation and Common Law Judicial Review One year after its first ruling on a devolution issue that had concerned reserved and devolved powers,150 the Supreme Court had to deal with another challenge that was brought before it, like most devolution issues, by private parties. Insurance companies claimed that the Damages (Asbestos-related Conditions) (Scotland) Act 2009 was unlawful. This Act of the Scottish Parliament stipulated, unlike a House of Lords decision,151 that asbestos-related conditions constituted an actionable personal injury so that damages could be claimed. One of two grounds that the appellants brought forward was that this act of Scotland’s legislature would be subject to judicial review as an unreasonable, irrational and arbitrary exercise of legislative power. Thus, they claimed that the list of grounds on which such an act may be challenged, which is contained in Section 29 of the Scotland Act, is not exhaustive. The Supreme Court held that an act of the Scottish Parliament could not be challenged as if it were a decision of an ordinary public body and was therefore not subject to ordinary common law judicial review.152 However, the judges also underlined that it would strike down such an act as unlawful if it violated the rule of law. Interestingly, the reasoning behind this ruling appeared to put acts of the Scottish Parliament, in certain regards, on an equal footing with legislation of the UK Parliament. They would be ‘self-standing’ and have within the areas of Scottish competence ‘the highest legal authority’ (para 46). With respect to such a fully fledged legislature, judicial review was seen as inevitably limited, as ‘the democratic process is liable to be subverted if, on a question of political or moral judgment, opponents of an act achieve through the courts what they could not achieve through parliament’ (para 49).

We have seen that judicial review, the core function of nearly all constitutional courts, has had important implications for both the federal system in general and the resolution of intergovernmental disputes in particular. First, a national or a subnational government may directly challenge acts of another government as ultra vires under abstract review. Second, it may in the case of concrete review act as party in legal proceedings regarding cases of an adversarial nature. Still, judicial review is not the only court function related to intergovernmental conflicts. Many federal constitutions indeed provide constitutional courts, in recognition of such conflicts being inherent to all systems with multiple governments, a broad competence for the judicial settlement of disputes of different kinds. Several Supreme Courts are granted, in addition to their ordinary appellate jurisdiction, original jurisdiction regarding disputes both in the vertical and horizontal dimension. In the United States, this follows from Article III, section 2, clause 1 of the Constitution, according to which judicial power extends not only to certain cases but also ‘to controversies to which the United States shall be a party; to controversies between two or more states’. As the Supreme Court’s jurisdiction in these cases has original and exclusive character, it may be petitioned directly. The Court

150

Martin & Miller v Lord Advocate [2010] UKSC 10. Rothwell v Chemical & Insulating Co Ltd [2007] UKHL 2. 152 AXA General Insurance Ltd v Lord Advocate (Scotland) [2011] UKSC 46. 151

Constitutional Adjudication 279 then attempts to settle the dispute by finding a ‘fair and equitable solution’.153 Since constitutional reforms in the mid-1990s, the Mexican Supreme Court is authorised to hear ‘constitutional controversies’ between and within levels of government (Article 105 of the Constitution). With this function the Court is expected to ensure the integrity of law throughout the country in view of potentially conflicting partial legal orders in the states.154 Similarly, Article 131 of the Indian Constitution foresees original and exclusive jurisdiction of the Supreme Court over vertical and horizontal disputes. After the judges had first made a decision under this clause in 1963,155 they were soon thereafter again asked to settle a conflict. 8.M India 1978: Parallel National Inquiries into Actions of a State Government? In a memorandum to the Indian Government, the members of the State Assembly of Karnataka had accused their Chief Minister of corruption and nepotism. In response, the Government of Karnataka established in 1977 a one-man commission consisting of a retired judge of the Karnataka High Court to inquire into these allegations. When the Government of India set up a commission as well on the basis of Section 3 of the 1952 Commissions of Inquiry Act, the state government replied that this interfered with a state domain and therefore claimed under Article 131 of the Indian Constitution that the notification establishing the national commission was ultra vires. The Supreme Court did not follow the plaintiff’s argument that the Indian Constitution would only provide for the collective responsibility of state governments to their legislative assemblies (Article 164(2)), but not permit parallel oversight by a national commission.156 In making this judgment, the Court upheld its power under Article 131 to resolve disputes ‘between the Government of India and one or more states’. It did so, in particular, by emphasising that the supposed distinction between the ‘state’ (as an abstract entity) and the ‘state government’ (as a concrete institution representing the state) was irrelevant.

In federal countries with specialist Constitutional Courts, the judicial settlement of intergovernmental disputes of various kinds is a widespread phenomenon. It is today a task assigned to institutions otherwise as different as the Swiss Federal Court (Article 189(2) of the Constitution) and the Italian Constitutional Court (Article 134 of the Constitution). In fact, this function has sometimes even preceded that of constitutional review by many years. In the German case, its earliest precursors have been traced back to dispute settlement between otherwise warring principalities before the Imperial Chamber Court (Reichskammergericht) established in 1495.157 During Belgium’s incremental process of federalisation, a court was set up in 1985 with the sole function of settling legislative conflicts between governments (Articles 141 and 142(2)(1) of the Constitution).158 It hardly comes as a surprise,

153

Texas v New Mexico 482 US 124, 134 (1987). Mexican Supreme Court of Justice, Constitutional Controversy 31/1997. 155 State of West Bengal v Union of India, 1963 AIR 1241. 156 State of Karnataka v Union of India, 1978 AIR 68. 157 See Kommers and Miller, The Constitutional Jurisprudence (2012) 4. 158 P Popelier and K Lemmens, The Constitution of Belgium: A Contextual Analysis (Oxford, Hart Publishing, 2015) 196f. 154

280 Prevention and Resolution of Conflicts therefore, that this institution was initially called the Court of Arbitration (court d’arbitrage/arbitragehof) and only in 2007 was renamed the Constitutional Court. This new label followed the transfer of additional functions, starting with the 1989 Special Law on the Court of Arbitration. Regarding competences of judicial dispute settlement, it is important to note that, from a functional perspective, there may be a certain overlap (abstract) constitutional review. In the German case, dispute settlement relates, in view of administrative federalism to conflicts ‘especially in the execution of federal law by the Länder’, but it also concerns more generally ‘disagreements concerning the rights and duties of the Federation and the Länder’ (Article 93(1)(3)), as well as ‘other disputes involving public law between the Federation and the Länder, between different Länder, or within a Land’ (Article 93(1)(4)). These functions of the Federal Constitutional Court are broadly designed mechanisms for general conflicts of various kinds that supplement the review power regarding special conflicts about the constitutionality of national or subnational law. Indeed, the settlement of vertical disputes under Article 93(1)(3) has, after an early heyday with some highly controversial disputes,159 increasingly lost in importance, not least because conflicts have often been decided under the procedure for abstract judicial review.160

159

1 BVerfGE 14 (Southwest State). See box 6.F. 12 BVerfGE 205 (First Broadcasting case). W Heun, The Constitution of Germany: A Contextual Analysis (Oxford, Hart Publishing, 2010) 171. 160 See

9 Local Government 9.1 LOCAL GOVERNMENT IN A CHANGED FACTUAL AND LEGAL CONTEXT

F

EDERAL SYSTEMS WERE traditionally construed according to a paradigm of dualism that prompts them to focus on the national and subnational levels of government. From this perspective, particularly common among old and aggregative federations, local governments do not directly matter for this dyadic constellation, but are left to the realm of subnational entities. Indeed, they were traditionally considered, as the US Supreme Court famously put it more than a century ago, ‘the creatures—mere political subdivisions—of the state, for the purpose of exercising a part of its power’.1 Yet, this seems to have changed, as national constitutions of more recent devolutionary systems have tended to explicitly recognise local governments. This has in some cases effectively provided them with greater autonomy, but in others, through rather extensive regulation, made them ‘creatures of the national government’.2 Anyway, local governments have recently been attracting increasing interest on the part of both constitution-makers and, consequently, scholars conducting research on federalism. This is not least the result of how strongly diverse local institutions have been impacted by fundamental changes of the factual and, at least in Europe, also legal context that they are embedded in. From a systematic point of view, the institutions of local government are far from uniform and vary significantly from case to case.3 Most common is, of course, the basic general-purpose municipality, which provides a wide range of essential services to households and the local community. But beyond that, local government institutions also include elected single-purpose bodies such as school boards, which, while largely an Anglo-American phenomenon, also exist, albeit to a lesser extent, in Switzerland. Such local governments are typically directly elected and thus different in nature from bodies of inter-municipal cooperation, which municipalities may equally create for a specific purpose.4 Another form of local government are umbrella entities like the counties (Kreise) in Germany, whose territorial scope encompasses

1

Atkin v Kansas 191 US 207 (1903). See box 9.C. See section 9.2.3 below. 3 For a systematic categorisation, see especially N Steytler, ‘Comparative Conclusions’ in N Steytler (ed), Local Government and Metropolitan Regions in Federal Systems (Montreal, McGill–Queens University Press, 2009) 398ff. 4 See section 9.4.2 below. 2

282 Local Government the area of several basic general-purpose municipalities.5 Rather specific for some cases is a fourth category, namely indigenous local governments. In certain instances, like in the Mexican state of Oaxaca, there is the possibility to nominate local authorities according to non-democratic customary law practices.6 Other countries like South Africa have proven less inclined to accommodate customary local authorities, although traditional leaders may still be granted non-voting seats of up to 20 per cent of council members.7 This variety of institutions demonstrates that ‘local government’ is anything but uniform, even if the term is sometimes wrongly used as a synonym for ‘municipality’. Notwithstanding this diversity, all institutions of local government have been strongly impacted in recent decades by several fundamental changes of the environment they are embedded in.8 One of these changes is the phenomenon of metropolitan growth. Indeed, ‘the last half century was characterized by population growth and parallel urbanization in which city areas physically overflowed the jurisdictional limits of their original human settlements’.9 Even though the actual degree of urbanisation varies, the depopulation of rural areas has become a general trend and a matter of concern for local governments worldwide. As we shall see, the fact of a growing urban–rural divide has had immediate and forceful repercussions on both the powers and structure of local government.10 The growing number and importance of metropolitan areas, which are commonly defined as coherent territories with a large urban core and a socially and economically integrated periphery,11 has made them core areas, which to a large extent determine the ‘wealth and health of nations’.12 These are not only in and of themselves tightly integrated as agglomeration economies, but they are also nodal points of a globalised economy and, as such, have an interdependent relationship with similar metropolitan areas, which are both partners and competitors. On the other hand, metropolitan areas have to bear many specific burdens like high concentrations of pollution, as well as the coexistence of wealth and poverty and different kinds of social distress.

5

See section 9.4.3 below. this state introduced this possibility in 1995, more than two-thirds of municipalities have applied the usos y costumbres system, which does not guarantee secret ballot and in some cases systematically excludes certain sections of the population, such as women. Apart from concerns regarding legitimation, this practice has also been criticised for isolating these municipalities and thus subjecting them to the influence of the national government. See A Xanthaki, ‘Indigenous Autonomy in the Americas’ in TH Malloy and F Palermo (eds), Minority Accommodation through Territorial and Non-territorial Autonomy (Oxford, Oxford University Press, 2015) 209. 7 See J de Visser, ‘Republic of South Africa’ in Steytler (ed), Local Government (2009) 273f. 8 See B Denters and LE Rose, ‘Local Governance in the Third Millennium: A Brave New World?’ in B Denters and LE Rose (eds), Comparing Local Governance: Trends and Developments (London, Palgrave Macmillan, 2005) 2ff. 9 B Graizbord, ‘Governance of Megacities in Federal Orders’ in J Kincaid and R Chattopadhyay (eds), Local Government in Federal Systems (New Delhi, Viva Books, 2008) 72. 10 See sections 9.3 and 9.4 below. 11 For an overview, see R Stren and R Cameron, ‘Metropolitan Governance Reform: An Introduction’ (2005) 25 Public Administration and Development 275, 275f. 12 Steytler, ‘Comparative Conclusions’ (2009) 394. 6 Since

Local Government in a Changed Context 283 All of this implies new challenges for local government, which is not only reflected in the recent focus of federalism research on these areas13 but also in constitutional law. In Switzerland, for instance, Article 50(2–3) of the Constitution has, since 1999, obliged the national Government to consider in all its activities the consequences for municipalities, thereby bearing in mind ‘the special position of the cities and urban areas’.14 In the cases of India and South Africa, this recognition of metropolitan growth has tangible institutional implications, as the constitutions of both countries foresee the establishment of different local government types for urban and rural areas.15 Another major change is the trend towards privatisation of services, which is epitomised by increased outsourcing of formerly typical local government responsibilities through public–private partnerships in the name of what has become known as New Public Management. These measures are sometimes heavily contested, in particular by structurally disadvantaged groups and areas. Even though, for instance, Article 30(5) of the Brazilian Constitution expressly authorises municipalities to render public services of local interest ‘directly or by concession or permission’, the latter option is for this reason chosen rather rarely, at least for essential services. Nonetheless, there is, overall, a clear trend towards the involvement of private actors from outside local government proper. While the rise of such constellations of local governance was first observed in the United Kingdom,16 it represents a more general pattern. This new environment has significantly modified the role of basic generalpurpose municipalities, which are (in line with the motto that government should ‘steer and not row’) increasingly engaged in coordination rather than doing everything themselves. Anyway, the rise of metropolitan areas and privatisation are global changes of local government’s factual context and have thus affected it, to a greater or lesser degree, worldwide. As far as Europe is concerned, local institutions have additionally found themselves, since the late 1980s, in a changed legal context created by the European Union and the Council of Europe. Within the EU’s legislative process, local governments may be granted a voice, albeit not a very strong one, as part of the Committee of the Regions.17 However, those Member States that do have subnational entities clearly tend to nominate, in line with the name of the assembly, representatives from the regional level.18 Moreover, the subsidiarity principle was recently extended by the Lisbon Treaty to include, now explicitly, the local level alongside the regional and national levels.19

13 See especially E Slack and R Chattopadhyay (eds), Governance and Finance of Metropolitan Areas in Federal Systems (Oxford, Oxford University Press, 2013). 14 This constitutional amendment was soon followed in 2001 by the creation of the Tripartite Agglomeration Conference (for details on this conference, see section 9.3.2.2 below). 15 See sections 9.3.1 and 9.4.3, respectively. 16 See R Leach and J Percy-Smith, Local Governance in Britain (Basingstoke, Palgrave, 2001). 17 On the Committee of the Regions, see ch 1.2.2. 18 Germany, for example, merely reserves three out of its 24 seats for associations of local government and Austria three out of 12. Belgium even allocates all of its 12 seats on the committee to its regions and communities. 19 See box 1.B.

284 Local Government Compared to these EU initiatives, the European Charter of Local Self-Government (ECLSG),20 which was adopted in 1985 under the auspices of the Council of Europe’s Congress of Local and Regional Authorities (CLRAE),21 has arguably been a more powerful instrument. Even though its enforcement is based, unlike the European Convention on Human Rights, upon a reporting and monitoring system instead of a formal court procedure, the Charter seems to be promising mainly for three reasons. First, it has been recognised and reinforced through its ratification by all Council of Europe members and by the adoption of an additional protocol on the very topical issue of democratic participation in local affairs.22 Second, the Charter’s content comprehensively covers all essential aspects of local self-government.23 Third, it represents an exceptional case of an international treaty that contains, at least within the limits of an ‘à-la-carte’ system,24 binding provisions regarding the usually domestic issue of state organisation. As to its subject matter, Article 13 ECLSG states, instead of providing a legal definition of ‘local authorities’, that the Charter applies ‘to all the categories of local authorities existing within the territory of the Party’. This provision thus leaves it to the discretion of the states whether or not to include, if they exist, the above-mentioned local governments other than municipalities.25 9.2 LEGAL RECOGNITION OF LOCAL SELF-GOVERNMENT

It is one of the key stipulations of the ECLSG that ‘the principle of local selfgovernment shall be recognised in domestic legislation, and where practicable in the constitution’ (Article 2). In keeping with this provision, a number of countries in Europe, but also elsewhere, indeed accord such recognition in constitutional law. Such entrenchment raises three main questions. The first one is about the source of such protective provisions, which may be the national and/or subnational constitution. The second question relates to the scope and quality of this recognition, that is,

20 See C Himsworth, The European Charter of Local Self-Government (Edinburgh, Edinburgh University Press, 2015). 21 On the CLRAE, see ch 1.2.2. 22 After being signed in 2009, the Additional Protocol to the European Charter of Local SelfGovernment on the Right to Participate in the Affairs of a Local Authority entered into force in 2012 and has so far been ratified by eight Member States of the Council of Europe. 23 See D Schefold, ‘Der Schutz der kommunalen Selbstverwaltung durch den Europarat’ in S Breitenmoser et al (eds), Menschenrechte, Demokratie und Rechtsstaat: Liber amicorum Luzius Wildhaber (Baden-Baden, Nomos, 2007) 1062. 24 According to Art 12(1) ECLSG, ‘each Party undertakes to consider itself bound by at least twenty paragraphs of Part I of the Charter’. For an overview of how differently the charter therefore impacts the legal systems of its parties, see the country case studies in A-M Moreno (ed), Local Government in the Member States of the European Union: A Comparative Legal Perspective (Madrid, INAP, 2012). 25 Germany, for instance, made a declaration that the Charter generally applies to both its municipalities and its counties. Belgium officially stated that the scope of the Charter includes its municipalities and provinces, but not the Centres publics d’Aide sociale (Public Centres for Social Welfare) on the territory of the Brussels-Capital Region. Switzerland and the United Kingdom also issued declarations that excluded certain categories of local government.

Legal Recognition of Local Self-Government 285 the extent of self-government, as well as its firm entrenchment and justiciability. The third one is linked with the problem that overregulation of local government often tends to thwart its autonomy. 9.2.1 Source of Recognition From a historical point of view, it appears to be beyond controversy that older constitutions fail to acknowledge local self-government as a result of the abovementioned national–subnational dualism. They are either completely silent on local governments or, like Section 92(8) of Canada’s 1867 Constitution Act, merely mention them within the distribution of powers as objects of subnational legislation. One old constitution, however, namely that of Austria of 1920, entrenches and protects municipalities not only as local administrative units but also as institutions of self-government (Article 116(1)). In certain other instances, like Switzerland in 1999, a recognition clause was belatedly added to an old constitution (Article 50(1)). Whereas constitutional recognition was twice rejected in Australia by unsuccessful referendums in 1974 and 1988, this failure only applied to the national level. All Australian states have, instead, added such safeguards to their constitution since the 1980s. Yet, these mostly guarantee little more than the mere existence of local governments and underline their belonging to the realm of the states rather than that of the national Government.26 Such weak guarantees could not prevent their marginalisation or, in the 1990s in the state of Victoria, even the outright replacement of elected municipal councils with councils run by an appointed chief executive officer.27 In the United States, subnational recognition is older and as a general rule stronger. Initially, the constitutions of the US states were as silent on local government as the national Constitution, even if they de facto accepted the established prerogatives of existing local authorities. Only in the mid-nineteenth century did the states come to see themselves as unitary sovereigns and these authorities as subordinated to them.28 As states soon became, in the face of growing cities, increasingly overwhelmed by the task of regulating every little aspect of local government and confronted with grievances about excessive intrusion, they started to recognise local ‘home rule’ as an alternative. Since 1875, they have, for these quite practical purposes, included more or less far-reaching guarantees of self-government in their constitutions.29 In contrast to the lack of subnational recognition in most countries with older constitutions, there has been, in the post-World War II era, a veritable trend towards guarantees in national constitutional law. This development is mostly attributed to

26 See C Saunders, ‘Constitutional Recognition of Local Government in Australia’ in N Steytler (ed), The Place and Role of Local Government in Federal Systems (Johannesburg, Konrad-Adenauer-Stiftung, 2005) 54. 27 The elected institutions were, however, later reinstated. 28 See ME Libonati, ‘Home Rule: An Essay on Pluralism’ (1989) 64 Washington Law Review 51. 29 See GA Tarr, Understanding State Constitutions (Princeton, Princeton University Press, 2000) 19f.

286 Local Government the transformation processes in a number of formerly authoritarian states, which all made local governments, only in their constitutions or also in actual practice, central pillars of new democracy.30 Article 28(2) of the German Basic Law of 1949, for example, restored local self-government, after it had been abolished in 1933.31 In Spain, municipal councils had already played a central role during the process of democratic transition (1975–78) so that the subsequent recognition of municipalities as self-governing institutions alongside the autonomous communities and provinces in Article 137 of the 1978 Spanish Constitution was anything but unexpected. Measures aimed at local self-government likewise preceded in Brazil the formal recognition of local authorities in Article 18 of the 1988 Constitution. In the Brazilian case, in particular, democracy from below was seen, as it was by India’s independence movement, not only as a bulwark against the return of authoritarian rule, but also as a catalyst for development. As mandated by Article 40 of the 1949 Indian Constitution, the states re-established, for these reasons, between 1959 and the late-1960s, the traditional panchayat councils as transformed democratically elected assemblies. 9.2.2 Scope and Quality of Recognition As far as the scope and quality of constitutional recognition is concerned, lengthy declarations do not necessarily imply a greater extent of self-government in practice. The Constitution of Brazil, for example, not only enshrines local autonomy in the above-mentioned Article 18, it also foresees already in Article 1 an explicitly threetiered state structure and moreover dedicates the entire Chapter IV to the role of the municipalities. Yet, the autonomy of municipalities is, as we shall see, more limited than in most other countries. South Africa takes a similar approach of particularly comprehensive constitutional recognition. First, Article 40(1) holds that ‘government is constituted as national, provincial and local spheres of government which are distinctive, interdependent and interrelated’. Then, the entire Chapter VII constitutionally protects democratic institutions, powers and local governments’ own revenue sources. While one-party dominance at all government levels has often been detrimental to genuine local autonomy, these constitutional guarantees have proved their importance in several court cases.32 Even if such comprehensive recognition may seem to provide an indication of the extent of protection, there is by no means an inherent link. This is illustrated by the German Basic Law, which essentially concentrates its treatment of the local government issue in one single provision, the above-mentioned Article 28(2).33 This

30

See Steytler (n 3) 396f and 406f. BVerfGE 167, 174 (Local Self-Governance) confirmed the conformity of Art 127 of the 1919 Weimar Constitution with the new Art 28(2). 32 See, eg, CDA Boerdery (Edms) Bpk v Nelson Mandela Metropolitan Municipality 2007 SCA 1 (RSA); City of Cape Town v Robertson 2005 (2) SA 323 (CC). 33 This clause is merely supplemented by provisions about the financial autonomy of municipalities in Arts 105–07. 31 1

Legal Recognition of Local Self-Government 287 provision simply states that ‘[w]ithin the limits prescribed by law, municipalities shall be guaranteed the right to regulate all local affairs on their own responsibility’, and yet endows municipalities with a comparatively high degree of autonomy.34 This is the case even if this clause is only an institutional guarantee that protects local self-government in principle and views municipalities as constituent parts of the Länder rather than a third order of government.35 Explicitly and without ambiguity, it grants this autonomy, like provisions in other constitutions, only ‘within the limits prescribed by law’, that is, law enacted by either the national or subnational parliament. Equally important for the scope of recognition as the clearly defined extent of selfgovernment is its justiciability. In fact, several more recently adopted or amended local government regimes, like South Africa’s in 1996 and Spain’s in 1999, refer to local autonomy not only as a loose guiding principle but also as a judicially enforceable right. A case in point is also the German Basic Law, which guarantees the enforceability of self-government before the constitutional court of the respective Land or, in a subsidiary manner, the Federal Constitutional Court (Article 93(1)(4b)).36 9.2.3 The Problem of (Over)Regulation The aforementioned examples of Brazil and South Africa demonstrate that there is a thin line between the constitutional recognition of local self-government and the (over)regulation of municipalities. The extensive treatment of local governments in both cases effectively constrained, compared to the dualism of older federal constitutions, the regulatory influence of subnational government. Instead, local governments are now under the influence of the national level through often excessively detailed constitutional and/or ordinary legislation. A glaring example of overregulation is Article 29(4) of the Brazilian Constitution, which specifies, according to population size, as many as 24 different categories of municipalities and establishes for each of them a maximum number of councillors. Austria is a similar, but much older, case. Articles 115–20 of the Constitution predetermine in an extensive way the organisation of municipalities, their powers and intergovernmental relations. This tight national constitutional regime vastly reduces the complementary power of the Länder to autonomously regulate local government under Article 115(2) of the Austrian Constitution and makes their legislation very homogeneous, if not identical.37 This tendency towards uniformity has also been further reinforced by extensive judicial interpretations of constitutional principles.

34 See M Goldsmith, ‘Local Politics in Europe’ in R Balme et al (eds), Les Nouvelles Politiques Locales (Paris, Presses de Sciences Po, 1999). 35 86 BVerfGE 148, 215 (Financial Equalization III). See M Burgi, ‘Federal Republic of Germany’ in Steytler, Local Government (n 3) 154. 36 According to this provision, the Federal Constitutional Court shall rule ‘on constitutional complaints filed by municipalities or associations of municipalities on the ground that their right to selfgovernment under Art 28 has been infringed by a law; in the case of infringement by a Land law, however, only if the law cannot be challenged in the constitutional court of the Land’. 37 See P Pernthaler and A Gamper, ‘Local Government in Austria’ in Steytler (ed), The Place (2005).

288 Local Government

9.A Austria 1993: Direct Election of Mayors Unconstitutional? In 1991, the Land of Tyrol foresaw in its Municipal Elections Act the direct elections of mayors. This created doubts as to whether such immediate electoral legitimation would be in line with the fundamental constitutional principle of representative democracy, which only features few exceptions and follows, according to the Constitutional Court, from the definition of Austria as ‘a democratic republic’ (Article 1 of the Constitution) and from the overall structure of the Constitution. The Court’s judgment recognised the power of the Länder, by virtue of the abovementioned Article 115(2), to regulate local government.38 In doing so, however, they would be bound to observe the implicit limits set by the abstract constitutional principle of (predominantly representative) democracy, which in this case overrides another such principle, that of federalism. Even an entrenchment of direct mayoral elections in the Constitution of the Land of Tyrol would not have made any difference because this subnational basic law is, under Article 99(1) of the Austrian Constitution,39 not allowed to affect the national Constitution as a whole and, even less so, its fundamental principles. In their reasoning, the judges referred to the fact that the directly elected local assembly would, in both organisational and functional terms, be the central institution of local self-government, to which all other municipal institutions are, through Article 118(5), accountable. In line with the general principle of (representative) democracy, this would create, at the local level, a parliamentary system that is also indicated by the silence of the Constitution on direct elections beyond those for the local assembly in Article 117(1–2). A change towards a dual system with two directly legitimated institutions would need explicit authorisation by the national Constitution. In response to this ruling, such authorisation was actually provided with the adoption of the new Article 117(6) in 1996. So far, six out of nine Länder have enacted constitutional provisions allowing for direct mayoral elections.

In Spain, detailed regulation stemming from the national level is not related to the Constitution but to pervasive basic legislation.40 Although the statutes of the autonomous communities expressly grant them legislative power over local government, Article 149(1)(18) of the Spanish Constitution still allows national law to set the ‘basic rules of the legal system of public administrations’ and thus also for the municipalities and provinces. As a result, there is an interplay of both national and subnational regulation, which entails a ‘two-fold nature’41 of the local government system. The case law of the Constitutional Court has been instrumental in identifying the boundaries of the Spanish Parliament’s basic legislation as against the constitutional guarantee of local government autonomy.

38

VfSlg 13500/1993. See box 5.D. 40 The relevant pieces of legislation are Law 7/1985 Regulating the Basis of the Local System and the Royal Legislative Decree 2/2004 of 5 March, which approved the Restated Text of the Local Tax Authorities Act of 1988. 41 STC 84/1982; STC 214/1989, FJ 11. 39

Legal Recognition of Local Self-Government 289

9.B Spain 1981–2004: The Guarantee of Local Administrative Autonomy In 1980, a Catalan law transferred all responsibilities from the four provincial councils to the government of the autonomous community. In effect, this total centralisation of powers would have abolished these provincial governments as institutions that have own tasks. Thus, it soon became a highly controversial issue whether such regional legislation would be in compliance with Article 137 of the Spanish Constitution, according to which the provinces enjoy, along with the municipalities and the autonomous communities, ‘selfgovernment for the management of their respective interests’. The Constitutional Court found the law in question to be unconstitutional and justified this decision with broader explanations about the different meanings of selfgovernment within the state of autonomies (Estado de las autonomías).42 According to the judges, the territorial framework outlined by the Constitution would grant political autonomy to the autonomous communities and merely administrative autonomy to the provinces and municipalities. This differentiation was, in their view, implied by a contextual reading of Article 137 that takes into account the overall structure of the Constitution. As to the provinces, their functioning as institutions with both delegated state tasks and autonomous powers exercised by representative councils is further entrenched in Article 141(1–2) of the Constitution. Even if administrative autonomy is evidently more limited than political autonomy, it still entails, according to the Court, an ‘institutional guarantee’ that protects provinces and municipalities in general as institutions not to be abolished. It does not, however, individually guarantee their existence or a specific content or scope of responsibilities. This means that the Constitutional Court has since then pointed out case by case the limits of local administrative autonomy against intrusions of regional and national basic legislation. In a judgment from 2004, for instance, the Court ruled that the constitutional guarantee of local autonomy prevents regional laws from denying municipal councils the power to initiate new urban plans.43 Regarding the provinces, the Constitutional Court has identified the classical function of cooperation with, and assistance for, municipalities, not least through the use of provincial spending power, as the minimum core of their administrative autonomy.44 In sum, the limits of both tiers of local government remain fluid.

In other cases, local governments are, as creatures of the subnational entities, exclusively subject to their regulation. This does not necessarily entail, from a local autonomy perspective, the absence of overregulation, but it certainly gives rise to more variation. This is epitomised by the example of the United States, where the states are entitled to regulate local government by virtue of the Tenth Amendment. The result is centralisation in Hawaii, for example, at one end of the spectrum, and decentralisation in New Hampshire, at the other.45 Whereas some states assign responsibility for education to counties and cities and others to single-purpose school districts, Hawaii is the only state that administers education directly.

42

STC 32/1981. STC 58/2004. 44 STC 109/1998. 45 See G Mathew, ‘The Functioning of Local Governments and their Relationship with Upper Levels of Government’ in Kincaid and Chattopadhyay (eds), Local Government (2008) 40. 43

290 Local Government 9.3 POWERS OF LOCAL GOVERNMENTS

9.3.1 Assignment of Local Powers The above-mentioned trend of increasing recognition of local self-government is certainly positive, though it remains insufficient if it is not accompanied by the actual transfer of significant powers. This link is underlined in Article 3 ECLSG, which defines local self-government as denoting ‘the right and the ability of local authorities, within the limits of the law, to regulate and manage a substantial share of public affairs under their own responsibility and in the interests of the local population’.46 Fundamentally, local government competences may be subdivided into autonomous and delegated powers, which they exercise on behalf of the subnational and, usually to a lesser degree, the national government. This duality is the result of both functional purposes, ie bringing administration closer to the people concerned, and also of historical patterns. In several cases, local governments were, like the US counties, initially local administrative units of subnational governments. Even if the subnational and local levels share this duality of autonomous and delegated powers, competences are, in the latter case, assigned quite differently. According to the above-mentioned paradigm of national–subnational dualism, there is a primary distribution of powers only between the national and subnational governments in the national constitution and a secondary allocation of powers to local government through ordinary legislation. A consequence of this large-scale deconstitutionalisation is less protection from interference. This holds true even though, for example, the constitutions of Brazil and South Africa expressly enumerate the powers of municipalities (Article 30 of the Brazilian Constitution and Section 156 of the South African Constitution). These appear, at first glance due to their broad wording, to be even potentially far-reaching. After all, a South African municipality has, beyond certain competences specified in Schedules 4 and 5 of the constitution, ‘the right to exercise any power concerning a matter reasonably necessary for, or incidental to, the effective performance of its functions’ (Article 156(5)). In practice, however, municipalities are in both cases vulnerable to national and subnational intervention through extensive and detailed framework legislation, even more so due to broad constitutional formulations.47 In the South African case, intervention occurs even though such legislation ‘may not compromise or impede a municipality’s ability or right to exercise its powers or perform its functions’ (Article 151(4)). Another problem concerning constitutionally listed local government powers is their need to be actualised, like in the case of India, by subnational legislation. It is the legislatures of the Indian states that ‘may, by law, endow the panchayats with such powers and authority as may be necessary to enable them to function as institutions of self-government’48 (Article 243G of the Indian

46 Art 4 ECLSG then goes on to emphasise the entrenchment of competences by the constitution or by statute, the principle of subsidiarity and the differentiation between autonomous and delegated powers. 47 For an overview and categorisation of local government powers, see de Visser, ‘Republic of South Africa’ (2009) 276ff. 48 Emphasis added.

Powers of Local Governments 291 Constitution) and thus transfer 29 subject matters listed in Schedule 11 of the Indian Constitution.49 This entrenchment of the states’ power to determine the extent of local autonomy was a concession to win sufficient political support for the local government reforms effected by the Seventy-Third and Seventy-Fourth Amendments of 1992. Presumably, it was also prompted by concerns that a radical change towards three-tiered federalism would, in view of the Supreme Court’s basic structure doctrine,50 risk being challenged and thus holding up the reforms for an indefinite period of time.51 Australian municipalities are, regarding their powers, similarly dependent on ordinary subnational legislation. But unlike those of India they have no formal constitutional support at either the national or the subnational level.52 Each state constitution leaves the determination of local government powers to ordinary state legislation. Whereas Article 71(2) of the Constitution of Queensland establishes at least an objective general standard for determining these powers, others merely refer to competences that the state legislature ‘considers necessary’.53 Regardless of which legal source determines local competences, this process is characterised, pursuant to diverse constitutional traditions, by two fundamentally different rationales: the concept of general competence or that of ultra vires. According to the latter concept, local governments only possess the powers that they are explicitly granted through enabling legislation. Particularly noteworthy is the specific variation of ultra vires in the United States, namely the so-called Dillon’s rule.54 Back in 1865, Judge John F Dillon famously stated in a ruling of the Supreme Court of Iowa that a municipal corporation possesses the following powers and no others: First, those granted in express words; secondly, those necessarily or fairly implied in or incident to the powers expressly granted; thirdly, those essential to the declared objects and purposes of the corporation, not simply convenient, but indispensable. Any fair, reasonable doubt concerning the existence of the power is resolved by the courts against the corporation, and the power is denied.55

Three years later, Dillon reinforced this rule in loftier words. Municipal corporations would ‘derive their powers and rights wholly from the legislature. It breathes into them the breath of life, without which they cannot exist. As it creates, so it may destroy. If it may destroy, it may abridge and control’.56 Even though the idea of local government being confined to expressly granted powers had already emerged in court judgments of the early nineteenth century,57

49 While this provision covers the panchayats established in rural areas, Art 243 W regulates, in a similar manner, the devolution of powers to the municipalities created in urban areas and refers to 18 subject matters enumerated in Schedule 12. 50 His Holiness Kesavananda Bharati Sripadagalvaru v State of Kerala, 1973 AIR 1461. See box 2.C. 51 V Vyasulu, ‘The Functioning of Local Government in Federal Systems: Perspectives from India’ in Kincaid and Chattopadhyay (n 9) 65. 52 See section 9.2.1 above. 53 See Saunders, ‘Constitutional Recognition’ (2005) 55. 54 See GE Frug, ‘The City as a Legal Concept’ (1980) 93 Harvard Law Review 1057. 55 Clark v Des Moines (City of) 19 Iowa 199 (1865). 56 Clinton (City of) v Cedar Rapids & Missouri River Railroad Co 24 Iowa 455 (1868). 57 See JG Grumm and RD Murphy, ‘Dillon’s Rule Reconsidered’ (1974) 416 The Annals of the American Academy of Political and Social Science 120, 122.

292 Local Government Dillon has to be credited with having popularised this rule, which thus came to be adopted by most state courts as a rule of statutory construction. As such, a judicially crafted non-mandatory rule to determine the legislative intent of statutes (in this case of state laws granting local government powers), Dillon’s rule is today practised to a greater or lesser extent in all except 10 US states.58 Although this rule has over time aroused considerable controversy, it enjoys broad recognition, which is not least the result of its confirmation by the US Supreme Court. 9.C United States 1903: US Supreme Court Endorsement of Dillon’s Rule? In 1891, the state of Kansas adopted a law that limited the working day to eight hours for all people ‘employed by or on behalf of the state of Kansas’ or a municipality of that state. Non-compliance with this rule was thereafter only allowed in certain situations of extraordinary emergency, namely war and the ‘protection of property or human life’. However, WW Atkin, a contractor with the municipal corporation of Kansas City for the construction of a brick pavement, required a labourer to work for 10 hours per day for the same pay. After Atkin had been sentenced to pay a fine, the matter was taken to the Supreme Court of Kansas, which upheld both the ruling and the law it was based on. The US Supreme Court eventually confirmed this decision.59 Atkin argued that the state’s eight-hour law would represent an unreasonable encroachment on his freedom of contract and would thus deny him the ‘equal protection of the laws’ guaranteed by Section 1 of the Fourteenth Amendment. The judges held, however, that this freedom does not allow a contractor of public works to decide on how to perform those works. Quite the contrary, it is the responsibility of the state to establish the conditions for such work. The judges clarified that the work was unquestionably of a public nature, even if it was not immediately taken care of by the state itself, because municipal corporations like those of Kansas City would be ‘the creatures—mere political subdivisions—of the state’ and ‘only auxiliaries of the state for the purposes of local government’.

There is no doubt that, historically, the principle of ultra vires has been closely linked with the common-law countries of the Anglo-American space. More recently, however, the concept’s impact has been weakened in some of these countries. Instead of expressly transferring single, narrowly defined competences, several Canadian provinces like Alberta, British Columbia and Ontario have started, over the last 15 years, to define wide areas of local jurisdiction.60 While the majority of Australian

58 For an overview of current practice, see JJ Richardson Jr, ‘Dillon’s Rule Is From Mars, Home Rule Is From Venus: Local Government Autonomy and the Rules of Statutory Construction’ (2011) 41 Publius 662. This article also offers an excellent treatment of the relations between Dillon’s rule, home rule and local government autonomy in general. 59 Atkin v Kansas 191 US 207 (1903). 60 See J Garcea and EC LeSage, ‘Municipal Reform Agendas and Initiatives: Analytical Framework and Overview’ in J Garcea and EC LeSage Jr (eds), Municipal Reform in Canada: Reconfiguration, Re-empowerment, and Rebalancing (Oxford, Oxford University Press, 2005).

Powers of Local Governments 293 states have similarly begun to circumscribe very broad responsibilities of local governments, the practical effect has been, through the prevailing dominance of the subnational level, much more limited. As a matter of fact, the states directly exercise a number of powers that are, in most other countries, left to the local level.61 Notwithstanding these cases of a trend from narrow to broader assignments, local governments are still dependent on powers allocated by the subnational level. Interestingly, this ultra vires logic is only fully reversed precisely in a few US states, which recently replaced Dillon’s rule by granting local governments the power to generally tax and regulate matters of local concern, unless this is disallowed by statute.62 In other words, these states switched to the concept of general competence, a paradigmatic example of which is the German constitutional tradition of the Allzuständigkeit. In sharp contrast to ultra vires, local governments in Germany may, and are indeed supposed to, assume responsibility for all affairs of local importance, even if they are not explicitly authorised by enabling legislation.63 9.D Germany 1988: Definition of the General Competence of Local Governments In Lower Saxony, a law from 1973 transferred the responsibility for waste management from the municipalities to the counties (Kreise). Thereafter, two municipalities made use of their right, according to the above-mentioned Article 93(1)(4b) of the Basic Law, to file a complaint before the Constitutional Court because they regarded this measure as violating their ‘right to regulate all local affairs on their own responsibility, within the limits prescribed by the laws’ (Article 28(2)). In their ruling, the judges broadly defined ‘local affairs’ as comprising, without regard to individual administrative capacity, all those matters that concern the living together of people in the municipality.64 Furthermore, they stated that the ‘limits prescribed by the law’ would not include merely the way local affairs were managed, but also municipal competence regarding these matters. In any case, this competence comprises all local affairs that have not been explicitly assigned by law to other levels of government. In others words, the municipalities are granted residuary power. Explicit authorisation by law is only required for municipal measures that regulate and restrict the ‘freedoms and property rights’ of individuals. Beyond this formally defined core of local self-government, Article 28(2) would also contain an obligation for both national and subnational legislation to distribute responsibilities in such a way that local affairs are only withdrawn from the municipalities on overriding grounds of public interest, above all the guarantee of their proper management. This principle in favour of municipal responsibility would apply in relation to all government levels, including the counties, and made the contested transfer of authority unconstitutional.

61 See DM Brown, ‘Federal-Municipal Relations in Australia’ in H Lazar and C Leuprecht (eds), Spheres of Governance: Comparative Studies of Cities in Multilevel Governance Systems (Montreal, McGill–Queen’s University Press, 2007) 103f. 62 For instance, Art 7 of the Constitution of Illinois and Art 10 of the Constitution of Alaska. 63 See V Mehde, ‘Steering, Supporting, Enabling: The Role of Law in Local Government Reforms’ (2006) 28 Law and Policy 164, 165. 64 79 BVerfGE 127 (Rastede).

294 Local Government Even though general competence is a main characteristic of Germany’s local government regime, the regulatory space of municipalities has been, in practice, considerably narrowed by the vast number of federal and Land statutes. In this light, the expectation that such general jurisdiction would necessarily entail far-reaching autonomy has thus been called by some a ‘mere fiction’.65 A very tangible impact of the Allzuständigkeit is, however, the absence of single-purpose institutions like school boards, which play an essential role, above all, in Anglo-American countries. This does not preclude that German general-purpose municipalities or the counties may agree to create joint administrative units to serve just a single purpose through various forms of inter-municipal cooperation.66 While ultra vires is anchored in the Anglo-American tradition, the concept of general competence is typical of several continental European countries. Apart from Germany, it is, for instance, firmly established in Switzerland and has more recently started to take root in Spain. Influenced by the German concept, the constitutional guarantee that local authorities ‘shall enjoy self-government for the management of their respective interests’ (Article 137) has been construed by several scholars as granting general competence.67 Several judgments of the Constitutional Court have, since the late-1990s, also pointed in this direction.68 Anyway, Spanish municipalities undertake today, in practice, significant activities beyond explicit authorisation, above all in the fields of social welfare and migrant integration. In Italy, municipalities are supposed to take care of all administration except that delegated to superior levels of government (Articles 114 and 118 of the Constitution). Within continental European systems, Austria is somewhat an outlier, as its municipalities do not have, with minor exceptions,69 their own original powers on the basis of the national Constitution. As a consequence, the municipalities may only act lawfully on the basis of competences that are expressly conferred upon them and circumscribed by either national or subnational legislation. However, this legislation must make them responsible for ‘all matters that exclusively or preponderantly concern the local community’ and are ‘suited to performance by the community within its local boundaries’ (Article 118(2) of the Austrian Constitution). But even if laws fail to comply with these criteria, they remain in force and are binding for municipalities until struck down by the Constitutional Court.70 This limitation evidently contrasts starkly with the general freedom of municipalities in Germany and other countries to become active in all affairs of local importance. In most cases, the distribution of powers is uniform for all municipalities and therefore fails to take into account the above-mentioned increasing urban–rural divide. The Austrian Constitution, for instance, still adheres to the ‘principle of the

65

See Mehde, ‘Steering, Supporting’ (2006) 168. See section 9.4.2 below. 67 See F Velasco Caballero, ‘Autonomía Local y Subsidiariedad en la Reforma de los Estatutos de Autonomía’ in T Font i Llovet (ed), Anuario del Gobierno Local 2004 (Madrid, Marcial Pons, 2005). 68 For example, STC 159/2001 concerning urban planning. 69 These are private-sector activities (Art 116(2) of the Austrian Constitution) and the issuing of local police ordinances (Art 118(6)). 70 VfSlg 6944/1972; VfSlg 8719/1979. 66

Powers of Local Governments 295 abstract uniform municipality’, as enshrined in 1920. This means that, with the minor exceptions of towns with their own statutes and the capital Vienna,71 all municipalities enjoy, also regarding their competences, equal legal status regardless of variations in territorial size, population or economic and administrative capacities. When the Constitutional Court makes the above-mentioned assessment about the powers to be transferred (Article 118(2)), it thus has to judge local concerns and local capacities uniformly with regard to both Vienna and tiny rural villages. A notable exception to the pattern of symmetry is India, where (rural) panchayats and (urban) municipalities, which are themselves merely types of local government with quite diverse subtypes,72 are treated differently. The above-mentioned separate lists of transferable subject matters in Schedules 11 and 12 each contain local government tasks specific to rural and urban areas, respectively. Though India’s differentiation is still an outlier, symmetry may come under more and more pressure in the long run. In fact, claims of increased competences for larger urban local governments have recently become a widespread phenomenon in countries as diverse as Canada, Brazil and South Africa.73 9.3.2 Restrictions of the Exercise of Local Powers As demonstrated in the previous section, legal rules concerning the allocation of powers to local governments have recently undergone profound reforms in a number of cases. Today, they therefore vary significantly in terms of legal entrenchment, autonomous and delegated competences, as well as asymmetrical or symmetrical designs. The question is then whether these reforms have also entailed in practice a broader pattern of extended authority and thus, contrary to the above-mentioned traditional focus on the national and subnational levels, reinforced the position of local governments. On the whole, the scope and effective exercise of local government powers appears to be limited by three widespread problems: first, the overburdening with delegated tasks; second, the role of subnational and/or national governments as strict supervisory authorities rather than cooperation partners on an equal footing; and third, scarce financial resources. 9.3.2.1 Overburdening with Delegated Tasks Nearly three decades ago, a comprehensive comparative study concluded, at least for industrialised countries, that local governments would play a ‘major role in the

71 Art 116(3) of the Austrian Constitution offers municipalities with at least 20,000 inhabitants the option to become towns with their own statute and thus to assume more responsibilities. The capital, Vienna, has different competences insofar as it is at the same time both a municipality and one of the nine Länder (Arts 108–12). 72 The 73rd Amendment concerned the (rural) panchayats to be established at the village, intermediate and district levels (Art 243 B(1)) and the 74th Amendment the (urban) municipalities. The latter include nagar panchayats for areas in transition from a rural area to an area with urban character, municipal councils for smaller urban areas and municipal corporations for larger urban areas (Art 243 Q(1)). 73 See Steytler (n 3) 432.

296 Local Government delivery of fundamental collective public and quasi-public goods’.74 This finding implies that local governments were already important back then, though they were largely confined to the delivery of basic services and thereby to implementing regulations rather than themselves playing a regulatory, even quasi-legislative, role. In more provocative terms, they were, as is sometimes said about municipalities in Australia, mainly regarded as being responsible for no more than ‘roads, rubbish and rates’.75 There appears to be no doubt that the core competences of municipalities are still centred on such issues as the provision of fundamental household utilities, the planning and building of infrastructure, culture and leisure activities, pre-school education, as well as sometimes primary and secondary schooling, social welfare and health services. The latter three areas are obviously particularly costintensive and therefore account in large part for significant variations concerning the local share of total government expenditures.76 Notwithstanding these persistent core functions, local authorities, among them especially larger urban municipalities, have started to go beyond these as a result of their changed role in a globalised context.77 This is illustrated in particular by the emerging role they play in several countries in areas such as environmental protection and economic development. Engagement in the latter field had for a long time been traditional only in the United States, but it has more recently become a general pattern.78 Nonetheless, there seems to be, in many cases, a mismatch between delegated and autonomous powers overall. The problem that local governments are increasingly and excessively consumed with delegated tasks and thus lack political and administrative capacities to adequately focus on their autonomous powers has recently come to the fore in a number of countries. South Africa and even Germany, where municipalities are usually deemed to be comparatively powerful, are just two prominent examples.79 As far as German municipalities are concerned, estimations implied already a decade ago that they perform up to 95 per cent of all their tasks on behalf of the other two government levels.80 Although the large-scale imposition of delegated tasks affects all local governments, smaller ones are evidently more prone to being overburdened. Such mandates tend to reinforce often strong variations between smaller and larger local governments regarding the levels of administrative

74 JJ Hesse and LJ Sharpe, ‘Local Government in International Perspective: Some Comparative Observations’ in JJ Hesse (ed), Local Government and Urban Affairs in International Perspective (Baden-Baden, Nomos, 1990) 608 (emphasis added). 75 Quoted in N Steytler, ‘Introduction’ in Steytler (n 26) 6. 76 For a comparative overview, see Steytler (n 3) 415f. 77 See section 9.1 above. 78 For comparative evidence of this, see B Denters and LE Rose, ‘Toward Local Governance?’ in Denters and Rose, Comparing Local Governance (2005) 247ff. 79 See R Hrbek and JC Bodenbender, ‘Municipal-Federal Relations in Germany’ in Lazar and Leuprecht (eds), Spheres of Governance (2007); N Steytler, ‘National, Provincial, and Local Relations: An Uncomfortable Ménage à trois’ in Lazar and Leuprecht (n 61). 80 See M Kuban, ‘Die Anliegen der Kommunen in der Kommission zur Modernisierung der bundesstaatlichen Ordnung (Föderalismus-Kommission)’ (2004) Zeitschrift für Kommunalfinanzen 113, 114.

Powers of Local Governments 297 capacities, equipment and skills that have been observed especially, but not exclusively, in developing countries.81 9.3.2.2 Supervision Prevailing over Intergovernmental Collaboration A second major restriction to the exercise of local government powers derives, in a number of cases, from the nature of intergovernmental relations with the subnational and national level. Put simply, these relations may be geared towards local governments as cooperation partners on a more or less equal footing or as subordinate units under the tutelage and tight supervision of superior government levels. Whereas in the United States and Canada, for instance, formal intergovernmental mechanisms, including local governments, are notably absent, such instruments are quite prominent in several other countries. These include the formal institutional participation of local governments in subnational or national institutions or the far more widespread functional participation through a consultative role during the legislative process. The latter approach is adopted by a three-level consultation mechanism in Austria that brings together the national Government, the Länder and organised local government, which is, unlike most functional participation, not informal but is even guaranteed by the Constitution.82 If the national or subnational government plans to pass legislation or by-laws entailing financial obligations for other government levels, the matter is, upon the request of one of the parties, relegated to a tripartite consultation committee. In default of an agreement within this body, the party considering the regulation is also responsible for its financing. Until recently, this consultation mechanism had always given rise to consensual political solution. As recently as 2014, the Constitutional Court found for the first time that the mechanism had been violated by the federal Government concerning costs for railway crossings for Austrian municipalities.83 Italy similarly relies on multilateral cooperation that includes local governments. The Standing Conference for the Relations between the State and Regions84 meets in three different compositions, two of which include the local level as well. These are the State-Cities and Local Autonomies Conference and the Joint Conference that brings together all three levels of government.85 Even though this comprehensive conference system primarily serves as a forum of debate between the executives, its opinions have, in certain cases, actually become binding.

81 Concerning developing countries, see Steytler (n 3) 418. An example of a developed country with similar discrepancies is Spain. See R Agranoff, ‘Local Government in Spain’s Multilevel Arrangements’ in Lazar and Leuprecht (n 61). 82 The mechanism is based on a constitutional law (BGBl I 1998/61) and a constitutional concordat concluded in accordance with Art 15a of the Austrian Constitution. 83 VfSlg 19868/2014. 84 See ch 6.2.2. 85 See R Bifulco, ‘The Italian Model of State–Local Autonomies Conferences (also) in the light of Federal Experiences’ in J Luther et al (eds), A World of Second Chambers: Handbook for Constitutional Studies on Bicameralism (Milan, Giuffrè, 2006).

298 Local Government In contrast to the Austrian and Italian cases, the South African Constitution provides an outstanding example of a much less common form of institutional participation. Based on a conceptual framework of ‘cooperative government’ of three distinct interrelated and interdependent spheres of government (Sections 40–41 of the South African Constitution), it not only foresees functional involvement in the legislative process86 but also a justiciable obligation on the part of the national and provincial governments to ‘support and to strengthen the capacity of municipalities to manage their own affairs’ (Section 154(1)). Section 67 also guarantees institutional participation by authorising organised local government, in practice the South African Local Government Association, to designate up to 10 members, albeit without the right to vote, of the National Council of Provinces. In recognition of the decreasing importance of second chambers,87 emphasis is thus placed, similar to other countries like Switzerland and Australia, on the involvement of local government in national institutions representing the executives.88 All of the aforementioned cooperation mechanisms appear to feature two common patterns. The first concerns the strong role of local government associations, which are, in Austria and South Africa, even recognised by the respective national constitution.89 These function not only as lobbying groups but also as central negotiating partners. Speaking with one voice is, however, complicated by the difficulties in reconciling the increasingly divergent interests of large urban and small rural local governments, which prompted countries like Austria and Switzerland, with many small municipalities, to form separate associations. A second pattern is the erosion through functional and sometimes direct institutional participation of federal systems’ traditional national–subnational dualism, according to which subnational governments are seen as the agents of local governments entrusted with promoting their interests.90 Such a role has, however, come to be seen as difficult due to the frequently contradictory interests of these government levels. Furthermore, dualism is often deemed unfit to reflect the political reality. The fact that intergovernmental conferences in Canada, for instance, include the premier of the tiny province of Prince Edward Island and exclude the politically more

86 Sec 154(2) provides the following: ‘Draft national or provincial legislation that affects the status, institutions, powers or functions of local government must be published for public comment before it is introduced in Parliament or a provincial legislature, in a manner that allows organised local government, municipalities and other interested persons an opportunity to make representations with regard to the draft legislation’. 87 See ch 6.2. 88 In Switzerland, the Swiss Union of Cities and the Swiss Union of Municipalities has, since 2001, formed part, together with the federal Government and the Conference of Cantons, of the tripartite agglomeration conference. The Australian Local Government Association is represented in the Council of Australian Governments alongside the prime minister of Australia, the premiers of the states and the chief ministers of the territories. 89 While Art 115(3) of the Austrian Constitution explicitly names the Austrian Association of Cities and Towns and the Austrian Association of Municipalities as representing the interests of local governments, the South African Constitution speaks, more generally, of ‘organised local government’. It does so, however, on several occasions and additionally requires national legislation to establish the details about the recognition of such organisations, at both the national and provincial level, as well as their involvement in intergovernmental relations (Art 163). 90 See H Risse et al, ‘Local Government in Federal Systems’ in Kincaid and Chattopadhyay (n 9) 11.

Powers of Local Governments 299 powerful mayor of Toronto, has been increasingly regarded as unjustified. While provincial governments in Canada still mediate the contacts between ‘their’ municipalities and the federal Government to a greater or lesser extent,91 there is in several other countries a clear trend towards the inclusion of local governments in tri-level mechanisms and/or direct local–national relations.92 Similar to cooperative intergovernmental relations of local authorities, their supervision is also still characterised by a predominant role of subnational governments. This holds true in spite of certain practices of, above all financial, control exercised by the national level. Taking both superior government levels together, the individual scope of their supervisory tools varies greatly, even if some control is common to all systems of local government. In this regard, the statement of the German Constitutional Court about supervision as ‘correlative of the right to self-government’93 can be generalised and illustrates the more limited nature of local self-government compared to subnational autonomy. At one end of the spectrum are countries like India, where the states exercise tight control themselves and/or empower, for this purpose, the district collector, that is, an office of the Indian Administrative Service and British legacy.94 At the other end is Spain, where the Constitutional Court largely excluded in one of its early landmark judgments at least rigid supervision that would have amounted to controlling political aspects or efficiency.95 Article 63 of the 1985 Law Regulating the Basis of the Local System even largely rules out controls of legality, that is, the only form, according to this ruling, of supervision compatible with the above-mentioned administrative autonomy of local governments,96 and thus leaves this exclusively to the courts. Only in extreme cases of unlawful actions is the national Government allowed to resort to strict measures such as the dismissal of a local council, which are more accessible in several other countries.97 In addition, formal supervision over expenditures of the Spanish municipalities is to some extent limited because the Constitutional Court construes the financial autonomy clause in Article 142 of the Constitution as ensuring their spending power.98 However, the above-mentioned problem of delegated tasks entails that their spending is, in practice, largely predetermined. Yet, stricter regimes of expenditure control, exercised mostly by subnational governments, are in place in several countries. They range from mere monitoring on the basis of local governments’ reporting

91 This mediation is stronger in the province of Quebec and weaker in Ontario, in particular regarding Toronto. The latter city is usually able to pursue its interests through informal political channels and sometimes even formalised tri-level mechanisms. Policies regarding settlement services for migrants are a case in point. See ch 13.2. 92 See C Leuprecht and H Lazar, ‘From Multilevel to ‘Multi-order’ Governance?’ in Lazar and Leuprecht (n 61) 7f. 93 26 BVerfGE 228 (Sorsum). 94 For an overview and categorisation of the far-reaching supervision powers, see G Mathew and R Hooja, ‘Republic of India’ in Steytler, Local Government (n 3) 189f. 95 STC 4/1981, FJ 2. 96 See section 9.2.3 above. 97 In practice, these have hardly ever been used. See F Velasco Caballero, ‘Kingdom of Spain’ in Steytler, Local Government (n 3) 319). 98 STC 48/2004, FJ 10.

300 Local Government duties to severe intervention, which may include, as in Brazil, control over spending quotas reserved for certain purposes99 or mandatory approval of expenditures exceeding certain limits.100 The extent to which supervision really restricts the exercise of local government powers is determined by a number of factors, including the stability and administrative capacity of local governments or their relative size and strength with regard to the respective subnational entity. But legal traditions and legal frameworks themselves are also of crucial importance. Diverse legacies, for instance, seem to account for the prevalence of legal supervision in Europe and political control in North America,101 as well as within Switzerland for the less rigorous oversight in Germanspeaking cantons than in French-speaking ones.102 In a number of countries, most notably in Germany, Austria and Switzerland, the degree of supervision also varies depending on the legal nature of local government competences. Whereas they are subject to instructions and controls of appropriateness regarding delegated powers, superior authorities are, regarding autonomous powers, typically confined to judging the lawfulness of local decisions and actions. This differentiation of supervision according to the type of competence is also stipulated by Article 8(2) ECLSG.103 9.3.2.3 Limited Financial Resources On the whole, supervision therefore does not seem to be the main restriction on the effective exercise of local government powers. In this regard, much more incisive constraints stem, in many cases, from limited financial resources that local authorities may spend freely according to their own preferences. This problem results from both heavy dependence on conditional transfers and generally scarce resources, which leads to a mismatch between local government powers and their funding. Both these aspects are addressed by Article 9 ECLSG.104 Quite frequently, the revenues of local authorities are heavily dependent on conditional transfers. Evidently, there are good reasons for intergovernmental transfers insofar as local governments exercise delegated competences and deliver, particularly in the case of large urban municipalities, services of wider regional or even national

99 Arts 77 and 212 of the Constitution, respectively, oblige municipalities to earmark at least 15% of their revenues from taxes and transfers for health and 25% for education. In case of a failure to comply with this rule, the state or, regarding municipalities in a federal territory, the national government has the power to intervene (Art 35(3)). 100 See S Jha, ‘Panchayats: Functions, Responsibilities and Resources’ (2004) National Institute of Rural Development Hyderabad, 8 fincomindia.nic.in/writereaddata/html_en_files/oldcommission_html/ predocs/speech/shikha%20jha.pdf. 101 See Steytler, ‘Introduction’ (2005) 7. 102 See A Ladner, ‘Swiss Confederation’ in Steytler, Local Government (n 3) 347f. 103 Art 8(2) reads as follows: ‘Any administrative supervision of the activities of the local authorities shall normally aim only at ensuring compliance with the law and with constitutional principles. Administrative supervision may however be exercised with regard to expediency by higher-level authorities in respect of tasks the execution of which is delegated to local authorities’. 104 Art 9 reads as follows: ‘(1) Local authorities shall be entitled, within national economic policy, to adequate financial resources of their own, of which they may dispose freely within the framework of their powers. (2) Local authorities’ financial resources shall be commensurate with the responsibilities provided for by the constitution and the law.’

Powers of Local Governments 301 interest. But comparative evidence points to excessive dependence, in several cases, on transfers as a share of their total resources.105 Of course, this is only a very rough measure of financial autonomy because it does not indicate the extent to which these grants are unconditional, earmarked for specific purposes or otherwise tied to more or less restrictive conditions on their spending. In some cases of particularly stringent conditions like Brazil, these reflect superior authorities’ profound distrust of local government officials’ financial management.106 The unconditional or conditional nature of these grants, which defies clear trends,107 is thus much more important than their origin from the national or subnational government. In accordance with the traditional understanding of local governments as being subordinated to the respective subnational government, transfers typically stem in cases like the United States or Switzerland from the states and cantons, respectively. In other cases, national governments, which are, due to the usual vertical fiscal imbalance, financially more powerful, have direct financial relations with local authorities or, like in India, transfer funds to them via the subnational level. In the latter context, subnational governments are in certain countries accused of having ‘sticky fingers’108 in passing on these funds to local governments and thereby further reducing their financial resources. The same problem has also emerged in Mexico. According to the country’s Supreme Court of Justice, Article 115 of the Constitution contains a legal obligation for the federal Government to ensure the timely and effective transfer of funds to Mexican municipalities, through the states.109 As far as the often-observed mismatch between local government competences and their financial resources is concerned, there are at least two main reasons. One concerns underfunded or even unfunded mandates; that is, the shift of responsibilities and their costs onto local authorities without adequate funding. De facto devolution involving at the same time the defunding of local government has been a matter of concern in the United States since at least the 1980s.110 It is true that the Congress passed the Unfunded Mandates Reform Act (UMRA) in 1995, after similar regulations had been enacted in state constitutions,111 and thereby obliged itself to estimate the fiscal effects of envisaged legislation. But the procedure very much centres on the Congressional Budget Office and does not involve, unlike the above-mentioned tripartite Austrian consultation mechanism, local governments

105 For an overview of local government with regard to their financial self-reliance, see Steytler (n 3) 418ff. The overview demonstrates the enormous variety regarding the share of local governments’ own revenues of total financial resources, which ranges from less than 10% in Nigeria and India to more than 75% in Switzerland, South Africa and Australia. 106 See LC de Queiroz Ribeiro and SG Braule Pinto, ‘Brazil’ in Steytler, Local Government (n 3) 90. 107 See Steytler (n 3) 422f. 108 Risse et al, ‘Local Government’ (2008) 12. 109 Mexican Supreme Court of Justice, Constitutional Controversy 5/2004. 110 See J Kincaid, ‘De Facto Devolution and Urban Defunding: The Priority of Persons over Places’ (1999) 21 Journal of Urban Affairs 135. Regarding the abandonment of cities by the national Government in the social welfare area through the retreat from the joint programmes of the 1960s, see ch 11.1. 111 See JF Zimmermann, ‘State Mandate Relief: A Quick Look’ (1994) 20 Intergovernmental Perspective 28.

302 Local Government themselves. For this reason and because of the exemption of certain forms of regulations, some observers have cast doubt on UMRA’s appropriateness to curb national interventionism.112 Among European countries, a decades-long trend in Germany of shifting costintensive responsibilities, above all social welfare, without sufficient funding to local governments, triggered a particularly intense debate and eventually a constitutional amendment. Since the federalism reform roughly a decade ago, federal legislation has been banned from entrusting municipalities and associations of municipalities with any tasks (Articles 84(1) and 85(1) of the Basic Law). Also in Spain, the national Constitution addresses the problem of unfunded mandates. Article 142 expressly declares that ‘local treasuries must have sufficient funds available in order to perform the tasks assigned by law’. The country’s Constitutional Court has construed this clause in such a way that it does not entitle them to determine their own revenue sources, but only refers to sufficient income whatever its source may be.113 The national Government is therefore responsible for ensuring, in the process of redistributing the total state revenues, the sufficiency of local funds stipulated by Article 142, as well as the territorial balance and solidarity mandated by Article 138.114 Their dependence on transfers from the national government made many local governments particularly vulnerable to austerity measures in response to the global financial crisis. In 2012, all but two EU Member States (the United Kingdom and the Czech Republic) signed the European Fiscal Compact,115 an intergovernmental international agreement outside the scope of EU law that requires all parties to introduce constitutional changes towards balanced-budget rules. In the Spanish case, this was done through the entrenchment in Article 135 of the Constitution of an internal stability pact that closely mimicked, in terms of its structure, the compact at the European level.116 In practice, austerity measures often aggravated the problem of local governments with underfunded or even unfunded mandates. Beyond these, a second reason for the imbalance between powers and funds is linked to the widespread lack of local governments’ profitable own revenue sources. In this regard, national constitutions foresee, usually in combination with much more detailed enabling legislation, three typical, but not very lucrative, sources. Fees, or user charges, play a significant role in some cases like US municipalities, counties and, understandably above all, special districts.117 However, they are in many countries very far from even covering the costs of the services provided. Borrowing is another typical source of local governments’ own revenues, but it is often subject

112

See P Posner, ‘The Politics of Coercive Federalism in the Bush Era’ (2007) 37 Publius 390, 402f. STC 4/1981, FJ 15. 114 STC 96/1990. 115 Officially called the Treaty on Stability, Coordination and Governance in the Economic and Monetary Union. 116 See V Ruiz Almendral, ‘Curbing the Deficit in Spain and its Autonomous Communities: A Constitutional Conundrum’ (2013) 1 Tijdschrift voor Constitutioneel Recht 68, 71. 117 See MA Pagano, ‘United States of America’ in Steytler, Local Government (n 3) 375. 113

Powers of Local Governments 303 to tight regulation and control. This may involve, for instance, restrictions regarding the maximum loan amount or the purpose of expenditure118 or even prior approval. The third and obviously most important source of local governments’ own revenues are taxes, most typically commercial taxes and property rates.119 The latter are the most traditional local government tax and commonly exclusively assigned to them. Even though they account for the bulk of tax revenues, especially in countries with a British legacy, they often do not generate large proceeds for two distinct reasons. First, as a tax on wealth, the revenues of property rates do not increase, unlike those of progressive taxes, with economic growth. Second, their potential is often not fully exploited due to the absence of enabling legislation, inefficient tax collection or the absence of the political will to increase the tax burden.120 The second main levy that most local governments have access to are various kinds of commercial taxes. These are even the mainstay of income for German municipalities, whose right to set tax rates ‘within the framework of the laws’ is, as with property taxes, even constitutionally guaranteed (Article 28(2) in conjunction with Article 106(6) of the Basic Law). The extent of this guarantee was a matter decided by the Federal Constitutional Court. 9.E Germany 2010: Minimum Assessment Rate of 200 Per Cent? Since 2004, German municipalities have been obliged by federal legislation to impose for taxes on trades an assessment rate of not lower than 200 per cent. Local governments are thereby prevented from freely choosing the level of their surtax and from raising no surtax at all. Two municipalities from the Land of Brandenburg regarded this restriction as a violation of local self-government and thus filed a complaint with the Constitutional Court by virtue of the above-mentioned Article 93(1)(4b) of the Basic Law. The judges stated that the 2004 law would clearly fall within the legislative power of the federal Parliament, as in this case ‘the maintenance of legal or economic unity renders federal regulation necessary in the national interest’ (Articles 72(2) and 105(2) of the Basic Law).121 Moreover, the power of municipalities to set the rates for taxes on trades is, as mentioned above, expressly bound to be exercised ‘within the framework of the laws’ (Article 106(6)). Nevertheless, none of these laws may, according to the judges, subject this local power to unreasonable restrictions that would encroach upon the core of financial autonomy as guaranteed by this provision and, more broadly, by Article 28(2). In this case, however, federal legislation would pursue two legitimate aims, namely the elimination of local tax havens and the avoidance of some municipalities receiving income tax shares without participating in their reciprocal financing through local taxes on trades. Even if the underlying rationale of the power to set assessment rates is tax competition between the municipalities, the common good would mandate restrictions in case of such competition

118 In Austria, borrowing is only allowed for capital investment and not for the coverage of operational expenditures. See A Kiefer and F Schausberger, ‘Republic of Austria’ in Steytler, Local Government (n 3) 54. 119 ‘Property taxes’ are known in the United States and ‘property rates’ in the United Kingdom, as well as in a number of countries with taxation systems deriving from the British system. 120 See Steytler (n 3) 421. 121 125 BVerfGE 141 (Trade Tax).

