The Foundations and Traditions of Constitutional Amendment 9781509908257, 9781509908288, 9781509908271

Citation preview

THE FOUNDATIONS AND TRADITIONS OF CONSTITUTIONAL AMENDMENT There is growing interest in constitutional amendment from a comparative ­perspective. Comparative constitutional amendment is the study of how constitutions change through formal and informal means, including alteration, revision, ­evolution, interpretation, replacement and revolution. The field invites scholars to draw insights about constitutional change across borders and cultures, to uncover the motivations behind constitutional change, to theorise best practices, and to ­identify the theoretical underpinnings of constitutional change. This volume is designed to guide the emergence of comparative constitutional amendment as a distinct field of study in public law. Much of the recent scholarship in the field has been written by the scholars assembled in this volume. This book, like the field it hopes to shape, is not comparative alone; it is also doctrinal, historical and theoretical, and therefore offers a multiplicity of perspectives on a subject about which much remains to be written. This book aspires to be the first to cover comprehensively the new dimensions of the study of constitutional amendment, and will become a reference point for all ­scholars working on the subject. The volume covers all the topics where ­innovative work is being done, such as the notion of the people, the trend of empirical ­quantitative approaches to constitutional change, unamendability, sunrise clauses, constitutional referenda, the conventional divide between ­constituent and constituted powers, among other important subjects. It creates a dialogue that cuts through these innovative ­conceptualisations and highlights scholarly disagreement and, in so doing, puts ideas to the test. The volume therefore captures the fierce ongoing debates on the relevant topics, it reveals the current trends and contested issues, and it offers a variety of a­ rguments elaborated by prominent experts in the field. It will open the way for further dialogue. Volume 17 in the series Hart Studies in Comparative Public Law

Hart Studies in Comparative Public Law Recent titles in this series: Constitutionalising Secession David Haljan Parliaments and Human Rights Redressing the Democratic Deficit Edited by Murray Hunt, Hayley Hooper and Paul Yowell 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

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 © Richard Albert, Xenophon Contiades and Alkmene Fotiadou 2017 Richard Albert, Xenophon Contiades and Alkmene Fotiadou 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: 978-1-50990-825-7 ePDF: 978-1-50990-827-1 ePub: 978-1-50990-826-4 Library of Congress Cataloging-in-Publication Data Names: Albert, Richard D., editor.  |  Contiades, Xenophon I., editor.  |  Fotiadou, Alkmene, editor. Title: The foundations and traditions of constitutional amendment / edited by Richard Albert, Xenophon Contiades, and Alkmene Fotiadou. Description: Oxford [UK] ; Portland, Oregon : Hart Publishing, 2017.  |  Series: Hart studies in comparative public law ; v. 17  |  Includes bibliographical references and index. Identifiers: LCCN 2017002144 (print)  |  LCCN 2017003153 (ebook)  |  ISBN 9781509908257 (hardback : alk. paper)  |  ISBN 9781509908264 (Epub) Subjects: LCSH: Constitutional amendments.  |  Constitutional law.  |  Law reform. Classification: LCC K3168 .F68 2017 (print)  |  LCC K3168 (ebook)  |  DDC 342.03—dc23 LC record available at https://lccn.loc.gov/2017002144 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.

Acknowledgements The seeds for this volume were planted in July 2014, as the three of us were exchanging emails about our shared scholarly interest in constitutional amendment. We had been longing for a forum to workshop our papers with colleagues in the field, and so we decided to convene a roundtable on comparative constitutional amendment in Chestnut Hill at Boston College. Our Workshop on Comparative Constitutional Amendment was held on 15 May 2015. We spent the entire day discussing and critiquing papers prepared specifically for the Workshop, and we continued our conversation over a dinner that stretched well into the night. We were hosted with great warmth and generosity by the Clough Center for the Study of Constitutional Democracy, directed by Professor Vlad Perju of Boston College Law School. We also benefited from our partnership with the Research Group on Constitution-Making and Constitutional Change in the International Association of Constitutional Law. We thank them for supporting our Workshop. We thank our friends at Hart Publishing—Bill Asquith, Francesca Sancarlo, Anne Flegel and Claire Banyard—for the opportunity to share the fruits of our Workshop in this published form. We reserve our biggest thanks for the contributors to this volume. We are inspired by their ideas, we are motivated by their hard work, and we are grateful for their friendship. We look forward to continuing the conversation we have begun in this volume, both with them and with our readers. The Editors

vi 

Contents Acknowledgements������������������������������������������������������������������������������������������������ v Notes on Contributors����������������������������������������������������������������������������������������� xv Introduction: The State of the Art in Constitutional Amendment��������������������������� 1 Richard Albert I. Defining the Field������������������������������������������������������������������������������������ 3 II. The Architecture of Constitutional Amendment Rules���������������������������� 4 A. Unamendability������������������������������������������������������������������������������� 7 B. The Operation of Constitutional Amendment���������������������������������� 8 C. Three Challenges in the Study of Constitutional Change��������������� 11 III. Amendment as Constitution������������������������������������������������������������������ 12 A. Perspectives on Constitutional Amendment����������������������������������� 14 B. The Future of Constitutional Amendment������������������������������������� 19 Part I: The Foundations of Constitutional Amendment 1. Amendment Power, Constituent Power, and Popular Sovereignty: Linking Unamendability and Amendment Procedures����������������������������������� 23 Yaniv Roznai I. Introduction������������������������������������������������������������������������������������������ 24 II. Unamendability and Constituent Power������������������������������������������������ 24 A. A Three-Track Democracy in a Nutshell���������������������������������������� 24 B. Primary Constituent Power and Popular Sovereignty��������������������� 26 III. The Constitutionalisation of Primary Constituent Power��������������������������������������������������������������������������������� 31 A. The Fallacy of Prescribed Constitution-Making Procedures����������� 31 B. We The ‘Limited’ People?��������������������������������������������������������������� 33 IV. The Spectrum of Constitutional Amendment Powers���������������������������� 37 A. Demanding and Facile Amendment Powers����������������������������������� 37 B. Linking Amendment Procedure and Unamendability��������������������� 41 C. The Spectrum of Amendment Powers and Judicial Review of Amendments����������������������������������������������������������������� 46 V. Conclusion�������������������������������������������������������������������������������������������� 48 2. Constitutional Theory and Cognitive Estrangement: Beyond Revolutions, Amendments and Constitutional Moments������������������������������� 51 Zoran Oklopcic I. Introduction: The Person of ‘The People’ and A Three-Fold Cognitive Estrangement��������������������������������������������������� 51 II. Beyond ‘The People’: New Tropes, Old Anxieties���������������������������������� 54

viii  Contents III. Three Forms of Estrangement-prevention: Holmes, Pettit, Dworkin���������������������������������������������������������������������� 56 IV. Tertium Datur: Mapping Constitutional Change Between the Revolution and the Amendment��������������������������������������� 59 V. Towards a Different Familiarity: ‘The People’, The Paradox and The Sacrifice������������������������������������������������������������� 69 3. Constraints on Constitutional Amendment Powers��������������������������������������� 73 Oran Doyle I. Introduction����������������������������������������������������������������������������������������� 73 II. A Doctrine of Unconstitutional Constitutional Amendments��������������� 74 A. Unconstitutional Amendments: Positive, Moral and Conceptual Claims������������������������������������������������������ 74 B. Morally Unconstitutional Constitutional Amendments���������������� 75 C. Conceptually Unconstitutional Constitutional Amendments�������������������������������������������������������������������������������� 77 D. A Problematic Rubric������������������������������������������������������������������� 80 III. Constraint and Powers of Constitutional Change�������������������������������� 81 IV. The Types of Constraint on Constitutional Amendment Powers����������������������������������������������������������������������������� 83 A. Process or Content����������������������������������������������������������������������� 83 B. Rule or Standard�������������������������������������������������������������������������� 84 C. Legislator or Court����������������������������������������������������������������������� 84 D. Values Served by Constraint: Foundational, Majoritarian, or Counter-Majoritarian����������������������������������������� 86 V. Distribution of Power and the Justification of Constraint�������������������� 87 VI. Justification of Constraints on Constitutional Amendment Powers����������������������������������������������������������������������������� 89 A. Parameters of Justification������������������������������������������������������������ 89 B. Illustrative Schema of Constraints������������������������������������������������ 89 C. Contextual Factors����������������������������������������������������������������������� 91 D. Majoritarian Constraints�������������������������������������������������������������� 92 E. Foundational Constraints������������������������������������������������������������� 94 F. Counter-Majoritarian Constraints������������������������������������������������ 94 VII. Conclusion������������������������������������������������������������������������������������������� 95 4. Comment on Doyle’s Constraints on Constitutional Amendment Powers��������������������������������������������������������������������������������������� 97 Mark Tushnet 5. Constituting the Amendment Power: A Framework for Comparative Amendment Law�������������������������������������������������������������� 105 Thomaz Pereira I. Introduction��������������������������������������������������������������������������������������� 105 A. Sieyès’ Two Different Problems�������������������������������������������������� 107 i. Constituting the Constituent Power������������������������������������ 108

Contents ix ii. Limiting the Constituent Power������������������������������������������ 109 iii. Lessons from Failure and Success��������������������������������������� 111 B. Constituting the Amendment Power������������������������������������������� 113 i. People Who?���������������������������������������������������������������������� 115 ii. People When?��������������������������������������������������������������������� 117 a. ‘The People’ has Left the House���������������������������������� 117 b. ‘We’ are Always Open������������������������������������������������� 118 c. Follow the Yellow Brick Road������������������������������������� 119 II. Conclusion����������������������������������������������������������������������������������������� 120 6. Sieyès: The Spirit of Constitutional Democracy?������������������������������������������ 121 Luisa Fernanda García López I. Introduction��������������������������������������������������������������������������������������� 121 II. Towards a Representative Democracy������������������������������������������������ 124 A. From the Tiers État to the National Constituent Assembly����������������������������������������������������������������� 125 B. From Citizenship to Constituent Power: The Foundation of Political Representation�������������������������������� 127 III. Towards a Constitutional Democracy������������������������������������������������ 129 A. The Sovereign People������������������������������������������������������������������ 130 B. The Sovereign Constitution�������������������������������������������������������� 131 IV. Conclusion����������������������������������������������������������������������������������������� 133 7. Revolutionary Reform in Venezuela: Electoral Rules and Historical Narratives in the Creation of the 1999 Constitution������������������� 137 Joshua Braver I. Introduction��������������������������������������������������������������������������������������� 137 II. Carl Schmitt’s Unfortunate Victory over Hannah Arendt in the Analysis of Popular Constitution-Making�������������������������������� 139 III. Hannah Arendt’s Revolutionary Reform�������������������������������������������� 141 A. The Dangerous Freedom of the People��������������������������������������� 141 B. Renewal and Revolutionary Reform������������������������������������������� 143 C. Extrapolating from Arendt: Unconventional Adaptation������������ 144 IV. Hugo Chávez’s Radical and Original Constituent Power������������������� 145 V. The Turning Point: The Electoral Rules for the Constituent Assembly������������������������������������������������������������������������ 148 A. The Difficulty of Fighting Against the Referendum��������������������� 149 B. Ex-Ante Control of the Constituent Assembly through Electoral Rules�������������������������������������������������������������� 150 VI. Radical Breaks and Exclusionary Mandates��������������������������������������� 151 A. Liberal Democracy’s Potential for Revolutionary Reform����������� 152 B. First Past the Post: The Revolutionary Mandate to Destroy the Past��������������������������������������������������������������������� 153 C. Proportional Representation: Inclusion and the Pluralised People������������������������������������������������������������������� 154 VII. Conclusion����������������������������������������������������������������������������������������� 156

x  Contents 8. ‘Revolutionary Reform’ and the Seduction of Constitutionalism���������������������������������������������������������������������������������� 157 Juliano Zaiden Benvindo I. A Revolution in Crisis: Braver’s Narrative of the Creation of the 1999 Constitution in Venezuela�������������������������������� 157 II. The Dilemma of the Sovereignty of the People���������������������������������� 162 III. The Seduction of Constitutionalism: When Constitutionalism Faces the Contingencies of Social Life������������������ 168 IV. Conclusion���������������������������������������������������������������������������������������� 172 Part II: The Traditions of Constitutional Amendment 9. Constitutional Sunrise������������������������������������������������������������������������������� 177 Sofia Ranchordás I. Introduction�������������������������������������������������������������������������������������� 177 II. Sunrise Clauses��������������������������������������������������������������������������������� 180 A. Automatic and Contingent Sunrise Clauses������������������������������� 182 i. Automatic Sunrise Clauses�������������������������������������������������� 182 ii. Contingent Sunrise Clauses������������������������������������������������� 183 B. Sunrise Clauses and the Use of Conditions�������������������������������� 184 C. Sunrise Clauses and By-Laws����������������������������������������������������� 186 III. Contingent Constitutional Change Between Retrospective and Foresight�������������������������������������������������������������� 187 A. Sunset Clauses��������������������������������������������������������������������������� 189 B. Sunrise Clauses and Aspirational Constitutionalism������������������ 190 IV. The Constitutionality of Sunrise Clauses������������������������������������������� 191 A. Riordan v An Taoiseach������������������������������������������������������������ 192 B. Sunrise Clauses and the Constituent Power������������������������������� 194 V. Conclusion���������������������������������������������������������������������������������������� 196 10. Constitutional Change and Interest Group Politics: Ireland’s Children’s Rights Referendum����������������������������������������������������� 199 Oran Doyle and David Kenny I. Introduction�������������������������������������������������������������������������������������� 199 II. Referendums and Constitutional Change������������������������������������������ 201 III. Political Cleavages and Narratives for Change���������������������������������� 202 IV. Children’s Rights: From Lawyers to Interest Groups and Back Again�������������������������������������������������������������������� 204 V. The Referendum Campaign�������������������������������������������������������������� 211 VI. Conclusion���������������������������������������������������������������������������������������� 216 11. Amendment-Metrics: The Good, the Bad and the Frequently Amended Constitution���������������������������������������������������������������������������������������������� 219 Xenophon Contiades and Alkmene Fotiadou I. Introduction: Does the Frequency of Amendment Relate to Constitutional Quality?����������������������������������������������������� 219

Contents xi II. Constitutional Length and the Economy: An Unexpected Relationship������������������������������������������������������������� 220 A. Using Metrics to Evaluate the Quality of Constitutions������������� 220 B. Persuasion through Numbers: Dubious Explanations���������������� 223 III. Poor Countries, Lengthy Constitutions and High Amendment Rates: In Search of an Explicable Correlation��������������� 228 A. Correlating History and Political Culture to Constitutional Length���������������������������������������������������������������� 228 B. When do Constitutions Change?����������������������������������������������� 232 C. The Greek Example: Testing the Neutrality Amendment Proposals for a Shorter Constitution����������������������������������������� 234 IV. Is There Such Thing as the ‘Ideal Constitution’?������������������������������� 236 V. Conclusion���������������������������������������������������������������������������������������� 239 12. Comment on Amendment-Metrics: The Good, the Bad and the Frequently Amended Constitution������������������������������������������������ 241 James E Fleming I. Introduction: Does the Frequency of Amendment Relate to Constitutional Quality?����������������������������������������������������� 241 II. The Notion of a Bad Constitution���������������������������������������������������� 242 III. Criteria for a Good Constitution������������������������������������������������������ 245 IV. The Purposes of Amendment������������������������������������������������������������ 248 V. The Fallacy of Confusing Correlation with Causation���������������������� 249 VI. Conclusion���������������������������������������������������������������������������������������� 251 13. Constituting ‘the People’: The Paradoxical Place of the Formal Amendment Procedure in Australian Constitutionalism���������������������������� 253 Lael K Weis I. Introduction�������������������������������������������������������������������������������������� 253 A. Australia’s Amendment Procedure in Context��������������������������� 255 B. The Place of the Formal Amendment Procedure in Australian Constitutionalism������������������������������������������������� 258 i. Section 128 as the locus of Australian Popular Sovereignty����������������������������������������������������������� 258 ii. Section 128 as the Site of Ordinary Politics����������������������� 262 II. Evaluation and Conclusion��������������������������������������������������������������� 268 14. Hard Amendment Cases in Canada����������������������������������������������������������� 273 Kate Glover I. Part V and Hard Amendment Cases������������������������������������������������� 275 A. The Logic of Part V������������������������������������������������������������������� 275 B. Hard Amendment Cases������������������������������������������������������������ 277 II. Principles of Application������������������������������������������������������������������� 278 A. Supporting Characteristics and Qualitative Assessments����������� 279 i. The Principle��������������������������������������������������������������������� 279 ii. The Principle in Action������������������������������������������������������ 279 iii. The Case of Supreme Court Reform���������������������������������� 281

xii  Contents B. Enhancements and Alterations of Architectural Interests����������� 283 i. The Principle��������������������������������������������������������������������� 283 ii. The Principle in Action������������������������������������������������������ 283 III. The Case of Mandatory Bilingualism at the Court���������������������������� 285 I V. Conclusion���������������������������������������������������������������������������������������� 291 15. Formal Amendment Rules and Constitutional Endurance: The Strange Case of the Commonwealth Caribbean���������������������������������� 293 Derek O’Brien I. Introduction�������������������������������������������������������������������������������������� 293 II. The Rationale for the Inclusion of Formal Amendment Rules in Commonwealth Caribbean Constitutions��������������������������� 295 A. Special Legislative Majorities����������������������������������������������������� 295 B. Referendum Requirements�������������������������������������������������������� 297 C. Constitution-Making in the Eastern Caribbean������������������������� 298 III. Post-Independence Constitutional Reform and Amendment Culture������������������������������������������������������������������� 301 A. Special Legislative Majorities����������������������������������������������������� 302 B. Referendums����������������������������������������������������������������������������� 302 i. Guyana������������������������������������������������������������������������������ 303 ii. Nevis��������������������������������������������������������������������������������� 304 iii. The Bahamas��������������������������������������������������������������������� 305 iv. St Vincent and the Grenadines������������������������������������������� 306 C. Amendment Culture������������������������������������������������������������������ 307 IV. Conclusion���������������������������������������������������������������������������������������� 312 16. The French People’s Role in Amending the Constitution: A French Constitutional Analysis from a Pure Legal Perspective������������������������������� 315 Jean-Philippe Derosier I. The Position of the People: Fundamental and Absolute, in Appearance����������������������������������������������������������������������������������� 317 A. The People’s Fundamental Desire for the Present Constitution������������������������������������������������������������������ 317 B. The People’s Central Role in the Constitution’s Amendment Procedure�������������������������������������������������������������� 319 C. The People’s Absolute will Reflected in the Constitution����������� 320 II. The Position of the People: Limited and Relative, in Law����������������� 322 A. Social People and Legal People�������������������������������������������������� 322 B. The 1962 Legal Revolution������������������������������������������������������� 323 C. The Amending Procedure of Article 11 Questioned������������������� 325 17. The Implication of Conflation of Normal and ‘Constitutional Politics’ on Constitutional Change in Africa���������������������������������������������� 327 Duncan Okubasu I. Introduction�������������������������������������������������������������������������������������� 327 II. Decision Making in a Polity�������������������������������������������������������������� 329

Contents xiii I II. IV. V. VI.

Expression of Normal Politics in Constitutional Politics������������������� 331 Revisiting Effectiveness and Stability������������������������������������������������ 336 Constitutional Irrelevance����������������������������������������������������������������� 338 Conclusion���������������������������������������������������������������������������������������� 340

18. Direct Democracy and Constitutional Change in the US: Institutional Learning from State Laboratories������������������������������������������ 343 Jurgen Goossens I. Introduction�������������������������������������������������������������������������������������� 343 II. Article V of the US Constitution������������������������������������������������������� 344 A. The Four Classic Paths to Amendment�������������������������������������� 344 B. Exclusive Reading of Article V�������������������������������������������������� 346 C. Non-Exclusive Reading of Article V������������������������������������������ 346 III. State Laboratories����������������������������������������������������������������������������� 350 A. Amendment by Proposal of the State Legislature����������������������� 351 i. One or Two Legislative Sessions—Supermajority vs Simple Majority Vote���������������������������������������������������� 352 ii. Special vs General Election������������������������������������������������ 353 iii. Size of Popular Vote���������������������������������������������������������� 354 B. Amendment by Popular Initiative���������������������������������������������� 355 C. Constitutional Convention�������������������������������������������������������� 357 D. Constitutional Commission������������������������������������������������������� 360 IV. Alterations to the Federal Amendment Procedure����������������������������� 361 A. Initiative Petitioning������������������������������������������������������������������ 361 B. Constitutional Referenda����������������������������������������������������������� 363 C. Simple Majority for Popular Votes�������������������������������������������� 365 V. Conclusion���������������������������������������������������������������������������������������� 366 VI. Attachment: Methods for Constitutional Amendment Provided by the State Constitutions�������������������������������������������������� 367 Conclusion: The Emergence of Comparative Constitutional Amendment as a New Discipline: Towards a Paradigm Shift���������� 369 Xenophon Contiades and Alkmene Fotiadou I. A Paradigm Shift in Comparative Constitutional Change�������������������� 369 II. The Amendability Factor: Understanding Constitutions through their Change���������������������������������������������������� 371 III. New Taxonomies of Constitutional Change���������������������������������������� 374 IV. Symbolism and Functionality of Formal Amendment Rules���������������� 377 V. Conscious Constitutional Design: Drafting Amendment Rules������������ 379 VI. Legitimacy and Constitutional Change������������������������������������������������ 383 VII. Conclusion: Toward a Holistic Conceptualisation of Comparative Constitutional Amendment���������������������������������������� 387 Index������������������������������������������������������������������������������������������������������������������ 389

xiv 

Notes on Contributors Richard Albert, Professor of Constitutional Law and Nicholson Scholar, Boston College Law School (United States) Juliano Zaiden Benvido, Professor of Constitutional Law, University of Brasília (Brazil) Joshua Braver, Doctoral Candidate, Yale University Department of Political Science Xenophon Contiades, Professor of Public Law, Panteion University of Social and Political Sciences (Athens, Greece) and Managing Director of the Centre for ­European Constitutional Law Jean-Philippe Derosier, Full Professor of Public Law, University Lille 2—Droit & Santé (France) Oran Doyle, Associate Professor, Trinity College Dublin (Ireland) James E Fleming, The Honorable Paul J Liacos Professor of Law, Boston University School of Law (United States) Alkmene Fotiadou, Dr.iur. and Research Associate at the Centre for European ­Constitutional Law (Greece) Luisa Fernanda García López, Professor of Constitutional Law, Universidad El Rosario in Bogota (Colombia) Kate Glover, Assistant Professor, Faculty of Law, Western University (Canada) Jurgen Goossens, Postdoctoral Fellow of the Research Foundation—Flanders (FWO), Ghent University, and Assistant Professor of Constitutional Law, Erasmus University of Rotterdam (Netherlands) David Kenny, Assistant Professor of Law, Trinity College Dublin (Ireland) Derek O’Brien, Reader in Public Law, Oxford Brookes University (United Kingdom) Zoran Oklopcic, Associate Professor at the Department of Law and Legal Studies, Carleton University (Canada) Duncan Munabi Okubasu, Director of the Centre for Jurisprudence & Constitutional Studies, Kabarak University School of Law (Kenya) Thomaz Pereira, Professor, FGV Direito Rio (Brazil) Sofia Ranchordás, Assistant Professor of Constitutional and Administrative Law, Leiden Law School (Netherlands)

xvi  Notes on Contributors Yaniv Roznai, Assistant Professor, Radzyner School of Law, Interdisciplinary Center (IDC) Herzliya (Israel) Mark Tushnet, William Nelson Cromwell Professor of Law, Harvard Law School (United States) Lael Weis, Lecturer, Melbourne Law School, University of Melbourne (Australia)

Introduction The State of the Art in Constitutional Amendment RICHARD ALBERT

I

N CODIFIED CONSTITUTIONS, amendment rules are the gatekeepers to the constitutional text. They give political actors a roadmap to alter a constitution,1 they identify what is subject to or immune from change,2 they encourage public deliberation on constitutional meaning,3 they foster stability by making a constitution harder to change than regular legislation,4 and they make it possible to make transformative constitutional changes without recourse to revolutionary means.5 By their nature, the rules of constitutional amendment reflect both faith and distrust in political actors: they authorise political actors to make alterations to a constitution,6 just as they limit how and when political actors may do so.7 Given the many essential functions served by constitutional amendment rules, we would expect constitutions to entrench them, and indeed almost all of them do. Yet the study of constitutional amendment has received little scholarly attention in a regime-centric fashion that brings together comparative, doctrinal, historical, and theoretical perspectives. Scholars have given significant attention to the study of informal amendment,8 which we can define as ‘the alteration of constitutional

1 R Dixon and R Holden, ‘Constitutional Amendment Rules: The Denominator Problem’ in T ­Ginsburg (ed), Comparative Constitutional Design (Cambridge University Press, 2012) 195 at 195. 2  J Elster, Ulysses Unbound: Studies in Rationality, Precommitment, and Constraints (Cambridge University Press, 2000) 101–04. 3  R Ku, ‘Consensus of the Governed: The Legitimacy of Constitutional Change’ (1995) 64 Fordham L Rev 535 at 571. 4 K Sullivan, ‘Constitutional Constancy: Why Congress Should Cure Itself of Amendment Fever’ (1996) 17 Cardozo L Rev 691 at 695. 5  W Dellinger, ‘The Legitimacy of Constitutional Change: Rethinking the Amendment Process’ (1983) 97 Harv L Rev 386 at 431. 6 BP Denning and J R Vile, ‘The Relevance of Constitutional Amendments: A Response to David Strauss’ (2002) 77 Tul L Rev 247 at 275. 7  DJ Boudreaux and AC Pritchard, ‘Rewriting the Constitution: An Economic Analysis of the Constitutional Amendment Process’ (1993) 62 Fordham L Rev 111 at 123–24. 8 See, eg, B Ackerman, We the People—Volume I: Foundations (Harvard University Press, 1991); B Ackerman, We the People—Volume II: Transformations (Harvard University Press, 1998); B Ackerman, We the People—Volume III: The Civil Rights Revolution (Harvard University Press, 2014); BP Denning, ‘Means to Amend: Theories of Constitutional Change’ (1997) 65 Tenn L Rev 155; P Oliver, ‘Canada, Quebec, and Constitutional Amendment’ (1999) 49 Toronto Law Journal 519; DA Strauss, ‘The Irrelevance of Constitutional Amendments’ (2001) 114 Harv L Rev 1457.

2  Richard Albert meaning in the absence of textual change’.9 But there has been a relative dearth of inquiry into formal amendment itself as a field of study probing how its various procedures interrelate with informal mechanisms of constitutional change. As Sandy Levinson has suggested, this is both surprising and problematic because few questions in the study and practice of constitutional design are more important than how to structure the rules of constitutional amendment.10 There is of course more to constitutional amendment than procedures alone. ­Zachary Elkins, Tom Ginsburg and James Melton have shown in their work on constitutional endurance that the study of constitutional amendment offers a window into the causes and consequences of constitutional stability and fragility.11 The study of constitutional amendment also raises foundational questions about the legitimacy of the constitutional order, the locus of sovereignty in a jurisdiction, whether we should understand formal amendment as part of a larger continuous category of constitutional change and whether there is something special about a popularlylegitimated process of altering a constitutional text.12 There are moreover deeply important normative implications to the chosen forms of constitutional change, whether it occurs formally or informally, and whether it entails substantial or minor changes to a constitution. In all cases, it is a constitution that changes, and this entails consequences vastly different from changing an administrative regulation, an ordinary statute, or a high court decision. In recent years, scholars have produced creative scholarship in constitutional amendment. They have asked what makes a constitution endure through crisis,13 they have explored unamendability in its various forms, whether formal,14 ­informal,15 or constructive,16 and they have taken critical approaches to the present-day implications of constitutional amendment rules.17 Scholars have also probed some of the major questions in constitutional change confronting constitutional designers, including the challenges of abusive constitutionalism,18 temporary constitutions,19 and stealth authoritarianism.20 This field of study is now thriving.

9  HK Gerken, ‘The Hydraulics of Constitutional Reform: A Skeptical Response to Our Undemocratic Constitution’ (2007) 55 Drake L Rev 925 at 929. 10 See S Levinson, ‘Designing an Amendment Process’ in J Ferejohn, JN Rakove and J Riley (eds), ­Constitutional Culture and Democratic Rule (Cambridge University Press, 2001) 271 at 275. 11  Z Elkins, T Ginsburg and J Melton, The Endurance of National Constitutions (Cambridge University Press, 2009) at 99–100. 12  R Albert, ‘Nonconstitutional Amendments’ (2009) 22 Canadian Journal of Law & Jurisprudence 5 at 9–10. 13 See X Contiades and A Fotiadou, ‘On Resilience of Constitutions: What Makes Constitutions Resistant to External Shocks?’ (2015) 9 ICL Journal 3. 14  See M Schwartzberg, Democracy and Legal Change (Cambridge University Press, 2007). 15  See S Krishnaswamy, Democracy and Constitutionalism in India: A Study of the Basic Structure Doctrine (Oxford University Press, 2009). 16  See R Albert, ‘Constructive Unamendability in Canada and the United States’ (2014) 67 Sup Ct L Rev (2d) 181. 17  See Y Roznai and S Suteu, ‘The Eternal Territory? The Crimean Crisis and Ukraine’s Territorial Integrity as an Unamendable Constitutional Principle’ (2015) 16 German L J 542. 18  D Landau, ‘Abusive Constitutionalism’ (2013) 47 U C Davis L Rev 189. 19  OO Varol, ‘Temporary Constitutions’ (2014) 102 Cal L Rev 409. 20  OO Varol, ‘Stealth Authoritarianism’ (2015) 100 Iowa L Rev 1673; GM Hahn, ‘Managed Democracy? Building Stealth Authoritarianism in St. Petersburg’ (2004) 12 Demokratizatsiya 195.

Introduction 3 I.  DEFINING THE FIELD

There is, with good reason, growing interest in constitutional amendment from a comparative perspective. Comparative constitutional amendment is the study of how supranational, national and subnational constitutions change through formal and informal means, including alteration, revision, evolution, interpretation, replacement and revolution. The field invites scholars to draw insights about constitutional change across borders and cultures, to uncover the motivations behind constitutional change, to theorise best practices and to identify the theoretical underpinnings of constitutional change. This volume is designed to guide the emergence of comparative constitutional amendment as a distinct field of study in public law. Much of the recent scholarship in the field has been written by the scholars assembled in this volume. This book, like the field it hopes to shape, is not comparative alone; it is also doctrinal, historical and theoretical, and therefore offers a multiplicity of perspectives on a subject about which much remains to be written. No other book to date has covered the ground we do here. There are currently four major volumes in the field, but none has accomplished what we offer in this collection. The first major volume is edited by Xenophon Contiades,21 and the second by Dawn Oliver and Carlo Fusaro.22 Both volumes gather country experts to describe constitutional change in specific jurisdictions. Their coverage of the world is impressive and useful, and each book has become a must-read for anyone in the field. The same is true of the third major volume in the field, edited by Mads Andenas; it is an informative collection of country-specific case studies on constitutional change.23 Our new volume builds on all three by offering a broad range of cross-jurisdictional perspectives and a theoretical analysis of the various foundations and traditions of constitutional change. The fourth major book in the field, edited by Sanford Levinson, is almost exclusively internal to the United States.24 It too is a must-read for the richness of its engagement with the theoretical underpinnings of constitutional amendment. The comparative, doctrinal, historical and theoretical approaches taken in this volume give readers a broader orientation to the complexity of constitutional change across the globe. The authors in this collection take a contextual approach to the study of ­constitutional amendment. The field has been gradually developing into a distinct area in public law. Both at the national level and from a comparative angle, there is an emerging set of related questions in constitutional change inquiring into the promise and peril of codifying constitutions, the meaning and formation of constitutional identity, and the purpose of constitutionalism itself. Much of the innovative

21 X Contiades (ed), Engineering Constitutional Change: A Comparative Perspective on Europe, ­Canada and the USA (Routledge, 2013). 22  D Oliver and C Fusaro (eds), How Constitutions Change: A Comparative Study (Hart Publishing, 2011). 23  M Andenas (ed), The Creation and Amendment of Constitutional Norms (British Institute of International and Comparative Law, 2000). 24  S Levinson (ed), Responding to Imperfection: The Theory and Practice of Constitutional Change (Princeton University Press, 1995).