304 Local Government

becoming detrimental to it. The Court held that, in view of these legitimate aims, the prescription of a minimum assessment rate of 200 per cent, which was below the average, would still leave municipalities considerable leeway and did not amount therefore to an unreasonable restriction and violation of the core of financial autonomy.

In contrast to their powers regarding property and commercial taxes, local governments are in most cases prevented from raising the particularly profitable income taxes. Swiss municipalities are a notable exception, as they are traditionally allowed to impose within the limits of cantonal legislation a surcharge on personal and corporate income taxes of the respective canton. This has led to considerable tax level variations not only across cantons but also within them.122 It is obvious that the combined effect of the twin problems of unfunded mandates and limited revenue generation may result in financially distressed local governments. Paradoxically, they are, despite their limited resources, expected and required to be financially viable entities. This is best expressed by Section 135(1) of South Africa’s 2003 Municipal Finance Management Act, according to which ‘[t]he primary responsibility to avoid, identify and resolve financial problems in a municipality rests with the municipality itself’. In line with this approach, the above-mentioned constitutional obligation to support and strengthen local governments does not include the duty to bail out a bankrupt municipality.123 Also in Switzerland, municipalities are expected to be financially self-reliant, which they effectively are to a greater degree than in other countries. If a local government fails to meet this expectation, it is not for other government levels to pay off its debts. Instead, the lack of selfsufficiency entails a limitation of financial autonomy. Cases of bankruptcy, which occur rather rarely, lead to all finance-related decisions requiring acceptance by the respective cantonal authorities.124 Even if the problem of limited financial resources concerns local governments in general, it affects small rural municipalities much more severely than others. Large urban municipalities typically benefit from above-average revenues from both user charges and the aforementioned main local taxes. They are also the ones that primarily engage in borrowing. Quite evidently, fiscal disparities are also closely linked to the urban migration of the mainly young and active population, which entails for rural municipalities at the same time a drain on revenues and an increase in costs. This gap widens as the process intensifies. Similar to subnational entities,125 the different economic and fiscal potentials of local governments may also be balanced through equalisation mechanisms, the establishment of which is stipulated by Article 9(5) ECLSG. Today, such mechanisms are in place in a number of countries and involve, according to different designs,

122 See IB Adamovich and G Hosp, ‘Fiscal Federalism for Emerging Economies: Lessons from Switzerland?’ (2003) 33 Publius 1, 7ff. 123 MEC, Mpumalanga v Imata 2002 (2) SA 76 (SCA). 124 See Ladner, ‘Swiss Confederation’ (2009) 348. 125 See ch 7.4.

Structure of Local Government 305 vertical transfers from the subnational to the local level, from the national to the local level mediated by the subnational entities, or from the national level to both other government levels.126 Although the main gap concerning financial resources usually runs along the urban–rural divide, this dualism is in certain cases an oversimplification. A differentiation is needed regarding the special situation of cities and their suburbs.127 Frequently, the local governments of cities have to cover higher expenditures, which are partly due to commuters from suburbs using their services. On the other hand, they may usually count, as mentioned above, on higher revenues from user charges and taxes. 9.4 STRUCTURE OF LOCAL GOVERNMENT

From a comparative point of view, it is puzzling how countries differ regarding the number of local governments in relation to population size. Spain and South Africa, for instance, have similar population figures but 8,151 and 234 municipalities, respectively.128 Notwithstanding urbanisation, Spain has so far shied away from significantly cutting down on the number of municipalities, which is, as in some other countries, still roughly at pre-industrialisation levels. By contrast, South Africa has undertaken, to some extent facilitated by the broader transformation process after apartheid, a complete territorial reform of local government and through consolidation created very large municipalities. The contrasting numbers of municipalities in both countries also reflect different responses to a salient challenge associated with the increasing demographic disparities between small rural and large urban local governments: to achieve the aim of administrative viability and efficiency, while at the same time maintaining autonomy of local governments and their accountability to the local population. In this regard, South Africa followed the more radical path of amalgamation. Spain has instead mostly relied on inter-municipal cooperation, as well as on umbrella entities. In essence, all these three responses touch upon the structural issues of the number and size of local governments and the relations between them. 9.4.1 Amalgamation It is evident that amalgamation is, from the viewpoint of self-government, the least preferable option. Yet, the growing urban–rural divide has in recent decades

126 See B Dafflon and F Vaillancourt, ‘Problems of Equalization in Federal Countries’ in R B lindenbacher and A Koller (eds), Federalism in a Changing World: Learning from Each Other (Montreal, McGill– Queen’s University Press, 2003) 395. 127 See E Slack and R Chattopadhyay, ‘Governance and Finance of Large Metropolitan Areas in Federal Systems: Summary of the Findings and Concluding Comments’ in Slack and Chattopadhyay (eds), Governance and Finance (2013) 351ff. 128 As to South Africa, this figure refers to the eight metropolitan municipalities and 226 local municipalities. It does not include the 44 district municipalities that comprise the local municipalities and are therefore umbrella units.

306 Local Government prompted a number of countries to undertake comprehensive territorial reforms. In Australia, extensive mergers, especially in the 1990s, have sparked contentious debates about the above-mentioned missing constitutional protection for municipalities. Also in Canada, provincial governments have typically imposed amalgamations without the approval of, and often even without consulting, the municipalities concerned.129 In the case of the 1997 City of Toronto Act, which consolidated the upper-tier Metro Toronto Government and six lower-tier municipalities into one single structure, this was even enforced against a negative referendum result.130 Particularly interesting is South Africa’s process of municipal demarcation initiated in 1998. In compliance with Section 155(3)(b) of the Constitution, an independent body was entrusted with the task of redrawing boundaries, ie the Municipal Demarcation Board. Even though the members of this institution are appointed by the President of South Africa for a term of five years, they do not merely have an advisory role but ultimate authority regarding boundary issues.131 Until 2000, the Municipal Demarcation Board reduced the number of municipalities from 843 to 283.132 As far as Europe is concerned, Article 5 ECLSG requires some form of consultation of the local population,133 but nonetheless allows imposed mergers. In fact, several countries, even those with historically deeply rooted municipalities, have carried out more or less consensual territorial reforms. Particularly noteworthy is Germany, which saw, during the late 1960s and early 1970s, a reduction in the number of municipalities from 24,300 to 8,500. This first wave, characterised by substantial procedural variations between the Länder, was then followed after unification by amalgamations in the new eastern part of the country.134 Remarkably, this long process was successful in striking a balance between efforts of the national governments towards reorganisation, on the one hand, and respect for autonomy of the Länder, on the other. In neighbouring Switzerland and Austria, by contrast, territorial reforms have been, despite similar problems with depopulated rural municipalities, less farreaching, albeit for different reasons. Even though in the Swiss case more than 300 municipalities have disappeared since the 1990s, this process has been cumbersome for procedural reasons because cantonal constitutions typically require a positive

129 See H Lazar and A Seal, ‘Local Government: Still a Junior Government? The Place of Municipalities within the Canadian Federation’ in Steytler (n 26) 31. 130 On this controversial merger and the ensuing legal dispute, see B Moore Milroy, ‘Toronto’s Legal Challenge to Amalgamation’ in C Andrew et al (eds), Urban Affairs: Back on the Agenda (Montreal, McGill–Queen’s University Press, 2002). 131 Executive Council of the Province of the Western Cape v Minister for Provincial Affairs and Constitutional Development, Executive Council of KwaZulu-Natal v President of the Republic of South Africa 2000 (1) SA 661, para 14. 132 See H Klug, The Constitution of South Africa: A Contextual Analysis (Oxford, Hart Publishing, 2010) 275. 133 Art 5 reads as follows: ‘Changes in local authority boundaries shall not be made without prior consultation of the local communities concerned, possibly by means of a referendum where this is permitted by statute’. 134 See H Wollmann, ‘The Two Waves of Territorial Reform of Local Government in Germany’ in J Meligrana (ed), Redrawing Local Government Boundaries (Vancouver, University of British Columbia Press, 2004) 111.

Structure of Local Government 307 popular vote in each municipality concerned. A particularly profound reform that reduced, in 2006, the number of municipalities in the canton of Glarus from 25 to three obtained the approval of the Landsgemeinde, that is, the assembly of all people entitled to vote. 9.F Switzerland 2006: Radical Territorial Reform in Glarus through Direct Democracy? Both the legislature and executive of the canton had endorsed at the Landsgemeinde a rather moderate reduction to 10 municipalities. During the debates at the popular assembly, however, a private citizen suggested not only rejecting the Government’s proposal, but immediately adopting a more radical merger into three municipalities. The participants of the Landsgemeinde then supported, by a narrow majority, this latter option. In the aftermath of the vote, another private citizen filed a legal complaint that was first rejected by the administrative court of the canton and then taken to the Federal Court. The appellant claimed that the vote had been held in disregard of Article 65(1–3) of the 1988 Constitution of Glarus. According to these provisions, the debate at the Landsgemeinde has to focus exclusively on the subjects put on the agenda, unless amendments proposed by a citizen at the assembly have an intrinsic material link to one of these subjects. The Federal Court, however, dismissed the claim that the proposal of a merger into three municipalities instead of 10 was a new subject without a material link.135 Thus, the ballot had not taken place without sufficient information of the voters. The judges thereby advanced the historical argument that the framers of the cantonal Constitution had expressly underlined that amendments should not be handled in a manner that was too restrictive. Furthermore, it would be the added value of assembly democracy, as represented by the Landsgemeinde, that voters can shape the debate through proposals on the spot and are thus not restricted to a mere decision of yes or no. The amendment in question could not be seen as being a new subject because it pursued the aim of strengthening local government through amalgamation for the same reasons, only in a more radical way. It was determined that voters had been sufficiently informed about the three-municipalities model, as it had already been discussed in the media and by the cantonal Parliament just before the popular assembly.

In Austria, by contrast, comprehensive territorial reform is absent less due to procedural obstacles than to a lack of political will. While a Land parliament is obliged to consult the municipalities concerned, it does not need, in case of a manifest public interest, the consent of their political representatives or even their citizens. After more or less half-hearted efforts on the part of some Länder in the 1960s to reduce the number of municipalities, Styria is a more recent exceptional case of territorial reform. After the Government of this Land had judged attempts until 2010 to promote inter-municipal cooperation and voluntary mergers as largely unsuccessful, it initiated such a reform and thereby reduced the number of municipalities from 542 to 287. Several municipalities challenged this reform before the Constitutional Court.

135

BGE 132 I 291.

308 Local Government

9.G Austria 2014: Reasonableness of Territorial Reform in Styria? Before 2014, Austria’s Constitutional Court had received more than 40 legal complaints from Styrian municipalities. These claimed that amalgamations would run contrary to the equality principle enshrined in Article 7 of the Austrian Constitution, which disallows, according to a broad interpretation of the Court, any legislation not based on reasonable or objective grounds (Sachlichkeitsgebot).136 Another recurrent argument was the incompatibility of the amalgamation law with the European Charter of Local Self-Government. The latter argument was summarily dismissed by the Court because this international treaty only has the rank of an ordinary law and therefore does not represent the (constitutional) standard that another such law has to be measured against.137 The Court likewise refuted the second argument focusing on an alleged lack of reasonable or objective grounds. It stated explicitly that the Austrian Constitution does not guarantee a single municipality a right to its existence but grants the Länder far-reaching leeway to determine the local government structure on their territory. Against this backdrop, the judges only regarded amalgamations as incompatible with Article 7 if they were ‘due to very particular circumstances foreseeably absolutely inappropriate’ to achieve the legitimate aim of improving the structure of municipalities. The disadvantages of territorial reform do not entail unconstitutionality as long as they are outweighed by the advantages. Quite evidently, this view grants the Länder substantial scope to merge municipalities even if their financial, demographic and infrastructural situation is stable and their populations oppose such a step.

It is obvious from all these experiences that amalgamations are nearly always highly contentious. Due to the incisiveness of such measures, valid arguments in favour of them need to surmount high thresholds of justification and usually clash with counterarguments like the historically evolved identity of municipalities or political traditions. In the United States, for example, the tradition of home rule and local competition between municipalities limits territorial reform in a similar way as in Switzerland, with its deeply rooted position of municipalities as strongholds of direct democracy. In other cases, like Quebec, amalgamations were instead opposed because they were looked at through an ethno-cultural lens. When the provincial Government initiated in 2001 a comprehensive territorial reform, a number of predominantly English-speaking cities resisted amalgamations with largely Francophone megacities. In response to the ensuing ‘de-merger’ movement, the Quebec Liberal Party subjected all imposed mergers to referendums. As a result, 32 amalgamated municipalities, most of them with Anglophone majorities, voted in 2004 to ‘de-merge’.138 Moreover, in other countries, both rich and poor ones,139 amalgamations have failed due to the refusal of wealthier suburbs to be united with financially distressed city centres. 136 See M Stelzer, The Constitution of the Republic of Austria: A Contextual Analysis (Oxford, Hart Publishing, 2011) 242ff. 137 VfSlg 19894/2014. 138 See Lazar and Seal, ‘Local Government’ (2005) 32. 139 Regarding Brazil, see, eg, M Arretche, ‘Governance and Finance in Two Brazilian Metropolitan Areas’ in Slack and Chattopadhyay (n 13) and concerning Switzerland, H Geser, ‘Die Gemeinden in der Schweiz’ in U Klöti et al (eds), Handbuch der Schweizer Politik (Zürich, NZZ Verlag, 1999) 426.

Structure of Local Government 309 While merging local governments is thus in any case complicated by a number of factors, it is arguably even more difficult in multilevel systems. In such cases, redrawing the map of local governments may be particularly challenging because metropolitan areas may transcend, like in the cases of Hamburg or Geneva, subnational boundaries.140 Another specific issue is sometimes a sceptical attitude of subnational entities towards amalgamations because they feel threatened, especially regarding metropolitan areas, by consolidated and more powerful large municipalities. As a result, their preference is then, according to the principle of divide et impera, for fragmentation of local government through a multitude of small municipalities. An obvious case of such dynamics is the rivalry between Catalonia and Barcelona, which ended with the disaggregation of the metropolitan Government into 32 municipalities, including the city of Barcelona.141 Not least for these reasons, amalgamations are implemented above all in rural areas. In metropolitan areas, by contrast, fragmentation in small local governments has so far remained the norm.142 In the few cases of amalgamations into one single metropolitan government, this was effected without popular consent, namely in Toronto through the provincial Government and in Cape Town through an independent body, the Municipal Demarcation Board. Interestingly, the mergers were in both cases balanced with the introduction of sub-metropolitan entities in an effort to preserve, as far as possible, local autonomy and accountability. The typical lack of metropolitan government does not imply, however, that these wider areas are denied legal recognition and special treatment of any kind.143 In some cases, they are granted special taxation powers or financial transfers in order to compensate them for services of wider regional or national interest. In others, metropolitan areas are, instead of being merely consolidated under a local government, awarded the status of a subnational entity and thus elevated to the second level of government. In such cases they either comprise subordinated municipalities (eg the Autonomous Community of Madrid) or they are themselves at the same time subnational entities and municipalities (eg the Länder of Berlin and Hamburg in Germany and Vienna in Austria). In the latter case, this double function and fusion of government levels obviously entails considerable asymmetry. 9.4.2 Inter-municipal Cooperation Amalgamation and inter-municipal cooperation are often portrayed as the two opposite ends on the spectrum of designing the structure of local government.144

140 In fact, the metropolitan area of Geneva not only sprawls from the homonymous Swiss canton into the canton of Vaud but also into France. 141 See Velasco Caballero, ‘Kingdom of Spain’ (2009) 305f. 142 For a classification of the institutional systems of metropolitan areas, see E Slack and R Chattopadhyay, ‘Introduction’ in Slack and Chattopadhyay (n 13) 5f. 143 See Slack and Chattopadhyay (n 13) 343ff and 350f. 144 For an often-quoted review of the arguments for both options, see E Ostrom, ‘Metropolitan Reform: Propositions Derived from Two Traditions’ (1972) 53 Social Science Quarterly 474.

310 Local Government Yet, in practice, they also complement each other, with the territorial reforms in Germany of the 1960s and 1970s being a case in point. In spite of the aforementioned examples of more or less comprehensive amalgamation processes, there appears to be, from a global perspective, an even stronger trend towards enhanced collaboration of local governments.145 In Europe, this trend is not only a result of practical necessities of efficient service delivery, but to some degree also influenced by the wider legal context. Article 10(1) ECLSG explicitly underlines the entitlement of local authorities ‘in exercising their powers, to co-operate and, within the framework of the law, to form consortia with other local authorities in order to carry out tasks of common interest’. As comparative evidence from Europe and beyond suggests, inter-municipal cooperation has to be understood as a broad umbrella term that encompasses an enormous variety of more or less institutionalised collaborative mechanisms under both public and private law and sometimes even including private actors. Legal frameworks enabling horizontal cooperation are in place in practically all multilevel systems. Even in the United Kingdom, where such collaboration is less common, due to comparatively large local governments, there is the legal possibility to outsource functions to another local authority or to establish joint committees.146 Very clearly, however, such mechanisms are used in particular in countries with many small and predominantly rural municipalities. In Austria, it is often regarded as the alternative to merging local governments and is carried out in manifold forms ranging from de facto cooperation or contracts, companies and associations under private law to formal municipality associations under public law.147 Interestingly, such associations may not only be established by municipalities themselves upon the consent of the supervisory authority (Article 116(a)(1)). Provision is also made for forced cooperation, as the national and the Länder governments may, for the execution of local tasks, create municipal associations after prior consultation with, but not the approval of, the municipalities concerned (Article 116(a)(2)). Also in both Germany and Switzerland, special-purpose associations (Zweckverbände), which are established under public law for specific tasks like waste disposal or public transport, are classical instruments of collaboration and are particularly widespread. There is, however, especially in the Swiss case, a notable tendency of municipalities resorting to mechanisms of private law because they leave them more room for flexibility.148 While both the extent and forms of inter-municipal cooperation thus greatly vary, many of its mechanisms share a common problem, that is, the issue of whether newly created supra-municipal bodies are sufficiently accountable and democratically legitimated. In recognition of this problem, the Austrian Constitution prescribes that ‘organs of the municipal associations which are to undertake matters pertaining to the municipality’s own sphere of competence are to be formed according to democratic principles’ (Article 116a(3)). Regarding the composition of such bodies, the

145

See Steytler (n 3) 432f. C Himsworth, ‘Local Government in the United Kingdom’ in Moreno, Local Government (2012) 673. 147 See A Gamper, ‘Local Government in Austria’ in Moreno (n 24) 34f. 148 See Ladner (n 102) 338. 146 See

Structure of Local Government 311 Italian Autonomous Province of Trento in Italy has taken with its valley communities (Comunità di Valle), which are based on a law of 2006, a middle approach. These particular institutions are supposed to tackle the problem of fragmentation into many small municipalities and consist, unlike similar bodies in other parts of the country, at least in part, that is, three-fifths, of directly elected representatives with the remaining members being nominated by the municipalities.149 In a number of other cases, however, all members of inter-municipal bodies are nominated, which raises the question whether they are still to a sufficient degree controlled by, and accountable to, the directly legitimated local institutions or actually undermine local autonomy. It seems to make sense that this degree should correspond with the number of tasks transferred to these bodies. This specific issue was at stake in the legal dispute regarding the Ämter of the German Land of Schleswig-Holstein. 9.H Germany 1979: Ämter of Schleswig-Holstein to be Composed of Elected Representatives? A subnational law of 1977 reorganised the traditional Ämter of the Land of SchleswigHolstein, which are entities under public law at the level between the municipalities and the counties (Kreise). With few exceptions, municipalities formed part of one of these entities, which should take care of certain tasks of public administration instead of their members each doing this alone. The main decision-making body of these Ämter was a committee (Amtsausschuss), composed of the mayors of the municipalities involved and of additional members nominated by the municipal councils. This committee then elected, usually from among its members, someone to represent the Amt and to be responsible for the implementation of its decisions (Amtsvorsteher). The Federal Constitutional Court had to decide whether the appointment and composition of these institutions was in compliance with Article 2(1–2) of the Constitution of Schleswig-Holstein. This provision emphasised the sovereignty of the people, which acts through its elected representatives at the levels of the Land, the municipalities and associations of municipalities. According to the Court’s judgment, the Ämter do not qualify as such associations of municipalities and are thus not subject to the requirement of direct electoral legitimation.150 The judges acknowledged that ‘associations of municipalities’ is a rather vague collective term that is used differently in legislation. For them, the decisive criterion is whether such an association was established predominantly for the purpose of fulfilling tasks of local self-government. In any case, this legal category would include the counties and exclude the above-mentioned Zweckverbände, which were created for specific purposes. The Ämter are rather similar to these Zweckverbände than to the municipalities and counties and do not, by its nature, predominantly fulfil tasks related to self-government. Even though the Court explicitly acknowledged that the Ämter would no longer amount to mere secretariats, they are, despite an extended range of tasks, still largely concerned with administrative support and managing single issues delegated by the municipalities.

149 See E Alber and A Valdesalici, ‘Framing Subnational “Institutional Innovation” and “Participatory Democracy” in Italy: Some Findings on Current Structures, Procedures and Dynamics’ in F Palermo and E Alber (eds), Federalism as Decision-Making: Changes in Structures, Procedures and Policies (Leiden, Brill–Nijhoff, 2014) 464f. 150 52 BVerfGE 95 (Ämter of Schleswig-Holstein).

312 Local Government 9.4.3 Umbrella Entities A third structural element of local government are, in several cases, umbrella entities that, as a second tier of local government, function within the same territorial jurisdiction as municipalities. Similar to institutions of inter-municipal cooperation, they are particularly useful and often indeed indispensable for small rural municipalities. It is therefore hardly surprising that the scope of action of these umbrella entities is frequently confined to non-urban areas. This holds true for the German counties, as well as for district panchayats in India, even if their status and significance is very different. Apart from the 116 ‘county-free cities’ (kreisfreie Städte), that are, according to most Land constitutions, cities with more than 100,000 inhabitants, all municipalities in Germany are constituent units of the 313 counties. The democratically elected administrators of these counties may gain significant power, even more so when they create, for a specific purpose, joint administrative units.151 A similar pattern of important coordination by umbrella entities outside urban agglomerations is characteristic of South Africa. There are both metropolitan municipalities and local municipalities, on the one hand, and second-tier district municipalities, on the other (Section 155(1) of the South African Constitution). The Constitution, however, does not say much about the distribution of responsibilities between local and district municipalities. Section 155(4) merely states that, in fulfilling this task, national legislation must ‘take into account the need to provide municipal services in an equitable and sustainable manner’. Moreover, provision is made for asymmetries because the power of local municipalities belonging to the same district municipalities must not be the same (Section 155(3c)). As to the purpose of the latter as umbrella entities within the overall structure of local government, the meaning of the constitutional text was clarified by seminal rulings of the Constitutional Court. 9.I South Africa 1996–97: Objectives of the District Municipalities In a decision of 1996,152 the Constitutional Court had found that the mere referral in the Constitution to the municipalities as the only type of local government would fail to comply with Constitutional Principle XXIV of the Interim Constitution of 1993. This principle, binding the drafters of the 1996 Constitution, stipulated that the ‘comprehensive powers, functions and other features of local government’ may be left to ordinary legislation, but required the ‘framework for local government powers, functions and structures’ to be treated in the Constitution itself. The Court had held that such a framework for the structure ‘should convey an overall structural design or scheme for local government’, including regulations about appointments and procedures for decision-making and thus much more than a mere referral to municipalities (para 301). In response, the Constitution was amended to include the above-mentioned differentiated structure with three categories of local government. When the province of KwaZulu-Natal challenged the certification of

151 152

See Burgi, ‘Federal Republic’ (2009) 141. Certification of the Constitution of the Republic of South Africa 1996 (4) SA 744 (CC).

Structure of Local Government 313

the amended text153 before the Constitutional Court, its counsel claimed that it would still be impossible to discern from the text what types of local governments the provinces may establish within the three categories of Section 155(1) of the Constitution and how the different types relate to each other, above all in terms of exercising their powers. The Court held154 that such a view ‘is to deduce too great a specificity from a phrase of such general and imprecise import as a “framework for local government … structures”’ (para 82). Moreover, it found the performance of ‘coordinating functions’ to be the overall objective of the district municipalities (para 77). In the aftermath of this judgment, a White Paper in 1998 further defined these coordinating functions so that the district municipalities were finally accorded a two-fold purpose.155 First, they have to assist local municipalities and make them capable of providing services. Second, they are responsible for an equitable redistribution of resources within the district according to needs, which is, from a comparative perspective, a not so common function of umbrella entities. But it is obviously due, in the specific South African case, to considerable territorial inequalities. In 2000, the Municipal Structures Amendment Act adjusted the role of district municipalities by making them, above all for rural areas, regular direct providers of services in areas like electricity, wastewater or municipal health services.

Rural India also features umbrella entities above the village panchayats, namely equally democratically elected panchayats at the levels of the district and, in more populous states, even between the village and district level. In contrast to Germany and South Africa, however, these are often substantially sidelined. At the district level, the institution of the district collector and his/her administrative units of senior officers of the Indian Administrative Service, which are directly responsible to the state government, do not merely coordinate. Often, they also closely supervise and, like rural development agencies established by both state and union governments, directly interfere with the competences of the panchayats. The latter have been, for these reasons, marginalised in many districts. The Spanish provinces differ from all three above-mentioned institutions insofar as they are not confined to rural areas. The provinces are the only umbrella entities in the sense of constitutionally entrenched institutions, even though the Constitution enables the creation of certain institutions between the municipalities and provinces, namely the comarcas. For example, on the basis of Articles 141(3) and 152(3), which authorise the statutes to form groupings of bordering municipalities with full legal personality that go beyond the provinces, both Catalonia and Aragon established the traditional comarcas. At the same level, there are also mancomunidades, that is, associations of local authorities constituting a distinct legal entity which has powers delegated by the participating municipalities. As opposed to these two institutions, the legal position of the provinces is stronger because their autonomy is

153

On certification, see chs 5.1.2 and 8.2.3.2. Certification of the Amended Text of the Constitution of the Republic of South Africa 1997 (2) SA 97. 155 See N Steytler, ‘Local Government in South Africa: Entrenching Decentralised Government’ in Steytler (n 26) 192f. 154

314 Local Government not only guaranteed by the Constitution but also clarified and reinforced by settled case law. The Constitutional Court identified as the minimum core of their administrative autonomy the classical function of cooperation with, and assistance to, municipalities, not least financially through the use of provincial spending power.156 Even though the provinces, in the wake of the 1978 Constitution, clearly lost competences to the autonomous communities, they may rely on this constitutional core, which is further specified by an explicit list of powers in the 1985 Law Regulating the Basis of the Local System.157 Unsurprisingly, this list has not been extended by the autonomous communities, as many of them, like Catalonia, perceive the provinces as competing institutions and remnants of centralised rule. 9.5 LOCAL GOVERNMENT AS A GENUINE THIRD LEVEL?

There is no doubt that federal systems are traditionally premised in both theory and practice on only two levels of government. A notable exception is, in this regard, the decentralisation theory of the Viennese School of Legal Positivism, which regarded both subnational entities and local governments as units on a sliding scale of decentralisation without any qualitative difference between them. While this view eschews dualism in favour of a monistic design, it might, of course, as well be rejected in favour of a three-level constellation. For instance, it has recently been argued that, from a normative perspective, a conception of local government as a genuinely equal third level of government would be, for all intents and purposes, compatible with a liberal democratic understanding of federalism.158 Yet, comparative evidence from current constitutional practice is, in this regard, at best mixed. The finding that municipalities in South Africa are caught in an ‘uncomfortable ménage à trois’159 appears to be equally applicable to a number of other cases in which local governments are the plaything of the other two still superior levels of government with their often conflicting interests. To be sure, the local level has in recent times been constitutionally reinforced in certain countries like Brazil and South Africa or also India, albeit through the still crucial role of the states to a lesser extent. But in all three cases, the strategy of the national governments to empower and reach out to municipalities, especially cities, was ultimately aimed, at least in part, at bypassing and indeed marginalising subnational entities.160 As to intergovernmental dynamics, these cases are in this regard strikingly similar to the alliance between the US federal Government and cities under Lyndon B Johnson’s Great Society programmes.161 Of course, such dynamics did not contribute to the emergence of a three-level system with partners on an equal footing. 156

STC 32/1981. See E Fossas and E Velasco, ‘Local Government in Spain’ in Steytler (n 26) 113f. 158 See D Halberstam, ‘Federalism: A Critical Guide’ (2011) 251 University of Michigan Public Law Working Paper 1, 39. 159 Steytler, ‘National, Provincial’ (2007) 229. 160 See JT Dickovick, ‘Municipalization as Central Government Strategy: Central–Regional–Local Politics in Peru, Brazil, and South Africa’ (2007) 37 Publius 1. 161 See ch 11.1. 157

Local Government as a Genuine Third Level? 315 From the perspective of subnational entities, on the other hand, the empowerment of local governments is often perceived, rightly or wrongly, as an existential threat. Even if they may at times form a strong alliance with local authorities vis-a-vis the national government regarding some issues of common interest,162 they mostly see subnational and local autonomy as being subject to a zero-sum game. Their ultimate fear is, from this viewpoint, the scenario of ‘hourglass federalism’,163 in which they are squeezed in the middle between the national and local government. For obvious reasons, this fear is particularly intense regarding large urban municipalities. The fact that certain Canadian cities make up roughly one third or, in the case of Winnipeg, even more than half of their province’s entire population, clearly gives them both economically and politically an immensely strong position towards the subnational government, even without constitutional reinforcement. Even though Canada provides particularly illustrative examples of big cities possessing considerable relative power vis-a-vis the respective provincial government, this is due to urbanisation, of course, a much broader phenomenon. In general, however, even the undisputed trend towards constitutional recognition of local self-government in at least some form does not give rise to a genuinely threetiered system. Many local governments suffer from overregulation and see their actual powers diminished by an overload with delegated tasks, tight supervision, as well as limited financial resources. In sum, local governments in general and even many cities may therefore be characterised as ‘policytakers, not policymakers’.164 It is true that the constitutions of some countries like Spain, Brazil and Italy explicitly mention local governments in the same breath with subnational entities as the autonomous units that constitute the multilevel system.165 Yet, the remainder of the constitution then reveals in all these cases that local autonomy is by no means equal to subnational autonomy. In these three and most other cases, local governments clearly remain in an inferior position so that they play at best a junior role in a system of ‘two and a half partners’.

162 In view of this possibility, it remains questionable whether local autonomy is, under federalism’s traditional paradigm of national–subnational dualism, generally weaker than in unitary states. See A Gamper, ‘Die Stellung der Gemeinden im Vergleich europäischer Bundesstaaten’ in Europäisches Zentrum für Föderalismus-Forschung Tübingen (ed), Jahrbuch des Föderalismus 2006 (Baden-Baden, Nomos, 2006). It is undeniable that certain unitary states, particularly Europe’s Nordic countries, feature very strong local governments, whereas others grant autonomy to neither the subnational nor the local level. 163 This notion was coined by Thomas Courchene after the Canadian Government had, through budget cuts in the mid-1990s, withdrawn federal funds to provinces and instead developed direct financial relations with cities. See T Courchene, ‘Hourglass Federalism’ (2004) Policy Options 12. The metaphor has since become used more widely. 164 Leuprecht and Lazar, ‘From Multilevel’ (2007) 2. 165 Such autonomous local governments are, in Spain, the municipalities and provinces (Art 137); in Brazil, the municipalities only (Art 18); and in Italy, the municipalities, provinces and metropolitan cities (Art 114).

316

Part III

Powers and Policies: Between Autonomy and Homogeneity Part III of this book deals with five areas that are highly relevant today for most governments, national and subnational ones alike: fundamental rights, social welfare and healthcare, environmental protection, immigration and migrant integration, as well as external relations. Some of these subject matters, like fundamental rights and external relations, have obviously been consistently important since the heyday of the liberal minimal state. Moreover, these two areas have a specific character, as autonomy regarding fundamental rights concerns subnational (quasi)constitutions rather than ordinary legislation, while external relations are very much about procedures to effectively use (domestic) subnational competences in a more and more interdependent international context. Other subject matters covered in the following chapters, ie social welfare and healthcare, environmental protection and immigration and migrant integration, became highly relevant through rather recent developments. Why we regard each of these five issues as very topical in general, and for federal systems in particular, is explained in the introductory part of each respective section. The overall aim of Part III is to examine, from a comparative perspective, how the distribution of powers responds in different federal countries to the common challenge of providing a legal framework for the co-management of the five abovementioned subject matters by different levels of government. The following part of the book ties in with chapter 5, which looked at the autonomy of subnational entities from a more structural and theoretical perspective. Part III now scrutinises the scope and content of this autonomy with regard to certain specific areas, each of them exemplifying how key aspects of federalism operate. Beyond that, the aim of this part is to respond to a second, albeit interrelated, question about the roles actually played by national and subnational governments1 in policy-making on the basis of the distribution of powers. The extent to which these roles are effectively determined by this legal framework or deviate from it will be of crucial importance. After all, the formal distribution of powers is in many federal systems merely indicative so that any analysis limited to legalistic formalism would be insufficient.

1 Although the role of the supranational and local levels is occasionally taken into account, the focus is on the interplay between the national and subnational levels of government. The principal reasons for this are Part III’s limited number of pages and the scant importance of the supranational level outside Europe. The European Union is a transversal issue which is, in one way or another, addressed in most chapters of the book. Local government is discussed separately in ch 9.

318 Powers and Policies Rather than constituting clear-cut powers belonging to only one government level, all five areas explored in Part III are, like most subject matters today, broad multifaceted areas that involve multiple levels of government. What is really needed, therefore, is an analysis of intergovernmental relations in these areas and constitutional adjudication. Both are, beyond the mere text of the constitution, decisive in defining the roles of national and subnational governments. Indeed, the analysis of broad subject matters regarding both their legal foundation in the distribution of powers and actual policy-making appears to emerge, in the context of an era of pluralism, as the most crucial challenge for research on federalism. This requires the academic community to support and guide decision-making, anticipating trends, and suggesting tools that can provide meaningful and effective responses to the challenges of pluralism. These tasks for federal studies can best be met if the research agenda becomes more technical, focusing on two main areas: procedures and policy fields.2

Against this backdrop, the comparative analysis in Part III focuses, as reflected in its title, on both powers and policies. Concerning the first dimension, the distribution of legislative and executive powers is scrutinised which typically follows from the national constitution. In some cases, however, like Belgium and Spain, competences are at least partially allocated by special legislation so that these sources are equally taken into account. The second dimension concentrates on how subnational policies are formulated and implemented, in practice, upon this legal basis. Quite obviously, both subnational powers and, as a consequence, subnational policies can only exist within a certain scope of autonomy. Yet, autonomy is delimited by legal and political instruments aimed at ensuring some country-wide homogeneity, a necessary minimum of it or more. This inherent tension is epitomised by the subtitle of this part. As to the actual use of autonomy within limits of homogeneity, particular emphasis will be placed on the potential of self-government not only for single subnational entities, but also on potentially beneficial spill-over effects of policy innovation to other jurisdictions. This presumed advantage of so-called competitive, experimental or laboratory federalism was emphasised with reference to the US example by James Bryce as early as in the late nineteenth century.3 Today, however, this rationale is even more associated with an often-quoted statement from a dissenting opinion by US Supreme Court Justice Louis Brandeis in 1932: It is one of the happy accidents of the federal system that a single courageous state may, if its citizens choose, serve as a laboratory, and try novel social and economic experiments without risk to the rest of the country.4

2 F Palermo, ‘Regulating Pluralism: Federalism as Decision-Making and New Challenges for Federal Studies’ in F Palermo and E Alber (eds), Federalism as Decision-Making: Changes in Structures, Procedures and Policies (Leiden, Brill–Nijhoff, 2015) 508f. 3 ‘Federalism enables a people to try experiments in legislation and administration which could not be safely tried in a large centralized country. A comparatively small commonwealth like an American State easily makes and unmakes its laws; mistakes are not serious, for they are soon corrected; other States profit by the experience of a law or a method which has worked well or ill in the State that has tried it.’ J Bryce, The American Commonwealth, 3rd edn (London Macmillan, 1893) 353. 4 New State Ice Co v Liebmann 285 US 262 (1932).

Powers and Policies 319 Brandeis’s laboratory argument boils down to the claim that multiple experiments at a smaller scale are more likely to produce good policies than one single nationwide effort. If such experiments fail, damage would be limited to a single jurisdiction. But if they prove successful, they could be emulated by other jurisdictions. Interestingly, a contextual reading of the Brandeis quote has revealed that, as a supporter of the then upcoming Scientific Management Movement, he actually emphasised the benefits of experiment-based public policy in general and not only in federal states.5 However, a federalism-friendly (mis)interpretation of his dissenting opinion has, over the years, become commonplace and has thus come to sustain the argument of federal systems as breeding grounds for innovation. Whether federal systems are really able to fulfil this role seems to depend on certain practical preconditions. First, the rationale of experimental federalism presupposes substantial legislative autonomy because, otherwise, there is just no room for experimenting with policies that are genuinely a subnational entity’s own to be then emulated by others. Administrative federalism6 in a very pure and extreme form, with subnational entities merely putting into practice nationally devised policies, seems therefore unlikely to make federal systems laboratories of innovation. Second, this rationale assumes that subnational entities are willing to bear the costs of inventing something new that others might then benefit from. Third, it presupposes that economic, social and political contexts are sufficiently similar to enable a transfer of innovation from one jurisdiction to another.7 Moreover, one should be cautious not to overestimate experimental federalism as the driving force behind a linear development towards more progressive societies. Just as subnational entities are not always the proponents of democratic reform but sometimes its opponents,8 they have sometimes also inhibited more progressive policies in other areas. This is because federalism is not ideological; instead, it is neutral with regard to ideology and just enables greater congruence of policies with subnational preferences. As a consequence, it also provides opportunities for conservative forces in subnational entities to resist the progressive agenda of a national government. From a historical perspective, this arguably applied to the United States, Canada and Australia in the late nineteenth and early twentieth centuries, with national governments attempting to achieve quick countrywide adaption in response to industrialisation-related social and economic change. At that time, ‘it seemed that the judiciary, backed by conservative business interests, was intent on using the division of powers to create a governmental no-go area and thereby enforce a regime of laissez-faire.’9 The backside of the theory that subnational entities are laboratories facilitating innovation is also reflected in more recent cases. A glaring example in the

5 See GA Tarr, ‘Laboratories of Democracy? Brandeis, Federalism, and Scientific Management’ (2001) 31 Publius 37. 6 See ch 5.3. 7 See TO Hueglin and A Fenna, Comparative Federalism: A Systematic Inquiry (Peterborough, Broadview Press, 2006) 347. 8 On the example of women’s suffrage in Switzerland, see ch 4.3.2. 9 Hueglin and Fenna, Comparative Federalism (2006) 342.

320 Powers and Policies United States was that federalism actually enabled the so-called Jim Crow Laws, ie, state legislation upholding racial segregation in the South adopted in the wake of the Civil War (1861–65) and in force until 1965. In order to understand how the interplay between national and subnational governments has responded in different countries to similar challenges,10 we will draw, in each of the following comparative analyses, on experiences and insights from five cases. In general, the selection of the countries follows the non-formalistic approach outlined in the introduction.11 This means that less attention is given to cases that are only nominally considered as federal or consider themselves as such and meet one or both of the following criteria. First, the practical effects of an at least formally federal constitution are in reality to a greater or lesser extent offset by a context of authoritarianism or continuous violent conflict. In such cases, the distribution of powers on paper counts little and seems thus irrelevant to be analysed here in great detail. Second, the actual interplay between government levels has not developed in the specific case solutions that are among the most relevant for a comparative functional analysis.

10 See C Saunders, ‘The Use and Misuse of Comparative Constitutional Law’ (2006) 13 Indiana Journal of Global Legal Studies 37, 37. 11 See ch 0.3.1.

10 Fundamental Rights

F

UNDAMENTAL RIGHTS ARE commonly understood as human rights transformed into positive law.1 As devices to limit governmental power, they are typical elements of national constitutions which normally contain, except for a few cases such as Australia, sets of civil and political rights.2 Also the entrenchment of social, economic and cultural rights, which is overall more common in civil law countries, plays a role in some countries covered in the following comparative analysis. In addition to these national guarantees, fundamental rights are also typical elements of many subnational constitutions.3 The autonomy of subnational entities to draft constitutions and to entrench within them their own rights guarantees is always limited, to a certain degree, by the national level.4 These limits of the subnational constitutional space are the first dimension explored in sub-section (1) of each country case study within this chapter, and they boil down to the following key questions. To what extent is the discretion of subnational entities to enshrine their own fundamental rights pre-empted by the national constitution? Legal instruments to accomplish such pre-emption may include detailed prescriptions regarding the contents of subnational constitutions, homogeneity clauses in order to stimulate conformity in a less specific way, as well as supremacy clauses that determine the precedence of national law over subnational law. A second question is then whether the autonomy of subnational entities is further limited by requirements to have their original constitution and/or subsequent amendments approved by the national legislature, executive or judiciary.5 A third question is to what degree the judicial interpretation of subnational rights by a national supreme court or constitutional court constitutes a restraining factor. While all these elements determine the limits of subnational autonomy to entrench fundamental rights, the second dimension of the comparative analysis deals with the utilisation of this constitutional space (sub-section (2)). To what extent do

1 For an overview of the formal, procedural and substantial concepts of fundamental rights, see R Alexy, ‘Rights and Liberties as Concepts’ in M Rosenfeld and A Sajó (eds), The Oxford Handbook of Comparative Constitutional Law (Oxford, Oxford University Press, 2012) 288ff. 2 See J Kincaid, ‘Comparative Observations’ in J Kincaid and GA Tarr (eds), Constitutional Origins, Structure and Change in Federal Countries (Montreal, McGill–Queen’s University Press, 2005) 420f. The Australian Constitution limits itself to the protection of only a few procedural rights such as the right to a jury trial under Sec 80. 3 For a list of common subjects of subnational constitutions, see GA Tarr, ‘Explaining Subnational Constitutional Space’ (2011) 115 Penn State Law Review 1133, 1134. 4 See ch 5.1.2. 5 See ch 8.2.2.

322 Fundamental Rights subnational entities exploit their leeway? Which kinds of rights do they focus on? Are there significant variations among subnational entities concerning the entrenchment of rights and patterns of emulation in the horizontal and/or vertical dimension? 10.1 UNITED STATES OF AMERICA

(1) The rather brief original US Constitution of 1787 is renowned for its lack of a fundamental rights catalogue. Even though the Bill of Rights was negotiated at the Philadelphia Convention, it was formally only integrated four years later with the first 10 amendments. These amendments were predated and significantly influenced by the bills of rights of the 13 founding states, most notably the Virginia Declaration of Rights of 1776.6 For more than a century, the provisions of the US Bill of Rights were held to be binding only against the national Government.7 As a result of the expansion of federal power after the Civil War, however, the US Supreme Court issued, particularly in the twentieth century, a series of decisions8 that made most provisions of this federal rights catalogue applicable to the states as well.9 For this judicial process of so-called incorporation, the judges relied on the Due Process Clause of the Fourteenth Amendment adopted in 1868. As a consequence of the Supremacy Clause (Article VI, clause 2 of the Constitution), these rights take precedence over conflicting rights in state constitutions and therefore establish an obligatory minimum of protection: In the United States the US Bill of Rights is often described as a floor rather than a ceiling. It establishes a standard below which constituent units cannot go, but it does not otherwise limit state initiatives in expanding rights—they can build on that floor.10

A potential restriction of subnational autonomy concerning the entrenchment of rights is relevant for all but the 13 founding states. According to Article IV, clauses 3 and 4 of the US Constitution, Congress has the power to set conditions on the admission of new states to the Union. To states that do not meet those requirements, admission can be refused either by Congress or the President. In practice, the conditions established have commonly been quite general like compliance with the Declaration of Independence or the separation of church and state. Their restraining impact on the freedom of states to provide constitutional rights at their own discretion has therefore been limited. As far as the judicial interpretation of these rights is concerned, the autonomy of the states is greatly facilitated by the dual court system, which restricts the US Supreme Court to the national Constitution. The ultimate interpretation of 6

See DS Lutz, ‘The State Constitutional Pedigree of the US Bill of Rights’ (1992) 22 Publius 19, 19. Barron v Baltimore 32 US (7 Pet) 243 (1833). 8 The first case was Chicago, Burlington & Quincy Railroad Co v Chicago (City of) 166 US 226 (1897), followed by Gitlow v New York 268 US 652 (1925) and five more cases in the 1940s. 9 The most recent incorporation concerned the 2nd Amendment and was effected by McDonald v Chicago (City of) 561 US 742 (2010). 10 GA Tarr, ‘Subnational Constitutions and Minority Rights: A Perspective on Canadian Provincial Constitutionalism’ (2008) 2 Revue québécoise de droit constitutionnel 174, 193. 7

United States of America 323 subnational constitutions is the sole responsibility of the respective state supreme courts, which are composed of state judges in terms of nomination and payment.11 (2) The states have made ample use of their autonomy to entrench rights in their constitutions. These rights can be categorised according to their relationship with the rights in the US Constitution in the following way.12 Some of them, like many criminal procedure protections, have identical wording, while others, such as the Cruel and/or Unusual Punishment Clauses,13 have only slightly different wording than their federal counterparts. The formulations of still other rights, like the much more detailed religion clauses, differ substantially from the corresponding right in the US Constitution or do not even have federal counterparts. The latter rights obviously exceed most clearly the minimum standard set by the above-mentioned incorporation of large parts of the US Bill of Rights. Rights found solely at state level include, for instance, guarantees of access to legal remedy, explicit privacy rights and positive rights to welfare and fair or quality education. But even regarding provisions with identical wording, the states have certain discretion even if through their judiciaries rather than their legislatures. It is today widely recognised that state supreme courts can use their independence within the dual court system to interpret such state rights as providing more protection than their federal counterparts by reference to diverse contexts such as different generating histories. This wide recognition is the result of what is today well known as New Judicial Federalism.14 The origins of this phenomenon can be traced back to 1969 when the retirement of Chief Justice Earl Warren triggered a turnaround of the US Supreme Court’s interpretation of fundamental rights towards a more restrictive approach.15 Prominent examples are its judgments to allow the states to reintroduce the death penalty16 and to permit the admission of improperly seized evidence in criminal proceedings if police officers had acted in good faith.17 In response to this shift, a number of state supreme courts started to ‘circumvent’ more restrictive judgments of the US Supreme Court by relying solely on state constitutional rights and interpreting them in an extensive manner. Therefore, the judicial activism of the Warren Court, which is often pictured as detrimental to states’ rights, actually set the stage, albeit indirectly, for the New Judicial Federalism. In other words, the latter started as a tactical backlash but has evolved over time into a lasting and defining characteristic of US federalism.18 11

See ch 5.4. See RF Williams, The Law of American State Constitutions (Oxford, Oxford University Press, 2009) 113ff. 13 For instance, Art I, cl 6 of California’s Constitution has a Cruel or Unusual Punishment Clause, whereas the 8th Amendment to the US Constitution speaks of cruel and unusual punishment. 14 See GA Tarr, ‘The Past and Future of the New Judicial Federalism’ (1994) 24 Publius 63. 15 The landmark decision on the right to abortion, Roe v Wade 410 US 113 (1973), is, of course, a notable exception. 16 Gregg v Georgia 428 US 153 (1976). 17 United States v Leon 468 US 897 (1984). 18 See Tarr, ‘The Past’ (1994) 72f. 12

324 Fundamental Rights In probably the most significant early case, the Supreme Court of California interpreted the Cruel or Unusual Punishment Clause of the state Constitution (Article I, clause 6) as outlawing the death penalty.19 More recent but equally prominent examples of extensive interpretation by state supreme courts concerned the recognition of a right to same-sex civil union,20 as well as same-sex marriage.21 Already in the early 1970s, state supreme court rulings in the wake of the New Judicial Federalism had gradually increased acceptance of the fact that state rights could be interpreted independently. Then-US Supreme Court Justice William J Brennan argued the case in a seminal article in 1977 for such judicial ‘double protection’ of rights: State constitutions, too, are a font of individual liberties, their protections often extending beyond those required by the Supreme Court’s interpretation of federal law. The legal revolution which has brought federal law to the fore must not be allowed to inhibit the independent protective force of state law—for without it, the full realization of our liberties cannot be guaranteed.22

Three years later, a judgment of the US Supreme Court itself followed this view. 10.A United States 1980: A More Extensive Right to Free Speech in California In a Californian shopping mall, a group of high school students collected signatures for a petition against the United Nations General Assembly Resolution 3379, which had defined Zionism as a form of racism. The owners of the shopping mall disallowed such activities and referred to the right to property under the Fifth Amendment to the federal Constitution. In an earlier ruling, the US Supreme Court had already held that the right to free speech under the First Amendment could not prevent the owner of a shopping centre from denying others free speech on his private property.23 On the other hand, whereas the federal First Amendment contains only a negative command to Congress not to abridge free speech, the Constitution of California guarantees an affirmative right of free speech, which had been interpreted extensively by the state’s supreme court. The US Supreme Court held that the obligation even of a private person to allow the collection of signatures, which had been derived from the more far-reaching Californian provision, could not be qualified as an unconstitutional infringement of property rights, as long as the value or use of the shopping centre was not unreasonably impaired.24 In this way, the ruling recognised a state right as an independent source of protection that can be interpreted by a state supreme court as granting more extensive protection than the parallel federal right.

19 People v Anderson 493 P 2d 880, 6 Cal 3d 628 (Cal 1972). Interesting explanations of the reasons for the early focus of the New Judicial Federalism on criminal proceedings can be found in J Friesen, ‘Recovering Damages for State Bills of Rights Claims’ (1985) 63 Texas Law Review 1269; and SS Abrahamson, ‘Criminal Law and State Constitutions: The Emergence of State Constitutional Law’ (1985) 63 Texas Law Review 1141. 20 Baker v Vermont 744 A 2d 864 (Vt 1999). 21 Goodridge v Department of Public Health 798 NE 2d 941 (Mass 2003). 22 See WJ Brennan Jr , ‘State Constitutions and the Protection of Individual Rights’ (1977) 90 Harvard Law Review 489, 491. 23 Lloyd Corp v Tanner 407 US 551 (1972). 24 Pruneyard Shopping Center v Robins 447 US 74 (1980).

United States of America 325 Three years after this decision, the US Supreme Court further contributed to the independence of state supreme courts by clarifying and limiting in Michigan v Long its own authority to review their rulings.25 In the frequent case of state court judgments, which mix federal and state doctrine, these would be considered final if they rely on ‘independent and adequate state grounds’ and contain ‘a plain statement … that the federal cases are being used only for the purpose of guidance, and do not themselves compel the result that the court has reached’.26 This ruling clarified the jurisdictions of supreme courts but still left room for federal–state judicial dialogue through the interpretation of the respective constitutions at each level. Recent examples include several state supreme court decisions that banned criminal punishment of homosexual conduct on the basis of Due Process Clauses. Although not binding for the US Supreme Court, these decisions provided points of reference for the latter to explicitly overrule in 2003 an earlier judgment in which it had upheld the constitutionality of a Georgia sodomy law criminalising oral and anal sex in private between consenting adults.27 According to the new ruling, the liberty protected by substantive due process had until then been interpreted too narrowly and would in fact comprise consensual sexual conduct.28 Judicial dialogue is also noticeable in the other direction. For instance, the Supreme Court of Delaware repeatedly invoked the case law of the US Supreme Court to interpret a press freedom clause extensively as encompassing freedom of speech even in spite of the complete absence of a textual basis.29 Beyond constitutional interpretation, dialogue and mutual influence in the vertical dimension has always been present in processes of constitution-making. From the beginning of US federalism, the states had spearheaded innovation regarding fundamental rights. Examples include the above-mentioned influence of state rights on the federal Bill of Rights of 1791, the federal Equal Protection Clause30 and the removal of suffrage restrictions concerning African-Americans and women, which had been pioneered by north-eastern and western states, respectively.31 In the horizontal dimension, the above-mentioned extensive interpretation of rights by trail-blazing state supreme courts regarding the death penalty, criminal defendants and homosexual couples has usually been emulated by several of their counterparts in other states. With regard to all these issues, however, progressive judgments have also sparked a conservative backlash instead of imitation. For instance, the prohibition of capital punishment by the Supreme Court of California in 1972 was met with such fierce domestic resistance that it was already overturned

25

Michigan v Long 463 US 1032 (1983). ibid 1040f. 27 Bowers v Hardwick 478 US 186 (1986). 28 Lawrence v Texas 539 US 558 (2003). See JA Gardner, ‘In Search of Subnational Constitutionalism’ (2007) 16 Buffalo Legal Studies Research Paper Series 1, 29. 29 See JA Gardner, Interpreting State Constitutions: A Jurisprudence of Function in a Federal System (Chicago, University of Chicago Press, 2005) 8. 30 See RF Williams, ‘Equality Guarantees in State Constitutional Law’ (1985) 63 Texas Law Review 1195. 31 See A Keyssar, The Right to Vote: The Contested History of Democracy in the United States (New York, Basic Books, 2000). 26

326 Fundamental Rights by a state constitutional amendment the same year. Such constitutional responses to reverse unpopular rulings, whether in the same or another state, have been greatly facilitated by generally very flexible amendment procedures. These make fundamental rights at the state level ‘more akin to statutory rights than to constitutional constraints’.32 Similar to capital punishment, a backlash against progressive judgments occurred when, more recently, several rulings on the lawfulness of same-sex marriage prompted a number of states to restrict marriage in their constitutions explicitly to male–female couples so as to pre-empt similar decisions by their own state supreme courts.33 However, this practice was affected by a landmark judgment of the US Supreme Court’s in 2015. 10.B United States 2015: Legalisation of Same-Sex Marriage throughout the United States After the US Supreme Court had struck down in 2013 parts of the federal Defense of Marriage Act of 1996 in United States v Windsor,34 state bans on same-sex marriage were the subject of several trials in district and circuit courts. Whereas the courts of appeal of four circuits found such prohibitions to be unconstitutional, that for the Sixth Circuit considered itself bound by the US Supreme Court judgment in Baker v Nelson,35 which had dismissed a same-sex couple’s claim to marriage for not presenting a substantial federal question. Consequently, it reversed the decisions of the district courts and upheld Ohio’s ban on such marriages. This created a circuit split, which then led the US Supreme Court to review the matter. The majority opinion found that the Fourteenth Amendment requires that each state both license a marriage between two people of the same sex and recognise such a marriage lawfully licensed by another state.36 Before presenting its reasoning, the Court stated that marriage is not static, but a matter of continuity and change, which is illustrated, for instance, by the elimination of the law of coverture. The Due Process Clause of the Fourteenth Amendment would justify the Court’s decision in several respects. This provision’s guarantee of liberty would comprise, with its concept of individual autonomy, a personal choice regarding marriage, which led, for instance, to the invalidation of bans on interracial marriage in Loving v Virginia.37 According to the majority opinion, the right to marry is protected under this clause as the most intimate and intense union between two people and is, although a safeguard for children and families, not conditioned on the capacity or will to have children. Finally, marriage has occupied from the outset a central position in the legal and social order of the United States so that many benefits linked to it are prohibited to same-sex couples. The latter’s right to marriage is also ensured by the

32 T Ginsburg and EA Posner, ‘Subconstitutionalism’ (2010) 62 Stanford Law Review 1583, 1605. See also Williams, The Law (2009) 21ff; JJ Dinan, ‘Court-Constraining Amendments and the State Constitutional Tradition’ (2007) 38 Rutgers Law Journal 983; and JJ Dinan, ‘State Constitutions and American Political Development’ in M Burgess and GA Tarr (eds), Constitutional Dynamics in Federal Systems: Subnational Perspectives (Montreal, McGill–Queen’s University Press, 2012) 56f. 33 For an overview, see Tarr, ‘Explaining Subnational’ (2011) 1147. 34 United States v Windsor 570 US ___ (2013). 35 Baker v Nelson 291 Minn 310, 191 NW 2d 185 (1971). 36 Obergefell v Hodges 576 US ___ (2015). 37 Loving v Virginia 388 US 1 (1967).

Switzerland 327

Equal Protection Clause of the Fourteenth Amendment, which is, according to Lawrence v Texas, concerning the legal treatment of same-sex couples closely interlinked with the Due Process Clause. By burdening the liberty of same-sex couples in contrast to other couples, the legislation in question subjects them to unequal treatment, which is to be evaluated in a different light as a result of societal changes that have occurred since Baker v Nelson. The four separate dissenting opinions denied the validity of the precedents cited for the above-mentioned reasoning on the basis of the Fourteenth Amendment because they had only removed unconstitutional limits to marriage as traditionally defined, but did not, unlike this decision, totally redefine marriage. They also advanced an argument related to both the separation of powers and federalism by claiming that the majority had not only engaged in legislation but, more precisely, also encroached upon the legislative prerogatives of the states. The Court would have stopped a vibrant debate that had witnessed the electorates of 11 states voting for the introduction of same-sex marriage and others refraining from it. The majority opinion acknowledged that democratic legislation is the means of effecting change. But it underlined that individuals whose (federal) fundamental rights are violated need not wait, like in the past as a consequence of Bowers v Hardwick,38 for (state) legislative action. 10.2 SWITZERLAND

(1) Compared to the oldest Swiss constitution, that of the canton of Geneva of 1847, the original federal constitution adopted one year later, contained only a few fundamental rights. Instead of drafting an extensive bill of rights, its rationale was rather to list only those rights whose scope had remained controversial due to differences in the pre-existing cantonal constitutions.39 The relative insignificance of rights from the national level remained unchanged until the Federal Court started, in the 1960s, to affirm the existence of a number of unwritten fundamental rights. It did so by referring to similar guarantees in the European Convention on Human Rights (ECHR) and/or to their general recognition in the cantonal constitutions. 10.C Switzerland 1995: An Unwritten Right to Subsistence? Three brothers from what was then Czechoslovakia had lived for several years in Switzerland as recognised refugees. After a court in Bern had condemned them in 1987 to a suspended prison sentence, the brothers were effectively expelled a few years later in the context of the changing political circumstances in Czechoslovakia and lost their refugee status. Soon thereafter, however, they re-entered Switzerland illegally and filed an application for social benefits. In response to the negative decision of the municipality and the appellate authorities, the brothers filed a constitutional complaint in 1994 to the Federal Court with reference to a

38

Bowers v Hardwick (n 27). D Thürer, ‘Verfassungsrechtlicher und völkerrechtlicher Status der Grundrechte’ in D Merten and H-J Papier (eds), Handbuch der Grundrechte in Deutschland und Europa: Grundrechte in der Schweiz und in Liechtenstein, vol 7/2 (Heidelberg, CF Müller, 2007) 42. 39 See

328 Fundamental Rights

federal and cantonal right to subsistence.40 In the latter regard, the Court held that the corresponding guarantee in Article 29(1) of the new Constitution of Bern was not applicable because it entered into force only after the contested decision had been issued. However, it found that even in the absence of an explicit provision in the federal Constitution, both Swiss citizens and foreigners, regardless of their residence status, were entitled to invoke an unwritten right to subsistence. The Court justified the affirmation of such a right by referring to similar guarantees in several cantonal constitutions, widespread individual rights of needy people to social assistance in ordinary cantonal legislation and the virtually unanimous endorsement of a right to subsistence among constitutional experts.

After the judicial recognition of several such unwritten federal rights,41 only the revised Swiss Constitution of 1999 entrenched a comprehensive catalogue of rights at the national level, which in part adopted the exact wording of ECHR provisions. Due to the comprehensiveness of this catalogue, the judicial recognition of further unwritten rights was in practice rendered unlikely.42 Moreover, this constitutional amendment also limited the space for the entrenchment of rights in the cantonal constitutions, as the Supremacy Clause of the Swiss Constitution declares the precedence of federal law over any conflicting provision of cantonal law (Article 49) and, even more specifically, requires that cantonal constitutions do not contradict federal law (Article 51(2)). In this regard, it is important to note that the constitutions of the cantons have to be consistent with the ECHR as well. This is because the latter forms, according to the Federal Court, an integral part of fundamental rights protection under the federal Constitution, with most of its provisions being directly applicable.43 That a canton has to respect these limits is safeguarded by the obligation to submit its constitution and every amendment to the federal Parliament for a legal oversight procedure. The Parliament is then obliged to formally ‘guarantee’, by means of a resolution, that the cantonal Constitution is not contrary to federal law (Article 51(2)).44 However, the few cases, in which this federal guarantee was refused, have never concerned fundamental rights provisions in cantonal constitutions. Although a few cantons, namely Jura, Vaud, Nidwalden and Graubünden, have their own constitutional courts, there is always the possibility of an appeal to the Swiss Federal Court (Article 189(1)(d)). (2) The cantons of Switzerland are not only allowed but are in fact obliged to have constitutions (Article 51(1) of the Swiss Constitution). Nearly all of them have 40

BGE 121 I 367. this manner, the Federal Court guaranteed the right to property (1960), freedom of expression (1961), personal freedom (1963), linguistic freedom (1965), freedom of assembly (1970) and the right to subsistence (1995). 42 See R Rhinow, Grundzüge des schweizerischen Verfassungsrechts (Basel, Helbing und Lichtenhahn Verlag, 2003) 174. 43 BGE 104 Ia 215, 221–22; BGE 101 Ia 67. 44 See ch 8.2.2. 41 In

Switzerland 329 adopted new ones since the 1960s. Arguably, the most important reason for this redrafting was the perceived need to update their fundamental rights guarantees.45 In this regard, the cantons are only allowed to exceed the minimum protection granted by federal rights because going below this standard is foreclosed by the above-mentioned Supremacy Clause (Article 49) and obligation to comply with federal law (Article 51(2)).46 In practice, the cantonal constitutions go beyond federal guarantees in different ways.47 In some cases, cantonal rights have been interpreted as being more far-reaching than their federal counterparts, even if they are identically worded. However, the number of these cases has remained limited because the majority of cantons have not established their own constitutional courts, which considerably diminishes the significance of cantonal rights guarantees in general.48 Other cantonal rights explicitly grant a wider scope of protection than the corresponding federal rights through a different formulation. One often-cited example is the right to petition, which in several cantons includes the right to a prompt response within a certain time limit.49 In other cases, cantonal constitutions contain entrenched rights unknown at the federal level, such as a right to work, housing, education or to run private schools.50 From a technical point of view, the entrenchment of fundamental rights in cantonal constitutions has followed three alternative paths. While Article 7 of the Constitution of Graubünden refers merely to the rights guaranteed by the federal Constitution and international treaties, Article 10(1) of Zurich’s Constitution makes a similar reference but also includes certain additional rights.51 Still others, like, above all, the Constitution of Jura (Titles II and III), contain comprehensive catalogues of cantonal rights and goals. However, this activism in some cantons cannot conceal the fact that the relationship between federal and cantonal rights has been reversed since 1848. Whereas the federal constitution used to supplement existing cantonal rights, these are today merely complements to the comprehensive federal bill of rights. While rights guarantees in subnational constitutions have therefore certainly become less important over time, it is questionable whether they have generally lost their pioneering role.52 In some areas, they still seem to have innovative potential with political rights

45 See N Schmitt, ‘New Constitutions for All Swiss Cantons: A Contemporary Challenge’ in Burgess and Tarr (eds), Constitutional Dynamics (2012) 146. 46 BGE 121 I 267. 47 See D Buser, Kantonales Staatsrecht: Eine Einführung für Studium und Praxis (Basel, Helbing and Lichtenhahn, 2004) 159ff. 48 See A Auer et al (eds), Droit constitutionnel suisse, 2nd edn (Bern, Stämpfli, 2006) 88. 49 For instance, one year according to Art 26 of the Constitution of Solothurn and Art 20.3 of the Constitution of Bern. 50 For a systematic overview of the length and content of the catalogues of cantonal fundamental rights, see Schmitt, ‘New Constitutions’ (2012) 148ff. 51 See G Biaggini, ‘Die neue Zürcher Kantonsverfassung: Gesamtbetrachtung im Lichte der Verfassungsfunktionen’ in LL Fosco et al (eds), Die neue Zürcher Kantonsverfassung (Zürich, Schulthess, 2006) 182. 52 See A Kley, ‘Der Grundrechtskatalog der nachgeführten Bundesverfassung—ausgewählte Neuerungen’ (1999) 135 Zeitschrift des Bernischen Juristenvereins 301.

330 Fundamental Rights being the most obvious example.53 All cantons have introduced legislative initiatives and direct elections of the executive and some of them have even introduced initiatives to recall the cantonal government or to abrogate acts of legislation. Other innovations that stem from the cantonal level are the entrenchment of social rights and social goals,54 the recognition of other forms of cohabitation besides marriage55 and the right of same-sex couples to a registered partnership.56 Several progressive initiatives on the part of some cantons were later copied by the federal Constitution of 1999 or other cantonal constitutions. These processes of mutual learning both in a vertical and horizontal dimension have produced a considerable degree of constitutional convergence and have been greatly facilitated by the fact that the federal and various cantonal constitutional reforms have taken place in a staggered manner over a period of nearly five decades.57 10.3 CANADA

(1) Right from the outset, it has to be emphasised that Canada is very much characterised by the still-prevailing British common law approach to fundamental rights. Following this approach, a person is free to do anything that is not explicitly prohibited, so individual rights actually ‘do not derive from positive law … but from an absence of positive law’.58 Moreover, following the doctrine of parliamentary sovereignty, these unregulated spaces of individual liberty should not have special constitutional protection. In contrast to the United States, the Canadian Constitution Act 1867 did not break with this legal heritage from the former colonial power and refrained from drafting a bill of rights with constitutional status.59 This remained unchanged for well over a century, as the Canadian Bill of Rights of 1960 was merely a statute and only applied to the federal level of government. Then the Constitution Act 1982 brought about two fundamental changes. First, it contained the comprehensive Canadian Charter of Rights and Freedoms (Sections 1–34). Second, it defined the Constitution of Canada (Section 52(2)) and declared that ‘any law that is inconsistent with the provisions of the Constitution is, to the extent of the inconsistency, of no force or effect’ (Section 52(1)). As part of this supreme law and by virtue of an even more explicit provision (Section 32(1)(b)), 53 See R Kägi-Diener, ‘Grundrechtsschutz durch die Kantone’ in Thürer et al (eds), Verfassungsrecht der Schweiz (Zürich, Schutlhess, 2001); K Nuspliger, ‘Wechselwirkungen zwischen neueren Kantonsverfassungen und der Bundesverfassung’ in U Zimmerli (ed), Die neue Bundesverfassung: Konsequenzen für Praxis und Wissenschaft (Bern, Stämpfli, 2000) 63ff. 54 The first differentiation between social rights and social goals was made in Arts 29 and 30 of the Constitution of Bern. 55 For example, Art 13(2) of the Constitution of Bern and Art 10(2) of the Constitution of Appenzell Ausserrhoden. 56 Art 14(2) of the Constitution of Fribourg. 57 See P Häberle, ‘Neuere Verfassungen und Verfassungsvorhaben in der Schweiz, insbesondere auf kantonaler Ebene’ (1985) 34 Jahrbuch des öffentlichen Rechts der Gegenwart 303, 339. 58 PW Hogg, Constitutional Law of Canada, 5th edn (Scarborough, Carswell, 2000) 34-2. 59 The Preamble to the Canadian Constitution even explicitly expresses the wish to have ‘a Constitution similar in Principle to that of the United Kingdom’.