4  Richard Albert dialogue currently ongoing in the field is driven by comparative perspectives on these questions and controversies. This volume engages with the major lines of inquiry and suggests new ones for the years ahead. We aim in this volume to create a reference point for scholars interested in the subject. The volume covers the major topics where innovative work is being done, including the contested notion of the people and the conventional divide between the constituent and constituted powers, empirical quantitative approaches to constitutional change, the forms and challenges of formal and informal unamendability, the design and limits of temporally-bound amendment rules, constitutional referenda and the inevitable impermanence of constitutional change. We have designed the volume as a dialogue that cuts across these innovative conceptualisations, highlights scholarly disagreement and in so doing puts big ideas to the test. This volume therefore captures the fierce ongoing debates on the frontier topics in constitutional change, it reveals the current trends and contested questions, and it probes a variety of arguments elaborated by experts in the field. This volume is intended to chart a path for rigorous study and constructive dialogue in the field. II.  THE ARCHITECTURE OF CONSTITUTIONAL AMENDMENT RULES

The study of constitutional amendment must begin with the way codified constitutions are actually altered. Lester Orfield long ago taught us that ‘[t]he idea of amending the organic instrument of a state is peculiarly American’.25 Early state constitutions in the United States entrenched formal amendment rules of their own. But the first national constitution in the modern era to entrench a formal amendment rule is the Articles of Confederation, America’s first constitution. Adopted in 1777, the Articles did not make amendment easy. It required approval from the unicameral Continental Congress along with the unanimous agreement of the 13 states in order to alter its text, no small feat for any federal state: And the Articles of this Confederation shall be inviolably observed by every State, and the Union shall be perpetual; nor shall any alteration at any time hereafter be made in any of them; unless such alteration be agreed to in a Congress of the United States, and be afterwards confirmed by the legislatures of every State.26

Every proposal to amend the Articles failed.27 There predictably came a point in time when some constitutional change became necessary for the improvement of the Union if not for its survival. Yet instead of abiding by the impenetrable unanimity rule in the Articles, the Philadelphia Convention that had been assembled ‘for the sole and express purpose of revising the Articles of Confederation’28 decided to

25 

LB Orfield, The Amending of the Federal Constitution (University of Michigan Press, 1942) 1. Articles of Confederation, Art XIII (1781). 27  BF Wright, ‘Consensus and Continuity—1776–1878’ (1958) 38 B U L Rev 1 at 19. 28  Journals of the Continental Congress, 1774–1789, Feb 21, 1787 (Washington C Ford et al, 1904–37) at 74. 26 

Introduction 5 cast aside the constraints in the Articles and instead to propose an altogether new constitution subject to a lower yet still considerably difficult amendment threshold: approval from two-thirds of the bicameral legislature and agreement from threequarters of the states. This is the amendment rule that now entrenches the United States Constitution.29 It was risky for the Philadelphia Convention to propose a new constitution. The Convention’s delegates were arguably violating the rules of the Articles by ignoring its rules of change. Further, the requirement they imposed for the ratification of the new constitution was a high bar to clear: ‘The Ratification of the Conventions of nine States, shall be sufficient for the Establishment of this Constitution between the States so ratifying the Same.’30 The outcome was far from guaranteed; s­ ecuring the agreement of nine out of 13 states took sustained advocacy and action. The Philadelphia Convention believed that the ratification of the Constitution in nine irregular assemblies of citizens—the Conventions of nine States—would retrospectively legitimate their contested choice to break with the rules of change in the Articles.31 The puzzle was how to ground the new Constitution in an authority higher than the Articles. The Articles possessed authority as a matter of law and also functionally in that they brought stability to the Union, if even to the point of stasis. The bridge from discarding the Articles to adopting the new Constitution was the idea of a convention, a revolutionary gathering of people in collective debate, deliberation and decision-making. If the people could make an informed choice to discard the Articles and to simultaneously accept the new Constitution, the formalist argument of the illegality of the new Constitution would lose its moral force. Jack Rakove put its well: the Philadelphia Convention ‘understood that a constitution adopted through some process of popular ratification could be said to have attained a superior authority’ than what the Articles could claim from its own ratification.32 The successful ratification of the new Constitution gave it an unassailable legitimacy: it became ‘regarded as the product of a process in which the ultimate source of legitimacy, the sovereignty of the people, was expressed as fully and as clearly as the accepted political beliefs and institutions of the time allowed’.33 The reason why the idea of a convention could bring legitimacy to the proposed Constitution and to the Philadelphia Convention’s decision to violate the Articles highlights the first major subject of the volume—a subject that doubles as the architectural foundation of constitutional amendment rules: the constituent power. The theory of constituent power springs from the thought of Emmanuel Joeph Sieyès, an 18th century French theorist who argued that the people were the fountain of legitimate authority.34 Sieyès distinguished the pouvour constituant from the pouvoir

29 

US Constitution, Art V (1789). ibid at Art VII. 31 Ackerman, We The People—Transformations (n 8) 85. 32  J Rakove, ‘Constitutional Problematics, circa 1787’ in J Ferejohn et al (eds), Constitutional Culture and Democratic Rule (Cambridge University Press, 2001) 41 at 65. 33  RS Kay, ‘The Illegality of the Constitution’ (1987) 4 Const Comment 57 at 75. 34  EJ Sieyès, Qu’est-ce que le Tiers état? (Éditions du Boucher, 2002) (originally published in 1789). 30 

6  Richard Albert constitué,35 the former referring to the people themselves acting in their constitutionmaking capacity and the latter to the institutions the people create—institutions that are authorised only to change the Constitution within the constitutional framework created by the people themselves. On this theory, the people of the United States acting deliberatively in state conventions across the Union exercised their constituent power to adopt the new Constitution and in so doing to legitimate their formally illegal break with the Articles. The theory of constituent power is not a legal concept but rather a sociological one. Yet it has been interpreted to entail implications in law for whether courts should have the power to declare constitutional amendments unconstitutional, for how we can identify the mystical body that is the people, and for how to differentiate constitution-making as a form of constitutional alteration from constitutional amendment. The constituent power features centrally in each of the chapters in Part I of this volume—a part we call The Foundations of Constitutional Amendment— written separately by Juliano Benvindo, Joshua Braver, Oran Doyle, Luisa Fernanda García López, Zoran Oklopcic, Thomaz Pereira, Yaniv Roznai and Mark Tushnet. Some constitutions seek self-consciously to constitutionalise the constituent power. They will, for example, create two separate tracks of rules for formal alteration: procedures that authorise constitutional amendment consistent with the existing framework and presuppositions of a given constitution and alternative procedures that authorise total replacement whether or not the changes cohere with that constitution as it exists and is presently interpreted. The Costa Rican Constitution, for example, creates separate rules for ‘partial amendment’ and ‘general amendment’,36 the first set deployable within exacting specifications that endeavour to ensure that the result will be an amended constitution that fits with the existing constitution and the second set useable for all manner of changes, including those that depart from and even violate the existing constitution. In the conventional theory of constitutional change, using the former procedures would amount to an exercise of an inferior constituted power while using the latter would represent the people marshalling their superior constituent power. The Constitutions of Austria, Spain and Switzerland are structured in a similar way.37 Yet whether a constitution self-consciously entrenches these two separate tracks of procedures for formal alteration, the conventional theory insists that they exist implicitly such that the inferior constituent power cannot be used to replace a constitution. This was the Indian Supreme Court’s justification for the basic structure doctrine, which the Court created to invalidate constitutional amendments that, in its view, created a new constitution without invoking the constitution-making power that only the constituent power can authorise.38 This distinction between constituent and constituted powers is also the basis for the Colombian Constitutional Court’s

35 

ibid at 53. Constitution of Costa Rica, Arts 195–96 (1949). 37  See Constitution of Austria, Art 44 (1920); Constitution of Spain, Arts 192–95 (1978); Constitution of Switzerland, Art 192 (1999). 38  See S Krishnaswamy, Democracy and Constitutionalism in India: A Study of the Basic Structure Doctrine (Oxford University Press, 2010). 36 

Introduction 7 substitution of the Constitution doctrine,39 an analytical framework whose aspirations are substantially similar to the basic structure doctrine, specifically to protect the Constitution from its replacement by amendment rules rather than the rules of replacement themselves. Colombia and India are today the leading sites for the study of unconstitutional constitutional amendment, though many courts around the world have taken to the idea that they can invalidate a constitutional amendment that goes too far.40 A. Unamendability For some, the idea of an unconstitutional constitutional amendment is problematic from a democratic perspective.41 It seems contrary to the very idea of a ­constitution—which ought to reflect the views and wishes of those bound by it—that there could be enforceable limits to what the people may do acting through their representatives when they amend their constitution in perfect conformity with the entrenched procedures for constitutional change. The question becomes one of institutional competence where a court identifies those limits to the amendment power against the backdrop of a codified constitution that in its own text makes no mention of such limits. The Court’s judgment about what is or is not unamendable cannot help but be rooted in the political philosophy of its individual members. Perhaps this line-drawing would be better assigned to a legislature than a court? The calculation may change where a constitution entrenches a formally unamendable provision. Constitutions often make principles, structures, and symbols impervious to formal amendment by removing them from the realm of amendable provisions. For example, republicanism,42 democracy,43 federalism,44 the separation of powers,45 and secularism46 are often made unamendable, as are rights and ­freedoms,47 pluralism,48 national borders,49 and even national flags.50 Constitutional designers may have many reasons to adopt formal unamendability as a focal feature of their constitution, whether to formalise a bargain or to preserve a founding norm, to transform the state or to reconcile previously warring groups, or quite simply to

39  C Bernal, ‘Unconstitutional Constitutional Amendments in the Case Study of Colombia: An Analysis of the Justification and Meaning of the Constitutional Replacement Doctrine’ (2013) 11 Int’l J Const L 339. 40  See A Barak, ‘Unconstitutional Constitutional Amendments (2011) 44 Israel L Rev 321. 41  See J Colón-Ríos, Weak Constitutionalism: Democratic Legitimacy and the Question of Constituent Power (Routledge, 2012); R Albert, ‘Constitutional Handcuffs’ (2010) 42 Ariz St L J 663. 42  See Constitution of Italy, Art 139 (1948). 43  See Basic Law of Germany, Arts 20, 79 (1949). 44  See Constitution of Brazil, Art 60 (1988). 45  See Constitution of Greece, Art 110 (1975). 46  See Constitution of Turkey, Art 4 (1982). 47  See Constitution of Namibia, Art 131 (1998). 48  See Constitution of Portugal, Art 288 (1976). 49  See Constitution of Cameroon, Art 64 (1972). 50  See Constitution of Timor-Leste, Art 156 (2002).

8  Richard Albert express a constitutional value, whether authentic or not.51 Whatever the motivation for making a provision formally unamendable, it will be necessary to interpret the scope of its formal unamendability against any set of facts that arise. Here, the case may be stronger for a judicial role on the front lines of enforcing unamendability because the task is squarely within the judicial competence of constitutional interpretation. But interpreting what is unamendable within the four corners of the codified constitution where its text identifies nothing as formally unamendable is a qualitatively different kind of undertaking, one that appears to be closer to the task of making a constitution than interpreting one. Yet formal unamendability does not obviate the question whether a legislature may be better positioned than a court to enforce the unamendable provision. The Norwegian Constitution, for instance, entrenches a formally unamendable provision but nonetheless assigns to the legislature the power to protect the contours and content of that provision.52 Whatever a constitution entrenches against amendment and whoever assumes the responsibility to protect it from violation, the very fact that constitutional designers chose to make one or more items unamendable tells us a lot about that constitution and its essential values. B.  The Operation of Constitutional Amendment One of the first building blocks in the procedures of constitutional amendment is the power to initiate a constitutional amendment. Designing the power to initiate is a binary choice. One option is to restrict the power to a single actor. In Germany, for instance, only the national legislature may initiate an amendment: the Basic Law is amendable only by two-thirds approval in each of the Bundestag and the ­Bundesrat.53 The power of initiation may alternatively be held by multiple actors, as in Canada, where three of its five amendment procedures may be initiated by either the national legislature or any of the subnational legislatures.54 The Spanish Constitution likewise assigns the power of initiation to several actors, including the government, the national legislature, and the country’s autonomous communities indirectly through their legislatures.55 Related to but distinct from the power of initiation is the scope of the procedures available to political actors to amend the Constitution. Some constitutions provide one or more avenues to amend all of the amendable parts of the text while others entrench multiple procedures that are expressly designated for use in relation to a limited category of amendable matters. We may identify these as either comprehensive or restricted in scope. In Japan, for example, the Constitution entrenches only one comprehensive amendment procedure that may be used to amend all of

51  R Albert, ‘The Unamendable Core of the United States Constitution’ in A Koltay (ed), Comparative Perspectives on the Fundamental Freedom of Expression (Wolters Kluwer Ltd, 2015) 13 at 15–19. 52  See Constitution of Norway, Art 121 (1814). 53  See Basic Law of Germany, Art 79. 54  See Constitution Act 1982, being Schedule B to the Canada Act 1982, ss 38, 41, 43, 46 (UK). 55  See Constitution of Spain, Art 166 (1978).

Introduction 9 the amendable parts of the text: in order to pass a valid amendment, the bicameral national legislature must approve an amendment proposal by a two-thirds supermajority, the proposal must then be approved in a national referendum by a simple majority vote, and the Emperor must then promulgate amendment.56 In France the Constitution entrenches multiple comprehensive procedures that may be used to amend all the amendable parts of the text: in order to pass a valid amendment, the President or any member of the national legislature may propose an amendment, and both houses of the legislature must then approve the proposal before it must go before the people in a referendum for their approval.57 But elsewhere, as in South Africa, the Constitution creates three procedures of formal amendment, and each is deployable only to amend certain provisions or principles.58 One procedure requires approval from three-quarters of the National Assembly and two-thirds of the National Council of Provinces in order to amend the Constitution’s declaration of constitutional values and the formal amendment rules themselves.59 Another procedure requires two-thirds approval in each of the National Assembly and the National Council of Provinces to amend the Bill of Rights and provincial rights, prerogatives and matters.60 The third and final procedure requires two-thirds approval in the National Assembly; this procedure must be used to amend all other items in the Constitution.61 Each of these three South African amendment procedures is restricted in scope, in contrast to the comprehensive scope of the Japanese and French amendment procedures. There may be good reason to entrench multiple procedures for amendment, be they comprehensive or restricted in scope. In constitutional democracies where the text is difficult to amend, it may be wise to offer political actors many routes to formal amendment so as to avoid the possibility that the text becomes frozen and unchangeable despite mounting political and popular will to update it. In the United States, for instance, all routes of formal amendment must pass through the Congress. Under Article V, two-thirds of both houses of Congress may propose an amendment, and to be valid three-quarters of the states must ratify it in either a legislative vote or a convention, the choice being up to Congress; alternatively, two-thirds of the states may petition Congress to call a convention to propose one or more amendments, and in order to be valid three-quarters of the states must ratify it or them either in a legislative vote or a convention, and again the choice is up to Congress.62 In all of these four procedures, Congress plays a gatekeeping role. In an age of congressional division and dysfunction, the requirement to assemble the required congressional majorities becomes a serious impediment to an amendment. Today, for

56 

See Constitution of Japan, Art 96 (1947). Constitution of France, Art 89 (1958). The President may bypass the referendum requirement for a governmental amendment bill. In these cases, Parliament convened in Congress must ratify the proposal by a three-fifths majority vote; ibid. 58  See Constitution of South Africa, s 74 (1996). 59 ibid. 60 ibid. 61 ibid. 62  Constitution of the United States, Art V. 57  See

10  Richard Albert example, it appears unlikely that any amendment could be proposed by the divided ­Congress even though it might well be ratifiable by a supermajority of states. A non-­ congressional track to amendment could perhaps make amendment possible in the United States. Returning to the basic building block of constitutional amendment procedures, constitutions sometimes disable the power to initiate an amendment. It may be disabled during periods of national emergency, martial law or a state of siege or war. The Estonian Constitution illustrates how: an ‘amendment of the Constitution shall not be initiated, nor shall the Constitution be amended, during a state of emergency or a state of war’.63 The same applies for periods of regency or succession. When the ruling monarch is absent or otherwise unable to serve, some constitutions might forbid their own amendment, as in Luxembourg where the Constitution declares that ‘during a regency, no change can be made to the Constitution concerning the constitutional prerogatives of the Grand Duke, his status as well as the order of ­succession’.64 The Belgian Constitution is similar.65 There may also be temporal restrictions on the power to initiate a constitutional amendment. Some constitutions require political actors to deliberate on an amendment proposal for a maximum amount of time. We see this in the Costa Rican Constitution.66 Others, like the South Korean Constitution,67 impose a minimum amount of time during which political actors must deliberate on a constitutional amendment. Constitutional amendment procedures may also be disabled in the period following a successful or failed amendment, as in Greece,68 and in the period immediately following the adoption of a new constitution, as in Cape Verde.69 The use and manipulation of time and contingency in the design of constitutional amendment rules is a fascinating subject that deserves more scholarly attention—and in this collection Sofia Ranchordás explores the subject in the context of sunrise clauses that few others have analysed so thoroughly.70 Another important device in the operation of formal amendment rules is the referendum. Whether entrenched in the rules of formal amendment themselves or used as a political imperative to supplement them, referenda are often viewed as a legitimating procedure without peer in constitutional democracies.71 In recent years, we have seen what seems like a proliferation of referenda with significant consequences for rights and liberties, and also for fundamental constitutional structures and arrangements. Think only of the Brexit referendum in the United Kingdom in 2016, the

63 

Constitution of Estonia, s 161 (1992). Constitution of Luxembourg, Art 115 (1868). Constitution of Belgium, Art 197 (1831). 66  See Constitution of Costa Rica, Art 195 (1949). 67  See Constitution of South Korea, Art 129 (1948). 68  See Constitution of Greece, Art 110 (1975). 69  See Constitution of Cape Verde, Art 309 (1980). 70  For a detailed study of the related subject of constitutional sunset clauses, see S Ranchordás, Constitutional Sunsets and Experimental Legislation: A Comparative Perspective (Edward Elgar Publishing Ltd, 2015). 71  For the best modern study of referenda, see S Tierney, Constitutional Referendums: The Theory and Practice of Republican Deliberation (Oxford University Press, 2012). 64  65 

Introduction 11 Peace Process referendum in Colombia in the same year, the Scottish independence referendum in 2014, and the marriage and Senate referenda in Ireland in 2015 and 2013, respectively. A referendum may in the years ahead become by default a requisite part of constitutional amendment procedures whether or not the constitutional text requires it. The many facets of referenda, in both theory and practice, form the core of the chapters in this volume written by Oran Doyle and David Kenny, Jürgen Goosens, and Jean-Philippe Derosier. C.  Three Challenges in the Study of Constitutional Change The study of the foundations and traditions of constitutional amendment necessarily entails comparative inquiry. We can of course undertake single-country studies, as we see in this volume, but we profit most from them when they are read alongside other jurisdictional accounts. This points to a serious limitation in the study of constitutional change specifically, and in comparative public law generally: we most often refer to the same comparator jurisdictions, with little attention given to other parts of the world where important developments may be underway but do not command our attention for a variety of reasons at times understandable and at others inexcusable. Ran Hirschl diagnoses the problem: [S]elf-professed ‘comparativism’ sometimes amounts to little more than a passing reference to the constitution of a country other than the scholar’s own or to a small number of overanalyzed, ‘usual suspect’ constitutional settings or court ruling. The constitutional experiences of entire regions—from the Nordic countries to sub-Saharan Africa to Central and South East Asia—remain largely uncharted terrain, understudied and generally overlooked.72

We seek in this volume to begin to remedy that problem in the field of comparative constitutional amendment. We feature papers on the Caribbean by Derek O’Brien and on Africa by Duncan Okubasu. There are of course many points of national distinction in these large regions but the authors nonetheless find threads that run across them. In their chapters, the authors invite scholars of comparative public law in the future to explore on their own the richness of the constitutional traditions in these underexplored regions. A second challenge in the study of constitutional change concerns the effect of constitutional amendments. We are familiar with the range of motivations why political actors choose to amend their constitution, for instance to correct a discovered flaw or to update a constitution in light of evolving social norms or expectations. But we know much less about the collateral consequences of constitutional amendment. Does a frequently amended constitution generate instability detrimental to the constitutional order or might it bring that constitution closer to the governed by making it seem more accessible and responsive to their wishes? And what about

72 R Hirschl, Comparative Matters: The Renaissance of Comparative Constitutional Law (Oxford University Press, 2014) 4.

12  Richard Albert an infrequently amended constitution? The reasons for its infrequent amendment would certainly matter to the analysis but it could, on one hand, signal a durable ­constitutional settlement where not much needs to change or it could, on the other hand, lead to a constitution’s own irrelevance if most change occurs informally and leads to a disjunction between the constitutional text and political reality. In a ­colloquy in this volume pairing their two chapters, James Fleming exchanges thoughts on these questions with Xenophon Contiades and Alkmene Fotiadou. A third challenge follows from the second: what is the role of the text in constitutional amendment and in the self-understanding the people and political actors have about their constitution? In some cases, for instance in Australia as Lael Weis explains in her volume, the constitutional text takes primacy over manifestations of popular will, a tension for any constitutional democracy with an entrenched constitution. In other cases, a constitution changes in ways not reflected in its constitutional text, and those changes may be accepted as valid and legitimate. These circumstances raise a different kind of tension: whether and how the constitutional text matters. In her chapter in this volume, Kate Glover explores these questions with respect to the Constitution of Canada. III.  AMENDMENT AS CONSTITUTION

Herman Finer famously wrote that ‘we might define a constitution as its process of amendment’.73 More metaphorical than empirical, Finer nonetheless put into words an intuition we have about the rules of constitutional amendment. No part of a constitution is more important,74 not only for formal and functional reasons but also for symbolic ones. Formally, constitutional amendment rules distinguish constitutional from ordinary law,75 the latter generally subject to simple majority legislative repeal while the former is commonly entrenched under heightened thresholds or special procedures.76 Functionally, constitutional amendment rules provide an orderly and transparent procedure to correct flaws or outright mistakes that time and experience reveal.77 Symbolically, the very design of amendment rules may reveal the deepest values of the constitution they entrench.78 What sets amendment rules apart from other constitutional rules are their twin constructive and destructive capacities. On one hand, amendment rules are essential for changing, in a lawful manner, the rules of the game that a constitution creates in

73 

H Finer, Theory and Practice of Modern Government (Henry Holt & Company, 1949) 127. J Burgess, Political Science and Comparative Constitutional Law: Volume I (Ginn & Co, 1893) 137. 75 See A Sajó, Limiting Government: An Introduction to Constitutionalism (Central European ­University Press, 1999) 39–40. 76  Over 95% of the world’s constitutions entrench amendment rules. See F Giovannoni, ‘Amendment Rules in Constitutions’ (2003) 115 Public Choice 37 at 37. 77  See n 6 above. 78  See R Albert, ‘The Expressive Function of Constitutional Amendment Rules’ (2013) 59 McGill L J 225 at 244. 74 

Introduction 13 the first place.79 They allow political actors to enact reforms they believe are needed for either intrinsic or instrumental reasons. Yet the idea of constitutional ­amendment need not be rooted in normative principles. Seen as a neutral device, perhaps as an empty vessel suited to any purpose, amendment rules expose themselves to the possibility of their own misuse. As Ulrich Preuss has warned, the amending power ‘is necessary to preserve the flexibility and sustainability of the constitutional order, but it can destroy it by amending the constitution in an anti-constitutional tenor’.80 Stated less ominously though no less accurately, amendment rules are, in the words of Akhil Amar, ‘of unsurpassed importance, for these rules define the conditions under which all other constitutional norms may be legally displaced’.81 Constitutional amendment rules are both sword and shield. Political actors may use them to build or break the constitutional order, to strengthen or weaken institutions, to expand or retrench rights, to ‘deconstitute and reconstitute’, as Finer well recognised.82 In the United States, for example, what was proposed and partially ratified as the original Thirteenth Amendment would have given states the unamendable power to enslave persons within their jurisdiction.83 The Civil War intervened to put a stop to the ratification vote, and what ultimately became entrenched in the Thirteenth Amendment was instead the abolition of slavery.84 An amendment, then, can be both democracy-promoting and democracy-demoting. As profound as Finer’s words may be, they do not quite capture the reality that amendment rules and their constitutions are rooted in meta-constitutional contexts shaped by history, law, politics and society. From a perspective internal to a constitutional regime it may well be difficult to contest the claim that we can learn much about any given constitution by looking at its amendment rules. These rules often create a constitutional hierarchy of entrenchment that indicates what is most important in a constitution and what is relatively more or less important. The South African Constitution’s escalating multiple thresholds of constitutional amendment is just one of many illustrations.85 But from an external perspective we can appreciate more immediately what political actors within a constitutional system can less clearly see: that whether, when, and how often or infrequently political actors choose to deploy amendment rules is driven by factors better explained by constitutional culture than the theoretical rigidity or flexibility of a constitution itself. The phenomenon of constitutional culture is sometimes reflected in constitutional texts, for instance in how powers are allocated, which institutions exercise them, and how

79 See BE Rasch and RD Congleton, ‘Amendment Procedures and Constitutional Stability’ in RD Congleton and B Swedenborg (eds), Democratic Constitutional Design and Public Policy (MIT Press, 2006) 319 at 319–21. 80  UK Preuss, ‘The Implications of “Eternity Clauses”: The German Experience’ (2011) 44 Israel L Rev 429 at 430. 81 See AR Amar, ‘The Consent of the Government: Constitutional Amendment Outside Article V’ (1994) 94 Colum L Rev 457 at 461. 82  Finer (n 73). 83 See M Vorenberg, Final Freedom: The Civil War, the Abolition of Slavery, and the Thirteenth Amendment (Cambridge University Press, 2001) 20–22. 84  Constitution of the United States, amend XIII (1865). 85  See Constitution of South Africa, s 74 (1996).

14  Richard Albert citizens interact with them. But sometimes the phenomenon of constitutional culture escapes formal entrenchment and perhaps also even recognition by political actors in their own understanding of its sources and consequences. A recent article by Tom Ginsburg and James Melton shows the way forward in the study of constitutional culture: it suggests that amendment culture may be more determinative of a constitution’s amendment rate than the ease or difficulty of its formal amendment rule.86 Derek O’Brien applies his learning from Ginsburg and Melton in his chapter for this volume on constitutional change in the Caribbean. A.  Perspectives on Constitutional Amendment The growing global interest in the study of constitutional change brought together two dozen scholars from around the world for a day-long workshop on the ­foundations and traditions of constitutional amendment. Their papers, now published as chapters in this book, double as an invitation and a challenge. They invite readers to learn about constitutional amendment in the languages of political science, history, law and philosophy. They also challenge readers to complicate their own study of constitutional amendment by looking beyond text and doctrine to mine the deeper reasons behind a given constitutional amendment, a trend of frequent or infrequent amendment, a history of amendment success or failure, and other inquiries both narrow and broad in the study of constitutional change. The authors in this volume hope by their example both to reinforce the growing interest in constitutional change and to light the path to new directions of exploration. This book is divided into two main parts. The first—on the foundations of constitutional amendment—explores the theoretical foundations of constitutional amendment, principally taking as its point of departure the theory of constituent power, an idea that has shaped the study of constitutional amendment and one that the authors in this first part variously reinforce, refine, challenge and interrogate. In the opening chapter on ‘Amendment Power, Constituent Power, and Popular Sovereignty: Linking Unamendability and Amendment Procedures’, Yaniv Roznai develops the distinction between primary constituent (constitution-making) power and secondary constituent (constitution-amending) power, the latter limited by unamendability and the former wholly unlimited by it. Roznai develops the distinction by introducing the ‘spectrum theory’ of constitutional amendment to identify when it is proper to invalidate a constitutional amendment. According to his spectrum theory, the more the democratic characteristics of the amendment power resemble those of the primary constituent power, the less the amendment power should be bound by limitations, and the more it resembles an ordinary legislative power, the more it should be bound by limitations.

86  See Tom Ginsburg and James Melton, ‘Does the Constitutional Amendment Rule Matter at All? Amendment Culture and the Challenges of Measuring Amendment Difficulty’ (2015) 13 Int’l J Const L 686.

Introduction 15 Zoran Oklopcic’s chapter on ‘Constitutional Theory and Cognitive Estrangement: Beyond Revolutions, Amendments and Constitutional Moments’, discusses two competing imaginaries of peoplehood: the Sieyèsian and the Lockean. Under the Lockean view, the people are caught in a binary: either they radically re-constitute the constitutional order as a result from oppression or they change it through regular constitutional channels. In his chapter, Oklopcic seeks to dissolve that binary to free us to study constitutional theory anew. Oklopcic suggests that this new beginning is possible if we estrange ourselves from the concept of the people’s personhood, if we personify the constitutional order itself, and if we see constitutional theory as a real person, with hopes, ideals and anxieties, just like us. Next, Oran Doyle’s chapter on ‘Constraints on Constitutional Amendment Powers’ begins from the proposition that the constraint of constitutional amendment powers, by rules or standards that determine the validity of constitutional amendments, is widespread. The fundamental question posed by this practice is whether the values served by constitutional constraints can justify a contemporary majority being subject to a past generation or a judicial elite. Doyle observes that the literature mostly fails to address those questions directly because it focuses instead on the quantum of change introduced by constitutional amendments. In his chapter, Doyle develops a set of four distinctions that establish a typology of ways to conceive of constraining the amendment power. His typology focuses attention on the extent to which these constraints disempower contemporary majorities in favour of past generations or judicial elites. In this way, it provides the baseline against which we might evaluate the justification of those constraints. Mark Tushnet reacts to Doyle’s ‘Constraints on Constitutional Amendment Powers’ in his comment and analyses Doyle’s arguments on limitations to amendment powers. Tushnet categorises his comments into two broad groups. The first set of comments deal with Doyle’s discussion of what Doyle calls the interpretative and moral approaches to identifying constraints on amendment powers. The second focuses on Doyle’s discussion of the costs to majoritarianism—or self-governance— when there are legal constraints on those powers. In ‘Constituting the Amendment Power: A Framework for Comparative Amendment Law’, Thomaz Pereira argues that the distinction between constituent power and constituted power is intrinsic to the very concept of constitutionalism, since it relates to the essential innovations brought out by the emergence of constitutional democracies. Pereira argues that comparative constitutional law can benefit from a better understanding of these concepts, which can bring insight into the relationship between how different constitutional systems regulate the amendment power, and the political history and constitutional culture of these different systems. One of the major takeaways from his chapter is that an adequate understanding of the divide between constituent and constituted power depends not on a choice between different abstract constitutional theories but on the political history and legal culture of different constitutional systems. Next, Luisa Fernanda García López turns to the origin of the theory of constituent power. She discusses the many contributions Emmanuel Sieyès has made to modern constitutionalism in her essay entitled ‘Sieyès: The Spirit of Constitutional Democracy?’ From eradicating royal prerogatives to shaping the precursor of the

16  Richard Albert inseparable relationship between national sovereignty and representative government, Sieyes’ contributions, she notes, were vital to modern constitutional theory. As García López writes, Sieyes’ understanding of the Nation institutionalised public law as a transfer of sovereignty from the people to the state. From here, she writes, the separation of powers followed as well as the primacy of law and the legislative power over executive and judicial powers. Then, in ‘Revolutionary Reform in Venezuela: Electoral Rules and Historical ­Narratives in the Creation of the 1999 Constitution’, Joshua Braver turns away from the traditional reliance on Carl Schmitt and towards Hannah Arendt to recover from the Venezuelan story an alternative vision of new beginnings called ­‘revolutionary reform’. Braver suggests that, like Schmitt, Arendt and revolutionary reformists celebrate momentous breaks, but they are partial and justified as necessary to realise the potential of old ideals. In Venezuela, as Braver explains, Chávez’s achievement of first-past-the-post electoral rules for the constituent assembly was the turning point that allowed him to eliminate the opposition through the purging of all other institutions. Braver argues that Chávez’s rule embodied a radical vision of revolution in which exclusion is necessary to completely rupture with all past Venezuelan history. Lacking its own vision of revolution, the Supreme Court struggled to contest the content of the electoral rule. Braver develops an alternative revolutionary reformist narrative of Venezuelan history that might have supported a more inclusive electoral rule. Juliano Zaiden Benvindo comments on Braver’s chapter in his own on ‘“Revolutionary Reform” and the Seduction of Constitutionalism’. He argues that constitutionalism may be inherently appealing but that it is limited by the historically contingent realities of social life. Constitutionalism, he writes, must face the reality of countries historically marked by semi-authoritarian practices. In this complex scenario, he explains, the dilemmas and paradoxes of constitutionalism are pushed to their very extremes. Zaiden Benvindo agrees with Braver’s understanding of how Venezuela lost sight of constitutionalism but suggests that any analysis of Venezuela, in order to be complete, must account for the nation’s challenges of economic development. The second part of the book focuses on the traditions of the constitutional amendment. It includes ten country-specific or regional comparative studies on various subjects related to constitutional amendment. We cover vast ground: Africa, Australia, Canada, the Caribbean, France, Ireland, the Netherlands, the United States, as well as important inquiries into the temporal dimensions of constitutional change and the frequency and difficulty of constitutional amendment. We begin with Sofia Ranchordás’s exploration of contingent constitutional change in her chapter on ‘Constitutional Sunrise’. Her inquiry probes constitutional changes that are contingent upon the verification of a legal or factual condition. She begins by explaining that sunrise clauses are forward-looking instruments that do not take effect until a later event or fact materialises. Ranchordás defines sunrise clauses in broad terms that encompass different types of contingent provisions, ranging from the delay of the coming-into-force of a provision to sunrise clauses that ‘tie’ the coming into effect of a provision to the enactment of statute. She draws from the

Introduction 17 Belgian, Indian, Irish and United States Constitutions to inquire into the legitimacy of sunrise clauses. We then move to Oran Doyle and David Kenny’s co-authored chapter on ‘Constitutional Change and Interest Group Politics: Ireland’s Children’s Rights Referendum’. Their chapter analyses the 2012 children’s rights amendment in Ireland, which, according to its proponents, addressed problems that had emerged in Irish constitutional law where ‘parents’ rights’ were held to be superior to ‘children’s rights’. Doyle and Kenny draw from the constitutional referendum on this amendment proposal to explore how processes of formal constitutional amendment can be used to alter intricate matters of constitutional law doctrine, and to illustrate the difficulties that these attempts encounter. They argue that the calls for this reform were based on a misunderstanding of this area of constitutional law, and that the changes wrought by the amendment were minimal. Despite this, they observe, interest groups that had called for the amendment insisted that it was a major change, leading to a confusing referendum campaign that failed to engage the public and resulted in the referendum passing by a surprisingly narrow margin. They highlight several lessons from this amendment experience—lessons relating to the difficulties of amending constitutions to achieve changes in legal doctrine, to the role of legal academics in highlighting constitutional problems and correcting misunderstandings of constitutional issues, and to the underexplored role of interest groups in effecting constitutional change. In ‘Amendment-Metrics: The Good, the Bad and the Frequently Amended Constitution’, Xenophon Contiades and Alkmene Fotiadou, my two co-editors for this volume, ask an important question that constitutional designers ought to consider when they build constitutions: can we identify a substandard constitution on the basis of its amendment rate? They note that the scholarly literature connects long constitutions with a higher amendment rate and therefore less stability. Their chapter evaluates this claim and suggests why a high amendment rate need not necessarily be associated with a substandard constitution in terms of the quality of its design. They moreover offer criteria for evaluating constitutional quality. Then, James Fleming comments on Contiades and Fotiadou’s chapter. He assesses their critique of the argument that long, frequently amended constitutions tend to be bad constitutions. He also critiques Contiades and Fotiadou’s analysis of the purposes of amendment, arguing that most amendments, in some way, aim to respond to imperfections or to correct flaws in existing constitutions. Fleming furthermore draws on the analysis of John Marshall to sketch some general criteria on what a good constitution might look like: it should be a ‘great outline’, not a detailed legal code; it should be difficult to amend, and it should not be amended frequently. ­Fleming closes his chapter with a final point: that a constitution, in order to be a good one, should serve as ‘basic law’, ‘higher law’, and ‘our law’. The next two chapters develop the important tension between constitutional text and political practice discussed above, with a focus on Australia and Canada, respectively. In ‘Constituting “the People”: The Paradoxical Place of the Formal Amendment Procedure in Australian Constitutionalism’, Weis inquires whether and how popular sovereignty is recognised and exercised in Australia. On one hand, she

18  Richard Albert writes, the idea that the use of the country’s formal amendment rules reflects the will of people is attractive for a tradition that places primacy on the constitutional text. On the other, she observes that the practice of constitutional change in Australia makes it difficult to conclude that the country’s formal amendment rules can fulfill this critical normative function of reflecting the consent of the governed. Next, in ‘Hard Amendment Cases in Canada’, Glover lays out the extraordinarily complex formal amendment rules in the Canadian Constitution and walks readers through a controversial hypothetical constitutional change of Canada—an example that highlights the uncertainty inherent in the structure of Canada’s formal amendment rules. For the example she chooses—mandatory bilingualism for Supreme Court judges—a plain reading of the constitutional text could suggest that Parliament has the authority to impose this requirement by ordinary law alone. Yet Glover offers a deeper contextual reading of the Constitution’s formal amendment rules and concludes that Parliament cannot act unilaterally to impose such an amendment because the magnitude and nature of the change it would affect amounts to a constitutional amendment in relation to the Court’s composition. Returning to the theme of constitutional flexibility and rigidity, Derek O’Brien introduces us to the Caribbean in his chapter on ‘Formal Amendment Rules and Constitutional Endurance: The Strange Case of the Commonwealth Caribbean’. O’Brien raises a paradox: the Constitutions of the Commonwealth Caribbean have exceeded the average lifespan for democratic constitutions but not because their formal amendment rules promote flexibility; it is instead because the formal amendment rules have themselves been a barrier to constitutional amendment. What is more, explains O’Brien, despite the availability of readily deployable formal amendment rules, the constitutional culture in the region is resistant to constitutional reform. Moving east from the Caribbean to France, Jean-Philippe Derosier investigates the idea of the people in his chapter on ‘The French People’s Role in Amending the ­Constitution: A French Constitutional Analysis from a Pure Legal Perspective’. Derosier suggests two ways to understand the body that is the people: the social or political people, on one hand, and the legal people on the other. Returning to the theory of constituent power that began the volume, Derosier classifies the social/ political people as the collectivity that exercises constituent power in a democracy. However, he writes, the legal people cannot act outside of the confines of the law. He draws from a controversial judgment of the French Conseil constitutionnel that reviewed the constitutionality of a national referendum to develop this distinction and to evaluate its implications for constitutional change. We next travel to Africa. In his chapter on ‘The Implication of Conflation of Normal and Constitutional Politics on Constitutional Change in Africa’, Duncan Okubasu exposes us to the remarkable fact that virtually all countries in Africa have changed their constitution in the last decade. He shows that most constitutions have been significantly altered for what he considers insignificant reasons, namely to secure an immediate political advantage. He also shows that major changes in political leadership are often if not always accompanied by constitutional change. Okubasu offers an explanation why: normal and constitutional politics have become indistinguishable, and this has rendered constitutions largely irrelevant and it has in turn negatively impacted the stability of constitutions in Africa.

Introduction 19 The final chapter in this second part is Jurgen Goossens’ ‘Direct Democracy and Constitutional Change in the US: Institutional Learning from State Laboratories’. Goosens begins with an observation: the paths to constitutional change in Article V of the United States Constitution involve only legislatures and do not formally authorise any form of direct democracy. In contrast, he notes, the constitutions of the states reflect a strong tradition of direct democracy and majoritarian voting rules. Why this difference, he asks? Goosens suggests that the United States Constitution stands to learn from state constitutional experience. For three reasons, he argues, the United States Constitution should be reformed to include mechanisms of direct democracy: first, the core of the American tradition of constitutional ­amendment is rooted in popular sovereignty; second, a non-exclusive reading of Article V makes it possible that referendal change in the United States would be both valid and legitimate; and third, evidence from state practice is persuasive evidence of the transferability of direct democratic mechanisms of change to the federal level. The volume closes with a conclusion by Contiades and Fotiadou. Entitled ‘The Emergence of Comparative Constitutional Amendment as a New Discipline: Towards a Paradigm Shift’, Contiades and Fotiadou explore in this concluding ­chapter how the concept of constitutional amendment has evolved through the years to encompass a matrix of notions, theories and practices. They argue that comparative constitutional amendment has become a distinct field, and has provided a conceptual lens through which to understand constitutions, constitutional law and constitutionalism more generally. Their chapter doubles invaluably as a forward-looking research agenda for the field in the years ahead. B.  The Future of Constitutional Amendment The chapters in this volume cover new ground and chart new lines of inquiry that augur well for the future of the study of constitutional amendment and also for the design of constitutional amendment rules. We as contributors and editors very much hope that the field of comparative constitutional amendment continues to attract and sustain the interest of scholars of public law. The very best scholarship in the field must be attentive equally to formal and informal procedures of constitutional change, how the two interact, and also to the costs and consequences of privileging one over the other. We have begun in this volume by taking the first step of probing the traditions and foundations of formal amendment, recognising that much more work remains to be done on informal amendment and how it may be driven, preempted and in any case is deeply connected to formal amendment. Many of the chapters already explore these interconnections. We hope that future scholarship in the field will take our invitation to describe and theorise them even more ­concretely. For now, we take encouragement from the growing interest in the field of ­comparative constitutional amendment, and we look forward to learning from and also contributing to the new scholarship that follows from this book.

20 

Part I

The Foundations of Constitutional Amendment

22 

1 Amendment Power, Constituent Power, and Popular Sovereignty Linking Unamendability and Amendment Procedures YANIV ROZNAI*

T

HE THEORY OF constitutional unamendability identifies a fundamental distinction between primary constituent (constitution-making) power and secondary constituent (constitution-amending) power. The latter is limited by unamendability and the former—perceived as the people’s democratic constitutionmaking power—is unlimited by unamendability. This chapter develops the distinction by supplementing it with a further one, between various shades of secondary constituent powers along a ‘spectrum’; a theoretical construct that links constitutional amendment procedures and limitations which ought to be imposed upon constitutional amendment powers. According to this spectrum theory, constitutional systems are polymorphic: the more similar the democratic characteristics of the amendment powers are to those of the primary constituent power, the less it should be bound by limitations; and vice versa: the closer it is to a regular legislative power, the more it should be fully bound by limitations. This examination is an important step towards a better understanding of the nature and scope of constitutional amendment powers.

*  An earlier version of this chapter was presented at the IACL-BC Workshop on Comparative Constitutional Amendment at Boston College Law School on 15 May 2015. I would like to thank the organisers of the workshop—Richard Albert, Xenophon Contiades and Alkmene Fotiadou for inviting me to present the paper and the participants for their valuable comments, especially to Zoran Oklopcic for his insightful discussion. I would also like to thank Aharon Barak, Andreas Kalyvas, Conor Gearty, Gábor Halmai, Grégoire Webber, Kim Lane Scheppele, Lawrence Sager and Oran Doyle for helpful conversations and Alexander Somek, Bartosz Marchiniak, Martin Loughlin, Michael Wilkinson, Neil Walker, Richard Kay, Thomas Poole and the anonymous reviewer for useful remarks on earlier drafts. An elaborated argument appears in Yaniv Roznai, Unconstitutional Constitutional Amendments—The Limits of Amendment Powers (Oxford University Press, 2017).

24  Yaniv Roznai I. INTRODUCTION

In earlier writings, I have claimed that constitutional amendment powers are increasingly conceived as limited (explicitly or implicitly) and that this unamendability rests on the theoretical distinction between primary and secondary constituent powers.1 This chapter makes an additional move forward towards a theory of constitutional unamendability. The first section summarises the presupposition that the people are free to change even unamendable elements of the Constitution; yet this power resides not in the amendment power but in the exercise of the primary constituent power, a concept strongly linked to popular sovereignty. The second section describes and analyses the current trend of prescribing a constitutional process for exercising primary constituent power. This method, it is argued, is a fallacy. Following the two previous sections, the third section sketches a spectrum of constitutional amendment powers. It considers whether the amendment power is equally limited in jurisdictions where the amendment process attempts to imitate the re-emergence of the primary constituent power, by incorporating elements such as referendums and constitutional conventions, as in jurisdictions where it is more similar to regular legislative power. It argues that the more similar the characteristics of the secondary constituent power are to those of a democratic primary constituent power (‘demanding amendment power’)—that is, inclusive, deliberative, and time-consuming—the less it should be bound by limitations, and vice versa. This examination links limitations upon amendment powers and the amendments’ procedures. The framework which contextualises this chapter’s approach is both comparative by seeking to identify the character of actual existing constitutional arrangements of constitutional amendments, and theoretical by offering a characteristic explanation of these arrangements and evaluating how they work against their own internal logic.2 II.  UNAMENDABILITY AND CONSTITUENT POWER

A.  A Three-Track Democracy in a Nutshell Once a constitution is constituted and the primary constituent power has accomplished its task, a constitutional organ—the amendment authority—is granted with

1 See, eg, Y Roznai, Unconstitutional Constitutional Amendments—The Limits of Amendment Powers (Oxford University Press, 2017); Y Roznai and S Yolcu, ‘An Unconstitutional Constitutional Amendment—The Turkish Perspective: A comment on the Turkish Constitutional Court’s headscarf decision’ (2012) 10(1) Int’l J Const L 175; Y Roznai, ‘Unconstitutional Constitutional Amendments—The Migration and Success of a Constitutional Idea’ (2013) 61(3) Am J Comp L 657; Y Roznai, ‘The Migration of the Indian Basic Structure Doctrine’ in M Lokendra (ed), Judicial Activism in India—A Festschrift in Honour of Justice V. R. Krishna Iyer (Universal Law Publishing Co, 2012) 240. For a typology of limitations on constitutional amendment powers see O Doyle, ‘Constraints on Constitutional A ­ mendment Powers’, in this volume. 2 See, eg, M Loughlin, ‘Constitutional Theory: A 25th Anniversary Essay’ (2005) 25(2) Oxford ­Journal of Legal Studies 183, 186; S Tierney, Constitutional Referendums—The Theory and Practice of Republican Deliberation (Oxford University Press, 2012) 2.

Amendment Power, Constituent Power, and Popular Sovereignty 25 the legal competence of revising the Constitution.3 What happens to the constituent power? Since constitutions are embedded within the idea of populism—the liberty of people to shape and reshape their society,4 the presupposition is that the people always retain the power to establish and change their constitutional order.5 As Carl Friedrich notes, ‘no matter how elaborate the provisions for an amending power may be, they must never … be assumed to have superseded the constituent power’.6 Therefore, constituted organs, including the amendment process, do not consume the primary constituent power which is neither exhausted nor bound by the existing constitutional limitations—including explicit or implicit unamendability.7 Put differently, the secondary constituent power is limited by unamendability. But as a delegated authority, it can be set aside, just as it can be created. The authorising primary constituent power thus remains in the constitutional background and can re-emerge to take its role.8 Unamendability thus limits the delegated amendment power but it cannot block the primary constituent power—the ‘sovereignty at the back of the Constitution’—from its ability to amend even the basic principles or structure of the constitutional order.9 Consequently, what unamendability means is that certain amendments establishing a ‘new constitution’ or a new ‘constitutional identity’ cannot be achieved through the regular amendment procedure but require a different constituent process.10 Unamendability should therefore not be viewed as blocking all the democratic avenues, but rather merely as proclaiming that one such avenue—the amendment

3  For the classical distinction between constituent and constituted powers see EJ Sieyès, ‘What is the Third Estate?’ in Political Writings (Hackett Publishing Company Inc, 2003) 136. For a contextualisation of Sieyès’ doctrine, see T Pereira, ‘Constituting the Amendment Power: A Framework for Comparative Amendment Law’, in this volume. 4  RD Parker, ‘“Here, the People Rule”: A Constitutional Populist Manifesto’ (1993) 27 Val U L Rev 531, 583. 5  J Wilson, Commentaries on the Constitution of the United States of America (J Debrett, J Johnson and JS Jordan, 1792) 38–39; WW Willoughby, An Examination of the Nature of the State—A Study in Political Philosophy (BiblioBazaar, 2009) 215–18; WT Han, ‘Chain Novels and Amendments Outside Article V: A Literally Solution To a Constitutional Conundrum’ (2010) 33 Hamline L Rev 71, 79. 6 CJ Friedrich, Constitutional Government and Politics—Nature and Development (Harper & Brothers Publishers, 1937) 117. See also T Murphy, ‘Constituent/Constituted Power’ in J Protevi (ed), Edinburgh Dictionary of Continental Philosophy (Yale University Press, 2006) 105. 7  The US Constitution itself was adopted in violation of the Articles of the Confederation. See generally RS Kay, ‘The Illegality of the Constitution’ (1987) 4 Const Comment 57; B Ackerman and N Katyal, ‘Our Unconventional Founding’ (1995) 62 U Chi L Rev 475; J Frank, ‘“Unauthorized Propositions” The Federalist Papers and Constituent Power’ (2007) 37(2–3) Diacritics 103. 8 On the amendment power as a delegated authority see Roznai, Unconstitutional Constitutional Amendments (n 1); AR Amar, ‘Philadelphia Revisited: Amending the Constitution Outside Article V’ (1988) 55 U Chi L Rev 1043, 1054–58; AR Amar, ‘The Consent of the Governed: Constitutional Amendment Outside Article V’ (1994) 94 Col L Rev 457, 458–500. 9 CV Keshavamurthy, Amending Power Under The Indian Constitution—Basic Structure Limitations (Deep & Deep Publications, 1982) 80–1. This approach was accepted, for instance, by the Brazilian Federal Supreme Court, which held that the constitution’s ‘immutable provisions’ impose limits on the secondary constituent power, but do not subordinate the primary constituent power itself. See AZ Melo, ‘A limitação material do poder constituinte derivado’ (2008) 8(1) Revista Mestrado em Direito 31, 48. 10  E Smith, ‘Old and Protected? On the “Supra-Constitutional” Clause in the Constitution of Norway’ (2011) 44(3) Israel L Rev 369, 375–76; SK Guha and M Tundawala, ‘Constitution: Amended it Stands?’ (2008) 1 Nat U Jur Sci L Rev 533, 537.

26  Yaniv Roznai process—is unavailable. In order to legitimately achieve the sought after constitutional change, other procedures ought to be used.11 The power to change unamendable principles does not reside within the constitutional amendment procedure rather it is appropriately part of the primary constituent power of the sovereign people, from which all legitimate authority springs.12 And so, within a constitutional democracy, one has to recognise three kinds of powers: primary constituent power, secondary constituent power (amendment power), and constituted power (eg legislative power).13 The legislative track is the sub-constitutional track of political life in which the legislature and executive bodies enact, enforce, and implement political decisions through ordinary legislation. The amendment track is the track of constitutional politics, through which the bodies entrusted with the authority to amend the Constitution, may enact, add, annul, or amend constitutional provisions. This track usually involves different—and separate—processes and organs apart from the legislative track. Finally, the primary constituent track is the third track of democracy which is not part of the ordinary constitutional politics. It is unbound by prior constitutional rules and may even create a new constitutional order. Hence, primary and secondary constituent powers are related but distinct powers, as the latter is a power established in the Constitution, which is bound by explicit and implicit limitations.14 The Constitution cannot restrict the primary constituent power, which resides outside of it and can ‘exercise its authority de novo’.15 B.  Primary Constituent Power and Popular Sovereignty The understanding of the people’s primary constituent power to constitute and re-constitute the Constitution is compatible with commitment to democratic selfgovernment. The people conceive themselves as a single sovereign in order to

11  V Jackson, ‘Unconstitutional Constitutional Amendments: A Window into Constitutional Theory and Transnational Constitutionalism’ in A Wallrabenstein, P Dann and M Bäuerle (eds), DemokratiePerspektiven, Festschrift für Brun-Otto Bryde zum 70 (Mohr Siebeck, 2013) 47. 12 AW Machen, ‘Is The Fifteenth Amendment Void?’ (1909–1910) 23 Harv L Rev 169, 170; WF Harris II, The Interpretable Constitution (Johns Hopkins University Press, 1993) 193. Of course, the assumption that constituent power resides in ‘the people’ carries a contingent character as it is the consequence of a certain political belief which may change in different periods and places. 13  cf S Weintal, ‘The Challenge of Reconciling Constitutional Eternity Clauses with Popular Sovereignty: Toward Three-Track Democracy in Israel As a Universal Holistic Constitutional System and Theory’ (2011) 44(3) Israel L Rev 449; A Kalyvas, ‘Carl Schmitt and The Three Moments of Democracy’ (1999–2000) 21 Cardozo L Rev 1525; S Prateek, ‘Today’s Promise, Tomorrow’s Constitution: “Basic Structure”, Constitutional Transformations and The Future of Political Progress in India’ (2008) Nat U Jur Sci L Rev 417, 458–61. 14  See Roznai, Unconstitutional Constitutional Amendments (n 1); U Baxi, ‘Some Reflections on the Nature of Constituent Power’ in R Dhavan and A Jacov (eds), Indian Constitution—Trends and Issues (NM Tripathi Private Ltd, 1978) 122, 136. 15 C Saunders, ‘The Constitutional Credentials of State Constitutions’ (2011) 42 Rutgers L J 853, 870. Of course, this idea of free and unbound constituent power carries its own dangers, as identified by H Arendt, On Revolution (Penguin Group, 1965) 163. See the debate in J Braver, ‘Revolutionary Reform in Venezuela: Electoral Rules and Historical Narratives in the Creation of the 1999 Constitution’ in this volume; JZ Benvindo, ‘“Revolutionary Reform” and the Seduction of Constitutionalism’ in this volume.

Amendment Power, Constituent Power, and Popular Sovereignty 27 a­ ttribute the Constitution to a single act of will.16 However, the very idea of the people as holders of the primary constituent power is perplexing and has given rise to heated debates within constitutional theory.17 The idea of a ‘power of the people’ acting in an identifiable, comprehensible, and unmediated way in order to constitute for themselves a constitution is a kind of a myth, a fiction.18 Even if ‘the people’ are recognised as holders of the primary constituent power, the people do not necessarily take part in the constitution-making process,19 and with old constitutions there is no ‘we the people’ as the constitution-makers are long gone.20 Since ‘the people’ simply cannot speak directly as a whole, the will that we attribute to ‘the people’ ought to be revealed through some kind of representation and deliberation processes.21 For the theory of unamendability, regardless of how historically accurate is the story we tell ourselves about ‘the people’ as constitution-makers, ‘the very plenipotentiary scope of the people as the normative constitutional author provides the limits on the normal amending apparatus’.22 I do not elaborate here on the various complexes of the constituent power concept, an exercise which deserves a separate treatment of its own.23 I only briefly point to the relationship between constituent power and popular sovereignty which is necessary for the further development of the chapter. ‘The dirty little secret’ of contemporary jurisprudence, as Roberto Unger describes it, is the discomfort with democracy and fear of popular action.24 Popular sovereignty’s understanding of the primary constituent power challenges this discomfort as it allows the people’s primary constituent power to re-emerge in ‘extraordinary moments, [when] politics opens up to make room for conscious popular p ­ articipation

16 M Loughlin, Foundations of Public Law (Oxford University Press, 2010) 224; S Levinson, Constitutional Faith, 2nd edn (Princeton University Press, 2011) 193–94. 17  See, eg, M Loughlin and N Walker (eds), The Paradox of Constitutionalism: Constituent Power and Constitutional Form (Oxford University Press, 2007); Z Oklopcic, ‘Three Arenas of Struggle: A Contextual Approach to the Constituent Power of “the People”’ (2014) 3(2) Glob Const 200; Z Oklopcic, ‘Constitutional Theory and Cognitive Estrangement: Beyond Revolutions, Amendments and Constitutional Moments’ in this volume. 18  See debates, eg in ES Morgan, Inventing The People: The Rise of Popular Sovereignty in England and America (WW Norton, 1988) 14, 58; B Darby ‘Amending Authors and Constitutional Discourse’ (2002) 25(2) Dalhousie L J 215, 221; FI Michelman, ‘Constitutional Authorship By The People’ (1998–1999) 74 Notre Dame L Rev 1605, 1628; E Oxman, The Constitution of the People (PhD Dissertation, Princeton University, 2009); Benvindo (n 15). 19  DJ Galligan, ‘The Sovereignty Deficit of Modern Constitutions’ (2013) 33 O J Leg Stud 1, 9–11. 20 DA Strauss, ‘We the People, They the People, and the Puzzle of Democratic Constitutionalism’ (2013) 91 Texas L Rev 1969. 21  RS Kay, ‘Constituent Authority’ (2011) 59 Am J Comp L 715, 739; See also Jackson (n 11) 65. Of course, the relationship between constitutional law and popular sovereignty differs between different constitutional systems which have diverse relationship with ‘the people’ in their constitutional design and constitution-making. See Pereira (n 3). 22  Harris (n 12) 193. 23  See generally M Loughlin, ‘The Concept of Constituent Power’ (2014) 13(2) European Journal of Political Theory 218; JI Colón-Ríos, ‘Five Conceptions of Constituent Power’ (2014) 130 Law Quarterly Review 306; M Spång, Constituent Power and Constitutional Order (Palgrave Macmillan, 2014); Y Roznai, ‘“We the People”, “Qui, the People” and the Collective Body: Perceptions of Constituent Power’ in G Jacobsohn and M Schor (eds), Comparative Constitutional Theory (Edward Elgar, 2017, forthcoming). 24  RM Unger, What Should Legal Analysis Become? (Verso, 1996) 72.

28  Yaniv Roznai and extra-institutional, spontaneous, collective intervention’.25 Constituent power, Lior Barshack writes, belongs to the transcendent and absent corporate body. ‘The people’, as the sovereign in a democracy, remains a transcendent entity, which appears only during special constitutional moments such as declarations of independence, revolutions, and referenda. In these instances of communitas, the communal body is dissolved of structural boundaries as it descends back into the social; the transcendence is collapsed into the immediate presence.26 These extraordinary moments, which carry a temporary character, define the corporate group and within them all generations are present.27 Importantly, for Barshack, ‘it is not the democratic principle of popular sovereignty that places popular will above constitutional procedure, but the fact of sovereign incarnation and the concomitant relaxation of all principles. The communal body wields supreme legislative power whenever it is enacted, in democratic as well as non-democratic contexts…’.28 It is the suspension of political parties and a high degree of communal body’s involvement that allows sovereignty to step forward in constitutional moments such as referendums.29 In contrast with this transcendental conception, according to Akhil Amar, through Article V, the people, who are the source of the Constitution, delegated the amendment power to ordinary government, without limiting themselves. Article V thus supplements but does not replace popular sovereignty. It is non-exclusive as the immanent people retain their reserved and inalienable right to revise the Constitution themselves, by a majority of voters, via referendum or special constituent convention, even outside of the amendment process.30 These conceptions resemble one another in one important sense: in Barshack’s conception of a corporate body, the transcendence collapses into the immanent present in extraordinary temporary constitutional moments. Likewise, Amar’s immanent conception of popular sovereignty is manifested in special constitutional moments. It is therefore not necessary, for the sake of the argument advanced here, to choose between these two conceptions. Yet, both accounts correspond with Sheldon Wolin’s conception of democracy as a political practice that involves the manifestation of popular sovereignty. Democracy, Wolin advocates, is not a ‘form’, but rather a representative moment in the nation’s life that dictates the substance of the Constitution. In these rare and episodic ‘moments of commonality … through public deliberations, collective power is used to promote or protect the well-being

25 A Kalyvas, Democracy and the Politics of the Extraordinary: Max Weber, Carl Schmitt, and Hannah Arendt (Cambridge, Cambridge University Press, 2008) 7. 26 L Barshack, ‘The subject of ideals’ (2000) 4(1) Cultural Values 77, 82; L Barshack, ‘Notes on The Clerical Body of The Law’ (2003) 24(3) Cardozo L Rev 1151, 1155, 1164; L Barshack, ‘Political Theology and the Authenticity of Modern Experience’ (2012) 5 Jerusalem Rev Leg Stud 12, 19. 27  L Barshack, ‘Time and the Constitution’ (2009) 7(4) Int’l J Const L 551, 553, 557, 566–67. On time and constitution-making/change see S Ranchordás, ‘Constitutional Sunrise’, in this volume. 28 L Barshack, ‘Constituent Power as Body: Outline of a Constitutional Theology’ (2006) 56 Uni Toronto L J 185, 202. 29  ibid 212–13. 30  AR Amar, ‘Of Sovereignty and Federalism’ (1987) 96 Yale L J 1425; Amar, ‘Philadelphia Revisited’ (n 8) 1054–58; AR Amar, ‘The Consent of the Governed’ (n 8) 457–500; AR Amar, ‘Popular Sovereignty and Constitutional Amendment’ in S Levinson (ed), Responding to Imperfection: The Theory and Practice of Constitutional Amendment (Princeton University Press, 1995) 89, 89–101.

Amendment Power, Constituent Power, and Popular Sovereignty 29 of the collectivity … power returns to “the Community” and agency to “the People”’.31 If we believe that the source of political authority rests with the people; that ‘the ability to engage in constitutional change is a fundamental act of popular sovereignty’,32 we also need to bridge the gap between the imaginary people and the real people. ‘Constitutionalism’, Carl Friedrich writes, ‘and more especially democracy, presupposes an active group of citizens who are ready to assume responsibility and become the “constituent power”’.33 Therefore, recent scholarship has called for the development of democratic tools in order to ‘return the epicenter of sovereignty to the people’, especially in constitutional decisions.34 Ali Riza Coban calls for the design of legal arrangements that would ensure a maximum level of the people’s democratic participation during constitution-making.35 Likewise, Joel Colón-Ríos argues that significant constitutional changes require a process that endeavours to reproduce and maximise popular participation and democratic openness at the same degree presented during constitution-making moments.36 Therefore, he claims, in order to acquire democratic legitimacy, fundamental constitutional changes, episodical by their nature, should take place through the most participatory process possible, which allows citizens the opportunity to propose, deliberate, and decide upon such changes. Participatory mechanisms such as elected constituent assemblies whose outcomes shall be approved in a popular referendum facilitate the exercise of constituent power.37 This understanding allows the mythical people to re-activate its constituent power.38 Indeed, the modern conception of primary constituent power is strongly associated with the notion of popular sovereignty and the recent proliferation of referendums is an indicator of the broader trend towards engaging the people themselves in constitutional matters.39 According to studies of late, among current existing

31 S Wolin, ‘Fugitive Democracy’ in S Benhabib (ed), Democracy and Difference: Contesting the Boundaries of the Political (Princeton University Press, 1996) 31, 39, 41. See also S Wolin, Politics and Vision (Princeton University Press, 2004) 602–03. 32  M Schwartzberg, Democracy and Legal Change (Cambridge, Cambridge University Press, 2009) 6. 33  CJ Friedrich ‘Military Government and Dictatorship’ (1950) 267 Annals of the American Academy of Political and Social Science 1, 4. 34  AAI Ninet, ‘From Popular Sovereignty to Constitutional Sovereignty?’ (2010) Buffalo Legal Studies Research Paper 2011-007 accessed 1 August 2016. 35 AR Coban, ‘The Continuity Problem of the Constituent Power’ in A Scherzberg, O Can and I Doğan (eds), Verfassungstheorie und Verfassungsgebung: Überlegungen anlässlich der Diskussion um eine Verfassungsreform in der Türkei (LIT Verlag, 2012) 49, 56, n 13. 36  J Colón-Ríos, ‘The Second Dimension of Democracy: The People and Their Constitution’ (2009) 2(2) Baltic J L & Pol 1. 37  J Colón-Ríos, Weak Constitutionalism—Democratic Legitimacy and the Question of Constituent Power (Routledge, 2012); Colón-Ríos, ‘The Legitimacy of the Juridical: Constituent Power, Democracy, and the Limits of Constitutional Reform’ (2010) 48(2) Osgoode Hall L J 199; Colón-Ríos, ‘Notes on Democracy and Constitution-Making’ (2011) 9(1) The New Zealand J Pub & Int’l L 17; Colón-Ríos, ‘The Counter-Majoritarian Difficulty and the Road Not Taken: Democratizing Amendment Rules’ (2012) Canadian J L Juris 53. See also AC Hutchinson and J Colón-Ríos, ‘Democracy and Constitutional Change’ (2011) Theoria 43, 52–53. 38  Colón-Ríos, ‘The Legitimacy of the Juridical’ (n 37) 240. 39  S Tierney, ‘Whose Political Constitution? Citizens and Referendums’ (2013) 14(12) German L J 2185, 2194; Tierney ‘Sovereignty and Crimea: How Referendum Democracy Complicates Constituent Power in Multinational Societies’ (2015) 16(3) German L J 523, 525.

30  Yaniv Roznai c­ onstitutions more than 40 per cent were publicly ratified by referendums and many others involved different forms of popular participation in the constitution-making process.40 As Cheryl Saunders notes, ‘popular participation is a distinctive feature of constitution-making in the 21st century’.41 There are many familiar difficulties associated with popular mechanism such as referendums: the designation of the individuals who qualify to participate; the drafting of the ballot question; the lack of knowledge of the voters; fear of tyranny of the majority; and the historical associations of abuse of plebiscites.42 Moreover, referendums do not necessarily truly express the people’s will due to political elites or interest groups’ manipulations of the people or control of the political agenda which cast doubts on the ability of referendums to allow a genuine exercise of deliberative democracy.43 Therefore, some, such as Antoni Ninet, claim that: ‘the legitimacy and validity of the Constitution requires not only popular ratification, but also real (or true) democratic involvement. A constitution made through ordinary parliaments and representatives is unacceptable’.44 In other words, popular participation in ‘constitutional moments’ should be throughout the constitutional norms-creating process and not be limited to a solely ‘yes’ or ‘no’ vote in a referendum.45 This understanding refuses to reduce primary constituent power to a mere acclamation— a ‘soccer-stadium democracy’, in the words of Holmes.46 This resembles Christoph Burchard’s criticism of Schmitt’s conception of constituent power for ‘there is no discourse, no rational consideration, only irrational masses cheering or booing’.47 Process matters.48 It is the manifestation of ‘we the people’, not simply ‘oui the people’.

40  See, eg, J Blount, ‘Participation in Constitutional Design’ in T Ginsburg and R Dixon (eds), Comparative Constitutional Law (Edward Elgar, 2011) 49; Z Elkins, T Ginsburg, and J Blount, ‘The Citizen as Founder; Public Participation in Constitutional Approval’ (2008) 81 Temple L Rev 361. See also M Versteeg, ‘Unpopular Constitutionalism’ (2014) 89(3) Indiana L J 1133, 1142 (‘today, 34% of all constitutions require ratification by popular referendum, while in 1950, only 7% did’). 41  C Saunders, ‘Constitution Making in the 21st Century’ (2012) 4 Int’l Rev L. See also Kay (n 21) 746; JA Lenowitz, ‘Why Ratification? Questioning the Unexamined Constitution-making Procedure’ (PhD Thesis, Columbia University, 2013) 119. 42  D Butler and A Ranney, ‘Theory’ in D Butler and A Ranney (eds), Referendums Around the World: The Growing Use of Direct Democracy (DC, AEI 1994) 11, 17–21; Friedrich (n 6) 115–17. 43  For a discussion of interest groups’ roles in referenda see O Doyle and D Kenny, ‘Constitutional Change and Interest Group Politics: Ireland’s Children’s Rights Referendum’, in this volume. I thank Mark Tushnet for emphasising that due to elite’s control of referendums parliamentary setting is more deliberative than referendums. Indeed, referendums are traditionally considered as ‘elite driven’. See, eg, A Lijphart, Democracies, Patterns of Majoritarian and Consensus Government in Twenty-One Countries (Yale University Press, 1984) 203. However, there are examples of failed referendums in which governments were unable to secure the outcome of referendums they have initiated. See D O’Brien, ‘Formal Amendment Rules and Constitutional Endurance: the strange case of the Commonwealth ­Caribbean’, in this volume. 44  A Ninet, Constitutional Violence: Legitimacy, Democracy and Human Rights (Edinburgh University Press, 2013) 14. 45  Hutchinson and Colón-Ríos (n 37) 53; Coban (n 35) 73. 46  S Holmes, The Anatomy of Antiliberalism (Harvard University Press, 1996) 49. 47  C Burchard, ‘Carl Schmitt on Democracy and International Relations’ (2006) 19 Leiden J Int’l L 9, 13. 48  See R Levy, ‘“Deliberative Voting”: Reforming Constitutional Referendum Democracy’ (2013) Pub L 555, 566 (proposes to improve deliberative voting in constitutional referenda); EJ Leib, ‘Can Direct Democracy Be Made Deliberative?’ (2006–2007) 54 Buffalo L Rev 903, 911–12 (‘deliberative democracy offers a potential blueprint that can help us devise a better way to undertake direct democracy’);

Amendment Power, Constituent Power, and Popular Sovereignty 31 A democratic primary constituent power must be committed to popular sovereignty which may exercise itself in forms such as special constitutional assemblies and constitutional referenda.49 These forms must have a special character, that is, separate from other public functions, thereby replacing revolution with peaceful means incorporating actual, deliberate, free choice by society’s members.50 While it is true that ‘in the end, there can be no precise algorithm specifying the conditions for defining a people capable of exercising constituent authority’,51 an exercise of primary constituent power should be inclusive, participatory, and deliberative.52 After all, the word constituere, Kalyvas reminds us, marks the act of founding together, jointly.53 III.  THE CONSTITUTIONALISATION OF PRIMARY CONSTITUENT POWER

A.  The Fallacy of Prescribed Constitution-Making Procedures In the previous section, I noted that even after the establishment of a constitution, the primary constituent power, the ‘sleeping giant’ so to speak,54 can wake-up and be exercised. Being external to the pre-existing constitutional order, this power cannot be bound by the prior and existing constitutional rules. Consequently, constitutions cannot (and most do not) regulate its emergence. Constitutions, in other words, ‘contemplate their amendment but almost never their replacement’.55 Nonetheless, there is often a desire to institutionalise the primary constituent power within the

S Tierney, ‘Constitutional Referendums: A Theoretical Enquiry’ (2009) 72(3) Modern L Rev 360, 382 (it ‘seems intuitively plausible that a referendum, carefully tailored to meet the specificities of a particular society, can help bring a populace together in a deliberative, constitutional moment’.) 49  cf Barshack (n 28) 190; B Ackerman, ‘The New Separation of Powers’ (2000) 113(3) Harv L Rev 633, 665–67. 50 D Conrad, ‘Limitation of Amendment Procedures and the Constituent Power’ (1970) 15–16 Ind Y B Int’l Aff 347, 404–10; Prateek (n 13) 454. 51  Kay (n 21) 742. Since there is no clear mechanism for the exercise of primary constituent power, the decision whether a constitutional norm was indeed a manifestation of the primary constituent power can only be given retrospectively, such as by a judiciary’s legitimation of the extra-constitutional or the recognition of the governed. See M Schwartzberg, ‘Should Progressive Constitutionalism Embrace Popular Constitutionalism? (2011) 72(6) Ohio State L J 1295, 1303, 1314. 52  An important aspect is therefore the maintenance of freedoms such as freedom of speech, free and fair election, freedom from arbitrary arrest, and freedom of assembly and association. See Guha and Tundawala (n 10) 543; D Conrad, ‘Constituent Power, Amendment and Basic Structure of the Constitution: A Critical Reconsideration’ (1977–1978) 6–7 Delhi L Rev 1, 12; K Chryssogonos, ‘Popular Involvement in Constitution-Making’ (2008) 20(4) Revue Européenne de Droit Public 1299–1316; Friedrich (n 6) 116. 53  A Kalyvas, ‘Popular Sovereignty, Democracy, and the Constituent Power’ (2005) 12(2) Constellations 223, 235. 54  I borrow this term from H Cyr, Canadian Federalism and Treaty Powers: Existential Communities, Functional Regimes and the Canadian Constitution (LL.D Thesis, Université de Montreal, 2007) 45, n 74. 55  D Landau, ‘The Importance of Constitution-Making’ (2011–2012) 89 Denver University L Rev 611, 616–17. See also C Winterhoff, Verfassung—Verfassunggebung—Verfassungsänderung. Zur ­Theorie der Verfassung und der Verfassungsrechtserzeugung (Mohr Siebeck, 2007) 150–51; Kay (n 21) 745.