Canada 331 the Canadian Charter is binding for the provinces, although the Notwithstanding Clause in Section 33 allows them, regarding a certain set of rights,60 to declare a provincial law as operative even in the face of a conflicting Charter provision. Despite this peculiar provincial power to make its ordinary law prevail over national constitutional law, the comprehensive catalogue of fundamental rights has substantially restricted the autonomy of the provinces to entrench their own fundamental rights. Until then, they had been almost completely free to amend the ‘provincial constitutions’, which had been provided by Sections 58–90 of the Constitution Act 1867. The only requirement had been to retain the office of Lieutenant Governor61 as the Queen’s representative (Section 92(1)) and, as a consequence, some form of representative democracy.62 By contrast, amendments according to the new Section 45 of the Constitution Act 1982 are subject to the federal standard of fundamental rights. Compliance with this standard is enforced by the Supreme Court of Canada, which deals today in most of its cases with conflicts between federal or provincial law and the Canadian Charter. A particularly prominent judgment was issued in 1988. 10.D Canada 1988: Outdoor Commercial Signs Only in French? An appeal to the Supreme Court of Canada that was launched by the Government of Quebec consolidated a number of cases initiated by Montreal-area merchants, including wool shop owner Valerie Ford. Following complaints, the Office québécois de la langue française, which is responsible for the enforcement of Quebec’s 1977 Charter of the French Language (Bill 101), had requested that they replace their bilingual outdoor commercial signs because the Charter requires that such particularly visible signs use French exclusively (Sections 58, 69 and 205–08). The shop owners argued that such an order would violate their freedom of expression as guaranteed by Section 2 of the Canadian Charter of Rights and Freedoms. In a landmark judgment, the Supreme Court decided in favour of the shop owners and struck down those sections of the Charter of the French Language that compelled business owners to use monolingual French outdoor signs.63 Nonetheless, in an attempt to find a broadly acceptable compromise solution, the Court held that the Quebec Government could still legitimately require French to have ‘greater visibility’ or ‘marked predominance’ on exterior commercial signs. Quebec’s National Assembly responded to the ruling by making minor amendments to the Charter of the French language (Bill 178, the ‘insideoutside bill’) that did not live up to the requirements set by the Court. In recognition of this, it then made use of the Notwithstanding Clause in Section 33 of the Canadian Charter to shield the new regulation from review by the Supreme Court. Only in 1993 was an amendment passed (Bill 86) that stated, in accordance with Ford v Quebec, that in spite of an obligation for the predominance of French on outdoor signs, the use of any second language is permitted.

60 Exceptions to the applicability of Sec 33 are democratic rights (Sec 3–5), mobility (Sec 6), language rights (Sec 16–23) and sexual equality (Sec 28). See Hogg, Constitutional Law (2000) 36-11. 61 Sec 92(1) of the Constitution Act 1867. 62 Reference Re Initiative and Referendum Act [1919] AC 935 (JCPC); R v Nat Bell Liquors Ltd [1922] 2 AC 128 (JCPC). 63 Ford v Quebec (AG) [1988] 2 SCR 712.

332 Fundamental Rights This and many other cases epitomise the significance of the Supreme Court of Canada as the ultimate interpreter not only of federal law but also of provincial law. While this importance certainly increased through the above-mentioned constitutional changes of 1982, it is rooted in the original constitution of 1867. On the basis of the federal competence to establish ‘a general court of appeal for Canada’ under Section 101 of the Constitution Act 1867, the Supreme Court’s appellate jurisdiction is explicitly declared to be general ‘within and throughout Canada’ (Section 35 of the Supreme Court Act). This makes it a powerful unitary institution at the top of the 10 provincial court hierarchies, which culminate in the provincial courts of appeal. The lack of a dual court system and the fear of centralising constitutional interpretation is regarded by some observers as an important reason why the provinces have so far been largely inactive with regard to constitution-making in general and the entrenchment of fundamental rights in particular.64 (2) It is indeed striking that ‘provincial constitutions barely dwell in the world of the (Canadian) subconscious. They are too opaque, oblique, and inchoate to rouse much interest, let alone passion’.65 The fact that constitutions in a formal sense as procedurally qualified fundamental legal documents do not exist in any province is mostly attributed to a sceptical attitude towards constitutionalism owed to the British heritage of parliamentary sovereignty.66 However, as ‘all subnational units must have constitutions of some kind, formal or informal’,67 the question remains then where typical contents of constitutional law, chiefly among them fundamental rights, are enshrined. Although ‘provincial constitutions’ are only mentioned but not defined by Section 45 of the Constitution Act 1982, it is widely recognised that they are, in the specific Canadian sense, not single codified documents but made up of different sources like constitutional conventions, portions of the federal constitution and certain ordinary provincial statutes.68 As far as these statutes are concerned, the scope of the provincial constitutions is problematically ill-defined.69 However, it is generally accepted that ordinary legislation dealing with fundamental rights certainly forms part of them. Already during World War II, the provincial legislatures had begun to adopt anti-discrimination laws and later developed them into human rights codes, typically enforceable by human rights commissions with the option of appealing to the courts. This step was first taken by Ontario in 1962 and then emulated by all other provinces, as well as the federal Government. Unlike the Canadian Charter, which is only binding for 64 See FL Morton, ‘Provincial Constitutions in Canada’ (2004) Subnational Constitutions and Federalism: Design and Reform Papers, 4 statecon.camden.rutgers.edu/sites/statecon/files/subpapers/morton.pdf. 65 N Wiseman, ‘Clarifying Provincial Constitutions’ (1996) 6 National Journal of Constitutional Law 269, 270. 66 See G Baier, ‘Canada: Federal and Subnational Constitutional Practices’ in Burgess and Tarr (n 32) 174. 67 C Saunders, ‘The Relationship between National and Subnational Constitutions’ in KonradAdenauer-Stiftung (ed), Subnational Constitutional Governance (Pretoria, KAS South Africa, 1999) 26. 68 See Tarr, ‘Subnational Constitutions’ (2008) 175f. See also ch 5.1.1. 69 See N Wiseman, ‘In Search of a Québec Constitution’ (2008) 2 Revue québécoise de droit constitutionnel 174, 142f.

Germany 333 federal and provincial governments (Section 32 of the Constitution Act 1982), they also apply to private actors in particular, so that they have significance even after the adoption of the Canadian Charter. The opposite is true for statutory bills of rights, which go beyond anti-discrimination. The first such statute, the 1947 Saskatchewan Bill of Rights Act, still focused on a few basic negative rights. The Alberta Bill of Rights and the Quebec Charter of Rights and Freedoms, enacted in the 1970s based on the model of the abovementioned Canadian Bill of Rights, were then already much more comprehensive and had the effect of overriding inconsistent provincial legislation. Since 1982, they have remained relevant in the rather few instances in which they exceed the Canadian Charter.70 Alberta and, above all, Quebec are also the only provinces to raise the idea of consolidating all statutes concerning typical constitutional contents, as well as law from the above-mentioned other sources of provincial constitutional law, into a single constitution and of formally entrenching it. In the latter case, this idea has given rise to much more tangible results as illustrated by the numerous initiatives in Quebec’s National Assembly.71 The most recent draft was introduced by Bill 196 in 2007 (which interestingly does not mention the words ‘Canada’ or ‘Canadian’ at all) and also dealt with fundamental rights. Instead of formulating a comprehensive catalogue, this bill declared Quebec’s Charter of Rights and Freedoms of 1975 and its Charter of the French Language of 1977 as integral parts of the Constitution and mandated their interpretation with due regard for ‘ensuring the predominance of the French language, protecting and promoting Québec culture’ (Section 8). This mandate illustrates that the cultural impetus has played a crucial role for Quebec’s— compared to other provinces’—exceptional constitutional activism. Even there, however, no draft of a formally entrenched constitution that includes fundamental rights has so far succeeded in passing the provincial Parliament. 10.4 GERMANY

(1) The German Basic Law of 1949 is very much shaped by the experience of totalitarian dictatorship.72 In this context, it entrenched a comprehensive catalogue of fundamental rights, emphasised its importance by placing it at the top of the Constitution and made it explicitly binding for both the federal Government and the Länder (Article 1(3) of the Basic Law). In order to safeguard homogeneity between

70 Examples of Supreme Court cases involving provincial rights with a wider scope than their federal counterparts are Godbout v Longueuil (City of) [1997] 3 SCR 844 (right to private life in the Quebec Charter); Syndicat Northcrest v Anselem [2004] 2 SCR 551 (application of Quebec Charter to a private contract); Bruker v Marcovitz [2007] 3 SCR 607 (application of Quebec Charter to a private contract). 71 While, in Alberta, a government-sponsored commission was not against the idea of a provincial constitution, it also did not vigorously support it, MLA Committee, Strengthening Alberta’s Role in Confederation (Edmonton, Government of Alberta, 2004) 59. 72 See B-O Bryde, ‘Fundamental Rights as Guidelines and Inspiration: German Constitutionalism in International Perspective’ (2007) 25 Wisconsin International Law Journal 189, 189.

334 Fundamental Rights these rights and those that may be enshrined in the constitutions of the Länder, the Basic Law obliges the federal Government to guarantee the consistency of both constitutional orders (Article 28(3)) and establishes the general rule of precedence of federal law over Land law (Article 31).73 On the basis of the latter provision, the abolition of the death penalty through Article 102 of the Basic Law overrides the possibility of such punishment formally still foreseen in the 1946 Constitution of Hesse for particularly serious crimes until today. Another case concerning the precedence of the national Constitution over Land law involved Bavarian legislation on crucifixes in public schools. 10.E Germany 1995: Crucifixes in Bavaria and Religious Freedom The 1983 Elementary School Regulations of the Land Bavaria stipulated that a crucifix or at least a cross had to be displayed in every classroom of public schools. Several parents challenged this provision. In 1995, the Federal Constitutional Court held that the children’s negative freedom of faith under Article 4 of the Basic Law had been violated.74 Ordinary legislation of a Land is not allowed to restrict this fundamental right entrenched in the national Constitution. The judges emphasised that the crucifix was not merely a cultural symbol of humanitarianism but the symbol of a specific religion. However, the practical impact of this ruling remains limited as the Bavarian Government subsequently declared that the crucifix should only be removed in atypical exceptional cases and in response to single lawsuits. Soon after the judgment, the Land’s legislature adopted the 1995 Law on Education and Schooling which states that ‘in view of Bavaria’s historical and cultural heritage a cross is displayed in each classroom’. The Constitutional Court of Bavaria has rejected individual fundamental rights complaints under Article 98 of the Land Constitution on the grounds that the new law would provide for a mechanism to resolve conflicts regarding this issue. The Federal Constitutional Court has refrained from accepting complaints against this decision of the Bavarian court.

If the supremacy clause collides with fundamental rights in subnational constitutions, the Basic Law contains an important exception to the rule of federal precedence. Such subnational rights ‘remain in force insofar as they guarantee basic rights in conformity with Articles 1 to 18 of this Basic Law’ (Article 142).75 It is widely held that this exception does not save these Land rights, which fall short of the protection guaranteed by the federal rights, from being overridden by the abovementioned Article 31 regarding the precedence of federal law.76 A case in point was the controversial Bavarian law on the right to assembly.

73

See ch 5.1.2. 93 BverfGE 1 (Classroom Crucifix II). 75 Notwithstanding the wording ‘remain in force’, which refers to the Länder rights that existed before the Basic Law, it is generally accepted that Art 142 also covers Länder rights enshrined after 1949. See A Stiens, Chance und Grenzen der Landesverfassungen im deutschen Bundestaat der Gegenwart (Berlin, Dunker & Humblot, 1997) 39. 76 See S Storr, Verfassunggebung in den Ländern: Zur Verfassunggebung unter den Rahmenbedingungen des Grundgesetzes (Stuttgart, Boorberg, 1995) 220ff. 74

Germany 335

10.F Germany 2009–12: A More Restrictive Law on the Right to Assembly for Bavaria? The 2006 reform of German federalism transformed the previously concurrent competence to regulate the right to assembly into a competence of the Länder (Article 74(1)(3) of the Basic Law). In 2008, Bavaria became the first Land to make use of this new provision by adopting legislation supplanting the federal law previously in force. The Bavarian law, however, soon met harsh criticism because, among other things, it imposed on organisers very onerous reporting duties, even exceeding the assembly in question itself, and enabled police to produce and analyse ‘overview images’ of public meetings. In response, an alliance of 13 organisations, composed of political parties, trade unions and non-governmental organisations, decided to file a constitutional complaint against the law before the Federal Constitutional Court and to request a preliminary injunction. As to the latter request, the judges soon reacted by suspending several provisions of the Bavarian law and mandating a narrower interpretation of others.77 For example, it held that police could produce ‘overview images’ only if doing so is required by the size and confusing situation of the individual public meeting in question. Moreover, these images may only be analysed within two months after the assembly and must then be deleted. In the wake of this decision, the Bavarian law was amended in 2010. Two years later, the Federal Constitutional Court held that the constitutional complaint, referring partly to the old and partly to the new version of the law, was inadmissible and thus, in the end, refrained from deciding on the merits of the case.78

While the Article 142 exception to the supremacy clause fails to cover cases of more restrictive Land laws, it certainly does encompass those cases in which the federal standard of rights protection is exceeded. This holds true unless the Basic Law itself contains a ceiling for this protection. For instance, its explicit prohibition of associations that ‘contravene the criminal laws, or that are directed against the constitutional order or the idea of international understanding’ (Article 9(2)), also bars the Land constitutions from extending the freedom of association to these cases. A third category of subnational rights, namely those with identical content to that of their federal counterparts, likewise remains in force because the Basic Law was clearly intended to open up the possibility of recourse to the respective Land constitutional court.79 The supremacy clause of Article 31 is relevant not only in the relationship between the Basic Law and Land rights. The latter are also overridden by ordinary federal legislation.80 Interestingly, Article 142 is, in such a case, not applicable so that Land rights colliding with a federal law are always null and void. For example, Article 4(3) of the Constitution of Thuringia, requiring a judge’s order 24 hours at the latest after arrest, is overridden by the 48-hour period established by Article 13(1) of the Federal Law on the Deprivation of Liberty.81 77

122 BVerfGE 342 (Bavarian Assembly Act). BVerfG 21 March 2012, 1 BvR 2492/08. 79 See Storr, Verfassunggebung (1995) 220. 80 See J Dietlein, Die Grundrechte in den Verfassungen der neuen Bundesländer (München, Vahlen, 1993) 55. 81 See PM Huber, ‘Die neue Verfassung des Freistaats Thüringen’ (1994) 4 Landes- und Kommunalverwaltung 121, 124. 78

336 Fundamental Rights As far as the interpretation of subnational rights is concerned, both the Federal Constitutional Court and the constitutional courts of the Länder82 play a role. Most Länder grant individuals the possibility to file before their constitutional court a complaint concerning violations of subnational rights.83 An individual can then choose freely, unless prevented by more restrictive Land regulation,84 between seeking protection of Land rights at the Federal Constitutional Court or its counterpart at the subnational level or simultaneously before both of them.85 In the latter case, the judgment of the Land constitutional court takes precedence if the subnational right grants more protection than its federal equivalent.86 If the two rights provisions have identical content, it is arguably bound to the interpretation of the Federal Constitutional Court.87 Although the Court in Karlsruhe usually does not mix federal and subnational constitutional law, its efforts towards systematic interpretation of a Basic Law right have sometimes led it to discuss a Land right with identical content.88 (2) The autonomy to guarantee fundamental rights has been used by the Länder to a rather varying degree. In this regard, the period in which a constitution was drafted has clearly played a major role.89 The five constitutions of the first generation, which were passed between 1945 and 1949, the year in which the Basic Law was adopted, were envisaged in the absence of federal limitations as ‘complete constitutions’ (Vollverfassungen). As such and in response to Nazi rule, they strongly embodied the political leanings of the Land Government to either conservatism90 or social democracy91 and contained extensive bills of rights. When the second generation of constitutions was drafted between 1949 and 1953,92 the Basic Law, with its comprehensive rights catalogue, was already in force, so they put their focus clearly on matters of institutional organisation. 82

See C Starck and K Stern (eds), Landesverfassungsgerichtsbarkeit (Baden-Baden, Nomos, 1983). These are all ‘new Länder’, Bavaria, Berlin, Hesse, Saarland and Rhineland-Palatinate. A few other Land constitutions also foresee the instrument of the constitutional complaint but restrict the initiative to certain government bodies. Art 98 of the Bavarian Constitution even grants individuals whose rights are not violated the right to file a complaint concerning the asserted inconsistency of any Bavarian legal act with a fundamental right guaranteed in the Land constitution (‘Popularklage’). 84 Several Länder assign to the constitutional complaint before their respective constitutional courts only subsidiary status. 85 22 BVerfGE 267, 272 (Uniform Fundamental Right). 86 See, eg, 53 BVerfGE 185 and StGH Hess 30 December 1981, PSt 880. 87 This view is reflected in 22 BVerfGE 267, 270 and 36 BVerfGE 342, 363 (Remuneration Law of Lower Saxony), albeit contested by a significant number of legal scholars. For an overview of this debare, see Stiens, Chance und Grenzen (1997) 189f. 88 See, eg, 2 BVerfGE 237, 262 (Mortgage Security Act) and 27 BVerfGE 71, 80 (Leipzig Daily Newspaper). 89 The classical differentiation of three generations of Land constitutions goes back to C Starck, ‘Verfassungsgebung in den neuen Ländern’ (1992) 7 Zeitschrift für Gesetzgebung 1, 1. For an overview of the three generations, see AB Gunlicks, The Länder and German Federalism (Manchester, Manchester University Press, 2003) 141ff. 90 The constitutions of Rhineland-Palatinate, Bavaria and the Saarland. 91 The constitutions of Bremen and Hesse. 92 The constitutions of Hamburg, North Rhine-Westphalia, Schleswig-Holstein, Baden-Württemberg and Lower Saxony (new constitution in 1993). Only the Constitution of Berlin (new constitution in 1995) did not follow this pattern due to its particular Cold War status as a divided city. 83

Germany 337 After constitution-making had ceased to be an issue for nearly four decades, German reunification gave rise to a third generation of subnational constitutions. Those of the five ‘new Länder’,93 which acceded to the federation in 1990, are unique insofar as their fundamental rights provisions reflect modern trends such as rights to environmental protection and data protection. On the other hand, they also emphasise their historical distinctiveness related to the socialist past with such rights as the right to work, housing, education and social security.94 The constitutions of the third generation were certainly influenced by the fundamental right provisions of the Basic Law and the constitutions of the ‘old Länder’, but they also influenced subsequent constitution-making projects of the 1990s like those in Lower Saxony and Berlin. The fundamental rights provisions of the three generations of Land constitutions differ not only in terms of their content but also from a technical point of view. Quite understandably, those of the second and third generations do not contain comprehensive autonomously drafted bills of rights like those adopted before 1949. They typically include several self-formulated guarantees, but otherwise they either refer to the catalogue of the Basic Law95 or, like most of the ‘new Länder’, largely repeat the federal rights often with almost the same wording. When Land rights exceed federal rights, they either merely extend the latter’s scope of protection or they create completely new claims. Examples of the first category are a right to petition, including a formal assessment, or a right of access to data, which comprises environmental data. The autonomy of the Länder to ‘invent’ new rights is restricted by a decision of the Federal Constitutional Court. This judgment effectively prevents them from introducing new negative rights because this category of rights is regarded as being covered exhaustively by the Basic Law.96 10.G Germany 1957: General Freedom of Action as Subsidiary Provision of the Basic Law? In 1953, German authorities refused, on the basis of § 7(1)(a) of the Federal Passport Act, to renew the travel documents of Christian Democratic Union of Germany (CDU) politician Wilhelm Elfes, who was known for his staunch opposition to policies of the CDU-led federal Government regarding military defence and reunification. This denial was justified, at the successive stages of appeal, by his participation in conferences and demonstrations abroad in which he had voiced sharp criticism of these policies. Upon a constitutional complaint by Elfes, the Federal Constitutional Court found that the freedom of movement under Article 11 of the Basic Law did not encompass the right to

93 This term refers to the five re-established Länder in the former German Democratic Republic which acceded to the Federal Republic of Germany upon reunification on 3 October 1990. 94 See K Stern, Das Staatsrecht der Bundesrepublik Deutschland: Allgemeine Lehren der Grundrechte, vol 3/2 (München, CH Beck, 1994) 1447. 95 This approach is taken by Art 2(1) of the Constitution of Baden-Württemberg, Art 3(2) of the Constitution of Lower Saxony, Art 4(1) of the Constitution of North Rhine-Westphalia and Art 5(3) of the Constitution of Mecklenburg-Vorpommern. 96 See DP Kommers and RA Miller, The Constitutional Jurisprudence of the Federal Republic of Germany, 3rd edn (Durham, Duke University Press, 2012) 315.

338 Fundamental Rights

travel abroad.97 However, it derived such a right from Article 2(1) of the Basic Law, which guarantees every person ‘the right to free development of his personality insofar as he does not violate the rights of others or offend against the constitutional order or the moral law’. It did so by construing this article as a guarantee of a general freedom of action (allgemeine Handlungsfreiheit) and as a subsidiary provision (Auffanggrundrecht). As such, Article 2(1) is relevant whenever none of the special negative rights enshrined in the Basic Law is violated but government activity nevertheless interferes with personal freedom. Such interference is only justified if it is covered by the ‘constitutional order’ (Article 2(1)), which in this sense also encompasses ordinary legislation. In this specific case, the Federal Passport Act provided such a justification.

In view of this decision, it is entirely logical that new Land rights are typically positive guarantees concerning work, housing, education and social security. These stem, above all, from the third-generation constitutions and take, in most cases, the form of goals with a programmatic nature rather than legally enforceable social rights.98 The fact that a clear distinction between these two categories is missing in a number of cases, above all in the Constitution of Brandenburg, has ignited considerable controversy because this threatens to overburden the courts with making a distinction that ought to be made in the constitution itself. 10.5 SPAIN

(1) The statutes of Spain’s autonomous communities are, from a formal perspective, characterised by their dual nature. On the one hand, they are ‘within the terms of the present [Spanish] Constitution … the basic institutional rule of the autonomous communities’ (Article 147(1) of the Spanish Constitution). On the other hand, the original statutes and all amendments are not purely regional acts of legislation but are, at the same time, Spanish organic laws, which are bilaterally negotiated and finally approved by both the respective regional parliament and the national Parliament (Article 147(3)).99 While some scholars explicitly qualify the statutes as ‘subnational constitutions’,100 most observers rather regard them as ‘secondary constitutional laws’ or ‘quasi-constitutional norms’.101 According to the Constitutional Court, the statutes are—together with the Spanish Constitution as well as certain organic and ordinary laws with relevance to the distribution of powers—part of the ‘constitutional block’ (bloque de constitucionalidad), which is superior to all

97

6 BVerfGE 32 (Elfes case). See C Starck, ‘The Constitutionalisation Process of the New Länder: A Source of Inspiration for the Basic Law?’ (1994) 3 German Politics 118, 127f. 99 On the procedures of adopting and amending the statutes, see chs 2.3.2 and 5.1.3. 100 See, eg, C Colino, ‘Constitutional Change without Constitutional Reform: Spanish Federalism and the Revision of Catalonia’s Statute of Autonomy’ (2009) 39 Publius 262. 101 See, eg, F Rubio Llorente, ‘El bloque de la constitucionalidad’ (1989) 27 Revista Española de Derecho Constitucional X, 25f. 98

Spain 339 remaining national and regional law and forms the standard for the constitutional conformity of any legislation.102 Within this block, they are nonetheless subordinated to the Spanish Constitution as the supreme source of all law in Spain,103 even though the Constitutional Court recently acknowledged that the statutes include typical constitutional contents and can thus be classified as materially constitutional.104 Of such nature are certainly the compulsory core contents of each statute, which are, according to Article 147(2) of the Spanish Constitution, the name, territorial boundaries, institutional organisation and powers of the autonomous community. Moreover, each statute can also have optional additional contents like those explicitly authorised by the Constitution, namely regulations concerning a co-official regional language (Article 3(2)) and a flag of the autonomous community (Article 4(2)). Beyond that, still other issues not expressly indicated by the Constitution can, according to the prevailing case law, also be included as optional additional contents if they are an adequate complement in view of the statutes’ function as the ‘basic institutional rule of the autonomous communities’ (Article 147(1)).105 Provisions that entrench rights belong to this third category of statute contents. In any case, the freedom of the autonomous communities to devise their statutes at their own discretion is limited by both the aforementioned mandatory approval of the Spanish Parliament and the Constitutional Court as the ultimate interpreter of regional rights provisions. The latter recently emphatically stressed its general jurisdiction ‘over the whole Spanish territory’ (Article 161 of the Spanish Constitution) when it declared binding opinions of the Catalan Consell de Garanties Estatutàries regarding bills affecting statutory rights (Article 76(4) of the 2006 Statute of Catalonia) as inconsistent with this monopoly of constitutional interpretation.106 (2) Compared to the first statutes adopted after 1978, the reformed statutes that have been passed since 2006 are, in many cases, much more comprehensive and—as far as the content is concerned—constitution-like documents.107 To some degree, this is also the result of the active entrenchment of rather extensive rights guarantees that is typical of nearly all statutes. A case in point is the Statute of Catalonia. Whereas Article 8 of the old statute of 1979 had merely referred to the rights enshrined in the Spanish Constitution, the reformed one of 2006 includes an entirely new Title I on 102

STC 66/1985, FJ 1. 4/1981, FJ 3. The position was recently confirmed in STC 31/2010, FJ 3: ‘Statutes of Autonomy are rules subordinated to the Constitution, as it corresponds to normative provisions that are not an expression of a sovereign power, but of a devolved autonomy based on the Constitution … As the supreme rule of the Legal System, the Constitution admits no equal or superior, only rules that are hierarchically subjected to it in all regards.’ 104 STC 31/2010, FJ 3. 105 STC 89/1984; STC 247/2007, FJ 12; STC 31/2010, FJ 4. 106 STC 31/2010, FJ 32. 107 For an overview of the particularly important new statutes of Valencia, Catalonia and Andalusia, see R Serra and P Oñate, ‘The Reform of the Spanish Subnational Constitutions: Rules and Regulations and Political Context’ (VII World Congress on Constitutional Law, Athens, 2007) 12ff. 103 STC

340 Fundamental Rights ‘Rights, Obligations and Governing Principles’.108 This title contains as many as 40 provisions grouped in civil and social rights (Articles 15–28), political and administrative rights (Articles 29–31), linguistic rights and obligations (Articles 32–36) and governing principles, which shall guide public policy (Articles 37–54). At the same time, the statute explicitly recognises the complementary nature of Title I, as none of its provisions ‘shall be enacted, applied or interpreted in any way that reduces or restricts the fundamental rights recognized in the Constitution and in international treaties and conventions ratified by Spain’ (Article 37(4) of the 2006 Statute of Catalonia). Although civil rights in the strict sense of the term are derived rather from the national and international levels and are thus largely neglected by the Catalonian statute, its broad range of rights is nevertheless impressive. Although the statute certainly places an emphasis on social rights, it still has very individual Catalonian characteristics such as, above all, the right and obligation to speak the Catalan language. The fact that many Title I rights had already been contemplated in ordinary legislation clearly indicates that their elevation to the statutory rank is to some degree also a symbolic act. It is an attempt to sustain, through a comprehensive bill of rights in the more visible statute, a sense of regional quasi-citizenship.109 While the focus on social rights is also a typical feature of the other recently reformed statutes, which were often modelled on the Catalonian example, these nonetheless exhibit an individual character resulting from different priorities. The statutes of Valencia and the Balearic Islands, for instance, include a right to a sufficient quantity of high-quality water. The Andalusian bill of rights is unique because of its extensive equality guarantees, for instance in terms of gender and for cohabiting couples, as well as the explicit right to a public and secular education. These characteristics reflect the specific origins of the Statute of Andalusia, which was, unlike all other statutes, not a compromise at the regional level with conservative parties but adopted by a left-wing coalition. Regardless of the individual thematic focal points of the different regional rights catalogues, the legal nature of their provisions has been controversial from the beginning in both politics and academia. Among constitutional lawyers, positions ranged from those who did not consider these provisions to be generally incompatible with the Constitution110 to others who regarded asymmetrical rights protection in different parts of Spain as conflicting with constitutional equality standards.111 In a landmark decision concerning the Statute of Valencia, the Constitutional Court refused to grant certain guarantees the character of fully fledged individual rights while it accorded such character to others.

108 See JM Castellà Andreu, ‘Spanish Autonomic State Reform: Catalonia’s 2006 New Statute of Autonomy (2008) 102 The Journal of Law and Political Science 89, 101. 109 See M Keating and A Wilson, ‘Renegotiating the State of Autonomies: Statute Reform and Multilevel Politics in Spain’ (2009) 32 West European Politics 536, 550. 110 See, eg, F Caamaño Domínguez, ‘Sí, pueden (declaraciones de derechos y Estatutos de Autonomía)’ (2007) 79 Revista Española de Derecho Constitucional 33. 111 See, eg, LM Díez-Picazo, ‘¿Pueden los Estatutos de Autonomía declarar derechos, deberes y principios?’ (2006) 78 Revista Española de Derecho Constitucional 63.

Comparative Conclusions 341

10.H Spain 2007–10: Justiciable Fundamental Rights in the Autonomy Statutes? During its elaboration, the Statute of Valencia was presented by the two main Spanish parties, the Partido Socialista Obrero Español (PSOE) and the Partido Popular (PP), as an example worth following. It was eventually approved by an organic law in 2006 as the first of all reformed statutes. However, the Government of the Autonomous Community of Aragon challenged Article 17(1) of the Valencian Statute, which guaranteed a right to sufficiently high-quality water. The Constitutional Court upheld the appealed provision and took the judgment as an opportunity for more general reflections on the nature of statutory rights provisions. Although such provisions are not expressly foreseen in the Spanish Constitution as either compulsory core content or as optional additional content, the Court deemed them lawful as an adequate complement in view of the statutes’ function as the ‘basic institutional rule of each autonomous community’ (Article 147(1) of the Spanish Constitution).112 However, with regard to the legal nature of these guarantees, the ruling only recognises those that have, like participatory rights, a direct connection to the compulsory statute content concerning institutional organisation (Article 147(2)(c) and 152(1)), as fully fledged individual rights. Other guarantees, such as social rights, above all, which are far more extensive in the statutes, are merely principios rectores. As such, they are addressed to the respective autonomous community and shall only form guiding principles for its social and economic policies. Albeit accorded legal character, they are not regarded as directly enforceable but as needing legal implementation. The more recent judgment of the Constitutional Court regarding the Statute of Catalonia confirmed this line of reasoning, although it did not explicitly indicate which provisions were individual rights and which ones were only principios rectores.113 10.6 COMPARATIVE CONCLUSIONS

(1) In certain countries covered in our comparative analysis, such as the United States, Switzerland and Germany, several subnational bills of rights are older than those at the national level so that they naturally enjoyed a high degree of discretion. However, subnational constitutional space concerning autonomous rights guarantees is typically restricted to a greater or lesser degree by the national constitution. First and foremost, this limitation is effected by general supremacy clauses,114 which stipulate the precedence of national law, sometimes only of the national constitution, over subnational law, and/or a specific requirement for subnational constitutions to be compliant with their national counterpart, sometimes also with ordinary national legislation.115 In these cases, the national constitution does not have to explicitly 112

STC 247/2007. STC 31/2010. 114 See F Palermo, ‘Annährungen an den ‘Verfassungskern’ der Europäischen Union: Die Homogenitätsund Integrationsklauseln im integrierten Verfassungsrecht des Europäischen Verfassungsraumes’ in K Weber and N Wimmer (eds), Vom Verfassungsstaat am Scheideweg: Festschrift für Peter Pernthaler (Wien, Springer, 2005). 115 Art 28(3) of the Basic Law, STC 4/1981, FJ 3 and most recently STC 31/2010, FJ 3, as well as Art 51(2) of the Swiss Constitution (also compliance with ordinary national legislation). 113

342 Fundamental Rights prescribe the contents of subnational constitutions because the national bill of rights then anyway sets a binding minimum standard that subnational entities are not allowed to fall short of. Thus, the more extensive national rights catalogues are as ‘engines of symmetrisation’, the lesser is the discretion of subnational entities and the potential of differentiation among them. Quite evidently, there has indeed been, over time, a clear tendency towards more extensive national catalogues, which is, in some cases, the result of judicial interpretation. Prime examples are the incorporation of large portions of the federal Bill of Rights by the US Supreme Court or the recognition of unwritten fundamental rights by the Swiss Federal Court, although the latter’s case law is based on widespread recognition in cantonal constitutions and thus contains a certain subnational element. In other cases, national bills of rights are more extensive today than several decades ago because of the recent constitutional entrenchment of comprehensive rights catalogues. Cases in point are the Swiss Constitution of 1999, as well as the 1982 Canadian Charter of Rights and Freedoms. To be sure, the Notwithstanding Clause in Section 33 of the Constitution Act 1982 allows the provinces to shield their legislation against certain rights of the Canadian Charter with which it is incompatible or to re-enact such legislation. An override power of this kind makes a judicial veto merely suspensive and obviously significantly strengthens subnational entities. In the United States, for example, certain states would, with a similar provision, have most likely overridden some of the progressive rulings of the Warren Court.116 Yet, there is no doubt that Canada’s 1982 Charter has clearly restricted provincial discretion particularly regarding language rights,117 for which the Notwithstanding Clause is not applicable. This is illustrated by a set of crucial Supreme Court cases, ranging from the above-mentioned case Ford v Quebec in 1988 to Solski v Quebec in 2005.118 Overall, the extension of national fundamental catalogues has thus been, in Canada and even more so in the other countries compared, an important factor in centralisation and symmetrisation. Apart from national law, a further limitation of the autonomy to entrench fundamental rights may arise from an obligation to obtain the approval of a subnational constitution from national institutions. The US Congress has mostly used its authority to establish arbitrary conditions on the admission of new states to the Union (Article IV, clauses 3 and 4 of the US Constitution) only to guarantee the consistency of their constitutions with very broad criteria so that its constraining effect has clearly been limited. Compared to this congressional power, the mandatory approval by the Swiss Parliament is more far-reaching because it has to be sought not only for original cantonal constitutions but also for each amendment (Article 51(2) of the Swiss Constitution). At the same time, however, it is less extensive insofar as it cannot be denied on the basis of arbitrary criteria but only in case of an inconsistency with federal law. In practice, it has proven to be equally insignificant, as none of

116

See Hogg (n 58) 36-12. S Choudhry, ‘Bills of Rights as Instruments of Nation-Building in Multinational States: The Canadian Charter and Quebec Nationalism’ in JB Kelly and CP Manfredi, Contested Constitutionalism: Reflections on the Charter of Rights and Freedoms (Vancouver, University of British Columbia Press 2009). 118 Ford v Quebec (n 63); Solski (Tutor of) v Quebec (Attorney General) [2005] 1 SCR 201. 117

Comparative Conclusions 343 the rare refusals of the federal guarantee has ever been related to rights provisions. The opposite is true for the statutes of Spain’s autonomous communities and their amendments, which always have to be adopted as organic laws by the national Parliament (Article 147(3) of the Spanish Constitution). The power of the national Parliament to modify or reject a draft statute at will makes subnational ambitions of rights entrenchment essentially dependent on what is achievable in bilateral negotiations, which is again dependent on the respective parties in government at both the regional and national level and whether they are able to reach a consensus. A third constraint on subnational autonomy may result from restrictive judicial interpretation of subnational rights by a national apex court. In the US case, the limitation of the Supreme Court to the interpretation of the national Constitution has greatly facilitated states’ rights protection in the wake of the New Judicial Federalism. It has done so by providing sufficient space for state supreme courts to act as the ultimate and independent interpreters of state constitutions, even if they also invoke—for the purpose of guidance—federal doctrine. Other subnational constitutional courts, if they exist at all, lack such space because their national counterparts are, in relative terms, clearly more powerful. For instance, the Supreme Court of Canada is explicitly granted general appellate jurisdiction throughout the country (Section 101 of the Constitution Act 1867 and Section 35 of the Supreme Court Act), while the jurisdiction of the Spanish Constitutional Court ‘over the whole Spanish territory’ explicitly includes the check of constitutionality of all legislation within this territorial scope (Article 161 of the Spanish Constitution). The Federal Court in Switzerland is likewise a unitary and harmonising institution, which is the ultimate interpreter even regarding the four cantons with their own constitutional courts (Article 189(1)(d) of the Swiss Constitution). By contrast, the German case requires a more differentiated view. Unless a differing Land regulation is in place, a decision of a Land constitutional court is, as a general rule, final and takes priority if the protection granted by the subnational right in question exceeds its federal counterpart. In conclusion, all countries analysed except the United States lack a dual system of ultimate constitutional interpretation either fully or partially and are consequently prone to a centralised view of subnational rights. (2) Today, it seems widely recognised that the under-utilisation thesis, which holds that subnational entities make little use of their constitutional space, is not generally valid but requires differentiation because their level of activism depends heavily on the issue at stake.119 Concerning the entrenchment of fundamental rights, the subnational units of this comparative analysis have—except for Spain’s autonomous communities without reformed statutes and Canadian provinces—used their autonomy, albeit even within the same country, to a quite varying extent. In some cases, like the United States, Switzerland and Germany, these differences are linked with the time of adoption of subnational constitutions, as some of them predated the national

119 See JJ Dinan, ‘Patterns of Subnational Constitutionalism in Federal Countries’ (2008) 39 Rutgers Law Journal 837. On the under-utilisation thesis, more generally, see ch 5.1.3.

344 Fundamental Rights constitution and were therefore still unrestrained by the latter. In other cases, such as Catalonia and Quebec, autonomy within the same national constitutional limits is simply utilised to a greater extent than in other subnational entities because, for them, their own bill of rights is regarded as a symbol of identity and of a nationbuilding process. From a technical perspective, comprehensive self-formulated rights catalogues are not the only form of entrenchment. In several instances, such as typically the third generation of German Land constitutions and some Swiss cantons, subnational constitutions incorporate the bill of rights of the national constitution by repeating most provisions literally or simply by reference and only then go on to enshrine certain additional rights. As far as the content of subnational rights is concerned, they are obviously determined by the binding standard of protection that national constitutions set through their rights catalogues in conjunction with supremacy clauses. Subnational entities are only permitted to exceed this minimum, and they usually do this in three different ways: they may provide for completely new rights without counterparts in the national constitution, rights with such counterparts but guaranteeing more extensive protection, and finally rights with identical or only slightly changed wording. To include rights from the latter category evidently makes sense if they can be interpreted independently and differently than their national counterparts by subnational courts such as the US state supreme courts. The completely new subnational rights obviously exceed the national minimum standard most clearly. In the cases analysed, these are, first and foremost, social rights such as the rights to education, work, welfare and housing.120 This focus is linked to the generally more extensive regulation of other rights categories than social rights at the national level. In Germany, the emphasis of Land constitutions on positive rights also follows from the fact that the introduction of new negative rights is effectively precluded because they are regarded by the Federal Constitutional Court as being covered exhaustively by the Basic Law. Other recurring examples of rights without national counterparts are rights of same-sex couples to a registered partnership or even marriage. However, these are usually restricted to only some subnational entities and often face a backlash in more conservative ones. Further examples of rights that are specific to some subnational entities are language rights in Quebec and Catalonia or political rights to participate in certain instruments of direct democracy, which are typical of Swiss cantons and several US states such as California. With regard to the entrenchment of rights, the patterns of some pioneering subnational entities have often been followed by other ones and/or the national government. Emulation in the vertical direction is most obvious in the case of those early subnational rights catalogues that preceded their national counterparts and therefore had a considerable bearing on them. While the US Bill of Rights of 1791 was arguably influenced most immediately by its counterparts in the state constitutions, the rights provisions of the Swiss cantons have been an important source, first for

120 Note, however, that the Spanish Constitutional Court explicitly denied social rights guarantees the character of fully fledged individual rights and defined them as guiding principles. See box 10.H above.

Comparative Conclusions 345 the jurisprudence of the Federal Court on the recognition of unwritten fundamental rights, and then for the revised federal Constitution of 1999. An example of horizontal emulation is the role played by the 1993 Constitution of Bern and the 2006 Statute of Catalonia as points of reference for reforms in other cantons and autonomous communities. In summary, such emulation has frequently been a progressive ‘levelling-up process’ that started from the asymmetrical constellation of ‘peaks and valleys of rights protection within a nation’.121 However, this cannot conceal the fact that there is certainly no linear path of subnational entities towards extended rights protection. This is illustrated most vividly by the United States, where some state initiatives in the wake of the New Judicial Federalism have provoked staunch opposition and a conservative backlash rather than emulation. Such state constitutional counter-reactions in the same and other states were facilitated decisively by the weaker entrenchment of subnational fundamental rights compared to their national counterparts which is the result of more flexible amendment procedures. The same holds true for Canada and Germany, whereas in Switzerland constitutional amendment is relatively flexible at both the national and international levels. Interestingly, the rights enshrined in the statutes of Spain’s autonomous communities are more protected than those of subnational entities in fully fledged federations, because any amendment is required to pass the consensus-based bilateral negotiation procedure under Article 147(3) of the Spanish Constitution. Also in terms of fundamental rights the statutes therefore have more of a character of stable fundamental laws than, for example, the US state constitutions, which are used ‘as tools of policy-making to supplement or supplant ordinary lawmaking’.122

121 J Kincaid, ‘Foreword: The New Federalism Context of the New Judicial Federalism’ (1995) 26 Rutgers Law Journal 913, 946f. 122 Williams (n 12) 1125.

11 Social Welfare and Healthcare

T

HERE CAN BE little doubt that the main policy areas of the modern welfare state, ie, social welfare and healthcare,1 are among today’s major political challenges. What these two areas have in common at their core is their rationale ‘to modify the play of market forces’.2 By providing income support and generally accessible public services, they reduce individual risks like those stemming from unemployment, ill health or old age. While efforts in this direction were first made from the 1880s onwards, the welfare state was consolidated during the inter-war period and experienced its ‘golden age’ or the trentes glorieuses from 1945 until 1975.3 A retrenchment of this Keynesian welfare state then followed for various reasons such as external economic shocks, an ageing population, globalisation and the rise of neoliberalism. Under these circumstances, governments have found it increasingly difficult to ‘square the welfare circle’,4 ie to meet the rising needs of, and demands from, their people and at the same time contain their expenditures. Due to the involvement of governments at different levels, this seems to be a particularly challenging and topical issue in multilevel systems. But while research has focused quite early on the identification of different welfare regime types, ie, Bismarckian vs Beveridgean regimes5 and the distinction of the corporatist Continental, liberal Anglo-Saxon and social-democratic Nordic regimes,6 it did not initially dedicate much effort to exploring the nexus between welfare-state policies and federalism. This neglect had already been remedied in Canada in the 1980s7 and only more recently in Europe, where the topic has aroused increasing

1 Apart from these two, public education is also sometimes included among welfare-state policies. The ambiguous term ‘social welfare’ is here taken to mean income-support programmes covering social insurance and social assistance, ie payments and services to people in need. 2 A Briggs, ‘The Welfare State in a Historical Perspective’ in C Pierson and F Castles (eds), The Welfare State Reader (Cambridge, Polity Press, 2000) 18. 3 See C Pierson, Beyond the Welfare State (Cambridge, Polity Press, 1998). 4 V George and P Taylor-Gooby (eds), European Welfare Policy: Squaring the Welfare Circle (Houndmills, Macmillan, 1996). 5 This fundamental distinction marks the difference between contributory social insurance provision (Bismarckian) and tax-financed provision of welfare by the state (Beveridgean). 6 See G Esping-Andersen, The Three Worlds of Welfare Capitalism (Cambridge, Polity Press, 1990). In the meantime, a fourth type, namely the ‘familialistic’ Mediterranean model, has been added, M Ferrera, ‘The “Southern Model” of Welfare in Social Europe’ (1996) 6 Journal of European Social Policy 17. 7 See especially KG Banting, The Welfare State and Canadian Federalism (Montreal, McGill–Queen’s University Press, 1982).

Social Welfare and Healthcare 347 attention on the part of federalism scholars,8 on the one hand, and of welfare-state experts,9 on the other. The crucial importance of this research is illustrated by the fact that all three main functions of welfare-state policies generate in multilevel systems quite distinctive and arguably more complex dynamics.10 The first of these functions is redistribution, which is the instrument used to realise the above-mentioned rationale of modifying the play of market forces. The second is political legitimation, which usually goes along with welfare-state policies as normally highly visible and popular policies from people’s perspective. It is therefore only natural that national and subnational governments are often rivals in occupying the field of welfare-state policies, particularly in countries with competing nation-building projects. A third function of these policies is therefore the consolidation of identity(ies). Historically, ‘the welfare state served to “crystallize” the nation-state’,11 as it called for a greater regulatory and redistributive role of national governments and thus helped to generate an overarching sense of nationhood. Against this background, it is hardly surprising that TH Marshall’s influential social citizenship concept of 1950,12 which has been at the core of the welfare state, has the nation state as its logical frame of reference. Since the rise of New Regionalism,13 however, competing nation-building projects have been initiated by a number of subnational governments with their own welfare policies being a key element of these projects. Typically, they aim to acquire new competences in this area with references to common identity and solidarity, which they then try to reinforce precisely by using these powers. At both the subnational and national levels, the relationship between welfare-state policies, on the one hand, and identity and solidarity, on the other, is therefore by no means unidirectional; it is, rather, dialectic: ‘Not only does a sense of common identity help sustain the values of mutual help, but the welfare state itself helps foster national identity and unity’.14 For some, such competing regional nation-building has per se a negative impact on countrywide solidarity, which is clearly presupposed by any effort to defend redistribution on the basis of needs against rational self-interest and strict reciprocity.15

8 See, eg, N McEwen and L Moreno (eds), The Territorial Politics of Welfare (Abingdon, Routledge, 2005). 9 See, eg, S Leibfried et al, Federalism and the Welfare State: New World and European Experiences (Cambridge, Cambridge University Press, 2005). 10 See J Poirier, ‘Multilevel Governance and Social Policy: Observations from the Perspective of Comparative Federalism’ in Actes des Ateliers 2008 sur la Gouvernance à niveaux multiples au sein de l’Union européenne (Brussels, Comité des Régions, 2009). 11 L Moreno and N McEwen, ‘Exploring the Territorial Politics of Welfare’ in McEwen and Moreno (eds), The Territorial Politics (2005) 2. 12 TH Marshall, ‘Citizenship and Social Class’ in TH Marshall and TB Bottomore (eds), Citizenship and Social Class (London, Pluto Press, 1992). ‘Extension of social services is not primarily a means of equalising incomes. In some cases it may, in others it may not … What matters is that there is a general enrichment of the concrete substance of civilised life, a general reduction of risk and insecurity, an equalisation between the more and the less fortunate at all levels—between the healthy and the sick, the employed and the unemployed, the old and the active, the bachelor and the father of a large family’ (p 33). 13 See ch 1.2.2. 14 M Keating, Nations against the State: The New Politics of Nationalism in Quebec, Catalonia and Scotland (New York, Macmillan Press, 2001) 40. 15 See D Miller, On Nationality (Oxford, Clarendon Press, 1995) 71ff.

348 Social Welfare and Healthcare This view is closely linked to one of the main arguments made by those who see federalism as being detrimental to the welfare state, namely that multilevel systems would inevitably entail a race to the bottom. This would result from the fact that subnational governments are—compared to their usually more powerful national counterpart—more inclined to lower welfare standards in response to pressure from mobile businesses and to fiscal constraints. This is one of the main arguments used in the so-called ‘declinist debate’16 about the alleged constraining effects of federalism on welfare-state development. Another one blames it for having such effects by creating multiple veto points to delay, dilute or defeat expansionist policies. A second major discussion, the ‘variation debate’, sees multiple government levels in a negative light because divergent policies would lead to some regions being disproportionately affected by ‘disinvestment’ and others becoming ‘welfare magnets’.17 This latter issue of reconciling the conflicting goals of variation and equity presents indeed the fundamental challenge of any multilevel welfare state. It is indeed at the heart of such a state’s two crucial policies regarding each country explored in this chapter, ie social welfare (sub-section (1)) and healthcare (sub-section (2)). In principle, the balance between the goals of variation and equity is influenced by two main determinants that limit the subnational scope of action: the limiting force of a common national framework, as well as the extent of intergovernmental and/or interpersonal financial transfers, which are, due to the enormous costs of social welfare and health, arguably more crucial than in any other policy field.18 It is important to note that a standardising national framework may be established not only by programmes devised and implemented directly by the national government but also by restrictive nationally defined parameters that ensure the similarity of subnational programmes throughout the country. The second key determinant is the extent of financial transfers, which are indispensable for equal or at least comparable access to social welfare and health without having to pay considerably different taxes. These reduce the divisive effects resulting from the fact that richer regions profit by a virtuous circle of less need for social support and at the same time greater fiscal revenues, while poorer regions are caught in the converse vicious circle.19 In programmes implemented by the national government, equalising financial transfers are implicit, as they follow from the greater use of these programmes in poorer regions and sometimes from differentiated funds allocation according to the wealth of regions. Whereas these transfers are essentially interpersonal and only indirectly interregional, national financing of programmes implemented at the subnational level typically involves formal transfers.

16 KG Banting and J Costa-Font, ‘Decentralization, Welfare, and Social Citizenship in Contemporary Democracies’ (2010) 28 Environment and Planning C: Government and Policy 381, 383. 17 PE Peterson and MC Rom, Welfare Magnets: A New Case for a National Standard (Washington, The Brookings Institution, 1990) 127. 18 See KG Banting and S Corbett, ‘Health Policy and Federalism: An Introduction’ in KG Banting and S Corbett (eds), Health Policy and Federalism: A Comparative Perspective on Multi-Level Governance (Kingston, McGill–Queen’s University Press, 2002) 19. 19 See KG Banting, ‘Social Citizenship and Federalism: Is a Federal Welfare State a Contradiction in Terms?’ in S Greer (ed), Territory, Democracy and Justice: Regionalism and Federalism in Western Democracies (London, Palgrave Macmillan, 2006) 50f.

United States 349 11.1 UNITED STATES

(1) The move towards social welfare expansion in the United States is, from a historical perspective, inextricably entwined with the New Deal. Notable examples of programmes in this realm preceding the 1930s are few and, as to their scope, they were restricted to benefitting only a few particularly ‘deserving’ groups of people, ie, veterans as well as mothers and infants.20 New Deal legislation on social welfare then started in 1933 and culminated in the Social Security Act of 1935, which included a host of new programmes.21 Whereas the Supreme Court upheld several early statutes,22 it later started to issue a number of negative rulings23 so that Franklin D Roosevelt felt constrained in 1937 to propose his infamous Judicial Procedures Reform Bill (the ‘Court Packing Plan’).24 Even though this plan fell victim to widespread opposition, the Supreme Court eventually turned around and became more sympathetic to New Deal initiatives, first by affirming the constitutionality of a minimum-wage law in the state of Washington25 and then by approving its centrepieces like the above-mentioned Social Security Act.26 For almost three decades thereafter, new social welfare initiatives remained scarce. During World War II and at the outset of the Cold War, the focus of politics shifted to foreign affairs and, instead of undoing the New Deal legislation, the Republicans accepted its basic principles. New activism only surfaced with Lyndon B Johnson’s Great Society programmes, which did not only include payments for medical attendance for elderly and poor people. They also substantially improved the coverage and benefits of the Social Security programme and famously launched a ‘War on Poverty’. Most of the decades that followed the Great Society were characterised by a divided government and increasing partisan polarisation, which significantly hampered efforts to both extend social welfare and to retrench it.27 Under precisely such circumstances, however, inter-institutional bargaining between Bill Clinton and a Republican Congress 20 A Civil War pension system, for Union soldiers only, was established as early as in 1862, whereas the maternal and infant health plan under the 1921 Sheppard–Towner Act only predated the New Deal by several years. See T Skocpol, Protecting Soldiers and Mothers: The Political Origins of Social Policy in the United States (Cambridge MA, Harvard University Press, 1992). 21 In concrete terms, these were Old Age Insurance, later expanded into Old Age, Survivors and Disability Insurance; Old Age Assistance and Aid to the Blind; Unemployment Insurance; as well as Aid to Dependent Children. 22 Home Building & Loan Association v Blaisdell 290 US 398 (1934); Nebbia v New York 291 US 502 (1934); Norman v Baltimore & Ohio Railroad Co 294 US 240 (1935). Interestingly, the first two cases were—unlike most New Deal initiatives—not about national laws, but about legislative projects on the part of single states, ie Minnesota and New York, respectively. 23 ALA Schechter Poultry Corp v United States 295 US 495 (1935); United States v Butler 297 US 1 (1936); Carter v Carter Coal Co 298 US 238 (1936). 24 The idea was to nominate one new justice for each one over the age of 70, which would have enabled Roosevelt to alter the composition of the Court in his favour. See WE Leuchtenburg, The Supreme Court Reborn: The Constitutional Revolution in the Age of Roosevelt (New York, Oxford University Press, 1995). 25 West Coast Hotel Co v Parrish 300 US 379 (1937). 26 Steward Machine Co v Davis 301 US 548 (1937); Helvering v Davis 301 US 619 (1937). On the latter case, see also box 7.J. 27 See M Tushnet, The Constitution of the United States of America: A Contextual Analysis (Oxford, Hart Publishing, 2009) 37f.

350 Social Welfare and Healthcare led in the mid-1990s to a comprehensive welfare reform that marked another milestone in the development of US social policy. The above-mentioned historical changes towards expansion or retrenchment occurred against the background of a Constitution that, back in 1787, unsurprisingly failed to allocate competences regarding the then-private matter of social welfare. As responsibilities were also not assigned by later formal constitutional amendments, the distribution of powers in this area is essentially ‘what the Supreme Court says it is’.28 Until the early twentieth century, case law still did not perceive social welfare as a field of governmental action and generally ruled legislation on such issues unconstitutional, be it from the national Government or from the states.29 This has, of course, changed in the course of time. As for the national Government, the two above-mentioned Supreme Court decisions of 1937, which upheld the Social Security Act, clearly paved the way for legislative action in this area. Case law played an equally decisive role at later critical junctures of policymaking in this area. The Great Society programmes, for instance, benefitted enormously from the Warren Court’s expansive interpretation of federal power.30 Another federal lever to determine social welfare policy is even more powerful than direct legislation. Today’s widespread use of conditional grants to states emerged in the wake of the Sixteenth Amendment of 1913. This amendment authorised the national Government to levy an income tax without being subject to the constitutional obligation regarding ‘direct taxes’ to apportion the revenue among the states on the basis of their respective populations (Article I, section 2, clause 3 of the US Constitution). In 1923, at a time when the Supreme Court still disapproved of many national initiatives that were based on the residual clause of the Tenth Amendment, the judges explicitly allowed the practice of conditional grants in Massachusetts v Mellon.31 In general, subnational governments cannot afford to deny such grants and the strings attached for two main reasons: the absence of a formal fiscal equalisation system and the presence of balanced-budget requirements in nearly all state constitutions, which force them to procyclically cut benefits during economic downturns.32 Beyond the limitations to state governments’ autonomy through national legislation and conditional grants, they are not very restricted by the Constitution itself. The latter only enforces equity to some extent by requiring, for instance, that subsistence allowances and emergency medical services be provided in an equal manner also to newcomers to a state’s territory.33 Interestingly, this equal access to the 28 K Finegold, ‘The United States: Federalism and its Counterfactuals’ in H Obinger et al (eds), ederalism and the Welfare State: New World and European Experiences (Cambridge, Cambridge F University Press, 2005) 176. 29 See DB Robertson, ‘The Bias of American Federalism: The Limits of Welfare-State Development in the Progressive Era’ (1989) 1 Journal of Policy History 261, 277ff. 30 See LA Powe, The Warren Court and American Politics (Cambridge MA, Harvard University Press, 2000) 445ff. 31 Massachusetts v Mellon 262 US 447 (1923). 32 See SL Greer, ‘How Does Decentralisation Affect the Welfare State? Territorial Politics and the Welfare State in the UK and US’ (2010) 39 Journal of Social Policy 181, 193. 33 Shapiro v Thompson 394 US 618 (1969) and Memorial Hospital v Maricopa County 415 US 250 (1974).

United States 351 rovision of ‘the basic necessities of life’34 is not derived from the argument of counp trywide solidarity, but from the right to unimpeded freedom of movement. Against a thus primarily judge-made distribution of powers, the respective roles of the national Government and the states concerning social welfare have been subject to radical changes over time. For example, Lyndon B Johnson’s ‘War on Poverty’ simply bypassed—potentially obstructive—states in order to channel grants in exchange for the administration of federal programmes directly to other beneficiaries. These were, above all in deprived rural areas, the local governments (‘direct federalism’) and in cities often community groups (‘private federalism’).35 Only a few years later, Richard Nixon’s vision of federalism again turned around intergovernmental power relations. His administration granted the states, through General Revenue Sharing enacted in 1972, albeit only temporarily, more financial autonomy,36 which could be used in the area of social welfare. Beyond legislative and fiscal levers, national governments have also relied on another, quite particular, instrument to shape their relations with the states. The widespread use of waivers, ie provisions allowing states to obtain exemptions from obligations enshrined in federal programmes, was used by the Reagan administration as a means to bring about reforms of the social welfare systems in the states, as long as these were in line with the vision of the national Government.37 Much more importantly, waivers later paved the way for the comprehensive countrywide welfare reform that culminated in the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) of 1996. This piece of legislation remodelled a number of existing welfare programmes in a manner that emphasised a ‘workfare’ approach and reduced access to benefits, for instance, for legally resident immigrants.38 Another characteristic of PRWORA was, however, new leeway for the states to devise their own policies even without having to obtain waivers. This is why this welfare reform came to be known as a ‘devolution revolution’,39 even though both the devolutionary and revolutionary character of this reform are sometimes called into question.40 Today, the degree of state autonomy and, as a consequence, of potential variation ultimately depends on the design of each programme. For instance, benefits of the Social Security programme for retired and disabled workers have been, since 1935,

34

Memorial Hospital, ibid 259. these two forms of federalism, see, respectively, DJ Elazar, American Federalism: A View from the States, 2nd edn (New York, Crowell, 1972) 80; JW Fesler, Public Administration: Theory and Practice (Englewood Cliffs, Prentice Hall, 1980) 292. 36 The state portion of the General Revenue Funds was already abolished in 1980, the local governments’ portion in 1986. See RW Rafuse Jr , ‘Fiscal Federalism in 1986: The Spotlight Continues to Swing toward the States and Local Governments’ (1987) 17 Publius 35. 37 On the use of waivers by Presidents Ronald Reagan, George HW Bush and Bill Clinton, see Finegold, ‘The United States’ (2005) 170ff. 38 This was the first time in US history that this group became ineligible for most benefits. See M Fix and JS Passel, ‘Assessing Welfare Reform’s Immigrant Provisions’ in A Weil and K Finegold (eds), Welfare Reform: The Next Act (Washington, The Urban Institute Press, 2002). 39 RP Nathan, ‘The ‘Devolution Revolution’—An Overview’ (1996) Rockefeller Institute Bulletin 5. 40 See J Kincaid, ‘Devolution in the United States: Rhetoric and Reality’ in K Nicolaïdis and R Howse (eds), The Federal Vision: Legitimacy and Levels of Governance in the United States and the European Union (Oxford, Oxford University Press, 2001). 35 For

352 Social Welfare and Healthcare exclusively federal with no role for the states in either financing or administration. Another pattern is that states, more precisely only some states, supplement benefits provided by the national Government.41 The Food Stamps Program is a good example of this and of states acquiring more and more influence over time. Although eligibility and benefit levels are basically regulated nationally, the states are responsible for implementation and they are by virtue of the PRWORA allowed to reduce or extend eligibility. Beyond this, Temporary Assistance for Needy Families (TANF) even allows states to establish their own eligibility and benefit criteria, which is also true for Unemployment Insurance42 and General Assistance programmes. But while with TANF the states manage a federal block, they partially or fully finance the latter two instruments, which make them, not least due to the scarce financial resources of many states, chronically underfunded. Altogether, the national Government has undeniably dominated social welfare policy, first and foremost as a result of its superior fiscal position. This does not mean, however, that the states have been irrelevant in this process. They served, for instance, in several cases as sources of innovation. Wisconsin had already introduced an unemployment insurance scheme in 1932, which was then largely emulated by the federal Social Security Act three years later. The same state much later again became, with its ‘welfare-to-work’ approach, a model for the national welfare reform of the 1990s.43 The more pronounced legislative activism of certain states, directed either towards expansion or retrenchment, also explains the considerable variations between the states regarding single programmes, which above all illustrates the distinctly more expansionist approach in Northern states, for instance with regard to General Assistance programmes.44 Whereas the New Deal had still left the paternalistic labour system and racial discrimination in the South unchallenged,45 the social welfare legislation of the Great Society was inseparably and explicitly linked with the adoption of the civil rights legislation of the same years, 1964–65. Insofar as US welfare policy has had a nation-building function in respect of socially integrating the Southern states and African Americans living there, it has largely only done so since the 1960s.46

41 This is the case, for example, regarding the Supplemental Security Income (SSI) for elderly, blind and disabled people and concerning the Earned Income Tax Credit (EITC), which reduces federal income tax liability. 42 In this regard, the Unemployment Insurance programme stands out in comparison with the other programmes created with the Social Security Act of 1935, which were all decidedly more centralised. 43 See LM Mead, Government Matters: Welfare Reform in Wisconsin (Princeton, Princeton University Press, 2004). 44 See Finegold (n 28) 142 and 159. 45 Interestingly, Congressional committees with (often Democratic) chairmen from the South succeeded in deleting a provision from Roosevelt’s initial bill that would have obliged states in return for receiving federal grants to provide ‘reasonable subsistence compatible with decency and health’. Instead, benefits in the South were much lower than in the North and higher for white people than for African Americans. See W Cohen, ‘The Social Security Act of 1935: Reflections Fifty Years Later’ in A Pifer and F Chisman (eds), The Report of the Committee on Economic Security, 1935 (Washington, National Conference on Social Welfare, 1985) 9. 46 By contrast, African Americans moving to Northern states had already benefitted from the New Deal, which over time contributed greatly to changing their affiliation from Abraham Lincoln’s party to the Democrats.

United States 353 On the whole, the relationship between US federalism and social policy seems to be much more complex than the standard argument of federalism as the main inhibiting factor for welfare-state development suggests.47 First, observers have cited a number of other reasons for limited social protection in the United States such as the traditionally weak labour movement, the mobility of the working class, racial divisions within the group of potential beneficiaries and the predominance of two parties, which cannot both be termed welfarist according to European standards.48 Second, even if it is true that federalism has in some cases incited a race to the bottom because states have sought to make themselves more attractive to companies and individuals by lowering taxes and benefits,49 it triggered welfare expansion in other instances. More concretely, it has done so by offering an alternative to the often politically unfeasible option of extending programmes throughout the country. Thus, the delegation of social policy-making to the states in a situation of deadlock at the national level resulted in expansion, for instance, in the 1930s regarding the introduction of several New Deal programmes like Unemployment Insurance and again in the 1990s concerning the health insurance coverage of children from lowincome families through the State Children’s Health Insurance Program (SCHIP).50 (2) Much of the above concerning social welfare equally applies to healthcare. In the absence of an explicit distribution of powers in written constitutional law, Supreme Court decisions and intergovernmental relations have assumed outstanding importance. In these circumstances, the adoption in 1965 of Medicare for the elderly and disabled, as well as Medicaid for the poor, within the framework of Lyndon B Johnson’s Great Society was certainly a milestone because it extended healthcare benefits to at least a substantial part of the population. Even though only Medicare is a centralised federal programme, the national Government also wields influence over Medicaid, which is regulated, within the limits of certain national rules regarding the coverage of mandatory groups, by the states and characterised by marked interstate variations concerning benefit levels.51 This influence is based on the strings attached to the federal matching funds that the states rely on to a greater or lesser extent.52 In particular, states in times of recession or those with structurally scarce fiscal resources53 very much depend on this financial support and are, from an economic perspective, hardly able to refuse these funds. In comparison to Medicaid, restrictions on the states through conditional funding are less severe concerning the SCHIP, which was created in 1997 to overcome 47 For this argument, see especially N Glazer, ‘The American Welfare State: Exceptional no More?’ in H Cavanna (ed), Challenges to the Welfare States: Internal and External Dynamics for Change (Cheltenham, Edward Elgar, 1998) 10. 48 For an overview of these alternative reasons, see Finegold (n 28) 139 and 148. 49 See PE Peterson, The Price of Federalism (Washington, Brookings Institution Press, 1995). 50 See Finegold (n 28) 145 and 166. 51 On these variations, see Banting and Corbett, ‘Health Policy’ (2002) 24. 52 See LA Baker, ‘Constitutional Federal Spending and States’ Rights’ (2001) 574 Annals of the American Academy of Political and Social Science 104. 53 Medicaid costs are shared between the federal Government and the states, albeit with different individual shares. While some richer state governments contribute 50% (eg, New York), others cover only 25% (eg, Mississippi).

354 Social Welfare and Healthcare a stalemate at the national level and as a surrogate for Bill Clinton’s unsuccessful plan for universal healthcare. Whereas Clinton’s proposal foresaw a crucial role for the states as to the individual organisation of their healthcare systems, it failed to take up, in its design, the innovative experiences of the states, which had already implemented such systems.54 Instead of universal healthcare, the SCHIP was aimed at covering at least children not eligible for Medicaid and granted states remarkable flexibility, together with substantial funding. While some states decided to do little more than comply with the not very harsh conditions, others used this freedom to experiment and to also cover children’s parents or even unrelated adults with diverse combinations of services.55 After democratic presidents had, since Harry Truman, repeatedly failed to introduce universal healthcare, the administration of Barack Obama made another effort in this direction. At its core, the Patient Protection and Affordable Care Act (PPACA) of 2010 aimed to both keep in check the drastically soaring costs of healthcare provision and to incrementally reduce the rather varying number—from state to state— of uninsured people.56 Even if Republicans in Congress finally left the coalition that passed the PPACA legislation, the latter very much reflects their preference for a design that shifts implementation to a complex network that includes state governors and legislatures, regulatory agencies and the Department of Health and Human Services.57 This design has, in the end, enabled states dominated by Republicans to contest the PPACA and to impede its implementation.58 This is less true for the part concerning the regulation and oversight of the private health insurance market in order to decrease insurance rates, but very much so for the remainder of the legislation. A second pillar of the PPACA was the creation of health insurance exchanges for people to be able to purchase high-quality healthcare coverage at affordable prices. As roughly half of the states, nearly all under Republican control, refused to set up these exchanges, these had to be enacted, as foreseen by the PPACA for such a situation, by the Department of Health and Human Services, which did not have sufficient funding for this task. Moreover, the negotiations between this department and the states about the rules of these exchanges, such as standards for essential healthcare benefits, proved, in a climate of partisan polarisation, to be extremely difficult. The third pillar of the PPACA, the expansion of Medicaid, likewise created a dispute between the national Government and several states that had to be decided by the Supreme Court.

54 For further analysis, see JS Hacker, ‘Learning from Defeat? Political Analysis and the Failure of Health Care Reform in the United States’ (2001) 31 British Journal of Political Science 61. 55 See Greer, ‘How Does Decentralisation’ (2010) 194f. 56 At the time when the PPACA was adopted, the number of people without health insurance was above 50 million, see D Béland et al, ‘Implementing Health Care Reform in the United States: Intergovernmental Politics and the Dilemmas of Institutional Design’ (2014) 116 Health Policy 51, 51. Among the states, the share of people under the age of 65 without health insurance ranged in 2007 from 11% in Minnesota to 30% in Texas, see Commission on a High Performance Health System, Aiming Higher: Results from a State Scorecard on Health System Performance (New York, Commonwealth Fund, 2007) 22. 57 By contrast, more radical reform proposals like a single-payer system, which are popular with at least parts of the Democratic Party, have been sidelined. See LR Jacobs and T Skocpol, Health Care Reform and American Politics (New York, Oxford University Press, 2012). 58 See Béland et al, ‘Implementing Health’ (2014) 51ff.