32  Yaniv Roznai constitutional form and some constitutions attempt to regulate the re-emergence of the primary constituent power.56 For instance, the Constitution of Venezuela of 1999 states that ‘The original constituent power rests with the people of Venezuela. This power may be exercised by calling a National Constituent Assembly for the purpose of transforming the State, creating a new juridical order and drawing up a new Constitution’.57 Other constitutions prescribe a procedure for their ‘total reform’, ‘complete revision’, or ‘adopting a new constitution’.58 Since by its nature the primary constituent power is unbound by any constitutional rules, this method, as Schmitt claimed, seems to be a fallacy: ‘no constitutional law, not even a constitution, can confer a constitution-making power and prescribe the form of its initiation’.59 If constituent power is considered superior and external to positive law, law cannot prescribe it. Therefore, Richard Kay claims that these attempts are not only ‘paradoxical’, but might also ‘be dismissed as rhetorical decoration’.60 Instead of their dismissal, it might be more valuable to regard these provisions not as constituting, but rather as recognising or declaring, existing powers; hence different from amending provisions. If a new constitution is constituted through these mechanisms, what are the implications for constitutional theory? Two plausible answers; either this process was an exercise of the instituted secondary constituent power,61 or it was indeed the primary constituent power that played the constitution-making role, yet it simply decided to act according to the existing procedures rather than being obliged by them.62 Carl Friedrich gives the example of the constitutional reform that occurred in Switzerland in 1874 when the Constitution of 1848 was ‘entirely overhauled and democratized’ through the ordinary amendment procedure. Constituent power, according to Friedrich: manifested itself through the amending power; but that does not mean that it is identical with it; in fact even to say that it manifested itself through it is something of a misstatement. It would be more accurate to say that the group which might otherwise develop into the constituent power manifests itself through, acts through the amending power.63

56 Kay (n 21) 725–26; Coban (n 35) 56; M Tushnet, ‘Constitution-Making: An Introduction’ (2012–2013) 91 Texas L Rev 1983, 1988; D Landau and R Dixon, ‘Constraining Constitutional Change’ (2015) 50 Wake Forest L Rev 859. 57  Art 347. Arts 348–349 of the Constitution further regulate this process. 58 See, eg, Argentine Const (1994) Sec 30; Nicaraguan Const (1986) Arts 191–193; Swiss Const (1999), Arts 138, 193, and Bulgarian Const (1991) Arts 158–162. See Kay (n 21) 725; ‘Report on ­Constitutional Amendment’, European Commission for Democracy Through Law (Venice Commission) (Venice, 11–12 December 2009) 13 , accessed 1 August 2016. 59  C Schmitt, Constitutional Theory (J Seitzer tr, Duke University Press, 2008) 132. 60  Kay (n 21) 727–28. 61  See, eg, Kalyvas (n 53) 228; R Guastini, ‘On the Theory of Legal Sources’ (2007) 20(2) Ratio Juris 302, 305: ‘in no legal system can constituent power […] be regarded as an institutionalized source: other­ wise, it would not be “constituent but “constituted.”’ 62  Lenowitz (n 41) 87. 63  Friedrich (n 6) 118.

Amendment Power, Constituent Power, and Popular Sovereignty 33 Since the primary constituent power has extra-juridical dimensions, it cannot be fully regulated or stipulated legally.64 ‘Like it or not’, Kay correctly notes, ‘a true constituent authority must act without the comfort of legal authorization’.65 This, however, does not mean that a constitution cannot stipulate the means by which a new constitution would be constituted. It only means that by its nature, primary constituent power does not have to abide by it, although it can act accordingly if it so wishes.66 These mechanisms can be viewed not as containing primary constituent power, but rather, simply as vehicles for its exercise.67 Not to be mistaken, these mechanisms may carry benefits. Exercising primary constituent power behind ‘a façade of legality’ may serve significant political ­interests,68 and may enhance legal certainty, continuity, and the legitimacy of the new document.69 Therefore, the Venice Commission ‘strongly endorse[d]’ the use of a legal procedure even for the adoption of ‘new constitutions’, as such a procedure would ‘strengthen the stability, legality and legitimacy of the new system’.70 Finally, regulating a relatively demanding constitutional–replacement process, would obviate abusing the amendment process or ignoring any constitutional procedures or guaranties whatsoever in the name of pouvoir constituant.71 Since the people’s will is divided, such legal rules may create ‘organized, complex procedures of deliberation and voting’, which would preserve the primary constituent power’s credibility.72 B.  We The ‘Limited’ People? The German Basic Law of 1949 is an interesting example of an attempt to legally positivise the primary constituent power.73 The final article of the Basic Law anticipated the Basic Law’s own destruction: ‘This Basic Law will lose its validity on the effective date of a constitution that has been chosen by the German people in a free decision’.74 Contrary to some of the examples noted earlier, this provision—

64  R Cristi, ‘The Metaphysics of Constituent Power: Schmitt and the Genesis of Chile’s 1980 Constitution’ (1999–2000) 21 Cardozo L Rev 1749, 1758, 1765; M Goldoni and C McCorkindale, ‘Why we (still) need a revolution’ (2013) 114(12) German L J 2197, 2214–15; PW Kahn, ‘Political Time: ­Sovereignty and The Transtemporal Community’ (2006) 28(1) Cardozo L Rev 259, 268. 65  Kay (n 21) 735. 66  Tushnet (n 56) 2006. 67  cf Tierney (n 2) 133. 68  Kay (n 21) 733. On the use of positive law argument for justifying revolutions, see RS Kay, ‘Legal Rhetoric and Revolutionary Change’ (1997) 7 Caribbean L Rev 161. 69  Coban (n 35) 56. 70  ‘Report on Constitutional Amendment’ (n 58) 15. 71  Of course, such mechanisms themselves are also open for abuses. D Landau, ‘Abusive Constitutionalism’ (2013) 47(1) UC Davis Law Review 189; Landau and Dixon (n 56). 72  Chryssogonos (n 52) 1299–1316. 73  C Möllers, ‘“We are (afraid of) the people”: Constituent Power in German Constitutionalism’, in Loughlin and Walker (eds) (n 17) 87, 97–98. 74  Grundgesetz (1949) Art 146. At the moment of unification, Art 146 was not invoked and the Basic Law was instead amended and was kept in place. See Kay (n 21) 727. In the acts of the reunification of 23.09.1990, Art 146 was amended as follows: ‘after the union and freedom of Germany have been finalized this constitution shall be valid for all the German people and will continue to be valid until the day when a new constitution is accepted by the free will of the German people’.

34  Yaniv Roznai while recognising the people’s constituent power—did not stipulate conditions or procedures for its exercise, which might be seen as a confirmation of its extra-legal character.75 This particular constitutionalisation of primary constituent power raises interesting questions as to the theory of unamendability. Famously, the German Basic Law includes in Article 79(3) an unamendable provision which prohibits amendments affecting the division of the Federation into Länder, human dignity, the constitutional order, or basic institutional principles describing Germany as a democratic and social federal state.76 Is the emergence of a new primary constituent power, constitutionalised by Article 146, restricted by Article 79(3)? Some claim that the unamendable principles would bind future constitution making while others remark that Article 146 is a legal manner with which to overcome the unamendable provision.77 In the Lisbon Case, the Federal Constitutional Court expressly left this question unanswered: It may remain open whether, due to the universal nature of dignity, freedom and equality alone, this commitment even applies to the constituent power, i.e. to the case that the German people, in free self-determination, but in a continuity of legality to the rule of the Basic Law, gives itself a new constitution.78

I agree with Jo Murkens that Article 79(3) addresses only the amendment power, whereas Article 146 foresees a new constitution adopted by the primary constituent power, which by its nature cannot be bound by the rules of the prior constitution. The new constitution-drafter may take Article 79(3) into account, but that would depend on its own ‘goodwill’, rather than on the nature of the unamendable provision as a legal obligation.79 Article 79(3) guarantees are unamendable, not eternal (as they are often wrongfully referred to).80 Consequently, although the primary constituent power is constitutionalised within the German Basic Law, Article 79(3) is unable to bind later generations when exercising their primary constituent power.81 One often noted example to the claim that the exercise of constituent power is hardly bound by limitations, is the 1962 Amendment to the French Constitution over the form of presidential elections which passed through a referendum initiated by President de Gaulle despite its violation of the amendment procedure.82

75 

Coban (n 35) 58. See H Goerlich, ‘Concept of Special Protection for Certain Elements and Principles of the Constitution Against Amendments and Article 79(3), Basic Law of Germany’ (2008) 1 Nat U Jur Sci L Rev 397. 77  See debates in Kay (n 21) 727; Möllers (n 73) 97; JEK Murkens, From Empire to Union: Conceptions of German Constitutional Law since 1871 (Oxford University Press, 2013) 173–75; PE Quint, The Imperfect Union: Constitutional Structures of German Unification (Princeton University Press, 1997) 49. 78  The Lisbon Case, BVerfg, 2 be/08 of 30 June 2009, para 217 accessed 1 August 2016. 79  Murkens (n 77) 175. See also Coban (n 35) 59. 80  See Y Roznai, ‘Unamendability and the Genetic Code of the Constitution’ (2015) 27(2) Eur Rev Pub L 775. 81  Goerlich (n 76) 404; UK Preuss, ‘The Implications of “Eternity Clauses”: The German Experience’ (2011) 44(3) Israel L Rev 429, 440, 443. 82  L Jaume, ‘Constituent Power in France: The Revolution and its Consequences’, in Loughlin and Walker (eds), (n 17) 67, 82–83; J P Derosier, ‘The French People’s Role in Amending The Constitution’, in this volume; Tierney (n 2) 130–36. 76 

Amendment Power, Constituent Power, and Popular Sovereignty 35 The amendment was challenged before the French Constitutional Council, which held that it had no competence to review laws passed by the people in a referendum since they are a direct expression of national sovereignty.83 This reflects the idea of le peuple-roi—‘the people’ is the new sovereign king who always retains the power to revise the Constitution. As one advocate stated in 1849 before a Versailles court, ‘the people never violate the constitution’.84 It is also compatible with Barshack’s supposition that ‘the fuller the sovereign presence, the more relaxed the constitutional structure and the formal procedure that governs the referendum’.85 Since this extra-constitutional change is to be regarded as a ‘constitutional violation’,86 its authoritative legitimacy could only be granted retrospectively.87 This postulation raises a thorny question: what are the implications of an amendment process which includes the primary constituent power’s characteristics of directness and speciality, like popular referendums or elections for special constituent assemblies? Here the people are part of the instituted amendment power. Are the people in that capacity limited? A positive answer would subordinate not only decisions of the people’s representatives, but also those of the people themselves to the judiciary. It seems that there are two prevailing approaches to this challenge.88 According to the first approach, when the amendment power is exercised by the people it is unlimited. This is the general approach of Irish jurisprudence. Due to the Christian character of the 1937 Constitution, there is a heated debate within Ireland on whether natural law sets limits to the amendment power.89 In its jurisprudence, the Irish Supreme Court rejected the claim that natural law is superior to the Constitution, holding that the people, not God, are the creator of the Constitution and the supreme authority. Hence, amendments made by the people in a referendum

83  CC Decision No. 62-20 DC, 6 November 1962 Referendum Act, Rec., 1962, 27. See D Baranger, ‘The Language of Eternity: Constitutional Review of the Amending Power in France (Or the Absence Thereof)’ (2011) 44(3) Israel L Rev 389, 392–94; Derosier (n 82). 84  Cited in F Lieber, On Civil Liberty and Self-Government (JB Lippincott, 1859) 388. 85  Barshack (n 28) 212–13. 86  See P Carrozza,’ Constitutionalism’s Post-Modern Opening’ in Loughlin and Walker (eds), (n 17) 169, 174. Whereas the factual tends to have a normative power, an effective exercise of constituent power does not mean that the extra-constitutional action was constitutional under the previous constitutional rules. See Y Roznai and S Suteu, ‘The Eternal Territory? The Crimean Crisis and Ukraine’s Territorial Integrity as an Unamendable Constitutional Principle’ (2015) 16(3) German L J 542, 569, n 161. Therefore, Derosier (n 82) is correct in his contribution to this volume when he refers to the 1962 amendment as a legal revolution. Nevertheless, one may argue that the employment process may establish a constitutional precedent for a new amendment procedure to be availed of in future occasions. cf B Ackerman, We the People: Transformations (Harvard University Press, 2000) 415. 87  Kalyvas (n 523) 231. 88  K Gözler, Pouvoir constituant (Ekin Press, 1999) 102. 89 See A Kavanagh, ‘Unconstitutional Constitutional Amendments from Irish Free State to Irish Republic’ in E Carolan (ed), The Constitution of Ireland: Perspectives and Prospects (Bloomsbury Professional, 2012) 331; Y Roznai ‘The Theory and Practice of “Supra-Constitutional” Limits on Constitutional Amendments’ (2013) 62(3) Int’l & Comp L Q 557, 566–69; R O’Connell, ‘Guardians of the Constitution: Unconstitutional Constitutional Norms’ (1999) 4 J C L 48, 61–66; A O’Sullivan and PCW Chan, ‘Judicial Review in Ireland and the Relationship Between The Irish Constitution and Natural Law’ (2006) 15 Nott L J 18, 32; O Doyle, ‘Legal Validity: Reflections on the Irish Constitution’ (2003) 25 Dublin University L J 56, 65–67.

36  Yaniv Roznai which expresses their will become the fundamental and supreme law of the land, and cannot be reviewed or nullify by courts.90 According to this approach, even if the primary constituent power rests with the people, albeit essentially anarchic and lawless, it may choose, so to speak, to be exercised within the constitutional framework of constitutional amendment. According to the second approach, the amendment power is limited even when exercised directly by the people. The people in that capacity of inclusion in the amendment process represent a legal organ of the state. Since there can be no sovereign within the constitutional order, the people’s power is necessarily limited.91 An example of this approach comes from Switzerland, where 100,000 people eligible to vote have the right to propose constitutional revisions (Volksinitiative). Following a Volksinitiative, according to which asylum seekers who enter the state unlawfully would be deported immediately and without the option of appeal, the Federal Council proposed that the Federal Assembly invalidate the Volksinitiative, which it did accordingly on 14 March 1996. According to the Federal Council, respect for international law fundamental norms, which include the peremptory norm of non-refoulement, is inherent to the Rechtsstaat principle of ‘rule by law’.92 In 1999, Switzerland granted explicit constitutional recognition to the proposition that international law jus cogens norms limit constitutional reforms, whether total or partial.93 Therefore, in Switzerland, even when the people are directly involved and the Constitution allows for its total revision, this faculty is still limited. According to this approach, ‘the people’ may be regarded in two distinct capacities: as a source of absolute power (primary constituent power) and as a constitutional organ established by the Constitution for its amendment (secondary constituent power).94 Stephen Tierney’s analysis appears lucid. Tierney claims that one has to distinguish between referendums that operate wholly within existing constitutional structures, thereby internal to the normative constitutional order, and referendums that transcend the existing order, which are external to the Constitution and include the power to ‘bring about a new order’.95 Tierney urges us to be cautious when referring

90 See O’Sullivan and Chan ibid; Roznai ibid; Kavanagh ibid; GJ Jacobsohn, ‘An Unconstitutional Constitution? A Comparative Perspective’ (2006) 4(3) Int’l J Const L 460, 469. Doyle (n 1), in this volume, notes that ‘The Irish rule that the Government cannot spend public money supporting just one side of a constitutional referendum is a majoritarian process-constraint, designed to ensure that the real “will of the people” is ascertained’. 91  O Beaud, La puissance de l'état (PUF, 1994) 437–38; O Beaud, ‘Le Souverain’ (1993) 67 Pouvoirs 33, 37. 92  See E de Wet, ‘The Prohibition of Torture as an International Norm of jus cogens and Its Implications for National and Customary Law’ (2004) 15(1) Eur J Int’l L 97, 101–05; A Zimmermann, ‘Is It Really All about Commitment and Diffusion? A Comment on Commitment and Diffusion: How and Why National Constitutions Incorporate International Law’ (2008) U Ill L Rev 253, 258. 93  Swiss Const (1999), Arts 193(4), 194(2). See also G Biaggini, ‘Switzerland’ in D Oliver and C Fusaro (eds), How Constitutions Change—A Comparative Study (Hart Publishing, 2011) 303, 316–17. 94 C Klein, ‘Is There a Need for an Amending Power Theory?’ (1978) 13 Israel L Rev 203, 213. See also Note, ‘The Faith To Change: Reconciling The Oath To Uphold With The Power To Amend’ (1995–1996) 109 Harv L Rev 1747, 1759. 95  Tierney (n 2) 12.

Amendment Power, Constituent Power, and Popular Sovereignty 37 to ‘people sovereignty’ simply due to an exercise of a referendum, and to pay attention to the mode of that exercise and the role that the people have played within it.96 On this account, when the people have a role within the amendment process, such an exercise is ‘constrained to operate within mainstream representative democracy, subordinate to the constitutional rules and subject to constitutional institutions, including courts’.97 It is only when the people ‘act as original constitutional authors, bringing a clear break in the old order; the referendum manifests the “people’s” direct democratic capacity to act as the supreme source of constitutional law in foundational constitutional acts’.98 Within this dilemma, Claude Klein remarks, lies ‘the crux of the problem of the theory of the amending power’.99 In the next section, I deviate from the dichotomy of the two approaches mentioned above and offer a more subtle account of a spectrum of constitutional amendment powers. IV.  THE SPECTRUM OF CONSTITUTIONAL AMENDMENT POWERS

According to the dichotomy described above, there is a binary constitutional code of constitutional amendments: an amendment originating via the constitutional process through the secondary constituent power, which is limited (even if includes the people), and an amendment that is constituted in a constitutional moment through the re-emergence of the primary constituent power and thus unlimited. Constitutional systems are more complex than this. They are polymorphic. Just as in materials science, a solid material can exist in multiple forms, so does the amendment power. Not only do constitutions have different procedures, mechanisms, and actors involved in constitutional amendments, but also a same constitution might incorporate different procedures for amending different provisions and principles. This section argues that constitutional amendment powers ought to be regarded not in a binary manner (limited/unlimited), but rather as a spectrum of scope. The more similar the characteristics of the secondary constituent power are to those of a democratic primary constituent power (‘demanding amendment power’), the less it should be bound by limitations, and vice versa. The closer it is to a regular legislative power (‘facile amendment power’), the more it should be bound by limitations and judicial scrutiny. This calls for an examination of the link between amendment procedures and the limitations that ought to be imposed upon amendment powers and their judicial enforcement. A.  Demanding and Facile Amendment Powers There is no single unified method or process for amending constitutions. Constitutions differ between dissimilar degrees of amendability. Some are ‘flexible’ in that 96 

ibid 13; Tierney (n 48) 360. ibid 13. See similarly Lenowitz (n 41) 85; Smith (n 10) 376. ibid 14. See also Tierney (n 48) 364. 99  Klein (n 94) 213. 97  98 

38  Yaniv Roznai the amendment process is relatively easy, such as ordinary legislative majorities, and some are more ‘rigid’ in that they require high barriers, such as super-majority threshold in parliament, a higher quorum than ordinary legislation, time delays, state ratification in federal systems, or additional requirements such as constituent assemblies, intervening elections and popular referendums.100 These hurdles not only make constitutional change more difficult than that of ordinary law, but also reflect the notion that the Constitution is a special kind of law.101 As described in the first part IIA of this chapter, the amendment power is situated in a grey area between the ordinary legislative power (constituted power) and the extraordinary constituent power.102 In this grey area, a spectrum of amendment power exists. Some amendment procedures are facile amendment powers in the sense that the amendment process is similar (or relatively similar) to the ordinary legislative process in terms of the organs involved, and the temporal and procedural constraints. Others, which significantly deviate from the ordinary legislative process with regard to these features are demanding amendment powers in the sense that their exercise resembles (or almost resembles) a constitutional moment—nearly an invocation of the primary constituent power. The most facile amendment powers are those in which a simple legislative majority is enough to bring about constitutional amendments.103 An amendment power more demanding than the ordinary majority is the one that requires a qualified majority in Parliament for the adoption of amendments. Supermajority requirements expand the range of interests to be taken into account and promote careful considerations, deliberation and persuasion.104 Requirement of multiple readings in Parliament, time delays between the initiative and the first debate in Parliament or between the readings are additional techniques for making the amendment process longer and more difficult thereby making the amendment power more ‘demanding’.105 But all of these procedures are still ‘facile’ from the perspective of a democratic constituent power in the sense that they exclude the people from the process. These procedures of ordinary and qualified majorities in Parliament may be reinforced, for instance with a requirement of a popular referendum, intervening

100  See R Albert, ‘The Structure of Constitutional Amendment Rules’ (2014) 49 Wake Forest L Rev 913; M Andenas (ed), The Creation and Amendment of Constitutional Norms (BIICL, 2000); Oliver and Fusaro (n 93); X Contiades (ed), Engineering Constitutional Change: A Comparative Perspective on Europe, Canada and the USA (Routledge, 2012); ‘Report on Constitutional Amendment’ (n 58); DS Lutz, ‘Toward a Theory of Constitutional Amendment’ (1994) 88 American Political Science Review 355, 363–64; J Elster, Ulysses Unbound. Studies in Rationality, Precommitment and Constraints (Cambridge University Press, 2000) 100–04. 101  cf JE Lane, Constitutions and Political Theory (Manchester University Press, 1996) 114; AV Dicey, ‘The Referendum’ (1894) 23 National Review 65, 69. 102  I elaborate on this in Roznai, Unconstitutional Constitutional Amendments (n 1). 103 See EV Schneier, Crafting Constitutional Democracies: The Politics of Institutional Design (Rowman & Littlefield, 2006) 223–24. 104  R Ku, ‘Consensus of the Governed: The Legitimacy of Constitutional Change’ (1995) 64(2) Ford L Rev 535, 542; LG Sager, ‘The Birth of Logic of a Democratic Constitution’ in J Ferejohn et al (eds), Constitutional Culture and Democratic Rule (Cambridge University Press, 2001) 110, 132–33; J Ferejohn and L Sager, ‘Commitment and Constitutionalism’ (2003) 81 Texas L Rev 1929, 1957. 105  See ‘Report on Constitutional Amendment’ (n 58).

Amendment Power, Constituent Power, and Popular Sovereignty 39 e­ lections, convening a special constitutional convention or a combination thereof.106 When constitutions require intervening elections for amendments, this is a demanding amendment power provided that the subject matter is put as a central issue in the elections’ agenda thereby asking the people’s opinion of it, and allowing enough opportunity for public and political debates on the proposed amendments. It is inclusive, deliberative, and attempts to minimise abuses of the amendment power— as the amending authority would not necessarily be the same before and after the amendment’s enactment. Demanding amendment powers are also those amendment procedures that require elections to a special constituent assembly. This process is inclusive of the people (through the elections, at the very least) and supplies a deliberative setting.107 The US Constitution is an interesting case. On the one hand, comparatively, Article V is unusually onerous as its hurdles are very demanding and time-consuming.108 These consensus requirements arguably label it as a demanding amendment power.109 On the other hand, this procedure is so demanding that some commentators note that ‘from the perspective of … the constituent power … the banishing of sovereignty from the internal life of the republic was perhaps too successful. … it is almost impossible to legally change the American constitution’,110 but this cumbersome procedure does not directly involve the people.111 The argument I advance is simple: demanding amendment powers should be awarded wider scope than facile amendment powers. This idea is compatible with existing constitutional arrangements, which utilise a ‘constitutional escalator’. As noted earlier, some constitutions incorporate different procedures for amending different constitutional subjects.112 These procedures may include, again, increased qualified majorities, referendums, intervening elections, or convening special constituent assemblies. To mention some examples, certain principles may only be amended by a referendum in Belarus (1994, Article 140), Estonia (1992, Article 162), Latvia (1992, Article 77), Lithuania (1992, Article 148), Singapore (1963, Article 5), Serbia (2006, Article 203), and Vanuatu (1980, Article 86). A different procedure exists in the Russian Constitution in which amendments to fundamentals

106 

ibid; Schneier (n 103) 224. Elster, ‘Deliberation and Constitution Making’, in J Elster (ed), Deliberative Democracy (Cambridge University Press, 1998) 97. 108  A Vermeule, ‘Second Opinions and Institutional Designs’ (2011) 97 Virginia L Rev 1435, 1438; R Dixon, ‘Partial Constitutional Amendments’ (2011) U Penn J Const L 643, 651–54. 109  cf LG Sager, ‘The Incorrigible Constitution’ (1990) 65 N Y U L Rev 893, 951–53. 110 A Arato and J Cohen, ‘Banishing the Sovereign? Internal and External Sovereignty in Arendt’ (2009) 16(2) Constellations 307, 317. Critics have argued that the amendment procedure is the main democratic defects of the US Constitution. See S Levinson, Our Undemocratic Constitution: Where the Constitution Goes Wrong (And How We the People Can Correct It) (Oxford University Press, 2006) 159–66; SM Griffin, ‘The Nominee is … Article V’ (1995) 12 Const Comm 171. 111  Interestingly, in contrast with the US Federal Constitution, direct democracy in the form of referenda is inherently part of US States’ constitutional tradition. See J Goossens, ‘Direct Democracy and Constitutional Change in the US: Institutional Learning from State Laboratories’, in this volume. 112  See, eg, South African Const. (1996), Art 74; Canada Constitution Act, 1982, Part V. See H Corder, ‘The Republic of South Africa’ in Oliver and Fusaro (eds), (n 93) 261, 270; AC Hutchinson, ‘Constitutional Change and Constitutional Amendment—A Canadian Conundrum’ in Contiades (ed) (n 100) 51, 64–66; R Albert, ‘Amending Constitutional Amendment Rules’ (2015) 13(3) Int’l J Const L 655; R Albert, ‘The Difficulty of Constitutional Amendment in Canada’ (2015) 53(1) Alberta L Rev 85. 107 J

40  Yaniv Roznai of the constitutional system require conveyance of a Constitutional Assembly (1993, Article 135).113 This is also often the case when the Constitution allows for a ‘total revision’ of the Constitution or its replacement with a new one. For example, the Austrian Constitution requires a popular referendum for its total revision (1920, Article 44[3]).114 In Spain, a total revision of the Constitution or amendments to certain basic provisions demand a more robust process, including the dissolution of Parliament and a subsequent approval by a referendum (Article 168(1)).115 In Costa Rica, a general amendment of the Constitution can be effected only by a constituent assembly (1949, Article 196), and in Bolivia, the total reform of the Constitution or that which affects its fundamental premises, can take place through a constituent assembly, put into motion by a popular referendum (2009, Article 411(1)). In line with my argument, except with the Russian Constitution, none of the abovementioned constitutions includes explicit unamendable provisions.116 The rationale behind this ‘constitutional escalator’ is clear; those provisions that are deemed more fundamental or protection-worthy enjoy a special protection from hasty changes through heightened requirements.117 This ‘selective rigidity’ mechanism was recently advocated for by constitutional scholars, such as Richard Albert and David Landau, for allowing greater protection to the core parts of the democratic order, thus reducing the possibilities of abusing the amendment process, while simultaneously allowing relatively easily amendments of non-fundamental principles. The two scholars also emphasise the importance of the temporal dimension; that is, intervening elections or time delays before the adoption of constitutional amendments, which resist the ability of powerful political forces to abuse their power and take advantage of their temporary popularity to amend the Constitution in a way that would damage democracy.118 From the perspective of constituent power theory, the ‘constitutional escalator’ is not only a practical safeguard for better protecting certain constitutional principles or institutions but means for generating legitimacy, as I elaborate in the next section.119

113 

See Coban (n 35) 57–66. A Somek, ‘Constitutional Theory as a Problem of Constitutional Law—On the Constitutional Court’s Total Revision of Austrian Constitutional Law’ (1998) 32 Israel L Rev 567. 115  See V Ferreres-Comella, ‘A Defense of Constitutional Rigidity’ in P Comanducci and R Guastini (eds), Analisi e diritto (G Giappichelli Publisher, 2000) 45, 62, n 42; A Elvira, ‘Spain’ in Oliver and Fusaro (eds) (n 93) 281, 282–84; AB Ortega and IS Guijarro, ‘Constitutional Change in Spain’ in Contiades (ed) (n 100) 299, 302–08. 116 This of course does not exclude the possibility of implicit unamendability. For example, in ­Resolución 2010–1331, the Supreme Court of Justice of Costa Rica (Constitutional Chamber) stated, in an obiter, that even the derived constituent power cannot amend the Constitution in a way that violates ‘the essence of fundamental human rights’. 117 P Suber, ‘Amendment’ in CB Gray (ed), Philosophy of Law: An Encyclopaedia I (Garland Publishing Co, 1999) 31–32. 118  Landau (n 71); R Albert, ‘Constitutional Handcuffs’ (2010) 42 Ariz St L J 663, 707–11; R Albert, ‘The Expressive Function of Constitutional Amendment Rules’ (2013) 59(2) McGill L J 225. For an argument that selective rigidity distorts the balance between flexibility and stability see RP Plato, ‘Selective Entrenchment against State Constitutional Change: Subject Matter Restrictions and the Threat of Differential Amenability’ (2007) 82 N Y U L Rev 1470, 1489–93. 119 A Arato, ‘Multi-Track Constitutionalism Beyond Carl Schmitt’ (2011) 18(3) Constellations 324, 340. 114  See

Amendment Power, Constituent Power, and Popular Sovereignty 41 B.  Linking Amendment Procedure and Unamendability Demanding amendment processes are devices that aim to imitate constitutional moments in which the primary constituent power is incarnated. Through formal mechanisms, such as referendums and summoning constituent assemblies, they aim to create an environment in which the people are ‘awaking’, in a sense, to resume their role as constituent authors.120 As Xenophon Contiades and Alkemene Fotiadou explain: The people are traditionally considered to have spoken during the exercise of the pouvoir constituant. Amending formulas may be described as replications of the constitutional moment where the pouvoir constituant was exercised, being attempted simulations of that primordial, constitution-making function.… This original constitution-making process is embellished with great symbolic force, the reproduction of which during every constitutional revision would be unfeasible. Yet, desire to somehow preserve the spirit of that moment is often apparent in constitutional arrangements that risk sacrificing practically for symbolism.121

Since primary constituent power is unlimited by constitutional limitations, a process which imitates the re-emergence of primary constituent power should not be considered as equally bound to restriction as facile constituent power, solely activated by one constitutional organ in the same procedure as in ordinary politics. This is because, as Barshack clarifies: Theoretically inelegant as this result may be, the binding force of constitutional procedure varies in every constitutional moment in proportion to the intensity of sovereign ­presence.… When the communal body asserts itself in the amendment of a constitution as intensely as it was involved in its original adoption, it is hardly bound by constitutional procedure at all and hardly subject to judicial review over the constitutionality of the amendment.122

Therefore, from the perspective of constituent power, it is not merely the protection of principles or the fear of abuse which justify ‘constitutional escalator’, that is, allowing the more fundamental principles to be amended through more heightened procedures, it is also the notion that ‘the more exuberant the sovereign presence, the less bound is the collective body by […] the non-amendability of certain constitutional principles…’.123 Different constitutional procedures can aim to create fuller or hollower bodies of ‘popular sovereignty’. It is in this way that we may understand the unique requirement in the Lithuanian Constitution that more than three-fourths of the electorate must participate in a referendum for amending Article 1 of the Constitution, according to which ‘Lithuania is an independent democratic republic’.124 The spectrum of 120  Here, the myth of the people as holders of constituent power serves as the guiding narrative for constitutional design, even if only formally. See A Sajó, ‘Enthusiasm and Acquiescence in Constitutionmaking’ (2009) 3 Indian J Const L 24, 27. 121  X Contiades and A Fotiadou, ‘Models of Constitutional Change’ in Contiades (ed) (n 100) 417, 430. 122  Barshack (n 28) 201. 123  ibid 201–02. 124  Lithuania Const. (1992) Art 2.

42  Yaniv Roznai amendment powers links the amendment process and limitability of the amendment power; the more demanding the amendment power, the less limited it should be. Climbing along the escalator of amendment powers depends upon the various procedures included in the amendment process, which enable inclusive, participatory and deliberative involvement of the people. Yet, the primary constituent power itself is unbound by any constitutional procedures. Thus, at a certain point, with the appearance of the ‘genuine’ primary constituent power the ‘constitutional escalator’ becomes a ‘constitutional roller-coater’ so to say. This suggests that to some extent demanding amendment powers might paradoxically be more difficult to exercise than primary constituent power.125 The normative requirement that in order to acquire a wide scope of action, secondary constituent powers must be demanding, in the sense of their inclusiveness, participatory and deliberative processes, surely implies the assumption that a legitimate exercise of primary constituent power should indeed conform to such requirements. This theoretical construct of a spectrum may be supported by two cumulative rationales. The first rationale is normative. Owing to the democratic nature of the primary constituent power, amendments that are enacted through demanding amendment powers carry greater legitimacy. An author-based theory of legitimacy considers constitution as ‘respect-worthy’ in light of their maker, and the people are the most ‘legitimate’ authors of democratic constitutions. True, when the people are involved in the amendment procedure, they act in their capacity as an institutional organ. Nonetheless, when the people are involved they act as a ‘legitimation elevator’ of the constitutional change.126 For example, referendums maximise legitimacy,127 and special assemblies carry a higher degree of ‘popular legitimacy’ than ordinary legislatures.128 A notable example is the South-African exceedingly participatory constitution-making process which enjoyed a high level of legitimacy.129 Importantly, the binding power of constitutions does not rest solely on ‘procedural legitimacy’, but on other substantive factors as well. However, if the people’s involvement indicates that the current generation accepts the constitutional framework and that the Constitution reflects its values, there is a greater claim for the Constitution’s ‘democratic legitimacy’.130 I am not claiming that a democracy cannot function without strong popular involvement,131 rather that since ‘inclusiveness

125 

I thank Zoran Oklopcic and Alkmene Fotiadou for pointing these issues. IM Young, Inclusion and Democracy (Oxford University Press, 2000) 5–6. This is social and political legitimacy, not a legal or moral one. See generally RH Fallon, ‘Legitimacy and The Constitution’ (2004–2005) 118 Harv L Rev 1787. 127  Butler and Ranney (n 402) 14–15. 128 LE Miller, ‘Designing Constitution-Making Processes: Lessons from the Past, Questions for the Future’ in LE Miller and L Aucoin (eds), Framing The State in Times of Transition: Case Studies in Constitution Making (US Institute of Peace Press, 2010) 601, 612. 129  H Klug, ‘Participating in the Design: Constitution-Making in South Africa’ (1996) 3 Review of Constitutional Studies 18. 130  cf B Medina, ‘“Foundational” Originalism? On Jack Balkin’s Living Originalism’ (2013) 7(1) Jerusalem Rev Leg Stud 1, 8. 131  See, eg, L Vinx, ‘The Incoherence of Strong Popular Sovereignty’ (2013) 11(1) Int’l J Const L 101, 124 (claiming that ‘The view that strong popular sovereignty is essential to democracy is false’). 126 

Amendment Power, Constituent Power, and Popular Sovereignty 43 is the contemporary mechanism for ensuring that a constitution actually is an exercise of the constituent power’,132 then participation and inclusiveness of the people during exceptional moments of constitutional change increases the Constitution’s democratic legitimacy. One might argue that this theory has an inner contradiction. On the one hand, it is argued that whether a primary constituent power is a genuinely ‘sovereign’ exercise can, to a large extent, only be judged retrospectively, that is, if a vast majority of the people accepts the exercise of power as legitimate, then one might argue that the state experienced a ‘constitutional moment’. On the other hand, it is claimed that the presence of demanding participatory and deliberative ex-ante procedures that include the people will increase the likelihood of social legitimacy. The alleged problem with characterising amendment powers as ex-ante facile or demanding (based on the constitutional text), is that it overlooks the backward-looking view of whether the people conceived of the exercise of power as legitimate. In other words, if the legitimacy of the primary constituent power is determined on an ex-post basis, can the position of types of constitutional amendments along the spectrum be determined on an ex-ante basis? The response to this claim is based of course on the distinctions between the two types of powers. A decision whether there was a legitimate exercise of primary constituent power must be judged ex-post facto since this is most commonly the result of illegality or extra-constitutionality (such as amending the Constitution outside the amendment procedure or amending unamendable provisions).133 This act cannot be judged upon standards of legality hence the need for legitimacy claims. Actions by the secondary constituent powers, in contrast, already carry a certain legal basis since they are brought about through the constitutional amendment procedures. There is a lesser need for ex-post considerations when one can infuse, so to say, within the constitutional mechanisms, procedures which would assist in elevating the democratic legitimacy of the amending exercise ex-ante. The second rationale is a practical one. Amendment procedures aim, inter alia, to provide mechanisms for deliberations and to subject the Constitution and proposed amendments to critical scrutiny.134 Presumably, the more deliberative, multiinstitutional and prolonged the processes of amendments are, the less the likelihood of abuse of the amendment power.135 This echoes Jon Elster’s argument that special constituent assemblies should make constitutions, rather than ordinary legislatures, since the latter are more likely to be influenced by group and institutional self-­interests while due to the irregularity of the former, they are presumed to be

132 

Tushnet (n 56) 2000. See also Kalyvas (n 53) 237. See, eg, M Tushnet, ‘Peasants with Pitchforks, and Toilers with Twitter: Constitutional Revolutions and the Constituent Power’ (2015) 13(3) Int’l J Const L 639, 642–43 (according to whom, successfully amending an unamendable principle through a supposedly unconstitutional amendment is an exercise of constituent power). 134  H Kong, ‘Deliberative Constitutional Amendments’ (2015) 41 Queen’s L J 105; Compare with CR Sunstein, Designing Democracy: What Constitutions do (Oxford University Press, 2001) 7, 240. 135  cf Landau (n 71) 226. 133 

44  Yaniv Roznai i­mpartial bodies insulated from short-term political bargaining.136 Interestingly, a similar argument has been made with regard to direct democracy.137 A referendum device may provide a safeguard against an executive which controls the Parliament and can abuse the amendment power for its short-term political interests.138 In other words, the referendum plays a role both as fostering a wide civic participation and assisting the check over the executive power in constitutional change.139 Some argue that exceptional popular mechanisms, especially in the absence of ordinarily political institutions, have actually aided charismatic leaders to impose authoritarian constitutions.140 Others point out that popular inclusiveness in a ‘we the majority’ form risks minority rights,141 which might even justify robust judicial review of direct democracy.142 In contrast, others claim that ‘direct democracy and the protection of minorities are not mutually exclusive’.143 The question is not whether mechanism of direct democracy risk minority rights, but whether such mechanisms increase or reduce such risks compared to other representative mechanisms, but this remains indecisive.144 Importantly, this chapter does not engage with ordinary legislative referendums, but with constitutional referendums taking place in exceptional moments. The two types of referendums are different.145 Moreover, a stand-alone referendum does not 136  J Elster, ‘Forces and Mechanisms in the Constitution-Making Process’ (1995) 45 Duke L J 364, 370–71, 395; J Elster, ‘Legislatures as Constituent Assemblies’ in RW Bauman and T Kahana (eds), The Least Examined Branch: The Role of Legislatures in The Constitutional State (Cambridge University Press, 2006) 181, 185; J Elster, ‘Arguing and Bargaining in Two Constituent Assemblies’ (2000) 2 U Pa J Const L 345, 419–20. Of course, the construction of constituent assemblies established for constitutional revision can also be manipulated or abused as demonstrated in this volume by Braver (n 15). 137  cf TE Pettys, ‘Popular Constitutionalism and Relaxing the Dead Hand: Can the People Be Trusted?’ (2008) 86 Wash U L Rev 313, 350–51. 138  KM Pillai, ‘Amendability of Fundamental Rights under the Constitution of India’ in Dhavan and Jacob (n 14) 192, 199. 139 E Daly, ‘A Republican Defence of the Constitutional Referendum’ (2015) 35(1) Leg Stud 30; AV Dicey, ‘Ought The Referendum To Be Introduced Into England?’ (1890) 57 Cont Rev 489, 505. 140 W Partlett, ‘The Dangers of Popular Constitution-Making’ (2012) 38 Brook J Int’l L 1, 4; D Landau, ‘Constitution-Making Gone Wrong’ (2013) 64(5) Ala L Rev 923. 141 HP Monaghan, ‘We The People[s], Original Understanding, and Constitutional Amendment’ (1996) 96 Colum L Rev 121. This might find some support in empirical studies. See, eg, BS Gamble, ‘Putting Civil Rights to a Popular Vote’ (1997) 41(1) Am J Pol Sci 245, 254, 261; B Frey and L Goette, ‘Does the Popular Vote Destroy Civil Rights?’ (1998) 42 Am J Pol Sci 1343; ZL Hajnal, ER Gerber and H Louch, ‘Minority and Direct Legislation: Evidence from California Ballot Propositions Elections’ (2002) 64(1) Journal of Politics 154; RT Moore and N Ravishankar, ‘Who Loses in Direct Democracy?’ (2012) 41 Social Science Research 646. But see JG Matsusaka, ‘The Eclipse of Legislatures: Direct Democracy in the 21st Century’ (2005) 124 Public Choice 157, 168 (arguing that the question of risk to minorities in direct democracy lacks ‘rigorous empirical work on this issue, and the work that does exist rests on flawed methodologies’). 142  JN Eule, ‘Judicial Review of Direct Democracy’ (1990) 99 Yale L J 1503. 143 W Marxer, ‘Forward’ in W Marxer (ed), Direct Democracy and Minorities (Springer Verlag, 2012) 7, 10. See also CA Johanningmeier, ‘Law & Politics: The Case Against Judicial Review of Direct ­Democracy’ (2007) 82 Indiana L J 1125, 1150. 144  JG Matsusaka, For the Many or the Few: The Initiative, Public Policy, and American Democracy (University of Chicago Press, 2008) 116; S Spadijer, ‘A Hardcover Case Against (Strong) Judicial Review of Direct Democracy’ (2012) 31 U Queensland L J 55. 145  See Tierney (n 2); Tierney (n 48) 360. Interestingly, referendums on constitutional revisions are less subject to substantive judicial review than referendums on ordinary legislation. See L Morel, ‘Referendum’ in M Rosenfeld and A Sajó (eds), The Oxford Handbook of Comparative Constitutional Law (Oxford University Press, 2012) 501, 522.

Amendment Power, Constituent Power, and Popular Sovereignty 45 engender demanding amendment power, as I have described it. Since a demanding amendment process is meant to be an inclusive, deliberative, and time-consuming, the referendum should be an additional platform to the political process, perhaps together with a special constituent assembly. The risk of abuse of the amendment power arises especially with facile amendment powers, where the amendment body is the same body that decides the everyday political decisions. The identity of bodies causes the mingling of longer-range issues of constitutional planning with short-term interests of political power.146 Blount, Elkins, and Ginsburg remark that ‘higher levels of participation are presumed to function like supermajority rules, restricting the adoption of undesirable institutions and protecting prospective minorities in the democratic processes that are established’.147 This is not necessarily a matter of ­difficulty,148 but of inclusiveness and deliberations. As Christopher Eisgruber notes, ‘by establishing a separate and difficult track for some political issues, the Constitution may focus public attention upon those decisions and improve deliberation about them’.149 An inclusive and deliberative process aims to improve the quality of the constitution-making/amending outcome.150 Accordingly, inclusive and deliberative amendment procedures which allow time for public and institutional deliberations reduce the possibility of abuse and enhance the legitimacy of the endorsed constitutional change.151 Such procedures, which imitate the invocation of primary constituent power should be given a greater margin of change: ‘the more significant a change may be, more likely is it that the procedure for affecting it would be more arduous, difficult and demand a higher level of deliberative legitimacy’.152 Importantly, even demanding amendment powers still act as limited constitutional organs. Yet, this limitability is not one-dimensional but moves along the spectrum of the amendment powers: ‘not all amending routes are equal. The closer amending routes are to the citizenry, the more freedom there is to alter the foundational terms of our political life’.153 Amendment processes must be linked to unamendability: The nature of the amending power as well as the level of deliberative legitimacy which the amending process imbibes in itself jointly [determines] … what is allowed to be changed

146 Conrad (n 52) 14–16; D Grimm, ‘The Basic Law at 60—Identity and Change’ (2010) 11(1) German L J 33, 40. 147 J Blount, Z Elkins, and T Ginsburg, ‘Does the Process of Constitution-Making Matter?’, in T Ginsburg (ed), Comparative Constitutional Design (Cambridge University Press, 2012) 31, 36. 148  A requirement of a popular ratification certainly adds another hurdle to the amendment ­process, but approvals of constitutional amendments via referendums is not necessarily more difficult than approving it in state legislatures. See DS Lutz, Principles of Constitutional Design (Cambridge University Press, 2006) 166–68; J Ferejohn, ‘The Politics of Imperfection: The Amendment of Constitutions’ (1997) 22 L & Soc Inq 501, 523. 149  CL Eisgruber, Constitutional Self-Government (Harvard University Press, 2007) 44. 150  cf Blount, Elkins and Ginsburg (n 147) 31, 54; AM Banks, ‘Expanding Participation in Constitution Making: Challenges and Opportunities’ (2008) 49 William & Mary L Rev 1043, 1050; K Samuels, ‘Post-Conflict Peace-Building and Constitution-Making’ (2006) 6(2) Chi J Int’l L 663, 668. 151 Deliberative settings may also increase public trust in the constitutional reform process. See R Levy, ‘Breaking the Constitutional Deadlock: Lessons From Deliberative Experiments in Constitutional Change’ (2010) 34 Melbourne U L Rev 805, 838. 152  Prateek (n 13) 459. 153  Note (n 94) 1762–63.

46  Yaniv Roznai and what is not through an amendment. So it is not that the limits on amending power are sketched without any reference to the procedure of amendment. Both the nature and procedure of amendments are critically important to truly understand what these limits may be.154

Considering the idea of a spectrum of amendment powers, William Harris’s constitutional theory carries much force. Similar to my argument, Harris distinguishes between the people’s unlimited constituent power and the instituted amendment authority which is a constitutional agent, holding its power in trust and thus bound by limits.155 Harris links the scope of constitutional amendments with the ‘wholeness’ of the collective people as the source of the constitutional authority. The more amendments seek to influence people’s rights or large scale revisions, the more necessary it is to seek popular approval through ‘sovereignty-reinforcing’ mechanisms such as special constituent conventions. Put differently, there is a reciprocal relationship between the ‘wholesale’ of the sought change and the ‘wholesale’ of the required sovereignty’s presence, which is identified through the criteria of ‘wholeness and deliberateness’.156 The spectrum of amendment powers advances Harris’ theory by recognising a systematic scale of amendment powers based upon the proximity to a democratic primary constituent power. C. The Spectrum of Amendment Powers and Judicial Review of Amendments The ‘spectrum of amendment powers’ is an analytical description of various amendment procedures. This explanatory theoretical model carries with it normative aspects: First, a constitutional design aspect directed at constitution-makers and urging them to design constitutional amendment-rules in an escalator way. In doing so, the more basic constitutional principles would be amendable in a more participatory process, which is time-consuming, deliberative and inclusive. This is in contrast with the less foundational provisions of the Constitution, which should be amendable relatively easily. Secondly, it is aimed at the judiciary. The intensity of judicial scrutiny of constitutional amendments should be connected to the amendment process. True, since we are dealing with secondary constituent powers, judicial review of amendments should be based on its compliance with the still-binding constitutional rules laid down explicitly or implicitly.157 But even once amendment powers are conceived as

154 

Prateek (n 13) 472. Harris (n 12) 167. 156  ibid 194–203. 157 On the authority to review constitutional norms and the proper standard of such a review see Y Roznai, ‘Legisprudence Limitations on Constitutional Amendments? Reflections on The Czech Constitutional Court’s Declaration of Unconstitutional Constitutional Act’ (2014) 8(1) ICL-J 29, 33–40. On judicial review of amendments see generally K Gözler, Judicial Review of Constitutional Amendments— A Comparative Study (Ekin Press, 2008); A Barak, ‘Unconstitutional Constitutional Amendments’ (2011) 44(3) Israel L Rev 321; G Halmai, ‘Unconstitutional Constitutional Amendments: Constitutional Courts as Guardians of the Constitution?’ (2012) 19(2) Constellations 182. 155 

Amendment Power, Constituent Power, and Popular Sovereignty 47 delegated authorities, holding their power in trust, fiduciaries’ duties are enforced with different degrees of strictness, depending on the nature of the specific relationship under consideration.158 As I elaborate in this chapter, not all amendment p ­ owers are similar. A spectrum of amendment powers exists. Some are facile amendment powers, which resemble ordinary legislative powers, some are demanding amendment powers, which allow for popular participation and deliberation and involve a great deal of time. One can expect (although this is not always the case) that the easier the amendment process, the higher the rate of formal amendments.159 An extremely facile secondary constituent power, where a dominant executive/legislature may control the amendment process, coupled with short-term political interests and temporary majorities, increase the fear of abuse of the amendment power.160 It is here where the strictest judicial oversight is required.161 Even Lester Orfield, ordinarily an antagonist of unamendability, states that ‘undoubtedly, where a simple majority is required, it is not an especially serious matter for the courts to supervise closely the amending process both as to procedure and as to substance. But when so large a majority as three-fourth has finally expressed its will in the highest possible form outside of revolution, it becomes perilous for the judiciary to intervene’.162 Therefore, the legitimacy of judicial review of amendments depends, at least in part, on the amendment process: ‘the fuller the enactment of sovereignty, the less justiciable the sovereign action’.163 The more the amendment is the product of inclusive and deliberative demanding amendment powers, which enjoy a high degree of democratic legitimacy and minimise risks of abuse, the less intensive the judicial review of amendments should be, and vice versa.164 Of course, this is merely a call

158  cf EJ Leib, DL Ponet and M Serota, ‘Translating Fiduciary Principles Into Public Law’ (2013) 126 Harv L Rev Forum 91, 93. 159  CM Fombad, ‘Limits on the Power to Amend Constitutions: Recent Trends in Africa and Their Potential Impact on Constitutionalism’ (2007) 6 University of Botswana Law Journal 27, 59; B ­Ackerman, ‘Constitutional Economics—Constitutional Politics’ (1999) 10 Constitutional Political Economy 415, 423. Amendments-rate depends on many other criteria. See R Dixon, ‘Constitutional Amendment Rules’ in Ginsburg and Dixon (n 40) 96; R Dixon and R Holden, ‘Constitutional Amendment Rules: The Denominator Problem’ in Ginsburg (n 147) 195. For a critical analysis of quantitative amendment metrics see X Contiades and A Fotiadou, ‘Amendment-Metrics: The Good, the Bad and the Frequently Amended Constitution’, in this volume and a response by JE Fleming, ‘Comment on Amendment-Metrics: The Good, the Bad and the Frequently Amended Constitution’, in this volume. See also T Ginsburg and J Melton, ‘Does the Constitutional Amendment Rule Matter at All? Amendment Cultures and the Challenges of Measuring Amendment Difficulty’ (2015) 13(3) Int’l J Const L 686. 160  Conrad (n 50) 415. 161  PJ Yap, ‘The Conundrum of Unconstitutional Constitutional Amendments’ (2015) 4 Glob Const 114, 131 (‘we should … weigh the risks of any legislative abuse of the amendment process against the dangers of any judicial abuse that may follow from unelected judges enforcing a nebulous ‘essential features’ doctrine that can frustrate legitimate constitutional revisions designed to meet changing times’.) 162 LB Orfield, Amending the Federal Constitution (The University of Michigan Press, 1942) 120. See also LB Orfield, ‘The Scope of the Federal Amending Power’ (1929–1930) 28 Mich L Rev 550, 558. 163  Barshack (n 28) 198, n 24. 164  cf ‘Report on Constitutional Amendment’ (n 58) 46–47; Jacobsohn (n 90) 487; Prateek (n 13) 465–67; Yap (n 161) 131–5; C Bernal, ‘Unconstitutional Constitutional Amendments in the case Study of Colombia: An analysis of the Justification and Meaning of the Constitutional Replacement Doctrine’ (2013) 11 Int’l J Const L 339, 357.

48  Yaniv Roznai for judicial restraint not the elimination of judicial oversight over actions of the amendment power. The Constitution’s substance is at least as important as the procedure by which amendments are enacted, and even demanding amendment powers are still limited constitutional organs. It has been remarked that jurists ‘must either learn to trust the amending process or repose their faith in non-elected judges’.165 I claim that these two variables— judicial scrutiny and the amendment process—are mutually reinforcing. Since the spectrum of amendment power is linked to the question of unamendability, the mirror picture of the spectrum of amendment power is the spectrum of intensity of judicial scrutiny and restraint, which should be exercised over constitutional amendments. V. CONCLUSION

The basic theoretical presupposition behind the theory of unamendability is the distinction between the people’s primary constituent power and the instituted secondary constituent power.166 In contrast with the primary constituent power, the secondary constituent power is limited by unamendability so that certain constitutional decisions require the re-emergence of the primary constituent power. They force ‘the real sovereign to return from its retirement in the clouds’167 in extraordinary constitutional moments. Amendment procedures often try to imitate these constitutional moments by creating extraordinary procedures, such as referendums and constituent assemblies, which would distinguish these constitutional politics from ordinary ones.168 As Contiades and Fotiadou write in this volume: In recent years, there is a growing tendency to involve the people in constitutional amendment through processes enhancing deliberation and popular participation. This trend envisages constitutional amendment as a participatory process aimed at creating a bridge between constituent and constituted power.169

These are demanding amendment powers. This chapter argued that the more an amendment process contains inclusive and deliberative democratic mechanisms, the closer it resembles the people’s primary constituent power. Congruently, since primary constituent power is unlimited by constitutional limits, demanding secondary constituent powers, which present a fuller presence of peoples’ sovereignty—while still limited—should be allowed greater latitude of constitutional changes. This postulation finds support in comparative constitutional amendments in the form

165 R Dhavan, ‘The Basic Structure Doctrine—A Footnote Comment’ in Dhavan and Jacob (n 14) 160, 178. 166 Roznai, Unconstitutional Constitutional Amendments (n 1). 167  Harris (n 12) 198. 168  Tierney (n 48) 367 (referendums can be regarded ‘as some kind of reversal of the original act of transference, or at least as a temporary return of power to the people’). 169  Contiades and Fotiadou (n 159), in this volume.

Amendment Power, Constituent Power, and Popular Sovereignty 49 of ‘escalator amendment procedures’. As the puzzle surrounding constitutional unamendability concerns a deeper conflict between substantive versus procedural approaches to constitutionalism; the former focuses on the Constitution’s fundamental principles and the latter on the Constitution’s procedures,170 the theory of spectrum of amendment powers links amendment procedures to unamendability thereby harmonising substance and procedure.171

170 

JK Tulis, ‘The Demon at the Center’ (1988) 55(2) U Chi L Rev 548. cf ME Brandon, ‘The “Original” Thirteenth Amendment and the Limits to Formal Constitutional Change’ in Levinson (ed) (n 30) 216, 228. 171 

50 

2 Constitutional Theory and Cognitive Estrangement Beyond Revolutions, Amendments and Constitutional Moments ZORAN OKLOPCIC

Freedom is always coming in the hereafter. But you know what, though—‘the hereafter’ is a hustle. We want it now.*

I.  INTRODUCTION: THE PERSON OF ‘THE PEOPLE’ AND A THREE-FOLD COGNITIVE ESTRANGEMENT

C

ONSTITUTIONAL THEORY IS the home of two partially overlapping, occasionally mutually reinforcing, but nonetheless competing visions of ­ peoplehood, both of which delineate the contours of our imagination of legitimate constitutional change. According to the first, ‘Sieyèsian’ vision, radical constitutional change is implicated in a well-known, (allegedly paradoxical) dynamic between the unconstrained pouvoir constituant of a sovereign people, and the constraining ­pouvoir constituée of a constitutional order, which gives ‘the people’ its political identity.1 Since it imagines ‘the people’ as a political body that can, at any point in time, rightfully rupture existing constitutional order and create a new one, ‘Sieyèsian’ view does not find it necessary to offer further portrayals of people’s personality traits. While the ‘Sieyèsian’ vision doesn’t concern itself with ‘the people’s’ character or temperament, the second, ‘Lockean’ vision portrays ‘the people’ as a fallible, but ultimately reasonable person. ‘The people’, memorably described by some theorists as Ulysses, binds itself to the mast of a constitution, in order to fend off the seductive

*  Jesse Williams, BET Humanitarian Award acceptance speech (26 June 2016). I transcribed his remark from a video of his speech, available at , accessed 20 July 2016. 1  See generally, M Loughlin and N Walker, ‘Introduction’ in M Loughlin and N Walker (eds), Paradox of Constitutionalism: Constituent Power and Constitutional Form (Oxford University Press, 2007) 4.

52  Zoran Oklopcic call of populist Sirens.2 In order for such people to exercise its constituent power, liberal democrats first imagine it as slow to engage in radical political action, thoroughly exhausted with suffering grave oppression before it exercises its revolutionary constituent power. Such view of ‘the people’s’ character gives rise to a revolution/ amendment [R/A] binary: ‘the people’ either radically re-constitutes constitutional order as a result of oppression, or otherwise changes it through regular constitutional channels. Contemporary constitutional theorists have engaged the postulate of Lockean reasonable peoplehood in different ways. Contemporary radical democratic defenders of the Sieyesian view, such as Joel Colón-Riós, have rejected it, by defending the idea of ‘weak’ constitutionalism, which does away with procedural restrictions on the activation of ‘the people’s’ pouvoir constituant.3 From within the liberal democratic camp, Jeremy Waldron has rejected the gesture of anthropomorphisation, finding it unhelpful in defending an attractive account of liberal-democratic constitutional politics.4 In different ways, and for different reasons, both have undermined the R/A binary. In this essay I also seek to problematise, and, ultimately, dissolve it. However, rather than recasting, or critiquing the R/A binary from the from the vantage point of radical democracy or political liberalism, my intervention starts off with a ­proposal not to simply embrace Weberian disenchantment [Entzauberung] with the idea of ‘the people’, but to actively engage in, what theorists of science fiction call cognitive estrangement.5 For Darko Suvin, a literary theorist who coined the term, ‘[a] representation which estranges is one which allows us to recognise its subject, but at the same time makes it seem un-familiar’.6 Simultaneously creative and cognitive, ‘the look of estrangement’, always must assume ‘the possibility of other strange, co-variant coordinate systems and semantic fields’.7 The concrete cognitive estrangement I propose in this essay is three-fold. Quite uncontroversially, I first propose that we estrange ourselves from ‘the people’s’ collective personhood. By abandoning people’s personhood, we would likewise distance ourselves from the corresponding portrayals of what constitutes (un)desirable constitutional change, otherwise connected to the image of a reasonable popular sovereign. Unsurprisingly, this would immediately seem to push us in a Sieyèsian direction: on the one hand, without the image of a fallible but reasonable Ulysses, we would have no reason not to succumb to quotidian populist temptations. By the same token, we would we have no reason not to engage in premature radical

2  S Holmes, Passions and Constraint: On the Theory of Liberal Democracy (University of Chicago Press, 1996). 3 JI Colón-Ríos, Weak Constitutionalism: Democratic legitimacy and the question of constituent power (Routledge, 2013). 4  J Waldron, ‘Precommittment and Disagreement’ in L Alexander (ed), Constitutionalism: Philosophical Foundations (Cambridge University Press, 2001). 5  D Suvin, ‘On the Poetics of the Science Fiction Genre’ (1972) 34(3) College English 372. I borrow the term from the theory of science fiction, where it is has been influential in articulating its distinctiveness, vis-à-vis other literary forms. 6  ibid 374. 7 ibid.

Constitutional Theory and Cognitive Estrangement 53 constitutional overhauls, against undesirable, but not necessarily oppressive, constitutional regimes. The second act of cognitive estrangement, however, equally goes against the idea of both Lockean and Sieyèsian peoplehood. Here I suggest we also estrange ourselves from the idea of a constitutional order, as the machine, temple, or vehicle— in any event from the idea that order is a lifeless product of people’s constituent power. More controversially, I suggest that instead of personifying ‘the people’, we provisionally personify constitutional order itself. For the purposes of this chapter this strikes me as immediately useful: having done that, we will appreciate the strange—and until recently dominant—but never really problematised, attitude of contemporary constitutional theory towards radical constitutional change. With the Ulysses-like reasonable ‘people’ out of the picture, we will notice that mainstream foundational constitutionalism portrays ‘constitution’ as a stubborn teenager, or as a credibility-obsessed mafia boss. With marked recent exceptions that I will discuss later in the text aside, foundational constitutionalism would rather have a constitution die in a revolution, than survive, but ‘lose face’, by accommodating concrete radical constitutional demands, out of turn, by dishonouring the conditions prescribed in its own amendment procedure. The final, third, move of cognitive estrangement rests on anthropomorphising not only the Constitution, but also the field that theorises it. To say that constitutionalism ‘portrays’, or ‘describes’ as I have said above, is a commonplace rhetorical shorthand, most often with no meaningful theoretical and conceptual implications. Instead, however, I propose that we take such casual personifications of the field of constitutional theory a step further and, for the sake of this article, imagine the field of constitutional theory as a person itself, harbouring not only hopes and ideals, but also political and moral anxieties.8 In the context of problematising the R/A binary, this three-fold constitutional estrangement will serve four, more concrete, objectives. First, a perspective that abandons ‘the people’s’ personhood and focuses on discipline’s anxieties will help us uncover the tropes that remain obfuscated behind the ideas of ‘the people’ and the R/A binary. Behind them—as I will argue in section 1—we will encounter the tropes of acceleration, desire, space, responsiveness, fragmentation and erosion. Secondly, by upfronting discipline’s anxieties we will establish a further productive detachment from the ways in which contemporary theorists attempt to maintain our faith in sovereign peoplehood. In section 2, I will discuss three strategies of

8  I use the word anxieties deliberately, for three reasons. First, anxieties entail predictions about future. These predictions carry with them both an element of that what is reasonably expectable, but also exaggerated anticipations of future dangers. Secondly, I call them anxieties because constitutional theory doesn’t openly discuss them as worries, but rather suppresses them through its constitutional imaginary. Finally, calling those fears anxieties suggests not only that they might be blown out of proportion, but also that something productive can be done about them either at the level of conceptual re-imagination, or institutional prescription. For a humorous treatment of anxieties in the production of constitutional subjectivity that does not end on a constructive note, see P Schlag, ‘The Empty Circles of Liberal Justification’ (1996) 96(1) Mich L Rev 1. For a more general treatment of Cartesian Anxiety, that informs much of our contemporary (including constitutional) thinking, see R Bernstein, Beyond Objectivism and Relativism: Science, Hermeneutics and Praxis (University of Pennsylvania Press, 1983) 19.

54  Zoran Oklopcic estrangement-prevention through which modern foundational constitutionalism negotiates its ideals and suppresses its anxieties—implicitly or explicitly re-enacting the R/A binary. Thirdly, by undermining the naturalness of the R/A binary, we will be ­better equipped to map the ways in which extra-constitutional constitutional change can be imagined in the space between them. In section 3, I set the stage for the re-imagination of that conceptual space by returning to the idea of ‘weak’ constitutionalism and by discussing the idea of ‘constitutional moment’, as two contemporary ways in which the R/A binary was either collapsed, or undermined. Though different in the way they understand the relationship between ‘the people’ and its constituent power, both ways of challenging the R/A binary nonetheless continue relying on the idea of a unified, sovereign people. This is problematic, I argue in this section, for two reasons. On the one hand, both prevent us both from identifying tendential responsiveness, as the common denominator embraced both by ColónRiós’s ‘weak’ constitutionalism and Ackerman’s theory of ‘constitutional moments’. On the other hand, by continuing to take ‘the people’ for granted, both obfuscate the three analytically distinct, if mutually intertwined, domains of tendential responsiveness: spatial, temporal and personal. Making this analytical distinction will enable us to enrich and complement the debate about constitutional amendments by introducing forms of radical constitutional change that either conceptually, or institutionally don’t rely on the will of the sovereign people. Finally, fourthly, cognitive estrangement will not only help us detect the tropes and possibilities behind the R/A binary, but will also help reframe other enduring debates in constitutional theory. On the one hand, in the final section of the essay, I propose a new way of understanding the so-called paradox of constitutionalism. On the other hand, I rearticulate the function of ultimate political sacrifice in constitutional imaginary. I end the essay, not only with a hopefully more nuanced understanding of the conceptual space between revolution and amendment, but also with a call for a more contextually sensible ‘marketing’ of the conceptual and institutional heritage of modern foundational constitutionalism. II.  BEYOND ‘THE PEOPLE’: NEW TROPES, OLD ANXIETIES

To claim there is something curious about modern constitutional orders—which would rather violently die from ‘the people’s hand’ than be impelled to accommodate, (but survive and lose face) the demands from one of ‘the people’s segments’—assumes that we can think of constitutions and constitutionalism differently. Thinking about them differently, in turn, itself rests on certain assumptions. To imagine a constitution not as a teenager, but as a generous adult, willing to respond with care to demands posed with urgency, but out of turn, would suggest that constitutional orders might be not only be capable of structuring big and small rhythms of political change, but that they could (or should) in principle speed up towards the satisfaction of discrete political demands. Equally, to imagine a constitutional order, which is in principle willing to accommodate such segmental demands,

Constitutional Theory and Cognitive Estrangement 55 implies that their satisfaction would have something to do not only with the quality of reasons that support them, but with the intensity of desires behind them. To imagine, finally, a fragment of population provoking, and successfully obtaining, such a change would gesture towards a potentially different constellation of responsiveness that remains obfuscated under the idea of a territorial people—which otherwise only through an amendment or a revolution effects a comprehensive constitutional change within a designated, integral territory. I will revisit these concepts in section 3, but at this point it is worth provisionally highlighting: acceleration, desire, configuration, responsiveness—these tropes appear as worthy of constitutionalist attention in their own right, only once we have imaginatively estranged ourselves from the idea of ‘the people’s’ personhood. Equally, however, by treating the field of constitutional theory as an anxious individual helps explain the remarkable resilience of the idea of a sovereign people, together with the corresponding reluctance to undermine the R/A binary. Without ‘the people’, and the R/A binary reliant upon it, accelerated responsiveness might make the very idea of constitutionalism empty and dangerous. While a robust conception of collective self-government assumes that ‘consent would flow continuously’ so that the ‘business of the self [is] not … displaced onto any distant force’,9 such a view rouses the anxieties of modern constitutionalism, which, above all, fears all ‘temporary impulses [and] manipulated desires’.10 Being responsive to intense, unqualified desires, unmediated by public reason, might make constitutional order accommodate wicked, stupid or unjust political projects. As famously stated by John Rawls in a different context, ‘intensity of desire, or strength of conviction is irrelevant when questions of justice arise’.11 Also, responding to such desires that come from discrete fragments of population may lead not only to sharp polarisation among discrete interest groups, but, more ominously, to territorial fragmentation— the destruction of constitutional order’s territorial integrity. A generous, responsive constitution might respond favourably to the demands posed out of turn (as we will see in our discussion of the Secession Reference) but its anxieties undercut the explicit recognition of acceleration, desire and variable constellations of responsiveness as constitutive building blocks of foundational constitutionalism, behind the vocabulary of sovereign peoplehood. Finally, this initial sketch of the tropes behind sovereign peoplehood would not be complete without an extended temporal perspective: the reasons for preserving the picture of a constitutional order as a face-obsessed, inflexible adolescent, however, rely not only on the putative intrinsic demerits of approaching speed, desire and responsiveness affirmatively, without prior prejudices. The anxiety of constitutional theory is critically informed by the anticipation of the future ripple effects of the present-day accommodation of the notionally unconstitutional behaviour.

9 

F Michelman, ‘Foreword: Traces of Self-Government’ (1986) 100 Harv L Rev 4, 42. Thompson, ‘Democracy in Time: Popular Sovereignty and Temporal Representation’ (2005) 12(2) Constellations 245, 246. 11  J Rawls, A Theory of Justice (Harvard University Press, 1971) 361. 10 DF

56  Zoran Oklopcic III.  THREE FORMS OF ESTRANGEMENT-PREVENTION: HOLMES, PETTIT, DWORKIN

We will inquire whether constitutional theory must capitulate before these worries in section 3, but before that, it is important to appreciate the idea of a sovereign people—and the R/A binary as its corollary—in a well-rounded way: not only as the manifestation of disciplinary and grassroots anxieties, but also as the manifestation of their ideals. Suppressing the spatiotemporal anxiety of constitutional theory is not the only reason why ‘the people’—either as a collective ‘person’, or as a name for a more abstract principle of constitution-making12—continues its sway over our imagination, preventing a productive cognitive estrangement I proposed in the ­introduction. There are other reasons, to which I now turn, whose effect is to uphold the idea of ‘the people’’s personhood that go beyond the suppression of spatiotemporal anxiety. For Stephen Holmes, conjuring the idea of ‘the people’ as a person (Ulysses) is useful not only because it enables him to harness the evocative power of Greek mythology to fend off seductive threats of anti-liberal populism, but also to provide ideational support for the protection of private property, and capitalist economy in general. Relying on James Madison, who refused to meet anti-constitutionalists on the plane of ‘high theory’, Holmes sees the virtues of trans-temporal, constitutionally pre-committed people political subjectivity in the prevention of ‘squander[ing] scarce resources’.13 ‘Some essential national purposes’ can be achieved only through an inter-temporal division of labour.14 Though Holmes rejects the proposition that the ‘inevitability of latent conflict’ would necessarily lead to ‘[l]ong term advantage of … debtors’, he sides with Madison in that the absence of collective pre-­commitment would ‘discourage every useful effort of steady industry’ to the detriment of all.15 For Philip Pettit, the idea of collective peoplehood is the logical outcome of the deeper republican ideal of non-domination. We need to posit the existence of the territorially-bounded people because only in that way would we be able to justify the political equality of citizens, who possess equal control and influence over the direction of their government.16 However, the reason why republican constitutionalists should reject the critical salience of the paradox of constitutionalism lies both in his assessment that democratic theory cannot offer guidance about the judging the legitimacy of ‘the people’, but also in his peculiar way in which he portrays external political environment. The reason why trapped minorities should accept their existence in states they dislike, as a matter of ‘tough luck’17 is because the

12 See J Habermas, Between Facts and Norms: Contributions to a Discourse Theory of Law and Democracy (William Rehg tr, The MIT Press, 1998) 133. 13  Holmes (n 2) 155. 14 ibid. 15 ibid. 16 P Pettit, On the People’s Terms: A Republican Theory and Model of Democracy (Cambridge University Press, 2012). 17  ibid 153.

Constitutional Theory and Cognitive Estrangement 57 ‘states [are] locked into a pattern’ which would enable the external predatory states to exploit domestic attempts to undermine the unity of ‘the people’.18 Assuming the ever-­present potential of external predation, forces you not only to ‘[y]ou … to comply with the state’ but also ‘assume its perspective’.19 Once the imagined external threat has corralled citizens together, Pettit mobilises the ideal of legal equality to further cement the imperative of belonging to the extant people. He does so by conjuring a binary between the desirable general obedience of the law and individual obedience ‘a-la-carte’. The latter is in turn interpreted as an unfair official exemption that privileges some, at the expense of all others. Under this scheme, allowing the secession of a segment of population would be problematic not simply because that group would take away a portion of the territory that belongs to all, but, more importantly, because such move would create a situation of legal inequality between secessionists and the population of the remainder of the state. Finally, Ronald Dworkin’s defense of law as integrity provides further, both direct and indirect, support for people’s personhood, even though Dworkin never directly concerned himself with the idea of popular sovereignty. For Dworkin, not only is the personification of groups widespread and a widely accepted feature of social life, but personifying a political community is a political and moral imperative. We ought to do it, if we believe that ‘community, as a whole, has obligations of impartiality towards its members, and that its officials must act as agents for the community in acquitting that responsibility’.20 We have to do it since those obligations cannot be grounded in ‘private morality’.21 Beyond the idea of ‘self-legislation’, a citizen’s understanding of political community as a moral person has practical and moral benefits for two types of relationships: the relationship between the governors and the governed, on the one hand; and, the relationship among the governed themselves, on the other. First, the perspective of community serves as a protection against partiality, deceit, corruption, favouritism, or vindictiveness on behalf of public officials.22 Secondly, assuming the perspective of political community contributes to the solidification of political friendship amongst citizens as it reminds them of the long game of their political relationship. It invites them to ‘treat relations among themselves as characteristically not just spasmodically governed by these standards’.23 In reminding citizens of the long game, it perforce invites them to decentre their moral perspective: ‘it insists that each citizen must accept demands on him, and may make demands on others, that share and extend the moral dimension of any explicit political decisions’. And that, in turn, has potential for engendering

18 

ibid 161.