Canada 355

11.A United States 2012: Medicaid Expansion against (many) States’ Resistance? Twenty-six states, several private citizens and the National Federation of Independent Business challenged the constitutionality of two elements of the PPACA. The first one was the individual mandate that required nearly all individuals legally residing in the US who are not exempt and not covered by health insurance through an employer or a government programme to buy insurance from a private company for ‘minimum essential coverage’ (Section 1501(b) PPACA). Non-compliance with this mandate entailed a ‘shared responsibility payment’, to be made with the individual’s taxes to the US Internal Revenue Service. The second element was Medicaid expansion, which obliged state programmes to cover not only certain specific groups but all adults whose income does not exceed 133 per cent of the federal poverty level. While the PPACA generally foresaw an increase in federal funding to compensate for the rising costs, states opposed to expansion may be deprived by the Secretary of Health and Human Services of their existing federal Medicaid funds. As to the individual mandate, the Court rejected its justification with the Commerce Clause (Article I, section 8, clause 3 of the US Constitution) because the mandate does not, as the established case law on this provision requires,59 regulate an existing commercial activity but forces people to undertake such an activity, that of purchasing insurance.60 The judges, however, accepted the federal Government’s alternative argument, which referred to the congressional power to ‘lay and collect taxes’ (Article I, section 8, clause 1).61 They found that the ‘shared responsibility payment’ is not a penalty for unlawful conduct but may be indeed qualified as a tax under this provision because it is not so high that it absolutely compels individuals to buy health insurance and it is collected by the Internal Revenue Service through the normal taxation procedure. In contrast to the individual mandate, the Court struck down the part of Medicaid expansion that threatened states in case of non-compliance with the loss of existing Medicaid funding, typically amounting to more than 10 per cent of a state’s total budget. Even if the Spending Clause (Article I, section 8, clause 1) authorises Congress to set up federal–state programmes, this provision requires states to accept the terms of these programmes on a voluntary basis. Such a sharp cutback in funding, by contrast, would leave them no choice. The Court also rejected the argument that the expansion would fall within the Congress’s reservation in the Social Security Act, which includes the original Medicaid legislation, to modify the programme. The expansion would have to be qualified rather as a dramatic change in kind from covering only specific groups to including all people on low incomes. 11.2 CANADA

(1) Back in 1867, the architects of the Canadian Constitution did not contemplate that governments, be they national or provincial, would ever provide social welfare to individuals. At that time, any support for disadvantaged people was still the domain of private actors such as churches and local charitable societies.62 Only 59

See box 5.G. National Federation of Independent Business v Sebelius 567 US ___ (2012). 61 See chs 7.2.2 and 7.3.2. 62 See AM Linden, ‘Flexible Federalism: The Canadian Way’ in J Fedtke and BS Markesinis (eds), Patterns of Regionalism and Federalism: Lessons for the UK (Oxford, Hart Publishing, 2006) 51. 60

356 Social Welfare and Healthcare during the Great Depression did it become widely accepted that a comprehensive effort on the part of public authorities in this area would be absolutely necessary. With poorer provinces having the least financial resources for such efforts and richer provinces not willing to become ‘welfare magnets’, the national government was seen as the logical provider of social welfare. However, its unemployment insurance plan of 1935 was struck down as unconstitutional by the Judicial Committee of the Privy Council (JCPC),63 like other acts of the ‘Canadian New Deal’, because the issue would fall within provincial jurisdiction over ‘property and civil rights in the provinces’ (Section 92(13) of the 1867 Constitution Act). Similar problems emerged in respect of old-age pensions, for which the national Government first co-financed provincial programmes in the late 1920s, but then intended to create its own scheme. As with unemployment insurance, the crucial point of the matter, which made the JCPC invoke Section 92(13), was again the plan to co-fund the programme with compulsory premiums paid by employers and employees. It is important to note here that a legislative design relying fully on the tax revenues of the national Government would not have created constitutional problems. It would have been covered by its spending power, which is not restricted to areas of legislative competence.64 A programme including compulsory contributions, however, required in light of the aforementioned JCPC judgments a constitutional amendment. These were eventually effected by adding in 1940 an exclusive national power over ‘unemployment insurance’ (Section 91(2A)) and in 1951 a c oncurrent competence with provincial precedence over ‘old age pensions’ (Section 94A) in 1951.65 The Constitution Act 1982 then proclaimed a formal commitment on the part of both government levels to social security (Section 36(1)) and of the national Government to ‘the principle of making equalization payments to ensure that provincial governments have sufficient revenues to provide reasonably comparable levels of public services at reasonably comparable levels of taxation’ (Section 36(2)). It is still an open question and is highly contested among scholars whether this provision is justiciable.66 In spite of these constitutional amendments, the precise delineation of legislative powers has, in several instances, remained unclear. This problem is arguably exacerbated by the tendency of the Supreme Court to interpret subject matters under the ‘living tree’ and the ‘double aspect’ doctrines as falling within provincial jurisdiction in one aspect and national jurisdiction in another.67 These doctrines, which have been criticised above all in Quebec for covertly expanding federal power,68 play out in particular in many areas of the complex subject matter

63

Canada (AG) v Ontario (AG) [1937] AC 355 (JCPC). See ch 7.3.2. 65 A further amendment in 1964 then extended Sec 94A to also include ‘supplementary benefits, including survivors’ and disability benefits irrespective of age’. 66 In favour of enforceability in court, see A Nader, ‘Providing Essential Services: Canada’s Constitutional Commitment Under Section 36’ (1996) 19 Dalhousie Law Journal 306, 349; and rather against it, PW Hogg, Constitutional Law of Canada, 5th edn (Scarborough, Carswell, 2000) 6-10 and 33-3. 67 See box 5.I. 68 For a very critical stance, see the opinion of Justice Jean Beetz: ‘The reason for this caution is the extremely broad wording of the exclusive legislative powers listed in ss. 91 and 92 of the Constitution 64

Canada 357 of social welfare. For instance, there has been a controversial legal dispute between Quebec and the national Government regarding the competence for maternity and parental benefits. 11.B Canada 2005: Maternity and Parental Benefits—a Provincial or Federal Prerogative? In order to finance its own programme of maternity and parental benefits, Quebec pressed the federal Government to reduce the contribution of Quebec employees and employers to the national employment insurance scheme. After negotiations had proven fruitless, the provincial Government started to pursue its plan unilaterally and referred the issue of constitutionality to the Quebec Court of Appeal, which found that the above-mentioned federal jurisdiction over unemployment insurance (Section 91(2A)) did not include maternity and parental benefits because the scope of this competence is limited to a loss of income due to job loss for economic reasons. It would not cover, therefore, mere interruption of employment for personal reasons. The Supreme Court held, however, that the disputed federal maternity benefits linked to pregnancy and actual birth and parental benefits related to care for newborns do fall within Section 91(2A).69 The judges recognised the difference between these benefits and unemployment allowances in the stricter sense, as they would not require an inability to work and do give parents some choice as to whether and when they claim these benefits. Yet, the Court ruled that the term unemployment insurance would have to be construed according to the ‘living tree’ doctrine progressively in the context of its (broader) meaning today. It emphasised that the federal Employment Insurance Act did not regulate maternity or parental leave or job security, which would indeed encroach upon provincial powers, but granted maternity and parental benefits with the primary aim of providing temporary income replacement during child-related interruptions in employment. The fact that these federal benefits effectively allowed parents to go on leave was only a secondary effect and not regarded as an overstepping of competences.

Whereas Quebec sought to limit national unemployment legislation in court, it was absolutely free, by virtue of the above-mentioned concurrent competence with provincial precedence, to devise its own programme regarding old-age pensions. After early national welfare-state policies had encountered the resistance of conservative Quebec governments that aimed at preserving the role of the Catholic Church in this area,70 the turn in the 1960s through the Quiet Revolution towards a more progressive nationalism entailed a change of strategy. The province would now proactively pursue its own welfare-state policies. To be sure, the Quebec Pension Plan (QPP) of 1965 (Régie des rentes), which covers people employed in the province, is very

Act 1867 and the risk that these two fields of exclusive powers will be combined into a single more or less concurrent field of powers governed solely by the rule of paramountcy of federal legislation’ in Bell Canada v Quebec (Commission de la santé et de la sécurité du travail) [1988] 1 SCR 749. 69

Reference Re Employment Insurance Act, ss 22 and 23 [2005] 2 SCR 669. D Marshall, ‘Nationalisme et politiques sociales au Québec depuis 1867: un siècle de rendezvous manqués entre l'État, l'Église et les familles’ (1994) 9 British Journal of Canadian Studies 301. 70 See

358 Social Welfare and Healthcare similar to the Canada Pension Plan (CPP) of the same year so as to avoid obstacles to countrywide mobility.71 But on the other hand, the provincial Government had already wielded considerable influence over the creation of the CPP. Moreover, it has been able to use the surpluses of its own programme to stimulate its economy.72 Whereas Quebec is unique in having its own pension plan, the area of social assistance is characterised by the actions of the other provinces as well. Even though provincial payments and services clearly dominate this field, the national Government has also introduced, by virtue of its spending power, some direct money transfers to people in need such as childcare benefits. Beyond these few social assistance programmes, it has gained influence above all through sharedcost programmes, ie the co-funding of provincial programmes, whose lawfulness is equally based on the national Government’s spending power.73 For example, the Canada Assistance Plan (CAP) of 1966 offered to cover half of the provinces’ costs, so-called ‘50-cent dollars’, but in return demanded compliance with several conditions. When rising costs prompted the introduction in 1990 of a ‘cap on CAP’ concerning Ontario, Alberta and British Columbia, the latter province challenged this measure. The Supreme Court, however, confirmed the authority of the national Government to make conditional grants in fields of provincial jurisdiction and to modify them unilaterally.74 In 1996, CAP was replaced by the Canada Health and Social Transfer, which has since been split up into the Canada Social Transfer (CST) and the Canada Health Transfer (CHT), which function as block grants in a completely different way. Instead of covering 50 per cent of the actual programme costs, these transfers depend on the population and gross domestic product of each province. Even though the national Government relaxed conditions for the CST,75 the overall effect of the reform on the provinces was rather negative because the new funding formula reduced federal co-financing to, on average, roughly 10 per cent.76 Thus, while CAP was mainly dismissed by the national Government as creating disincentives to responsible-minded spending on the part of the provinces, its abolishment was also a strategy of welfare retrenchment.77 Through unilateral decentralisation, the provinces were less limited by conditions, but due to having fewer financial resources, they were also hardly able to make use of this greater autonomy. 71 The two plans are coordinated insofar as benefits are transportable, payroll tax and replacement rates identical and both are financed exclusively through contributions of employers and employees. 72 See DC Thomson, Jean Lesage and the Quiet Revolution (Toronto, Macmillan, 1984) 185ff. 73 According to Banting, this shared-cost federalism was as a prevalent pattern preceded by classical federalism with modifications of competences through constitutional amendments and followed after by joint-decision federalism with particular regard to Quebec. See KG Banting, ‘Canada: Nation-building in a Federal Welfare State’ in Obinger et al (eds), Federalism (2005) 95. 74 Reference Re Canada Assistance Plan (BC) [1991] 2 SCR 525. See box 8.F. 75 The CST maintained only one of CAP’s four initial conditions, namely the ban on minimum residency requirements. 76 Parliamentary Budget Office, ‘Fiscal Sustainability Report 2011’ (2011) 16, www.parl.gc.ca/pbodpb/documents/FSR_2011.pdf. 77 Other strategies of retrenchment that had started in the late 1980s and intensified in the second half of the 1990s have been the target of previously universal benefits (eg, family allowances) and the introduction of stricter eligibility criteria (eg, employment insurance), see D Béland and A Lecours, ‘Sub-State Nationalism and the Welfare State: Québec and Canadian Federalism’ (2006) 12 Nations and Nationalism 77, 86f.

Canada 359 The retrenchment period of the late twentieth century was accompanied by the intergovernmental Social Union Framework Agreement (SUFA) of 1999, which the national Government regarded as an initiative for more coordination and effectiveness of welfare-state policies. For critical observers, SUFA meant ‘collaboration of rowing agents who follow the indications of a steering principal’, namely ‘the prime minister’s “court”’,78 and for the Parti Québécois it was even a primary argument for secession. In the end, Quebec opted to abstain from the agreement because it considered a coordinating body for social security matters as a mechanism of federal intrusion. Decades after the first steps towards asymmetry concerning social welfare, ie, the opt-out from the CPP, Quebec’s non-participation in the SUFA framework exemplifies what has aptly been termed ‘9-1-1 federalism’.79 Interestingly, this differentiation, owed to the utilisation of welfare-state policies for two competing nation-building projects, did not result in a race to the bottom. Instead, levels of social protection in Quebec are usually equal to those in the rest of the country and in some cases even higher, such as universal day care for children or parental leave.80 In light of the more egalitarian and collectivist leanings, which are in Quebec seen as a core element of its specific identity, more extensive social welfare in this province is hardly surprising. More contested is the impact of asymmetry on welfare-state development in Canada as a whole. Some claim that it would have increased levels of social protection because of the Canadian and Quebec governments ‘outbidding’ each other, at least in times of budget surpluses, and due to the potential of Quebec’s innovative policies, such as day care for children, to serve as points of reference for advocates of similar programmes in other provinces.81 Others contend that these dynamics of diffusion have remained weak and that supporters of more extensive social welfare throughout Canada have lost, through the abstention of the reliably welfarist Quebec Government from certain national programmes, a powerful ally.82 (2) Just like social security, healthcare did not figure prominently in negotiations about the distribution of powers in 1867 because it was equally considered as a domain of the private sector. The only clear and explicit reference to this area is the assignment of jurisdiction over the ‘Management of Hospitals, Asylums, Charities’ to the provinces (Section 92(7) of the 1867 Constitution Act). Two other less obvious heads of powers have been interpreted to confer upon them a more extensive authority regarding healthcare. The provinces’ competence regarding ‘property and civil rights in the province’ (Section 92(13)) has empowered them 78 A Noél, ‘Without Québec: Collaborative Federalism with a Footnote’ (2000) 1 IRPP—Policy Matters 1, 11. 79 H Bakvis and G Skogstad, ‘Canadian Federalism: Performance, Effectiveness and Legitimacy’ in H Bakvis and G Skogstad (ed), Canadian Federalism: Performance, Effectiveness and Legitimacy (Oxford, Oxford University Press, 2001) 12. 80 See D Béland and A Lecours, ‘Does Nationalism Trigger Welfare State Disintegration? Social Policy and Territorial Mobilization in Belgium and Canada’ (2010) 28 Environment and Planning C: Government and Policy 420, 425f. 81 ibid 426f. 82 See Banting and Costa-Font, ‘Decentralization’ (2010) 384f.

360 Social Welfare and Healthcare to regulate such diverse issues as the provision of both public and private health insurance, healthcare professions, drug standards and, as part of labour relations, also employment-related h ealthcare benefits regarding most businesses.83 A second provision is provincial power concerning ‘all matters of merely local or private nature in the province’ (Section 92(16)), which the Supreme Court has invoked regarding several healthcare issues.84 The national Government, on the other hand, has authority in relation to Canadawide health issues, either because they are of national concern or are emergency cases.85 In conclusion, the constitutional distribution of powers entrusts the provinces with substantial responsibility, but also leaves ample scope for national action. This finding is epitomised in the Supreme Court’s characterisation of healthcare as an ‘amorphous topic’, whose assignment to one of the two government levels depends in each single case on the nature or scope of the specific issue at stake.86 Just like in the area of social welfare, the spending power has a considerable impact on actual intergovernmental power relations. Even though provincial legislatures adopt their own publicly funded healthcare insurance plans that regulate the services of hospitals and physicians, as well as certain other medical services, most Canadians think of Medicare as a national programme.87 This is because the federal Government makes its contribution to the funding of the provincial plans dependent on their fulfilment of five criteria specified in the 1985 Canada Health Act,88 which is a lawful use of its spending power.89 Although this results in rather uniform standards throughout the country today, the provinces have been instrumental in the development of public healthcare. Saskatchewan was the pioneering province, which introduced, for the first time in Canada, a system of hospital insurance in 1947 and Medicare in 1961.90 Quebec was then, after its transformation by the Quiet Revolution into a progressive driving force behind welfare-state policies, one of the key actors in extending these programmes throughout the country. To bring this province into the fold was actually one of the reasons why the national Government established Medicare in 1968 as a quite generous shared-cost programme that covered, like the above-mentioned CAP, 50 per cent of provincial programme costs. Also similar to CAP, the programme was later modified to be based on population and gross domestic product, which caused a decline of national funding from 50 percent to on average approximately 19 per cent.91 Particularly in this context, 83 Such benefits are only regulated by the national Government for some specific businesses over whose labour relations it has been explicitly assigned jurisdiction. See Hogg, Constitutional Law (2000) 32-33. 84 Schneider v British Columbia [1982] 2 SCR 112. 85 Toronto (City of) Electric Commissioners v Snider [1925] AC 396 (JCPC) 412. 86 Schneider (n 84) 142. 87 See Hogg (n 66) 32-34. 88 These criteria are public administration, comprehensiveness (coverage of all ‘medically necessary’ services of hospitals and physicians), universality (coverage of all residents of the province on the same terms), portability and accessibility (service provision free of charge). 89 Eldridge v British Columbia (AG) [1997] 2 SCR 624, para 25. 90 See Linden, ‘Flexible Federalism’ (2006) 50f. 91 Parliamentary Budget Office, ‘Renewing the Canada Health Transfer: Implications for Federal and Provincial-Territorial Fiscal Sustainability’ (2012) 2, www.parl.gc.ca/pbo-dpb/documents/Renewing_CHT.pdf.

Belgium 361 Quebec started, in 2001, to raise the issue of ‘fiscal imbalance’ between the government levels. In 2004, it achieved a special role insofar as the national Government agreed to sign a bilateral accord on healthcare funding that exempted the province with full financial compensation from both the specific priorities and accountability measures obliging all other provinces. From a broader perspective, this bilateral agreement is remarkable because it is the first official document to explicitly invoke the concept of asymmetrical federalism.92 11.3 BELGIUM

(1) Social welfare in Belgium is, albeit to a decreasing degree, still a centralised cornerstone of a highly decentralised state. As in other countries, the matter was, at its earliest stages during the interwar period, a clearly national domain. Back in those times, a model emerged that featured national employers’ associations and labour unions even more than elected politicians as the key actors in designing and implementing social insurance schemes. This unitary and strongly ‘Bismarckian’ imprint remained dominant after World War II, as fragmentation, not along linguistic or territorial lines, but along the three pillars of Belgian society, ie Catholic, socialist and liberal, rendered comprehensive reform impossible.93 Instead, pre-war structures continued to manage the four big ‘social security’ schemes of the post-war period, ie family allowances, old-age pension, unemployment and health insurance. While the so-called progress phase of expansion until 1975 witnessed a gradual shift from hitherto employment-related social welfare to a universalistic approach, it only featured few and rather marginal debates about the question of decentralisation.94 This changed radically with the state reform of 1980 when the first steps were taken towards the decentralisation of social welfare. Henceforward, the national Government would remain in charge of ‘social security’, ie essentially the four abovementioned big social insurance schemes, while the communities would acquire certain competences regarding ‘social assistance’ to individuals. Even though in the latter area the national Government retained the right to set standards, for instance, concerning minimum income and measures for disabled persons,95 the communities were granted, as a result of the subsequent state reforms, considerable powers (Article 128(1) of the Belgian Constitution in conjunction with Article 5(1)(2) of the 1980 Special Law on Institutional Reforms). Today, this includes, always with

92 See J Leclair, ‘Reforming the Division of Powers in Canada: An (Un)Achievable Endeavour?’ in G Robbers (ed), Reforming Federalism—Foreign Experiences for a Reform in Germany (Frankfurt am Main, Peter Lang, 2005) 115. 93 See Béland and Lecours, ‘Does Nationalism’ (2010) 427f. 94 For instance, in the first years after World War II there was some discussion about interregional transfers of unemployment benefits and family allowances, which would have flowed, in contrast to today’s constellation, from Wallonia to Flanders. Around 1970, claims of the Flemish nationalist party Volksunie reignited a debate about the regionalisation of social welfare. See R Dandoy and P Baudewyns, ‘The Preservation of Social Security as a National Function in the Belgian Federal State’ in McEwen and Moreno (n 8) 153ff. 95 See B Cantillon et al, ‘Social Redistribution in Federalised Belgium’ (2006) 29 West European Politics 1034.

362 Social Welfare and Healthcare certain competences explicitly reserved to the national Government, important areas like social assistance for families and children, welfare aid, measures for disabled people and elderly people, as well as the protection of minors. As the regions have become responsible for issues like employment promotion or housing policy, jurisdiction over the broad area of social welfare is actually divided among the national and two subnational government levels. Overall, it is certainly true that subnational powers, above all on the part of the communities, have increased considerably. Nonetheless, the national Government still regulates the big social insurance schemes, collects their contributions and redistributes them. In sum, ‘social security’ has regularly accounted for roughly two-thirds of the national Government’s overall expenditures.96 Although the architects of the Belgian state reforms deliberately favoured the decentralisation of exclusive competences in order to avoid disputes over the respective areas of authority,97 the practice of policy-making in a complex transversal subject matter like social welfare demonstrates the limits of this approach. This holds true even though special legislation substantiates the constitutional distribution of powers by dissecting each subject matter and assigning the numerous aspects to the national Government, communities or regions. Yet, this mosaic of competences is unfit to prevent overlaps and disputes in practice.98 For instance, the family allowances, recently decentralised in 2011, are regulated and financed differently for various occupational categories and regarding civil servants still intertwined with other allowances like unemployment benefits that are still an exclusively national prerogative. In particular, the principle of exclusivity has been challenged by the Constitutional Court’s case law regarding the Flemish Care Insurance (FCI) that was introduced to cover the non-medical costs of care for elderly people. 11.C Belgium 2001–06: Complementary Flemish Care Insurance? In 1995, the federal Government extended the Allowance for Assistance to Elderly Persons programme (AAEP) to include a new system of service vouchers. In response, Flanders claimed that such vouchers would fall within the powers of the communities and started to devise the above-mentioned FCI. This insurance was eventually introduced by a decree of 1999 and involved monthly allowances for care-dependent people resident in Flanders or the Brussels-Capital Region and for people only employed there, although resident in another Member State of the European Union. It thus excluded people residing in the Walloon Region even if they were working in Flanders or Brussels. Interestingly, but in line with the bipolarism of the Belgian federation, the law concerned was not challenged

96 See P Popelier and B Cantillon, ‘Bipolar Federalism and the Social Welfare State: A Case for Shared Competences’ (2013) 43 Publius 626, 627. 97 See F Delpérée, ‘Le Fédéralisme de Confrontation à la Belge’ in J Kramer (ed), Föderalismus zwischen Integration und Sezession (Baden-Baden, Nomos, 1993) 136. 98 With specific regard to social welfare, see P Popelier, ‘Social Federalism and the Allocations of Powers in a Comparative Law Perspective: The Case for Shared Power’ in B Cantillon et al (eds), Social Federalism: The Creation of a Layered Welfare State (Cambridge, Intersentia, 2011). On this mosaic in general, see M Uyttendaele, Le fédéralisme inachevé—réflexions sur le système institutionell Belge, issu des réformes de 1988–1989 (Bruxelles, Bruylant, 1991) 631.

Belgium 363

by the federal Government but by the French Community and the Walloon Region, which led to two important judgments by the Belgian Court of Arbitration.99 The judges ruled in case no 33/2001 that the matter in question would for some aspects fall within federal jurisdiction and for others within that of the communities. As a social welfare initiative, the FCI would be constitutional because it is not divergent from, or contradictory to, the federal AAEP but merely complements it. While the Court upheld the competence of both government levels, the relationship between national and subnational legislation and precise delineation remains unclear. The line of argument in the second case, no 51/2006, opened up an even broader legal dispute involving EU law because it was based on the unequal treatment of people migrating within Belgium and from other Member States of the European Union. As this involved issues beyond the competence of the Court of Arbitration, the judges referred several preliminary questions to the European Court of Justice. The latter held that the legislation would concern, against the arguments of the Flemish Government, ‘internal situations’ only insofar as it excludes Belgian nationals who work in Flanders or the Brussels-Capital Region but live in other parts of Belgium and have never exercised their freedom of movement within the EU.100 By contrast, the right to free movement within the EU and the principle of non-discrimination (Article 18 TFEU) would indeed apply to all situations that are not purely internal. Therefore, these EU legal guarantees protect all people employed in Flanders or the Brussels-Capital Region but residing in other parts of Belgium who are citizens of other Member States or Belgian nationals having used their freedom to move within the EU before returning.101

The judgments regarding the FCI not only call into question the exclusivity of competences, but also point to another problem associated with this principle. As the national and subnational governments are allowed to have programmes with similar purposes alongside each other, they have in effect parallel powers. Moreover, they are, according to the logic of the exclusivity principle, pressured to also make use of these powers and engage, like with the FCI, in a competition of similar policies, which has been deemed to have two negative effects.102 First, these rival policies form part of a costly ‘race to the top’. Second, they are often uncoordinated, which implies a lack of intergovernmental cooperation. This lack is also visible with regard to policies that do not have a similar purpose, but are nevertheless intertwined like, for instance, job placement and unemployment insurance.103 Moreover, the FCI rulings also whet the Flemish appetite for further decentralisation. With the 2011 state reform in the aftermath of the Belgian Government crisis, 99

Belgian Constitutional Court Nos 33/2001 and 51/2006. C-212/06 Government of the French Community and Walloon Government v Flemish Government [2008] ECR I-1683. 101 See K Lenaerts, ‘The Law of the European Union and the Exercise by Regions of their Tax Powers’ in Flemish Department of Finance and Budget (ed), Fiscal Federalism in the European Union: EU Presidency Seminar (Gent, Groep De Boeck Departement Larcier Gent, 2012) 22ff. 102 See J Poirier, ‘Layered Social Federalism: From the Myth of Exclusive Competences to the Categorical Imperative of Cooperation’ in Social Federalism: How is a Multi-level Welfare State Best Organized?, e-book 9 (Re-Bel Initiative, 2011) 17, www.rethinkingbelgium.eu. With regard to the FCI and other examples, see Popelier and Cantillon, ‘Bipolar Federalism’ (2013) 640. 103 See P Popelier et al, ‘On Division of Power and the Belgian Layered Welfare State’ in Social Federalism, ibid 7f. 100 Case

364 Social Welfare and Healthcare the communities assumed, for the first time, jurisdiction over a subject matter, ie certain family allowances, that belongs, according to the definition of the Constitutional Court, to the previously strictly national realm of ‘social security’.104 The desire to decentralise this realm and not only ‘social assistance’ already arose at the end of the 1980s when first social partners and then nationalist parties in F landers started to perceive interpersonal transfers from healthy active adults to less productive groups of society as implicit interregional transfers to the less affluent Wallonia.105 Analyses of these money flows by Flemish academics, namely the ‘Club van Leuven’, were only the first of many studies contending that Walloons would have after ‘social security’ transfers more or only slightly less disposable income than Flemings.106 Since the 1990s, all Flemish parties except for the Greens have demanded the decentralisation of ‘social security’, albeit to varying extents. The fact that some of these parties have been arguing for this transfer of powers not only in terms of promoting policy innovation and better correspondence to divergent subnational preferences,107 but also see it as an intermediary step towards independence, makes Walloon parties feel uneasy about such demands. In this light, the recent decentralisation of family allowances is feared to be a watershed that will spur debates about further decentralisation of ‘social security’, even more so in times of financial crisis.108 While the course thus seems to be set for further conflict, it has to be borne in mind that old-age pensions are exceptional among ‘social security’ schemes because Flemings are net receivers and, with a more rapidly ageing population, they are increasingly so.109 This limits the interest in decentralisation of pensions and also influences debates about ‘social security’ in general.110 Another critical issue are the practical difficulties linked with the decentralisation of social insurance schemes

104 For this official definition of ‘social security’ as the traditionally national prerogative par excellence, see J Velaers, ‘Social Federalism and the Distribution of Competences in Belgium’ in Cantillon et al, Social Federalism (2011). 105 These discussions emerged in the Flemish Economic Union (Vlaams Economisch Verbond) and the Flemish Christian Mutual Insurance Company (Caritas Catholica Vlaanderen). Then they were taken up by parties like the Volksunie and the Vlaams Blok. See S Leblanc, ‘La fédéralisation de la sécurité sociale’ (1997) 1282–83 Courrier Hebdomadaire du CRISP 1, 22). 106 A Alen et al, Vlaanderen op een kruispunt: sociologische, economische, en staatsrechtelijke perspectieven (Leuven, KU Leuven, 1989). While in this first report the ‘Club van Leuven’ even claimed a higher Walloon income after transfers, more recent analyses suggest only a reduction of the income gap, albeit a quite considerable one from 19% to 10%. See B Cantillon, ‘On the Possibilities and Limitations of a Layered Social Security System in Belgium’ in Cantillon et al (n 98). 107 Whereas people in Wallonia, for instance, tend to conceive of unemployment benefits as social rights, they are seen by Flemings as conditional entitlements to incentivise people’s return to the labour market. See J Poirier and S Vansteenkiste, ‘Le débat sur la fédéralisation de la sécurité sociale en Belgique: le miroir du vouloir-vivre ensemble?’ (2000) 2 Revue Belge de Sécurité Sociale 331, 348f. 108 See W Swenden, ‘Conclusion: The Future of Belgian Federalism—Between Reform and Swansong?’ (2013) 23 Regional and Federal Studies 369, 370; K Deschouwer and M Reuchamps, ‘The Belgian Federation at a Crossroad’ (2013) 23 Regional and Federal Studies 261, 265. 109 Already in 2010, the Flemish community received a bigger share of pension expenditures (62.13%) than its share of the Belgian population (57.68%). See Cantillon, ‘On the Possibilities’ (2011) 87. 110 Only Flemish nationalist parties advocate the decentralisation of old-age pensions. But even for them the other areas of ‘social security’ are priorities. See N Bouteca et al, ‘The Future of Belgian Federalism as Seen through the Eyes of the Social Partners: A Continuing Obstacle to Social Policy Decentralization?’ (2013) 23 Regional and Federal Studies 293, 299.

Spain 365 to both the communities and regions. In the first case, the need to assign inhabitants of the Brussels Region to the scheme of either the Flemish or French Community would require them to make a thus far politically unpopular self-declaration of belonging.111 The second option would run contrary to general Flemish preferences for self-government based on the communities and create financial problems in the Brussels Region of having, at the same time, a concentration of socially vulnerable people and wealth created to a considerable extent by commuters. (2) As far as health insurance is concerned, much of the above about the decentralisation debates regarding ‘social security’ equally applies because it forms part of this realm according to the aforementioned definition of the Constitutional Court. Starting in the 1990s, much of these debates actually centred on health insurance, with Flemish claims that medical consumption would be higher in Wallonia. Recently, however, these disparities in terms of expenditures seem to have decreased, while Flemings, of course, still contribute more to the insurance scheme as a result of a higher employment rate.112 Even though health insurance has remained a national prerogative, the communities have acquired, since the state reform of 1980, and always with certain powers explicitly reserved for the national Government, responsibility for certain other healthcare issues, including nursing homes and home care, the organisation of basic healthcare, health education and preventive medicine (Article 128(1) in conjunction with Article 5(1)(1) of the 1980 Special Law on Institutional Reforms). This made it possible to some extent to design subnational policies in accordance with diverging preferences like the greater emphasis in the Flemish community on preventive medical care and on primary healthcare by general practitioners instead of special consultations.113 Yet, the decentralisation of health insurance so far still remains an unmet Flemish demand. 11.4 SPAIN

(1) In the aftermath of Spain’s transition to democracy, the country’s system of social welfare underwent fundamental changes, which were characterised overall by expansion and consolidation.114 These occurred in the context of large-scale changes, ie the massive entrance of young women and immigrants into the labour market since the 1990s, which both fuelled economic dynamism in the 1997–2008 period and had important structural repercussions for welfare-state policies.115 Back in the time 111 Political will has so far been lacking due to problems of categorising an extremely diverse population and perhaps also the fear of uncovering the overrepresentation of Flemings in the political institutions of Brussels. See Popelier and Cantillon (n 96) 636 and 644. 112 See Cantillon (n 106) 85. 113 See Bouteca et al, ‘The Future’ (2013) 297ff. 114 For an overview of the development in the post-Franco era, see G Rodríguez Cabrero, ‘The Consolidation of the Spanish Welfare State (1975–2010)’ in AM Guillén and M León (eds), The Spanish Welfare State in European Context (Farnham, Ashgate, 2011). 115 See M León and AM Guillén, ‘Conclusions’ in Guillén and León (eds), The Spanish (2011) 311.

366 Social Welfare and Healthcare of the Franco regime, social welfare had been left, except for a modest contributory scheme of social security, to the Catholic Church, charitable organisations and families. Against this backdrop, the Constitution of 1978 was faced with the challenge of distributing competences regarding a subject matter that had hitherto hardly existed. While all autonomous communities were offered an opportunity to assume, in their statutes, responsibility for social assistance (Article 148(1)(20)), their jurisdiction in this area was at the same time limited by the national Government’s exclusive jurisdiction over the ‘basic legislation and financial system of social security, without prejudice to implementation of its services by the Self-governing Communities’ (Article 149(1)(17)). Moreover, social assistance excluded such important fields as assistance for disabled people and old-age pensions. The Constitution only makes reference to these by listing them among the ‘Principles governing Economic and Social Policy’ (Article 49 and Article 50, respectively). Both areas had already been transferred in 1978 to the Instituto de Servicios Sociales (INSERSO), a quasiautonomous public body within the Spanish Ministry of Labour.116 This constitutional framework, which did not clearly delineate the competences in the realm of social welfare, enabled the autonomous communities to employ an extensive interpretation of their jurisdiction and thus to establish, from 1982 onwards, regional systems of social welfare.117 Common traits of this first wave of regional legislation in this new area were the principles of public responsibility and universalism regarding access to services, as well as the delegation of much of their provision to local governments with the autonomous communities limiting themselves to planning and coordination. This period, however, also witnessed several national government initiatives in this area. For example, the latter passed a law in 1982 regarding benefits for disabled people and made, in 1983–84, an eventually futile attempt to consolidate the, from its viewpoint, excessively fragmented social welfare provision in the National Act on Social Services. Nonetheless, it was in the end more for the Constitutional Court to impose limits on the policies of the autonomous communities. 11.D Spain 1986: Supplementary Social Assistance Initiatives of the National Government? Two resolutions of the Spanish Government had foreseen in 1983 and 1985 national grants for the financing of social assistance programmes provided by non-profit organisations and the national Government. Such a testing of the limits of competences was typical for this early period of regionalisation and soon resulted in opposition. It was the autonomous community of Galicia, which challenged these acts on the ground that ‘social assistance’ would be an exclusive power and thus rule out any action by the national Government.

116 See L Moreno and A Arriba, ‘Decentralization, Mesogovernments, and the New Logic of Welfare Provision in Spain’ (1999) Instituto de Estudios Sociales Avanzados Working Paper 1, 6. 117 This process began with the Basque Country (1982), Navarra (1983) and Madrid (1984). It ended with Cantabria, which only passed legislation in this area in 1992. See R Gallego and J Subirats, ‘Spanish and Regional Welfare Systems: Policy Innovation and Multi-level Governance’ (2012) 22 Regional and Federal Studies 269, 284.

Spain 367

The Constitutional Court held that regarding a subject matter as complex and of crucial importance in a welfare state, the meaning of exclusivity must not be construed in a narrow sense.118 The judges argued that insofar as particular social problems are identified that require a more comprehensive response than that of only one single region, the national Government is authorised to intervene. In doing so, it has to respect the competences of the autonomous community concerned, but only as far as possible (FJ 4). Moreover, the Court deemed it useful, but not mandatory, to let this region participate in the process of specifying the concrete terms of implementation of the national programme. As a result, this judgment recognised, with only minimal restrictions, that supplementary welfare initiatives on the part of the national Government may be constitutional if they respond to a social problem calling for a solution at the level above the autonomous communities (FJ 5). Notwithstanding these constraints, the autonomous communities have clearly remained the protagonists in the area of social welfare.119 The national Government, on the other hand, started after the above-mentioned failed attempt at recentralisation to rely on a strategy of intergovernmental coordination and conditional grants, with the 1987 Concerted Plan for the Development of the Basic Provision of Social Services by the Local Authorities being the most prominent example. This agreement regulated administrative cooperation between national, regional and local governments and tied national funding to the adherence of the autonomous communities to certain common standards. Even though the Basque Country finally chose not to sign the agreement, its format has proved paradigmatic for subsequent accords regarding social welfare and other policy fields as well.120 This trend towards intergovernmental coordination was greatly facilitated, and indeed epitomised, by the creation in 1988 of a Spanish Ministry of Social Affairs, which focused primarily on its role as the central point of contact for the autonomous communities. Significant policy initiatives on the part of the national Government have since then indeed remained scarce, with the 2006 Law on Dependence and Long-Term Care being a major exception.121 Otherwise, social welfare has remained a genuinely regional affair, as intergovernmental agreements paved the way for the transfer of executive authority over assistance to disabled people and old-age pensions from

118

STC 146/1986. See Moreno and Arriba, ‘Decentralization’ (1999) 8f. 120 See R Agranoff, ‘Intergovernmental Relations and Management of Asymmetry in Federal Spain’ in R Agranoff (ed), Accommodating Diversity: Asymmetry in Federal States (Baden-Baden, Nomos, 1999) 107. 121 This crucial piece of legislation was aimed at facilitating the labour-market participation of traditional care-givers, ie mainly women, and thus reflected a critical turn away from traditional ‘familialism’. Nonetheless, reliance on family support has so far remained, also due to precarious finances during the crisis, the normal case. See S Sarasa, ‘Long-term Care: The Persistence of Familialism’ in Guillén and León (n 114). 119

368 Social Welfare and Healthcare the INSERSO to the autonomous communities and did not obstruct a second wave of regional legislation regarding social welfare throughout the 1990s. Compared to the above-mentioned first wave, the new regulations, pioneered by Galicia and Catalonia, reflected a move from universalism to selective access and from public responsibility to enhanced involvement of non-governmental organisations (NGOs) and private companies.122 Similar to this fundamental shift of principles, also one of the most highly praised policy innovations in social welfare was stimulated by one of the ‘historic nationalities’. With its 1988 Plan to Fight Poverty, the Government of the Basque Country introduced a minimum-income scheme (renta mínima de inserción) that was unprecedented in Spain and referred mainly to EU recommendations and the experiences of the French Revenu minimum d’insertion.123 Within a few years, however, the governments of all other autonomous communities adopted, in the face of outright resistance on the part of the national Government, similar programmes either upon their own initiative or in reaction to pressure from the parliamentary opposition, trade unions and NGOs such as Caritas. These programmes have been marked by considerable variations in terms of funding level and coverage,124 which most likely result from different policy priorities and financial resources. On a more general level, spending for social welfare as a whole similarly demonstrates clear disparities, as the northern-central axis of autonomous communities has higher coverage rates than the southern-Mediterranean axis, including the northern regions of Galicia and Cantabria.125 This has prompted some observers to claim that decentralisation would entail, regarding certain areas, excessive disparities and major problems of intergovernmental coordination.126 Others, by contrast, see the transfer of power to the autonomous communities in a more positive light as having ‘triggered a process of innovation in social policies bearing an intense territorial spill-over effect’,127 which on the whole produced considerable interregional convergence. Another factor limiting divergence are financial transfers. Article 158(1) of the Constitution stipulates that an ‘allocation may be made in the State Budget to the Self-governing Communities … to guarantee a minimum level of basic public services throughout Spanish territory’. It is evident from the wording of this clause that it merely alludes to the possibility of such financial transfers towards more countrywide equity. Since an amendment in 2009, however, Article 15(1) of the Organic Law on the Financing of Autonomous Communities (LOFCA) obliges the national Government to provide for such allocations so as to ensure a minimum level of basic 122

See Gallego and Subirats, ‘Spanish and Regional’ (2012) 284. See Moreno and Arriba (n 116) 10ff. 124 M Laparra and M Aguilar, ‘Social Exclusion and Minimum Income Programs in Spain’ in B Palier (ed), Comparing Social Welfare Systems in Southern Europe, vol 3, Florence Conference (Paris, MIRE, 1997). 125 Gallego and Subirats (n 117) 283f. 126 With reference to minimum income and long-term care, respectively, see L Ayala, ‘Tackling Poverty’ in Guillén and León (n 114); Sarasa, ‘Long-term Care’ (2011). 127 AM Guillén and S Álvarez, ‘The EU’s Impact on the Spanish Welfare State: The Role of Cognitive Europeanization’ (2004) 14 Journal of European Social Policy 285, 294; Similarly, L Moreno and C Trelles, ‘Decentralization and Welfare Reform in Andalusia’ (2005) 15 Regional and Federal Studies 519, 524. 123

Spain 369 public services regarding education, healthcare and essential social services. Just for this purpose, the special legislation established a new fund (Fondo de Garantía de Servicios Públicos Fundamentales) that is supported by payments from the national Government and, through revenues from ceded taxes, also from the autonomous communities (Article 15(2)).128 (2) Responsibility for healthcare has been decentralised as well, albeit to a clearly lesser extent. While Article 148(1)(21) of the Spanish Constitution appears at first glance to assign to the autonomous communities comprehensive jurisdiction over ‘health and hygiene’, this is immediately put into perspective by an extensive list of enumerated special competences belonging to the national Government in this area. Article 149(1)(16) makes the latter not only responsible for particular issues such as ‘external health measures’ or ‘legislation on pharmaceutical products’, but also assigns to it a general competence regarding ‘basic conditions and general coordination of health matters’. In contrast to social welfare, the decentralisation of healthcare was initially a highly asymmetrical process. It started with Catalonia in 1981, followed by Andalusia and the Basque Country several years later. This asymmetry was a deliberate design because the national Government did not believe all regions to have sufficient capacities to assume comprehensive healthcare responsibilities. The few privileged autonomous communities, on the other hand, were involved in a number of conflicts with the Spanish Government regarding the delineation of competences, which naturally made the Constitutional Court in this period a pivotal actor. From 1983 to 1990, the latter held in several rulings that the national Government bears, in relation to healthcare and other matters, a responsibility for promoting ‘general interests’ and thereby reinforced its dominant position.129 The Government of the Partido Socialista Obrero Español (PSOE) used this position in the mid-1980s to carry out a comprehensive reform towards a national healthcare system that featured, in stark contrast to the model of the Franco regime, regionalised management and universal access to services.130 In this constellation, the national Government merely tasked the autonomous communities with the provision of health services and continued to coordinate them. Soon thereafter, the initial asymmetry regarding healthcare competences was largely removed, as the remaining autonomous communities assumed similar powers in this realm. Politically, this symmetrisation was based on the acuerdos autonómicos of 1992, ie on a settlement reached between the two major national

128 See J Ruiz-Huerta Carbonell and A Herrero Alcalde, ‘Fiscal Equalization in Spain’ in N Bosch and JM Durán (eds), Fiscal Federalism and Political Decentralization: Lessons from Spain, Germany and Canada (Cheltenham: Edward Elgar, 2008) 150. 129 For an overview of this case law, see R Agranoff, ‘Federal Asymmetry and Intergovernmental Relations in Spain’ in 2005 Special Series on Asymmetric Federalism (Kingston, Queen’s University, 2006) 10. 130 The legal cornerstones that effected the shift from a contributory model to a universal model financed through progressive taxation were the 1984 Royal Decree on Basic Health Care Structures and the 1986 General Health Care Law. See R Gallego et al, ‘Spain: From State Welfare to Regional Welfare?’ in McEwen and Moreno (n 8) 110f.

370 Social Welfare and Healthcare parties, PSOE and the Partido Popular (PP), without any regional involvement.131 Legally, it was effected by applying Article 150(2) of the Constitution, according to which ‘the State may transfer or delegate to the autonomous communities, through an organic law, powers appertaining to it which by their very nature lend themselves to transfer or delegation’. Even though healthcare has since developed into one of the key jurisdictions of all autonomous communities, the national Government still has financial levers and influences regional policies through its efforts to ensure, for instance with the 2002 Law on Quality and Cohesion of the National Health System, countrywide standards.132 Within this national framework, which is certainly narrower than concerning social welfare, there is nonetheless space for regional innovation. A prime example is the Basque Country, which is distinguished by such innovations as the introduction of an electronic health card. This and other innovations were later emulated by other autonomous communities.133 Even though some interregional divergences prevail, these emulation processes, in the Spanish context famously described as ‘ethnoterritorial mimesis’,134 as well as the above-mentioned national basic legislation, have kept differences within certain limits. This has also been facilitated by intensive intergovernmental coordination through multilateral sectoral conferences. While such meetings had been largely restricted until the late-1980s to the realms of education and economy, this ‘joint policy-making mode’135 was then extended to healthcare. As to the often voiced concern that decentralisation would endanger equity, the thus overall limited divergence of regional healthcare models has rather reduced inequalities between autonomous communities than instigated a ‘race to the bottom’.136 11.5 UNITED KINGDOM

(1) At the time of its establishment, the UK welfare state was regarded as one of the most powerful symbols of British nationhood and is deemed to have replaced in this unifying role its then-declining Empire.137 In particular, it was closely linked to the notion of union as espoused by the Labour Party. Following the seminal report by the economist William Beveridge in 1942,138 the Labour Government of the first post-war years established, on the basis of wartime solidarity, a welfarestate model that featured, compared to Bismarckian systems, an exceptionally high 131

See Agranoff, ‘Intergovernmental Relations’ (1999) 109. See Guillén and Álvarez, ‘The EU’s Impact’ (2004). 133 See Gallego et al, ‘Spain’ (2005) 111. 134 L Moreno, ‘Asymmetry in Spain: Federalism in the Making?’ in Agranoff (eds), Accommodating Diversity (1999) 165. 135 Agranoff, ‘Intergovernmental Relations’ (n 120) 109. 136 See J Costa-Font and A Rico, ‘Vertical Competition in the Spanish National Health System (NHS)’ (2006) 128 Public Choice 477. 137 See F Williams, Social Policy: A Critical Introduction (Cambridge, Polity Press, 1989) 162. 138 Social Insurance and Allied Services (1942) proposed a structure of flat-rate, contributory insurance that would provide a universal safety net while not jeopardising private savings. 132

United Kingdom 371 degree of universalism. The decidedly all-British character of this model is reflected in the (re)labelling of a number of institutions such as the Ministry of National Insurance or the National Health Service. Thus, the UK welfare state was at its outset fully in line with TH Marshall’s classic conception of national social citizenship, presupposing a central government that equalises the particularities of local service provision. In light of this drive for uniformity, it is evident that devolution had to be perceived as a threat to the very basis of the welfare state. Consequently, the Labour Party’s focus shifted from advocating home rule for Scotland and Wales to promoting the interests of these two parts of the UK at the national Government level.139 Unsurprisingly, however, the unifying force of the welfare state diminished along with its capacity to deliver benefits due to deteriorating economic conditions in the 1960s. This eventually fuelled the rise of Scottish and Welsh parties pursuing a combined nationalist–welfarist agenda, in particular against Conservative British governments (1979–97). However, even in these times before the initiation of the devolution process in 1998, ie in the period of what was known as administrative devolution140 with central government departments being responsible for all parts of the UK, there was some divergence across the United Kingdom and even within England. Relative administrative autonomy was arguably enabled by a national government that was hardly interested in the details of policy implementation in the peripheral parts of the UK, which it sought to offload in order to focus on ‘high politics’.141 Even though such autonomy in a realm as important as social welfare demonstrates the functioning of the UK even before 1998 as a ‘union state’142 rather than a unitary state, it is beyond doubt that devolution was a critical juncture because it greatly increased the potential to vary policies. The Scotland Act of 1998 devolved all social-welfare competences that are not explicitly identified as reserved matters of the UK Government in Schedule 5(II)(F) ‘Social Security’. For instance, social security schemes143 are, with certain exceptions, reserved matters. These include, in particular, National Insurance, which provides benefits in case of illness, unemployment, old age and other cases, or housing benefits. Northern Ireland has jurisdiction over all areas that are not listed as excepted or reserved matters, which comprise, similar to Scotland, National Insurance and

139 In Scotland, this new focus was the explicit rationale for Labour’s formal abandonment of home rule in 1958. As to Wales, it is epitomised by Aneurin Bevan, the leading representative of the Welsh Labour Party, who became the chief architect of the National Health Service. See N McEwen and R Parry, ‘Devolution and the Preservation of the United Kingdom Welfare State’ in McEwen and Moreno (n 8) 51. 140 Administrative devolution started in Scotland with the establishment of the Scottish Office in 1885 and in Wales with the Welsh Office in 1964. It then lasted in both cases until 1998. See J Mitchell, Devolution in the UK (Manchester, Manchester University Press, 2009) 16ff. 141 See M Moran, The British Regulatory State: High Modernism and Hyper-innovation (Oxford, Oxford University Press, 2003) 64. Such relations between the centre and the periphery have been described with the image of ‘Court and Country’. See J Bulpitt, Territory and Power in the United Kingdom (Manchester, Manchester University Press, 1983)). 142 S Rokkan and D Urwin, The Politics of Territorial Identity: Studies in European Regionalism (London, Sage, 1982) 11. 143 They are defined as ‘schemes supported from central or local funds which provide assistance for social security purposes to or in respect of individuals by way of benefits’ (F.1).

372 Social Welfare and Healthcare a few additional excepted matters in Schedule 2(10). In contrast to these two parts of the UK, the National Assembly for Wales assumed the power to make primary legislation only more recently. Since a positive referendum in 2011, the National Assembly for Wales has finally been granted in all 20 devolved matters the power to enact primary legislation, so-called Acts of the Assembly, without the need for the consent of the UK Parliament.144 The social welfare matters that may be regulated through such acts of the Welsh legislature are enumerated in a very detailed manner in Schedule 7(1)(15) of the 2006 Government of Wales Act.145 This legislative autonomy of the devolved governments is underpinned by their spending autonomy, which is, especially in a cost-intensive policy field like social welfare, of paramount importance. The so-called Barnett formula146 has from the perspective of the parts of the UK with a devolution regime the advantage that they may freely allocate funds from a block grant as they see fit. Of course, their leeway in decision-making regarding their own expenditures is to a certain degree limited by the fact that finance in general is determined by spending priorities in England.147 The question is then whether this legislative and financial autonomy in the wake of devolution has really been utilised by Scotland, Northern Ireland and Wales in order to pursue own policies. Above all, the narrative of Scottish nationhood is centred around being particularly welfarist so that devolution was expressly portrayed in the 1990s as necessary for public policy to be more in line with its ‘collective and community traditions’.148 In fact, there are today some examples of divergence. As to Scotland, a case in point is the gradual abolition before 2008 of up-front university tuition fees and the introduction of free personal care for the elderly in 2001, both of which have inflamed long-lasting tensions with the UK Government.149 The removal of tuition fees ran contrary to the UK policy, because the Labour Government had introduced them at the national level just before devolution. The second case of policy divergence proved similarly controversial because free provision of non-nursing personal care in private homes and residential establishments was guaranteed in Scotland just after the UK Government had refused to follow a 1999 recommendation to that effect of the Royal Commission on Long-Term Care

144 Arts 107–09 and Sch 7 of the Government of Wales Act 2006; Government of Wales Act 2006 (Commencement of Assembly Act Provisions, Transitional and Saving Provisions and Modifications) Order 2011, SI 2011/1011. The initial devolution act of 1998 had only granted the assembly the power to pass secondary legislation. While the act of 2006 empowered the assembly to make primary legislation by means of ‘Assembly Measures’, this power had to be transferred in each single case through a Legislative Competence Order of the UK Parliament, which subsumed a specific ‘Matter’ to one of the 20 devolved ‘Fields’ enumerated in Sch 5 of the Government of Wales Act 2006. 145 Precisely this provision devolves jurisdiction over ‘Social welfare including social services. Protection and well-being of children (including adoption and fostering) and of young adults. Care of children, young adults, vulnerable persons and older persons, including care standards’. 146 See ch 7.1.3. 147 See D Heald and A McLeod, ‘Embeddedness of UK Devolution Finance within the Public Expenditure System’ (2005) 39 Regional Studies 495. 148 Scottish Constitutional Convention, ‘Scotland’s Parliament, Scotland’s Right’ (Edinburgh, 1995) 2f, paulcairney.files.wordpress.com/2015/09/scc-1995.pdf. 149 See J Curtice, ‘Policy Divergence: Recognising Difference or Generating Resentment?’ in G Lodge and K Schmuecker (eds), Devolution in Practice 2010 (London, Institute for Public Policy Research, 2010).

United Kingdom 373 for the Elderly.150 Although free Scottish personal care diminished the costs of the countrywide means-tested attendance allowance, the UK Government refused to compensate the devolved government for these savings. Such financial implications of policy divergence are part of the explanation as to why Wales and Northern Ireland eventually did not follow Scotland’s universalistic approach. Nor did they emulate the Scottish policy regarding the abolition of university tuition fees, concerning which they only foresaw means-tested compensation and exemptions. By contrast, a noteworthy Welsh innovation, namely the institution of the Children’s Commissioner, was basically copied by Scotland, Northern Ireland and even by the UK Government.151 Interestingly, the radicalism of breaking new ground seems to be even more pronounced in Wales than in Northern Ireland, as illustrated by its fundamental reform of early-childhood education. In summary, the overall impact of devolution on social welfare remains as yet difficult to assess because of different pre-devolution legacies and the fact that highly publicised examples such as those mentioned above are contrasted by substantial convergence on other issues.152 Fears of excessive divergence associated with a ‘welfare nationalist’ approach in Scotland and Wales,153 which would consider the subnational entity as the primary units of social solidarity, seem therefore largely unfounded. Since the early days of devolution, a strong contrarian current of ‘welfare unionist’ thinking has emphasised UK-wide social citizenship and equity to be enforced through a common policy framework.154 In some cases, the UK Government has sought to achieve this equity merely by lobbying the devolved governments so as to get them to commit to Britishness as being ‘cemented by the common endeavour of … building the welfare state’.155 In other cases, convergence has been produced by legislation passed by the UK Parliament on devolved matters of social welfare, which it has enacted, however, according to the ‘Sewel convention’ only after prior consent of the respective devolved legislature through a legislative consent motion (or ‘Sewel motion’).156 It is true that such (consented) intervention has cleared the way for far-reaching changes like the 2012 Health and Social Care Act and the Welfare Reform Act of the

150 See R Simeon, ‘Free Personal Care: Policy Divergence and Social Citizenship’ in R Hazell (ed), The State of the Nations 2003 (Thorverton, Imprint Academic, 2003). 151 See A Trench and H Jarman, ‘The Practical Outcomes of Devolution: Policy-Making across the UK’ in A Trench (ed), Devolution and Power in the United Kingdom (Manchester, Manchester University Press, 2007). 152 See D Wincott, ‘Social Policy and Social Citizenship: Britain’s Welfare States’ (2006) 36 Publius 169, 175f. 153 On ‘welfare nationalism’ and ‘welfare unionism’, see M Keating, ‘Intergovernmental Relations and Innovation: From Co-operative to Competitive Welfare Federalism in the UK’ (2012) 14 British Journal of Politics and International Relations 214, 221f. 154 See, eg, R Hazell and B O’Leary, ‘A Rolling Programme of Devolution: Slippery Slope or Safeguard of the Union’ in R Hazell (ed), Constitutional Futures: A History of the Next Ten Years (Oxford, Oxford University Press, 1999) 43. Similarly, the Calman Commission on Scottish devolution stated ‘there may be a case for a broadly common social citizenship across the UK’, Commission on Scottish Devolution, ‘The Future of Scottish Devolution within the Union: A First Report’ (Edinburgh, 2008) 4.60 news.bbc. co.uk/1/shared/bsp/hi/pdfs/02_12_08_calman.pdf. 155 G Brown and D Alexander, New Scotland, New Britain (London, The Smith Institute, 1999) 19. 156 See ch 8.1.3.

374 Social Welfare and Healthcare same year. But beyond such legislation, considerable convergence may also be attributed to other factors like the operation of interest groups on a UK-wide basis, the tight integration of Scottish and Welsh Labour with the UK Labour Party, the above- mentioned tight limits to fiscal autonomy and in the particular case of Northern Ireland the lasting impact of direct rule from 2002 to 2007.157 Intergovernmental coordination, by contrast, has hardly been a factor for generating uniformity. Such coordination has been weak and informal overall, a ‘projection forward of pre-devolution practices into the post-devolution era’,158 which is arguably the result of the piecemeal approach to devolution. To be sure, formal institutions have been in place since the late 1990s. But mechanisms for multilateral coordination like the Joint Ministerial Committee, either composed of the heads of government or of ministers with corresponding portfolios, only met regularly in their early days. Moreover, the 1999 memorandum of understanding that was supposed to guide intergovernmental coordination largely restrains itself to establishing fundamental principles such as the ‘4 Cs’ of communication, consultation, cooperation and confidentiality. The therefore marginal role of such mechanisms involving all governments has given rise to informal and largely bilateral relations—a state of affairs that has been repeatedly criticised as insufficient not only by academic observers, but also by the UK Parliament.159 (2) Similar to social welfare, devolution has also empowered subnational governments concerning healthcare. According to the Scotland Act of 1998, this subject matter is almost entirely decentralised, with only few matters reserved to the UK Government by virtue of Schedule 5(II)(J) ‘Health and Medicines’. Northern Ireland also enjoys ample general competence, which is only slightly limited by a few reserved matters listed in Schedule 3(31–35). The 2006 Government of Wales Act, by contrast, explicitly details, again with certain exceptions, subnational competences regarding ‘health and health services’ in Schedule 7(I)(9). But the resulting broad autonomy has to be seen, of course, in the context of certain influential historical legacies. First, there is the influential National Health Service (NHS), created in 1946–47, as a Beveridgean-style system with an exceptionally high degree of free and universal services.160 This service was overhauled comprehensively precisely in the two years prior to devolution. Second, the devolved governments are still impacted by the inheritance of pre-1998 administrative devolution. Even though the NHS has always 157 During this time, policies were closely aligned with those for England. Coupled with the inertia of devolved policy-making after this period, this has resulted in overall less variation. See M Keating et al, ‘Policy Convergence, Transfer and Learning in the UK under Devolution’ (2012) 22 Regional & Federal Studies 289, 295f. 158 C Jeffery, ‘Devolution in the United Kingdom: Problems of a Piecemeal Approach to Constitutional Change’ (2009) 39 Publius 289, 304. 159 See A Trench, ‘Washing Dirty Linen in Private: The Processes of Intergovernmental Relations and the Resolution of Disputes’ in Trench, Devolution (2007); Select Committee on the Constitution, Devolution: Inter-institutional Relations in the United Kingdom (second report) (2002–03, HL 28); Welsh Affairs Committee, Wales and Whitehall: Eleventh Report of Session 2009–10 (2010, HC 256). 160 National Health Service Act 1946, creating the National Health Service in England and Wales; The National Health Service (Scotland) Act 1947, creating the National Health Service in Scotland.

United Kingdom 375 appeared to be a bulwark of central control and was famously cast by Gordon Brown as epitomising Britishness,161 it had not even realised uniform s ervice provision before 1998.162 Back then, this was due to the crucial roles assigned to medical professionals and to institutions of administrative devolution like the Scottish Board of Health established as early as in the aftermath of World War I. This entailed that the distinctiveness of the various parts of the UK regarding healthcare was taken into account and promoted long before devolution. As a consequence, even though today’s subnational variation regarding healthcare is seen as impressive overall,163 this is only partly a result of post-1998 devolution itself. Although the costs and visibility of healthcare incited a lively public debate about devolution leading to ‘medical apartheid’,164 devolution surely enabled certain radical changes that were unthinkable before. A case in point is Scotland, which first opposed some of the Blair Government’s reforms aimed at greater involvement of the private sector and then completely reorganised its until-then Thatcherite health system.165 Scotland did so mainly by eliminating competition between ‘trusts’, ie firm-like healthcare providers like hospitals, and integrating them into the health boards from which the ‘trusts’ had previously commissioned care. Health service provision falls within the responsibility of these boards, which, similar to an earlier reform in Wales, cooperate closely with local governments. Both systems differ radically from the English one, which is characterised by the involvement of the ‘trusts’ as private actors and by strong centralisation.166 This diversity in terms of healthcare’s institutional structure makes clear why observers see greater variation regarding health policy-making than concerning health policies.167 However, there are also several examples of divergent policies such as the Welsh focus on public health and preventive medicine and moves on the part of all devolved governments to abolish health service prescription charges. Similar to social welfare, coordination issues following from this divergence are largely managed by informal networks, mostly between civil servants and relevant healthcare professionals, above all chief medical officers. Formal multilateral mechanisms such as the Joint Ministerial Committee and the memorandum of understanding only play a minor role. Under these current circumstances of weak coordination,

161 ‘Today, when people talk of the National Health Service, whether in Scotland, Wales or England people think of the British National Health Service: here national is unquestionably British … And its most powerful driving force is that every citizen of Britain has an equal right to treatment regardless of wealth, position or race, and indeed, can secure treatment in any part of Britain’—Gordon Brown (1999), quoted in N McEwen, ‘The Territorial Politics of Social Policy Development in Multi-level States’ (2005) 15 Regional and Federal Studies 537, 548. 162 See Moran, The British (2003) 139f. 163 SL Greer, ‘The Territorial Bases of Health Policymaking in the UK after Devolution’ (2005) 15 Regional and Federal Studies 501, 503ff. 164 On this debate, see SL Greer and A Trench, ‘Intergovernmental Relations and Health in Great Britain after Devolution’ (2010) 38 Policy and Politics 509, 516. 165 See K Woods, ‘Health Policy and the NHS in the UK 1997–2002’ in J Adams and P Robinson (eds), Devolution in Practice: Public Policy Differences within the UK (London, IPPR, 2002). 166 See SL Greer, ‘A Very English Institution: Central and Local in the English NHS’ in R Hazell (ed), The English Question (Manchester, Manchester University Press, 2005). 167 See SL Greer, Territorial Politics and Health Policy: UK Health Policy in Comparative Perspective (Manchester, Manchester University Press, 2004) 222ff.

376 Social Welfare and Healthcare which is at best appropriate for resolving low-intensity conflicts, ‘devolution and policy divergence can become political problems when they should be managerial ones’.168 11.6 COMPARATIVE CONCLUSIONS

Probably few other historical developments have had such a great impact on power relations between national and subnational governments as the expansion of the welfare state. The assumption by the national level of many social-welfare competences and, importantly, of competences over taxation to finance countrywide redistributive programmes has, without doubt, left a lasting imprint on intergovernmental power relations. The subnational level, on the other hand, often finds itself playing a complementary role of being reduced to functions of implementation, the provision of infrastructure and services. Such a new division of roles appears to have replaced in various cases the logic of many older federations as a ‘historical compromise between nationalizing modernizers and regional traditionalists’169 with economic competences allocated to the national level and regulating social and cultural affairs left to subnational government. However, the advent of the social agenda has not only strengthened national governments, it has also intensified, on the whole, intergovernmental collaboration and thereby, arguably, facilitated a trend towards cooperative federalism, albeit, in certain instances, also competitive federalism.170 The basis of how these relations look today is to some degree certainly still influenced by timing, more precisely, by the question of whether the introduction of social welfare and health policies preceded federalism or vice versa. It seems evident that national governments had a head start if they could, as in Belgium during the interwar period, occupy the field of welfare-state policies as the only legislator in a still unitary setting. However, even if a federal structure was already in place during the formative years of the welfare state, the national Government had an initial edge over their subnational counterparts for two reasons. First, old federal constitutions like those of the United States and Canada, which preceded the age of social welfare and health policies and thus failed to explicitly allocate competences in this area, tended to be interpreted in a manner favourable to the national Government. This critical moment of centralising interpretation was in the US case the upholding of the Social Security Act in 1937.171 It is true that, for Canada, the JCPC struck down in the same year the unemployment insurance plan172 because the subject would fall within provincial jurisdiction. But this was overruled just three years later by a constitutional amendment establishing the exclusive power of the national Government. A second typical advantage of national governments in the early days of the welfare

168

Greer and Trench, ‘Intergovernmental Relations’ (2010) 523. Hueglin and A Fenna, Comparative Federalism: A Systematic Inquiry (Peterborough, Broadview Press, 2006) 350. 170 See Keating, ‘Intergovernmental Relations’ (2012) 216f. 171 Steward Machine (n 26); Helvering v Davis (n 26). 172 Canada (AG) v Ontario (n 63). 169 TO

Comparative Conclusions 377 state was their command of most of the tax room. This superior fiscal position enabled it, for example, in the United States after the Sixteenth Amendment of 1913, to establish countrywide programmes that became, as expressions of uniform national social citizenship, deeply engrained and thus difficult to undo.173 To be sure, this path dependency174 created by the initial dominance of the national Government in either a federal or still unitary setting before or during welfare-state formation has left an imprint. However, it does not account entirely for how power relations look today. In fact, these relations have, in many instances, varied significantly at different points in time. Changes in the distribution of powers have been effected either through constitutional jurisprudence and/or in a written form through formal amendments. The latter may be in part, as with the statutes in Spain and special acts in Belgium, relegated to the sub-constitutional level. Even though the powers of the Spanish autonomous communities are thus regulated in great detail, they have still been influenced significantly by case law. In concrete terms, the Constitutional Court has clearly reinforced the constitutional authority of the national Government to adopt basic legislation (Articles 149(1)(16) and 149(1) (17)) by underlining, on several occasions, the latter’s responsibility for promoting ‘general interests’ and countrywide equity.175 In the United States, the influence of the Supreme Court has been much more pervasive, as demonstrated by its crucial rulings enabling the New Deal and Great Society legislation. In the United Kingdom, by contrast, neither case law nor written law in the devolution acts provides a definitive distribution of powers. There is still a lot of flexibility, albeit only for the central Government, to adopt legislation outside of its actual legislative competence, provided that the devolved legislatures give their consent to this pursuant to the Sewel convention. The eminently important Health and Social Care Act and the Welfare Reform Act of 2012 are cases in point. Regardless of whether the distribution of powers has been developed primarily by written law or by court judgments, there seems to be a general pattern of subnational authority over ‘social assistance’ and of national jurisdiction over ‘social security’, as these fields are defined and differentiated in Belgium.176 Admittedly, there are a few exceptions to this rule. While the responsibility of the US states for the Unemployment Insurance scheme does not really empower them due to the only marginal importance of this programme within the whole system of social welfare,177 recent developments in Belgium seem more noteworthy. The transfer of the competence for certain family allowances to the communities in 2011 and the upholding of the Flemish Care Insurance as an additional element complementing national ‘social security’ have clearly extended the scope of subnational legislation beyond mere ‘social assistance’. Nevertheless, it still remains to be seen whether these 173

See Banting, ‘Social Citizenship’ (2006) 54. idea that precisely influencing factors at earlier stages have disproportionately greater and self-reinforcing effects and thus to some degree predetermine future developmental trajectories is central to theories of path dependency. See P Pierson, ‘Increasing Returns, Path Dependency, and the Study of Politics’ (2000) 94 American Political Science Review 251. 175 STC 146/1986; STC 124/1989. 176 See ch 11.3. 177 See Finegold (n 28) 159. 174 The

378 Social Welfare and Healthcare steps will really turn out to be, against practical and political obstacles, a s lippery slope towards comprehensive decentralisation of ‘social security’. More generally, it should be borne in mind that the value of subnational jurisdiction over ‘social assistance’ schemes is typically reduced by nationally established basic common standards and by the status of these programmes as the residual component of social welfare. After all, it is ‘social security’ that encompasses the major income-transfer programmes like old-age pensions, unemployment benefits, health insurance and family allowances. Where ‘social assistance’ has more than a residuary function, as in Australia, it tends to be a realm of the national government.178 Regardless of whether the distribution of powers has been changed formally or judicially, a clear delineation of competences has proven to be difficult due the enormous complexity of transversal subject matters. If the Canadian Supreme Court famously characterised healthcare as an ‘amorphous topic’,179 which would necessarily require a casuistic approach depending on the context of the specific issue at stake, this appears to be equally true beyond Canada and for the field of social welfare as well. Arguably, the latter area has become even more transversal and complex since governments, both national and subnational, started in the 1990s to subscribe to the idea of ‘welfare to work’ or ‘workfare’. This is due to the linkage in this context of traditional social welfare, ie passive income support, with active labour-market policies, which has ‘blurred the distinction between redistributive and developmental policies’.180 The resulting increased complexity is particularly evident in the case of Belgium, where labour-market policies are a core responsibility of the regions, which now need to be coordinated with traditional social-welfare policies of the communities and the national Government in a tripolar constellation. To be sure, the distribution of powers is aimed in certain countries to assign exclusive competences by dissecting these complex subject matters, either through detailed special legislation like in Belgium or, as with the Canadian double aspect doctrine, through constitutional jurisprudence. The result is, however, in both cases a mosaic of competences, which still fails to prevent overlaps and disputes over the respective areas of authority. Instead, the complexity of social welfare and health, which is perhaps greater than of any other subject matter, needs to be confronted with enhanced intergovernmental cooperation, above all in a multilateral format. This cooperation, however, is sometimes absent, perhaps most evidently in the United Kingdom. Above all, in such cost-intensive areas as social welfare and health, the formal distribution of powers may be considerably distorted by the national government’s exercise of its spending power. Essentially, who holds the purse dictates policy choices and, as a consequence, legislation. In some cases, like the United States, this financial instrument supplements ample law-making powers. In others, national governments were forced to recognise the limits of direct legislative intervention and have, as a result, turned to their spending power as a surrogate. Such a deliberate shift of strategy occurred, for example, in Canada, as the national Government introduced in the

178

See Banting (n 19) 49f. Schneider v British Columbia (n 84) 142. 180 Keating (n 153) 218. 179

Comparative Conclusions 379 1960s, aided by favourable case law,181 the Canada Assistance Plan and Medicare as shared-cost programmes. Another case in point is the reliance on conditional grants in Spain since the mid-1980s after a futile attempt to recentralise social welfare by legislative means. In light of the considerable vertical fiscal imbalance in many countries,182 the use of the spending power remains a powerful instrument. It is true that the conditions attached to payments have become less harsh in some cases such as the Canada Social Transfer, which, in 1996, abolished most requirements that had still been obligatory under the Canada Assistance Plan. However, it is important to note that this greater autonomy was accompanied by a drastic reduction of funding from 50 per cent to roughly 10 per cent so that provinces have found it difficult to make use of this autonomy due to the scarcity of financial resources. The devolved governments in the United Kingdom find themselves in a similar situation. On the one hand, they are free to spend their block grant under the Barnett formula on whatever purpose they choose. On the other hand, they face the practical constraints of the size of this block grant being determined by English expenditure priorities. In an attempt to avoid these problems of underfunding, which are at odds with genuine autonomy, constitutional equity clauses more or less oblige in certain cases the national government to redistribute transfers to subnational entities. A case in point is Canada, where such payments are supposed to guarantee provinces ‘sufficient revenues to provide reasonably comparable levels of public services at reasonably comparable levels of taxation’ (Section 36(2) of the Constitution Act 1982). Whereas the judicial enforceability of this provision is still an unsettled issue, a similar obligation on the part of the Spanish Government to ensure minimum countrywide standards (Article 15(1) LOFCA) has so far clearly had a stronger practical impact. A new fund established precisely for this purpose has, since the reform of 2009, evolved into a primary instrument of equalisation. Whereas these two equity clauses address the national Government in its redistributive function, the case law of the US Supreme Court obliges the states to observe certain basic equity standards183 and it does so, interestingly, by invoking the freedom of movement instead of solidarity. To be sure, the national Government of the United States also makes, like any other government, even without a constitutional obligation, at least some equalising transfers. These are inherent, for example, in the needs-based differentiation of matching funds within Medicaid. In comparison, however, it certainly assumes to a lesser extent the central fiscal role, which many economists demand for national governments regarding redistribution and counterbalancing the differential impact of economic shocks across a country through active macroeconomic management.184 Beyond conditional and redistributive transfers, variation between subnational entities regarding social welfare and health is also limited by a number of other

181

Angers v MNR [1957] Ex CR 83; Reference Re Canada Assistance Plan (n 74). See ch 7.4. 183 Shapiro v Thompson (n 33); Memorial Hospital (n 33). 184 See especially WE Oates, ‘Assignment of Responsibilities and Fiscal Federalism’ in R Blindenbacher and A Koller (eds), Federalism in a Changing World: Learning from Each Other (Montreal, McGill–Queen’s University Press, 2003) 40. 182

380 Social Welfare and Healthcare factors. Among these are, most prominently, scarce own revenues, top-down intergovernmental coordination dominated by the national Government, legacies of uniformity from before decentralisation like in Belgium and the emulation of innovative policies, which again produces convergence. In this light, the general statement that ‘uniformity is antithetical to federalism’ because the freedom of subnational units to disagree with each other and with the national government ‘inevitably leads to differentiation’,185 has to be put into perspective. It is, of course, true that a federal system creates the potential for differentiation. But in practice this potential is often barely realised due to the above-mentioned limiting factors. Counterexamples of effective divergence are often more affluent subnational units, which pursue, like in the northern parts of the United States and Spain, quite consistently more extensive social welfare and health policies. Similar efforts towards one’s own path are being made by some entities that have recognised the popularity and visibility of these policies and thus conceive of them as an integral part of a competing regional nation-building project. In this sense, elements of a distinct ‘welfare region’ are often deliberately constructed to serve in everyday life as ‘unmindful reminders of nationhood’.186 A case in point is Quebec with its opt-out from the 1999 Social Union Framework Agreement and the bilateral Health Care Agreement of 2004. In comparison, policy divergence on the part of Scotland has so far remained more limited and confined to certain prestigious and highly publicised initiatives like free personal care for the elderly. While variation is sometimes truly the result of the more welfarist leanings of Quebec and Scotland, it follows in other instances just from the desire to have one’s own policies rather than one’s own, more extensive policies. Evidence, for instance, of broad support for contributory pensions in the 1960s throughout Canada suggests that the drive for a separate Quebec Pension Plan was really more about welfare nationalism than about correspondence to divergent preferences.187 Finally, a crucial issue remains the question of whether federalism facilitates a race to the top, ie towards welfare expansion, or rather a race to the bottom, ie welfare retrenchment. Although the latter view is much more common, comparative evidence is clearly mixed. The dynamics of national and subnational governments outbidding each other with similar policies, thereby triggering a costly trend towards higher standards of social protection, is a problem in Belgium and usually blamed on the constellation of parallel exclusive powers. More often, however, national governments have instead rather emulated and thus spread innovative subnational policies than engaged in the dynamics of outbidding. This was the case with the unemployment insurance scheme of Wisconsin, as well as Saskatchewan’s hospital insurance and Medicare programme. Beyond this vertical pattern, welfare expansion has also occurred at least as frequently through the horizontal diffusion of innovative policies. The most notable case is probably the spread of the minimum

185 A Wildavsky, ‘Federalism Means Inequality: Political Geometry, Political Sociology, and Political Culture’ in RT Golembiewski and A Wildavsky (eds), The Costs of Federalism: Essays in Honor of James W Fesler (New Brunswick, NJ, Transaction Publishers, 1984) 57. 186 M Billig, Banal Nationalism (London, Sage, 1995) 174. 187 See Banting (n 19) 61f.

Comparative Conclusions 381 income scheme, introduced by the Basque Country in 1988, to the other autonomous communities of Spain. The ‘race to the bottom’ argument, on the other hand, relies on the reasoning that subnational governments would, in response to real or perceived pressure from mobile companies and individuals, lower taxes and limit social protection. On a normative level, such competition is for some a key reason for praising federalism in Hayek’s tradition as a ‘market preserving’ mechanism that would limit the ‘intrusion’ of government.188 Such thinking is based on the assumption that the choice of mobile citizens, willing to ‘vote with their feet’,189 between varying offers of taxes and social protection in different subnational entities would ‘discipline’ governments like consumer choice does with producers. Yet, empirical evidence of a race to the bottom is not even conclusive for the United States,190 which seems to provide, due to people’s mobility and in default of a formal fiscal equalisation system, the most fertile ground for this kind of competition. Furthermore, the section on the US case demonstrated that the comparatively low standards of social protection are linked to a number of other causes and that the federal system has in several instances rather facilitated welfare expansion than its retrenchment. A third related argument does not claim that a federal design triggers a race to the bottom, but that it at least hampers welfare expansion due to inherently cumbersome procedures of decision-making with more veto players. It is true, for example, that Ontario and Quebec initially obstructed the development of the Canadian welfare state and that a number of US states have played a similar role concerning the Patient Protection and Affordable Care Act. First, however, there are in both instances other subnational entities that are vigorously in favour of welfare expansion. Second, the real veto point is, in the US case, undoubtedly rather located in the presidential–congressional separation of powers linked with a situation of divided government than at the state level. In summary, it thus appears fair to say that each of the three above-mentioned arguments about the link between a federal design and welfare-state development only holds true in some cases so that a generally either positive or negative impact is unlikely.

188 See, eg, B Weingast, ‘The Economic Role of Political Institutions: Market Preserving Federalism and Economic Growth’ (1995) 11 Journal of Law, Economics, and Organization 1. Friedrich Hayek was arguably the first to claim that in federal states there would be less government intrusion. See FA Hayek, ‘The Economic Conditions of Interstate Federalism’ in FA Hayek (ed), Individualism and Economic Order (London, Routledge, 1976). Wallace Oates, by contrast, concluded that the degree of centralisation has little effect on the size of government, WE Oates, ‘Searching for the Leviathan: A Reply and Some Further Reflections’ (1989) 79 American Economic Review 578, 582. Jonathan Rodden, in turn, finds that decentralisation only limits governmental intervention if subnational entities have to fund their expenditures through their own taxes. See J Rodden, ‘Reviving Leviathan: Fiscal Federalism and the Growth of Government’ (2003) 57 International Organization 695, fn 79. 189 CM Tiebout, ‘A Pure Theory of Local Expenditures’ (1956) 65 Journal of Political Economy 416. 190 For the claim that there is a race to the bottom, see PE Peterson and MC Rom, ‘American Federalism, Welfare Policy and Residential Choices’ (1989) 83 American Political Science Review 711; sceptical towards this finding are C Weissert and W Weissert, Governing Health: The Politics of Health Policy (Baltimore, Johns Hopkins University Press, 2002) and S Schram et al, ‘Without Cause or Effect: Reconsidering Welfare Migration as a Policy Problem’ (1998) 42 American Journal of Political Science 210.

12 Environmental Protection

F

ROM A HISTORICAL point of view, the relevance of environmental protection as a task for public authorities only emerged in the context of the Industrial Revolution, which entailed an unprecedented level of pollution. Even though some legislation was in this regard adopted well before the twentieth century— most notably the British Alkali Acts of 1863—modern environmental law is widely regarded as dating back to the late 1960s.1 In response to the increasing ecological activism of that time, the United States enacted, in 1969, the seminal National Environmental Policy Act and established the Environmental Protection Agency (EPA) one year later.2 In 1971, the European Commission issued a first political document regarding a common environmental policy,3 and the United Nations Conference on the Human Environment (‘Stockholm Conference’) was held the following year. In this early period of environmental protection, the focus was clearly on restrictive legislation that relied on compulsory limits and on penalties in case of noncompliance. However, this approach, meanwhile pejoratively termed ‘command and control’, came to be massively criticised in the 1980s as overly bureaucratic, inflexible and too costly for businesses.4 Such criticism set the stage for a new trend towards deregulation, which concentrated on non-compulsory measures such as information and voluntary agreements,5 as well as on economic instruments like environmental taxes or tradable permit schemes. More recently, another shift has taken place that has given rise to so-called new environmental governance.6 This novel approach is seen as being characterised by pluralism due to the inclusion of a wide range of stakeholders in decision-making processes and by increased regulatory flexibility through the combination of classical legal tools and previously economic and voluntary instruments, so-called New Environmental Policy Instruments (NEPIs). This trend towards diversification in terms of both decision-makers and 1 See N Gunningham, ‘Environment Law, Regulation and Governance: Shifting Architectures’ (2009) 21 Journal of Environmental Law 179, 182. 2 See section 12.1 below. 3 See European Commission, ‘First Communication of the Commission about the Community’s Policy on the Environment’, SEC (71) 2616. The first legally binding instruments then followed in 1975 regarding water and waste law, eg Council Directive 75/442/EEC of 15 July 1975 on waste [1975] OJ L194/23; Council Directive 75/440/EEC of 16 June 1975 concerning the quality required of surface water intended for the abstraction of drinking water in the Member States [1975] OJ L194/26. 4 See Gunningham, ‘Environment Law’ (2009) 182ff. 5 See Organisation for Economic Co-operation and Development (OECD), Voluntary Approaches for Environmental Policy: Effectiveness, Efficiency and Usage in Policy Mixes (Paris, OECD, 2003). 6 See K Holzinger et al, ‘Rhetoric or Reality? New Governance in EU Environmental Policy’ (2006) 13 European Law Journal 403.

United States 383 instruments is considered to have been facilitated by the forces of globalisation, marketisation and, most importantly for our purposes, decentralisation.7 The new mode of governance involving multiple actors does not mean, of course, that the state and its environmental protection legislation have become irrelevant. After all, alternative instruments such as tradable permit schemes also operate within, and are bound by, this legislation. In other words, ‘NEPIs have supplemented rather than supplanted traditional regulation’.8 As a policy field, environmental protection has several particular characteristics that also impact on the dynamics between the different levels of government. While legislation in this area was first adopted at the national level less than 50 years ago, the topic is even newer for subnational governments. This entails that intergovernmental dynamics are still much more in flux than in other—more traditional—fields of governmental action. Another important feature is the difficulty of precisely circumscribing environmental protection as a clear-cut field due to both its intrinsic interrelations with other areas, such as agriculture, spatial planning or energy, and its internal fragmentation into specific legal regimes for distinct subfields, which are often assigned to different government levels. This character as a complex transversal issue has entailed, rather than a holistic approach, stark differentiation of regulatory frameworks for the protection of water, air, soil and many other subfields both internationally and within states. While this development has facilitated the finding of solutions for specific issues, it has arguably given rise to a ‘balkanisation of scholarly expertise’9 and the challenge of grasping the interactions between these specific regimes.10 This seems even more important when, as is usually the case, multiple levels of government are involved. A third characteristic follows from the fact that many environmental problems transcend boundaries, both within states and between them. As these problems therefore require transboundary responses, efficient environmental protection is not only about autonomy of different government levels but also about their coordination.11 12.1 UNITED STATES

It is hardly surprising that the drafters of the US Constitution back in 1787 did not contemplate environmental protection as an issue to be dealt with in the distribution 7 See MA Delmas and OR Young, Governance for the Environment (Cambridge, Cambridge University Press, 2009) 71ff. 8 RKW Wurzel et al, Environmental Governance in Europe: A Comparative Analysis of New Environmental Policy Instruments (Cheltenham, Edward Elgar, 2013) 223. 9 E Fisher, ‘Maturity and Methodology: Starting a Debate about Environmental Law Scholarship’ (2009) 21 Journal of Environmental Law 213, 242. Such ‘balkanisation’ might be the case even for two issues that are relevant for a common goal: ‘an expert on the rules governing carbon sinks may have trouble communicating with an expert on international emissions trading, notwithstanding the fact that both issues fall under the Kyoto Protocol on climate change’, D Bodansky et al, ‘International Environmental Law: Mapping the Field’ in D Bodansky et al (eds), Oxford Handbook of International Environmental Law (Oxford, Oxford University Press, 2007) 4. 10 See Fisher, ‘Maturity’ (2009) 230. 11 See A Nollkaemper, ‘Responsibility of Transnational Corporations in International Environmental Law: Three Perspectives’ in G Winter (ed), Multilevel Governance of Global Environmental Change: Perspectives from Science, Sociology and the Law (Cambridge, Cambridge University Press, 2006) 199.

384 Environmental Protection of powers. Nor did subsequent amendments create an explicit and evident legal basis for the national or state governments to act as the primary regulator in this area. Quite naturally, it was thus for the Supreme Court to provide clarification. At the outset of environmental legislation in the late-1960s, the national Government used its treaty-making power (Article II, section 2, clause 2)12 to link environmental issues, until then widely recognised as local affairs, to international obligations and thus make it its prerogative.13 Later on, the Supreme Court facilitated centralisation by subsuming environmental protection under its broad interpretation of the Commerce Clause (Article I, section 8, clause 3).14 Consequently, competences in this area have been subject to the fluctuations of the Supreme Court’s case law regarding this crucial provision. Since the Court restricted, for the first time in six decades, in United States v Lopez,15 the reach of national jurisdiction under the Commerce Clause in 1995, the limiting rationale behind this decision and other judgments issued soon thereafter16 has affected the field of environmental protection in particular. This is because all these rulings require, for the decisive test of whether intrastate activities substantially affect interstate commerce, that these activities must be economic in nature. After all, the prevention of Congress from regulating non-economic intrastate activities provides, even though ‘economic’ was defined very broadly in Gonzales v Raich, some protection for the states against national legislation,17 at least more than during the decades before these cases. One area, however, in which the Supreme Court still seems to consistently uphold national authority is the regulation of depletable natural resources.18 Beyond that, the picture is quite mixed and confusing because the Court has not only curtailed national power by limiting, since United States v Lopez, the reach of the affirmative Commerce Clause. At the same time, it has also curtailed state power by expanding the scope of the dormant Commerce Clause, which prohibits a state from unjustifiably burdening or discriminating against interstate commerce.19 Thus, the Supreme Court has demonstrated on the whole a rather restrictive attitude towards environmental legislation on the part of both national and state authorities.20 On the basis of this rather dynamic distribution of powers through constitutional jurisprudence, both the national Government and state governments have become active in the field of environmental protection. Soon after World War II, Congress adopted some first legislative measures concerning water quality, the Federal Water

12

See ch 14.1. Fitzgerald, ‘The Constitutional Division of Powers with Respect to the Environment in the United States’ in K Holland et al (eds), Federalism and the Environment: Environmental Policymaking in Australia, Canada and the United States (Westport, Greenwood Press, 1996) 25. 14 See box 5.G. 15 United States v Lopez 514 US 549 (1995). 16 United States v Morrison 529 US 598 (2000); Gonzales v Raich 545 US 1 (2005). 17 See B Hudson, ‘Commerce in the Commons: A Unified Theory of Natural Capital Regulation under the Commerce Clause’ (2011) 35 Harvard Environmental Law Review 375, 378. 18 ibid 431. 19 Oregon Waste Systems Inc v Department of Environmental Quality of Oregon 511 US 93 (1994). 20 See the long-term case law analysis by CA Klein, ‘The Environmental Commerce Clause’ (2003) 27 Harvard Environmental Law Review 1. 13 EA

United States 385 Pollution Control Act of 1948. More critically for states’ autonomy, it initiated in the following years a grant programme to steer the process of developing plans for water pollution control in the states.21 However, comprehensive regulatory activism at the national level did not start until the crucial National Environmental Policy Act of 1969 and the subsequent wave of laws concerning specific subfields: the Clean Air Act of 1970, the Federal Water Pollution Control Act of 1972,22 the Endangered Species Act of 1973 and the Resource Conservation and Recovery Act of 1976. In hindsight, much of this national legislation was justified by the need for countrywide minimum standards to avoid confusion among businesses resulting from too different rules applied in different states and to prevent a race to the bottom concerning environmental protection standards.23 Yet, even though national regulation has since then definitely increased dramatically, states still have made use of their legislative autonomy. Even the constitutions of states have been complemented over the last half century with environmental provisions which then provided encouragement for legislative action in this field.24 The extent of this action is illustrated by estimates that roughly two-thirds of states’ environmental laws are adopted upon their own initiative and not merely specifications of national standards.25 Of course, some state legislatures are more active than others in this regard. This horizontal divergence particularly concerns those areas like climate policy that have not been pre-empted by the national Government. Mostly notably, California seized the opportunity created by a stalemate in Congress to establish its own capand-trade programme concerning greenhouse gases.26 In several areas that, unlike greenhouse gas emissions, are well covered by national legislation, the laws of California and other states have set more stringent standards, which again illustrate certain asymmetries regarding environmental protection. The fact that some subnational entities have gone further than others defies the general validity of the above-mentioned race-to-the-bottom argument.27 Section 209 of the Clean Air Act adopted by Congress in 2010 even acknowledged a special status of California insofar as it allows this state, and only this state, to impose stricter rules and allows the remaining states to follow either these regulations or the national ones. As far as the implementation of environmental law is concerned, this task is in some areas vested exclusively in national administrative bodies and carried out along cooperative patterns in others. Whereas the federal role had been largely

21 See Committee on Public Works, Federal Water Pollution Control Act Amendments of 1971—Report of the Committee on Public Works United States Senate together with Supplemental Views to Accompany S 2770 (Washington, US Government Printing Office, 1971) 2. 22 Later renamed the Clean Water Act. 23 See HN Butler and JR Macey, Using Federalism to Improve Environmental Policy (Washington, AEI Press, 1996) 21ff. 24 See BH Thompson Jr , ‘The Environment and Natural Resources’ in GA Tarr and RF Williams (eds), State Constitutions for the Twenty-First Century: The Agenda of State Constitutional Reform (Albany, SUNY Press, 2006). 25 See ME Kraft, Environmental Policy and Politics, 3rd edn (New York, Pearson Longman, 2004) 90. 26 Global Warming Solutions Act (AB 32) 2006. 27 For a critical assessment of this argument, see RL Revesz, ‘The Race to the Bottom and Federal Environmental Regulation: A Response to Critics’ (1997) 82 Minnesota Law Review 535.

386 Environmental Protection limited before the 1969 National Environmental Policy Act to providing research and technical support, this pivotal law declared environmental protection to be a matter of concern for all national agencies. The above-mentioned wave of legislation in the 1970s then effectively entrenched their position as principal administrators in most subfields of environmental protection. Notwithstanding this dominance, there are significant examples of vertical intergovernmental cooperation concerning implementation processes. For instance, even though the limits under the Clean Air Act are national standards, the states are assigned the crucial task of developing—within certain deadlines for compliance and subject to the approval of EPA—their own individual state implementation plans. Intergovernmental relations under this act were also the subject of a recent legal dispute in which some states attempted to force the EPA as a national body to take measures against climate change. 12.A United States 2007: EPA Standards for Greenhouse Gas Emissions against Climate Change? Section 202(a)(1) of the Clean Air Act (CAA) states that the Administrator of the EPA shall by regulation, prescribe and revise ‘standards applicable to the emission of any air pollutant from any class or classes of new motor vehicles or new motor vehicle engines, which in his judgment cause, or contribute to, air pollution which may reasonably be anticipated to endanger public health or welfare’. In 2003, the EPA denied a petition to regulate emissions of greenhouse gas from motor vehicles. The agency argued that it had no authority to do so for climate change purposes. The following legal dispute illustrated very well how approaches to environmental protection vary from state to state. Whereas the petitioners included, in addition to several environmental groups, a number of states (predominantly governed by the Democrats), the respondents included, apart from the EPA and some automobile industry associations, a roughly equal number of other states (predominantly governed by the Republicans). Crucial then before the Supreme Court was the question of whether carbon dioxide could be qualified under the above-mentioned provision of the CAA as an ‘air pollutant’ that causes ‘air pollution’ that plausibly ‘endangers public health or welfare’.28 If not, the EPA would indeed lack authority to set standards. If yes, it could refuse such regulation only within the limits of the CAA. When the EPA administrator justified the decision not to regulate, at least at that moment, with scientific uncertainty about man-made carbon dioxide emissions as the cause of global warming, the Court therefore did not have to assess the existence of this causal nexus per se but only whether this justification was based on a valid reason under the CAA. In this regard, Justice Scalia argued in his dissent that the text of the provision in question referred, regarding the air-pollutant–air-pollution–danger to public health or welfare link, explicitly to the judgment of the EPA and ‘says nothing at all about the reasons for which the Administrator may defer making a judgment’ (p 552). The majority, by contrast, interpreted this provision more broadly and held that ‘greenhouse gases fit well within the CAA’s capacious definition of an air pollutant’ (pp 528–29). It therefore found the then-current justification for non-regulation inadequate and remanded the issue to the EPA for the provision, if it so desired, of an alternative rationale.

28

Massachusetts v Environmental Protection Agency (EPA) 549 US 497 (2007).

Switzerland 387 An important case of pragmatic collaboration is, at least since the 1990s, water pollution.29 This area of environmental protection revolves, at its core, around permits to discharge pollution into waters under the Clean Water Act, which are based on limits established by both national and state administrative bodies.30 Although the EPA is by default in charge of administrating the permit programme, the agency is authorised to delegate this task. As most environmental protection laws are based on similar permits to pollute up to a certain limit and in a certain manner, it would be practically impossible for the EPA to handle these directly, which naturally makes delegation a widespread phenomenon. If a state demonstrates that it has adequate resources, it may request the delegated administration of a national programme or in some areas even the authority to apply a state programme that is in substance equivalent to the national one.31 Under Section 402b of the Clean Water Act, the EPA is in such a case even obliged to agree to the application of state programmes, which currently exist in as many as 46 states. Even though the EPA is then confined to an advisory role, it still bears ultimate responsibility and therefore maintains, in case of insufficient state implementation, full authority to step in or to withdraw delegation altogether. However, such interventions occur quite rarely due to both procedural difficulties and potential political controversy in response to them.32 Recently, subnational administrative autonomy has rather been threatened by the impact of the economic crisis, which has reduced the ability of states to dedicate adequate resources, ie the crucial requirement for delegation.33 As to intergovernmental cooperation in the horizontal dimension, recent years have witnessed several initiatives. Again the state of California has been one of the driving forces behind a regional framework, the Western Climate initiative established in 2007, which is aimed at developing a regional target for the reduction of greenhouse gas emissions and to achieve that with market-based instruments. An interesting feature of this initiative is that it recognises the above-mentioned fact that pollution knows no boundaries, and it has thus extended cooperation on this issue to also include the subnational entities of neighbouring countries, ie Canadian provinces and Mexican states. 12.2 SWITZERLAND

From a constitutional point of view, environmental protection began to matter in Switzerland in 1971 when Article 24septies was introduced following a positive 29 See AK Gerlak, ‘Federalism and US Water Policy: Lessons for the Twenty-First Century’ (2005) 36 Publius 231, 248. 30 The National Pollution Discharge Elimination System permits the former set technology-based limits and the latter water-quality-related limits. 31 See C Rechtschaffen and DL Markell, Reinventing Environmental Enforcement and the State/ Federal Relationship (Washington, Environmental Law Institute, 2003) 93ff. 32 See JA Mintz et al, Environmental Enforcement: Cases and Materials (Durham, Carolina Academic Press, 2007) 21. 33 See L Paddock and J Bowmar, ‘Environmental Governance in the United States’ in M Alberton and F Palermo (eds), Environmental Protection in Multi-Layered Systems: Comparative Lessons from the Water Sector (Leiden-Boston, Brill, Martinus Nijhoff, 2012) 50f.

388 Environmental Protection r eferendum vote. The fact that over 92.7 per cent of the electorate endorsed a provision that gave ample legislative powers in this area to the national Parliament not only reflects a recent awareness of the need for governmental action, it also illustrates a preference that this action be spearheaded at the national level. This constitutional provision, which formed the basis for the adoption of the Environmental Protection Act in 1983, foreshadowed the future distribution of power. This penchant for comprehensive national authority was then reinforced by Article 74(1) of the new Constitution of 1999, according to which the federal Parliament ‘shall legislate on the protection of the population and its natural environment against damage or nuisance’.34 What further adds to federal predominance is the fact that the notion of ‘environment’ underlying this provision has been interpreted in a very broad manner to include ‘everything linked to the natural surroundings of the human being’.35 Similarly, the ‘damage or nuisance’ that federal legislation must protect against has been construed extensively. Following Article 7(1) of the above-mentioned Environmental Protection Act, these comprise everything except man-made nuisances and infringements that are able to threaten people’s life, health or well-being or to damage their natural environment. Even though Article 74(1) thus provides the national Government with comprehensive legislative power over a number of issues, it is still a concurrent power so that the cantons can make laws as long as the Swiss Parliament does not.36 On the basis of this distribution of competences, the national Government has clearly assumed the leading role through a vast amount of legislation concerning various aspects of environmental protection.37 In doing so, it has occupied so much space in this area that the concurrent competence of the cantons is almost completely empty as a result. As a consequence, the cantons are today basically confined to certain temporary measures and excluded from making any essential regulations with long-term significance. However, the relative weakness of the cantons regarding environmental protection is not always due to restraints from the national Government, but sometimes also originates from self-restraint. 12.B Switzerland 1995: A Valid Popular Initiative for Cantonal Air Pollution Control? In 1993, the Government of the Canton of Lucerne opposed a popular initiative that requested stricter measures for the prevention of air pollution. It formally advised the cantonal Parliament to invalidate this initiative because it would collide with the federal Environmental Protection Act of 1983, which specified immission limits. According to the Government of Lucerne, the cantons would not be allowed to set divergent limits and

34 Several special provisions (Arts 76–80) only supplement and sometimes limit the general rule of Art 74 by adding further details on how and to what end the national Government should use its legislative power in certain specific areas such as the protection of water (Art 76), forests (Art 77), natural and cultural heritage (Art 78) and animals (Arts 79 and 80). 35 See N Schmitt, ‘Environmental Governance in Switzerland’ in Alberton and Palermo (eds), Environmental Protection (2012) 86. 36 See P Hänni, Planungs-, Bau- und besonderes Umweltschutzrecht (Bern, Stämpfli, 2002) 344. 37 For a detailed account of these acts of legislation, see A Petitpierre, Environmental Law in Switzerland (The Hague, Kluwer Law International, 1999).

Switzerland 389

make them the basis of further, more stringent measures. In 1995, the cantonal Parliament indeed invalidated the popular initiative, which was then challenged in court by the Green Alliance as the driving force behind the initiative and by a private citizen. The Court stated that a decree issued by the Government of Lucerne in 1988 already foresaw certain short-term measures against excessive immission from air pollutants and had done so fully in line with both the federal Environmental Protection Act of 1983 and recommendations of the federal Government.38 Compared to that, the popular initiative sought to introduce measures that were clearly more incisive and would have to be implemented at lower threshold levels of immission. The judges held that, against the reasoning of the cantonal Parliament, the above-mentioned 1983 law and other federal legislation refrained from regulating protective measures exhaustively and therefore did not completely preclude legislative action on the part of the cantons. Indeed, federal law had been reluctant to foresee short-term measures because they did not contribute much to achieving its principal aim of ongoing compliance with immission limits. Instead, the 1983 law would implicitly permit cantons to take such measures, as does, more broadly, their competence for environmental policing in situations of crisis. However, such shortterm measures would still have to pass the test of proportionality. The Court deemed proportionality unquestionable for the proposed instrument of an appeal to the population for more frugal use of fossil fuels. But it declared some of the more incisive measures (eg, partial or complete driving bans, room temperature limits for private households, emission limits for industrial plants) partly reasonable and partly unreasonable to achieve their legitimate aim. On this basis, the Court also found that the valid parts of the popular initiative could effectively contribute to the aim of acting against short-term peaks of smog and it assumed that these parts alone would have been equally supported by a sufficient number of citizens. Therefore, the Parliament of the Canton of Lucerne had illegally invalidated a popular initiative in its entirety that was at least in part valid.

Even if the national Parliament thus overall clearly dominates environmental legislation, it is, like its cantonal counterparts, influenced and to some extent limited by popular initiatives and guiding principles. Not surprisingly, the first popular initiatives concerning environmental protection were proposed in the 1970s, with a 1972 initiative against the pollution of water preparing the ground. Ever since, roughly one-fifth of all initiatives have focused on environment issues.39 Other constraints for the discretion of environmental legislation may arise from guiding principles that are partly constitutionally entrenched and partly outlined in the Environmental Protection Act or developed in the case law of the Federal Court.40 Given the extensive pre-emption of legislative space through the national Parliament, the role of the cantons is essentially reduced to implementation. In line with 38

BGE 121 I 334. See Schmitt, ‘Environmental Governance’ (2012) 97. 40 Most important are the principles of sustainability (Arts 2 and 73 of the Swiss Constitution), prevention (Art 74(2) of the Swiss Constitution; Arts 1(2) and 11 of the 1983 Environmental Protection Act; BGE 116 Ib 265), causality (Art 74(2) of the Swiss Constitution; Art 2 of the 1983 law), proportionality (Art 17(1) of the 1983 law), information (Art 6 of the 1983 law) and cooperation (Art 41a of the 1983 law). 39

390 Environmental Protection Article 46(1) of the Swiss Constitution, which stipulates that national laws are, as a general rule, implemented by the cantons, Article 74(3) restates that this rule also applies—except where the law reserves this task to the national level41—to the area of environmental protection. Regarding the process of implementation, however, the national Government is granted by Article 38 of the Environmental Protection Act not only the competence to coordinate the multiple government levels involved, but also vested with comprehensive powers of supervision. In spite of this hierarchy of oversight, relations between national agencies for environmental protection and their cantonal counterparts are in general characterised by a cooperative spirit. The fact that the administrative structure in the cantons in most cases mirrors the subdivision of the Federal Office for the Environment in eight sectors42 has effectively facilitated this collaboration.43 Still, the environmental activism of certain cantons has at times given rise to conflicts with the national Government. While the enshrinement of environmental protection as an objective in certain cantonal constitutions is of rather symbolic importance,44 other provisions have sparked greater controversy. For instance, the more concrete Article 115(2) of the Constitution of Basel-Landschaft states that the canton works towards avoiding the establishment of nuclear power plants on its own territory and in its vicinity. At first glance, this openly clashes with nuclear energy being explicitly designated by Article 90 of the Swiss Constitution as a jurisdiction of the national Government. Yet, after intensive debates both chambers of the national Parliament finally decided not to refuse regarding this constitutional provision of Basel-Landschaft the federal ‘guarantee’ according to Article 51(2) of the Swiss C onstitution.45 The reasoning was that the wording of the provision in question would leave enough room for an interpretation in conformity with the national Constitution. After all, ‘working towards the avoidance of nuclear power plants’ would not mean actually prohibiting them. In practical terms, cantonal referendums against specific administrative decisions have proven to be more of a restraint for federal implementation of environmental law. For example, the Canton of Nidwalden has refused in four votes to store nuclear waste on its territory and thus created a problem for the national Government that is difficult to manage. In view of the deeply rooted Swiss tradition of direct democracy and cantonal autonomy, it has been politically difficult for the latter to disregard both these principles and simply assert itself against the declared will of Nidwalden’s citizenry. In summary, however, this has been one of the rather few instances of open conflict. In most cases, implementation of environmental law by the national and cantonal governments continues to work relatively smoothly and in a cooperative manner.

41

These reservations are listed in Art 41(1) of the 1983 Environmental Protection Act. These are the protection of water, air, the question of waste, hazardous substances, major accidents, soil, noise and the protection of nature and landscape. 43 See Schmitt (n 35) 102. 44 Art 14(1) of the Constitution of Ticino, for eg. 45 See ch 8.2.2. 42

Germany 391 12.3 GERMANY

In Germany, environmental protection was, with the introduction of Article 20a of the Basic Law in 1994, expressly formulated as a constitutionally entrenched objective. This provision explicitly provides that action in this regard shall be taken by ‘the state’ as a whole, ie by all its levels of government. Similar and sometimes more far-reaching clauses have also been inserted in several Länder constitutions. A case in point is Brandenburg, whose Constitution identifies environmental protection, much like Article 20a of the Basic Law, as a goal (Article 38(5)) and thus creates a rather vague objective duty for public authorities to work towards its achievement.46 But beyond that, it also guarantees a concrete subjective ‘right to protection from damage and intolerable danger that result from a change in the natural basis of life’ (Article 38(2)). The scope and significance of any of these constitutional provisions obviously depends to a considerable degree on their interpretation by the Federal Constitutional Court. In a remarkable recent decision, the judges held, for example, that Article 20a would justify, concerning the regulation of genetically modified organisms, certain restrictions of the fundamental rights to economic freedom and private property as long as there is inconclusive scientific evidence about the longterm environmental impact of these organisms.47 While constitutional provisions about environmental protection are thus more or less binding, they do not say anything, of course, about the distribution of power between the national and Länder governments. On a technical level, this distribution is not effected by a single provision regulating the competence for ‘environmental protection’ in its entirety, but through the assignment of jurisdiction over each of its manifold subfields. Put briefly, only very few of these subfields fall within the residuary powers of the Länder under Article 70(1) of the Basic Law. Some important environmental issues, like the ‘production and utilisation of nuclear energy’, are exclusive jurisdictions of the national Government (Article 73) and even more crucial matters, such as waste disposal, air pollution and water resources, are defined as concurrent powers (Article 74).48 The central question is therefore whether this distribution of powers effectively provides the Länder with meaningful legislative autonomy. After all, they are, regarding concurrent powers, authorised to make laws ‘so long as and to the extent that the Federation has not exercised its legislative power’ (Article 72(1)). From a constitutional point of view, the position of the Länder is further strengthened by the deviation clause (Article 72(3)), which empowers them to adopt their own laws even in the face of existing national laws. In fact, five of the six matters subject to the deviation clause are related to environmental protection.49 One of several

46

See G Winter, ‘Environmental Governance in Germany’ in Alberton and Palermo (n 33) 56. 128 BVerfGE 1, 37 (Genetic Engineering Act). 48 For details on the assignment of the various subfields, see Winter, ‘Environmental Governance’ (2012) 59ff. 49 These are hunting, the protection of nature and landscape management, land distribution, regional planning and water resources. The sixth matter is admission to institutions of higher education and their requirements for graduation. 47

392 Environmental Protection r easons for the federalism reform of 2006 to grant the Länder such an escape clause for precisely these matters50 is that they had not been exclusive national domains before. They had been subjects of federal framework legislation to be then specified by legislation of the Länder so that the latter claimed to maintain at least some legislative role beyond 2006.51 Yet, in reality the impact of the deviation clause on subnational autonomy concerning environmental protection is at best ambivalent. On the one hand, the prospect of legal fragmentation motivates the federal Government to consider the views of the subnational governments a priori.52 On the other hand, there is a tendency of the Länder in areas where their own regulations had specified national framework laws not to legislate in deviation of national standards but to accept them. A case in point is nature protection, where the Land of Hesse had earlier required on the basis of a framework law the consent of the nature protection agency to any relevant encroachment. When Article 17 of the Federal Nature Protection Act of 2009, thus after the federalism reform, only required consultation of this institution, the Land did not reintroduce, although it could have done so, through a deviating law, its former stricter standard. To sum up, the national Government has clearly taken the lead in environmental protection legislation as a result not only of its exclusive powers but, even more importantly, also due to its dominant position regarding concurrent competences. Some of the national laws concerning environmental protection have even exerted significant influence on the drafting of EU legislation. At the same time, they have been influenced by the latter in areas like the trading of greenhouse gas emissions.53 As environmental law in Germany is thus basically the result of interrelated legislation from the national and European levels, the Länder are with few exceptions confined to their role regarding implementation. But even in this area, which is, according to Article 83 of the Basic Law, their domain, their role has diminished. First, the execution of national laws by national administrative bodies has been increasing over time, above all through the establishment of organisationally separate but still supervised agencies in areas like environmental data collection or market placement of products. Second, even when the Länder are in charge of implementing environmental legislation, their leeway is often limited. It is therefore crucial whether subnational administrative bodies execute national laws on the basis of a genuine right to do so (Article 84) or only on federal commission (Article 85).54

50 See A Eppler, ‘Das materielle Abweichungsrecht in der deutschen Umweltpolitik als Element föderaler Asymmetrie’ in F Palermo et al (eds), Auf dem Weg zum asymmetrischen Föderalismus? (BadenBaden, Nomos Verlag, 2007) 189ff. 51 See MT Neumann, ‘The Environmental Law System of the Federal Republic of Germany’ (1996) 3 Annual Survey of International & Comparative Law 69, 74. 52 See H Schulze-Fielitz, ‘Umweltschutz im Föderalismus—Europa, Bund und Länder’ (2007) 26 Neue Zeitschrift für Verwaltungsrecht 249, 254. 53 On the interplay of German environmental law with relevant regulations at the European level, see L Knopp, International and European Environmental Law with Reference to German Environmental Law: A Guide for International Study Programs (Berlin, Lexxion, 2008) 77ff. 54 See ch 5.3.

Italy 393

12.C Germany 1990: Federal Instructions for Implementation by the Länder? From 1972 onwards, the Land of North Rhine-Westphalia had conducted a licensing procedure regarding a nuclear power plant in the town of Kalkar. While an earlier case before the Federal Constitutional Court had involved a conflict about building the power station,55 another controversy arose after the Chernobyl disaster in 1986. Even if the Land authority had already approved the safety concept as sufficient, it deemed it necessary to make a reassessment in light of the lessons to be learned from that catastrophic event before granting any further partial permission. This was opposed by the federal environment minister, who claimed, based on an expert commission, that there were significant technical differences between the power plants in Kalkar and Chernobyl and that safety concerns were therefore not justified. The dispute eventually culminated when the federal minister issued an instruction under Article 85(3) of the Basic Law, which North RhineWestphalia challenged as unconstitutional. By virtue of Article 87c of the Basic Law and § 24(1) of the Federal Nuclear Energy Act, the law falls within those federal laws that are executed by the Länder on federal commission. In its judgment on whether the instruction was in line with the constitution, the Federal Constitutional Court made more general observations regarding Länder execution of federal commissions and distinguished between two kinds of executive competences.56 The authority to act vis-a-vis third parties (Wahrnehmungskompetenz) would reside exclusively with the Länder and would not leave room under Article 85 for any federal interference. By contrast, the power to make substantive decisions (Sachkompetenz) is, even though vested primarily in the Länder, not exclusive and may indeed be taken over by the federal Government, which is done precisely through an instruction under Article 85(3). The Court even emphasised that this is neither restricted to exceptional cases nor does it require justification. Instead, it regarded such a takeover as a normal instrument to make, in case of conflict, the federal Government prevail. The instruction of the federal environment minister was therefore deemed fully consistent with the Constitution.

Regardless of commissioned or genuinely their own administration, the Länder typically need to coordinate their activities to a certain extent with national bodies. This is true, in particular, concerning measures with effects on different subfields of environmental protection that are assigned to different government levels. In such cases, one entity is usually entrusted with the final decision under the obligation to prior consultation with, but rarely the consent of, other administrative units.57 This approach is usually successful in preventing potential and resolving acute intergovernmental conflicts. 12.4 ITALY

According to the text of the Italian Constitution, the national Parliament has exclusive power over the ‘protection of the environment’ and of ‘the ecosystem’ (Article 117(2)(s)). At first glance, this provision seems to state unequivocally that 55

49 BVerfGE 89 (Kalkar I). 81 BVerfGE 310 (Kalkar II). 57 For numerous examples, see Winter (n 46) 68. 56

394 Environmental Protection there is no legislative autonomy at all at the subnational level. In reality, national jurisdiction is certainly prevalent but not as exclusive as it may seem. First, there is the possibility that national laws can empower the regions to specify the legal framework laid down in these laws. A notable example is Legislative Decree 152/2006, which grants regions the power to have their own detailed regulations regarding such important issues as waste treatment and environmental impact assessment. Even though the Constitutional Court has in recent judgments defended and emphasised this subnational power of delegated legislation,58 the latter’s significance is in practice often limited because national framework legislation is so detailed and far-reaching that it actually leaves the regions with very little complementary autonomy.59 Second, subnational autonomy extends in several cases to non-delegated regional legislation as well. In concrete terms, Article 117 of the Constitution assigns some environmental issues to the regions as residuary competences, like agriculture or fishing and hunting, and identifies others as falling within the concurrent competences. Concurrency means that legislative power resides, in principle, with the regions but is subject to the determination of fundamental principles in national legislation. Cases in point for such concurrency are ‘land-use planning’ or the ‘national production, transport and distribution of energy’. It is obvious that the definition of these specific matters vis-a-vis the above-mentioned general national power for environmental protection (Article 117(2)(s)) is not an easy task. In fact, it illustrates the great influence of the Constitutional Court. Even before the introduction of Article 117(2)(s) through the constitutional reform of 2001, the Court had identified ‘environmental protection’ through progressive interpretation60 as a broad transversal issue and had divided jurisdiction over it along the pattern of prevailing national or regional relevance.61 Interestingly, this interpretation of the distribution of powers endured, in contradiction to the constitutional text, even during the first years after 2001.62 Only since 2007 have several judgments regarding the new Article 117(2)(s) changed that pattern.63 In this recent case law, the Court confirmed national exclusive competence for those laws that have ‘environmental protection’, construed quite extensively, as their ultimate objective and thus moved away from the criteria of national or regional relevance.64 The regions, on the other hand, are authorised by virtue of their above-mentioned concurrent and residual competences to establish a higher level of environmental protection, but they are prohibited from undercutting the basic national standard.65

58 Italian Constitutional Court No 186/2010 concerning environmental impact assessment, judgments nos 234/2010 and 44/2011 regarding wastewater. 59 See A Ferrara, ‘Le politiche regionali dell’ambiente tra concorrenza e collaborazione’ in A Ferrara (ed), La tutela dell’Ambiente nella Legislazione Regionale (Milan, Giuffré, 1999) 12. 60 See G Cordini, ‘Principi costituzionali in tema di ambiente e giurisprudenza della Corte Costituzionale italiana’ (2009) 5 Rivista Giuridica dell’Ambiente 611. 61 Italian Constitutional Court Nos 273/1998 and 382/1999. 62 See E Orlando, ‘Environmental Governance in Italy’ in Alberton and Palermo (n 33) 144ff. 63 Italian Constitutional Court Nos 367/2007 and 378/2007. 64 See P Maddalena, ‘L’interpretazione dell’Articolo 117 e dell’Articolo 118 della costituzione secondo la recente giurisprudenza costituzionale in tema di tutela dell’ambiente’ (2011) 6 Rivista giuridica dell’ambiente 735. 65 Italian Constitutional Court No 61/2009.

Italy 395 Even though it was the national Parliament that adopted early environmental legislation in the 1960s, especially regarding industrial pollution,66 the regions later became increasingly active. However, the above-mentioned exclusive power of the national Government under Article 117(2)(s), as interpreted since 2007, seems to have restricted subnational legislative autonomy. This has become evident especially in cases of conflict between this provision with concurrent competences (Article 117(3)). 12.D Italy 2009–10: National Limits for Concurrent Regional Power? Several Italian regions have attempted to regulate and sometimes prevent by law the construction of renewable energy plants in certain areas, which they regard as particularly precious and worthy of being protected. One of these initiatives was Article 2 of Law 15/2008 of the Molise Region, concerning the establishment of solar power plants and windmills on its territory. With this act of legislation, the regional Parliament identified certain areas as inappropriate for such constructions because they were deemed particularly precious in terms of their specific landscape and ecosystem. The Italian Constitutional Court held that such legislation indeed falls within the abovementioned concurrent power regarding the ‘production, transportation and national distribution of energy’ under Article 117(3).67 Concurrency would mean, however, that the autonomous law-making power of the regions is limited by the obligation to observe the fundamental principles established by national legislation. The Court thus reinforced the position of the Italian Parliament as the key institution in balancing the conflicting interests of energy supply and environmental protection. This balance enshrined in national law must be accomplished, according to the judges, in compliance with the constitutional principle of loyal cooperation, which mandates certain mechanisms of intergovernmental coordination. Still, the attempt on the part of the Region of Molise to shield certain areas from the construction of renewable energy plants was ruled unconstitutional.68

Beyond the legislative sphere, the exclusive state competence of Article 117(2)(s) of the Italian Constitution even has implications for the implementation of environmental law. In a recent judgment, the Constitutional Court held that this provision also encompasses the power to allocate administrative functions, at least as long as it thereby observes the principles of subsidiarity, differentiation and adequacy (Article 118(1)) and acknowledges the proportionality principle.69 But already in the years before the introduction of Article 117(2)(s) in 2001, the implementation of environmental law had undergone profound and lasting changes.70 First, the administrative structure at the national level had become increasingly integrated in the wake of the

66

For example, Law 615/1966 on atmospheric pollution; Law 319/1976 on water pollution. Italian Constitutional Court No 282/2009. 68 Similar rules in Law 31/2008 of the Region of Apulia were likewise declared unconstitutional in Italian Constitutional Court No 119/2010. 69 Italian Constitutional Court No 225/2009. 70 See JL Bermejo Latre, ‘Le politiche ambientali in Italia nella transizione del ventesimo secolo’ (2008) 5 Rivista Giuridica dell’Ambiente 755. 67

396 Environmental Protection creation, in 1986, of a distinct Ministry for the Environment, which overcame the previous fragmentation into separate sector-based administrations. Second, the decentralisation process of the late-1990s,71 above all Legislative Decree 112/1998, entrusted the regions with significant administrative tasks regarding environmental protection.72 Coordination between these two levels of government is mainly effected in this field as in others through the Standing Conference for the Relations between the State and Regions.73 Moreover, the principle of loyal cooperation shall play, according to the Constitutional Court case law regarding Articles 5 and 114 of the Italian Constitution, a particular role for the implementation of environmental law.74 This principle not only mandates, regarding overlapping competences, the participation of all government levels concerned. It also forms the basis for concrete instruments to realise this participation. In its jurisprudence, the Constitutional Court has developed and put forward a set of graduated instruments that range, according to different levels of concernment, from mere exchanges of information to obligatory formal accords.75 This again vividly illustrates the great judicial influence on the definition of competences not only regarding environmental legislation, but also its implementation. 12.5 UNITED KINGDOM

Unlike the above-mentioned national or subnational constitutions of Germany and Switzerland, the devolution acts in the United Kingdom do not contain with regard to environmental protection any statements of objectives or individual rights. Instead, they focus pragmatically on just the more technical issue of distributing powers.76 With regard to Wales, devolved law-making power regarding environmental protection is circumscribed by comprehensive lists in Schedule 7(1)(6) of the 2006 Government of Wales Act.77 Concerning Scotland and Northern Ireland, by contrast, these powers were transferred not by enumeration but as residuary competences. The field of environmental protection was, for the most part, simply not included in the respective catalogues of matters remaining within the responsibility of the UK Government. As a matter of fact, the environment-related powers retained by the latter are few. They follow first and foremost from the national competence for energy, ie nuclear energy regarding Northern Ireland (Schedule 2(18) of the 1998 Northern Ireland Act) and energy matters more generally concerning Scotland (Schedule 5(II)(D) of the 1998 Scotland Act). 71

See ch 2.3.2. See Title III ‘Environment, Territory and Infrastructures’ of Legislative Decree 112/1998. 73 See ch 6.2.2. 74 For this principle regarding environmental protection, see Italian Constitutional Court Nos 359/1985 and 151/1986. 75 See B Caravita, Diritto dell’ambiente (Bologna, Il Mulino, 2001) 142. 76 See O Pedersen, ‘A Bill of Rights, Environmental Rights and the British Constitution’ (2011) Public Law 577. 77 On how Wales only assumed the power to enact primary legislation without the consent of the UK Parliament in 2011, see ch 11.5. 72

United Kingdom 397 On the basis of this distribution of powers, a considerable amount of subnational legislation has been enacted in less than two decades,78 which now supplements national laws. As a consequence, environmental law is today primarily a ‘statutory matter’79 with little room for doctrines of common law, which naturally limits the role played by the courts. However, in spite of considerable legislative activism at the subnational level, there has been rather little differentiation overall in terms of content so that environmental protection standards have remained rather homogeneous across the UK.80 Of course, this can be attributed to a certain extent to the harmonising effect of the need to comply with EU regulations. But this is only part of the explanation. Interestingly, many of the most characteristic differences today actually do not result from subnational laws but originate from the pre-devolution era. The most notable example is probably the establishment of distinct non-departmental public bodies, like the setting up of the Scottish Environment Protection Agency and the Environment Agency covering England and Wales, in the mid-1990s. As far as Scotland is concerned, however, it has, since devolution, certainly followed its own path more than Wales and Northern Ireland. This is illustrated, for instance, by the fact that it applied the EU-initiated environmental assessment, unlike the other parts of the UK, to all its policies81 and enacted a substantial reform of its water law in 2003.82 The leading position of Scotland over Wales can be attributed to the stronger Scottish tradition of autonomous policy-making and, from a legal point of view, to the more limited legislative competences of the Welsh Assembly. As to Northern Ireland, one major reason for its position for a long time as a relative laggard concerning environmental protection was the neglect of the issue in the face of its violent political conflict between nationalists and unionists. Recently, however, the Northern Ireland Assembly has been very active in reducing this legislative backlog through a set of comprehensive reforms.83 The special political circumstances in Northern Ireland also account in part for asymmetries concerning the implementation of environmental law. The UK administration is still dealing with a number of issues there that have already been in the hands of Scotland’s own bodies for a long time.84 Similar to environmental legislation, Scotland also enjoys a relatively higher degree of autonomy regarding its implementation. Actually, its executive’s scope of powers is even broader than that of the Scottish Assembly because the UK Government transferred, through so-called

78 Regarding Scotland, see CT Reid, ‘Environment and Sustainable Development’ in E Sutherland et al (eds), Law Making and the Scottish Parliament: The Early Years (Edinburgh, Edinburgh University Press, 2011). 79 S Bell and D McGillivray, Environmental Law, 7th edn (Oxford, Oxford University Press, 2008) 94. 80 See CT Reid and A Ross, ‘Environmental Governance in the United Kingdom’ in Alberton and Palermo (n 33) 171f. 81 Environmental Assessment of Plans and Programmes Regulations 2004, SI 2004/1633. 82 Water Environment and Water Services (Scotland) Act 2003. 83 See S Turner, ‘Transforming Environmental Governance in Northern Ireland. Part One: The Process of Policy Renewal’ (2006) 18 Journal of Environmental Law 55; S Turner, ‘Transforming Environmental Governance in Northern Ireland. Part Two: The Case of Environmental Regulation’ (2006) 18 Journal of Environmental Law 245. 84 See Reid and Ross, ‘Environmental Governance’ (2012) 167.

398 Environmental Protection executive devolution by means of an order under Section 63 of the 1998 Scotland Act, quite extensive further responsibilities. For instance, energy matters are, as mentioned above, essentially reserved to the UK in terms of legislation but fall, concerning their executive dimension, to a considerable extent within the responsibility of the Scottish Government. This concerns such important issues as the construction of power stations or transmission lines.85 Executive devolution in these areas has enabled Scotland to follow to some degree its own path concerning energy and climate policy, which is characterised above all by active support for electricity generated from renewable sources and stricter targets for the reduction of greenhouse gas emissions. However, fiscal incentives, which are a crucial instrument of climate policy, fall within the scope of powers of the UK Government.86 As is also the case in other fields, most conformity regarding the implementation of environmental law exists without doubt between England and Wales. This is reflected, for instance, in a common environmental permissions regime based on a joint regulatory framework.87 Intergovernmental cooperation also plays a role in other cases such as the so-called cross-border public authorities regulated in Sections 88–90 of the 1998 Scotland Act. The activities of these bodies are transboundary and touch upon both reserved and devolved matters. In this light, they are obliged to report to both the legislatures of the UK and Scotland, and the UK minister can act, in relation to these bodies in several cases, only after consultation with their Scottish counterpart. 12.6 COMPARATIVE CONCLUSIONS

Since environmental protection emerged as a subject matter too late to be addressed by the distribution of powers of older constitutions, the question of jurisdiction had to be clarified belatedly. From a comparative point of view, it is interesting then how this clarification has taken place. Whereas explicit textual provisions have been introduced in some cases by constitutional amendment or, in the specific British case, through ordinary acts of legislation, the distribution of competences has been shaped in other cases by judicial interpretation. The latter hold true, in particular, for the United States, where the respective jurisdictions over environmental protection have been delineated entirely by constitutional jurisprudence. This once again illustrates the enormous influence of the US Supreme Court and, more precisely, of its Commerce Clause doctrine. Even if the Court has since its seminal 1995 ruling in United States v Lopez restricted the competences of Congress under the affirmative Commerce Clause, it has, through an extensive interpretation of the dormant Commerce Clause, also limited state competences. By applying both these doctrines simultaneously, the judges have thus overall limited the power of the entire legislative

85 Scotland Act (Transfer of Functions to the Scottish Ministers etc) Order 1999, SI 1999/1750, Art 2 and Sch 1. 86 Climate Change Levy under the Finance Act 2000. For an analysis, see BJ Richardson and KL Chanwai, ‘The UK’s Climate Change Levy: Is It Working?’ (2003) 15 Journal of Environmental Law 39. 87 Environmental Permitting (England and Wales) Regulations 2010, SI 2010/675.

Comparative Conclusions 399 branch. This contrasts with the entrenchment in other countries of environmental protection as a constitutional objective, albeit not always judicially enforceable, but to be furthered by the ordinary legislator (and the executive) anyway. Cases in point are Article 20a of Germany’s Basic Law, as well as similar provisions in the constitutions of some German Länder and Swiss cantons, which virtually urge parliaments to take legislative action in this field. To be sure, the United States is, through the exclusively judicial definition of environmental protection powers, an outlier. However, cases in which the roles in this emergent policy field have been basically determined by written law also vary significantly concerning the degree of this determination. In Italy, for instance, the explicit constitutional provisions, namely the exclusive competence of the national Government (Article 117(2)(s)) and several environment-related concurrent powers of the regions (Article 117(3)), have remained in uneasy tension in relation to each other so that they could not provide sufficient certainty per se. It thus needed the Constitutional Court to further clarify, in several important decisions from 2007 onwards, the actual distribution of powers. It did so, in essence, by interpreting the legislative autonomy of the regions in a rather restrictive way. By contrast, Article 74(1) of the 1999 Swiss Constitution much more clearly affirmed the pre-existing legislative predominance of the national Government. Although the cantons do have concurrent law-making powers regarding environmental protection, they are effectively preempted by abundant national legislation. Overall, they are largely confined, according to the Federal Court, to short-term measures of environmental policing in the event of acute crises.88 From a technical perspective, the German Basic Law applies, compared to Switzerland, a very different approach but still arrives at similar outcomes. In Germany, there is no single main provision making environmental protection in broad terms a federal prerogative. Instead, it divides this vast area into various subfields and mostly identifies them as of exclusive national competences (Article 73) or concurrent powers (Article 74), thereby leaving little room for residuary powers of the Länder (Article 70(1)). Evidence as to whether the deviation clause (Article 72(3)) has effectively promoted the adoption by the Länder of their own legislation regarding environmental protection is mixed at best. By contrast, the United Kingdom not only provides the possibility to adopt differentiated legislation of subnational entities, but it also endows them with differentiated legislative powers. This profound asymmetry is, of course, related to the bilateral case-by-case approach inherent in devolution. As a result, Scotland and Northern Ireland enjoy today, apart from the important national competence for energy, very extensive law-making powers, compared to the subnational entities of the classical federations examined in this section. Even though Scotland has clearly displayed the most activism, differentiation between the UK’s subnational entities has remained rather limited altogether. To some degree, this is certainly due, as in Germany and Italy, to the harmonising effect of EU law,

88

BGE 121 I 334.

400 Environmental Protection whose intensity depends on the specific subfields of this broad area.89 In the United States, on the other hand, both the absence of significant supranational environmental regulation and, presumably, an exceptionally strong ideological polarisation between the states have arguably given rise to considerable horizontal differentiation. The deep divisions between leaders and laggards in environmental protection are vividly illustrated by the two rival factions of states in Massachusetts v EPA,90 as well as by states’ own climate change policies and stricter environmental standards on the part of California and others. A particularly remarkable example of differentiated autonomy is the special position of California under the Clean Air Act, which foresees two parallel regimes with other states essentially having the choice to follow either federal standards or Californian standards. Notwithstanding the differences between the cases analysed in terms of the distribution of powers and degree of asymmetry, environmental legislation has overall largely remained a national domain. This has been justified, from the earliest US laws of the late 1960s and the Swiss 1971 referendum in favour of such predominance until the recent case law of the Italian Constitutional Court,91 by similar reasoning. Environmental protection would require—so the argument goes—basic countrywide minimum standards, which subnational entities are allowed in some cases to exceed, but never to undercut. Since environmental protection is, as a policy field, perhaps more than any other, transboundary and transversal in character, it seems undeniable that the adoption of certain regulations on larger territorial scales, which may be national or even supranational, indeed makes good sense. Moreover, one must not forget, also with regard to the EU, that harmonisation of environmental law in at least certain subfields seems, as already underlined in the United States around 1970,92 quite relevant for the effective working of a common market. From a comparative perspective, autonomy is by and large clearly more extensive regarding implementation than in the legislative sphere. Yet, there are considerable differences between the cases examined as to both the extent of this autonomy and the obstacles to it. While Article 74(3) of the Swiss Constitution declares the implementation of environmental law to be with some exceptions a matter for the cantons, ordinary national legislation still foresees certain instruments of supervision for the national Government (Article 38 of the 1983 Environmental Protection Act). Germany is similarly marked by a combination of limited legislative powers of subnational entities with rather comprehensive, albeit not unrestrained, responsibility for implementation (Article 83). Restraints on autonomy arise in particular regarding the execution of federal environmental law on federal commission (Article 85), which allows the national Government to oversee the appropriateness of implementation and, according to the Federal Constitutional Court,93 to issue instructions regarding substantive decisions.

89

See, in particular, L Krämer, EU Environmental Law, 7th edn (London, Sweet & Maxwell, 2011). Massachusetts v EPA (n 28). 91 Italian Constitutional Court No 61/2009. 92 See Butler and Macey, Using Federalism (1996) 21ff. 93 81 BVerfGE 310 (Kalkar II). 90

Comparative Conclusions 401 In spite of certain constraints, the national constitutions reflect, in both the Swiss and German cases, a decision of principle in favour of subnational implementation as a general rule. In Italy, by contrast, the Constitutional Court recently held that Article 117(2)(s) would vest the ordinary national legislator with the power, subject to certain principles, to assign executive functions to the different levels of government.94 In the United States and the United Kingdom, responsibility for the implementation of environmental law is, in the absence of explicit constitutional entrenchment, transferred, by different legal instruments, directly from the national government to the subnational entities. In the UK case, this instrument is besides the complete devolution of a matter, also concerning its legislative dimension, the transfer of merely executive functions from a UK minister to a member of a devolved government. This executive devolution has augmented the competences of Scotland in particular. But as this transfer is merely effected through an Order in Council tabled by the cabinet minister responsible, Scotland is far from the firm entrenchment of implementation as a normally subnational prerogative in the above-mentioned Swiss and German cases. In the United States, subnational executive power is not derived, unlike in the UK, directly from a cabinet member but from the EPA. Although this agency’s Administrator is a high-ranking official of the national executive appointed by the US President and approved by Congress, he/she is not part of the Cabinet. Despite considerable direct federal administration, the EPA has partly delegated the implementation of national or sometimes equivalent state programmes to the states. Even if subnational entities enjoy more or less comprehensive executive autonomy on the basis of constitutional entrenchment, ordinary legislation, devolution or delegation, they are naturally still forced to engage in vertical intergovernmental cooperation with the national Government. This follows not only from the latter’s above-mentioned powers of supervision and/or intervention, but also from the character of environmental protection as a transversal and transboundary issue. As such, it requires and indeed often naturally brings about patterns of intergovernmental collaboration that are more or less efficient. In any event, the strengthening of these patterns is, for such a broad and fragmented field, much more crucial than attempts, anyway futile, to neatly subdivide competences and assign them between levels of government. Thus, the shift from dual to (necessarily) cooperative federalism seems to be particularly evident in environmental protection. In practice, the cases analysed have developed quite diverse mechanisms of collaboration. In Switzerland and Germany, these are commonly deemed particularly successful, not least because their proper functioning is, in addition to the largely efficient design of these structures, sustained by an underlying cooperative spirit. As to the Italian case, the Constitutional Court has ruled from the mid-1980s onwards that the principle of loyal cooperation, derived from Articles 5 and 114 of the Constitution, would be central in particular to the implementation of environmental law.95 It even took this principle as a starting point to elaborate a system of

94 95

Italian Constitutional Court No 225/2009. See Italian Constitutional Court Nos 359/1985 and 151/1986.

402 Environmental Protection specific mechanisms of intergovernmental collaboration. This example epitomises the crucial role that the courts sometimes play in developing law, in addition to their basic function of deciding disputes over the respective areas of competence. Often resulting from the failure of cooperation, such disputes are, in light of the transversal and transboundary nature of environmental protection, a rather normal phenomenon. Consequently, the significant function of the ‘courts as ultimate conflict-solvers has to be considered as a physiologic rather than a pathologic element of environmental governance’.96 Conflict and cooperation are, of course, similarly salient in the horizontal dimension of intergovernmental relations, ie the dimension between subnational entities. In this regard, the cross-border public authorities under Sections 88–90 of the 1998 Scotland Act are a good example of bilateral mechanisms that deal with both devolved and reserved matters of shared concern beyond subnational boundaries. By contrast, the United States has, in recent years, witnessed quite significant multilateral collaboration, which is by comparison also less formalised and less linked to the national Government. While alliances such as the Western Climate Initiative and others only comprise several US states, they interestingly decided to extend intergovernmental cooperation beyond international boundaries to subnational entities of neighbouring countries. Thus, they are reminiscent of the potential especially in the field of environmental protection for cross-border cooperation.97 But the fact that these alliances are territorially limited also illustrates that, precisely in this field, the policy preferences of subnational entities vary significantly. This holds true, as mentioned above, in particular for the United States but similarly for other cases and thus makes it a challenge to strike a balance between subnational autonomy and national, or even supranational, coherence.

96 97

M Alberton and F Palermo, ‘Concluding Remarks’ in Alberton and Palermo (n 33) 526. On cross-border cooperation, more generally, see chs 1.2.3 and 14.

13 Immigration and Migrant Integration

T

HE ISSUES OF immigration and the integration of migrants have come to the forefront of the political agenda in a number of countries, and they are likely to remain there. Quite obviously, this increasing importance is linked to the steadily growing share of the foreign-born population in many countries, in Europe even more so since the onset of the migrant crisis in 2015. In numerical terms, the world’s migrant population increased from 100 million, or 2 per cent, in 1980, to 214 million, or 3.2 per cent, in 2010.1 Some classical federations such as the United States, Canada and Australia have been traditional immigration countries right from the outset and continue to have exceptionally high net migration rates. In Europe, by contrast, many traditional countries of emigration, most notably Spain and Italy, have experienced a sharp increase in their foreign-born population during the last three decades. These developments have gradually transformed Europe into a ‘continent of immigrants’.2 Apart from the mere extent of international migration, its complexity in terms of mobility patterns and profiles of migrants has also greatly increased and has given rise, at least in the main metropolitan areas, to what has been termed ‘super-diversity’3 or ‘diverse diversity’.4 This phenomenon of quantitatively and qualitatively increased international migration has evolved unevenly not only in different continents and states but also in different subnational entities of the same state. Some have become extremely multicultural in the descriptive sense of the term,5 while others have been less impacted. Although immigration (sub-section (1) of the various country case studies within this chapter) and integration (sub-section (2)) as policy fields are obviously closely connected, it is imperative to draw an analytical distinction between them that is today widely recognised. Whereas the first area deals with ‘policies of immigration’, the second focuses on ‘policies for migrants’.6 As a consequence, immigration is essentially concerned with the regulation of international migration flows, which boils down to authoritative decisions on the selection and admission of migrants. 1 See K Koser et al, World Migration Report 2010—The Future of Migration: Building Capacities for Change (Geneva, International Organization for Migration, 2010). 2 M Okólski, ‘Europe, a Continent of Immigrants’ in M Okólski (ed), European Immigrations. Trends, Structures and Policy Implications (Amsterdam, Amsterdam University Press, 2012). 3 S Vertovec, ‘Super-diversity and its Implications’ (2007) 30 Ethnic and Racial Studies 1024. 4 KG Banting, ‘Canada’ in C Joppke and FL Seidle (eds), Immigrant Integration in Federal Countries (Montreal, McGill–Queen’s University Press, 2012) 82. 5 The descriptive dimension of multiculturalism refers to the mere existence of cultural diversity, while its normative dimension means that cultural diversity is actively endorsed as something valuable and indeed desirable. See A Heywood, Key Concepts in Politics (Basingstoke, Palgrave Macmillan, 2000) 227. 6 T Hammar, Democracy and the Nation State (Aldershot, Avebury, 1990).