19 ibid. 20 

R Dworkin, Law’s Empire (Belknap Press, 1986) 175. Note that the belief in the necessity of personification cuts across the liberal-democratic and radical democratic political divide and is shared by both. See C Lefort, Democracy and Political Theory (Polity Press, 1988) For Lefort, ‘any attempt to define objective impersonal entities implies a ­personification of those entities’ (254). 22  Dworkin (n 20) 188. 23  ibid 189. 21 ibid.

58  Zoran Oklopcic political generosity in situations of heightened moral conflict. The ‘expressive value’ of political community is ‘confirmed when people in good faith try to treat one another in a way appropriate to common membership in a community … and to see each other and making this attempt, even when they disagree about exactly what integrity requires in particular circumstances’.24 Drawing attention to the assumptions and implications of Holmes’, Pettit’s and Dworkin’s argument is important, as it helps establish who and why might find the continued personification of ‘the people’ problematic. While Holmes doesn’t specifically defend liberal democratic capitalist constitutionalism, his reliance on Madison’s arguments suggests that constitutional pre-commitment might be particularly useful in societies that have internalised the Western idea of ‘civilisation’. We need to constitutionally pre-commit as ‘the people’, not just to take up any kind of loan, but rather those that one constitutional generation will not be able to repay within its 20-year, Jeffersonian, lifespan. Big loans are needed to undertake large-scale, but spatially, or demographically focalised projects of socioeconomic development. For those societies that do not share such visions of socioeconomic development, the stark R/A binary will perforce be less compelling. While in the existing world of foundational constitutionalism this objection is relatively harmless, recent economic crises in the global (semi-)peripheries shed light on the dark side of the virtuous link between pre-commitment and the ‘useful effort of the steady industry’. In countries such as Greece, the fiction of people’s pre-­committment provides moral(istic) support to the arguments for fiscal austerity, which demand unfair long-term painful sacrifices from the most vulnerable sections of the population which have neither had meaningful influence over the course of government’s economic policy, nor have they profited from the corruption among the bureaucracy and the ruling elite. Unless we adopted the Sieyèsian understanding of ‘the people’, people’s personhood has a potential to systematically shortchange the poor, by moralistically, but politically unproductively, impelling them to internalise the idea that they, together with everyone else, are in the same ‘boat’ together. While ‘extending the moral dimension of any explicit political decision[]’ may cut both ways, the same risk plagues Dworkin’s defense of political community’s personification. In addition, the assumptions behind it leave national minorities within existing states without the theoretical apparatus that would enable them to challenge the territorial status quo, without succumbing to ethno-nationalist visions of popular self-determination. In defending the idea of personification, Dworkin exclusively focuses only on ‘United States and Britain’, treating the question of political boundaries among different nations as a technicality. While the existence of fraternal, associative political obligations beneath Dworkin’s legitimate political community leaves the door ajar for the dismemberment of the existing collective persona of ‘the people’, Pettit has, as we have seen, vehemently rejected any such possibility. This rejection, however, is problematic not only because he wrongly assumed that democratic theory has remained silent in front of

24 

ibid 190.

Constitutional Theory and Cognitive Estrangement 59 the paradox of constitutionalism,25 but also because he didn’t provide reasons as to why accept the general conceptual assumption of the predatory outside, as the political truth in concrete radical political conflicts that put in doubt the identity of ‘the people’. Concretely: why would the Catalans or the Quebecois have to accept their ‘tough luck’, if the larger context of the EU within which their secession would occur is decidedly not predatory. By the same token, why would we have to accept Pettit’s binary between general obedience and obedience a-la-carte when negotiating towards secession—within an existing constitutional order, as envisaged by the Canadian Supreme Court—would have violated no one’s political and legal equality. In giving us a richer defense of Lockean Ulysses, the critique of Holmes’, Dworkin’s and Pettit’s defense of ‘the people’s’ personality also carries ethical, constructive, and re-constructive implications. First, the critique hazarded above can be understood as a wishful disruption of the dominant ‘social imaginary’ and the grassroots political ethos, based on it. By engaging in cognitive estrangement, and by charting the ways in which this estrangement is combated in theory, we would contribute to a more hesitant acceptance of the vocabulary of peoplehood, understood not simply as a sum of political and moral aspirations, but also as a political ideology animated by disciplinary anxieties of foundational constitutionalism. By stressing who may stand to lose from Holmes’, Dworkin’s and Pettit’s embrace of people’s personality, we are better positioned to understand peoplehood as a device for narrowing ‘the credibility gap’ between the governors’ assertions of legitimate authority, and the fragile belief ‘in the regime’s legitimacy held by its subjects’.26 IV.  TERTIUM DATUR: MAPPING CONSTITUTIONAL CHANGE BETWEEN THE REVOLUTION AND THE AMENDMENT

In this section, however, I wish to focus on the second, constructive, implication of the engagement with Holmes, Dworkin and Pettit, while leaving the third implication for the conclusion. The critique of their reasons for personification of peoplehood has shown that our embrace of the R/A binary will depend not only on whether we find the stylisations of peoplehood compelling, and not only on our contingent positions in concrete constitutional struggles, but also on the price we are otherwise willing to pay for the putative benefits of collective peoplehood. In other words, making conceptual room between the revolution on the one side, and the amendment on the other, will depend not only on whether we find the idea of Ulysses, or docile Lockean rebel attractive, but also on whether we are persuaded by Holmes’, Dworkin’s and Pettit’s conceptual assumptions, normative aspirations and prudential warnings all of which inform their commitment to the idea of popular sovereignty. This is important to stress because recent attempts to collapse the R/A binary start from definitional commitments to a particular idea of ‘the people’ and its pouvoir constituant, and proceed without a finer grained mapping of the promises and perils

25  26 

Pettit (n 16). M Loughlin, ‘The Constitutional Imagination’ (2015) 78(1) Modern L Rev 1, 14.

60  Zoran Oklopcic of peoplehood in its different contexts. The way in which Colón-Riós collapses the R/A binary is simple: constituent power of ‘the people’, read in a democratic key, should mandate radical political openness, and, widest possible direct popular participation in the process of constitutional change. During this process, established constitutional forms shouldn’t enjoy precedence: irrespective of amending procedures, citizens should be allowed to propose, deliberate and decide on a new constitution, unconstrained by strictures of the existing constitution.27 Such, ‘weak’, constitutionalism is pitted against another influential contemporary project of undermining the R/A binary, Bruce Ackerman’s ‘constitutional moments’. Ackerman’s argument is well known and doesn’t require much exposition: For him, the heightened degree of political mobilisation in the United States in the late 1860s and 1930s has led to constitutional ‘moments’—profound extra-constitutional political changes which amounted to a de facto constitutional amendment.28 Both during the reconstruction in the late 1860s, and during the New Deal in the 1930s, mobilised political deliberations over contentious political issues led to a constitutional impasse, which was broken by a decisive electoral victory won by one of the branches of government, which led her to claim the mandate to pursue radical constitutional change. In both cases, the meaning of such electoral victory was disputed by the branch defending the constitutional status quo. The branch that claimed the decisive electoral mandate constitutional autonomy was then challenged, forcing it to reckon whether or not to endure in its challenge of a ‘decisive electoral mandate of another branch of government’. The result in both cases was the ‘shift in time’ where the dissenting branch called off its resistance and acceded to the interpretation of the branch that initially claimed popular mandate for itself.29 Ackerman’s model has been both remarkably influential, as well as subject to sustained critiques about its historical accuracy. What is more important for our purposes is to emphasise that while constitutional moment undermined the R/A binary, both Ackerman’s innovation—as well as ‘weak’ constitutionalism that Colón-Riós offered as its alternative—have continued to operate within the monolithic conception of ‘the people’, and its will. In both cases, concrete political contexts devoid of problem-generating radical diversity helped in in upholding the monolith. In Ackerman’s case, the contingent facts of the defeat of the Confederation in 1865, and, by the 1930s, solidified Pan-American national consciousness, have critically contributed to the portrayal of the radical political conflicts as the conflicts over the meaning, not the existence of the unified people’s will. Equally, the factual absence of politically salient fragmentation within Venezuela and Ecuador helped Colón-Riós assume the identity of ‘the people’ as the bearer of pouvoir constituant. At this point, however, the second act of cognitive estrangement becomes critical for reinterpreting the hidden commonality between weak constitutionalism and constitutional moment, and—more importantly—for setting the stage for a richer understanding of the conceptual space between the revolution and the amendment. 27 

Colón-Ríos (n 3). B Ackerman, ‘Higher Lawmaking’ in S Levinson (ed), Responding to Imperfection: The Theory and Practice of Constitutional Amendment (Princeton University Press, 1996). 29  ibid 78. 28 

Constitutional Theory and Cognitive Estrangement 61 If constitutional order—and not ‘the people’—is a person, then Colón-Riós’ and Ackerman’s constitutional order establishes itself as in principle responsive to the demands for radical constitutional change posed out of prescribed constitutional turn. What is more, such constitutional order establishes itself as tendentially responsive, towards the satisfaction of popular desires. Further to this, such constitutional order, as I suggested in section 1, shows that it in working towards the satisfaction of popular aspirations it is capable of accelerating the process of constitutional change. Traces of that are visible in Ackerman’s account of constitutional moments. In worrying about the possibility that constitutional order might remain unmoved— producing ‘false negative’ answers to authentic expressions of ‘the people’s’ will— and by embracing an account of human psychology which starts from a spasmodic character of individual political mobilisation, Ackerman comes close to recognising responsiveness and its speed as constitutional ideals in their own right. If instead of ‘tracks’ of constitutional change we imagined ‘gears’, and if instead of ‘higher’ and ‘lower’ modes, we saw ‘faster’ and ‘slower’ speeds of constitution making, we would be compelled to admit that the accelerated pace of responsiveness must be understood affirmatively for constitutional moments (and weak constitutionalism) to make sense. However, as Sujit Choudhry rightly noted, Ackerman’s template confronts difficulties in the contexts where constitutional change proceeds without an underlying agreement about the identity of ‘the people’.30 Still, for Choudhry, ‘the Secession Reference is best understood as part of a constitutional moment—an extralegal move to amend the Canadian Constitution to add a secession clause, prompted by the failure of the rules for constitutional amendment under the Constitution’.31 However, by insisting that the Reference can still be seen as part of a constitutional moment, Choudhry has neglected how the Supreme Court made tendential, accelerated, responsiveness the centrepiece of its judgment. This, in turn, presents the Reference not only as an alternative to Ackerman’s constitutional moment, but also a challenge to the idea of the sovereignty of a unified people itself, irrespective of the lip-service the Court paid to it in the judgment itself. What makes the Secession Reference exceptional are neither the invocation of the four unwritten principles, nor its description of the interplay between the principles of federalism and democracy.32 What is truly innovative, however, is the way in which the Court linked the idea of the consent of the governed with the duty of responsiveness on behalf of the constitutional order to be responsive to its changes. Rather than treating consent as a counterfactual heuristic device like many contemporary constitutional theorists, the Court stated that ‘[t]he continued existence and operation of the Canadian constitutional order cannot remain indifferent to the clear expression of a clear majority of Quebecers that they no longer wish to remain in Canada’.33

30 S Choudhry, ‘Ackerman’s higher lawmaking in comparative constitutional perspective: Constitutional moments as constitutional failures?’ (2008) 6(2) Int’l J Const L 193. 31  ibid 221. 32  For a discussion see Z Oklopcic, ‘The Migrating Spirit of the Secession Reference in Southeastern Europe’ (2011) 24(2) Canadian Journal of Law & Jurisprudence 347, 365–76. 33  Reference re Secession of Quebec [1998] 2 SCR 217, para 92.

62  Zoran Oklopcic The Court concluded that doing so ‘would amount to the assertion that other constitutionally recognised principles necessarily trump the clearly expressed democratic will of ‘the people’ of Quebec’.34 Elsewhere I have argued that the latter invocation of the Quebecois peoplehood obfuscates the theoretical importance of the Reference.35 I have argued that what animates the Court’s opinion is its commitment to the improvement in the aggregate satisfaction of individual constitutional attachments, not the right of ‘the people’ of Quebec to determine its destiny.36 I have argued that while the majority of ‘the people’—the population—of Quebec has a right to trigger the process of constitutional change, it cannot unilaterally determine neither the territorial nor personal sphere of independent Quebec’s jurisdiction. The important thing is that a segment of the population can trigger the process of constitutional change, which should then unfold towards the satisfaction of their constitutional demands. By positing the requirement of ‘good faith’, the Court can also be interpreted as suggesting that the tempo of constitutional negotiations should correspond to the urgency of the democratic demand for secession, or, at the very least that constitutional negotiations should not arbitrarily slow down, without consequences for the legitimacy of the Canadian constitutional order. The contrast with Ackerman’s constitutional moment cannot be starker: while constitutional moment arises from the game of chicken between different branches of government, Secession Reference imagines constitutional order as radically, and, what is more, tendentially, responsive to the changes in the degree and constellation of political allegiance. While constitutional moments reveal the will of the united people, and can be understood as moments of nation-building, the Secession Reference responds to the desires of a segment of the entire Canadian population, by providing the framework for nation-dismantling. Finally, constitutional moments leave the Constitution’s territory intact, while the Reference portrayed the Canadian constitutional order as spatially responsive. Treating the Secession Reference as a challenge to Ackerman’s constitutional moment, in a different way from the one proposed by Choudhry, is important because it leads us to the critical analytical insight. What Secession Reference forces us to recognise is that tendential responsiveness can appear not just in one, but also in three registers. In other words, what otherwise passes under the name of people’s will is a constellation of spatial, personal and temporal registers of tendential responsiveness. In principle, and sometimes in practice, constitutional order can also speed up to satisfy the aspirations not only of the whole (territorial) people, but also those of one of its territorial or non-territorial segments. The constitutional change may result in spatial or substantive responsiveness towards the desires of a segment of a population. At this point, however, it is important to insist on the scope of radical responsiveness. Intense political desires are relevant in all three, temporal and spatial, 34 ibid.

35  Z Oklopcic, ‘The Anxieties of Consent: Theorizing Secession between Constitutionalism and SelfDetermination’ (2015) 22(2) International Journal on Minority and Group Rights 259. 36 ibid.

Constitutional Theory and Cognitive Estrangement 63 personal domains of responsiveness. By circumventing Part V of the Constitution Act 1982, secessionists can trigger and effectively speed up the process of constitutional change towards the spatial reconfiguration of Canada according to their desires. Their desire is legitimate, not because they of their substantive plans of what to do with independent Quebec, but simply because spatial reconfiguration (of the prescribed provincial magnitude, and under stipulated circumstances of ‘clear majority and clear question’) of Contemporary Canadian constitutional jurisprudence, however, does not make such tendential responsiveness available neither to imaginary socialists demanding negotiations over abolishing capitalist economy, nor to environmentalists demanding that Canada re-commits to Kyoto protocol, nor to, finally, Aboriginal peoples demanding radical reconstitution of Canada. Having resorted to a narrative of Canadian constitutional history and relevant jurisprudence, the Court has not only insulated secessionist demands from other radical demands, but has further restricted them to province-wide majorities demanding secession. In the case of the Secession Reference we have seen that the spatial and personal register of tendential responsiveness are inextricably intertwined. Spatial responsiveness is the result of the personal responsiveness, and the personal responsiveness is the result of no concrete, specified constitutional project. The Court deemed the secession of Quebec a legitimate project, not because it was convinced that it is necessarily a good idea, but simply because enough Canadian citizens (would) have revoked its consent to the Canadian constitutional order. However, there are historical examples that show us how constitutional orders can, in principle, be responsive to the segmental demands which have general constitutional implications, not in the name of the entire people, but rather in the name of political desires of that segment itself, in a way in which such radical constitutional responsiveness does not carry with itself permanent territorial consequences. To help us imagine such radical constitutional responsiveness we must travel further in time, and consider the phenomenon of secessio plebis, periodic constitutional crisis in ancient Rome. Rather than a territorial ‘secession’ in a contemporary sense, the first secessio plebis was a combination of an Occupy Movement and a general strike. In 494 BC, disenfranchised plebeians abandoned the confines of Rome, and encamped on the adjacent Mons Sacer, vowing to remain there until their fundamental demands were met. Their purpose was not secession, the extraction of concrete material benefits, or the establishment of a new constitutional order—but rather radical constitutional reform that would create new institutions (tribunate) giving them more power and, by implication, more equality vis-à-vis ruling patricians. The constitutional moment they provoked was not effectuated in the name of, (or by), Ackerman’s ‘We the people’, but rather in response to the demands of plebeians, those who stood outside (both literally and conceptually) of the political body. As a form of extra-constitutional constitutional change distinct from constitutional moments, and popular constituent power, secessio plebis has both insti­ tutional and conceptual implications. John McCormick has, for example, suggested that modern constitutional orders could find inspiration in the office of tribunes,

64  Zoran Oklopcic which would have a right to veto legislation, in the name of protecting the interests of the downtrodden, plebeian sections of the general population.37 Such institutional innovation, however, comes at a price. According to Miguel Vatter, for secessio plebis to be coherent, we would have to abandon the republican idea of a unified people as the body of co-equal individuals that lives under the same constitutional framework.38 Rather than heeding Dworkin’s advice to personalise political community in order to be able to conjure political equality among all of its citizens, we should keep the plebs separate from the populus as the condition of their ‘constituent power to make equal law’.39 Splitting the unified people between two camps (rich and poor, powerful and powerless in this case) is the conceptual condition for reimagining constitutional order as temporally and substantively responsive to radical, segmental political demands. I conclude the project of dissolving the R/A binary and nuancing the forms of extra-constitutional change by discussing ‘existential’ exception, proposed by Hans Lindahl. While in the previous cases, constitutional change occurs at a macro scale, applying to the entire population and to the entire territory, or at least to their large discernible segment, (progressive) ‘existential’ exception applies at the level of the individual citizen itself. The change in this case doesn’t result in new constitutional provisions (14th Amendment) political institutions (tribunate) or the inscription of new fundamental rights. Rather, the change in question is the silence or a seemingly misdirected reaction of the Constitution in response to the behaviour that is otherwise understood as unconstitutional. Like with previously discussed cases, in order to restrict itself (this time at the micro level) constitutional order needs neither revolution nor constitutional amendment. Unlike Vatter and McCormick, however, Lindahl undermined the R/A binary not by abandoning the figure of ‘the people’, but rather by drawing our attention to something that Ackerman’s account swept under the rug, by simply accepting the results of historical contingencies behind the (re-)founding moments of the ­American republic. While the secession of the Southern states and their subsequent re-­incorporation into the Union can be narrated as part of a larger debate about the moral identity of the American people, these historical periods are also manifestations of a deeper conceptual difficulty that concerns the legitimate scope of the American demos. In approaching this question through the lens of the debates between compact and nationalist theories of federation, most constitutional scholars have under-­ appreciated the extent of the challenge posed by the question about the legitimate scope of democratic unit. Rather than presenting itself in the context of two-level compound polities, this challenge extends all the way through.40 Rather than

37 

JP McCormick, Machiavellian Democracy (Cambridge University Press, 2011) 183. Vatter, ‘The quarrel between populism and republicanism: Machiavelli and the antinomies of plebeian politics’ (2012) 11(3) Contemporary Political Theory 242. 39  ibid 244. 40  Think of the recursive secession of West Virginia in 1861, for example. 38  M

Constitutional Theory and Cognitive Estrangement 65 a­ ccepting that this question is settled by contingent historical events, Lindahl sees conflicts over the identity ‘the people’ as insuppressible and endemic.41 Irrespective of the ways in which constitutional ‘closure’ occurs, the legitimacy of new polity will always remain tainted by constitutional over- and under-inclusion.42 As a result, realising that the Constitution of ‘the people’ always carries traces of heteronomy, should impel the members of a polity not to deny the existence of this collective subject (such as in Vatter and MacCormick), but rather to understand it as unavoidably ‘strange’.43 By implication, this wished-for cognitive distance has direct political consequences: in being aware that our collective existence is ‘strange’, ‘collectives ought to acknowledge that they have a normative blind spot which they can neither fully justify nor remove, and they ought to take this into account when responding to a-legality in the process of setting legal boundaries’.44 Welcome cognitive dissonance that arises out of the recognition of this blind spot should lead polity’s officials to exercise ‘collective self-restraint’ in confronting unconstitutional challenges. Such a principled stance can justify constitutional strategies, which Lindahl—recognising his debt to Carl Schmitt—dubbed existential exceptions. While Lindahl appropriated Schmitt’s affirmative account of exceptions, he turned Schmitt’s argument on its head. Rather than being used to suspend constitutional order in order to neutralise internal enemies, they are carved out in order to ‘preserve and sustain’ radical challenges to constitutional order which constitutional order itself cannot code as intelligible.45 Such exception will occur as the result of legal acts that are always at the disposal of legal officials, but which are generally not perceived as devices, or are—worse still—seen as aberrations from the ideals of constitutionalism and the rule of law. Such existential exceptions can be carved out in a myriad of ways: by not initiating criminal proceedings and waiting for the statute of limitations to pass, by strategically invoking political questions doctrine, or simply by refraining to declare unconstitutional behaviour unconstitutional. Lindahl’s perspective on peoplehood not only undermines the R/A binary, but also allows us to understand the quotidian behaviours of legal officials—those that we would otherwise criticise as deviations from the ideals of the rule of law and constitutionalism— as mechanisms of constitutional change. Drawing attention to the way in which we may re-envision quotidian legal acts as mechanisms of constitutional change brings us to the question of perspective in our attempts to complicate, and ultimately dissolve the R/A binary. Before tackling

41  I’ve emphasised the constitutional theory here because there are burgeoning approaches in democratic theory, which purport to be able to provide comprehensive arguments about the scope of a legitimate people. I think these approaches are for different reasons mistaken, and that Lindahl is right to stress the irreducible agonism in the context of polity formation. 42  H Lindahl, Fault-lines of Globalization: Legal Order and the Politics of A-Legality (Oxford University Press, 2013). 43 ibid. 44  ibid 249. 45  ibid 251.

66  Zoran Oklopcic this question, here is another, schematic, glance at the conceptual space between the revolution and amendment:

R/A binary

Theoretical approach

The identity of ‘the people’

Registers of responsiveness Spatial Personal (fragmentation) (beneficiary)

Temporal (acceleration)**

Preserved

‘Standard’ foundational constitutionalism

upheld/ unproblematised

no

‘the people’

no

Dissolved

‘Weak’ constitutionalism

upheld/ unproblematised

no

‘the people’

yes

Constitutional moments

upheld/ unproblematised

no

‘the people’

yes

Secession reference

superficially upheld/ disaggregated

yes permanent

the segment

yes

Secessio plebis

explicitly rejected/ fractured in two camps

yes temporary

the segment

yes

‘Existential’ exception

upheld/put in doubt

no

segments and individual citizen

no

**By

‘acceleration’ I understand speeding up the process of constitutional change in a way not envisaged by the constitutional provisions themselves. So, for example, triggering the process of constitutional change through a referendum based on popular initiative would accelerate the process of constitutional change, but would not count as acceleration in the way I am using the term in this essay.

Pausing on the question of perspective is important because it impels us to confront the political, concrete productivity of providing the taxonomy of extra-­ constitutional changes that have complicated, and ultimately undermined, the R/A binary. This problem doesn’t present itself if we approached the dissolution of the R/A binary either from the perspective of a theorist, or from the perspective of a legal official. In the first case, the theoretical perspective leaves the object of its inquiry untouched, contributing at most to a more self-aware understanding of the conceptual field within which constitutional change becomes intelligible. In the second case, however, this heightened awareness of what may (should) count as constitutional change, may engender mutations in the officials’ so-called internal point of view’. Without altering their patterns of justification that rely on foundational constitutionalism, such officials could calibrate their concrete arguments so as to contribute to, say, acceleration of constitutional order’s responsiveness towards the radical demands of discrete segments of population. Finally, beyond the theorists and the officials, the dissolution of the R/A binary speaks also to grassroots, radical challengers to constitutional status quo. Scholars

Constitutional Theory and Cognitive Estrangement 67 most critical to (republican) foundational constitutionalism, such as Emilios Christodoulidis, have suggested that they engage in ‘strategies of rupture’ and ‘identify a spectrum of possible [radical] political interventions in relation to law, rather than under its auspices’.46 Within that spectrum there is a need for ‘a strategic decision … whether to play the system or to confront it’.47 In Christodoulidis’ enigmatic phrase, strategies of rupture call for a ‘militant attention to the points of tension upon which the management of consensus depends, and the introduction of a ‘heterogeneity’ or incongruence capable of generating and sustaining itself against the management of consensus and the order of representation that it serves’.48 However, Christodoulidis’ suggestion to move beyond the rebellion and reformism in imagining avenues of radical political change presents us, then, with the question of productivity of the analytical dissolution of the R/A binary for such political movements. The crucial question is whether the analytics that went into the dissolution of the R/A binary can cash out not only in the register of perception and self-awareness, but also discursively, as a political demand. In other words, is it possible to demand acceleration of responsiveness, spatial reconfiguration, or the creation of new, partisan, constitutional institutions without the mediating vehicle of sovereign peoplehood, and without demanding constitutional amendment? If the three registers of responsiveness are the suppressed analytical truth behind popular, foundational constitutionalism, can they do argumentative work, without references to either peoplehood, fundamental rights,49 or some over-arching justicebased argument? Why shouldn’t various movements simply perform ‘exhaustion with suffering’ in the pursuit of their radical agenda, as they have done so many times in the past? Put differently, strategies of rupture may choose to ignore their own discursive implications. The question, however, is can they: existing ‘in relation to [constitutional] law’, these strategies will constantly be provoked to explain it, and in the process, explain themselves. It is worth stressing three questions that will continue to haunt them. The first question concerns patience. At the ‘meta-level’ of the strategy of rupture, the partisans of rupture will decide between provoking oppression, and setting the stage for the objective exhaustion with suffering, or performing exhaustion with constructed suffering. Pausing on the trope of patience should make us wonder about a third possibility: embracing impatience as a legitimate constitutional sentiment, and demanding the acceleration of responsiveness, beyond the rhythms of constitutionalised democratic politics. Only under this assumption does Slavoj Žižek’s oftenrepeated strategic advice make sense: The lesson here is that the truly subversive thing is not to insist on ‘infinite’ demands we know those in power cannot fulfill. Since they know that we know it, such an ‘infinitely

46 

E Christodoulidis, ‘Strategies of Rupture’ (2009) 20 Law & Critique 3.

47 ibid. 48 ibid.

49  For an influential account of the intimate relationship between the vocabulary of rights and popular sovereignty in the context of the struggles for social emancipation see C Lefort, The Political Forms of Modern Society Bureaucracy, Democracy, Totalitarianism (The MIT Press, 1986) 261–62.

68  Zoran Oklopcic demanding’ attitude presents no problem for those in power: ‘So wonderful that, with your critical demands, you remind us what kind of world we would all like to live in. Unfortunately, we live in the real world, where we have to make do with what is possible.’ The thing to do is, on the contrary, to bombard those in power with strategically well-selected, precise, finite demands, which can’t be met with the same excuse.50

Importantly, however: it is not that those ‘precise, finite demands’ cannot be met in reality, but rather that they cannot be met soon. In other words, an intelligent foundational constitutionalist (or a republican adherent of popular sovereignty) will only argue that such demands have to be processed in the naturally slow rhythm of democratic politics and constitutional change.51 So should they choose to speak back, strategists of rupture would need to argue two things. First, that the acceleration of responsiveness is an ideal already implicit in the vocabulary of peoplehood and foundational constitutionalism. And secondly, that patience is not simply as a positive virtue but also as a negative response. The demand to be patient is a refusal to (continue) explain(ing) why the process cannot be sped up, concealing the arbitrary budgeting of political energy of those in power. The second question concerns face. Constitutional theorists have long drawn attention to how personation—the mask of ‘the people’, ‘commonwealth’, ­‘sovereignty’—enables political and legal agency.52 But we should also think of the ‘face’ of constitutional order literally. Spatiotemporal anxieties of fragmentation and fickleness rely on an assumption that constitutional order won’t be credible, once it has made an exception. Republicans, such as Pettit, don’t make that argument explicitly, but those who seek to confront the ‘tough luck!’ dismissal, must be ready to engage those anxieties constructively. Christodoulidis’ injunction to ‘play[] the system’ here doesn’t only mean a decision of whether or not to engage in legal argument, but more expansively, it gestures towards the rhetoric of how to help constitutional order save face, and diminish its own anxieties, after radicals have forced it to make a Lindahlian ‘existential’ exception. The third question concerns affectedness. A spatiotemporal ‘provocation’ such as secessio plebis challenges the understanding of ‘the people’ as the body of all those who are ‘in it together’. But here the analogy between secessio plebis and any such present-day strategy of rupture, breaks down. Plebeians actually left the territory of the state, and weren’t part of the Roman political body, to begin with.53 In contrast, the contemporary 99 per cent are part of the larger ‘people’, and, as a result, have to fight counter-charges of NIMBY-ism, irresponsibility, and socio-economic parasitism—the charges predicated on an idea that satisfying their localised demands must come at the expense of other fellow citizens’ well-being.

50 

S Žižek, ‘Resistance is Surrender’ (2007) 29(22) London Review of Books 7. cf W Scheuerman, ‘Constitutionalism in an Age of Speed’ (2002) 19 Const Comment 353. Scheuerman claims that ‘the age of speed’ has ushered in an appetite for executive action and that it disrupts the slow rhythms of democratic politics. Rather than mourning the acceleration of constitutional politics, we should be attentive to how the hegemony of liberal-democratic constitutionalism continues to rely on the idea of a slow pace of democratic political change. 52  For a discussion of the face, ‘persona’ and personation in the context of Hobbes’ political thought see D Runciman, Pluralism and the Personality of the State (Cambridge University Press, 1997). 53  Vatter (n 38) 256. 51 But

Constitutional Theory and Cognitive Estrangement 69 If they wanted to engage those counter-claims discursively, strategies of rupture would face two theoretical avenues. On the one hand, they could pursue a neo-­ Sieyesian route, and construct a polit-economic complement to a radical constitutional imaginary that would dignify the figure of the plebs, the poor, or the excluded. What gave force to Sieyes’ pamphlet were not three sentences on the meaning of constituent power that most constitutional theory quotes today, but an elaborate socioeconomic argument supporting the struggle of the then ‘99 per cent’. On the other hand, they could try to demask the very idea of popular collective self-government that continues to put the demands of those disenfranchised in its place. But here comes a crucial difference. To re-enact secessio plebis today, we would need to do it in the name of a principle other than either ‘the people’s or the plebs’ ‘constituent power’. In giving dignity to the strategy of rupture in relation to the idea of a constitutional order—this time without an overarching account of political economy—constitutional theory would need to challenge the idea of ‘being in it together’, and perforce the idea that socioeconomic policies, conducted in the name of ‘the people’, affect everyone equally. In doing so, it would need to find a way to speak to those who find Dworkin and Pettit persuasive, constructively: that is, construct imaginative ways which would maintain equal dignity of everyone who belongs to a polity, while at the same time selectively disenfranchising those who use the vocabulary of peoplehood and affectedness—as the ultimate line of defense—to protect their visible and invisible, concrete and diffuse, socioeconomic privilege. V.  TOWARDS A DIFFERENT FAMILIARITY: ‘THE PEOPLE’, THE PARADOX AND THE SACRIFICE

Perhaps this can be done by embracing an image that clearly broadcasts the fact that ‘the people’ is a fictional shorthand for a bundle of moral and prudential imperatives.54 Such an account of peoplehood might be capacious enough to embrace the legitimacy of the impatient desires of different minorities that cannot otherwise be rendered intelligible by invoking ‘the will’ of a sovereign people, or its constituent power. Likewise, such an account might allow us to continue enjoying the benefits of the imagination of collective agency—useful not only for the ethical purposes described by Dworkin—but also for more prosaic purposes we have otherwise learned to take for granted. As Raino Tuomela argued, [t]he value of the group-level description of group agents lies largely in its economizing and systemizing, but also its explanatory value (concerning, e.g., groups’ actions and members’ collective action), in general its epistemic and practical advantages. A group-level description makes it much easier to discuss interaction and interdependence between various macro social groups and social structures … Such a group agent basically has the capacity

54  For one such recent proposal, see M Tushnet, ‘Peasants with pitchforks, and toilers with Twitter: Constitutional revolutions and the constituent power’ (2015) 13(3) Int’l J Const L 639.

70  Zoran Oklopcic to act as a unit; while a citizen-level description of its particular activities would be an overwhelming task, a group-level description is viable.55

I stress this because the aim of this essay has not been to advocate for abandoning the concept of the (sovereign) people, nor for the rejection of any of its attributes, such as constituent power or self-determination. Rather, in calling for the estrangement from the vocabulary of peoplehood, my aim was to call for a renewed imaginative effort that would both lend constitutional intelligibility and political dignity to the sentiments expressed by Williams in the epigraph to this chapter, and that would contribute to finding ways of inscribing that dignity into whatever figures end up mediating our ‘popular’ expectations. By the same token, however, in this chapter I also found it productive to radicalise this cognitive estrangement in the hope of making three theoretical implications of doing so sharper. The first concerns the problem that most constitutional theorist today call the paradox of constitutionalism. With the benefit of the preceding discussion, the so-called paradox ought to be seen as the product of a conscious, or semi-conscious decision that the alternatives to the vocabulary of peoplehood (that generated it in the first place) are not, for whatever reason, acceptable. Seen from this perspective, ‘the people’ should be seen neither as Ulysses, nor as a temporary occupant of the ‘empty space of power’, but rather as that imaginative and institutional leftover space, delineated by our conscious refusals to embrace the alternatives, some of which I outlined in the preceding section of the chapter. In other words, the affirmations of the paradox’s existence should not be seen as inconvenient conceptual truths of modern constitutional theory, but rather as symptoms of unreflective intimations, or half-considered decisions, that the price for abandoning ‘the people’ is too steep, or that the benefit of doing so is too small. Secondly, confronting the defenses of peoplehood with its alternatives also unearths another suppressed dimension in the pragmatics of its invocation. Simply put, ‘the people’, its sovereignty, rights and constituent power make more sense in certain contexts, less so in others. In Dworkin’s ‘United States and Britain’, assuming the moral personhood may be fruitful in all these numerous cases where the bonds of friendship can be taken for granted, and where extending our moral perspective is helpful in overcoming ‘the torment of heteronomy’56 involved in legal and political arbitration of momentous, but quotidian ethical conflicts. Abortion, equitable taxation, vote suppression, gay rights: it may well be that it is prudentially and morally beneficial to assume the aspirational perspective of the long popular durée among all those whom we have sentenced to sharing the same constitutional ‘boat’. On the other hand, in extraordinary conflicts—those that concern the nature and existence of a polity—republican insistence on sovereign people is, to put it bluntly, unconvincing, but in the insistence on its credibility infuriating, shell game. The insistence on it is either self-defeating (by adopting assumptions such as Dworkin’s fraternal obligations), or entails a caricatured vision of external environment and the postulation

55 

R Tuomela, Social Ontology (Oxford University Press, 2013) 22. H Kelsen, ‘On the Essence and Value of Democracy’ in AJ Jacobson and B Schlink (eds), Weimar: A Jurisprudence of Crisis (University of California Press, 2000) 84. 56 

Constitutional Theory and Cognitive Estrangement 71 of crude binaries between general obedience and inequality-generating exceptions (Pettit). The insistence on sovereign peoplehood is not only surprisingly disrespectful towards those who want to radically change their political status peacefully (‘Tough luck!’, says Pettit), but fails to register moral costs of doing so, as it pushes those who want to establish a radically different constitutional order towards revolution. It encourages nationalists to continue insisting on their sovereign peoplehood since this is the only language in which modern constitutionalism may eventually understand them. Moreover, it encourages them, together with other radicals, to stage their oppression—perform their exhaustion with suffering—all in order for their project to be morally and conceptually legible to modern constitutionalism. Finally—and this is my last point—it encourages them to offer evidence for the credibility of their assertions through sacrifice. Against culturalist approaches to constitutional theory (such as those of Paul W Kahn), self- and other- sacrifice should not be seen as part and parcel of some deep theological infrastructure of modern popular sovereignty.57 Rather, political sacrifice should be seen as a functional, quite secular, imperative of modern constitutionalism, stemming from its unwillingness to move past the revolution/amendment binary. We cannot accelerate; we cannot be responsive to your desires, not only because doing so would make you more ‘equal’ than your co-citizens, but also because your desires may be unauthentic, their motive exploitative, and your claim of being exhausted with oppression, fabricated or premature. Revolutionary sacrifice is a secular ordalium—a functional response to this epistemic anxiety.