404 Immigration and Migrant Integration The integration of migrants, sometimes also called inclusion of migrants,7 then in terms of time sequence follows immigration. Moreover, it is much more complex than immigration, as we can distinguish a social, economic and political dimension of integration.8 Though having set points in time, such as the moment of naturalisation regarding its political dimension, integration is a longer-term process with a scope that is much more difficult to define. Rather than being a single and precisely circumscribed policy field, integration is a cross-sectoral responsibility, whose aforementioned dimensions are encompassed by many areas such as the labour market, education, housing, welfare, language use and citizenship. Although the roles of different government levels are still considered under-researched for both immigration and integration,9 this section cannot cover all aspects pertaining to the latter area. Therefore, it focuses, regarding social and economic integration, on welfare, education, housing, the labour market and language training and, in terms of political integration, on naturalisation, multicultural policies and voting rights at the subnational and local level. 13.1 SWITZERLAND

(1) Immigration to Switzerland is clearly a federal domain both in terms of the constitutional distribution of powers and the related political processes. Article 121(1) of the Swiss Constitution makes it clear that the national Parliament is responsible for ‘legislation on entry to and exit from Switzerland, the residence and the permanent settlement of foreign nationals and on the granting of asylum’.10 As to the process of drafting ordinary legislation on immigration, two things are remarkable that to some degree diminish the power of the Parliament: the vital importance of direct democracy and of pre-parliamentary bargaining, which involves the Federal Office for Migration in the Federal Department of Justice and Police, experts’ commissions and interest groups.11 These processes have given rise to a Swiss immigration policy that has been characterised since a major change in the early 2000s by a comparatively lenient approach towards EU and European Free 7 The use of the terms ‘integration’ and ‘inclusion’ varies significantly among international organisations, eg the European Union and the Council of Europe, among states and even, such as in Italy, among subnational entities within the same state. Yet, ‘integration’ is preferred in this chapter because it seems to be the overall prevalent expression and is used by several standard publications like Joppke and Seidle (eds), Immigrant Integration (2012). 8 See K Kössler, ‘Immigration and Integration in Multilevel Systems: A Challenge between Regional Autonomy and Intergovernmental Cooperation’ in R Medda-Windischer and A Carlà (eds), Migration in Territorial Autonomous Units: The Case of South Tyrol and Catalonia (Leiden, Brill–Nijhoff, 2013). 9 See G Zincone and T Caponio, ‘The Multilevel Governance of Migration’ in R Penninx et al (eds), The Dynamics of International Migration and Settlement in Europe: A State of the Art (Amsterdam, Amsterdam University Press, 2006) 291. 10 It is striking that beyond circumscribing the federal authority in the immigration field, the Swiss Constitution then dedicates five paragraphs to regulating in detail such issues as offences by migrants, their expulsion and ban on re-entry. 11 Whereas pre-parliamentary negotiations are in general more compromise-oriented, campaigning related to direct democracy decisions has contributed to the intense politicisation of immigration issues. See H Mahnig and A Wimmer, ‘Integration without Immigrant Policy: The Case of Switzerland’ in F Heckmann and D Schnapper (eds), The Integration of Immigrants in European Societies: National Differences and Trends of Convergence (Stuttgart, Lucius & Lucius, 2003).

Switzerland 405 Trade Association (EFTA) citizens in the wake of the 1999 Bilateral I Agreement on the free movement of persons12 and an increasingly restrictive tendency towards third-country nationals. According to the federal Aliens Law of 2006, the latter are only admitted after a favourable assessment of the economic and demographic needs and if their special professional skills are indispensable.13 (2) Despite a considerable influx of foreign workers soon after the end of World War II, Switzerland denied for a long time the obvious fact that it was an immigration country. Although the rotation model, which foresaw the regular substitution of workers instead of permanent residence, was abandoned in the early 1960s, proactive integration remained absent. It was assumed that integration would occur naturally over time and through the general public services provided at cantonal and local level so that a proactive policy would be dispensable. Only the 1990s saw a shift towards genuine integration measures and official reports started to underline that these measures should be taken at all levels of government.14 As a matter of fact, this shift was to a considerable extent the result of subnational initiative because it was closely linked with debates on urban development in general and in particular with the newly drafted integration guidelines of the three city cantons of Bern, Zurich and Basel-Stadt. Yet, the federal Government also increased its activities in this field. First, in 1999 it inserted Article 25a, the so-called integration article, in the old Aliens Law, which foresaw joint funding with the cantons and municipalities for social integration initiatives. Then it adopted a new Aliens Law in 2006. Whereas this piece of legislation promotes the integration of permanent legal residents, it marks a turn to a more restrictive approach towards third-country nationals.15 In financial terms, this promotion is reflected in the allocation of funds by the national Government for the language and integration courses provided by the cantons, which go along, however, with its increasing influence on the content of these courses. By contrast, concerning many integration-relevant general public services, the cantons continue to play the leading role.16 Cases in point are the social assistance regimes,17 which are a cantonal responsibility with only a few common basic principles, and, above all, primary and secondary education. The different school systems make migrants adopt the dominant language of the canton and deviate in some cases a great deal from the federal recommendations regarding the educational integration of immigrant children.18 12 But see the Swiss federal popular initiative ‘Against mass immigration’ which also affected EU and EFTA citizens and was approved in 2014 by a narrow majority. 13 See N Blattner and G Sheldon, ‘Foreign Labour, Growth and Productivity: The Case of Switzerland’ in I Gordon and AP Thirlwall (eds), European Factor Mobility: Trends and Consequences: Proceedings of the Conference of the Confederation of European Economic Associations (Houndmills, Macmillan, 1989). 14 See G D’Amato, ‘Switzerland’ in Joppke and Seidle (n 4) 176ff. 15 Art 64(1)(b) in conjunction with Art 5(1)(b) provides, for example, that third-country nationals who are no longer financially self-sufficient can be expelled. 16 See N Wichmann et al, Gestaltungsspielräume im Föderalismus: Migrationspolitik in den Kantonen (Bern-Wabern, Eidgenössische Kommission für Migrationsfragen, 2011). 17 On the definition of social assistance, see ch 11.6. 18 See Mahnig and Wimmer, ‘Integration’ (2003) 144.

406 Immigration and Migrant Integration As far as political integration is concerned, the cantons play an important role regarding voting rights, which some of them, typically the predominantly Frenchspeaking cantons, have extended to foreigners in respect of local and sometimes cantonal elections.19 From a comparative perspective, the involvement of the cantons in the area of naturalisation is rather exceptional. It follows from Article 37(1) of the federal Constitution that every Swiss is a ‘multilevel citizen’ because nationality relates not only to the national level but also to the cantonal and municipal levels.20 Thus, a candidate for naturalisation has to meet criteria stipulated at each of these levels. The federal role is thereby restricted, according to Article 38(2) of the Swiss Constitution, to granting the final naturalisation permit and establishing minimum conditions. The latter are regulated in the Citizenship Law, whose partial liberalisation was rejected in a 2004 referendum not only by a majority of all Swiss voters but also by a clear majority of the single cantonal electorates.21 Therefore, this law still stipulates an exceptionally long residence requirement of 12 years (Article 15) and a comprehensive assessment of the applicant’s suitability for citizenship (Article 14). The latter also includes, in addition to certain clear criteria, like adherence to the Swiss legal system and the absence of a security risk, some rather vague terms, namely ‘integration in the Swiss living conditions’, as well as ‘familiarity with Swiss habits and customs’. In addition to these federal requirements, the cantons establish very different conditions like residence in the canton for two years in Geneva and for the entire 12-year period in Nidwalden. By the same token, the requirements stipulated at the local level vary greatly. As the municipalities assess not only these but also the two above-mentioned vague federal conditions, they have much leeway in decision- making. As a consequence, the approval at the local level, granted either by the municipal administration or the entire citizenry, is normally most difficult to obtain.22 The latter decision-making method revealed tensions between local direct democracy, on the one hand, and rule of law and equal treatment, on the other. These tensions are reflected in the 2003 judgment of the Federal Court regarding the much debated Emmen case.23

19 See S Cattacin and K Bülent, ‘Die Politik der Integration von Migrantinnen und Migranten im föderalistischen System der Schweiz’ in L Akgün and D Thränhardt (eds), Integrationspolitik in föderalistischen Systemen—Jahrbuch Migration 2000/2001 (Münster, LIT, 2001) 206. 20 Art 37(1) reads as follows: ‘Any person who is a citizen of a commune and of the Canton to which that commune belongs is a Swiss citizen.’ 21 Similar referendums about liberalisation had already failed in 1983 and 1994. The changes proposed in 2004, namely easier access to Swiss nationality for the second generation and the ius soli for the third generation, found even less support than those proposed in the referendum a decade earlier. It is interesting to note that apart from Basel-Stadt all other predominantly German-speaking cantons, which had been favourable to liberalisation in 1994, switched camps in 2004. 22 See M Helbling, ‘Naturalisation Politics in Switzerland: Explaining Rejection Rates at the Local Level’ in T Caponio and M Borkert (eds), The Local Dimension of Migration Policymaking (Amsterdam, Amsterdam University Press, 2010) 35ff. 23 This case (BGE 129 I 217) and the case BGE 130 I 140 of 2004 are discussed in detail in D Thürer and M Frei, ‘Einbürgerung im Spannungsfeld zwischen Demokratie und Rechtsstaatlichkeit: Zu zwei historischen Entscheiden des Schweizerischen Bundesgerichts’ (2004) 123 Zeitschrift für Schweizerisches Recht 205.

Canada 407

13.A Switzerland 2003: Secret Ballot Decisions on Naturalisation Requests? In the small town of Emmen, located in the canton of Luzern, the citizenry approved on 12 March 2000 by secret ballot all naturalisation requests by Italians and rejected all those by candidates from the former Yugoslavia and Poland. While several of the rejected applicants regarded this method of decision-making as discriminatory, apologists for the procedure pointed to the Swiss traditions of local autonomy and direct democracy. The Federal Court found that the secret ballot decision was unconstitutional for two reasons.24 First, it could not guarantee the prevention of arbitrary decisions, which is at the heart of the concept of the rule of law. In particular, the procedure did not fulfil the obligation to state reasons enshrined in Article 29(2) of the Swiss Constitution, whose purpose is precisely to safeguard against arbitrariness. Second, the decision violated the principle of non-discrimination guaranteed by Article 8(2) of the Swiss Constitution because the secret ballot did not give any justification for positive outcomes in some cases and negative ones in others. 13.2 CANADA

(1) With regard to immigration competences, the Canadian provinces in general and Quebec in particular certainly have an exceptional position that has gradually evolved over time. Though this subject matter was identified by Section 95 of the Constitution Act 1867 as one of only two concurrent competences with federal paramountcy, the provinces soon left the leading role to the federal Government.25 In the 1960s, however, the increasing recruitment of migrants by Quebec, and to a lesser degree Alberta, Manitoba and Saskatchewan, foreshadowed provincial emancipation in this field. Quebec became particularly active because the tendency of migrants to assimilate to the English-speaking community in the province rendered the French-speaking population more opposed to the influx of foreigners than other Canadians.26 Instead of responding to this mood with a call for a reduction in immigration, the provincial Government chose to demand a role in steering it through involvement in the selection of migrants. To this end, it established its own government department for immigration in 1968 and negotiated with the federal Government successive bilateral agreements from 1971 onwards.27 These accords steadily extended their competences from mere consultation to full control over the selection of all economic migrants to the province and the permission of preferential recruitment of Frenchspeaking migrants.28 These two significant innovations of the 1991 Canada–Québec 24

BGE 129 I 217. F Hawkins, Canada and Immigration: Public Policy and Public Concern, 2nd edn (Montreal, McGill–Queen’s University Press, 1988). 26 See R Lambert and J Curtis, ‘Opposition to Multiculturalism among Québécois and English- Canadians’ (1983) 20 Canadian Review of Sociology and Anthropology 193. 27 These are the Lang-Cloutier Agreement of 1971, the Andras-Bienvenue Agreement of 1975 and the Cullen-Couture Agreement of 1978. 28 According to the points system currently in force, candidates for selection can be awarded 0 to 6 points for their English skills, but 0 to 16 points for their French skills. 25 See

408 Immigration and Migrant Integration Accord effectively reduced the corresponding federal competences to the selection of remaining classes of migrants, ie family members and refugees, as well as to security checks and final admission concerning all migrants. Initially, the special immigration powers granted to Quebec did not provoke the envy of other provinces, which regarded the subject matter in the absence of a comparable cultural-linguistic imperative rather as a burden. Instead, it was the federal Government’s interest in avoiding excessive asymmetry that initiated a broader decentralisation process involving all provinces. This started with mandatory consultation of the provinces regarding the federal immigration plan under Section 7 of the 1976 Immigration Act and several, albeit not far-reaching, intergovernmental agreements.29 Only in the aftermath of the above-mentioned 1991 Canada–Québec Accord did a number of provinces join Manitoba, a campaigner for increased immigration powers for decades, in expressing their interest in the matter. The common goal of these provinces was to attract more migrants because their hitherto extreme concentration in Ontario, British Columbia and Quebec threatened to reduce their economic and political weight. In response, the federal Government created, in 1995, the Provincial Nominee Program (PNP) as a framework for bilateral agreements that have, in the meantime, been concluded with all provinces. The PNP establishes a second track for the selection of economic migrants by empowering the provinces to identify nominees according to their own selection criteria, which in fact differ substantially from the federal standards.30 Although a move by the federal Government towards recentralisation in 2012 forces semi- and low-skilled PNP candidates to comply with Canada-wide minimum language standards, the less appealing provinces still have considerable leeway to attract a sufficient number of migrants with the profiles they need to fill their labour-market gaps. In general, the combination of priority processing of provincial nominees versus federally selected migrants and removing the cap on provincial nominees in 2008 resulted in an increased share of provincial nominees Canada-wide. In recent years, they have constantly more or less equalled the number of economic migrants admitted through the federal track. In line with the rationale of the PNP, the less attractive provinces rely much more on their own nominees, who, in 2012, accounted for 98 per cent of all newcomers to Prince Edward Island and 94 per cent to Manitoba. Ontario, by contrast, is comfortable with both its share of migrants and their profiles such that the provincial share of own nominees is only 3 per cent there.31 To guarantee a long-term steering effect, the challenge is then to retain these migrants in the less attractive provinces. As all people with the status of a permanent resident have, according to Section 6(2) of the Charter of Rights and Freedoms, the right to free movement within Canada,32 these provinces

29 See N Kelley and MJ Trebilcock, The Making of the Mosaic: A History of Canadian Immigration Policy (Toronto, University of Toronto Press, 1998) 392. 30 Citizenship and Immigration Canada, ‘Evaluation of the Provincial Nominee Program’ (2011) 31f www.cic.gc.ca/english/resources/evaluation/pnp/index.asp. 31 Citizenship and Immigration Canada, ‘Annual Report to Parliament on Immigration’ (2012) 21 www.cic.gc.ca/english/pdf/pub/annual-report-2012.pdf. 32 Sec 6(2) reads as follows: ‘every person who has the status of a permanent resident of Canada has the right (a) to move to and take up residence in any province; and (b) to pursue the gaining of a livelihood in any province’.

Canada 409 have had an incentive to convince newcomers by offering an effective and migrantfriendly framework of integration. (2) Similar to immigration, the area of migrant integration is characterised by a multitude of intergovernmental agreements that supplement scarce constitutional foundations. As far back as the nineteenth century, integration was not seen as a field for governmental action. Thus, it is perfectly understandable that the Constitution Act 1867 remains largely silent on the issue except for declaring ‘naturalization and aliens’ to be a federal prerogative (Section 91(25)). While naturalisation meant at that time the granting of British-subject status to an individual born outside the Empire, Canadian citizenship was only introduced by a federal act in 1947. Since then, the national Government’s approach to citizenship has been much less restrictive than in many other countries with a permanent-residence requirement of only three years and relatively high pass rates on the citizenship test. The naturalisation regime applied by Citizenship and Immigration Canada also impacts on another element of political integration, namely access to the rights enshrined in the Canadian Charter of Rights and Freedoms. Although most of these, like the non-discrimination rule (Section 15), are not bound to citizenship, some such as voting rights (Section 3) and minority-language education rights (Section 23) are restricted to Canadians. Overall, because of its paramount importance,33 the Canadian Charter certainly serves, in respect of migrants, as a powerful instrument of political integration into a pan-Canadian community of people holding the same rights.34 An equally strong instrument of political integration is multiculturalism, which was adopted as official policy during the 1970s and is reflected in the 1988 Multiculturalism Act. The targeted funding programmes in this context are considered to have contributed significantly to migrants’ participation in Canadian political institutions more quickly and effectively by supporting their self-organisation capacity and familiarity with Canadian politics.35 While most provinces do not object to the unquestionable dominance of the federal Government regarding political integration, each of its above-mentioned three main instruments arouses controversy in Quebec. There, the approach of interculturalism is proposed as an alternative to multiculturalism, and claims have emerged for the province to have its own constitutionally entrenched bill of rights, as well as officially recognised Quebec citizenship.36 As far as social and economic integration is concerned, a distinction has to be made between settlement services and longer-term integration.37 The latter is overwhelmingly within the responsibility of the provinces and includes, for instance,

33

See ch 10.3. A-G Gagnon, ‘Immigration in a Multinational Context: From Laissez-faire to an Institutional Framework in Quebec’ in R Zapata Barrero (ed), Immigration, Self-Government of Minority Nations (Brussels, Peter Lang, 2009). 35 See I Bloemraad, Becoming a Citizen: Incorporating Immigrants and Refugees in the United States and Canada (Berkeley, University of California Press, 2006) 6. 36 See A-G Gagnon and R Iacovino, Federalism, Citizenship and Quebec: Debating Multinationalism (Toronto, University of Toronto Press, 2006) 122. 37 See Banting, ‘Canada’ (2012) 95f. 34 See

410 Immigration and Migrant Integration traditionally quite successful integration in primary and secondary education. Other typical provincial prerogatives are vocational training and social assistance, which are, in light of the recently decreasing economic success of migrants,38 more and more important. While the provinces are crucial regarding the regulation of access to the labour market, a recent Supreme Court ruling suggests that sometimes the federal level may also play a role. 13.B Canada 2001: Non-lawyers as Counsellors in Immigration Processes? Jaswant Singh Mangat was a legal adviser in immigration processes working for Westcoast Immigration Consultants Ltd, a company located in British Columbia. Mangat offered his services, although he had not studied law in Canada and was not a member of the British Columbia Law Society. This prompted the latter association to challenge the validity of Sections 30 and 69(1) of the federal Immigration Act, which allowed representation in the immigration process by non-lawyers, because they would violate the provincial Legal Profession Act that restricts appearance before administrative tribunals to lawyers. The Supreme Court dismissed this appeal.39 According to its reasoning, the decision regarding who foreigners may choose as a representative pertains to the delineation of the procedural rights of those individuals in quasi-judicial proceedings and as such falls within the exclusive federal competence to regulate ‘naturalization and aliens’ under Section 91(25) of the Constitution Act 1867. As it had been widely thought until then that laws concerning the process of admitting foreigners into Canada would pertain to the concurrent immigration jurisdiction under Section 95 of the same act, the decision effectively reduced the scope of provincial jurisdiction.

In contrast to the aforementioned general public services like education or social assistance, the largely federally funded settlement programmes, which involve language courses or the twinning of migrants with volunteers as informal mentors, are only available in the period before naturalisation and are usually restricted to permanent residents.40 Provincial competences concerning these settlement services are usually determined in the same bilateral agreements that regulate their powers on immigration. In this regard, Quebec again has an exceptional position, as the federal Government, through the 1991 Canada–Quebec Accord, withdrew completely from the regulation and provision of settlement services. It did so by allocating federal funds for these henceforth provincial services according to a quite favourable agreement that foresees an annual increase according to the growth rate of the Canadian Government’s overall spending. Moreover, the federal Government did not stipulate any reporting and accountability duties, in fact not even an obligation to spend the funds on settlement services.41

38 See M Frenette and R Morissette, ‘Will They Ever Converge? Earnings of Immigrant and Canadianborn Workers over the Last Two Decades’ (2005) 9 International Migration Review 363. 39 Law Society of British Columbia v Mangat [2001] 3 SCR 113. 40 Only some provinces extended such services to temporary residents like seasonal workers, international students and refugee claimants. 41 See Banting (n 4) 95f.

Australia 411 Devolution of settlement services to other provinces was launched in the mid1990s in the context of a financial crisis in order to offload a financially burdensome task.42 In 1999, after complete control over settlement services had been transferred to Manitoba and British Columbia, however, the federal Government changed course and began to favour more limited devolution. This gave rise to a number of different models such as a co-management arrangement in Alberta, a tri-level model including the City of Toronto in Ontario and mere consultation of the provinces in the remaining cases. Similar to the above-mentioned immigration competences, 2012 saw partial recentralisation, as the federal Government cancelled the particularly far-reaching settlement agreement with Manitoba and British Columbia with reference to the importance of integration for nation-building and the need for more symmetry.43 13.3 AUSTRALIA

(1) On the basis of its exclusive jurisdiction over ‘naturalization and aliens’ (Section 51(xix) of the Australian Constitution), the federal Government clearly dominates the immigration field through the powerful Department of Immigration and Citizenship.44 In the 1990s, however, the territorial concentration of migrants in urban areas and, above all, in the state of New South Wales45 gave rise to a labour shortage in certain geographical areas and technical professions. This eventually led to a growing federal–subnational consensus about a partial decentralisation of immigration. The joint response to these needs was the 1996 initiative towards a State S pecific Regional Migration programme (SSRM), which ultimately established a second track of skilled migrant selection. Bearing the specific regional needs in mind, states and employers can nominate candidates who would not meet the points-based requirements of the federal Government for final selection by the Department of Immigration and Citizenship. These people can then be admitted either as temporary or permanent workers as long as they apply for a job outside of indicated areas with already high numbers of migrants. The partial decentralisation of selection is clearly reflected in the numbers from 2006 and 2012, which show an increase in

42 See R Vineberg, ‘History of Federal-Provincial Relations in Immigration and Integration’ in J Biles et al (eds), Integration and Inclusion of Newcomers and Minorities across Canada (Montreal, McGill– Queen’s University Press, 2011). 43 Recentralisation should serve the aim of avoiding ‘the development of a patchwork approach to the important work of settling new Canadians’. Citizenship and Immigration Canada, ‘Government to Strengthen Responsibility for Integration of Newcomers: “Integration Services are About Nation Building” says Kenney’ (2012) www.cic.gc.ca/english/department/media/releases/2012/2012-04-12.asp. 44 See N Aroney, ‘Australia’ in L Moreno and C Colino (eds), Diversity and Unity in Federal Countries (Montreal, McGill–Queen’s University Press, 2010) 27. 45 In 1998–99, New South Wales was the destination of 41.8% of all newcomers to Australia, while Victoria received 20.5%, Queensland 19.0% and Western Australia 12.7%. The shares of the remaining states and territories were minimal. See Department of Immigration and Citizenship, Settler Arrivals 1998–99 to 2008–09: Australia. States and Territories (Belconnen, Department of Immigration and Citizenship, 2009) 3.

412 Immigration and Migrant Integration employer-sponsored newcomers from 15.230 to 46.554 (37 per cent of all skilled migrants) and state-sponsored newcomers from 8.020 to 22.247 (18 per cent).46 (2) In terms of integration policies, Australian subnational governments have longstanding expertise. It was from this level, more specifically from the 1983 Ethnic Affairs Policy Statement of New South Wales, that multiculturalism emerged and then spread throughout the country.47 Similar to Canada, the integration field has to be subdivided into the early period of settlement services and long-term measures provided to a large extent by the states through their responsibility for general public services. Settlement services, on the other hand, are regulated and funded primarily by the federal Government, while only their delivery is usually left to nongovernmental organisations (NGOs) and, especially in the states with a high concentration of migrants, to state agencies. Federal funding concentrates in particular on programmes for acquiring and improving the English skills of migrants and their offspring. Both the Adult Migrant English Program and the New Arrivals Program for children are based on a decades-long tradition of similar initiatives and have expanded considerably over the years.48 Less smooth are federal-state relations in respect of the recognition of foreign credentials, which is essential to the integration of migrants into the labour market. Several attempts by the federal Government to make the prevailing fragmented assessment procedures, which are dominated by professional bodies at the state level,49 less concentrated on formal qualifications and more coherent have resulted in certain improvements. But the issue remains challenging for intergovernmental relations. Regarding the conferral of citizenship, on the other hand, the federal Government is in no way restricted by the states because naturalisation is encompassed by its exclusive competence under the above-mentioned Section 51(xix) of the Constitution. Recently, the national Government tightened conditions for obtaining Australian nationality, for example by extending the residence criteria from three to four years. Through the introduction of a citizenship test in 2007, which checks candidates’ knowledge about Australia and their English-language skills, the federal Government has reinforced its clear domination in this area.50 13.4 GERMANY

(1) Article 73(3) of the Basic Law makes it unambiguously clear that immigration is an exclusive competence of the federal Government. For decades, the latter pursued, on 46

See L Hawthorne, ‘Australia’ in Joppke and Seidle (n 4) 33. A Whelan, 25 Years of EAPS: Review of EAPS Operation in New South Wales (Sydney, NSW Government, 2009) 14. 48 See Hawthorne, ‘Australia’ (2012) 37ff. 49 See Committee to Advise on Australia’s Immigration Policies, Immigration, a Commitment to Australia: The Report of the Committee to Advise on Australia’ Immigration Policies (Canberra, Australian Government Publishing Service, 1988) 54. 50 See Aroney, ‘Australia’ (2010) 27. 47 See

Germany 413 this constitutional basis, a policy that, in practice, admitted two main groups of people into the country—temporary workers (Gastarbeiter) from Southern Europe and descendants of German settlers in the countries of the communist bloc—but officially recognised neither of these two processes as immigration. Individuals belonging to the latter group, which Article 116(1) of the Basic Law included next to holders of German citizenship in the definition of who is ‘a German’, were not considered as aliens who immigrate but rather as ‘resettlers’.51 Concerning temporary workers, reference was made to the fact that they entered the country in the context of a rotation policy only for a certain period of time so that they were equally denied the label of being a migrant. Thus, Germany could be seen by policy-makers as a nonimmigration country. This long-prevailing self-perception, which ‘articulates not a social or demographic fact but a political-cultural norm’,52 found legal expression in the rudimentary and restrictive 1965 Aliens Law and was only corrected with the reform of this law in 1990.53 During the 25 years in between, the Federal Constitutional Court had countered the non-immigration-country paradigm by granting foreigners, on the basis of the general freedom of action under Article 2(1) of the Basic Law, incremental fundamental rights protection, which gradually approximates their legal status as a function of the length of their stay to that of German citizens.54 In line with the general principle of administrative federalism and the aforementioned Article 73(3), the Länder governments did not act as autonomous actors in framing immigration policy but were confined to executive tasks. (2) The long-standing self-perception of Germany as a non-immigration country also had important repercussions on its approach to the integration of migrants. This official denial of an obvious reality entailed a lack of an explicit and proactive policy on the part of the federal Government and to a lesser extent also Länder governments. As a result, the challenges of integration were in practice left for a long time to the municipalities. Through pragmatic problem-solving in such areas as housing and social counselling, Germany’s local governments acted together with the general social security system, in which migrants were included through employment, as a ‘silent integration machine’.55

51 It has been pointed out that ‘perceiving the Federal Republic as a country of immigration with ultiple national minorities would run contrary to the Basic Law’s conception of a provisional state m which is headed towards the recovery of national unity’. K Hailbronner, ‘Ausländerrecht und Verfassung’ (1983) 36 Neue Juristische Wochenschrift 2105, 2113. 52 R Brubaker, Citizenship and Nationhood in France and Germany (Cambridge, Harvard University Press, 1992) 174. 53 See C Joppke, ‘The Legal-Domestic Sources of Immigrant Rights: The United States, Germany, and the European Union’ (1999) EUI Working Paper SPS 3, 19ff cadmus.eui.eu/bitstream/handle/1814/303/ sps99_3.pdf?sequence=1. 54 49 BVerfGE 168 (Legitimate Expectation Regarding the Renewal of Residence Permits). For a discussion of this judgment’s rationale, see G Schwerdtfeger, Welche rechtlichen Vorkehrungen empfehlen sich, um die Rechtsstellung von Ausländern in der Bundesrepublik Deutschland angemessen zu gestalten? Gutachten A zum 53. Deutschen Juristentag Berlin 1980 (München, Beck, 1980). 55 M Bommes and H Kolb, ‘Germany’ in Joppke and Seidle (n 4) 117.

414 Immigration and Migrant Integration After this traditional distribution of roles was shattered by the new political circumstances after reunification, a series of events gave rise to a new allocation of tasks. While municipalities have remained powerful, the role of the Länder in respect of social and economic integration has certainly been strengthened by these changes. First and foremost, this is due to the extension of their competences in the education field through the reform of German federalism in 2006. This left as concurrent competences only the regulation of admission to universities and of university degrees with the Länder having to deviate from federal legislation (Article 72(3)(6) of the Basic Law).56 As far as the federal Government is concerned, it recently diverted from its course of decades of passivity and began to discover integration policy. At the heart of the new commitment lie the language and orientation courses that were introduced by the federal Immigration Law of 2004. As the Länder proved to be unwilling to contribute to the funding of these courses, their implementation was eventually put entirely under direct federal authority. Against the German tradition of administrative federalism,57 they are therefore managed by the Federal Office for Migration and Refugees and its branch offices throughout Germany. A second federal measure in the integration field is the organisation of public and highly publicised national integration summits. The first such event in 2006 gave rise to the publication of a national integration plan one year later, which involved mutually concerted obligations for each level of government and beyond that for migrant organisations, trade unions and employers’ associations, religious communities, media and important non-profit organisations such as the German Football Association.58 Although the personal involvement of federal government members might be interpreted as a sign of political commitment, the summits and the plan have been criticised for having little more than symbolic importance.59 By contrast, concerning political integration the federal level clearly has the most leverage because Article 73(2) of the Basic Law lists citizenship as one of its exclusive prerogatives. On this constitutional basis and in line with the nonimmigration-country paradigm, the German Government’s position had been opposed for decades to ius soli elements, which were only introduced by the new Citizenship Law of 1999. This piece of legislation also established a right to naturalisation after, as a rule, eight years of residence if the candidate met several additional requirements such as passing a citizenship test. At this point, the Länder come into play because they can apply this federal law with considerable discretion.60 For instance, their respective naturalisation procedure can accelerate or slow down the processing of applications and their assessment of legal criteria such as language skills or loyalty to the Constitution can be decisive. To verify the latter requirement, Baden-Württemberg used, until recently, in-depth interviews instead 56

On the deviation clause, more generally, see ch 12.3. See chs 3.3.3 and 5.3. 58 See M Siemiatycki and T Triadafilopoulos, ‘International Perspectives on Immigrant Service Provision’ (2010) 2 Mowat Centre for Policy Innovation Paper 1, 13ff. 59 See Bommes and Kolb, ‘Germany’ (2012) 129. 60 See I Michalowski, ‘Required to Assimilate? The Context of Citizenship Tests in Five Countries’ (2001) 15 Citizenship Studies 749. 57

Belgium 415 of merely relying on declarations. Altogether, the Länder therefore have substantial leeway to follow a more or less restrictive approach, which is reflected in statistics on naturalisation rates, which are consistently higher in some subnational entities and lower in others such as Baden-Württemberg and Bavaria.61 While the Länder thus play an important role in applying federal citizenship legislation, they have very little leeway regarding the extension of voting rights to migrants. Attempts to make residents with citizenship of certain foreign countries part of the electorate at the local level were undermined by the Federal Constitutional Court in 1989. 13.C Germany 1990: Voting Rights of Foreigners at the Municipal and County Level? In 1989, the Land of Schleswig-Holstein amended the electoral law for municipalities and counties (Kreise) in such a way that the right to vote would be granted in case of five years of legal residence in Germany to citizens of six other countries. The selection of these countries, ie Denmark, Ireland, the Netherlands, Sweden, Norway and Switzerland, was made on the basis of reciprocity. The Federal Constitutional Court held that this violated Article 28(1) of the Basic Law, which stipulates that the people shall be represented in each county and municipality by an elected body.62 According to the Court’s reasoning, the notion of ‘the people’ in this provision includes, just like Article 20(2), which characterises all state authority as derived from ‘the people’, only Germans as defined in Article 116(1) of the Basic Law. Non-citizens were considered to be only part of ‘the population’ (Bevölkerung) but not of ‘the people’ (Volk). The Federal Constitutional Court held that only such an interpretation of the people in Article 28(1) of the Basic Law would guarantee for the various entities at all levels of government a uniform source of democratic legitimacy. As the Treaty of Maastricht later introduced the right to vote in local elections as a corollary of EU citizenship, Article 28(1) was amended. It now states that ‘persons who possess citizenship in any member state of the European Community are also eligible to vote and to be elected in accord with European Community law.’ 13.5 BELGIUM

(1) Although the Belgian state has, since 1970, experienced an unprecedented decentralisation process, the federal Government has so far been reluctant to cede to the regions and communities any authority concerning immigration. Thus, this area is still covered by Article 36 of the Belgian Constitution as a federal residual power. As a consequence, it is up to national legislation and administration to adapt immigration policy to the very different economic realities in the country and the rather unequal distribution of immigrant population, which accounts for a relatively small proportion of the population in Flanders and Wallonia, but for a very large share

61 See H Hagedorn, ‘Föderalismus und die deutsche Staatsangehörigkeit: Die Einbürgerungspolitik der Bundesländer’ in Akgün and Tränhardt, Integrationspolitik (2001). 62 83 BVerfGE 37 (Foreign Voters I).

416 Immigration and Migrant Integration in the Brussels-Capital region. However, this federal immigration policy does not always correspond to subnational demands. In particular, before the current economic crisis, the Flemish region was much more open towards recruiting specifically skilled migrants than the federal Government, which does not pursue a consistent or proactive policy of economic immigration. On the basis of these labour-market considerations and in line with its nation-building efforts, the Flemish region has repeatedly demanded a transfer of powers on the selection and admission of foreigners.63 (2) As Article 128(1) of the Belgian Constitution had only established in very general terms the jurisdiction of the communities over ‘person-related matters’, Article 5(1)(2)(3) of the 1980 Special Law Institutional Reforms specified that the integration of migrants is one of these matters. Without doubt, the most significant subnational initiative targeted at migrants is the 2003 Flemish decree on inburgering (‘citizenisation’)64 which reflects an integration policy with a strong emphasis on the language, culture and values of the receiving society. The integration courses foreseen under this decree follow a two-step approach. First, they offer basic Dutchlanguage instruction, career orientation and social orientation classes about Flemish norms and values and then advanced language and professional training. According to Article 5 of the decree, participation in the courses is mandatory for certain categories of people such as new arrivals, people with refugee status and—quite notably—ministers of recognised religions. But the non-fulfilment of this obligation is sanctioned only by low administrative fines. From a comparative point of view, it is interesting to see the stark contrasts between the decree on inburgering and the integration policies of the federal Government65 and of the other subnational units. While it is obvious that the Flemish policy regarding language training is aimed at countering the prevailing tendency of migrants to prefer learning French over Dutch,66 its integration policy is more demanding in general. The substantial differences between the Flemish and Walloon policy approaches are to some extent reproduced on a smaller scale in the Brussels-Capital region, even though the integration course organised by the Flemish part of the Government is, unlike the one in Flanders, not compulsory. The dominant role that the subnational level plays today in terms of social and economic integration rests not only on the explicit authorisation provided by the above-mentioned Special Law of 1980, but also on other competences with crucial relevance for integration. These are, above all, the responsibility of the c ommunities for almost the entire field of education (Article 127(1)(2) of the Belgian Constitution),

63

See M Martiniello, ‘Belgium’ in Joppke and Seidle (n 4) 67f. uses in English the term ‘citizenisation’, but at the same time underlines that a literal translation of inburgering, which differs from formal naturalisation, is impossible (ibid 71 and 77). 65 See M-C Foblets and Z Yanasmayan, ‘Language and Integration Requirements in Belgium: Discordance between the Flemish Policy of “Inburgering” and the Federal Legislator’s View(s) on the Integration of Newcomers and Migrants’ in R Van Oerset al (eds), A Re-Definition of Belonging? Language and Integration Tests in Europe (Leiden, Martinus Nijhoff Publishers, 2010). 66 See M Martiniello, ‘Belgium’s Immigration Policy’ (2003) 37 International Migration Review 225, 228. 64 Martiniello

Belgium 417 their power over parts of social welfare (Article 5(1)(2)(2) of the 1980 Special Law)67 and the jurisdiction of the regions concerning housing (Article 6(1)(4) of the 1980 Special Law). In all these areas, the subnational entities have implemented comprehensive and rather different policies. The competences regarding education have empowered the Flemish community to actively promote language acquisition by migrant children, whereas the French community has preferred to avoid targeting them with specific measures. As for the position of migrants under social welfare regimes, the communities have considerable autonomy in designing their policies, although the Constitutional Court recently pointed out certain constitutional requirements to be observed by all government levels. The Court ruled that the federal Government had to extend all social assistance measures except guaranteed income for the elderly to long-term residents68 because this category of migrants, and only this one would enjoy a right to equal treatment with Belgian citizens.69 Even though no case has since then concerned social assistance measures of the communities, they are indeed characterised by a clearly more inclusive approach towards permanently settled migrants than newcomers. Interestingly, the 1999 decree introducing Flemish Care Insurance70 requires a certain period of residence in Flanders while social assistance benefits in the other communities refer to residence in all of Belgium. Moreover, the Flemish housing policy impacts migrants since a reform in 2006 made the willingness of all residents of the region to learn Dutch a condition for access to social housing. Whereas the Constitutional Court ruled that this requirement cannot apply to French speakers in municipalities with constitutionally guaranteed language facilities, it upheld the requirement with regard to all other individuals.71 The variations between the subnational units concerning social and economic integration policies would not be very problematic per se if a minimum of intergovernmental coordination was ensured. In practice, however, the subnational entities consider this field their ‘private hunting ground’72 and do not use existing institutional structures for cooperation. The cleavages between them are so great that they are even reluctant to merely exchange best practices and information.73 Efforts by the federal Government to facilitate cooperation, in concrete terms the Commission for Intercultural Dialogue established in 2004 and the Interculturality Sessions launched in 2009, have failed due to the unwillingness of the subnational entities to implement recommendations in the first case and a lack of granting them proper representation in the other.

67

See ch 11.3. is important to note that, regarding long-term residents, the European Court of Human Rights focuses on factual elements rather than a formally recognised status. See Koua Poirrez v France App no 40892/98 (ECtHR, 30 September 2003). 69 Belgian Constitutional Court No 5/2004. 70 See box 11.3. 71 Belgian Constitutional Court No 101/2008. 72 Martiniello, ‘Belgium’ (2012) 71. 73 See M Martiniello et al (eds), Nouvelles migrations et nouveaux migrants en Belgique (Gent, Academia Press, 2010) 311ff. 68 It

418 Immigration and Migrant Integration Whereas the federal level is largely sidelined with regard to the social and economic dimensions of integration, it still controls the area of political integration. Flemish discontent with this current situation has been related not only to the dominance of the national Government as such but also to its liberal approach. This approach is primarily reflected in a 2004 federal law granting migrants with five years of residence the right to vote in local elections and the relaxation of naturalisation criteria in 2000. The latter was part of a reform of the naturalisation regime, which falls, according to Article 74(1) of the Belgian Constitution, within the exclusive responsibility of the national Government. In effect, the amendments of 2000 were the culmination of the ‘integration through citizenship’ strategy that can be traced as far back as to the early 1980s.74 The reform of that year removed the willingness to integrate to be proven by knowledge of one of the three national languages as a requirement. This liberalisation is considered to have provoked as a backlash the introduction of a demanding Flemish integration policy under the above-mentioned 2003 decree on inburgering.75 However, a recent reform of Belgian citizenship legislation in 2012 marked a turning point towards a more restrictive approach. Among other things, it extended the residence requirements for naturalisation upon discretion and entitlement from three to five years and from seven to ten years, respectively. For both of these tracks towards citizenship, the reform reintroduced conditions relating to social and economic integration, which have to be met in terms of language skills, participation in integration courses or employment records. Overall, this reform thus points in a direction that corresponds to the Flemish vision of integration policy.

13.6 COMPARATIVE CONCLUSIONS

(1) It has been demonstrated that both the selection of migrants and their admission to the state territory are usually listed by constitutions as exclusive competences of national governments.76 Among the cases analysed, only Section 95 of the Canadian Constitution Act 1867 establishes concurrent jurisdiction. Regardless of whether the national level then determines immigration policy essentially on its own or is, as in Switzerland, entangled in processes of pre-parliamentary bargaining and direct democracy, subnational entities commonly do not play an important part. The two notable exceptions in terms of policy formulation are Australia and Canada. Starting from different constitutional points of departure, ie an exclusive federal competence in the first case and a concurrent jurisdiction in the latter, subnational entities

74 See M-C Foblets and S Loones, ‘Belgium’ in R Bauböck et al (eds), Acquisition and Loss of ationality: Policies and Trends in 15 European Countries (Amsterdam, Amsterdam University Press, N 2006) Vol 2, 91. 75 See F Delmartino et al, ‘Kingdom of Belgium’ in Moreno and Colino (eds), Diversity (2010) 71. 76 Art 121.1 of the Swiss Constitution, Sec 51.19 of the Australian Constitution, Art 73.3 of the Basic Law and Art 36 of the Belgian Constitution. Only Sec 95 of the Constitution Act 1867 establishes concurrent jurisdiction.

Comparative Conclusions 419 have, in these two countries, taken on a certain role in immigration policies. While in 2011 as many as 55 per cent of all skilled workers admitted to Canada were selected at the provincial level,77 the comparable figure for Australia has never been that high.78 Of course, it has to be taken into account that this difference is in part due to the fact that, under the Australian selection regime, employers can also sponsor potential migrants, which accounted in 2012 for 37 per cent of all skilled newcomers to the country. It is obvious that this decentralisation is in both federations linked to the assumption that subnational entities are in a better position to assess local labour-market needs so that their involvement through a second track of selecting skilled migrants would facilitate their optimal territorial distribution. In both cases, the previously less attractive subnational units have really made use of these new selection powers and have thus produced the envisaged steering effects. Although they still lag behind the most attractive cities and subnational entities, ie Toronto, Ontario, and S ydney and New South Wales, respectively, their relative share of newly arrived skilled migrants has clearly increased in recent years. While labour-market considerations are evidently the only rationale for decentralisation in Australia, this holds true in Canada only for the PNP. By contrast, the bilateral transfer of immigration powers to Quebec since the 1970s, has a second, and at least at the outset, predominant, cultural-linguistic rationale. A similar combination of labour-market considerations and nation-building imperatives has prompted calls by the Flemish Government for competences regarding immigration that have, however, so far been rejected by the national Government. (2) It has been aptly pointed out that ‘immigrants are naturalized as citizens of Belgium, Germany, … but they are socialized as Walloons or Flemish, Bavarians or Hamburgers’.79 This pattern is also reflected in the distribution of competences regarding integration. As far as the social and economic dimension is concerned, subnational entities usually play a crucial role because they have responsibilities by implication from the considerable powers that they typically have concerning the labour market,80 education, housing, social welfare and healthcare. Although the public services usually provided in these areas by them and local governments are intended for the general public, they are also very relevant specifically to the integration of migrants. Therefore, for instance, the increase of subnational jurisdiction over one of these subject matters also entails, as an ancillary effect, an extension of integration competences. The clearly dominant position that the Länder assumed regarding education in the wake of the reform of German federalism in 77 Twenty-five per cent of them were Quebec-selected skilled workers and 30% provincial nominees, see Citizenship and Immigration Canada, ‘Annual Report’ (2012) 10. 78 See Hawthorne (n 46) 32ff. 79 D Thränhardt, ‘Immigration and Integration in European Federal Countries: A Comparative Evaluation’ in D Thränhardt (ed), Immigration and Federalism in Europe: Federal, State and Local Regulatory Competences in Austria, Belgium, Germany, Italy, Russia, Spain and Switzerland (Osnabrück, Institut für Migrationsforschung und Interkulturelle Studien, 2013) 7. 80 But see box 13.B above.

420 Immigration and Migrant Integration 2006 is a case in point. Overall, these implied powers have enabled subnational entities to design and implement autonomous policies according to their own and often different preferences, which is particularly evident in the case of Belgium. There, the more demanding Flemish approach towards integration is clearly recognisable in its education, housing and social welfare policies. Moreover, in contrast to other subnational units, which typically have to rely on implied powers, the Belgian communities are also endowed with explicit jurisdiction over the integration of migrants (Article 128(1) of the Belgian Constitution in conjunction with Article 5(1)(2)(3) of the 1980 Special Law on Institutional Reforms). Beyond the authority to provide general services that are not available exclusively but also for migrants, these subnational implied powers commonly also include the provision of specific measures intended only for migrants.81 Key among the latter are language and orientation courses that have, since the 2000s, become a common trend, especially in European countries. As a general rule, national governments have significant bearing on these—in some cases mandatory—courses because they are, like in Switzerland, in charge of regulating their content and funding or, like in Germany, additionally in charge of implementation. A noteworthy counterexample is Belgium, where the language and orientation courses in the Flemish region are a subnational initiative in the specific context of a practically absent federal integration policy. Another example of measures targeted specifically at migrants are the settlement services in Canada and Australia. These deal likewise with language acquisition and social orientation but go well beyond these issues. Whereas in A ustralia there is a clear and uniform division between the federal legislation on these settlement measures and their implementation by the states together with NGOs, the Canadian situation is much more complex. Even after the recent unilateral termination of the far-reaching settlement agreements with Manitoba and British Columbia, there are still several different models in place with the various provinces. Exceptional among them is, as usual, Quebec, which according to the 1991 Canada–Quebec Accord not only provides these services but also regulates them. In respect of political integration, it is usually taken for granted that one of its main instruments, namely naturalisation, is a domain of the federal level. Generally speaking, most constitutions indeed define citizenship as an exclusively national responsibility in terms of both legislative and executive power, with the latter being delegated to local branches of the national administration. Therefore, the exceptionally powerful position of the cantons of Switzerland in this area remains certainly unequalled. To be sure, the concept of multilevel citizenship enshrined in Article 37(1) of the Swiss Constitution and substantiated in ordinary legislation actually makes the municipalities the most decisive level concerning the naturalisation process. But the cantons still have considerable leeway to establish their own additional criteria, like a requirement of residence for a certain period not just in Switzerland but in the canton. Some room for choosing at the subnational level a more or less restrictive approach also exists in Germany. Even if the Länder, unlike the Swiss cantons, do 81 A distinction can therefore be made between general/indirect and specific/direct integration measures. See JP Lynch and RJ Simon, Immigration the World Over: Statutes, Policies and Practices (Lanham, Rowman and Littlefield, 2003) 32.

Comparative Conclusions 421 not have the competence to impose their own naturalisation requirements, they have substantial discretion in applying federal citizenship legislation both concerning procedures and the interpretation of criteria. By contrast, in Belgium the subnational entities do not have—much to the disappointment of the Flemish Government—a significant formal role regarding naturalisation. To what extent the recent turn of the federal citizenship legislation in 2012 towards a more restrictive approach was in practice a response to corresponding Flemish preferences, is difficult to assess. While naturalisation should not be equated with political integration, it is certainly a crucial instrument to promote it, because many rights are still reserved for citizens despite an undeniable increase of rights based on residence. A case in point is, in this regard, the stance of the German Federal Constitutional Court, which frustrated Länder efforts to include certain categories of migrants in the electorate of municipal and county elections.82 In some countries, the non-EU citizens actually do have the right to vote in local elections, albeit upon the initiative of different levels of government. Whereas in Belgium this right to vote in local elections goes back to a federal law of 2004 and was explicitly opposed by the Flemish Government, the Swiss cantons can take this decision autonomously. Thus, the predominantly French-speaking cantons have proved much more open towards extending voting rights for migrants concerning local and sometimes even cantonal elections. Patterns are likewise divergent from country to country in the area of multicultural policies, which have recently been despised by a number of political leaders in Europe but reinforced in Canada and Australia. Whereas in Canada the federal Government was the pioneer of multiculturalism, such policies have emerged in Australia to a greater degree as a result of subnational initiatives, with the 1983 Ethnic Affairs Policy Statement of New South Wales paving the way. Regardless of whether the social and economic dimension or the political dimension of integration is concerned, decentralisation has taken place in the five countries analysed either on a voluntary or involuntary basis. The latter was the case in Switzerland and Germany, where the subnational entities—but even more the municipalities—were forced for a long time to bear the main burden of integration, as the respective national governments, due to the non-immigration-country paradigm, remained reluctant to develop an explicit policy. In other cases, subnational entities end up with responsibility for integration not because of the inactivity of the national Government but, quite to the contrary, because the latter actively devolves this responsibility. A prime example of such a move on the part of a national Government in order to strengthen itself ‘by off-loading the less gratifying functions and externalizing difficult tasks’83 is the devolution of the settlement services to the provinces in the context of Canada’s financial crisis in the mid-1990s. Nonetheless, subnational entities seem to be, in most cases, the driving force behind decentralisation of integration and to acquire competences voluntarily. This is particularly true for regions that engage in nation-building and in this context push most vigorously

82

See box 13.C above. Keating, ‘Europeanism and Regionalism’ in B Jones and M Keating (eds), The European Union and the Regions (Oxford, Oxford University Press, 1995) 10. 83 M

422 Immigration and Migrant Integration for jurisdiction over integration. Efforts to make newcomers integrate in the subnational context, such as the Flemish decree on inburgering of 2003 and Quebec’s monopolised settlement services, are obviously in competition with the respective national integration projects. As migrants are to be won over as linguistic, cultural and, in the end, political allies,84 integration policies in these subnational entities tend to be more proactive and demanding. This difference and other factors give rise to considerable asymmetries in terms of both subnational integration powers and policies. For instance, the latter vary dramatically in Belgium, where the Flemish and Walloon approaches to integration could not be much further apart. This asymmetry in relation to policies stems from symmetrical decentralisation of powers just because the two subnational entities make use of their equal competences to a different extent and in a different manner. Although in Canada Quebec’s pioneering role in acquiring far-reaching responsibility for integration triggered a catching-up process in the remaining provinces, the latter clearly lag behind, even more so after the recentralisation efforts of 2012. In both states, asymmetry is regarded as having contributed to a certain opaqueness of integration policies85 because intergovernmental cooperation, practically absent in Belgium and largely bilateral in Canada, has not even ensured a minimum of coherence. To find this balance remains a challenge in these two and other countries.

84 See E Hepburn, ‘“Citizens of the Region”: Party Conceptions of Regional Citizenship and Immigrant Integration’ (2011) 50 European Journal of Political Research 504, 521. 85 See Martiniello (n 63) 75f; Banting (n 4) 103ff.

14 External Relations

S

Internal diversity, external unity.1

INCE A NINETEENTH-CENTURY Swiss politician characterised, with this concise formula, the pre-eminence of national governments in the international arena, the nature of both states and this arena have changed fundamentally. Although states continue to take on the leading role in an international setting that some observers even regard as having neo-Westphalian tendencies,2 they are now embedded in a complex and fragmented universe of diverse actors, including international organisations, non-governmental organisations (NGOs), multinational companies, etc. Moreover, the states themselves, as still primary actors, are no longer black boxes that demonstrate external unity, as subnational units have become increasingly involved in international affairs. This phenomenon, usually termed ‘paradiplomacy’3 or ‘constituent diplomacy’,4 is most often attributed to the growing interdependence that has followed from globalisation and, above all in Europe, from supranational integration.5 It is obvious that, in the context of ‘intermestic affairs’,6 in which the boundaries between international and domestic affairs are increasingly blurred, nearly all domestic powers of subnational units can only be exercised in a meaningful way if they are granted at least some authority regarding the international dimensions of these competences. The nature of external relations thus differs from the primarily domestic powers explored in the preceding chapters, such as social welfare and healthcare or environmental protection, because external relations are very much about procedures to effectively use subnational competences in a more and more interdependent international context. Without such an external projection of internal powers, subnational entities risk a creeping erosion of their autonomy.7 1 Alfred Escher as quoted in D Thürer, ‘Federalism and Foreign Relations’ in R Blindenbacher and A Koller (eds), Federalism in a Changing World: Learning from Each Other (Montreal, McGill–Queen’s University Press, 2003) 56. 2 See, eg, A Cassese, ‘Gathering Up the Main Threads’ in A Cassese (ed), Realizing Utopia: The Future of International Law (Oxford, Oxford University Press, 2012) 646. 3 ID Duchacek, ‘Perforated Sovereignties: Towards a Typology of New Actors in International Relations’ in H Michelmann and P Soldatos (eds), Federalism and International Relations: The Role of Subnational Units (Oxford, Clarendon Press, 1990). 4 J Kincaid, ‘Foreign Relations of Subnational Units’ in Blindenbacher and Koller (eds), Federalism (2003). 5 See ID Duchacek et al, Perforated Sovereignties and International Relations: Trans-Sovereign Contacts of Subnational Governments (Westport, Greenwood Press, 1988). 6 The term was coined by B Manning, ‘The Congress, the Executive and Intermestic Affairs: Three Proposals’ (1977) 55 Foreign Affairs 306. 7 See F Palermo, Die Außenbeziehungen der italienischen Regionen in rechtsvergleichender Sicht (Frankfurt am Main, Peter Lang, 1999) 9f.

424 External Relations From a systematic point of view, the increasing external relations of subnational entities with partners as different as foreign countries, their constituent units, international or supranational organisations, etc include many activities. These also structure this section’s comparative analysis of five countries. After an introduction to the distribution of external powers in each case (sub-section (1)), it will go on to the first and fundamental subnational activity abroad, the continuous maintenance of external relations, (sub-section (2)) through a wide range of measures such as running representation offices or undertaking foreign visits. A third important international activity concerns the role of regions in c reating international and supranational law (sub-section (3)), including the exercise of their own treaty-making power, participation in negotiating treaties concluded by their national government and in decision-making within a supranational organisation like the European Union. Even though there has been the option since the Maastricht reforms for Member States to be represented in the Council of the EU by a member of a subnational government (Article 16(2) of the Treaty on the European Union (TEU)),8 the latter have remained in large part marginalised. The dominance of national governments is obvious in decision-making regarding both primary EU law (through treaty revisions) and secondary EU law (through the Council of the EU). As the competences of the European Union overlap considerably with typical subnational powers, but national governments negotiate in the Council, there is a significant risk of the latter striking bargains that negatively affect competences of the subnational entities.9 In other instances, however, Member States have countered a perceived creeping shift of powers to the EU by emphasising the importance of the internal distribution of powers.10 Cases in point are the Maastricht judgment by Germany’s Federal Constitutional Court11 and the same country’s successful challenge of certain EU directives.12 Besides the creation of non-domestic law, a fourth dimension of external relations is the role of subnational entities concerning the implementation of international and supranational law (sub-section (4)). Regarding the special case of EU law, it is important to note that the Member States may not justify a lack of its implementation with internal constitutional obstacles like exclusive jurisdiction of subnational entities over the subject matter in question. Similar to the implementation of international law,13 the state alone, perceived as a unitary subject, is responsible for fulfilling this obligation.14 Generally, subnational entities are obliged not to apply

8

See ch 1.2.1. Of course, this risk is to some extent reduced by the often careful limitations of EU powers regarding some of these typical prerogatives of subnational entities. Cases in point are ‘Education, Vocational Training, Youth and Sport’ (Arts 165–66 of the Treaty on the Functioning of the European Union (TFEU)), Culture (Art 167 TFEU), Public Health (Art 168 TFEU) and Consumer Protection (Art 169 TFEU). 10 See S Weatherill, ‘The Challenge of the Regional Dimension in the European Union’ in S Weatherill and U Bernitz (eds), The Role of Regions and Sub-National Actors in Europe (Oxford, Hart Publishing, 2005) 6ff. 11 See box 4.B. 12 Case C-376/98 Germany v Parliament [2000] ECR I-8419. 13 Art 27 of the 1969 Vienna Convention on the Law of Treaties. 14 Case C-103/01 Commission v Germany [2003] ECR I-5369. 9

United States 425 any domestic law that is incompatible with applicable EU law.15 Another important point is that the standing of subnational units to challenge EU law is not better than that of a private individual and that, conversely, a challenged subnational law will be defended by the national government (Article 263 of the Treaty on the Functioning of the European Union (TFEU)). All this considered, it seems evident that the rules about the creation, implementation, application and judicial review of EU law have a strong direct impact on the subnational level of government. In this light, the EU’s alleged ‘regional blindness’, ie its indifference towards internal constitutional arrangements, appears to indeed be a rather ‘deceptive notion’.16 Yet, national constitutions do matter and make a difference for relations with the EU, albeit to a lesser extent than regarding classical external relations. After all, it is their task to provide the domestic legal framework, within limits imposed by non-domestic law, for subnational entities to operate on the external plane. The following sections will demonstrate that federal systems do this quite differently. 14.1 UNITED STATES

(1) Although the US Constitution does not contain a single subject matter called ‘external relations’ or anything similar, it establishes, without ambiguity, federal prevalence in this realm by explicitly centralising foreign economic policy and the power to enter into treaties. Article I, section 8, clause 3 of the Constitution says that Congress ‘shall regulate commerce with foreign nations’. Article I, section 10, clauses 1 and 3 then add that ‘[n]o state shall enter into any treaty, alliance, or confederation’ and that ‘[n]o state shall, without the Consent of Congress … enter into any agreement or compact with another state’. While the subnational governments are thus denied treaty- making power, Article II, section 2, clause 2 explicitly grants the ius contrahendi to the US President, acting with the ‘advice and consent’ of two-thirds of the Senate. To a certain extent, this constitutional entrenchment of federal dominance was a backlash to the mostly separate and sometimes divergent external relations under the Articles of Confederation. The perceived necessity to make the newly established federal state speak abroad with one voice already prompted James Madison to argue the case for centralisation17 and is clearly reflected in the distribution of powers. (2) Beyond these constitutional restrictions pertaining to foreign commerce and international treaties, the states nevertheless continued international activities in the nineteenth century, as evidenced, for example, by their close cooperation with several European countries over the important issue of attracting immigrants.18 This activism then began to slowly fade away in the wake of the Civil War and 15

Case 103/88 Fratelli Costanzo [1989] ECR 1839. Weatherill, ‘The Challenge’ (2005) 30. 17 The Federalist no 42. 18 See J Kincaid, ‘The International Competence of US States and their Local Governments’ (1999) 9 Regional and Federal Studies 111, 116f. 16

426 External Relations was only rediscovered in the 1960s.19 In this period, some states started to make regular visits to Europe and to station representatives there, in part to benefit from the continent’s burgeoning economic integration. After a steady increase of international activity especially since the 1980s, state bills and resolutions addressing external relations,20 visits of state delegations abroad and the establishment of trade offices has today become common practice. Though trade offices of states exist in a number of countries, they are often opened and closed depending on the individual international commitment of a governor and rarely provided with a comprehensive strategic programme.21 Thematically, the bulk of the states’ international actions indeed revolve around the promotion of their economic development, above all around attracting tourists and foreign direct investment, as well as export promotion. In this context, the practice of states to grant foreign investors fiscal privileges and thereby outbidding each other with ever more generous incentives22 has entailed a controversial public debate but was reinforced by a recent judgment of the Supreme Court in DaimlerChrysler Corp v Cuno.23 Economic motivation is also elementary to cross-border cooperation with Canada and Mexico, not least as a result of common membership in the North American Free Trade Association (NAFTA). However, the forums of transfrontier cooperation, such as the Border Governors’ Conference or the Council of Great Lakes Governors, also discuss environmental protection, energy management, infrastructure and security issues. The legal formalisation of this cooperation is often effected by agreements with congressional consent, which states are allowed to conclude by the above-mentioned Article I, section 10, clause 3 of the US Constitution. Whereas cross-border cooperation usually happens to be uncontroversial, state attempts to move from this limited geographical space onto the stage of world politics may provoke conflict with the federal Government as illustrated by the opposing stances of some states on the US policy in Iraq and concerning the Kyoto Protocol. For instance, California’s policy of reducing emissions of greenhouse gases involved significant international action, whose tangible outcomes are a bilateral environmental accord with the British Government and the Western Climate Initiative, including not only other US states but also Canadian provinces.24 A particularly vivid example of divergent positions and the precedence of the federal Government in such a case is the Supreme Court decision in Crosby v National Foreign Trade Council.25 In this judgment, the Court invalidated a Massachusetts law of 1996 that had restricted access to state procurement contracts for companies doing business in Myanmar because a federal law had already imposed economic sanctions and thus pre-empted 19

See J Kincaid, ‘The American Governors in International Affairs’ (1984) 14 Publius 95. See TJ Conlan et al, ‘Taking on the World: The International Activities of American State L egislatures’ (2004) 34 Publius 183. 21 EH Fry, ‘The United States of America’ in H Michelmann (ed), Foreign Relations in Federal Countries (Montreal, McGill–Queen’s University Press, 2009) 316. 22 See EH Fry, The Expanding Role of State and Local Governments in US Foreign Affairs (New York, Council on Foreign Relations Press, 1998) 117. 23 In DaimlerChrysler Corp v Cuno 547 US 332 (2006), the Supreme Court ruled that Ohio taxpayers were not entitled to challenge the granting of a USD 300 million tax break by its state government to Daimler Chrysler as an incentive for building a new assembly plant. 24 See ch 12.1. 25 Crosby v National Foreign Trade Council 530 US 363 (2000). 20

United States 427 legislative action by the states as a result of the Supremacy Clause (Article VI, clause 2 of the US Constitution). (3) According to the Supremacy Clause, the ‘supreme law of the land’ includes not only the US Constitution and federal laws, but also ‘all treaties made, or which shall be made, under the authority of the United States’. The fact that international treaties therefore take precedence over conflicting state laws, of course, enormously increases the importance of who has the power to conclude them. As outlined above, the constitutional distribution of power makes the ius contrahendi an exclusive prerogative of the federal Government. This evidently entails, as an immediate consequence, a substantial limitation of the states’ legislative scope.26 If, for example, their legal measures to protect the environment or public health collide with World Trade Organisation or NAFTA obligations, there is no doubt that the latter will prevail. In Medellín v Texas,27 the Supreme Court clarified that an international treaty may only bind and constrain states, if it is self-executing or the Congress has enacted statutes implementing the treaty. Beyond that, also the participation of states in shaping the content of treaties beforehand is only minimal. Even though some mechanisms such as the Intergovernmental Policy Advisory Committee exist, the states are only consulted sporadically and their views not effectively considered.28 The influence of a state depends, rather than on these institutions, on the power and commitment of its senators to represent its interests in bargaining processes with the US President and within Congress.29 (4) When it comes to the implementation of international treaties, the central question is in the US case whether this competence follows the distribution of domestic powers or whether it is included in the treaty-making power (of the national Government). It is obvious that in the latter case, the legislative scope of the states would be, beyond the above-mentioned effect of the Supremacy Clause, even further reduced. In the absence of any clear indications in the text of the US Constitution, it was up to the Supreme Court to decide this question. 14.A United States 1920: Congressional Power to Implement Treaties The US Congress had enacted laws concerning the hunting of the migratory waterfowl on the grounds that the natural movement of this bird across boundaries would require uniform regulation. Several states had then successfully challenged these federal laws because the issue at stake did not fall within one of the enumerated federal subject matters and

26 See PJ Spiro, ‘Globalization and the (Foreign Affairs) Constitution’ (2002) 63 Ohio State Law Journal 649. 27 Medellín v Texas 552 US 491 (2008). 28 See Fry, ‘The United States’ (2009) 305. 29 See GA Tarr, ‘United States of America’ in J Kincaid and GA Tarr (eds), Constitutional Origins, Structure, and Change in Federal Countries (Montreal, McGill–Queen’s University Press, 2005) 398f; Kincaid, ‘The International’ (1999) 114.

428 External Relations

therefore belonged to the residuary power of the states according to the Tenth Amendment. Unhappy with this limitation, the national Government decided to circumvent it by concluding with the United Kingdom, at that time still in charge of Canada’s external relations, an international treaty concerning the issue. In order to implement this treaty, it enacted, two years later, the Migratory Bird Treaty Act. The state of Missouri resolved to file a lawsuit in order to prevent the US game warden, Ray Holland, from enforcing this piece of legislation. The Supreme Court ruled that the Migration Bird Treaty Act was consistent with the Constitution.30 Congress has the power to give effect to an international treaty through federal legislation even if that statute standing alone, ie, without its function of implementing a treaty, would violate states’ rights under the Tenth Amendment. The residual power of the states is irrelevant because the power to make treaties is expressly enumerated as an exclusive federal competence (Article II, section 2, clause 2) and also comprises legislation ‘necessary and proper’ to exercise to this power (Article I, section 8, clause 18).31 A federal law with the same content can therefore be consistent or inconsistent with the Constitution depending on whether its function is to implement a treaty or not. 14.2 SWITZERLAND

(1) When Switzerland was established as a federal state in 1848, its cantons were in general accorded far-reaching powers. Quite contrary to this pattern, external relations have been, from the beginning, a domain of the federal Government. In order to preserve the independence of this small country, the framers of the Constitution deemed a uniform policy in the international arena indispensable. While the preservation of independence has indeed been the main objective of external relations, the revision of the Constitution in 1999 explicitly broadened their scope to also include such aims as the alleviation of poverty or the promotion of human rights and democracy (Article 54(2) of the Swiss Constitution). Anyway, it is still widely accepted that the pursuit of these new aims on the global stage shall be reserved to the federal Government. The latter’s prerogative regarding the conduct of foreign relations was since 1848 regarded as inherent to the Constitution32 and only in 1999 expressly confirmed by the new Article 54(1). (2) While external relations on the part of the cantons are not ruled out, they are clearly characterised by a much more limited geographical scope. They are focused above all on cross-border cooperation, with the areas around internationally shared lakes and rivers such as Lake Geneva and the Upper Rhine, as well as the cantons of

30

Missouri v Holland 252 US 416 (1920). On the Necessary and Proper Clause, see ch 5.2.3. 32 See L Wildhaber, ‘Aussenpolitische Kompetenzordnung im schweizerischen Bundesstaat’ in A Riklin et al (eds), Neues Handbuch der Schweizerischen Aussenpolitik (Bern, Haupt, 1992). 31

Switzerland 429 Basel-City and Zurich, being particularly active. The topics covered include typical concerns of cross-border cooperation like urban and rural development, natural resources, infrastructure and—especially between Zurich and the economically equally strong German Land of Baden-Württemberg—business cooperation.33 In a usually cooperative manner, the federal Government sometimes plays the role of a facilitator by concluding international treaties to create a framework for transfrontier activities. A case in point is the Karlsruhe Agreement between Switzerland, Germany, France and Luxembourg,34 which allows all subnational entities involved to enter into accords without approval. The fact that the cantons focus regarding their contacts on other subnational entities, arguably corresponds to their preferences, but in part also follows from a constitutional obligation. Article 56(3) of the Swiss Constitution only authorises them to ‘deal directly with lower ranking foreign authorities’, whereas the federal Government ‘shall conduct relations with foreign states on behalf of a canton’. Apart from relations with states, this restriction also applies to international organisations. It is for this reason that the cantons generally prefer EU membership to the current approach of concluding agreements on specific topics of mutual interest.35 In their view, membership would offer them more participation than the prevailing bilateral Swiss-EU relations, which are clearly dominated by the federal Government.36 However, apart from European policies, the cantons seem to feel comfortable with federal dominance in external relations. This is illustrated, for instance, by their reluctance to set up an adequate institutional structure. Only a few cantons established specialised departments at home and none of them permanent representative offices abroad. Self-restraint also characterises the priorities which external relations of cantons focus on. Apart from classical issues of cross-border cooperation, they largely concentrate on economic cooperation and do not have a broader political agenda. But even this area, which is obviously of vital importance for a small country like Switzerland, is dominated by the federal Government. Unlike other countries with significant linguistic diversity, the cantons are commonly not very active abroad concerning cultural affairs. For instance, the cantons with a French-speaking majority do not push strongly for recognition as members of the Organisation international de la Francophonie next to Switzerland as a whole. Overall, they show little interest in pursuing an autonomous agenda in the international arena. Instead, the ‘primary aim of the cantons as regards foreign policy seems to be to maintain their influence in domestic policy’.37

33 See D Thürer and M MacLaren, ‘Swiss Confederation’ in Michelmann (ed), Foreign Relations (2009) 284ff. 34 Agreement between Germany, France, Luxembourg, and Switzerland on Cross-Border Co-operation between Territorial Bodies and Local Public Authorities of 23 January 1996. 35 The Bilateral Agreements I of 1999 dealt with questions of market access. The Bilateral Agreements II of 2004 concern, above all, security and asylum policy, environment and cultural affairs. 36 See Thürer and MacLaren, ‘Swiss Confederation’ (2009) 279. 37 Thürer and MacLaren (n 33) 282.