57 

See PW Kahn, Putting Liberalism in its Place (Princeton University Press, 2005).

72 

3 Constraints on Constitutional Amendment Powers ORAN DOYLE* changed, changed utterly A terrible beauty is born. WB Yeats, Easter 1916

I. INTRODUCTION

T

HERE EXISTS A widespread practice of constraining constitutional ­amendment powers through rules or standards that determine the validity of constitutional amendments. These rules and standards arise both from ­constitutional texts and judicial decisions.1 This practice raises significant concerns about the distribution of power both between generations and between political actors. The fundamental question is whether the values served by constraints on amendment powers can justify a contemporary majority being subject to a past generation or a judicial elite. The literature mostly fails to address this question directly, focusing instead on the quantum of change introduced by constitutional amendments. Eliding different understandings of the word ‘constitutional’, problematic amendments are identified by the extent to which they depart from the moral value of constitutionalism, or the extent to which they go beyond what is claimed to be a proper conceptualisation of constitutional amendment, or by some combination of these approaches. In this chapter, I suggest a different approach. Four cross-cutting distinctions establish a typology of ways in which the power of constitutional amendment can be constrained. This typology focuses attention on how constraints disempower

*  I am grateful to the conveners and participants of the symposium on constitutional amendment in Boston College on 19 May 2015 and in particular to Mark Tushnet who presented the draft of this chapter for discussion. I am also grateful to the anonymous reviewers as well as to Aileen Kavanagh, Madhav Khosla, Yaniv Roznai and David Prendergast who read and provided most helpful comments. All constitutional provisions are cited from . 1  For a comprehensive and insightful account of this practice, see Y Roznai, ‘Unconstitutional Constitutional Amendments: The Migration and Success of a Constitutional Idea’ (2013) 61 Am J Comp L 657.

74  Oran Doyle contemporary majorities in favour of past generations or judicial elites, thus providing a baseline against which we can assess whether those constraints are justified. Constraints that seek to serve majoritarian values pose the relatively simple question of whether the detriment to majoritarian values in the here and now is justified by the protection of majoritarian culture over the medium to long term. Constraints that prevent change to the fundamental features of the polity are unjustified: there can be no objection to the current generation using constitutional processes to transform its polity. Constraints that seek to preserve counter-majoritarian values are the most difficult to assess, raising as they do a competition between the incommensurable values of majoritarian democracy and the protection of minorities from unjust laws. II.  A DOCTRINE OF UNCONSTITUTIONAL CONSTITUTIONAL AMENDMENTS

A.  Unconstitutional Amendments: Positive, Moral and Conceptual Claims The literature largely revolves around the question of whether particular legal systems have or should have a judicial doctrine of unconstitutional constitutional amendments, as well as the appropriate parameters of such a doctrine. This literature obscures and elides critical differences between positive, moral and conceptual senses of unconstitutional constitutional amendments. In the positive sense, an amendment is an unconstitutional amendment if it fails to meet standards posited by the constitutional text. In the moral sense, an amendment is an unconstitutional amendment if it fails to respect the value of constitutionalism, understood as the value of constraining governmental power.2 In the conceptual sense, an amendment is an unconstitutional amendment if it purports to make a change that falls outside the concept of constitutional amendment. Recognition of these different senses allows us to make sense of the otherwise self-contradictory ‘unconstitutional constitutional amendment’: an amendment may be constitutional in one sense but unconstitutional in another sense. Positively unconstitutional amendments are nearly always considered to be invalid. Indeed, it is linguistically unusual—although not senseless—even to refer to a positively unconstitutional amendment as a constitutional amendment at all.3 The claim of those who argue for a doctrine of unconstitutional constitutional amendments is that an amendment, regardless of what the constitutional text requires, should be deemed positively unconstitutional if it is unconstitutional in the moral sense or the conceptual sense. In other words, constitutional actors (including courts, absent a special rule on non-justiciability) should neither enforce nor respect an amendment

2  There are different understandings of constitutionalism but the one that pervades this literature is the value of controlling government, what Barber terms negative constitutionalism. Nick Barber, ‘Constitutionalism: Negative and Positive’ (2015) 38 Dublin U L J 249. 3  De Gaulle’s use of a referendum to amend the Constitution of the Fifth Republic in 1962 is arguably an example of a positively unconstitutional constitutional amendment that is still considered to be a constitutional amendment. See discussion of Jean-Philippe Derosier ‘The French People’s Role in Amending the Constitution’, in this volume.

Constraints on Constitutional Amendment Powers 75 that is unconstitutional in a moral or conceptual sense, irrespective of whether it is constitutional in the positive sense. The literature framed in terms of ‘unconstitutional constitutional amendments’ is problematic for three related reasons. First, there is a tendency to elide the three different meanings of ‘constitutional’, thereby obscuring the precise claim being made. Secondly, both the moral and the conceptual approaches often use the quantum of constitutional change introduced by an amendment as a proxy for determining whether an amendment is morally or conceptually unconstitutional. As well as further obscuring the difference between moral and conceptual approaches, this move—since lawyers are conveniently well placed to measure that quantum of constitutional change—deflects attention from the extent to which judges are politically empowered by a doctrine of unconstitutional constitutional amendments. Thirdly, the power of constitutional amendment involves moral values that have nothing to do with the value of constitutionalism: it is difficult to situate these values in a debate that focuses on whether a constitutional amendment is ‘constitutional’. B.  Morally Unconstitutional Constitutional Amendments The moral value of constitutionalism could—in principle—provide grounds for constraining the power of constitutional amendment. However, it is problematic to focus exclusively on constitutionalism to the exclusion of other values. This valuemonism leads to an interpretative approach that seeks to construct the best meaning of the moral value of constitutionalism rather than directly address the justification of constraints on the power of constitutional amendment. In this vein, Landau writes that the doctrine of unconstitutional constitutional amendments suggests ‘a more substantive conception of constitutionalism—one that states that a constitution is not really constitutional unless it actually works in certain ways and adheres to certain fundamental principles’.4 Walter Murphy adopts such an interpretative approach, suggesting that amendments that would destroy or cripple the values of constitutional democracy are invalid. Murphy argues that a polity that violates its fundamental principles destroys its justification for existence and its public officials lose their authority to speak as agents of the people. Any change that would transform the polity into a totalitarian or authoritarian political system would be illegitimate, in the sense of not being grounded in the existing system’s fundamental normative principles.5 Murphy here elides two discrete issues: the extent to which the amendment departs from the value of constitutionalism and the extent to which the amendment is inconsistent with the Constitution’s fundamental principles, ie the quantum of change introduced by the amendment. Jacobsohn makes a similar elision in reverse, moving from a quantum-of-change approach to a moral value of constitutionalism approach. He argues for a Burkean 4 

D Landau, ‘Abusive Constitutionalism’ (2013) 47 UCDL Rev 189, 260 (emphasis added). Murphy, ‘Merlin’s Memory: The Past and Future Imperfect of the Once and Future Polity’ in S Levinson (ed), Responding to Imperfection: The Theory and Practice of Constitutional Amendment (Princeton University Press, 1995) 163, 173 and 178–80. 5 W

76  Oran Doyle criterion of constitutional identity to determine whether a constitutional amendment is legitimate.6 An amendment can only be legitimate if the old constitution survives without loss of identity.7 This criterion of identity focuses not on the degree of change per se but rather ‘the content of this change insofar as it implicates the question of constitutional identity’.8 The difficulty here is that no argument is made as to why stability of constitutional identity is a value that should constrain the amendment power. For such an argument to be analytically discrete, it would have to be ambivalent as between benign and malign constitutional identities. But it is impossible to conceive of an argument as to why the power of constitutional amendment should be constrained in order to preserve a malign constitutional identity. For instance, the Nineteenth Amendment to the US Constitution is not objectionable simply because it changed the constitutional identity of the US by allowing women to vote. Nor is it an answer to suggest that the only solution to a malign constitutional identity should be the replacement of the entire constitution. The purpose of a constitutional amendment power is to allow for constitutional improvement; there is no reason why improvements to constitutional identity should be prohibited. Jacobsohn seeks to avoid the problem of malign constitutional identities by drawing on the moral sense of constitutionality. He argues that a rulership that fails to meet the minimum requirements of Fuller’s internal morality of law has no constitutional identity in the first place: ‘certain attributes of the rule of law … are necessary for generic constitutional governance’.9 In this guise, ‘constitutional identity’ is no longer an empirically observable feature of a constitution but rather a moral assessment of whether a positive constitution conforms to the value of constitutionalism. Even as a moral assessment, however, it is lacking since there are many ways in which a constitution’s identity might be malign (for instance, through the legitimisation of chattel slavery) without offending any of Fuller’s desiderata of the rule of law. Although the value of constitutionalism could justify a constraint on the power of constitutional amendment, it is problematic not to consider competing values. Most relevantly, all constraints on the power of constitutional amendment come at a cost to the value of democratic majoritarianism, the right of contemporary majorities to choose their own form of government. Murphy’s suggestion that public officials lose their authority to speak where a polity violates its fundamental principles seems to imply a judicial power to disregard amendments that judges perceive to be illegitimate in this sense. However, that judicial power is not separately justified: it is just taken to follow from the fact that too great a change is being made to the Constitution. The focus on quantum of change—an assessment that is well suited to the skillset of judges—obscures the extent to which judges are politically empowered by this doctrine. Jacobsohn is ultimately reticent about such a judicial power, not for

6  GJ Jacobsohn, ‘An Unconstitutional Constitution? A Comparative Perspective’ (2006) 4 Int’l J Const L 459. 7  ibid 483. 8  ibid 485. 9  ibid 481. This solution is only partial because it fails to grapple with less fundamental ways in which a constitutional identity could be malign.

Constraints on Constitutional Amendment Powers 77 moral reasons but rather due to a prudential concern that judicial intervention might be ineffective or might make things worse by causing people to leave these concerns to the judiciary.10 He also suggests that the degree to which judges should have a judicial review power over constitutional amendments depends on the relative ease or difficulty of altering the document.11 Yap more directly addresses this competition of values. For him, the best justification of constraints is that they may be necessary ‘to prevent a temporary dominant authoritarian political party/coalition from harnessing the amendment process’ to secure its own position indefinitely.12 However, this must be weighed against ‘the dangers of any judicial abuse that may follow from unelected judges enforcing a nebulous “essential features” doctrine that can frustrate legitimate constitutional revisions designed to meet changing times’.13 Despite recognising the competition in values, Yap’s resolution is similar to that of Jacobosohn: for malleable constitutions, judges should interfere with the substance of an amendment only where it would ‘substantially destroy the pre-existing constitution’ by being manifestly unreasonable.14 The criterion of ‘substantial destruction’, however, means that quantum of change (rather than the moral value of constitutionalism) re-emerges as the touchstone of legitimacy. C.  Conceptually Unconstitutional Constitutional Amendments Conceptual approaches claim that, irrespective of what a constitution says, there are certain types of change to a constitutional order that cannot be made through any amendment procedure prescribed by the constitutional text.15 Some conceptual approaches focus on how amendment is different from revision or replacement. For instance, Murphy has argued that the word ‘amend’ cannot mean ‘to deconstitute and reconstitute’ or ‘to replace one system with another or abandon its primary principles’. Such changes, for Murphy, would not be amendments at all, but revisions or transformations.16 But it is difficult to see why the concept of amendment necessarily precludes revisions or transformations. There are at least instances of constitution-drafters understanding the concept of amendment very broadly.17

10 

Jacobsohn (n 6) 486. ibid 487. 12  PJ Yap, ‘The conundrum of unconstitutional constitutional amendments’ (2015) 4 Global Constitutionalism 114, 128. Yap limits his consideration to judicially recognised implicit constraints on amendment powers, but the justification that he identifies is of broader application. 13  ibid 131. 14  ibid 132. 15  I suspect this must rest on a stronger conceptual claim that a constitution necessarily cannot prescribe a procedure for change (whether called amendment, revision or replacement) greater than its own amendment, understood in this limited way. However, as the more contentious issue in the literature is whether ‘amendment’ powers are necessarily constrained, I shall focus on that weaker claim. 16  Murphy (n 5) 177. 17  For instance, Art 46 of the Irish Constitution allows any provision to be amended, by way of variation, addition or repeal. 11 

78  Oran Doyle Although there is a conceptual distinction between replacement and amendment, this undermines rather than supports the claim that some significant amendments should actually be characterised as constitutional replacements: if it is an amendment, it cannot be a replacement. The Columbian Constitutional Court’s constitutional replacement doctrine illustrates the difficulties in drawing any distinction between amendment and replacement at any point short of the formal replacement of one constitution with another.18 The Court now holds that an amendment amounts to a replacement if it replaces an essential element of the Constitution. Although the Court provides some structural guidance for establishing what provisions might be essential, it does not, as noted by Bernal, make clear what an essential element is.19 This reflects the core problem: there is no threshold at which an amendment becomes a replacement before it actually (in the most ordinary sense) replaces the Constitution. Bernal offers a conceptual and normative justification for the replacement ­doctrine.20 His conceptual justification relies on the claim that constitutions, at least in civil law systems, necessarily imply the protection of constitutional rights, the rule of law, and the separation of powers. A power to amend the Constitution cannot include a power to alter its nature. However, this does insufficient work: one could still replace a constitution with a fundamentally different constitution provided that it is still a ‘constitution’. In truth, the force of Bernal’s conceptual justification relies on the force of the normative arguments in favour of the protection of constitutional rights, the rule of law and the separation of powers. These normative arguments cohere with his openly normative justification, namely that amendments are only permissible if they seek to preserve or improve the political system originally chosen by the constituent power, in Columbia’s case one based on deliberative democracy. However, these claims need to be assessed on normative grounds; they are not a plausible elaboration of any distinction between replacement and amendment. Rather than focus on the conceptual limits of amendment per se, some of the literature draws on a radically different intellectual tradition to explore limits on the concept of constitutional amendment. This revolves around the force of popular sovereignty and the distinction between constituent power and constituted power. The idea of popular sovereignty emerged in Britain at the time of the Civil War as a means of grounding political authority in opposition to the divine right of kings.21 Loughlin presents this as the birth of constituent power.22 During the French Revolution, Sieyès introduced the distinction between constituent and constituted

18  This discussion draws on the account provided by Bernal. See C Bernal, ‘Unconstitutional constitutional amendments in the case study of Colombia: An analysis of the justification and meaning of the constitutional replacement doctrine’ (2013) 11 Int’l J Const L 339. 19  ibid 344. 20  ibid 352. 21  ES Morgan, Inventing the People: The Rise of Popular Sovereignty in England and America (WW Norton & Company, 1988) 36–37. 22  M Loughlin, ‘Constituent Power Subverted: From English Constitutional Argument to British Constitutional Practice’ in M Loughlin and N Walker (eds), The Paradox of Constitutionalism: Constituent Power and Constitutional Form (Oxford University Press, 2008) 32–38.

Constraints on Constitutional Amendment Powers 79 power: ‘In each of its parts a constitution is not the work of a constituted power but a constituent power. No type of delegated power can modify the conditions of its delegation’.23 For Sieyès, the nation was the constituent power. As Thomaz Pereira argues in this volume, Sieyès should be understood as making a political argument directed towards particular goals in the context of the French Revolution.24 Carl Schmitt generalised Sieyès’ distinction between constituent and constituted power into a constitutional theory. Schmitt alternates between characterising the constitution-making power as an entity25 and as a capacity born by an entity.26 For Schmitt, the constitution-making power might be (or be borne by) the people or the monarch.27 A constitution is valid because it derives from the will of a constitution-making power or authority; the word ‘will’ denotes an actually existing power as the origin of a command.28 The constitution-making power is ‘the political will, whose power or authority is capable of making the concrete, comprehensive decision over the type and form of its own political existence’.29 This conception of constitutions leads Schmitt to distinguish between the constitution-making power and constituted powers, leading to constraint on all amendment powers: The boundaries of the authority for constitutional amendments result from the properly understood concept of constitutional change. The authority ‘to amend the constitution’, granted by constitutional legislation, means that other constitutional provisions can substitute for individual or multiple ones. They may do so, however, only under the presupposition that the identity and continuity of the constitution as an entirety is preserved.30

Schmitt’s claim that constituted authorities can only exercise the powers that they are given is not unusual. His distinctive claim is that restrictions on amendment result not from a decision of the bearer of the constitution-making power but rather ‘from the properly understood concept of constitutional change’. Importantly, this is a claim of conceptual limits on the constitution-making power, not on the constituted powers.31 At the moment of constitution-making, the bearer of the constitutionmaking power cannot grant a power to a constituted entity to make fundamental changes to the Constitution. Later, once a constitution is established, the bearer of the constitution-making power is unable to make fundamental changes through the existing constitutional processes—it can only establish a new constitution. It is very difficult to articulate any theory of legitimate authority that can justify such a restriction. If Schmitt is right that constitutions derive their authority from the brute force

23 Sieyes, Political Writings: Including the Debate between Sieyes and Thomas Paine in 1791 (Michael Sonenscher tr, Hacket Publishing Company Inc, 2003) 136. 24  T Pereira, ‘Constituting the Amendment Power’, in this volume. 25  C Schmitt, Constitutional Theory (Jeffrey Seitzer tr, Duke University Press, 2008) 64. 26  ibid 77. 27 ibid. 28  ibid 64. 29  ibid 121. 30  ibid 150. Schmitt notes that there may be differences between power and authority in some contexts but contends in a note that the distinction need not be elaborated for the exposition of his constitutional theory; ibid 458. 31  For a presentation of this argument see Yap (n 11) 116 and Roznai (n 1) 690–93.

80  Oran Doyle of political will, a claim that he repeatedly asserts rather than justifies, the bearer of the constitution-making power simply could not be circumscribed in the way that Schmitt claims in respect of constitutional amendment. Such a powerful entity must have the power at the moment of enactment to create an unfettered amending power; it must later have the power to act in any way through the prescribed amendment power, should it so choose. Notwithstanding the widespread acceptance of Schmitt’s arguments, the distinction between constitution-making powers and constituted powers does not imply any constraints on the power of constitutional amendment.32 D.  A Problematic Rubric One of the great ironies in the literature is how the work of Schmitt, who had at best a questionable commitment to the values of liberal constitutionalism, appears to fit so well with the liberal constitutionalist concerns of commentators such as Murphy, Jacobsohn and Bernal. This happens through the elision of conceptual and moral understandings of unconstitutional amendments. The elision is facilitated by the way in which conceptual and moral approaches converge on the same question: does the amendment introduce too much change? Amendments are claimed to be conceptually unconstitutional if they introduce so much change that they are really a revision or a replacement or beyond the scope of constituted power. Amendments are claimed to be morally unconstitutional if they depart too much from the value of constitutionalism generally presupposed to be adequately instantiated by the existing constitution. Such elision makes it more difficult to ascertain the discrete merits of the conceptual and moral approaches. It also facilitates a further elision with the positive sense of an unconstitutional amendment, whereby an amendment is unconstitutional if it makes too much change by reference to what the Constitution actually permits. This elision obscures the crucial move of the doctrine of unconstitutional constitutional amendments, namely the empowerment of constitutional actors to treat amendments that are claimed to be morally or conceptually unconstitutional as if they were positively unconstitutional. For all of these reasons, the rubric of unconstitutional constitutional amendments impedes rather than facilitates analysis. It elides critical differences between moral and conceptual claims while obscuring the empowerment of judges to transform moral or conceptual claims into claims about positive law. Conceptual claims about constraints on the power of constitutional amendment are unfounded and cannot be rescued by shifting the focus to the moral value of constitutionalism. Moral claims about constraints on the power of constitutional amendment may be justified but they are not limited to the value of constitutionalism and must be assessed against

32  For further criticisms of Schmitt see M Polzin, ‘Constitutional Identity, Unconstitutional Amendments and the Idea of Constituent Power: The Development of the Doctrine of Constitutional Identity in German Constitutional Law’ (2016) 14 Int’l J Const L 411, 419–21.

Constraints on Constitutional Amendment Powers 81 potentially competing values. This cannot be achieved if we limit ourselves to the value-monist question of whether the amendment is morally constitutional. For those who design or develop constitutions, the important questions are whether constraining a constitutional amendment power would serve important values, and whether that constraint undermines other important values. More specifically, do the constraints serve important values that justify the subjection of the present generation to the decisions of a past generation, alongside the empowerment of a judicial elite? These concerns percolate through the literature, but in a diluted and dispersed way. In order to address these concerns properly, we need a framework of analysis that allows us to focus directly on the values that are served and damaged by the constraint of constitutional amendment powers. Before that, however, it is necessary to explore in more detail the idea of constraint on constitutional amendment powers. III.  CONSTRAINT AND POWERS OF CONSTITUTIONAL CHANGE

Constraint involves a limitation or fetter on action. A person constrained from f-ing is in some sense unable to f. We might say that a person is constrained from f-ing if she is coerced not to f or is physically unable to f. Constraint in this empirical sense correlates with Albert’s notion of constructive unamendability: the constitutional politics that make certain provisions of the US and Canadian constitutions effectively unamendable through the formally prescribed amendment process.33 These empirically observable constraints are not the concern of this chapter.34 Our concern is normative constraints—obligations. The central case of obligation is moral obligation, the obligation that truly applies to people. I follow Joseph Raz in treating obligations as exclusionary or protected reasons for action. If X is under an obligation to f, then X has a reason to f that simultaneously excludes the force of other reasons.35 A legal obligation is an obligation claimed by law—an obligation that the law claims to have imposed on X.36 The law’s claim may be mistaken: it may lack authority to make any claims at all, or it may lack authority to make the claim in question. Nevertheless, at least where there is a functioning legal system, law’s claims demand special attention. They tend to be respected by powerful figures and therefore have significant real-world effects. A constitutional obligation is an obligation claimed by the Constitution, a species of legal obligation. When a constitution states (or is authoritatively interpreted to state) that X cannot f, then there is a legal obligation on X not to f. A constitution

33  R Albert, ‘Constructive Unamendability in Canada and the United State’ (2014) 67 Sup Ct L Rev (2d) 181. 34  Of borderline relevance are constraints that arise from another legal order (such as the European Union or international law) or from constitutional conventions (in the Canadian sense). I shall not complicate the analysis by adding these dimensions to the discussion. 35  See J Raz, Practical Reason and Norms, 2nd edn (Oxford University Press, 1990) and J Raz, Ethics in the Public Domain: Essays in the Morality of Law and Politics (Oxford University Press, 1994) 210–37. 36  J Gardner, Law as a Leap of Faith (Oxford University Press, 2012) 125–28.

82  Oran Doyle contains its own rules of change that both empower a constitutional legislator to amend the Constitution but simultaneously constrain it only to amend the Constitution according to a particular procedure. At least some procedural constraints are necessary corollaries of any power to change the constitutional rules since procedurally only the constitutional legislator may make the amendment. Typically, constitutions distribute the power of amendment among a number of collective actors who must follow prescribed processes. The greater the procedural constraints, the more difficult amendment becomes. A rule of change could also constrain the constitutional legislator from including certain content in constitutional amendments. These rules of change that empower and constrain the constitutional legislator are reflected in a rule of recognition that imposes a duty on political actors—and judges unless there is a non-justiciability rule—to recognise as constitutional law only those changes that respect the constraints. All of these constraints are contained (explicitly or implicitly) in the posited rules of the legal system. Constraints can be asserted from outside the legal system also. For instance, irrespective of what a constitution says, it could be argued that it is impermissible to amend that constitution in a way that contravenes the will of God. In precisely the same way, it could be asserted that it is impermissible to amend a constitution in a way that offends the moral value of constitutionalism. These constraints, if they are true in the sense of imposing real moral obligations, are relevant to those who write constitutions and to those who evaluate constitutions. However, they are not immediately relevant to the officials of the legal system who, provided the legal system is generally just, are under an obligation to apply every rule that is not radically unjust.37 This qualification is important because it suggests that there may be circumstances in which it is appropriate for judges to impose a new constraint on the amendment power. In doing so, however, judges exercise a very different power from when they apply the law. Indeed, in one sense, they act against the law. Their actions, which redistribute power among political actors, must be assessed with reference to the same sort of moral arguments that we would consider when deciding what textual constraints to include in a constitution in the first place. However, the threshold for justification is greater since judges, when they create new constraints, act against the law and can claim no direct authority from majoritarian democracy. This account provides us with a clearer picture of constraints on constitutional amendment powers: whether they relate to the process for changing the Constitution or the permissible content of a change, constraints make amendment more difficult. The fundamental question is whether and to what extent it is justified to make constitutional change more difficult. Rather than address that question at best indirectly by measuring the quantum of constitutional change in various ways, we should address it directly. This requires us to consider the full range of potential constraints on the power of constitutional amendment, rather than jump straight to the question

37  This is a slight development of Finnis’s account of the moral obligation to obey the law. See J Finnis, Natural Law and Natural Rights (Clarendon Press, 1980), chs XI and XII.

Constraints on Constitutional Amendment Powers 83 of whether it is permissible for judges to declare positively valid constitutional amendments to be unconstitutional. IV.  THE TYPES OF CONSTRAINT ON CONSTITUTIONAL AMENDMENT POWERS

A.  Process or Content Constraints on the amendment power affect either the process for exercising the amendment power or the content of amendments.38 As noted above, amendment powers necessarily constrain the process of amendment since only the assignee of the amendment power may exercise that power. In this sense, Article 368 of the Indian constitution that empowers the Indian Parliament to amend the Constitution is a process-constraint on the amendment power since that procedure must be followed for an amendment to be effective. Where several actors are involved, the effect of the process-constraint is greater. For instance, Article 41 of the Canadian Constitution establishes an amendment procedure for certain matters that requires approval by both Houses of the Federal Parliament and by the Legislative Assembly of each Province. Some process-constraints are more obviously constraints in that they can be characterised as extraneous to the identification of the constitutional legislator. Temporal limitations are one example of this. Albert identifies deliberation provisions that set upper or lower limits on the amount of time for which an amendment can be considered and safe harbour provisions that protect provisions from change for a certain period of time.39 We might also include dual-approval provisions that require the relevant body to approve the amendment twice, often with an intervening general election.40 Temporal limitations can turn on uncertain events, such as a preclusion of amendment during a time of war,41 or a state of emergency.42 An example of a content-constraint is the prohibition in the Honduras constitution on the removal of term limits for the President.43 Sometimes, process and content constraints can intersect. For instance, Article V of the US Constitution precluded the adoption of any amendment to the constitutional provision preventing Congress from prohibiting the importation of persons until 1808. This constrained the amendment power from being exercised in respect of certain content for a certain period of time and was thus both a content- and process-constraint. Indeed, as illustrated by the Canadian example considered above, any constitution that prescribes

38  Process is the best word here but there is still potential confusion. What is meant is the process through which amendments are adopted, not the content of an amendment prescribing another constitutional process. 39  R Albert, ‘The Structure of Constitutional Amendment Rules’ (2014) 49 Wake Forest L Rev 913, 952–54. 40  Art 29 of the Icelandic Constitution. 41  Art 191 of the Belgian Constitution. 42  Art 152.3 of the Romanian Constitution. 43  Arts 4, 237 and 274.

84  Oran Doyle different amendment processes for different types of amendment combines processand content-constraints in this way.44 For analytical purposes, however, we can still distinguish between the different types of constraint combined in the one provision. B.  Rule or Standard Constraints can be formulated either through rules or through standards.45 The distinction points us to a spectrum between the most formal rule and the most informal standard. Between these two extremes, rules have increasing levels of vagueness and standards increasing levels of precision. There is no precise boundary between the two but the spectrum is clear. I use ‘rule’ and ‘standard’ as shorthand to connote their orientation on this spectrum. The value of legal certainty suggests that processconstraints that identify the constitutional legislator should take the form of rules: it is important that there be little doubt as to whether the relevant entity has purported to exercise the amendment power. Once this is clear, there is more scope to formulate process-constraints in the form of standards. For instance, the Irish Supreme Court has created a rule that referendum processes must be fair, prohibiting the government from spending public money to support just one side of the campaign.46 Content-constraints can likewise be formulated as rules or as standards. The US slavery provision considered above was a rule, not a standard. In contrast, Article 79 of the German Constitution (with reference to Article 1) imposes the standardconstraint of prohibiting amendments that violate human dignity. Article 121 of the Norwegian Constitution employs an even vaguer standard-constraint, allowing only amendments that do not ‘contradict the principles embodied’ in the Constitution and that do not ‘alter the spirit of the Constitution’. C.  Legislator or Court This distinction has been anticipated by some of the examples above: constraints may be imposed by the constitutional legislator, whether in the original constitution or in an amendment, or by the body charged with applying the Constitution, usually a court. The distinction is clouded because a court has both a law-application and law-making role and judges may be mistaken about or indifferent to which role they are performing.47 Where a standard-constraint (such as respect for human dignity) is imposed, the import of the constraint is determined partly by the legislator and partly by the court. One can expect courts to claim that they are merely interpreting

44  See T Kahana, ‘Canada’ in D Oliver and C Fusaro (eds), How Constitutions Change: A Comparative Study (Hart Publishing Ltd, 2011) 9. 45  For a helpful account of the distinction between rules and standards and the reasons for using them to formulate normative directives, see F Schauer, Thinking Like a Lawyer: A New Introduction to Legal Reasoning (Harvard University Press, 2009), ch 10. 46  McKenna v An Taoiseach (No 2) [1995] 2 IR 10. 47  Gardner (n 35) 79–80.

Constraints on Constitutional Amendment Powers 85 a standard or perhaps identifying an implicit constraint rather than creating a new constraint. Whether this claim is justified depends on the constitutional text in question and the norms of interpretation that govern judicial activity in that jurisdiction. Nevertheless, there is always the conceptual possibility that a court has created a new constraint rather than interpreted an existing constraint. This distinction between constraints that are created by courts and constraints that are created by legislators is more analytically useful than a distinction between explicit and implicit constraints. The distinction between explicit and implicit constraints cannot bear on the legitimacy of the court’s decision: if the constraint is implicitly constitutionally endorsed, the court is under a presumptive obligation to apply it. As noted above, however, greater justification is required for any decision by a court to create a new constraint since this involves acting against the law. Our set of distinctions must capture this important difference. These issues are illustrated by the Indian basic structure doctrine.48 In LC Golak Nath v State of Punjab, the Indian Supreme Court had held that constitutional amendments were ‘laws’ that could be reviewed for their compatibility with fundamental rights.49 Parliament sought to put the scope of the amending power beyond doubt by amending Article 368 as follows: Notwithstanding anything in this Constitution, Parliament may in exercise of its constituent power amend by way of addition, variation or repeal any provision of this Constitution in accordance with the procedure laid down in this article.

Nevertheless, in Kesavananda Bharati v State of Kerala, the Indian Supreme Court held that Parliament’s amendment power did not extend to amendments that destroyed or emasculated the basic or fundamental features of the Constitution.50 The textual justification offered for the Court’s decision was the limitation inherent in the concept of amendment. However, this is difficult to support in light of the historical background to the new provision, as well as its sheer breadth—precisely the same wording has been held by the Irish courts to lead to a substantively unlimited amendment power. Following Kesavananda, Parliament again amended Article 368 explicitly removing any power on the part of the courts to review constitutional amendments and declaring that there was no limitation on the constituent power of Parliament to amend the Constitution. In Minerva Mills Ltd v Union of India, the Supreme Court held that these amendments were themselves unconstitutional on the basis that they upset the balance of power between the constitutional organs and allowed the Constitution to become uncontrolled.51 This more structuralist justification supports the view that the basic structure doctrine must be viewed as a courtconstraint on the amendment power.

48  For general discussion, see MP Jain, Indian Constitutional Law (LexisNexis Butterworths Wadhwa Nagpur, 2010), 2341–82 and P Bhanu Mehta, ‘The Inner Conflict of Constitutionalism: Judicial Review and the “Basic Structure”’ in Z Hasan, E Sridharan and R Sudarshan (eds), India’s Living Constitution: Ideas, Practices, Controversies (Permanent Black, 2002) 179. 49  AIR 1967 SC 1643. 50  AIR 1973 SC 1461. 51  (1980) 3 SCC 625.

86  Oran Doyle D. Values Served by Constraint: Foundational, Majoritarian, or Counter-Majoritarian Constitutions, at least in liberal democracies, typically perform three functions. They establish or specify the parameters of a polity, whether ethnically, nationally, federally or territorially. They establish some form of majoritarian government. And they protect some fundamental rights. In this way, constitutions limit and channel the power of political actors to change certain features of their polity; they set the domain and rules of the game that political actors must follow. The constitutional settlement of these foundational, majoritarian and counter-majoritarian issues is open to revision unless there are constraints on the amending power. Article 152.1 of the Romanian constitution provides an example of a foundational content-constraint: territorial integrity may not be the object of a constitutional amendment. The Constitution of Honduras provides a good example of a majoritarian content-constraint. Articles 4, 237 and 274 secure that the president can serve only one term and this rule cannot be changed by constitutional amendment.52 This protects against the concentration of power in one person’s hands, irrespective of current majoritarian support, in order to ensure that majoritarian processes can continue. The Indian basic structure doctrine has been applied to protect majoritarian values. For instance, in Indira Nehru Gandhi v Raj Narain, the Supreme Court held that the Thirty-Ninth Amendment offended the basic structure in the way that it both voided a High Court decision that set aside the election of Mrs Gandhi and immunised the election of the Prime Minister from judicial scrutiny.53 The US Constitution provides a good example of a counter-majoritarian content-constraint: the rule that prohibited Congress from banning slavery until 1808 prevented Congress from imposing a majority view against slavery. The examples in the previous paragraph were all content-constraints, but processconstraints can also reflect these three values. For instance, although the Republic of China (Taiwan) claims authority over the territory of the People’s Republic of China and Mongolia, the first additional article (2005) of its Constitution provides that electors of ‘the free area of the Republic of China’ may vote on a legislative proposal to amend the Constitution or alter the national territory. This provision, itself an amendment, constrains the process through which certain foundational decisions in Taiwan’s constitution can be changed. The Irish rule that the Government cannot spend public money supporting just one side of a constitutional referendum is a majoritarian process-constraint, designed to ensure that the real ‘will of the people’ is ascertained.54 I take Halmai to argue for a majoritarian process-constraint

52  For a detailed consideration of these provisions of the Honduras Constitution, see R Dixon and V C Jackson, ‘Constitutions Inside Out: Outsider Interventions in Domestic Constitutional Contests’ (2013) 48 Wake Forest L Rev 149. 53 (1975) Supp SCC 1. See discussion in M Khosla, ‘Constitutional Amendment’ in S Choudhry, M Khosla and P Bhanu Mehta, The Oxford Handbook of the Indian Constitution (Oxford University Press, 2016) 241–42. 54  McKenna v An Taoiseach (No2) [1995] 2 IR 10. The temporal limitations considered above are further examples of majoritarian process constraints.