430 External Relations (3) This is ensured above all through participation in the treaty-making of the federal Government as well as through their own treaties. In the 1990s at the latest, it became obvious that treaties concluded by the Swiss Government concerning areas of subnational responsibility presented an ever more serious threat to the legislative competences of the cantons as a result of increasing internationalisation and the Swiss way of European integration.38 This gave rise to demands for a comprehensive reform of treaty-making power. As a consequence, the revised Constitution of 1999 both guaranteed a cantonal right to enter into treaties and, even more importantly, created mechanisms of participation regarding treaties entered into by the federal Government. Article 56 of the Swiss Constitution explicitly states that a canton has the power to ‘conclude treaties with foreign states on matters that lie within the scope of its powers’ provided that they do not run contrary to the law or interests of the federal Government or the law of other cantons. Moreover, the canton must inform the federal Government before concluding such a treaty. The fact that the latter can raise objections not only in case of a conflict with federal law but also with its interests requires cantonal treaties to be appropriate in the context of general Swiss foreign policy. If a cantonal treaty is vetoed, either by the federal Government or by another canton, it is up to the Federal Assembly to approve it or not (Article 172(3)). Furthermore, since such a treaty has the status of cantonal law, it can also be abrogated, as a result of the supremacy clause in Article 49(1),39 by a treaty of the federal Government or by a federal law. In legal terms, these mechanisms restrict the ius contrahendi of the cantons quite considerably. In practice, however, they are of limited relevance because cantonal treaty-making power is also rather rarely used. If it is used, then it typically concerns merely a limited range of issues and partners from neighbouring jurisdictions. Even more important is therefore the enhanced participation of the cantons concerning the much more numerous treaties concluded by the federal Government. With regard to all foreign policy decisions, the federal Government is constitutionally obliged to respect the powers and to protect the interests of the cantons (Article 54(3)). Even more important are the procedural duties to consult them if their powers or essential interests are affected (Article 55(1)), to inform them fully and in due time (Article 55(2)) and to ensure their participation in international negotiations with an impact on their areas of competence (Article 55(3)). These provisions of Article 55 are further elaborated by ordinary legislation adopted back in 1999.40 According to the revised Constitution, treaty-making is, in certain cases, not even restricted to the interplay between the national and subnational governments but is a process that involves other actors too. Along with the cantons, political parties and interested groups shall be consulted at least with regard to significant treaties (Article 147). 38 See B Habegger, ‘Participation of Subnational Units in the Foreign Policy of the Federation’ in Blindenbacher and Koller (n 1) 166. 39 See ch 5.1.2. 40 See T Pfisterer, ‘Auslandbeziehungen der Kantone’ in D Thürer et al (eds), Verfassungsrecht der Schweiz (Zürich, Schulthess, 2001) 535ff.

Canada 431 Another Swiss peculiarity is, of course, the practice of referendums concerning certain international treaties. These are mandatory for treaties providing membership in collective security organisations or supranational communities (Article 140(1)(b)) and can be requested by 50,000 voters or eight cantons regarding other categories of treaties (Article 141(1)(d)). The latter referendums only require a majority of all voters, while the aforementioned compulsory referendums also need for approval a majority of voters in a majority of the cantons (Article 142(1–2)). Recent positive votes concerning the European Union and the United Nations41 indicated a decline of traditional Swiss scepticism towards too much integration in international organisations. In particular, the Bilateral Agreement II with the EU of 2004 demonstrated that the cantons have clout in treaty negotiations if they speak with one voice through the Conference of Cantonal Governments.42 This conference, established in 1993 and characterised by a permanent institutional structure, has developed into the most significant instrument of cantonal participation, which, in several instances, has succeeded in preventing the much feared large-scale erosion of cantonal competences through international treaties.43 (4) Apart from this feared erosion of powers, the constitutional emphasis on the participation of the cantons in federal treaty-making also follows from the wish to facilitate the process of treaty implementation in which they play a pivotal role. Treaties must be implemented in any case by the cantons if they concern an area of cantonal competence. Beyond that, the national Government can make them responsible for giving effect to a treaty if it pertains to a federal competence, which it actually does quite often. The cantons are then obliged to undertake all necessary measures in due time and in a cooperative manner following constitutional principles of intergovernmental collaboration (Article 44 of the Swiss Constitution). At the same time, the cantons have a certain scope of discretion during this implementation process. This is considered to be implied by the wording of Article 46 of the Constitution, which uses Umsetzung (implementation) instead of Vollzug (execution).44 14.3 CANADA

(1) When it comes to external relations, the Canadian Constitution is quite peculiar, for its distribution of powers does not assign competences in this area either to the federal Government or to the provinces. The reason for this is quite simple. Though the original Constitution of 1867 established Canada as a self-governing

41 The Swiss people gave their consent to several treaties deepening bilateral relations with the EU. In 2002, UN accession was approved by narrow margin after a clear rejection in 1986. 42 See B Münger, ‘Die Konferenz der Kantonsregierungen’ in D Freiburghaus (ed), Die Kantone und Europa (Bern, Haupt, 1994). 43 See N Schmitt, ‘Swiss Confederation’ in Kincaid and Tarr (eds), Constitutional Origins (2005). 44 See Thürer and MacLaren (n 33) 277.

432 External Relations entity in respect of domestic affairs, the conduct of its foreign affairs remained the responsibility of the Crown as a result of the country’s status as a dominion within the British Empire. Only in 1926 did the Balfour Declaration acknowledge that Great Britain and the dominions were ‘autonomous communities within the British Empire, equal in status, in no way subordinate one to another in any respect of their domestic or external affairs’.45 In a royal Letters Patent of 1947, the Governor General, thus effectively the federal Government, was formally authorised ‘to exercise all powers and authorities lawfully belonging to us [the King] in respect of Canada’,46 which also included foreign affairs. At the same time, it has always been beyond dispute that the provinces can also act in the international arena, albeit without h aving international legal personality. (2) Substantial external relations are nonetheless rather recent for most p rovinces, as their international activism only started to develop in the late 1970s. Only Quebec had already sent representatives to France, the United Kingdom and the United States soon after becoming part of Canada and expanded its international activities in the wake of the 1960s Quiet Revolution when it established further offices abroad and a government department in charge of external relations, the Ministère des relations internationals.47 Since then, Quebec has set up representative offices in 25 countries, which are, in contrast to the other provinces, usually separate structures and not located in Canadian embassies. Furthermore, it has signed several hundred agreements with states and subnational governments from all continents. As long as agreements are not binding under international law, there are without doubt no constitutional obstacles for the provinces to conclude such accords.48 Nevertheless, in the absence of international legal personality, both these agreements and representative offices are subject to framework arrangements concluded by the federal Government with the respective partner. The fact that Quebec’s many agreements do not only cover cultural affairs but a broad variety of areas in which it has domestic competences, clearly illustrates the ambition of the province to be perceived abroad as a nation.49 In spite of the broad agenda, it is obvious that for bilateral and multilateral relations the most important partners are still France and the Organisation international de la Francophonie (OIF). While these close links had in the past created considerable tensions,50 the federal Government has meanwhile recognised Quebec’s special direct relationship with France. Likewise, a compromise was found regarding representation in

45 Quoted in M Ollivier, The Colonial and Imperial Conferences from 1887 to 1937 (Ottawa, Queen’s Printer, 1954) Vol 3, 146. 46 Clause 2 of the ‘Letters Patent constituting the office of the Governor General of Canada’ of 1947. 47 See L Balthazar, ‘The Quebec Experience: Success or Failure?’ (1999) 9 Regional and Federal Studies 153. 48 See PW Hogg, Constitutional Law of Canada, 5th edn (Scarborough, Carswell, 2000) 11-18. 49 See A Lecours, ‘Canada’ in Michelmann (n 21) 128. 50 Examples of past controversies are Charles De Gaulle’s famous ‘Vive le Quebec libre!’ speech of 1967 in Montreal, as well as, more generally, the French policy of ‘non-interference, non-indifference’.

Canada 433 the OIF, of which both Quebec and New Brunswick, with its substantial Frenchspeaking minority, are ‘participating governments’ as opposed to Canada which is a ‘member’.51 Whereas the international activities of these two provinces also have a cultural dimension, their predominantly Anglophone counterparts largely focus on establishing trade links and attracting foreign investment. Most active among them are certainly Alberta, which has built economic ties in particular with the Asia-Pacific area, and to a lesser extent Ontario,52 which has established a special partnership, for instance, with the regions constituting the ‘Four Motors of Europe’.53 Apart from these particular partnerships, the essentially economic external relations of both these provinces and of the remaining ones are strongly focused on the United States. This essentially follows from proximity and common membership in the NAFTA, which has produced an unprecedented economic entanglement. In fact, NAFTA has significantly increased Canadian trade with the United States and correspondingly decreased interprovincial trade. In 2011, for instance, 73.7 per cent of all Canadian exports were destined for the United States, from where Canada also received 49.5 per cent of all imports and 47.2 per cent of its foreign direct investments.54 (3) In contrast to the agreements described above, the authority to enter into treaties that are binding under international law is reserved to the federal Government. This is derived from the aforementioned Clause 2 of the Letters Patent of 1947, which formally declared foreign affairs in its entirety, ie including the ius contrahendi as its traditional core competence, to be a prerogative of the Governor General.55 To some extent, the treaty-making power of each dominion government had already been recognised at the imperial conference of 1926. Between that time and 1947, only treaties in head-of-state form still required some formal steps in London. Less formal treaties, by contrast, had already been concluded in this period by the Canadian Government without any British interference.56 The fact that the power to enter into international treaties is granted exclusively to the national Government has drawn criticism from some provinces, above all Quebec. The argument advanced by Quebec is based on the undisputed provincial responsibility to implement international treaties concluded by the federal Government insofar as they affect provincial competences.57 The so-called Gérin-Lajoie 51 See A Lecours, ‘Paradiplomacy: Reflections on the Foreign Policy and International Relations of Regions’ (2002) 7 International Negotiation 91. 52 See JP Groen, Intergovernmental Relations and the International Activities of Ontario and Alberta (Kingston, Queen’s University Press, 1995). 53 Composed of Rhône-Alpes in France, Baden-Württemberg in Germany, Lombardy in Italy and Catalonia in Spain. 54 Department of Foreign Affairs, Trade and Development, ‘Canada’s State of Trade: Trade and Investment Update 2012’ (2012) 67ff www.international.gc.ca/economist-economiste/performance/state-point/ state_2012_point/2012_5.aspx?lang=eng. 55 See GL Morris, ‘The Treaty-Making Power’ (1967) 45 Canadian Bar Review 478, 482ff. 56 See Hogg, Constitutional Law (2000) 11-11. 57 See box 14.B below.

434 External Relations doctrine contends that there must be congruence between treaty-making and treaty implementation, as ‘these are simply two steps of the same process’.58 However, this argument has never been accepted by the federal Government. It was even explicitly rejected in two white papers.59 Due to the lack of own treaty-making power, the provinces are limited to participation in the negotiation of treaties concluded by the national Government. Although involvement of the provinces is by no means compulsory, the federal Government has a strong incentive to cooperate at least regarding treaties that affect jurisdictions of the provinces and thus require their legislative action for implementation. A case in point for problems without this involvement is the Kyoto Protocol. Both the presentation of a Canadian position without provincial consent and, in addition, the unilateral announcement that the national Parliament would soon ratify the protocol had seriously alienated the provinces and contributed to problems of implementation.60 Other tensions, albeit only between the federal Government and Quebec, are related to the United Nations Educational, Scientific and Cultural Organisation (UNESCO), which the predominantly Francophone province places great importance on. After several controversies, a 2006 intergovernmental agreement, granted Q uebec the right to have a permanent representative within Canada’s UNESCO delegation and to be consulted before a formal position is taken. By contrast, cooperation mechanisms have already been in place for a longer period of time in foreign economic policy, in the case of the General Agreement on Tariffs and Trade (GATT) negotiations even since the 1970s.61 This does not, of course, fully exclude conflict on the intensity of cooperation. For example, the promised ‘full participation’ of the provinces in the negotiations of the Canada–United States Free-Trade Agreement, the precursor of NAFTA, implied, in their view, formal presence in the Canadian delegation. However, they could not overcome the national Government’s resistance and eventually had to settle for regular consultation meetings.62 This compromise solution did not prevent Ontario and Manitoba from remaining in opposition to the agreement until the end. (4) As indicated above, the implementation of a treaty is a prerogative of the provinces insofar as one of their jurisdictions is concerned. This is anything but s traightforward 58 Quebec’s education minister, Paul Gérin-Lajoie, in 1965, quoted in C Morin, L’Art de l’impossible: La Diplomatie québécoise depuis 1960 (Montreal, Boréal, 1987) 28 (author’s translation). 59 See P Martin, Federalism and International Relations (Ottawa, Queen’s Printer, 1968); and MW Sharp, Federalism and International Conferences on Education: A Supplement to Federalism and International Relations (Ottawa, Queen’s Printer, 1968). 60 See CJ Kukucha, ‘From Kyoto to the WTO: Evaluating the Constitutional Legitimacy of the Provinces in Canadian Foreign Trade and Environmental Policy’ (2005) 38 Canadian Journal of Political Science 129, 146ff. 61 See R Simeon, ‘Important? Yes. Transformative? No: North American Integration and Canadian Federalism’ in H Lazar et al (eds), The Impact of Global and Regional Integration on Federal Systems: A Comparative Analysis (Montreal, McGill–Queen’s University Press, 2003) 154f. 62 See DM Brown, ‘The Evolving Role of the Provinces in Canadian Trade Policy’ in DM Brown and MG Smith (eds), Canadian Federalism: Meeting Global Economic Challenges? (Kingston, Institute of Intergovernmental Relations/Institute for Research on Public Policy, 1991) 93ff.

Canada 435 because an explicit and unambiguous constitutional regulation is again missing. Section 132 of the Constitution Act 1867 only provides that the Canadian Parliament and Government ‘shall have all powers necessary or proper for performing the obligations of Canada or any province thereof, as part of the British Empire, towards foreign countries, arising under treaties between the Empire and such foreign countries’. While this provision clearly identifies the implementation of treaties as a federal competence, it only refers to treaties of the British Empire, obviously because back in 1867 treaty-making power remained in London. But what was to happen when the Canadian Government assumed this power for most treaties in 1926 and formally for all of them in 1947? The Judicial Committee of the Privy Council (JCPC) had to decide this question. 14.B Canada 1937: Treaty Implementation Follows Domestic Distribution of Powers Between 1919 and 1928, the International Labour Organisation (ILO) adopted three conventions under which its members, among them Canada, committed themselves to enacting laws concerning a limitation of working hours, weekly rest and a minimum wage level. As part of the ‘Canadian New Deal’, the federal Parliament adopted three laws to fulfil this obligation. The JCPC held that Section 132 of the Constitution Act 1867 did not apply in this case because its wording restricted federal authority to perform treaty obligations to those binding Canada ‘as part of the British Empire’.63 Moreover, it argued that ‘there is no such thing as treaty legislation as such. The distribution is based on classes of subjects’ (p 351). In order to ascertain whether giving effect to a treaty comes under federal or provincial responsibility, one was therefore supposed to disregard, unlike in the above-mentioned US case,64 the implementation function of the laws in question and to focus instead on their content. The laws in question, dealing with conditions of employment in the industrial sector, would fall within the provincial power regarding ‘property and civil rights in the province’ (Section 92(13)). That the internal distribution of powers also has to be observed when Canada engages in external relations was vividly illustrated by the famous ‘watertight compartments’ metaphor:65 ‘While the ship of state now sails on larger ventures and into foreign waters she still retains the watertight compartments which are an essential part of her original structure’ (p 354).

The fact that the JCPC decided that the implementation of treaties followed the internal distribution of powers is still extremely relevant today. As outlined above, it does not only grant the provinces a role regarding the implementation of certain treaties, but also provides an incentive to the federal Government to consult them a priori. Moreover, it has prompted the national Government to often seek the inclusion of a ‘federal state clause’ in its treaties, which makes the fulfilment of its obligations dependent on proper implementation by the provinces.66

63

Canada (AG) v Ontario (AG) [1937] AC 326 (JCPC), (Labour Conventions). See box 14.A. 65 On the watertight compartments doctrine, see ch 5.2.1. 66 See R Knopff and A Sayers, ‘Canada’ in Kincaid and Tarr (n 29) 125. 64

436 External Relations 14.4 BELGIUM

(1) The Constitution of Belgium endows the king, ie the federal Government, with the power to direct international relations. However, it also grants communities and regions extensive competences to regulate international cooperation, even including the conclusion of treaties, for all matters that fall within their competences (Articles 127(1), 128(1), 167(1) and 167(3) of the Belgian Constitution). This alignment of internal and external powers, known as the in foro interno, in foro externo principle,67 was introduced for the communities in 1988 and for the regions in 1993. The said principle makes the scope of subnational external powers exceptional, even if their exercise is subject to certain restrictions:68 first, regions and communities have to act with respect to federal loyalty (Article 143(1)); second, they must remain in line with the broad orientations of Belgian foreign policy, for instance in case of an international embargo; third, they are obliged to inform the federal Government, which then decides within 30 days whether or not to suspend negotiations. The Interministerial Conference of Foreign Policy (Article 31 of the 1980 Special Law on Institutional Reforms), composed of both federal and subnational representatives, eventually makes the final decision by consensus.69 If these mechanisms fail and the international activities of a subnational entity contradicts international law or EU law, the federal Government can act as a substitute if the subnational entity concerned was convicted by an international court like the Court of Justice of the European Union (Article 169). In practice, however, the rule of substitution has never been invoked. (2) The principle of in foro interno, in foro externo gives rise to an unequalled presence of Belgium’s subnational entities in the international arena. In delegations to international organisations that deal with matters within their exclusive responsibility, they are even exclusively represented. In such cases, the federal Government has to rely fully on this representation because it lacks ministries for these matters, as well as corresponding administrative capacities. With regard to the many organisations, such as the Council of Europe, UNESCO and the Organisation for Economic Cooperation and Development (OECD), whose activities touch upon areas of both national and subnational competence, representation is regulated in an intergovernmental cooperation agreement of 1994. One organisation with exclusive relevance for the French Community is the Organisation international de la Francophonie, in which the community is represented by its own delegation in addition to the Belgian one. Similarly, Flanders focuses its external relations in the field of cultural affairs on the Nederlandse Taalunie (Dutch Language Union).

67 See H Dumont et al, ‘Kingdom of Belgium’ in A Majeed et al (eds), Distribution of Powers and Responsibilities in Federal Countries (Montreal, McGill–Queen’s University Press, 2006) 44ff. 68 See P Bursens and F Massard-Piérard, ‘Kingdom of Belgium’ in Michelmann (n 21)97f. 69 See Y Lejeune, ‘La conduite des relations internationales’ in F Delpérée (ed), La Belgique fédéral (Bruxelles, Bruylant, 1994) 321.

Belgium 437 The Flemish–Francophone bipolarity regarding subnational activities abroad is further reinforced by the integration of the international relations agencies of all predominantly French-speaking entities70 into a joint administrative unit called Wallonie-Bruxelles International. This body, on the one hand, and the corresponding Flemish bodies, on the other, maintain a large network of representatives in foreign countries and within international organisations. In most cases, these attachés are hosted by Belgian diplomatic missions and integrated into its hierarchy but are functionally instructed by the respective subnational entity.71 Most remarkable and most important are certainly the subnational representatives in Belgium’s Permanent Representation to the European Union, which deal with almost all EU policies of relevance to their competences and give them considerable influence on Belgian positions.72 Another indicator of the close European integration of Belgium’s subnational entities is their traditionally strong commitment to cross-border cooperation, with the Meuse-Rhine Euroregion being the earliest and probably most significant example. More recently, they have also been making use, mainly with French partners, of the possibility to establish European Groupings of Territorial Cooperation (EGTC), which enjoy legal personality under EU law.73 While the above-mentioned cultural relations of Belgium’s subnational entities parallel each other and are thus uncontroversial, controversy regarding the orientation of international activities has sometimes emerged from the diverse structures of their economies. For example, the interest of the Walloon industry to export arms clashed with the restraint of Flemish parties in the federal Government to issue the necessary licences. The typically Belgian solution was then in 2003 to transfer the competence regarding arms trade to the regional level. This pragmatic approach to solving problems as they arise and usually by means of decentralisation is typical of Belgian federalism in general and of external relations in particular.74 (3) As far as treaty-making is concerned, the substantial decentralisation in combination with the in foro interno, in foro externo principle provides extraordinary opportunities for the regions and communities. In areas of their exclusive competence, they also have the exclusive right to conclude treaties with foreign countries or subnational entities (Article 167(3) of the Belgian Constitution) and are only subject to the obligation to inform the King about treaty negotiations. Although the monarch is entitled to stop negotiations and even to cancel an already-concluded treaty, these powers are confined to certain specific cases, ie a lack of diplomatic recognition of

70 These are the French Community, the Walloon Region and the French Community Commission of the Brussels-Capital Region. 71 See D Criekemans, ‘The Foreign Policy and Diplomatic Representation of the Belgian Regions: Flanders and Wallonia Compared’ in F Requejo (ed), Foreign Policy of Constituent Units at the Beginning of 21st Century (Barcelona, Collecció Institut d’Estudis Autonòmics, 2010) 48. 72 State’ See J Beyers and P Bursens, ‘The European Rescue of the Federal (2006) 29 West European Politics 1057. 73 See ch 1.2.3. 74 See K Deschouwer, ‘Kingdom of Belgium’ in Kincaid and Tarr (n 29) 54.

438 External Relations the envisaged partner by the federal Government or a collision of the treaty concerned with Belgian obligations under international or supranational law.75 In the many cases of mixed treaties whose content pertains to both subnational and national competences, approval has to be granted by the parliaments of both government levels. Paradoxically, the regions and communities therefore have veto power concerning international treaties on matters for which constitutional amendments could be made entirely without their participation.76 The mandatory ratification of mixed treaties by the subnational parliaments concerned has also led to the practice of integrating their delegates already at the negotiation stage. The same rules also apply to the mostly mixed treaties in the context of the EU that are regularly co-negotiated by subnational representatives in the Belgian delegation at the ministerial and diplomatic level. It is quite evident that both exclusive subnational treaties and mixed treaties require intergovernmental cooperation as a guarantee for a coherent Belgian foreign policy. To this end, the 1993 Special Law on the International Relations of the Communities and Regions introduced the instrument of cooperation agreements. Though most of these agreements also include the federal Government, some of them foster otherwise relatively weak horizontal cooperation. For instance, an agreement concluded in 1993 obliged commercial attachés from the regions to provide mutual support and to take part in activities on behalf of other regions if they themselves were not represented. Even more important is a cooperation agreement of 1994 on policymaking within the EU because it established the Directorate of European Affairs at the Foreign Ministry. This central institution of coordination organises sectoral meetings of federal and subnational representatives. Usually, these meetings succeed in reaching a consensus on a common position for negotiations so that a transfer of a matter to higher levels, the Interministerial Conference for Foreign Policy or even the Concertation Committee of the prime ministers, is rather rare.77 The second major novelty of this fundamental 1994 agreement are the rules for Belgian representation in the Council of the European Union. These create a system with six different configurations and varying degrees of subnational involvement that depends on the extent of subnational competences concerning each subject matter.78 In mixed delegations, which are led according to the subject matter either by the federal or subnational representative, the head of delegation is responsible for 75 See F Delmartino, ‘Belgium After the Fourth State Reform: Completed Federalism or Confederalism in the Making?’ in G Färber and M Forsyth (eds), The Regions—Factors of Integration or Disintegration in Europe? (Baden-Baden, Nomos, 1996) 124. 76 According to the formula still based on the unitary original Constitution of 1830, an amendment merely requires two-thirds majorities in both newly elected houses of the Belgian Parliament (Art 195(5) of the Belgian Constitution). 77 See J Poirier, ‘Formal Mechanisms of Intergovernmental Relations in Belgium’ (2002) 12 Regional and Federal Studies 24. 78 Regions and communities have exclusive representation through a common delegate regarding exclusive subnational competences, one subnational representative in mixed delegations with federal leadership concerning predominantly federal powers, mixed delegations led by a subnational representative concerning predominantly subnational competences and no representation concerning exclusive federal powers. In 2003, two more categories were added. In the case of competences that are relevant, like fisheries for Flanders, only for one regional government, Belgium is represented by this government. Concerning agriculture, the federal Government takes the lead even though it is an exclusive regional power.

Spain 439 casting the vote for Belgium while this person’s actions are politically controlled by the delegate from the respective other level of government. A serious practical problem involving these mixed delegations is similar to the delegations for treaty negotiations that the regions and communities have to coordinate their often divergent positions and that their representatives must rotate every six months. In contrast to the continuous federal representation, which speaks with one voice, this usually weakens the subnational part of these delegations.79 (4) The central principle of in foro interno, in foro externo also applies to the implementation of international treaties80 and legislation of the European Union. On the other hand, the state reform of 1993 also introduced a new competence for the federal Government to give effect to a treaty for the sake of fulfilling international or supranational obligations (Article 169 of the Belgian Constitution). However, due to its cumbersome procedure,81 this substitution mechanism has so far only been of symbolic importance.82 First, it requires that a relevant international court, for instance the Court of Justice of the European Union, must have condemned Belgium. Only then can a special law be adopted that authorises implementation by the federal Parliament or Government, unless the necessary steps are taken by the subnational entities within three months.

14.5 SPAIN

(1) The external relations of Spain’s autonomous communities have developed significantly since the end of the Franco regime,83 whose centralism had suppressed their international activity. Quite remarkably, this revival has occurred in spite of a constitutional framework that generally reserves the conduct of international relations (Article 149(1)(3) of the Spanish Constitution) and, more specifically, foreign representation and diplomacy (Articles 56(1) and 63(1)) as well as treaty-making (Articles 93, 94 and 96) to the central Government. What nonetheless enabled a considerable role for the autonomous communities was a turnaround of the Constitutional Court’s jurisprudence. 79 See W Swenden, ‘The Belgian Regions and the European Union: Unintended Partners in Unravelling the Belgian State?’ in RW Jones and R Scully (eds), Regionalism and Constitutionalism in an Enlarged European Union (Basingstoke, Palgrave Macmillan, 2009) 138ff. 80 Belgian Constitutional Court No 33/1992. 81 Art 16(3) of the 1980 Special Law on Institutional Reforms, as amended by the 1993 Special Law on the International Relations of the Communities and Regions. 82 See B de Witte, ‘Regional Autonomy, Cultural Diversity and European Integration’ in S Ortino et al (eds), The Changing Faces of Federalism: Institutional Reconfiguration in Europe from East to West (Manchester, Manchester University Press, 2005) 213. 83 See L Moreno, ‘Ethnoterritorial Concurrence in Multinational Societies: The Spanish Comunidades Autónomas’ in A-G Gagnon and J Tully (eds), Multinational Democracies (Cambridge, Cambridge University Press, 2001).

440 External Relations In the early years after the Constitution of 1978, its interpretation of Article 149(1)(3) was highly restrictive, as it considered any international territorial dimension sufficient to invoke this provision even if the subject matter concerned came under exclusive regional competence.84 While the Court declared in 1989 a cooperation agreement on environmental protection between Galicia and Denmark to be unconstitutional, it also conceded for the first time in this seminal case that there was certain room for regional activity in the international arena either by participating in the federal treaty-making power or by regions exercising their own competences.85 In the aftermath of this decision, the Court delivered more and more judgments that pointed in this less restrictive direction.86 The most important one of these was a decision in 1994. 14.C Spain 1994: Official Representation of the Basque Country in Brussels Within the European Union, the Basque Country has been particularly active from the beginning. Before all other autonomous communities, it had established a private-law-based representation in Brussels as early as in 1986. The next step should then be a fully-fledged and official EU delegation. This met with fierce opposition on the part of the Spanish Government and finally resulted in a highly publicised legal case before the Constitutional Court. In a landmark decision, the Court ruled that the autonomous communities are in principle allowed to conduct external relations if this is necessary or at least highly relevant to the exercise of their competences.87 However, all international activities can only be undertaken on condition that they do not run contrary to general Spanish foreign policy and refrain from interfering with the core competences of the central Government. The latter include the conclusion of international treaties, as well as the establishment of diplomatic missions.

The case law of the Constitutional Court is not the only reason why subnational powers regarding external relations are greater than suggested by the text of the Constitution. Another factor is the partial deconstitutionalisation of the distribution of powers, ie its partial regulation in the statute of each autonomous community. Whereas some autonomy statutes like those of Catalonia and Andalusia foresee comprehensive special rights to be involved in treaty negotiations and implementation, as well as other matters of external relations, the statute of La Rioja, for instance, does not go beyond the standard established by the Constitution. The complementary function of the statutes and the fact that they are bilaterally negotiated with the Spanish Government entail a highly asymmetrical and dynamic legal framework. Although both the case law of the Constitutional Court and the statutes have reduced the restrictive impact of Article 149(1)(3), the Spanish Constitution still

84

STC 44/1982. STC 137/1989. 86 See, eg, STC 153/1989 and STC 80/1993. 87 STC 165/1994. 85

Spain 441 reserves the above-mentioned core competences of foreign affairs for the national Government.88 Beyond that, it also commits the latter to overseeing the external relations of the autonomous communities in order to guarantee the coherence of Spanish foreign policy (Article 97).89 (2) Within this legal framework, all autonomous communities have become active in the international arena, albeit with varying degrees of ambition.90 The driving forces behind this development have certainly been the Basque Country and Catalonia. The first has always espoused a vivid tradition of external relations91 and demanded recognition of competences regarding this area in the Constitution of 1978, albeit to no avail. Catalonia has far-reaching external powers enshrined in its reformed statute of 2006, which came to be emulated by other autonomous communities. First and foremost, increasing international activities of the regions are embodied by their representations abroad. All of them maintain delegations in Brussels, as well as a varying number of trade offices worldwide. Whereas fostering external relations for the sake of economic development is a common concern of all autonomous communities, the aim of cultural promotion abroad is only pursued by some of them like Andalusia, Galicia and, in particular, Catalonia through the Consorci Català de Promoción Exterior de la Cultura. Similar to many other EU countries, Spanish regions actively promote cross-border cooperation by using the instrument of establishing EGTCs, including French and Portuguese partners. By contrast, international activities that are quite peculiar to the Spanish autonomous communities are the exceptionally strong commitment to grant development aid, as well as the support of their diasporas. As one might expect, the historic nationalities of Catalonia, the Basque Country and Galicia have the most solid bonds with their communities abroad, even though these contacts are emphasised not only in their statutes but also in those of Andalusia, Asturias and Extremadura. (3) As far as treaty-making is concerned, the monopoly of the national Government is beyond dispute. Ever since the Constitutional Court’s above-mentioned ruling in 1994,92 the ius contrahendi has been recognised as belonging to the core competences of international relations covered by Article 149(1)(3) of the Spanish Constitution. The scope of the autonomous communities to conclude international agreements is therefore reduced to certain atypical accords that do not entail obligations for Spain under international law, such as joint declarations, protocols, etc. After initial

88 ibid. 89

STC 26/1982 and STC 44/1982. E Aja, ‘The Debate about Competences in Spain’ in G Robbers (ed), Reforming Federalism— Foreign Experiences for a Reform in Germany (Frankfurt am Main, Peter Lang, 2005) 143. 91 See A Ugalde Zubiri, ‘The International Relations of Basque Nationalism and the First Basque Autonomous Government (1890–1936)’ (1999) 9 Regional and Federal Studies 170. 92 STC 165/1994. 90 See

442 External Relations r eservations on the part of the Constitutional Court, this practice has recently gained widespread acceptance and has led to hundreds of such agreements with partners as diverse as foreign countries, other subnational entities, international organisations or NGOs.93 When the Spanish Government exercises its treaty-making power, the autonomous communities do not have, in general, a right to participate in negotiations. However, many of the recently reformed autonomy statutes introduced certain changes concerning this matter. Only a few of them allow representatives of the autonomous communities to take part in the delegation responsible for negotiations. Most statutes merely guarantee a right to be informed about international treaties with obligations for an autonomous community, as well as the right to ask the central Government to enter into negotiations on issues of regional interest. In practice, intergovernmental cooperation takes place first and foremost at sectoral conferences, ie thematically structured multilateral conferences that deal with both the domestic and international dimension of a specific subject matter. Within this framework, the representatives of the national Government and the autonomous communities discuss Spain’s position for the negotiation of international treaties, as well as EU decision-making. In the latter case, the Spanish Government is obliged to defend the common position of the autonomous communities if their exclusive competences are concerned.94 However, this mechanism is rather ineffective. As the consequences of disregarding this obligation are only vaguely formulated, coordination within both the sectoral conferences and the overarching Conference on Affairs Related to the European Union often rather has the character of a non-binding dialogue.95 Within the latter conference, an agreement was reached in 2004 that grants autonomous communities the right to nominate two delegates to Spain’s Permanent Representation to the EU and, in certain cases, even direct representation in the Council alongside the Spanish delegate. (4) Whereas the Spanish Constitution explicitly mentions the power to conclude treaties, it remains silent on their implementation. Therefore, the case law of the Constitutional Court is once again highly relevant. In the leading case, the Court declared that the authority to implement both international treaties and EU law would follow the distribution of domestic powers.96 In the same judgment, however, it also emphasised the importance of the implementation process being guided by the cooperation principle implied by Article 2 of the Constitution.97 That the autonomous communities do have certain competences in this regard is also derived from provisions concerning implementation in most statutes.

93

See F Aldecoa and N Cornago, ‘Kingdom of Spain’ in Michelmann (n 21) 257f. J Bengoetxea, ‘The Participation of Infra-State Entities in European Union Affairs in Spain: The Basque Case’ in Weatherill and Bernitz, The Role (2005). 95 See De Witte, ‘Regional Autonomy’ (2005) 211f. 96 STC 252/1988. 97 STC 18/1982 and STC 46/1990. 94 See

Comparative Conclusions 443 14.6 COMPARATIVE CONCLUSIONS

(1) Since Kenneth Wheare claimed that a monopoly of foreign affairs would be a minimum power of national governments,98 reality has changed significantly. Today, external relations by subnational entities are widely recognised, albeit in quite different ways and to different extents. In a number of cases, only the foreign affairs powers of the national government are explicitly entrenched in written law, ie in the United States and Switzerland in constitutional provisions and in Canada, for historical reasons, in a royal letters patent. In all these three classical federations, subnational external relations have been nonetheless accepted from the beginning as being complementary to, albeit limited by, those relations exercised by the national government. By contrast, in Spain and Belgium, the external powers of the subnational entities are also guaranteed by provisions of written law. Although the Spanish Constitution remains silent on this issue, relevant provisions have been included in almost all statutes. This has been enabled by a turnaround of the Constitutional Court’s jurisprudence that authorised autonomous communities to maintain external relations as long as they are linked to their domestic competences and do not encroach upon certain national core competences. In Belgium, the Constitution itself explicitly guarantees external powers of the subnational entities as well. It even recognises them as a result of the in foro interno, in foro externo principle to an exceptional extent. (2) On this legal basis, subnational entities maintain external relations with varying motivations and scope in terms of content and partners. The most pragmatic and arguably oldest motivation is what has been aptly termed ‘cross-border h ousekeeping’,99 ie jointly resolving issues that do not stop at international borders such as environmental protection and infrastructure. In this regard, there is an evident difference between the analysed cases. While cross-border cooperation has been facilitated in the Canadian and US case by common NAFTA membership, the EU provides a still stronger impetus for such collaboration. A second and particularly widespread motivation for external relations is, in all cases examined, the promotion of economic interests. It is obvious that activities in this regard are far more important for subnational entities in countries like Canada, whose prosperity depends significantly on foreign trade, than in the United States, which can rely much more on its large domestic market. This difference is reflected in the fact that the Canadian provinces have a clearly more strategic and continuous approach to the promotion of their economic interests abroad. Both cross-border housekeeping and economic external relations usually do not put a strain on national–subnational relations and are often even actively promoted by the

98 99

See KC Wheare, Federal Government, 4th edn (Oxford, Oxford University Press, 1967). Kincaid, ‘Foreign Relations’ (2003) 86.

444 External Relations federal Government. Cases in point are the international treaties to facilitate crossborder cooperation of Swiss cantons and the know-how and structures provided by Belgian embassies to trade offices of the regions. The opposite is obviously true for international activities fuelled by cultural or political motivations, even more so as these often focus on kin-states. It took the Canadian Government, for instance, more than a century to recognise direct relations between Quebec and France and to consent to the province’s representation in the Organisation international de la Francophonie. By contrast, in the Belgian case cultural external relations on the part of the subnational entities do not create controversy because there is no strong national government to oppose them. The Swiss cantons in turn are not even interested in high-profile cultural relations. This is illustrated by the fact that, unlike the subnational units of Canada and Belgium, they are not represented at UNESCO. In some cases, external relations are used by culturally diverse subnational entities also to underline their ambition to be recognised as a nation. This is illustrated by the fact that Quebec strives to maintain external relations that are thematically nearly as comprehensive as those of a state or the clear preference of Flanders for maintaining relations with states rather than other subnational entities. Regardless of their motivation, subnational entities have to employ a wide array of measures like foreign visits, promotional events, etc in order to maintain continuous external relations in an effective way. In this regard, they have to be able to rely on stable structures abroad, which in some cases collide with the aim of national governments to monopolise international representation. In general, national governments have therefore been rather hesitant to allow subnational units truly autonomous representation. Most far-reaching, structured and active are certainly Quebec’s offices abroad. But even these institutions, due to the province’s lack of legal personality under international law, can only function within framework agreements that the federal Government concludes for them. The Swiss cantons, constitutionally restricted to relations with subnational entities, do not even make use of this limited scope to set up a stable institutional structure abroad. This contrasts with the strong international presence of Belgium’s regions and communities, as the in foro interno, in foro externo principle allows them, in combination with the high degree of decentralisation, to be represented abroad regarding a number of subject matters. Though their attachés are, unlike the representatives of Quebec, housed in embassies, they are nonetheless instructed by their respective government and act in a quite autonomous way. But apart from these specific cases, it is hard to neglect the fact that the ius legationis is still largely a domain of national governments. (3) With regard to subnational entities’ own ius contrahendi, the cases analysed show quite a diverse picture. Besides the Swiss cantons, only the subnational entities of Belgium enjoy, again on the basis of the in foro interno, in foro externo principle, the right to conclude international treaties on subject matters within their own areas of jurisdiction. If exclusive competences of the regions or communities are concerned, this even excludes the corresponding federal treaty-making power. While duties to inform the national government have to be observed in both countries,

Comparative Conclusions 445 the legal p osition of the Swiss cantons is clearly weaker. First, unlike in Belgium their treaties can be prevented by the national Government in case of a collision with any federal or cantonal law and even federal interests. Second, treaties entered into by the cantons can be abrogated by federal treaties or legislation. As a result of these restrictions, but arguably also due to self-restraint, they rarely use their treaty-making power. This contrasts with the exceptionally numerous treaties of the Belgian regions and communities.100 In both the United States and Spain, the constitution reserves the right to enter into formal international treaties to the national government so that the states and autonomous communities are confined to concluding certain atypical agreements. The same is true for the Canadian provinces, even if the exclusive ius contrahendi of the federal Government, which is derived from the general transfer of royal authority through the 1947 Letters Patent, is heavily contested by Quebec on the basis of the so-called Gérin-Lajoie doctrine. While subnational entities of these three countries lack their own treaty-making power, their involvement in the process of treaty-making at the national level varies greatly. Effective consultation of US states occurs only sporadically and coordination within the sectoral conferences in Spain has a largely non-binding character. The potentially more powerful mechanism of direct subnational participation on the Spanish negotiation team is, in turn, merely foreseen in the statutes of a few autonomous communities. In Canada, by contrast, the provinces are involved to a considerable degree even in the absence of a corresponding constitutional obligation. This follows from the fact that the federal Government has to rely in areas of concurrent or exclusively provincial jurisdiction on provincial cooperation concerning the implementation of a treaty and thus grants them in exchange a say at the negotiation stage. As a rule, the controversial question is therefore not whether the provinces participate but to what degree. In contrast to de facto cooperation in Canada, procedures for participation are even formalised in Belgium and Switzerland. The cantons have demonstrated since 1999 in several instances that these procedures, laid down in detail in ordinary legislation, have enabled them to influence treaties, in particular when a consensus among them has endowed the Conference of Cantonal Governments with a strong bargaining position. Moreover, through the typical Swiss instrument of mandatory referendums concerning certain international treaties, the populations of the cantons are also given a direct say. The Belgian regions and communities enjoy, in the many cases of mixed treaties that affect areas of concurrent jurisdiction, not only a right to participation but even a veto right. This strong legal position led in turn to the practice of including their representatives on the negotiation team. In Belgium and Spain, the impetus for collaboration is equally evident in the domestic rules on EU decision-making, which are seen as an at least partial compensation for the transfer of subnational legislative powers to the European level. At the heart of the Belgian rules are the crucial role of intergovernmental consensusbuilding under the aegis of the Directorate of European Affairs, as well as the system 100 See D Criekemans, ‘Regional Sub-State Diplomacy from a Comparative Perspective: Quebec, Scotland, Bavaria, Catalonia, Wallonia and Flanders’ (2010) 5 The Hague Journal of Diplomacy 37.

446 External Relations of differentiated representation of the regions and communities in the Council of the European Union. By comparison, direct representation of the autonomous communities on the Council is less extensive and coordination in the sectoral conferences beforehand less binding. Nevertheless, this gives them, in combination with other channels such as the Committee of the Regions and their offices in Brussels, still clearly more influence than the Canadian provinces and US states have within NAFTA. Although NAFTA has important implications for subnational entities, it offers them only a few points of access. On the other hand, they also do not campaign for such representation as intensively as subnational entities within the EU. Considering the fact that NAFTA is, of course, much less integrated and thus absorbs fewer subnational competences, this is quite understandable. In summary, the participation of subnational entities in the exercise of national treaty-making power seems to be even more significant than those entities’ own ius contrahendi, which is, if it exists at all, in most cases subject to significant procedural limits and barely used. As the right of national governments to enter into treaties extends, apart from Belgium, even to areas of exclusive subnational jurisdiction, their treaty-making is normally far more frequent and thus simply more important in quantitative terms. In most cases, cooperation mechanisms work to the benefit of both the national and subnational governments. From the perspective of national governments, cooperation is an alternative, and more lenient, way of safeguarding coherence of foreign policy, which is in the face of its liability under international law certainly a legitimate interest. From a subnational point of view, their participation in the treaty-making of national governments is crucial in order to compensate for the loss of certain competences in the wake of globalisation and European integration. Following this rationale, subnational entities in Switzerland and Belgium campaigned successfully for guarantees for such participation whose entrenchment in the 1990s not only concerns external relations but also reflects a general shift towards cooperative federalism of these systems as a whole. (4) Regarding the implementation of international treaties, federal systems face a particular problem that does not emerge in unitary states. In the latter, once the national government has concluded a treaty it can, under normal circumstances, easily guarantee the passage of the legislation necessary to give effect to it. By contrast, the national government of a federal country may be forced to rely on implementation measures by the subnational entities if the issue at stake would fall internally within their competences. Whether the national government is able to fulfil its liability for implementation under international law therefore ultimately depends on the willingness of the subnational units to cooperate. Basically, there are two ways to solve this problem, as the distribution of powers can either emphasise the content or the function of the legislation that gives effect to a treaty. In the first case, the power to pass this legislation is assigned according to the domestic competence regarding the subject matter at stake. In the second case, it is included in the—as a general rule national—treaty-making power.

Comparative Conclusions 447 Since the Labour Conventions case of 1937, it has been clear that Canada follows the first pattern. The same is true for Spain and Belgium, where—as a result of constitutional jurisprudence in the first case and of the in foro interno, in foro externo principle in the latter—the content of a treaty determines the power to implement it. While also the cantons of Switzerland are responsible for the implementation of a treaty concerning their areas of jurisdiction, the country’s Constitution goes a step further. It provides the national Government with the possibility to make the cantons beyond that responsible for treaties affecting areas of federal jurisdiction. This clearly epitomises the traditional Swiss preference for administrative federalism.101 In contrast to all other cases analysed, the United States follows the second pattern, as the US Supreme Court regards the competence for the implementation of a treaty as being comprised by the exclusive federal power concerning treaty-making. In the context of the Supremacy Clause, the Court thereby reduced the legislative scope of the states significantly. Legally speaking, this judicial interpretation also had the peculiar effect that a federal law with the same content may be constitutional or unconstitutional depending on whether it has the function of implementing a treaty or not.

101

See ch 3.3.2 and 5.3.

15 Concluding Remarks 15.1 FEDERALISM IN THE TWENTY-FIRST CENTURY

F

EDERALISM MEANS DIFFERENT things in different contexts, depending on the legal tradition, on history, on political culture and on the interpretation provided by the political actors and not least by the courts. Any universal definition that goes beyond the very broad, almost generic formula by Daniel Elazar (‘self-rule plus shared rule’) faces exceptions and is contradicted by comparative evidence. In other words, there is no general recipe for the right mix between s elf-rule and shared rule. From a legal perspective, a pragmatic, institutional approach is mandated. The instruments of what we have labelled the ‘federal toolkit’ are in the end the features that explain the functioning of federal systems in practice: institutions; procedures; relations between norms; mechanisms for conflict prevention and resolution; guarantees for the existence and participation of subnational entities and financial relations. The ultimate test for this toolkit is provided by judicial interpretation (to which this book pays special attention), as well as by the practical way to manage complex policy fields, such as those analysed by way of sample in Part III of this book. The importance of federal studies in present times goes beyond the remarkable attention they attract in the academic community. Their relevance is growing for a number of other reasons too. The political success of the federal idea and even more so of the federal toolkit is due to its ability to tackle the challenges posed by factors such as ethno-cultural diversity and the management of large territories, of complex policy areas, or of multilevel democracy. Diversification of responses within a common constitutional framework also allows for tailor-made solutions, for experimentation, for wider participation, for enhancing democracy by dividing and sharing powers. Only the careful observation of such phenomena in a comparative perspective, having regard to their practical developments (thus looking carefully at the judicial interpretation of federal issues), makes it possible to understand the functioning of institutions and constitutional systems, and the challenges ahead. In contemporary societies, including non-federal ones, decision-making has become an extremely complicated task. Decisions are made by an increasing number of actors, whose diverse mandate often exceed mere political-electoral legitimacy (such as in the case of courts). These actors are arrayed both vertically (levels of government) and horizontally (parliaments, governments, agencies, courts, administrations, interest groups). All these norm suppliers coexist and shape decision-making and implementation in each subject area, and at each level. Subject matters become

A Federal Research Agenda: The Four ‘Ps’ 449 more articulated the more society and technology evolve and the claim for responses from public bodies becomes stringent. Thus, jurisdictions tend to overlap as no field can be clearly separated from others, and the legal and administrative regulation of each competence matter is subject to ever more entangled norms and procedures produced by several authorities at different levels, often in an manner that makes the necessary coordination extremely cumbersome. When the idea of federalism was born, it was primarily a philosophical concept and a principle of political organisation, essentially serving economic and military purposes. In legal terms, the early idea of federalism was closer to international rather than constitutional law, aimed at bringing together sovereign units that alone were no longer competitive in economic and military terms. This was the case for historic federations, notably the United States, Switzerland, and Germany. The more federal countries established themselves (nineteenth century) and significantly increased in number (twentieth century), the more federalism became noteworthy not only to political philosophers and political scientists, but also to constitutional lawyers and economists. All of these methodological perspectives facilitated the study of how federal systems (both classical federations and related manifestations such as regional or devolved states) work in practice, what elements they have in common, how their functioning can be improved, and, above all, what institutions and procedures are needed in order to make them work. In the twenty-first century, the major challenge is no longer the creation of new federations by pooling together sovereign states, as the example of the EU and its refusal to become a fully-fledged federal state clearly shows. And even the establishment of federal systems through decentralisation has lost the grip it had in the course of the second half of the twentieth century. The federal idea has been sufficiently explored: the box was assembled, the tools are inside. Institutional strengths and weaknesses are rather well-known, and so is the importance of cooperative instruments and of the readiness to cooperate by the involved actors, as there can be no federalism without a certain degree of cooperation. While all this seems to be consolidated, the critical test that remains is the effective management of pluralism and its inherent complexity. This means that federalism can no longer be seen as a pure institutional interplay, a system accommodating the coexistence of institutions belonging to different tiers of government but has to include the understanding of how federalism works (or may work) in practice, what its added value could be and how it could be developed. 15.2 A FEDERAL RESEARCH AGENDA: THE FOUR ‘Ps’

Federalism (and the federal toolbox) is so far one of the most developed constitutional instruments to deal with the pluralism of interests, actors, institutions, procedures and with the challenges this poses to contemporary societies. The lessons that can be learned from comparative federal studies help identify answers to complex governance issues. Thus, to familiarise with the instruments developed in federal systems is a gateway for a better understanding of social phenomena and for designing possible (legal) solutions to them.

450 Concluding Remarks While the ‘mechanics’ of federalism and its inherent role as an instrument to manage complexity have not changed in the course of history, the challenges of contemporary federalism (and consequently of federal studies) are obviously different as compared to centuries or decades ago. The study of institutional structures is nowadays no longer sufficient to design solutions to solve complex problems. And giving up the ambition to solve complex problems equals abandoning the ideal and philosophical mission of the federal thought. Against this background, the main challenges to federal studies for the years to come can be summarised in terms of the four ‘Ps’: managing pluralism by means of procedures, participation and policies. Managing pluralism, in its various territorial, ethno-cultural and other manifestations, has always been the core task of federal systems. Traditionally, however, this was done primarily using institutional structures, such as second chambers. While institutional structures have been and still are important in determining the nature and the operational mode of federal countries, they are with no exception undergoing re-thinking, reconsideration, and, frequently, processes of substantial reform. New, less institutionalised bodies or merely procedural arrangements are being established to supplement decision-making and more generally to overcome the structural limits of institutional arrangements. Focusing on procedures aimed at regulating decision-making, at coordinating the implementation of decisions, at preventing and settling conflicts, at including non-institutional actors (interest groups, experts, citizens, non-governmental organisations, etc) is of utmost importance in understanding federal dynamics. Participation has also been an essential element of the federal toolkit from the inception. However, traditional institutional forms of (territorial) participation are no longer sufficient to establish a workable system. Not only are rules on participation essentially procedural, determining who can participate, how, under which conditions, exercising what rights, and—not least—what the consequences of inclusion in or exclusion from the process are (for example, whether those excluded have the right to challenge the decision in front of a court or not). In present times, the pressure towards more democratic and participatory decision-making makes it necessary to look beyond merely institutional participation and to also include (and regulate) forms of societal participation. The ways non-institutional stakeholders can be included in decision-making processes can more easily be designed digging in the federal toolkit. Finally, it is necessary for federal studies to look more closely at policies, including how they are managed on the basis of legal norms and how they are interpreted by courts. Focusing on exemplary judicial cases helps understand how federalism is lived and how it impacts on political actors and private citizens and how these, in turn, prompt the development of the federal toolkit. Furthermore, the analysis of relevant policy fields, both related to the legal division of powers and to the political processes, illustrates the trends and challenges of governance. For this reason, both intergovernmental policymaking on one hand and case law on the other are given particular attention in this book. The detailed analysis of the management of competence areas provides an extraordinary litmus test for investigating and conceptualising what CJ Friedrich called the

A Federal Research Agenda: The Four ‘Ps’ 451 ‘trends of federalism in theory and practice’. Some of the areas analysed in this book, such as environmental protection, immigration and migrant integration, or social welfare and healthcare, are among the most significant areas of law that are subject to regulation and implementation by a plethora of governmental levels and actors, duly challenging the traditional approach to the vertical division of powers among levels of government. In fact, looking at how policies are managed throws into sharp relief the very idea of a rigid division of competences, as hardly any activity can be reduced to one single function to be assigned to one individual level. On the contrary, this analysis epitomises the modern trend affecting a growing number of issues that are better understood in terms of multifaceted competence matters rather than in terms of clear-cut powers belonging to only one level of government. A number of significant consequences follow from this trend. First, in order to more accurately describe and interpret the functioning of federal systems, it is helpful to look at how complex competence areas are managed rather than merely focusing on institutional settings and constitutionally established division of powers, which in many cases are just but indicative. Second, the demarcation of deeply rooted categories like federal and regional states fades away and has to be called into question when the rather similar management of many competence areas is taken into account. Third, a modern understanding of these areas requires forms of coordination that often go beyond political and administrative borders, by providing for instruments of cooperation among different public authorities. Fourth, in terms of policymaking, it seems more important to focus on how to coordinate actions and powers by different levels and actors rather than trying to rigidly divide the jurisdictions of each level, since overlap and contradictions are simply unavoidable. Fifth, the more inter-institutional and inter-governmental coordination is lacking, the more critical is the role played by the judiciary in finding rational ways to determine the division of work among the inevitably numerous players. Sixth, the claim for more participation goes far beyond the sole institutional sphere (government), and includes individuals and societal groups, thus requiring more inclusive, not necessarily simple majority-based, decision-making procedures (governance). To look further into these aspects, as this book has tried to do, will represent the next, fascinating chapter of the long journey of federalism. This is a journey that has been going on since the emergence of public power, and will continue as long as public power exists.

452

Index A accountability democracy 119 intergovernmental relations 249, 256 Adams, John 74 administrative efficiency competition 225 taxation 222, 225 administrative federalism generally 79, 156–159, 319 Germany 81, 156, 157, 158–159 Swiss Confederation 79, 156, 447 administrative powers autonomy of subnational entities 126, 157–159 legislative powers, relationship with 156–159 residual clauses 139–146 subnational participation at national level 178, 193 African Union 60 aggregative federalism (or coming-together federalism) constitutional autonomy 126–127 generally 9, 40, 42–43, 45–46, 52, 65 participation of subnational entities 164 use of term 9 Althusius, Johannes 19, 83, 86 Ambedkar, BR 38–39, 106, 154 Amphictyonic League 67 anarchism federalism and, generally 14, 16 APEC macro geo-economic region 30 appointment processes subnational participation at national level 178, 193–199 Aquinas, Thomas 19 Arab League 60 Argentina constitutional review 277 direct democracy 118 federal nature 45, 46 Fiscal Responsibility Law 237 intergovernmental relations 255 judicial appointments 195 judicial system 162 residual powers 140 second chamber 165, 169, 171, 191 Aristotle 19, 86 assembly, right to Bavaria 334–335 associated states generally 34, 59, 60

Association of Southeast Asian Nations (ASEAN) generally 60 macro geo-economic region 30 associative democracy 117 asymmetry asymmetric federalism generally 9, 47 constitutional autonomy 129 financial relations 208–210 Italy 209–210 North German Confederation and German Empire 80 Spain 209 United Kingdom 210 Australia aggregative federalism 45, 46 British colonial origins 6, 43, 46 conservative resistance to progressive agenda 320 Constitution 43, 321 constitutional adjudication 267–268, 272 constitutional amendment procedure 182, 183 constitutional anti-secessionism 106 constitutional autonomy 127 constitutional review 275, 276, 277 Council for the Australian Federation 255 Council of Australian Governments 255–256 creation 46 customs tariffs 212 delegation of powers 257 dual federal system 46, 156 equalisation 242 executive federalism 253 federal, use of term 9 federal nature 43, 45, 127 federal spending power 231–232 Financial Framework Legislation Amendment Act 231–232 fundamental rights 321 horizontal cooperation 255 immigration and migrant integration 403, 411–412, 418–419, 421 inconsistency of state legislation with a federal law 132 intergovernmental agreements 259 intergovernmental relations 255–256, 257 judicial appointments 195 judicial system 162 legalism 272 legislative power 146, 149 local government 285, 292–293, 296, 306 multicultural policies 412, 421 municipalities 296, 306

454 Index National School Chaplaincy Programme 231–232 naturalisation 412 participatory democracy 137, 182 popular sovereignty 182 residual powers 140 Senate 165, 169, 171–172, 191 state constitutions 127 state legislatures 137 tax-raising powers 223 territorial changes 187 Austria administrative federalism 156, 157 Conference of the Länder Governors 157, 177 Conference of Speakers the Länder parliaments 254 Congress of Vienna 80, 189 constitutional adjudication 6 constitutional amendment procedure 184 constitutional autonomy 134 Constitutional Court 39, 134, 275 constitutional review 274, 275 cooperative federalism 148 creation 43 definition of federal state 38–39 direct democracy 118 distribution of powers 140, 147, 157–158 enforcement of national legislation 157–158 environmental protection 250–251 EU membership 199 European Fiscal Compact (TSCG) 237, 302 European Union 23 Federal Council (Bundesrat) 166, 169, 170, 172, 177, 184, 191, 199 federal loyalty principle 250–251 federal nature 16, 39, 40, 45, 46 federal and Länder legislation 133 Financial Constitutional Law 206, 219 Financial Equalisation Law 206, 218–219 financial relations 206 homogeneity clause 134 indirect federal administration 158, 159 intergovernmental agreements 260 intergovernmental relations 254, 259 judicial appointments 197 judicial system 160 legislative procedures 191 local government 285, 287–288, 291, 294–295, 297, 298, 306–308, 310 metropolitan areas 309 municipalities 285, 287–288, 291, 294–295, 297, 306–308 participatory democracy 117 principle of mutual consideration 148, 250 referendums 184 residual powers 140 second chamber 50, 169 Länder legislation, scrutiny 262–263 tax-raising powers 218–219 territorial changes 187

autonomous regions financial autonomy and its limits 224–225, 234–235 Italy 59, 207 Spain 206–207, 234–235, 236 terminology 9 autonomy of subnational entities administrative powers 126, 157–159 asymmetrical autonomy 129 commerce clause doctrine 142–145 constitutional autonomy 126–139 subnational constitutions, power to adopt 126 double aspect doctrine 145, 356, 378 elements limiting 321, 341–343, 379 generally 48, 49, 58–59, 125–126, 317 government systems 137 homogeneity and 318 homogeneity clauses 19, 40, 130, 133–136, 321 identity affirmation 138–139 judicial powers 126, 159–163 legislative power 126, 139–156 meaning 58 national constitution guaranteeing 126 parliamentary systems 137–138 presidential systems 137 regional states 50 self-rule and shared-rule 3–4, 8, 10, 18, 39, 48, 66, 125–126, 448 sovereignty and 130 subnational legislatures 137 supremacy clauses see supremacy clauses territorial 58–59 utilisation of constitutional autonomy 344–345 B bald man paradox 3 Balfour Declaration 432 Barnett formula (United Kingdom) 207–208, 210, 372, 379 Basque Country external relations 440–441 financing arrangements 53, 209 healthcare 369, 370 minimum income scheme 368, 381 Plan to Fight Poverty 368 representatives at EU 440, 442 status 52–53 tax competition 226 tax-raising powers 222, 226 Bedouin tribes 67 Belau micro-federation 45 Belgium see also Flanders; Wallonia ‘alarm bell’ procedure 154 principle of comity 148, 249 communities and regions 9, 21, 23, 26, 170, 206, 213–214, 223, 247, 416–417 Constitution 126

Index 455 Constitutional Court 271, 279–280 cross-border cooperation 437 cultural promotion abroad 437, 444 decentralisation 361, 362–363, 377, 378, 380, 415, 444 devolutionary federalism 45, 46 distribution of powers 140, 146, 154, 318, 380 dual federal system 249 environmental protection 247 equalisation 240, 241, 244 ethno-cultural diversity 154, 170, 437 European Groupings of Territorial Cooperation (EGTC) 437 European Union 23, 437, 438–439, 446 external relations 436–439, 443, 444, 445–446, 447 federal nature 40, 43, 45 federal and subnational legislation 133 federal spending power 233 financial relations 206, 213–214, 223 Flemish Care Insurance 362–363 foreign relations see external relations above healthcare 365, 380 horizontal transfers 364–365 immigration and migrant integration 415–418, 419, 420, 421 in foro interno, in foro externo 436, 437–438, 439, 443, 444, 447 inburgering 416, 418, 422 intergovernmental agreements 259 intergovernmental cooperation 247 Interministerial Conference of Foreign Policy 436 international treaty negotiation and implementation 436, 437–439, 444–446, 447 judicial system 160 labour-market policies 378, 416 language issues 437 Maastricht criteria 235–236 monarch 436 nationalism 101 naturalisation 421 OECD 436 Organisation international de la Francophonie (OIF) 436 as regional state 55 residual powers 140 Senate 50, 170, 175 social welfare 361–365, 376, 377–378, 380 subnational autonomy 126, 129 subnational representatives to EU 437, 438 tax-raising powers 213–214, 223 UNESCO 436 Beveridge, William 370 bicameralism see second chambers Bismarck, Otto von 81 Bodin, Jean 84, 85, 86, 88 borrowing local government 302–303

Bosnia and Herzegovina federal nature 43, 44, 45 financial arrangements 223 internal minority population 104–105 legislative competence 147 multinational federalism 6, 99, 223 power-sharing 104–105 Republika Srpska 99, 104–105, 223 residual powers 140 second chamber 176–177 state and entities 9 tax-raising powers 222, 223 Brandeis, Justice Louis laboratory federalism 318–319 Brazil civil service appointments 198 conditional transfers 301 constitutional adjudication 268 constitutional anti-secessionism 106 democracy 118 dual federal system 156 federal nature 45 Fiscal Responsibility Law 237 intergovernmental relations 255, 258 judicial appointments 195 legislative power 149 local government 283, 286, 287, 290, 300, 301, 314, 315 municipalities 283, 286, 287, 290, 314 residual powers 140 revenue sharing 221 second chamber 165, 169, 171 tax competition 226 tax-raising powers 221, 226 Brennan, Justice William J 324 Brown, Gordon 375 Bryce, James 318 C Calhoun, John C 87, 89 Canada see also Quebec aggregative federalism 45, 46 Balfour Declaration 432 bi-lingual status 331 Bill of Rights 330, 333, 342 British colonial origins 6, 43, 46, 431–432 Canada Assistance Plan 261, 358, 379 Canada Health Transfer 358 Canada Social Transfer 358, 379 Charlottetown Conference 254 Charter of Rights and Freedoms 180, 330–331, 332–333, 342, 408, 409 City of Toronto Act 306 Commerce Clause 144–145 concurrent legislative powers 132–133 conservative resistance to progressive agenda 320, 381 Constitution Act (1867) 43, 169, 330, 343, 359, 418, 431 Constitution Act (1982) 180, 330–331

456 Index constitutional adjudication 267–268, 269–270, 271–272 constitutional amendment procedure 180–182 constitutional autonomy 126, 127 constitutional review 275–276, 277 Council of the Federation 255 creation 46 cross-border cooperation 433 customs tariffs 212 distribution of powers 132–133, 148, 149, 257 double aspect doctrine 145, 356, 378 equalisation 211, 240, 241–243, 245 ethno-cultural diversity 103, 234, 308, 331, 432–433 executive federalism 156n, 253 external relations 431–435, 443–444, 446, 447 federal, use of term 9 federal nature 45, 47, 49 federal spending power 230–231, 232–233, 234, 235 financial relations 202–203, 208–209, 211, 212 First Ministers’ Conferences 178, 255, 256 fiscal discipline 238–239 foreign relations see external relations above free movement within Canada 408 freedom of expression 331 fundamental rights 330–333, 343 Gérin-Lajoie doctrine 433–434, 445 graduated subnational participation 180–182 healthcare 359–361, 378–379, 380, 381 immigration and migrant integration 403, 407–411, 418–419, 421, 422 ‘inside-outside bill’ 331 intergovernmental agreements 208–209, 217, 221, 259, 261 intergovernmental bodies 177, 178 intergovernmental relations 254–255, 256, 257, 269–270 international treaty negotiation and implementation 433–435, 445, 447 judicial appointments 195–196 judicial system 160, 162–163 language rights 331, 342, 344 legislative power 132–133, 148, 149 living tree doctrine 145, 271–272, 356, 357 local government 285, 292, 297, 298–299, 306, 308 local government financing 202–203, 214–215 Medicare 360, 379 multicultural policies 409, 421 municipalities 306, 315 natural resources 211, 242 naturalisation 409 New Deal 356, 435 North American Free Trade Agreement (NAFTA) 25, 30, 426, 433, 443, 446 Notwithstanding Clause 270, 331, 342

parliamentary sovereignty 261, 332 participatory democracy 118, 120, 122 peace, order and good governance (POGG Clause) 149 pension provision 356, 357–358 provinces 9, 49 provincial constitutions 332–333, 343 Provincial Nominee Program (PNP) 408, 419 Quebec Conference 169 Quiet Revolution 432 reference jurisdiction 277 residual powers 140 Rowell-Sirois Commission 217 secessionism 101, 108–109 Senate 50, 165, 169, 177, 191 ‘seven-fifty’ formula 181, 187 shared-cost programmes 358, 379 Social Union Framework Agreement 359 social welfare 346, 355–359, 376–377, 378–379, 381 sovereignty-association 108–109 subnational autonomy 126 Supreme Court 196, 332, 343 tax decentralisation 242–243 tax-raising powers 212, 214–215, 217, 221–222, 223 territorial changes 187 unemployment insurance 356 urbanisation 315 capitalism federalism and 13–14 Catalonia ‘armour-plated’ competences 138, 141 bill of rights 344 centralisation of powers 289 Consell de Garanties Estatutàries 339 cultural promotion abroad 441 Declaration of Sovereignty 110 equalisation within Spain 244, 245 external relations 440–441 generally 26 identity affirmation 138–139 language laws 138–139, 340 local government 314 nationalism and regionalism 26, 101 residual clause 141 rivalry with Barcelona 309 secessionist movement 110, 141, 245 social welfare and healthcare 368, 369 Statute of Catalonia 138–139, 141, 244, 339–340, 345, 440 tax-raising powers 222 utilisation of autonomy 344 Catholic social doctrine Quadragesimo Anno 19 central government use of term 9 centralisation Catalonia 289 subnational resistance to 246

Index 457 trend towards centralised taxation 222–228 United States 425 China Special Economic Zones 22–23 civil law countries constitutional review 274–275, 277 fundamental rights 321 intergovernmental relations 248 judicial systems generally 163 sovereignty 87 civil partnerships same-sex 324, 330, 344 civil servants appointment process 193, 198 Clinton, Bill 349–350, 354 coercive federalism 148, 233–234 common law countries constitutional review 275–277 dual judiciary and federal common law 161–162 fundamental rights 330 intergovernmental agreements 260 intergovernmental relations 248 judicial systems generally 163 sovereignty 87–88 ultra vires principle 292–293 Commonwealth of Independent States status 37 Commonwealth of Nations status 37 Comoros 7, 45 comparative method generally 5–7 competition regarding taxation and social welfare horizontal tax competition 225–228 race to the bottom 348, 353, 359, 370, 380–381, 385 Swiss Confederation 242 United States 308, 353, 380 vertical tax competition 215–217 competitive federalism see laboratory federalism complex power-sharing 104 conditional transfers use, generally 300–301 condominiums 60 generally 34 confederations definition 35–36 Dutch Republic of the Seven United Provinces 68–69, 70 generally 34–35, 43, 48 historical background 68–72 legal elements 35 Old Swiss Confederacy 7, 9, 35, 36–37, 68, 69–70 Serbia and Montenegro 37 transitional nature 36–38 United States of America under the Articles of Confederation 36–37, 68, 70–72 conflict prevention and resolution Canada 269–270

conflicts over distribution of power 246 constitutional adjudication 266–280 cooperation principle 251–252 delegation 257–261 disputes over natural resources 210–211 Ethiopia 266 federal oversight 246, 261–266 first and second chambers, disagreements between 192–193 flexibilisation and centralisation through implied powers 149–156 generally 246, 448 intercircuit conflicts 269 intergovernmental relations 246–261 South Africa 252, 269 subnational constitutions, scrutiny 263–266 subnational legislation, scrutiny 246, 261–266 subnational resistance to centralising trends 246 Swiss Confederation 141, 200, 252, 269 United Kingdom 252–253 Congress of Vienna 80, 189 conservative resistance to progressive agendas 319–320, 344, 357, 381 constitutional pluralism see pluralism constitutional review civil law countries 274–275, 277 common law countries 275–277 constitutional tolerance 96 constitutions asymmetrical autonomy 129 constitutional adjudication 6, 266–280 constitutional amendment procedure 178–185 constitutional autonomy 126–139 contents of subnational constitutions 129 democratic see democracy flexibility and rigidity 270 form of subnational constitutions 128 homogeneity clauses 19, 40, 130, 133–136, 321 interpretation 271 limits of subnational constitutional space 321 national 126, 141 partisan 113 political or judicial scrutiny of subnational constitutions 263–266 subnational 126–139, 141 supremacy of national/federal 39, 48, 321 Cook Islands 60 cooperation see also cooperation principle; cooperative federalism cross-border 31–33, 60–61 distribution of powers 148 as element of federal states 39 cooperation principle see also cooperation; cooperative federalism administrative federalism 157 Austria 148, 250 between legislatures 253–256 Canada 269–270

458 Index generally 251–252, 449 Italy 148, 251 social welfare provision 376 South Africa 147, 148, 298 Spain 148, 251 cooperative federalism distribution of power 146–149 environmental protection 401 generally 9, 46 Germany 16, 46 ‘joint-decision trap’ 16, 148 legislative powers 146–149, 253–256 Council of Europe Congress of Local and Regional Authorities 23, 28, 284 Framework Convention for the Protection of National Minorities 32 Madrid Outline Convention 32 courts and judiciary appointment process 193, 195–198 autonomy of subnational entities 126, 159–163 civil law countries 163 common law countries 163 constitutional adjudication 6, 266–280 Constitutional Courts 266–280, 336 constitutional review 272–280 distribution of powers, adjudication 141–142 draft legislation, judicial scrutiny 261–263 dual systems 160–162 environmental policy, judicial interpretation 398–399, 402 European Court of Human Rights (ECtHR) 78 European Court of Justice (CJEU) 63–64, 93 fundamental rights at subnational level, judicial interpretation 321, 336 generally 159–163 integrated systems 160, 162–163 judicial restraint 271 judicial review 155, 267, 274, 277–280, 425 residual clauses 139–146 subnational constitutions, judicial scrutiny 263–266 Supreme Courts 267, 278–279 US dual system 322–323, 343 US Judicial Procedures Reform Bill (‘Court Packing Plan’) 349 cross-border cooperation see transboundary issues Czechoslovakia creation 43 ethno-federation 45, 98, 100 D data protection Germany 337 death penalty Germany 334 United States 323–324, 325–326 decentralisation autonomy and 58–59

Belgium 361, 362–363 devolution 55–58 economic theory 224–225 environmental protection 383 meaning 58–59 regional states 52–55 Spain 52–53 Switzerland 255 Viennese School of Legal Positivism 314 decision-making generally 448–449 declinist debate regarding social welfare and healthcare 348 Delors, Jacques 26 democracy see also subnational participation at national level; suffrage assemblies 77–78, 118–119, 122, 137, 307 associative 117 audience-democracy 116 Austria 117, 118, 134 Canada 118, 120, 122 constitutional autonomy 137–138 crisis of democracy 114–122 deliberation day 118 deliberative 117, 119 deliberative polls 118 deliberative process 118 digital 119 direct 77–79, 114–116, 118, 134, 137, 182–183, 307, 344 efficiency of participatory democracy 120–121 federalism and non-democratic countries 7 federalism and participatory democracy 114 Germany 118 India 118 indirect 120–121 international law and direct democracy 78 Italy 120–121, 122 Latin America 117, 118, 137 legitimacy and 121–122 liberal 115 majoritarian 112–113 parliamentary sovereignty 56–57, 74, 261, 271, 332 parliamentary systems 137–138 participatory 83, 111–122, 137 participatory budgeting 122 partisan constitutions 113 pluralism and participatory democracy 83, 111 political accountability 119 popular initiatives 137, 183 populism 114–115 post-democracy 116 problem of scale of participatory democracy 119 referendums 23, 37, 48, 53–54, 56, 77–78, 107–110, 114, 115, 117, 120, 137, 184, 189–190, 285, 308, 390, 400, 406, 431, 445

Index 459 representative 115–116 second chambers see second chambers subnational legislatures 137 Swiss direct democracy 115, 118, 344 Swiss Landsgemeinde 77–78, 118, 137, 307 unpolitical 116 US direct democracy 118, 137, 344 Denmark Greenland 59 devolution administrative 371 delegated powers 148, 257–258 devolved legislation 262, 269, 278 devolved states generally 8 distribution of power 55–56 ‘English question’ 56 executive 398 generally 449 parliamentary sovereignty and 56–57 reversibility 56 United Kingdom 6, 23, 55–58, 148, 207–208, 210, 252, 257–258, 262, 371, 379 Dicey, Albert V 15, 34, 115, 266 Dillon’s rule municipalities 291–292, 293 diplomacy see external relations direct democracy see democracy ‘direct federalism’ United States 351 disabled persons Belgium 362 Spain 366, 367–368 United States 351–352 disputes see conflict prevention and resolution distribution of powers adjudication 141–142 administrative powers 126, 157–159 autonomy see autonomy of subnational entities Belgium 146, 318 centralisation 149, 156, 289 commerce clause doctrine 142–145 concurrent competences 145, 146–147, 149, 388, 391, 392, 394–395 conflicts over 246 constitutional adjudication 267–268 constitutional autonomy 126–139 cooperative federalism 146–149 delegated powers 63, 148, 257–261, 290, 296–297, 377, 394 detailed or broadly defined 140–141 devolution 55–56 dual federalism 146–149, 151, 156 enumeration of powers 139 European Union 63, 144, 423, 424 external relations 424, 443 federal states, generally 48 flexibilisation and centralisation through implied powers 142, 149–156 framework legislation 147 generally 125, 317, 318 homogeneity and 140

implicit concurrency 149 implied powers 149–156 Implied Powers Theory 150–151 ‘joint-decision trap’ 16, 148 judicial powers 126, 159–163 legislative and administrative powers, relationship 156–159 legislative power see legislative power local government 290–305 national interest principle 153–154 political culture and 141–142 principle of comity 249 regional states 55 residual clauses 139–146 shared competences 10, 147–148 Spain 318, 379, 440 spending power and 378–379 taxation see taxation tripartite 146, 249 UK devolution 262, 377 United States 425 diverse societies see also ethno-cultural diversity pluralism and 112 draft legislation judicial scrutiny 261–263 dual federalism administrative powers 156 Australia 46, 156 Belgium 249 Canada 46, 145, 156 cooperative elements 148 distribution of power 146–149, 156 federal states 48 generally 9, 46, 156, 281 legislative powers 146–149, 151, 156 sovereignty 46 Swiss Confederation 46 United States 46, 151, 156, 158–159, 216 Dutch Republic of the Seven United Provinces 68–69, 70 E economic development local government 296 New Regionalism 25–26 economic organisation federalism as principle of 13–14 education as fundamental right 344 immigration and migrant integration 404, 405, 409, 410, 411, 416–417, 419–420 local government 296 Scotland 372 Spain 369 United States 289 elderly, provision for Belgium 362–363 United Kingdom 371–373 United States 351–352

460 Index emergency powers subnational participation 178, 200 energy matters United Kingdom 396, 399 environmental protection see also pollution Austria 250–251 Belgium 247 climate change policy 385, 386, 387, 400, 426 ‘command and control’ policies 382 as constitutional objective 399 cooperative federalism 401 decentralisation 383 delegated legislation 394 economic instruments 382 EU policy 382, 397, 399–400 generally 317–318, 382–383, 398–402 Germany 337, 391–393, 399, 400–401 globalisation 383 greenhouse gas emissions 387, 392 horizontal cooperation 402 intergovernmental regulation 247, 250–251 interrelation with other policy areas 383 Italy 393–396, 400–402 judicial interpretation 398–399, 402 local government 296 New Environmental Policy Instruments 382–383 nuclear energy/waste 390, 393 race to the bottom 385 renewable energy plants, siting 395 Swiss Confederation 387–390, 399, 400–401 tradeable permit schemes 382, 392 transboundary issues 383, 387, 398, 401, 426, 443 United Kingdom 396–398, 399, 401, 402 United Nations 382 United States 382, 383–387, 398–399, 400, 401, 426 vertical cooperation 401 voluntary agreements 382 Western Climate Initiative 255, 387, 402, 426 equalisation Australia 242 Austria 206, 218–219 Belgium 240, 241, 244 Canada 211, 240, 241–243, 245 Germany 239, 240, 241, 242, 243–244, 245 horizontal imbalances 240, 242, 244 horizontal transfers 241, 364–365 intergovernmental transfers 240 local government finances 304–305 reasonably comparable levels of public services 241–242 risk-pooling 240 second chambers 167, 169 Spain 243, 244, 245, 379 subnational taxing powers 240 Swiss Confederation 241, 242 United States 167, 169, 242, 379

vertical fiscal gap 201, 240 vertical transfers 240–241, 304–305 Ethiopia constitutional disputes 266–267 Eritrean secession 107–108 federal nature 45 legislative powers 149 multinational federalism 6, 99 second chamber 176–177 ethno-cultural diversity amalgamations 308 autonomous regions 59 autonomous tax collection 222 Belgium 154, 170, 437 Bosnia and Herzegovina 6, 99, 223 Canada 103, 234, 308, 331 Czechoslovakia 98 Ethiopia 6, 99 ethno-federations 45, 98, 100 federalism and, generally 97–111 generally 448 identity affirmation 138–139 intergovernmental relations 248 language rights 331, 340 legal status of internal minorities 101–105 local government 300, 308 Macedonia 97–98 minorities within minorities 102 multinational federalism 6, 98–101, 102 national interest principle 154 nationwide minorities 98–99, 102 power-sharing 104–105 regional minorities 98–99, 102 risk of secession 98, 100–101 second chambers 169–170 Soviet Union 98 Spain 52 Swiss Confederation 79, 300 Yugoslavia 98 Eubulides bald man paradox 3 euro zone Maastricht criteria 235–236 European Coal and Steel Community (ECSC) 92–93 European Convention on Human Rights (ECHR) direct democracy in the Swiss Confederation 78 EU accession 152 fundamental rights in the Swiss Confederation 327–328 European Union Austria 199 Basque representatives 440 Belgium 23, 437, 438–439, 446 ‘Brexit’ 57, 96, 110–111 Committee of the Regions 25, 28–29, 30, 283, 446 confederal elements 37–38 Conference of European Regions with Legislative Power (REGLEG) 28–29

Index 461 constitutional pluralism 96–97, 111 constitutional tolerance 96 Council of the European Union 175, 424 cross-border regions 31–33, 60–61, 437 delegated competences 63 Directorate of European Affairs 438 distribution of powers 63, 144, 423, 424 environmental protection 382, 397, 399–400 euro zone 235–236 ‘Europe of the regions’ 27–28 European Charter of Local Self-Government (ECLSG) 284, 290, 304–305, 306, 308, 310 European Convention on Human Rights (ECHR) 152 European Court of Human Rights (ECtHR) 78 European Court of Justice (CJEU) 63–64, 93, 152–153 European Groupings of Territorial Cooperation (EGTC) 32–33, 437, 441 European Regional Development Fund (ERDF) 24 European Social Fund (ESF) 24 European Stability Mechanism (ESM) 203–205 eurozone 203 federal traits 37–38, 61–64 federalists and intergovernmentalists 62–63 financial relations 203–205 European Fiscal Compact (TSCG) 204–205, 211, 236–237, 302 flexibilisation through implied powers 151–153 foedus pacificum 61 fundamental rights 63 generally 248, 423, 424, 446 German Länder 20, 199 ideas behind 61–62, 92–93 in foro interno, in foro externo 152–153, 439 integration 93–94, 95–96, 97, 449 judicial review 425 local government 283–284, 290 Maastricht criteria 235–236 macro-regional strategies 30–31 multilevel governance (MLG) 26–27 Nomenclature des unités territoriales statistiques (NUTS) 24, 27 Outright Monetary Transactions (OMT) 204 principle of loyal cooperation 63 proportionality principle 63 ‘regional blindness’ 425 regionalism 14, 23–29, 30–33 Schuman Declaration 62, 92 single market 28 sovereignty 42, 92–97 Spain 23, 440, 442 subsidiarity principle 19–20, 28, 63 supranational federalism 6 supremacy of EU law 62–64

Swiss-EU bilateral agreements 78, 429 theory of counter-limits 63, 64 Treaty of Amsterdam 28 Treaty of Lisbon 19, 20, 30, 94–95 Treat of Maastricht 19–20, 25, 94–95 Treaty of Nice 28 Ventotene manifesto 61–62 executive federalism 156, 156n, 253 see also administrative federalism executive branch increasing influence 165 institutions of intergovernmental relations 253–256 intergovernmental agreements 258 supervision over executive branch activities 200 experimental federalism see laboratory federalism external relations Belgium 436–439, 443, 444, 445–446, 447 Canada 431–435, 443–444, 446, 447 cultural promotion abroad 432, 433, 441, 444 distribution of powers 424, 443 EU integration and 423, 424, 446 European Union 423, 424 foreign investment 426 generally 317–318, 423–425, 443 globalisation and 423, 446 international and supranational law 424, 444–445 paradiplomacy 423 pre-eminence of national governments 423, 443 Spain 439–442, 443, 445–446, 447 Swiss Confederation 428–431, 443, 444–445, 447 transboundary issues see transboundary issues treaty negotiation and implementation 424–425, 444–447 United States 425–428, 443, 445, 446, 447 F federacies 34, 59–60 federal use of term 9 (federal) loyalty principle Austria 250–251 cooperation principle 251–252, 396 European Union 63, 250 Germany 250 Italy 396, 401–402 Spain 251 Swiss Confederation 249, 264 federal states archetypes 72–82 autonomy see autonomy of subnational entities coming-together (or aggregative) 9, 40, 42–43, 52, 164 cooperation among levels 39, 449 creation 42–47 decision-making process 48, 49–50

462 Index definition 38–40, 47–50 dispute resolution 48 distribution of power 48 emergent 44–45 financial arrangements 48 foedus (compact) 18, 38–42, 44 Germany 79–82 holding-together (or devolutionary) 9, 42, 43 homogeneity clauses 40 institutional elements 39, 47–50 mature 44–45 micro-federations 45 ‘quasi-federations’ 45 role of history in shaping 42–47 self-rule and shared-rule 3–4, 8, 10, 18, 39, 48, 66, 125–126, 448 sovereignty 41–42, 48, 86–87 structure 125 Swiss Confederation 76–79 United States 72–76 use of term 8 federal system use of term 8 federalism administrative 156 aggregative (or coming-together) 9, 45, 65, 126–127, 137, 164 Althusius 86 anarchist movements and 14, 16 Anglo-American understanding 10, 14–15, 16 asymmetric 9, 47 autonomy see autonomy of subnational entities capitalism and 13–14 classification of federal countries 3–4 conceptualising 14–20 constitutional jurisprudence 2 cooperative see cooperation principle; cooperative federalism definition 2, 3–5, 13, 14–15, 16, 18–19, 38–39, 64–66, 448 devolutionary (or holding-together) 9, 45–46, 65; see also devolution dual see dual federalism economic organisation 13–14 elements of, generally 6, 8, 10, 48–49, 448, 449–451 executive 156 federation compared 17–18 forms 34–35, 42–47 freedom and 14 functional 16–17 German-speaking tradition 16–17 historical background 13–17, 67–82 institutional pluralism see pluralism ‘integral’ 14 interlocking 148 ‘layer cake’ 46 legislative 46, 156 ‘marble cake’ 46 multinational 6 nature, generally 1–2, 14–17

negative connotations 17 non-democratic countries 7 number of federal countries 3, 7 power sharing 10; see also distribution of powers as process 15 regional states and 65 regionalism and 14 restless systems 9, 47 self-rule and shared-rule 3–4, 8, 10, 18, 39, 48, 66, 125–126, 448 social organisation 13–14, 15–16 subsidiarity 19–20 supranational 6 symmetric 9, 47 use of term 8–10, 13 The Federalist 14, 73–74, 73n federations generally 34 use of term 8, 18 financial relations see also spending power; taxation administrative efficiency 222, 225 asymmetrical 208–210 Australia 212 Austria 206 autonomy and 379 Belgium 206, 223 Canada 202–203, 208–209, 211, 212 conditional grants/transfers 228–229, 233–234, 300–301, 350 credit controls on subnational entities 211 debt ceilings and financial autonomy 236 deconstitutionalisation 207–208 equalisation see equalisation EU bailout mechanism 203–204 European Stability Mechanism (ESM) 203–204 European Union 203–205 eurozone 203 federal states 48 financial autonomy and its limits 224–225 fiscal discipline 235–239 fiscal equivalence 228 fiscal federalism 202, 225 generally 448 Germany 202, 204–205, 208, 212, 228, 239, 250 global financial crisis 211, 302 horizontal disparities 201, 208–210 horizontal imbalances 240 India 203, 212–213 intergovernmental agreements 208–209, 217, 221 Italy 206, 207, 209–210 legal sources 205–210 local government 202–204, 206, 295, 300–305 public borrowing 211 revenues from natural resources 210–211 shared-cost programmes 358

Index 463 South Africa 203, 207, 225 Spain 206–207, 209, 223–224 Swiss Confederation 201, 205, 228 United Kingdom 207–208, 210 United States 202, 205–206, 208, 212 vertical fiscal gap 201, 240 vertical tax competition 217–218 Finland Åland islands 59 fiscal discipline Argentina 237 Brazil 237 Canada 238–239 debt ceilings and financial autonomy 236 European Fiscal Compact (TSCG) 236–237, 302 Maastricht criteria 235–236 Swiss Confederation 238–239 United States 238 fiscal equivalence 228 fiscal federalism see also financial relations competition 225 generally 202 Flanders external relations 436–437, 444 generally 26, 126, 241, 363 immigration and migrant integration 416–422 intergovernmental cooperation 247 nationalism and regionalism 26 social welfare provision 362–363, 364–365, 377, 417 taxation 240 foedus (compact) 18, 38–42, 44, 67 foreign relations see external relations France collectivités territoriales 23 European Groupings of Territorial Cooperation (EGTC) 437 institutional representation of the subnational units 50 regionalism 21, 22–23, 55 sovereignty 89, 95 Frantz, Constantin 16 freedom of association 334–335 of expression 331 free speech 324–325 individual and collective 14 of movement 337–338, 351, 379 functional federalism 16–17 fundamental rights see also individual rights absence of positive law 330 approval of subnational constitutions 342–343 Australia 321 bills of rights 341–342 Canada 330–333, 343 civil law countries 321 common law countries 330

conservative resistance to progressive agendas 319–320, 326, 344, 345 data protection 337 European Convention on Human Rights (ECHR) 78, 152, 327–328 education 344 environmental protection 337 European Union 63 free speech 324–325 generally 317–318, 321–322, 341–345 Germany 333–338, 341, 343, 344 horizontal emulation 345 housing rights 344 judicial interpretation 321 limits of autonomy 321, 341–345 negative/positive rights 344 precedence of national legislation 334–335, 341–342 right to work 344 Spain 338–341, 343, 345 supremacy clauses 322, 328, 329, 334, 335, 341 Swiss Confederation 327–330, 341, 343, 344–345 United Kingdom 330 United States 322–327, 341, 343 unwritten 327–328, 342, 345 utilisation of constitutional space 321–322 vertical emulation 344–345 welfare rights 344 G Gandhi, Indira 41 General Agreement on Tariffs and Trade (GATT) 434 Gérin-Lajoie doctrine 433–434, 445 German Confederation 167 German Empire 80 Germany administrative federalism 81, 156, 157, 158–159 aggregative federalism 45 General competence of local governments (Allzuständigkeit) 293–294 Ämter of Schleswig-Holstein 311 bailouts for Länder 239 Bundesrat (Federal Council) 42, 82, 165, 167–168, 172–175, 177, 191–192, 193, 199, 200, 256 citizenship tests 414–415 Conferences of the Speakers of the Länder parliaments 254 constitutional adjudication 269, 279, 280 constitutional amendment procedure 183 constitutional autonomy 127–128, 134–135 Constitutional Court 163, 275 constitutional jurisprudence 208 constitutional review 274–275 cooperative federalism 16, 46, 141, 148, 157 counties (Kreise) 281–282, 293–294, 312 data protection 337

464 Index death penalty 334 debt brake 237 deviation clause 391–392, 399 direct democracy 118, 137 distribution of power 146–147, 157, 158–159, 391–393 enforcement of national legislation 157 environmental protection 337, 391–393, 399, 400–401 equalisation system 239, 240, 241, 242, 243–244 eternity clause 40, 183, 186 EU level in financial relations 204–205 EU membership 20, 23, 63–64, 94–95, 199, 424 European Fiscal Compact (TSCG) 237, 302 extra-jurisdictional spending 229 federal law, precedence of 334–335, 341 federal legislation 200 federal loyalty principle 250 federal nature 8, 16–17, 43, 45 federation, formation 79–82 financial relations 202, 204–205, 239, 250 fundamental rights 333–338, 341, 343, 344 generally 6, 449 homogeneity clause 40, 133, 134–135 immigration and migrant integration 412–415, 419–421 intergovernmental agreements 208, 258, 259–260 intergovernmental relations 248, 250, 254, 256 international treaties 199 ‘joint-decision trap’ 16, 148 judicial appointments 197 judicial interpretation of fundamental rights at Länder level 336 judicial system 160, 162, 163 Länder 9, 20, 81, 82, 87, 91, 127–128, 134–135, 167 legislative emergencies 200 legislative power 146–147, 157, 158–159 legislative procedures 191–193 local government 202, 281–282, 286–287, 293–294, 299, 302, 310, 311, 312 as ‘malleable federal state’ 186 metropolitan areas 309 municipalities 202, 287, 293–294, 296, 303–304, 312, 414, 415 national interest principle 154 naturalisation 414–415, 420–421 North German Confederation 80 nuclear energy/waste 391, 393 Parliament (Bundestag) 173, 193 political culture 141 president, appointment 193 Prussian hegemony 80–81, 82 referendums 137 residual powers 139, 141 reunification 245, 337, 414 social welfare 302

sovereignty 91, 94–95 special-purpose associations (Zweckverbände) 310, 311 spending power 228, 237 subnational constitutions 127–128, 336–338 subnational participation at national level 165, 167 supremacy clause 132, 159, 334, 335, 341 tax-raising powers 212, 218, 219 territorial reforms 186–187, 310 Weimar Republic 80 global financial crisis austerity measures following 302 local government funding 302 public borrowing 211 globalisation generally 423, 446 intergovernmental relations 248 international interdependence 317 multinational companies 423 multinational federalism and 101 municipalities 296 New Regionalism 25 paradiplomacy 423 government systems constitutional autonomy 137 grants coercive federalism 233–234 conditional 228–229, 233–234 H Hamilton, Alexander 14, 230, 267 Hanseatic League 68 healthcare see also social welfare Belgium 365, 380 Canada 359–361, 378–379, 381 declinist debate 348 financing 348 generally 317–318, 346–348, 376–381 horizontal variation 379–380 intergovernmental cooperation 378 intergovernmental regulation 247 local government 296 Spain 369–370 United Kingdom 374–376 United States 242, 350, 353–355, 380, 381 ‘welfare magnets’ 348, 356 Heller, Hermann 80–81 Helvetic Republic 70 Hindenburg, Paul von 80 Hobbes, Thomas 84–85, 86 holding-together federalism (or devolutionary federalism) generally 9, 42–43, 45–46, 65 use of term 9 Holy Roman Empire 80, 85, 86 homogeneity Austria 134 distribution of powers 140

Index 465 Germany 40, 134–135, 333 homogeneity clauses 19, 40, 130, 133–136, 321 national 318, 333 South Africa 136 Swiss Confederation 40 United States 134, 135–136 homosexuality same-sex civil union/marriage 324, 326–327, 330, 344 United States 324, 325, 326–327 horizontal relations environmental protection 402 equalisation 240, 241, 242, 244 financial relations 201, 208–210 horizontal cooperation 254–256 intergovernmental agreements 258 local government 309–311 Swiss Confederation 255 tax competition 225–228 United States 259, 400 hourglass federalism 315 housing immigration and migrant integration 404, 405, 413, 417, 419–420 UK housing benefits 371 human rights see fundamental rights I immigration and migrant integration Australia 403, 411–412, 418–419, 421 Belgium 415–418, 419, 420, 421, 422 Canada 403, 407–411, 418–419, 421 citizenship tests 409, 412, 413 democracy and 116 education measures 404, 405, 409, 410, 411, 416–417, 419–420 generally 317–318 Germany 412–415, 419–421 housing 404, 413, 417, 419–420 interculturalism 409 labour market 404, 411, 416, 419 language measures 404, 405, 407–410, 412, 416, 418, 420 metropolitan areas 403 migrant crisis 403 multicultural policies 404, 409, 412, 421 naturalisation 404, 406–407, 409, 412, 414–415, 418, 419, 420–421 pluralism 116 political integration 404, 406, 409, 415, 420–421 Swiss Confederation 404–407, 418, 420, 421 United States 403 urbanization 282, 304 voting rights 404 welfare measures 404, 405, 410, 414, 416–417, 419–420 impeachment powers subnational participation 178, 200

in foro interno, in foro externo Belgium 436, 437–438, 439, 443, 444, 447 European Union 152–153, 439 independence autonomy and 58 India basic structure doctrine 41 British colonial origins 6 Commissions of Inquiry Act 279 conditional transfers 301 Constitution 126, 153–155 constitutional amendment procedure 182, 184–185 constitutional anti-secessionism 106–107 Council of States (Rajya Sabha) 50, 165, 171, 184–185, 191, 192, 200 credit controls 211 definition of federal state 38–39 delegated powers 148 establishment 43, 43–44n federal nature 7, 38–39, 40, 41, 43, 45 financial relations 203 flexibilisation and centralisation through implied powers 153–155 Inter-State Council 256 intergovernmental relations 256 Jammu and Kashmir 49, 130, 189, 213 judicial review 155, 275, 276–277, 279 legislative power 146, 148 local government 283, 286, 290–291, 295, 299, 301, 312, 313, 314 municipalities 283, 295, 314 panchayats 286, 290–291, 295, 312, 313 participatory democracy 118 president, appointment 193–194 public borrowing 211 residual powers 140, 141 subnational autonomy 49, 126, 130 subnational legislatures 137 tax-raising powers 203, 212–213, 221 territorial changes 189 as union 38–39, 106–107 indigenous communities see also ethno-cultural diversity local government 282 indirect federal administration in Austria 158, 159 infrastructure distribution of powers 140, 296, 376 local government powers 296 participatory democracy 137 transboundary cooperation 31, 226, 426, 429, 443 innovation laboratory federalism 318–319 ‘integral’ federalism 14 integration see immigration and migrant integration integrative federalism see aggregative federalism intergovernmental agreements Australia 259

466 Index Austria 260 Belgium 259 bilateral 258 Brazil 258 Canada 208–209, 217, 221, 258, 259, 261 common law countries 260 delegation 257–261 financial relations 208–209, 217, 221 Germany 208, 258, 259–260 horizontal 258 instruments 257–258 legal basis 259–261 legislative implementation 260 ratification 260 Spain 258, 259, 260 Swiss Confederation 258, 260–261 United States 259, 260 vertical 258 intergovernmental bodies subnational participation at national level 164, 177–178 intergovernmental cooperation see cooperation principle intergovernmental relations accountability 249, 256 Australia 255–256, 257 Austria 254, 259 Belgium 247 Canada 208–209, 217, 221, 254–255, 256, 257 civil law countries 248 common law countries 248 cooperation between legislatures 253–256 cooperation principle 251–252 disputes see conflict prevention and resolution environmental protection 247, 250–251 European Union 248 executive branch 253–256 executive federalism 253 federal loyalty principle 250–252 Germany 208, 248, 250, 254, 256 globalisation 248 horizontal cooperation 254–256, 309–311 horizontal mechanisms 248 India 256 Italy 251 local government 295, 297–300 regulatory state 247 South Africa 248, 249, 256 Spain 248, 251, 258, 367 Swiss Confederation 248, 249, 252, 255 transparency 249, 256 tripartite distribution of powers in Belgium 249 UK devolved entities 257–258 United States 254, 255 vertical cooperation 253–254 vertical mechanisms 248 welfare state 247 interlocking federalism 148 International Atomic Energy Agency (IAEA) 60

international interdependence generally 317 International Labour Organization (ILO) 60, 435 international relations see external relations international and supranational law subnational powers generally 424, 444–445 international treaty negotiation and implementation Belgium 436, 437–439, 444–446, 447 Canada 433–435, 445, 447 Quebec 433–434 Spain 439, 441–442, 445–446, 447 subnational powers generally 424–425, 444–447 Swiss Confederation 430–431, 444–445, 446, 447 United States 199–200, 425–428, 445, 447 Iraq Kurdistan region 21, 211 natural resources 210, 211 Iroquois League 67 Israel, 12 tribes of 67 Italy asymmetries 209–210 autonomous provinces 9, 23, 50, 54, 59, 207, 209–210 Chamber of Deputies 190, 193 valley communities (Comunità di Valle) 311 concurrent competences 394–395 constitutional adjudication 269, 279 constitutional amendment procedure 185 constitutional autonomy 127, 128, 129, 137 Constitutional Court 394 constitutional review 274, 275 cooperation principle 148, 251 credit controls 211 delegated legislation 394 devolutionary federalism 45 distribution of powers 147, 153, 157 enforcement of national legislation 157 environmental protection 393–396, 400–402 European Fiscal Compact (TSCG) 237, 302 European Union 23 federal nature 7, 17, 43, 45, 50 financial relations 206, 207, 209–210 harmonisation of public budgets 237 horizontal cooperation 311 immigration 403 institutional representation of the subnational units 50 legislative procedures 190 local government 297, 311, 315 loyal cooperation principle 396, 401–402 national interest principle 153 participatory democracy 120–121, 122, 137 president, appointment 193 public borrowing 211 regional governors 137 as regional state 23, 53–54 renewable energy plants, siting 395

Index 467 residual clause 139, 140, 141 Sardinia 51, 54 Senate 175, 177, 185, 190, 193 Sicily 54 Standing Conference for the Relations between the State and Regions 177–178, 396, 397 statutes of ordinary and special regions 129 tax-raising powers 223 territorial changes 187–188 unification 53–54, 65 J Jacobinism 14 Jay, John 14 Jefferson, Thomas 74, 76 Jellinek, Georg 16 Johnson, Lyndon B Great Society programmes 314, 349, 350, 351, 353 joint functional authorities cross-border cooperation 60–61 generally 34, 60 ‘joint-decision trap’ 16, 148 judicial powers see courts and judiciary jurisprudence constitutional 2 financial relations 208 Germany 208 K Kant, Immanuel 61, 86 Kelsen, Hans 80, 274 Keynesian–Westphalian states 113 L laboratory federalism 318–320 Scientific Management Movement 319 social welfare provision 376 laissez-faire politics conservative resistance to progressive agenda 319 language diversity and rights Belgium 437 Catalonia 138–139, 340, 344 federal societies 15 immigration and migrant integration 404, 405, 407–410, 412, 416, 418, 420 Organisation international de la Francophonie (OIF) 429, 432–433, 436, 444 Quebec 331, 333, 342, 344, 432–433 Swiss Confederation 79, 429 Laski, Harold 112 Latin American federations creation 43 generally 45 non-democratic 7 participatory democracy 117 ‘layer cake’ federalism 46 leagues 34, 60, 67–68 legal pluralism see pluralism

legislative federalism 46, 156–159 see also dual federalism legislative power administrative powers, relationship with 156–159 autonomy of subnational entities 126 centralisation 149, 156, 289 commerce clause doctrine 142–145 concurrent powers 145, 146–147, 149 cooperative federalism 146–149 de facto concurrency 145 delegated powers 148, 257–261 distribution generally 126, 139–156 double aspect doctrine 145, 356, 378 dual federalism 146–149, 151, 156 elusion 131 enforcement of national legislation 157–158 enumeration of powers 139 European Union 424–425 exclusive powers 146 flexibilisation and centralisation through implied powers 142, 149–156 formal concurrency 145 framework legislation 147 implied 149–156 ‘joint-decision trap’ 16, 148 national interest principle 153–154 pre-emption 131–132 residual clauses/powers 139–146 shared competences 10, 147–148 subnational participation at national level 164–165, 178 supremacy clauses see supremacy clauses Lincoln, Abraham 106 linguistic diversity see language; language diversity and rights local government amalgamations 305–309 assignment of powers 290–295 austerity measures impacting 302 Australia 285, 292–293, 296, 306 Austria 285, 287–288, 291, 294–295, 297, 298, 306–308, 310 autonomous powers 290, 296–297 borrowing 302–303 Brazil 283, 286, 287, 290, 300, 314, 315 Canada 285, 292, 297, 298–299, 306, 308 conditional transfers 300–301, 302 conflicting interests 314–315 core competences 296 Council of Europe 283–284 counties 281–282 decentralisation theory 314 deconstitutionalisation 290 definition 290 delegated powers 290, 296–297 economic development 296 education 296 environmental protection 296 equalisation 304–305 ethno-cultural diversity 300

468 Index European Charter of Local Self-Government (ECLSG) 284, 290, 304–305, 306, 308, 310 European Union 283–284, 290 financial resources 202–204, 295, 300–305 general competence principle 291, 293–295 generally 281–284, 314–315 Germany 281–282, 286–287, 293–294, 299, 302, 310, 311, 312 global financial crisis 302 healthcare 296 horizontal cooperation 309–311 hourglass federalism 315 India 283, 286, 290–291, 295, 299, 301, 312, 313, 314 indigenous local governments 282 intergovernmental relations 295, 297–300 Italy 297, 311, 315 justiciability of right to self-government 287 legal recognition 284–289 local government associations 298 metropolitan areas 282, 283, 309 Mexico 301 municipalities see municipalities national-subnational dualism 281, 285, 290, 314 national intervention 290 New Public Management 283 overburdening with delegated tasks 295–297 overregulation 287–289 privatisation of services 283 restrictions on powers 295–305 single-purpose bodies 281 social welfare 296, 302 South Africa 282, 283, 286, 287, 290, 298, 312–313, 314 Spain 286, 287, 288–289, 294, 299–300, 302, 313–314, 315 structure 305–314 subnational entities and 281, 290 supervisory authorities 295, 297–300 Swiss Confederation 281, 283, 285, 298, 300, 301, 306–307, 310 ultra vires principle 291–293 umbrella entities 312–314 underfunding 300–305 United Kingdom 310 United States 281, 285, 289, 291–292, 293, 296, 297, 301–302 urban-rural divide 282, 304, 305–307 local self-government supervision as correlative of right to 299 Lombard League 68 loyalty principle see federal loyalty principle M Macedonia ethno-cultural diversity 97–98 macro geo-economic regions 30 Madison, James 14, 71, 73–76, 89, 230, 425

Malaysia British colonial origins 45 constitutional adjudication 271 constitutional autonomy 137 monarch 193 revenue from natural resources 211 Sarawak 130 Senate 170–171 ‘marble cake’ federalism 46 marriage same-sex 324, 326–327, 344 Mercosur New Regionalism 25, 30 metropolitan areas see also municipalities Austria 309 Germany 309 Greater London Authority Act 56, 57 growth 282 local government 282, 283 migration 403 Spain 153, 309 status as subnational entities 309 transcending subnational boundaries 309 Mexico civil service appointments 198 constitutional adjudication 279 constitutional amendment procedure 180 direct democracy 137 distribution of powers 147 dual federal system 156 federal nature 45 indigenous local government 282 judicial appointments 195 legislative power 147 local government 301 North American Free Trade Agreement (NAFTA) 25, 30, 426 Oaxaca 282 residual powers 140 second chamber 50, 165, 169, 171 Micronesia, Federated states of 7, 45 migration see immigration and migrant integration Mill, John Stuart direct and indirect taxes 215 minority groups see ethno-cultural diversity Mirkine-Guetzévitch, Boris 7–8 mixed government theory 76 Moldova Gagauzia 59 monarchies appointment process 193 constitutional autonomy 137 Monnet, Jean 62, 93–94 Montenegro confederation with Serbia 37 Montesquieu 76 multicultural policies migrant integration 404, 409, 412, 421

Index 469 multilevel governance (MLG) EU regionalism 26–27 multinational federalism 6, 98–101, 102 municipalities Australia 296, 306 Austria 285, 287–288, 291, 294–295, 297, 306–308 Brazil 283, 286, 287, 290, 314 Canada 306, 315 competition between 226 core competences 296 Dillon’s rule 291–292, 293 financing 202–204 Germany 202, 287, 293–294, 296, 303–304, 312, 414, 415 and globalisation 296 horizontal cooperation 309–311 India 283, 295, 314 inter-municipal cooperation 309–311 South Africa 203, 283, 287, 290, 296, 304, 305, 306, 312–313, 314 Spain 286, 294, 299–300, 305, 309 Swiss Confederation 226, 283, 304, 306–307, 308 United States 291–292, 293, 302, 308 N Napoleon Bonaparte 70 nation-building in connection with social welfare 347, 352, 359, 372, 373, 375, 376, 380 national terminology 9 national interest flexibilisation and centralisation through implied powers 153–154 nationalism see also secessionism devolution and 56 regionalism and 26 Native American tribes 67 natural resources Canada 211, 242 disputes over 210–211 revenues from 210–211, 220–221 United States 384 naturalisation 404, 406–407, 412, 418, 420–421 Navarra financing arrangements 209 tax-raising powers 222 Nepal 43 Netherlands Dutch Republic of the Seven United Provinces 68–69, 70 federacy 60 New Judicial Federalism 323–324, 343, 345 New Public Management 283 New Zealand Cook Islands 60

Nigeria constitutional amendment procedure 180 constitutional review 275 credit controls 211 federal nature 40, 45 Federation Account 220–221 intergovernmental relations 255 judicial appointments 195 judicial system 162 natural resources 210, 220–221 second chamber 165 tax-raising powers 220–221 Nixon, Richard 351 North American Free Trade Agreement (NAFTA) generally 426, 427, 433, 443, 446 New Regionalism 25, 30 North German Confederation 80 Northern Ireland Belfast Agreement 32 cross-border cooperation 32 devolution 56, 57, 207–208 devolved legislation 399 environmental protection 396, 397, 399 financial arrangements 210, 372, 373 healthcare 374, 379 Northern Ireland Act 56, 271 Sewel convention 257–258, 373–374 social welfare 371–372, 379 nuclear energy/waste Germany 391, 393 Swiss Confederation 390 United Kingdom 396 O Obama, Barack 198–199, 354 Organisation for Economic Cooperation and Development (OECD) 436 Organisation international de la Francophonie (OIF) 429, 432–433, 436, 444 Ottoman Empire 85 P Pakistan delegated powers 148 federal nature 45 Papen, Franz von 80–81 paradiplomacy 423 parliamentary sovereignty Canada 261, 332 United Kingdom 56–57, 271 United States denial 74 parliamentary systems bicameralism see second chambers constitutional autonomy 137–138 participation see democracy; second chambers; subnational participation at national level partnership principle EU regional policy 24 Pascal, Blaise 1 Peace of Westphalia 80, 85–86, 96, 113

470 Index pension provision Canada 356, 357–358, 380 Spain 366, 367–368 United Kingdom 371 petition, right to in Swiss cantons 329 pluralism constitutional 96–97, 111, 318 descriptive 112 diversity and 112 environmental protection 382 European Union 111 federalism and 6, 83, 111–122 generally 17, 111, 449–450 interest groups 116 legal 111 majoritarian democracies 112–113 migration 116 NGOs 116 normative 112 participatory democracy and 83, 111–122 suffrage and 116 territorial 34 use of term 111–112 policy formulation and implementation subnational entities 317, 318 ‘political constitutionalism’ in the United Kingdom 110 political legitimation in connection with social welfare 347 political rights see fundamental rights political scrutiny of subnational constitutions 263–266 pollution see also environmental protection generally 382 German regulation 391 greenhouse gas emissions 387, 392 metropolitan areas 282 Swiss regulation 388–389 US regulation 384–385 water quality regulation 384–385, 387 populism 114–115 Portugal 59 powers administrative see administrative powers autonomy see autonomy of subnational entities complex power-sharing 104 distribution see distribution of powers judicial see courts and judiciary legislative see legislative power residual 139–146 separation of 73 taxation see taxation pre-emption supremacy clauses 131–132 presidential systems appointment of heads of state 193–194, 198 constitutional autonomy 137 institutions of intergovernmental relations 255

principle of comity Belgium 249 ‘private federalism’ 351 privatisation healthcare and social welfare 368 local government services 283 New Public Management 283 Proudhon, Pierre-Joseph 13–14, 16 public services see social welfare Puerto Rico 59–60 Pufendorf, Samuel von 80, 85, 86 Q Quebec Charter of the French Language 331, 333 Charter of Rights and Freedoms 333, 344 conservative resistance to progressive agenda 357, 381 constitutional amendment procedure 180–181 cultural promotion abroad 432, 433, 444 cultural relations with France 444 external relations 432–434, 443–444 financing arrangements 208, 209, 214–215 General Agreement on Tariffs and Trade (GATT) 434 generally 356 Gérin-Lajoie doctrine 433–434, 445 healthcare 360–361, 380 immigration and migrant integration 407–411, 419, 420, 422 indigenous people 102 interculturalism 409 international treaty negotiation and implementation 433–434, 445 language rights 331, 342, 344, 432–433 local government amalgamations 308 nation-building 380 National Assembly 9 Organisation international de la Francophonie (OIF) 432–433, 444 pension provision 357–358, 380 Quiet Revolution 357, 360 secessionism 108–109 social welfare 357–359, 380 sovereignty-association 108–109 Supreme Court seats 196 tax-raising powers 214–215, 217, 222 UNESCO 434 utilisation of autonomy 344 R race to the bottom regarding social welfare and healthcare 348, 353, 359, 370, 380–381 Reagan, Ronald 351 redistribution in connection with social welfare 347, 376 region use of term 21–22 regional consociations 103 regional states definition 50–51

Index 471 distribution of power 55 federalism and 65 France 55 generally 449 historical background 51–52 Italy 23, 53–54 Serbia 55 Spain 21, 26, 52–53 subnational units 50–51 use of term 8, 51–52, 65 regionalism bottom-up 22–23 Council of Europe 23 cross-border cooperation 31–33 Europe 14, 21, 23 European Union 14, 23–33 generally 20–22, 52 historical background 22, 52 macro geo-economic regions 30 multilevel governance (MLG) 26–27 nationalism and 26 New Regionalism 25–29, 30, 347 regionalisation compared 23–25 top-down 22–23 use of term 30–33 religious diversity 15 religious freedom 334 revenue powers see taxation Rhodesia and Nyasaland 43n risk-pooling equalisation 240 Roosevelt, Franklin D 349 Rossi, Ernesto 61 Rousseau, Jean-Jacques 115 Rowell-Sirois Commission 217 Russian Federation constitutional amendment procedure 180 constitutional autonomy 126, 130 Council of the Federation 165, 176, 200 federal nature 45 federal and provincial legislation 133 impeachment powers 200 judicial system 160 natural resources 210 residual clause 139 sovereignty 128 subnational constitutions 127–128, 129–130 subnational legislatures 137 subnational units 129–130 S Saint Kitts and Nevis micro-federation 45 National Assembly 171 secession clause 108 Sardinia 51 Schmitt, Carl 80 Schuman Declaration 62, 92 Scientific Management Movement 319 Scotland Barnett formula 207–208, 210, 372, 379

cross-border public authorities 398, 402 devolution 23, 56, 57 devolved legislation 262, 269, 278, 399 Edinburgh Agreement 110 environmental protection 396, 397–398, 399, 401, 402 executive devolution 398 financial arrangements 207–208, 210, 372–373 generally 26 healthcare 374, 375, 379, 380 judicial system 160 nation-building 380 nationalism and regionalism 26 natural resources 211 Scotland Act 56, 271, 371 Scottish Environmental Protection Agency 397 secessionist movement 26, 101, 110–111, 115, 211, 371 Sewel convention 257–258, 373–374 social welfare 371, 379, 380 tax-raising powers 210 ‘welfare nationalism’ 373 secessionism see also nationalism Australia 106–107 Canada 101, 108–109 constitutional prohibition 105 Eritrean secession 107–108 multinational federalism 6, 98–101 natural resources and 211 Scotland 26, 101, 110–111, 115, 211, 371 Spain 101, 110, 141, 245 under federal constitutions 102, 105–111 United Kingdom 101, 110–111, 115, 211, 371 second chambers appointments to 169–171, 193–199 composition 171–172 constitutional amendment procedures 183–185 council (ambassadorial) model 165, 167–169, 172–175 declining role 165 direct elections to 165, 169, 170 disputes with first chambers 192–193 equal representation of subnational units 167, 169 ethno-culturally diverse countries 169–170 functions performed by 178–200 generally 137, 164–166 German Bundesrat 42, 82, 165, 167–168, 172–175, 177, 183 hybrid 165, 176–177 indirect elections to 170–171 judicial appointments 193–198 legislative and administrative functions 190–193 money or tax bills 191

472 Index origins and constitutional entrenchment 167–168 purpose 166 senate model 165, 168–172 territorial reforms 186–190 US Senate 42, 49, 75–76, 165, 166, 167, 168, 169, 179–180, 190–191, 194–195 self-rule and shared-rule as feature of federalism 3–4, 8, 10, 18, 39, 48, 66, 125–126, 448 senates see second chambers Senegambia 43n Serbia confederation with Montenegro 37 Vojvodina 55, 59 Seydel, Max von 87 Sieyès, Emmanuel-Joseph 72 social citizenship national 347, 371, 377 social citizenship concept 347, 371 social organisation federalism as principle of 13–14, 15–16 social welfare Belgium 361–365, 376, 377–378, 380 Beveridgean regimes 346, 370–371, 374 Bismarckian regimes 346, 361 Canada 346, 355–359, 376, 378–379, 380, 381 centralised provision 376–377 corporatist regimes 346 declinist debate 348 ‘direct federalism’ 351 federalism and 347–348 financing 348, 378–379 Flanders 362–363 functions 347 generally 317–318, 346–348, 376–381 Germany 302 healthcare see healthcare horizontal variation 379–380 immigration and migrant integration 404, 405, 410, 414, 416–417, 419–420 liberal Anglo-Saxon regimes 346 local government 296 nation-building function 347, 352, 359, 372, 373, 375, 376, 380 national framework 348 New Regionalism 347 political function 347 ‘private federalism’ 351 privatisation 368 Quebec 357–359, 380 race to the bottom 348, 353, 359, 370, 380–381 redistribution as function 347, 376 selective access 368 shared-cost programmes 358, 379 social assistance 377–378 social-democratic regimes 346 social security 377–378 Spain 258, 365–369

United Kingdom 370–374, 377 United States 349–353, 376, 377, 380 ‘welfare magnets’ 348, 356 ‘welfare-to-work’/‘workfare’ 351, 352, 378 South Africa constitutional adjudication 268–269 constitutional amendment procedure 182, 184 constitutional autonomy 128, 129, 137 constitutional review 277 cooperative government 147, 148, 298 coordination of legislative action 253–254 devolutionary federalism 45 dispute resolution 252, 269 distribution of powers 147 Financial Management of Parliament Act 253–254 financial relations 203, 207 framework legislation 147 homogeneity clause 136 intergovernmental relations 248, 249, 256 judicial appointments 195 judicial system 160 legislative procedures 191 local government 282, 283, 286, 287, 290, 298, 312–313, 314 Municipal Finance Management Act 304 municipalities 203, 283, 287, 290, 296, 304, 305, 306, 312–313, 314 National Council of Provinces 165, 176, 184, 191, 192, 256 National Revenue Fund 220 provinces generally 9 provincial budgeting 207, 220 provincial constitutions, scrutiny 264–265 residual powers 140 sovereignty 92 Speakers’ Forum 253–254 status of provinces 129, 264–265 tax-raising powers 203, 220, 222, 223, 225 territorial changes 188 traditional leaders 282 sovereignty Bodin, Jean 84, 85, 86, 88 civil law countries 87 common law countries 87–88 constitutional autonomy 130 divided 88–97 dual federal system 46 European Union 42, 92–97 federal systems 41–42, 48, 83–97 ‘floating’ 84 France 89, 95 Germany 91 Hobbes, Thomas 84–85, 86 Indivisible 83, 84–97 Keynesian–Westphalian states 113 late 96–97 legal and political 84 loss of, compensation for 164 parliamentary see parliamentary sovereignty Peace of Westphalia 80, 85–86, 96

Index 473 popular 182 Russia 128 secessionism see secessionism South Africa 92 Spain 92 subnational constitutions as remnants of 127–128 Swiss Confederation 91 uniform 86 United States 87–91 sovereignty-association 108–109 Spain see also Basque Country; Catalonia administrative federalism 156, 157 asymmetries 209, 226 autonomous communities 9, 23, 51, 52–53, 126, 153, 206–207, 209, 219–220, 222, 223–224, 225, 309, 338–339, 366–367, 377 ceded taxes 219–220 Ceuta 153 comarcas 313 conferences of ministers 157 Constitution 52, 92, 338–339, 366 constitutional amendment procedure 185 constitutional autonomy 128 ‘constitutional block’ (bloque de constitucionalidad) 338–339 Constitutional Court 92, 197–198, 275, 343, 439, 440 constitutional review 274, 275 cooperation principle 148, 251 cross-border cooperation 440, 441 decentralisation 52–53 devolutionary federalism 45, 46 ‘differential facts’ (hechos differenciales) 138–139 distribution of powers 147, 318, 440 equalisation 243, 244, 245, 379 ethno-cultural diversity 52 European Fiscal Compact (TSCG) 236–237, 302 European Union 23, 440, 442 external relations 439–442, 443, 445–446, 447 federal nature 7, 8, 43, 45, 50 federal spending power 232, 234–235 financial autonomy 206–207, 234–235 financial relations 206–207, 209, 223–224, 225, 226, 370, 379 First Republic 14 foreign relations see external relations above fundamental rights 338–341, 343, 345 healthcare 369–370, 379 immigration 403 institutional representation of the subnational units 50 intergovernmental acuerdos and convenios 260 intergovernmental coordination 367 intergovernmental relations 248, 258, 259, 260

internal stability pact 236 international treaty negotiation and implementation 439, 441–442, 445–446, 447 judicial appointments 197–198 judicial system 160 local government 286, 287, 288–289, 294, 299–300, 302, 313–314, 315 Organic Law on the Financing of the Autonomous Communities (LOFCA) 206–207, 209, 223–224, 243, 244, 368–369, 379 Organic Law on the Harmonisation of the Autonomy Process (LOAPA) 53 Maastricht criteria 236 Madrid 153 mancomunidades 313 Melilla 153 metropolitan areas 153, 309 minimum income schemes 368, 381 municipalities 286, 294, 299–300, 305, 309 ‘national interest’ principle 153 participatory democracy 137 principle of loyalty to the Constitution (fidelidad a la constitución) 251 principios rectores 341 provinces 313–314 as regional state 21, 26, 52–53 residual powers 140, 141 secessionist movements 101, 110, 141, 245 second chamber 175, 185 Second Republic 21, 52 Senate 170 social welfare 365–369, 377, 379 sovereignty 92 statutes of autonomous communities 126, 138–139, 141, 338–341, 343, 345, 440, 442 subnational legislation, scrutiny 265, 266 tax competition 225, 226 tax-raising powers 207, 214, 218, 219–220, 222, 223–224, 225, 226 spending power Australia 231–232 Belgium 233 Canada 230–231, 232–233, 234, 235 coercive federalism 233–234 conditional grants 228–229, 233–234 debt ceilings and financial autonomy 236 extra-jurisdictional spending 229–230 fiscal discipline 235–239 fiscal equivalence 228 Germany 228, 229, 237 Italy 237 legal basis 229–232 limits 232–235 national minimum standards 229 Spain 232, 234–235 Swiss Confederation 228 United States 230, 233–234 Spinelli, Altiero 61, 93 Stewart, Justice Potter 3

474 Index subnational entities see also subnational participation at national level areas transcending subnational boundaries 309 ‘armour-plated’ competences 138, 141 autonomy see autonomy of subnational entities bicameral legislatures 137 constitutions, scrutiny 263–266 credit controls 211 decentralisation theory 314 disputes see conflict prevention and resolution draft legislation, scrutiny 261–263 external relations see external relations financial relations generally 201 fundamental rights 321–345 international relations see external relations limits of autonomy 321 limits of constitutional space 321 limits on federal powers 272–274 local government and 281 metropolitan areas 309 nation-building 347 national constitutions limiting rights of 321 parliamentary systems 137–138 policy formulation and implementation 317, 318 regional states 50–51 revenue powers 201 sovereignty 130, 164 spending powers 201, 228–239 taxing powers see taxation terminology 9 unicameral legislatures 137 utilisation of constitutional space 321–322 vertical fiscal gap 201 ‘welfare magnets’ 348, 356 subnational participation at national level administrative functions 178, 193 aggregative federal states 164–165 appointments processes 178, 193–199 constitutional amendment 178–185 emergency powers 178, 200 generally 48, 49–50, 164–165, 448, 449, 451 graduated 180–182, 183–184 impeachment powers 178, 200 importance 166 indirect 179, 182, 183–185 institutional forms 164–178 intergovernmental bodies 164, 177–178 legislative functions 178, 190–193 procedural forms 164, 178–200 ratification 179 second chambers see second chambers territorial changes 178, 186–190 subsidiarity European Union 19–20, 28, 63 federalism and 19–20 suffrage see also democracy pluralism and 116

Swiss Confederation 122 United States 71, 135–136, 325 supremacy clauses federal states without 132–133 generally 19, 130–133, 321 Germany 132, 159, 334, 335, 341 pre-emption 131–132 prohibition of elusion 131 Swiss Confederation 70, 328, 329, 341 United States 75, 91, 131, 159, 322, 341, 427, 447 Swiss Confederation administrative autonomy 159 administrative federalism 79, 156, 447 aggregative federalism 45 balanced-budget rules 238 cantonal constitutions 328–330, 342–343, 344–345 cantonal constitutions, scrutiny 263–264 cantons 9, 226, 444–445 competition, focus on 217–218, 225, 226–227, 242 conditional transfers 301 Conference of Cantonal Governments 255, 431 Constitution (1848) 76–79 Constitution (1874) 77, 90 Constitution (1999) 35, 79, 328, 342, 345, 428 constitutional adjudication 271, 275, 279, 343 constitutional amendment procedure 182–183 constitutional autonomy 128 cooperative federalism 148 Council of States (Ständerat) 42, 77, 165, 169, 172, 191, 192 cross-border cooperation 226–227, 428–429 decentralisation 255 direct democracy 77–79, 118, 137, 182–183, 307, 344 dispute resolution 141, 252, 269 dual federal system 46 environmental protection 387–390, 399, 400–401 equalisation 241, 242 ethno-cultural diversity 79, 300 EU-Swiss bilateral agreements 78, 429 external relations 428–431, 443, 444–445, 447 Federal Assembly (Bundesversammlung) 200, 263, 264, 271, 430 Federal Court (Bundesgericht) 343 federal executive 194 federal loyalty principle 249, 264 federal nature 7, 16–17, 35–36, 43, 45, 47, 68 federation, formation 76–79 financial constitution 205 financial relations 205 fiscal competition 217–218, 225, 226–227 fiscal discipline 238–239

Index 475 foreign relations see external relations above fundamental rights 327–330, 341, 343, 344–345 general competence principle 294 generally 6, 449 Helvetic Republic 70 homogeneity clause 40 horizontal cooperation 255 immigration and migrant integration 404–407, 418, 420, 421 inter-cantonal legislation 260–261 intergovernmental agreements 258, 260–261 intergovernmental collaboration 248, 249, 255 international treaty negotiation and implementation 430–431, 444–445, 446, 447 judicial system 160, 162 Landsgemeinde 77–78, 118, 137, 307 languages 79, 429 legislative power 146, 157, 159 local government 281, 283, 285, 298, 300, 301, 306–307, 310 municipalities 226, 283, 304, 306–307, 308 naturalisation 406–407, 420 nuclear energy/waste 390 Old Swiss Confederacy 68, 69–70 Organisation international de la Francophonie (OIF) 429 popular initiatives 137, 183 popular sovereignty 182 ‘rainy day funds’ 238 referendums 77–78, 137, 182–183, 189–190, 388, 390, 400, 406, 431, 445 residual clause 139, 141 residual powers 141 right to petition 329 same-sex couples 330 sovereignty 91 spending and revenue powers 201, 205, 217–218, 221, 222, 225, 226–227, 228 suffrage 122 Supremacy Clause 77, 328, 329, 341 termed confederation 7, 9, 35–36 territorial changes 189–190 unwritten rights 327–328, 342, 345 special-purpose associations (Zweckverbände) 310 symmetry see also asymmetry symmetric federalism generally 9, 47 T Tanzania Zanzibar 59 taxation see also financial relations administrative efficiency 222, 225 allocation of taxing powers 212 Australia 223 Austria 218–219

autonomous 218–222 Belgium 213–214, 223 Bosnia and Herzegovina 222, 223 Brazil 221 Canada 212, 214–215, 221–222, 223, 242–243 ceded taxes 219–220 centralisation, trend towards 222–228, 240 concurrent powers 212, 214–218 costs of collection 222 customs tariffs 205, 212 direct and indirect taxes 211, 215 distribution of shared revenue 218–222 double taxation, prohibition 213–214, 227 environmental taxes 382 equalisation see equalisation ethno-culturally diverse countries 222 exclusive powers 212–214 financial autonomy and its limits 224–225 financing social welfare and healthcare 376 Germany 212, 218, 219 horizontal competition 225–228 India 221 indirect 211 Italy 223 local government 303–304 local taxation 203, 207 money or tax bills 191 natural resources 210–211, 220–221 Nigeria 220–221 Scotland 210 shared powers 212, 218–222 South Africa 203, 220, 222–223, 225 Spain 207, 214, 218, 219–220, 222, 223–224, 226 subnational entities generally 211–222 Swiss Confederation 217–218, 221, 225, 226–227 United States 212, 214, 215–217, 221, 222, 227–228 vertical competition 215, 217–218 Taylor of Caroline, John 89 territorial changes subnational participation 178, 186–190 Tocqueville, Alexis de 89 trading blocs New Regionalism 25, 30 transboundary issues see also European Union Belgium 437 Canada 433 cross-border cooperation 31–33, 60–61, 426 environmental protection 383, 387, 398, 401, 426, 443 infrastructure 31, 226, 426, 429, 443 metropolitan areas 309 New Regionalism 25, 30 regionalism 30–33 Scotland Act 398 Spain 440, 441 Swiss Confederation 226–227, 428–429

476 Index taxation 226–227 United States 31, 426 Western Climate Initiative 255, 387, 402, 426 transparency intergovernmental relations 249, 256 treaties see international treaty negotiation and implementation tri-level intergovernmental mechanisms 298–299, 314–315 Trinidad and Tobago 7 Trudeau, Pierre 180, 231 U ultra vires principle local government 291–293 under-utilisation thesis 343 unemployment insurance Canada 356 United Kingdom 371 United States 353, 377, 380 Union of Soviet Socialist Republics (USSR) creation 43 ethno-federation 45, 98, 100 federal status 7 Union of Utrecht 68–69 unions generally 34 India 38–39, 106–107 United Kingdom see also Northern Ireland; Scotland; Wales administrative devolution 371–372 asymmetries 210 Balfour Declaration 431–432 Barnett formula 207–208, 210, 372, 379 Beveridgean social welfare regime 370–371, 374 Children’s Commissioners 373 constitutional statutes 56–57 cross-border public authorities 398, 402 delegated powers 148, 257–258, 377 devolution and delegated powers 148, 257–258 devolution and the welfare state 371–375, 377 devolved entities 6, 23, 55–58, 207–208, 210, 252, 257–258, 396, 399 devolved legislation 262, 269, 278 dispute resolution 252–253, 269 distribution of powers 147 energy matters 396, 399 ‘English question’ 56 Environment Agency 397 environmental protection 396–398, 399, 401, 402 EU membership/‘Brexit’ 57, 96, 110–111, 204, 397, 399–400 executive devolution 398 federal nature 43 financial relations 207–208, 210, 379 fundamental rights 330

Greater London Authority Act 56, 57 health trusts 375 healthcare 374–376, 377, 379 House of Lords 169 intergovernmental cooperation 378 intergovernmental coordination 374 Joint Ministerial Committee 252, 374, 375 judicial review 277–278 judicial system 160 local government 310 local government privatisation 283 multinational federalism 101 National Health Service 374–375 National Insurance 371 natural resources 211 nuclear energy 396 parliamentary sovereignty 56–57, 271 ‘political constitutionalism’ 110 regionalism 23, 26 secessionist movements 101, 110–111, 115, 211, 371 Sewel convention 257–258, 373–374 social welfare 370–374, 377, 379 Statute of Westminster 43 Welfare Reform Act 377 United Nations Conference on the Human Environment (‘Stockholm Conference’) 382 UNESCO 434, 436, 444 United States administrative agreements 260 administrative compacts 260 admission of new states to Union 322, 342 Advice and Consent Clause 194–195, 198 aggregative federalism 45, 126–127 American Civil War 44, 89, 106, 169, 216 American Recovery and Reinvestment Act 238 Anti-Federalists 89 appointments process 198–199 Articles of Confederation 36, 68, 70–72, 74, 88, 425 Bill of Rights 73, 322, 323, 325, 341–342, 344 Case or Controversy Clause 274, 277 centralisation 425 civil rights legislation 143, 352 Clean Air Act 385, 400 Clean Water Act 387 climate change policy 385, 386, 387, 400, 426 coercive federalism 233–234 Commerce Clause 142–143, 144–145, 215, 227–228, 384, 398–399 competition 308, 353, 380 conditional transfers 301, 350 confederation 36–37, 68, 70–72 Connecticut Compromise 76, 167 conservative resistance to progressive agenda 319–320, 323–324, 326, 345, 381 Constitution 72–76, 88, 89, 322, 350

Index 477 constitutional adjudication 267–268, 269, 270, 272, 278–279 constitutional amendment procedure 74, 179–180 constitutional anti-secessionism 106, 107 constitutional autonomy 126, 128 Constitutional Convention 72–74 constitutional review 272–274, 278–279 Continental Congress 71 court system 322–323, 324–325, 343 credit controls 211 cross-border cooperation 31, 426 Cruel and/or Unusual Punishment Clauses 323, 324 customs tariffs 205, 212 death penalty 323–324, 325–326 Declaration of Independence 70–71 Dillon’s rule 291–292, 293 direct democracy 118, 137, 344 ‘direct federalism’ 351 distribution of powers 142–143, 146–151, 158–159, 425 Dorr Rebellion 135–136 dual federalism doctrine 46, 151, 156, 158–159, 216 Due Process Clause 322, 325, 326 economic development 296 education 289 Elastic Clause see Necessary and Proper Clause below elusion, prohibition 131 Endangered Species Act 385 environmental protection 382, 383–387, 398–399, 400, 401, 426 Equal Protection Clause 325, 327 equalisation 167, 169, 242, 379 external relations 425–428, 443, 445, 446, 447 federal common law 161–162 federal nature 7, 9, 43, 44, 45, 68 federal spending power 230, 233–234 Federalists 9, 88 The Federalist 14, 73–74, 73n federation, formation 72–76 financial relations 202, 205–206, 208, 212, 378, 379 fiscal competition 215–217, 225–226 fiscal discipline 238–239 Food Stamps Program 352 foreign economic policy 425, 426 foreign investment in 426 free speech 324–325 freedom of movement 351, 379 fundamental rights 322–327, 341, 343 General Assistance programmes 352 General Revenue Sharing 351 General Welfare Clause 230, 233–234 generally 5–6, 449 Governmental Affairs Committee 200 Great Society programmes 314, 349, 350, 352, 353, 377

Guarantee Clause 40, 135, 263, 266 healthcare 350, 353–355, 380, 381 homogeneity principle 134, 135–136 homosexual conduct 324, 325, 326–327 horizontal compacts 259 horizontal divergence 400 horizontal tax issues 227–228 House of representatives 169 immigration 403, 425–426 Implied Powers Theory 150–151 improperly seized evidence 323 incorporation of the Bill of Rights 322 intercircuit conflicts 269 intergovernmental agreements 259, 260 intergovernmental relations 248, 249, 252, 254, 255 internal consistency test 227–228 international treaty negotiation and implementation 199–200, 425–428, 445, 447 Jim Crow Laws 320 judicial appointments 194–195 judicial double protection of rights 324–325 Judicial Procedures Reform Bill (‘Court Packing Plan’) 349 judicial review 74–75 judicial system 160–162 legislative procedures 190–191 limits on federal powers 272–274 local government 281, 285, 289, 291–292, 293, 296, 297, 301–302, 308 local government financing 202, 206 Medicaid 242, 353–355, 379 Medicare 353, 380 ‘midnight judges’ 74–75 Migratory Bird Treaty Act 427–428 municipalities 291–292, 293, 302, 308 National Conference of State Legislatures 254 National Environmental Policy Act 382, 385, 386 natural resources 384 Necessary and Proper Clause 149–151 New Deal 349, 352, 353, 377 New Judicial Federalism 323–324, 343, 345 North American Free Trade Agreement (NAFTA) 25, 30, 426, 427, 433, 443, 446 originalism 272 participatory democracy 117 Patient Protection and Affordable Care Act (PPACA) 354–355, 381 Pennsylvanian Constitution (1776) 71–72, 74 Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) 351, 352 Philadelphia Convention 14, 88, 89, 167, 212, 322 political question doctrine 135–136 pre-emption 131–132 press freedom 325 ‘private federalism’ 351 Puerto Rico 59–60

478 Index race to the bottom regarding social welfare 353, 385 racial segregation 320 ‘rainy day funds’ 238 Recess Appointments Clause 198–199 referendums 137 reinforcement of federal power 89–90 Republican Clause see Guarantee Clause above residual clause 139, 140 same-sex couples 324, 325–326 Senate 42, 49, 75–76, 165, 166, 167, 168, 169, 179–180, 190–191, 192, 194–195, 199–200 senatorial courtesy 195 separation of powers 73 Social Security programme 349, 350, 351–352 social welfare 349–353, 376, 377, 378, 380 sovereign immunity 90–91 sovereignty 87–91 State Children’s Health Insurance Program (SCHIP) 353–354 state constitutions 71–72, 128, 263, 265–266, 270, 325, 343, 345 state court judgments 325 state debt crisis 238, 239 state legislatures 137, 254 states’ autonomy 322–327, 343 subnational participation at national level 165, 167, 179–180 suffrage 71, 135–136, 325 supervision of executive 200 Supremacy Clause 75, 91, 131, 159, 322, 341, 427, 447 Supreme Court 349–350, 353, 354, 377, 398 taxation 212, 214, 215–217, 221, 222, 227–228 territorial changes 187 unemployment insurance 353, 377, 380 Unfunded Mandates Reform Act (UMRA) 301, 302 urbanisation 285 Virginia Declaration of Rights 322 waivers 351 ‘War on Poverty’ 349, 351 Warren Court 323–324, 342, 350 water quality regulation 384–385, 387 welfare reform as ‘devolution revolution’ 351 ‘welfare-to-work’/‘workfare’ 351, 352, 378 Western Climate Initiative 255, 387, 402, 426 urbanisation see also metropolitan areas; municipalities Canada 315

United States 285 urban-rural divide 282, 283, 304, 305–307 V Venezuela federal nature 7, 45 second chamber 50 Ventotene manifesto 61–62 vertical relations equalisation 240–241, 304–305 fiscal gap 201, 240 generally 2 intergovernmental agreements 258 presidential systems 255 tax competition 217–218 vertical cooperation 253–254, 255–256 Viennese School of Legal Positivism decentralisation theory 314 W Wales Acts of Assembly 372 devolution 23, 56, 57, 207–208 devolved legislation 262, 269, 372 environmental protection 396, 397, 399 financial arrangements 210, 372, 373 Government of Wales Act 56, 271, 374 healthcare 374, 375, 379 Sewel convention 257–258, 373–374 social welfare 372, 379 ‘welfare nationalism’ 373 Wallonia external relations 437 generally 241 immigration and migrant integration 416, 422 social welfare 362–363, 364–365 Walloon Assembly 170 Warren, Chief Justice Earl 323–324, 342 Washington, George 76 water protection of water resources 391 quality regulation in US 384–385, 387 ‘welfare nationalism’ 373 welfare state see also healthcare; social welfare intergovernmental regulation 247 West Indies 43n Western Climate Initiative 255, 387, 402, 426 Y Yeltsin, Boris 133 Yugoslavia 45, 98, 100