Constraints on Constitutional Amendment Powers 87 when he criticises the Hungarian Constitutional Court for not considering whether the extreme haste and lack of deliberation with which a constitutional amendment was passed undermined the validity of that amendment.55 Article 182.3 of the Cypriot Constitution requires that amendments to that Constitution receive a majority vote comprising at least two-thirds of the total number of the Representatives belonging to the Greek Community and at least two-thirds of the total number of the Representatives belonging to the Turkish Community. In allowing a minority of each community to block constitutional change, this imposes a counter-majoritarian process-constraint. V.  DISTRIBUTION OF POWER AND THE JUSTIFICATION OF CONSTRAINT

A power of constitutional amendment typically empowers a contemporary majority in some form. Constraints on the power of constitutional amendment empower others at the expense of that contemporary majority. Across time, there is a distribution of power between past generations (T1), the present generation (T2) and future generations (T3). Within the present generation, there is a distribution of power between those whose power of amendment is constrained and those who either create new constraints or interpret and enforce the existing constraints. The set of distinctions analysed in section IV allows us to pinpoint the different ways in which constraints on amendment powers distribute power along these two dimensions. All legislator-constraints on amendment power necessarily empower past generations at the expense of present generations. However, there is a significant variation depending on whether the constraint is formulated as a rule or a standard. Ruleconstraints, such as that prohibiting the US Congress from banning the importation of persons for 20 years, leave little discretion to constitutional interpreters. In contrast, standard constraints, such as the German provision about human dignity, grant considerable power to constitutional interpreters. Standard-constraints therefore empower the past generation at the expense of the present generation far less than rule-constraints since the meaning of the standard (eg, human dignity) remains to be assessed by the current generation. Justiciability, while not affecting the normative claim of law, has significant implications for the distribution of power. If standard-constraints are justiciable, the current constitutional legislator loses power both to the generation that enacted the constraint and (perhaps more significantly) to the court that interprets the constraint. If the standard-constraint is non-justiciable, however, the current constitutional legislator may adopt its own interpretation of the constraint, perhaps rendering the constraint meaningless.56 In contrast, non-justiciable rule-constraints are, because of their clarity, more difficult

55  G Halmai, ‘Constitutional Courts as Guardians of the Constitution’ (2012) 19 Constellations 182, 195. 56 This might be described as constructive amendability, the converse of Albert’s constructive unamendability. Provisions that appear unamendable in law become amendable in fact. Albert, (n 32). Of course, courts retain some power to the extent that they can alter justiciability rules.

88  Oran Doyle to breach with impunity and therefore may succeed in empowering the past at the expense of the present. The previous paragraph considered content-constraints. Process-constraints empower the past at the expense of the present because they make constitutional change difficult. In principle, process-constraints disempower the current generation less than content-constraints: they do not limit what can be done but rather how it can be done. However, it is easy to imagine how a particular process-constraint could—in practice—constrain more than a particular content-constraint. The Canadian unanimity requirement, for example, is a greater constraint on the amendment power than a content-constraint that imposed a trivial restriction. Court-constraints tend to represent an empowerment of courts at the expense of legislators. Although it is conceptually possible that a court could create a constraint on the amendment power and then deem it non-justiciable, this seems to be an unlikely (if not quite contradictory) exercise of judicial power.57 The constitutional legislator could agree with the content of a court constraint or might rely on the constraint as a way to avoid politically unpopular decisions. Nevertheless, the constraint does affect the distribution of political power. The question of whether courtconstraints empower the past at the expense of the present is more complicated. At the moment of creation (T2), court-constraints typically constrain the current generation with reference to decisions of the past generation as interpreted by the courts. Although the Indian basic structure doctrine, on my analysis, was a judicial creation, it protected values endorsed by India’s founding generation (T1) in the constitutional text as interpreted by the courts at (T2). At first blush, process-constraints that protect majoritarian values remove less power from the current generation than do constraints that protect foundational or counter-majoritarian values. Majoritarian process-constraints, such as a rule requiring fairness in amendment processes, aim to ensure that the amendment made at time (T2) really represents the will of the current generation (T2). Double-approval requirements and deliberation requirements create a period of time around (T2) that moderately empowers the future at the expense of the present. Although these observations are valid, they are subject to the qualification that to the extent that the constraint just makes amendment more difficult, it actually empowers the past generation that made the decision. Thus double-approval requirements and deliberation requirements could, depending on how difficult they make constitutional change in practice, be characterised as empowering either the past or the future generation at the expense of the present generation.58 As the empowerment of a future generation is more distinctive from other process-constraints, I shall focus on that possibility in the rest of this chapter. In contrast, majoritarian content-constraints unequivocally empower the future generation since what they protect are constitutional processes that will continue to allow the majority will to be ascertained into the future.

57  The constitutional conventions recognised by the Supreme Court of Canada in re Patriation Reference [1981] SCR 753 are arguably examples of this. 58  These arguments do not cancel each other out: majoritarian-process constraints either empower the past generation or the future generation.

Constraints on Constitutional Amendment Powers 89 VI.  JUSTIFICATION OF CONSTRAINTS ON CONSTITUTIONAL AMENDMENT POWERS

A.  Parameters of Justification Constitutions serve competing values. The value of majoritarianism derives from the proposition that laws should serve the needs of those whom they govern and that those subjects are collectively the best judges of their own needs. Foundational values derive from the proposition that a political community should be able to determine its own parameters and establish the relevant people for the purposes of majoritarian rule. Counter-majoritarian values derive from the proposition that there are some actions that are so wrong that a majority should not be permitted to require or allow them. Constraints on amendment powers serve these values by distributing power away from contemporary majorities and towards past generations, sometimes future generations and judicial elites. Even constraints that serve majoritarianism in the medium to long run, through the protection of majoritarian procedures from majoritarian abrogation, do a disservice to majoritarianism in the here and now. The central question, therefore, is whether the disservice to majoritarianism can be justified by majoritarian, foundational or counter-majoritarian values.59 Posing the question in this way is not to give priority to majoritarianism over other values. Rather, it simply recognises that since the structure of constraints necessarily undermines the value of majoritarianism, the best way to approach the question of justification is to identify precisely the loss to majoritarianism and then explore whether it is justified. It is far beyond the scope of this chapter to engage in a full analysis of how these values should be reconciled or how all constraints on constitutional amendment powers should be evaluated. What follows, therefore, is an outline of the sort of arguments that are relevant. B.  Illustrative Schema of Constraints The following diagram, building on the analysis in section V, presents in a schematic way the extent to which constraints on constitutional amendment powers disenfranchise contemporary majorities. The X-axis represents time and the Y-axis represents the distribution of political power between majorities and courts at the time of constitutional amendment. The assumption here, as throughout this chapter, is that we

59 This question resonates with how Dixon and Landau present the doctrine of unconstitutional constitutional amendments. They argue the doctrine is potentially justified as a guard against abusive constitutionalism notwithstanding the extent to which it impedes change through democratic processes. R Dixon and D Landau, ‘Transnational Constitutionalism and a Limited Doctrine of Unconstitutional Constitutional Amendment’ (2015) 13 Int’l J Const L 606, 609–14. However, my formulation of the issue is more open-ended as it does not limit itself to judicial doctrines, nor to substantive constraints; moreover, it remains open to the possibility of foundational and counter-majoritarian justifications, as well as majoritarian justifications. By keeping all values in view, we sharpen our sense of each values argument.

90  Oran Doyle are concerned with constitutions that stipulate a broadly majoritarian process for their own amendment. The intersection of the X-axis and the Y-axis represents the point at which the contemporary constitutional legislator controls the amendment power. The further one moves from that point, the greater the justification that is required for a constraint on the amendment power. H

H1

F E

G

T1

F1

C

A

B

T2

D

T3

—— A: legislator-process-rule-majoritarian, the Indian textual rules on amendment. —— B: legislator-process-rule-majoritarian, the Icelandic double-approval requirement. —— C: legislator-process-rule-majoritarian, the Irish referendum requirement. —— D: legislator-content-rule-majoritarian, the Honduras term limits rule. Constraint-A is located closest to the intersection of the X-axis and Y-axis but not precisely at the intersection. The requirement of a parliamentary process to amend a constitution still constrains the power to a limited extent and, as such, distributes some power to the founding generation. Constraints-B and -C are similar. Both make it more difficult for the contemporary majority to amend the Constitution but they distribute power to future and past generations respectively. As my contribution with David Kenny to this volume illustrates, legally suboptimal rules are less likely to be changed if a referendum is required.60 Constraint-D—in a more pronounced way than the process constraint-B—transfers power to a future generation by precluding amendments that undermine certain majoritarian processes. —— E: court-process-standard-majoritarian, the Irish rule on fairness in referendum campaigns. I have located this constraint directly above T2: whatever amendment process is used, constraint-E tries to ensure that the outcome reflects the will of the contemporary generation. However, it also distributes power to the courts since they decide what is fair. —— F: legislator-content-standard-counter-majoritarian, the German human dignity provision (if justiciable). 60 

O Doyle and D Kenny, ‘Constitutional Change and Interest Group Politics’, in this volume.

Constraints on Constitutional Amendment Powers 91 —— F1: the German human dignity provision (if non-justiciable). —— G: legislator-content-rule-counter-majoritarian, the US slavery provision (before 1808). Constraints-F and -F1 illustrate the effect of a non-justiciability doctrine on how a constraint distributes power. A justiciable standard-content-constraint, particularly if it involves a concept as nebulous as human dignity, is much more an empowerment of a contemporary judiciary than a past generation. Constraint-G, in contrast, illustrates that more power is retained by the past generation where a counter-majoritarian rule (rather than a standard) is laid down. —— H: court-content-standard-majoritarian, the Indian basic structure doctrine.61 —— H1: the Indian basic structure doctrine 40 years after its judicial creation. Constraints-H and -H1 illustrate how a basic structure doctrine distributes power to courts but also the temporal distribution when viewed from the perspective of T3.62 At the moment of its pronouncement, the basic structure doctrine in India constrained the contemporary democratic majority in favour of (a) the Supreme Court, through its decision that there should be a basic structure doctrine, (b) the past generation that made the decisions that loosely determined the Indian constitution’s basic structure, and (c) the Supreme Court again through its power to interpret that basic structure. Constraint H1 shows what the basic structure looks like at time T3—still located in the past but, on account of it being a standard-constraint, somewhat amenable to contemporary judicial interpretation. C.  Contextual Factors The purpose of the schema is merely to represent the differing ways in which constraints on constitutional amendment powers distribute power away from contemporary majorities. Before analysing the majoritarian, foundational and counter-majoritarian justifications, it is necessary to consider some contextual factors that bear on those justifications. Although our concern is with normative constraints rather than physical constraints (see section III above), the justification for including such constraints in the Constitution is sensitive to the effect that constitutionalisation of the constraint will actually have in the particular political community. In this regard, attention should be paid to the following points. First, whether a constraint is justified depends on the country concerned: some countries may be at greater risk of democratic decay and therefore more in need of majoritarian-constraints, for example. Secondly, we cannot assume that the constitutional imposition of a constraint will necessarily lead to respect for that constraint.

61  The basic structure doctrine can be seen as majoritarian or counter-majoritarian in outlook depending on the values that it protects. I present it here in its majoritarian orientation. 62  If I were presenting a counter-majoritarian orientation of the basic structure rule, it would be closer to T1 than T3.

92  Oran Doyle I cited the Honduran protection of term limits on several occasions above precisely because it illustrates this point. The Supreme Court of Honduras has held that the constitutional prohibition on amendments to presidential term limits is itself unconstitutional, applying a basic structure doctrine to the original constitutional text as opposed to an amendment.63 Thirdly and also illustrated by the Honduran case, we cannot simply assume that judges will respect and correctly interpret constitutional constraints on amendment powers. There may be reasons why judges, as a relatively independent branch of government, are well placed to check contemporary majoritarian incursion on the three values under consideration. But it is also possible that judges could be under the sway of majoritarian forces or could act to protect the sectional interests of the relatively privileged social class from which judges are often drawn. These factors, and therefore the evaluation of constraints on constitutional amendment powers, will again vary from country to country. Fourthly, we cannot assume that contemporary majorities have no respect for counter-majoritarian or long-run majoritarian values. Jacobsohn criticises the Irish Supreme Court decision in State (Ryan) v Lennon for not constraining the constitutional amendment power of the Irish Free State or—at least—for not articulating for the benefit of the constitutional legislator that the amendment in question was unconstitutional.64 However, the most salient feature of this case is that it was ­followed within two years by a new constitution that, however imperfectly, has provided a reasonable basis for majority rule and the protection of minority rights for the following 75 years: a democratic majority checked itself more effectively than a court ever could have. Fifthly and finally, legal constraint of amendment powers does nothing to preclude constitutional replacement: there are limits to the extent to which political power can (and should) be controlled by constitutional constraints.65 D.  Majoritarian Constraints Majoritarian constraints seem to be the most prevalent in practice. Their justification is relatively easy to assess since there is no conflict between values but rather a conflict between respect for majoritarianism in the short term and the desire to preserve majoritarianism over the medium term. Process-majoritarian constraints, which seek to ensure that the true majoritarian will is being expressed, require less justification

63  For discussion, see D Landau and B Sheppard, ‘The Honduran Constitutional Chamber’s Decision Erasing Presidential Term Limits: Abusive Constitutionalism by Judiciary?’ (International Journal Constitutional Law Blog, 6 May 2015) accessed 1 August 2016 and L Marsteintredet, ‘The Honduran Supreme Court Renders Inapplicable Unamendable Constitutional Provisions’ (International Journal Constitutional Law Blog, 1 May 2015) accessed 1 August 2016. 64  [1935] IR 170. Jacobsohn (n 6) 470. For discussion, see A Kavanagh, ‘Unconstitutional Constitutional Amendments from Irish Free State to Irish Republic’ in E Carolan (ed), The Irish Constitution: Perspectives and Prospects (Bloomsbury, 2012) 331. 65 Oliver and Fusaro argue that if incremental change is difficult, rupture or forced interpretation becomes more likely. Oliver and Fusaro (eds) (n 43) 424.

Constraints on Constitutional Amendment Powers 93 than content-majoritarian constraints, which prevent contemporary majorities from making certain changes. This observation coheres with the insights of Yap and ­Roznai, the latter writing in this volume, that where amendment processes are rigid and/or approximate to involvement of the constitution-making power, courts should be slower to declare amendments unconstitutional. From the perspective of this chapter, such amendment processes should themselves be understood as majoritarian process-constraints, and therefore relatively easy to justify. In other words, a process-constraint serves the value of majoritarianism when it allows the ascertainment of a settled popular will. For this reason—and not because they approximate to constitution-making power—any further constraints require greater justification. As noted above, however, an extreme process-constraint might make a constitutional provision constructively unamendable and therefore require the same justification as a content-majoritarian constraint to similar effect. In this regard, the conundrum is how to guard against the risks to majoritarian democracy in the future while respecting majoritarian democracy in the here and now. This problem is particularly acute because democracy is itself a contested concept. It would be undemocratic to dictate what form of majoritarian decision-making process a particular polity must follow. Majorities must decide their own procedures for identifying their own majority will, subject only to the constraint that they cannot take decisions that lead to the disestablishment of any form of majoritarian governance.66 Nevertheless, there is a genuine dilemma here. Political elites have incentives to entrench their own power, incentives that may not cohere with the interests of the people who elect them. Content-constraints guard against this possibility but their justifiability will vary from country to country and depend as much on questions of political culture as institutional design. In this regard, Halmai has plausibly argued that constraints of this type are required in countries with a history of military dictatorship and totalitarian government.67 As noted above, court-constraints are more difficult to justify than legislator-­ constraints since they involve judges acting against the law that they are under a presumptive obligation to apply; otherwise, however, the structure of argument is the same. Process-constraints that require fairness—whether for referendums as in the Irish case or for parliamentary deliberations as suggested by Halmai in the ­Hungarian case—are justifiable on the basis that they make it more likely that an amendment actually reflects majority will. Content constraints are more difficult to justify but may—in principle—be justifiable. The Indian basic structure doctrine, although a significant expansion of judicial power at the cost of a contemporary democratic majority, appears to have been vindicated by subsequent events. The curtailment of Indira Gandhi’s state of emergency helped to preserve democratic culture within India.68 Nevertheless, there seems to be a troublesome tendency for

66  I am indebted to the work of David Prendergast for this understanding of democracy. PhD thesis on file with author. 67  Halmai (n 54) 192. 68  Jain (n 47) 2362.

94  Oran Doyle the scope of basic structure doctrines to expand.69 Any such risk must be factored into the justification of basic structure doctrines. E.  Foundational Constraints For Schmitt, the substance of the Constitution (as distinct from mere constitutional laws) is ‘the concrete political decisions providing the … people’s form of political existence’.70 Foundational constraints are therefore the paradigm case of restrictions on the constitution-making power. This view is reflected in provisions, such as Article 152.1 of the Romanian constitution, prohibiting amendments that affect territorial integrity. Roznai and Suteu argue that the people should be allowed to address issues such as territoriality through constitutional processes.71 This is correct but applies more broadly to all foundational questions.72 There is no sound reason to preclude the alteration of a polity through its existing constitutional processes. As noted above, in 2005 additional articles were added to the Taiwanese Constitution that, by altering the identity of the constitutional legislator, effectively shifted the people of the Republic of China to those living in the ‘Free Area’ only, implicitly excluding those in mainland China.73 In 1998, the Irish Constitution was amended to withdraw a territorial claim over Northern Ireland and adopt a more inclusive notion of the Irish people. For Schmitt, these amendments would have represented paradigm cases of a constituted power illegitimately purporting to exercise a constitution-making power. If the distinction between constituted and constituent power holds salience in this context, which I argue above it does not, it follows that the actions of the Taiwanese and Irish constitutional legislators were illegitimate. Conversely, if one considers those actions legitimate, one must abandon the claim that the concept of constitution-making power implies necessary limitations on the powers of constitutional amendment. F.  Counter-Majoritarian Constraints Counter-majoritarian constraints pose a straightforward conflict between incommensurable values. This is the fundamental debate of constitutional theory and cannot be resolved here. It suffices to note two important ways in which the issue of constraints on constitutional amendment powers differs from general debates in constitutional theory. First, a common way of resolving the competition between

69 

Landau (n 4) 237–8; Dixon and Landau (n 58) 620–23. Schmitt (n 25) 78. 71  Y Roznai and S Suteu, ‘The Eternal Territory? The Crimean Crisis and Ukraine’s Territorial Integrity as an Unamendable Constitutional Principle’ (2015) 16 German L J 542, 573. 72  This does not preclude such issues being addressed alternatively outside the constitutional order. 73  Yeh argues that these amendments have formed a new collective identity of Taiwanese people. See J Yeh, The Constitution of Taiwan: A Contextual Analysis (Hart Publishing, 2016) 48–49. 70 

Constraints on Constitutional Amendment Powers 95 majoritarian and counter-majoritarian values is to characterise a constitution as the embodiment of values that the majority has chosen to obey. The possibility of constitutional amendment is necessary for this characterisation to work. Accordingly, this resolution cannot justify the imposition of constraints on the amendment power. Secondly, as noted above, constraints on the constitutional amendment power may be avoided by constitutional replacement. Indeed, the more difficult it is to change a constitution, the more likely that rupture will occur. Counter-majoritarian constraints may have precisely the opposite effect to what was intended. As against this, it can be argued that the expressive function of unamendability might dissuade political actors from even considering certain measures.74 VII. CONCLUSION

Constraining the power of constitutional amendment raises fundamental normative questions. On the one hand, there is value in allowing contemporary democratic majorities determine their own laws. On the other hand, contemporary majorities may seek to undermine democratic processes in the medium run or act contrary to other fundamental values. The appropriate resolution of these competing values is heavily context-dependent. Moreover, given the incommensurability of some of the values in questions, any resolution may remain contestable. Nevertheless, it is important that the competition between values be directly addressed. Much of the literature impedes clarity of analysis because it focuses on a judicial doctrine of unconstitutional constitutional amendments. This deflects attention onto the largely irrelevant question of the quantum of change introduced by the constitutional amendment. Such an approach may be superficially attractive because it appears to align the judicial task (interpreting ‘constitutionalism’ or the limits of ‘constitutional amendment’) with the archetypal judicial role of interpreting the Constitution itself. However, this apparent alignment depends on incoherent arguments and conceals the judicial role in resolving a fundamental competition of moral values. In order to have any chance of reaching defensible answers, constitution-drafters, courts and scholars alike must focus directly on the values served and disserved by constraints on the power of constitutional amendment.

74 During the Parliamentary Council deliberations on the German Constitution, Thomas Dehler argued that constraints on constitutional amendment could prevent revolutionaries from relying on a constitution to legitimise their movement. See Polzin (n 31) 423–24.

96 

4 Comment on Doyle’s Constraints on Constitutional Amendment Powers MARK TUSHNET

O

RAN DOYLE’S DISCUSSION of the constraints on amendment powers draws many valuable distinctions. Indeed, perhaps it may draw too many distinctions. His ‘four cross-cutting distinctions’ generate a set of 24 possible categories into which amendment powers might fall. As a recent discussion of ‘nuance’ in sociological theorising suggests, the proliferation of distinctions may impede the development of even moderately general theory.1 Doyle’s text actually suggests as much when he offers a chart locating ten, not 24, possibilities along two dimensions, and indeed it might be interesting to work out which possible categories do not appear, and why. Doing so might provide some clues to possible generalisations. But, of course, we can also treat the proliferation of analytical categories as exemplifying one rhetorical strategy in long-standing methodological discussions in comparative law, in support of the ‘contextualising’ approach against the ‘functional’ one. My comments on the substance of Doyle’s argument fall into two groups. The first deals with his discussion of what he calls the interpretive and moral approaches to identifying constraints on amendment powers. The second focuses on his discussion of the costs to majoritarianism—or self-governance—when there are legal constraints on those powers.2 Doyle begins by emphasising how important the quantity of change is in discussions of permissible constitutional amendments. Gary Jacobsohn and Walter Murphy both work within that paradigm. For Jacobsohn, an amendment is legitimate only if the preexisting constitution ‘survives without loss of identity’; for Murphy, ‘amendments that would destroy or cripple the values of constitutional democracy are invalid’. One implication is that when such amendments occur—as they might—they should be regarded as a form of constitutional replacement or substitution. Seeing them in that way might be consequential

1 K Healy, ‘Fuck Nuance’, available at accessed 28 September 2015. 2  I note here my agreement with Doyle that these costs occur as a result of the legal nature of the constraints, and not entirely from their judicial enforceability, though of course enforceability increases the likelihood that those costs will actually be incurred.

98  Mark Tushnet in legal systems that offer different procedural paths to constitutional amendment and constitutional replacement. A common practice, for example, allows parliament to amend a constitution but requires a constituent assembly to replace one. So, now consider an ‘amendment’ adopted by parliament in the manner that amendments are to be adopted, but regarded by a court as altering the preexisting constitutional identity, or as crippling democratic values. It would not be surprising to find the court invalidating the ‘amendment’ on purportedly procedural grounds, saying that the people could indeed change the Constitution to accomplish the result sought by parliament, but only by convening a constituent assembly.3 As Doyle suggests, then, the quantitative approach to constitutional amendments may in some circumstances lead to judicial empowerment, which, he correctly argues, needs to be taken into account when we think about the legitimacy and lawfulness of constitutional amendments. There is, I think, an additional problem with Murphy’s approach. That the change in constitutional identity or the destruction of democratic values occurs by means of a purported amendment is not, on his account, a matter of independent interest. Perhaps Jacobsohn would say, ‘You want to change your national identity—fine with me, as long as you are clear about what you are doing (which is why I say you can do it but only after holding a constituent assembly)’. But I doubt that Murphy would say something similar about a new constitution that destroyed democratic values. For him, that document, no matter how strongly endorsed by a constituent assembly and a popular referendum, would not count as a ‘real’ constitution. ­Murphy’s account, then, is not really about constraints on constitutional amendments; it is about the concept of constitutionalism. There are hints in Doyle’s exposition that he believes the same to be true of all (or most) quantitative accounts. I share that intuition (if Doyle holds it), and think it would be productive for future research to develop it in more detail. Murphy’s approach and others in the same class raise another question about constitutional amendment. Suppose a nation has a constitution that satisfies minimal, or more than minimal, requirements of constitutionalism. There is no reason a priori to treat that constitution as perfect, and indeed the very existence of amendment procedures testifies to the framers’ judgement that their resolution of various questions of constitutional substance might be mistaken. Now suppose that this constitution contains provisions purporting to insulate some of its provisions against amendment. The inclusion of such provisions instantiates the framers’ judgement that, whatever might be true about their resolution of other constitutional questions, they believe that they are entirely, indeed unalterably, correct in their evaluation of the permanent importance of these provisions. Their judgement on that question,

3  There is a concealed problem of regress here: The people, acting through parliament, may regard the requirement that constitutional replacements occur only after a constituent assembly’s work as itself an illegitimate restriction on their power. Consistent with a more general argument I make below, on that understanding the parliamentary amendment could be seen as a pro tanto repudiation of the provisions purporting to require that replacements take place after the work of a constituent assembly. This view is, in my opinion, particularly forceful when the amendment deals with a relatively discrete, rule-like problem (such as, in some circumstances, increasing the number of terms a president may serve).

Comment on Doyle’s Constraints on Constitutional Amendment Powers 99 though, might be thought mistaken by a later generation. What can be done about such a perceived error? Here some of Doyle’s distinctions are helpful. First, we might want to distinguish between: (a) perceived errors with respect to clear (or rule-like) constitutional ­provisions, such as a numerical limit on the number of terms an elected president can serve, and (b) perceived errors with respect to vaguer (more standard-like) provisions, such as federalism and secularism. Secondly, particularly in connection with the vaguer provisions, we might want to distinguish between nominally unamendable provisions whose interpretation is remitted to the legislature alone, and provisions the courts will interpret. Finally, we must acknowledge that the purportedly unamendable provisions can in fact be eliminated from the Constitution, either through a legally authorised and channeled process of constitutional substitution or replacement, or through a revolutionary process taking place outside the framework of existing legality. Of course, these methods have associated costs—for substitution or replacement the costs of using the legally authorised mechanisms, which I assume are larger than the costs of using the ‘ordinary’ amendment process, and for revolutionary transformation the costs to legality itself.4 One pervading theme of my comments, and I think a theme implicit in Doyle’s presentation, is this: Under what circumstances should good constitutional design require that a nation’s people incur these costs so that they can govern themselves in a way different from the way previously prescribed by the Constitution’s framers? Consider first an unamendable term-limits provision. Inserted into the Constitution for what at the time of its adoption were seen to be good reasons, and insulated from amendment for what, again, were seen as good reasons, the provision comes to be seen as an obstacle to good governance—for example, because it prevents a good president from implementing a long-term program that would benefit the nation. The provision, that is, is not an ‘abuse’ of constitutional rights or a deep flaw in constitutional design, but simply one that is now reasonably regarded as flawed. My view, developed in detail elsewhere, is that changing the term-limits provision by using the prescribed form of constitutional amendment can sometimes be regarded, in law, as a pro tanto constitutional replacement or revolutionary transformation.5 The ‘entirely new’ constitution is word-for-word the same as the old one, except for the new provision on term limits. The thought is that it is a mistake, as a matter of constitutional design and theory, to require that a nation’s people roll out the big guns of a constituent assembly simply to make changes in a purportedly unamendable provision that is only loosely coupled with other provisions that the people continue to regard as worth preserving. I pause here to address an objection to the argument that unamendable provisions can be amended through ordinary amendments rather than full-bore constitutional

4  Violence may occur in connection with revolutionary transformations, but it need not, and I think ignoring the costs of violence promotes clarity in thinking about these matters of constitutional theory. 5  M Tushnet, ‘Peasants with pitchforks and toilers with Twitter: Constitutional revolutions and the constituent power’ (2015) 13 Int’l J Const L 639.

100  Mark Tushnet replacement. The objection is that packaging a major change as an ordinary amendment may conceal the significance of the change—or, perhaps more precisely, the significance the Constitution’s original drafters attributed to the provision, as evidenced by their decision to make it unamendable. But, that is precisely the point of the argument in favour of treating ordinary amendments of purportedly unamendable provisions as pro tanto constitutional revolutions. The purported unamendability signals the framing generation’s judgement about the provision’s significance. That signal remains for the amending generation to view, and for opponents of amendment to point to. But, the amendment’s proponents challenge the underlying judgement about the provision’s importance. For them, the purported unamendability is like a traffic signal at a once-dangerous intersection that has become entirely safe as a result of changes in traffic patterns. They see no point in digging up the entire intersection simply to remove the traffic signal. Next consider an unamendable standard-like provision—federalism, for example— whose interpretation is left in the legislature’s hands. Suppose the legislature changes some feature of existing federal arrangements, for example by shifting some topic from the list of powers held exclusively by the subunits to the list of concurrent powers. Some people will treat the shift as altering federalism in a way barred by the unamendable commitment to federalism. Because of the value, emphasised by Doyle, of allowing a nation’s people to govern themselves, I think it better to treat the shift as an alternative—and permissible—specification of federalism’s meaning, at least if there are reasonable grounds for thinking that the new allocation of powers leaves the nation as a federal system in the broadest terms.6 Finally, what of an unamendable standard-like provision—secularism, for ­example—that the courts are authorised to interpret? Suppose the legislature enacts a statute authorising religious institutions to participate in some general social welfare program, such as the nation’s system for placing children for adoption, and suppose that the constitutional court holds the statute unconstitutional because it is inconsistent with the court’s understanding of the constitutional commitment to secularism. Assume that that holding is not unreasonable (it would not be, in my judgement, on standard understandings of what secularism requires). Still, the holding is not compelled by the very notion of secularism: Secularism, again on standard understandings, is compatible with the inclusion of religious institution in general social welfare programs. Ordinary amendment processes cannot be used to put in place the alternative reasonable specification of secularism’s meaning, because the Constitution’s secularism provision is unamendable. What can the nation’s people do to reclaim their power of self-governance from the judiciary? As always, constitutional substitution or replacement remains possible, but as before, invoking those processes seems excessive as a means of changing the outcome of a single judicial proceeding. And again, the solution seems to me to lie in treating an ordinary constitutional amendment as a pro tanto substitution or replacement. 6  Shifting all of the powers allocated to the subunits to the list of powers exercised exclusively by the centre, for example, would not preserve the nation as a federal system in any sense. Whether shifting all the exclusive powers of the subunits to the list of concurrent powers would eliminate federalism is, I think, an interesting and difficult question in the theory of federalism.

Comment on Doyle’s Constraints on Constitutional Amendment Powers 101 This discussion of how we can reconcile a people’s ability to govern itself at time-2 with judicial interpretation of a provision subject to a time-1 limitation on amendment brings me to another important theme in Doyle’s argument. That is that unamendability coupled with judicial interpretation may result in a substantial loss to a people’s power to govern itself. Here I think it important to rid ourselves of a form of rhetoric that elides the real differences between the people at time-1—the founding generation, for example—and the people at time-2 who want to change the Constitution through which they govern themselves. The actual individuals who are the people at time-2 are, in all interesting cases, different from those at time-1. We can posit a certain kind of historical continuity between the two sets of different people. And, in so positing, we can include in the historically continuous people many who were not directly represented at time-1; the usual examples here are women and African Americans for the United States at the time of the framing, and the usual ‘solution’ or postulate is that they were virtually represented by the white males who framed the US Constitution. And, we can even posit such a continuity for some nations of immigration such as the United States; here the ‘solution’ lies in the observation that the Constitution’s principles are universalistic, capturing universal human rights.7 But, for all that, it remains true that it is only in a metaphorical sense that ‘we’ at time-2 find ourselves bound by decisions that ‘we’ at time-1 made. Rather, we are bound by decisions that some other people made, and to the extent that we are unable to alter those decisions we are not fully self-governing. As Doyle notes, in its strongest form this is a difficulty with constitutionalism, and is ordinarily addressed in constitutional theory by pointing to the fact that we can indeed govern ourselves by amending the Constitution. But, if the provisions we want to amend are stated to be unamendable, and if courts stand ready to enforce unamendability, the existence of the amendment process leaves us not fully self-governing. Doyle’s emphasis on how unamendability can transfer governing power to the judiciary may slight the insight that drives his argument overall: Responsibility for our inability to govern ourselves is actually divided between two institutions, the Constitution that says some provisions are unamendable, and the courts that enforce unamendability. I turn now to Doyle’s discussion of what he calls the most difficult case, where provisions are made unamendable in the service of countermajoritarian values. Suppose that a constitution makes unamendable provisions dealing with freedom of expression in connection with elections, or provisions dealing with the rights of religious minorities. The usual technique for promoting countermajoritarian values like these is to embed them in a bill of rights. But, ordinarily, even bill of rights provisions can be amended so long as an appropriate supermajority can be obtained. One can imagine, though, that a constitution’s framers, sensitive to their nation’s specific conditions and history, could believe that there is a worrisome risk that the requisite supermajority for constitutional amendment could actually be attained. Further, they might have the same view about the risks associated with creating a

7  I note, though, that this solution is almost certainly unavailable with respect to a constitution’s structural principles, including things like presidential term limits and, probably, federalism.

102  Mark Tushnet ‘ratcheted’ supermajority requirement of the sort suggested by Richard Albert, to the effect that, while some constitutional amendments can be made by a 60 per cent majority, others require at two-thirds or three-quarters majority. Their solution is to make unamendable the provisions that need the strongest possible protection against majoritarian overreaching. The intuition underlying Doyle’s concern is obvious, and in one sense obviously correct. Put most starkly, it does not make much sense to allow constitutional provisions whose point is to protect against majoritarian overreaching amendable by a majority. Requiring a supermajority to amend such provisions may not be sufficient either. But as the supermajority required for amendment increases, the risk of overreaching decreases, with the effect that fewer and fewer constitutional provisions need the strongest possible insulation from amendment via unamendability. And that is where the rub occurs. Return to Doyle’s distinction between standard-like and rule-like constitutional provisions. Most unamendable provisions in contemporary constitutions are standard-like. The countermajoritarian ones insulate institutions such as secularism, human dignity, and democracy against amendment. The difficulty, to which I alluded above, is that specifying the content of these ­standard-like provisions is almost always going to be contestable—and contested, when a nation’s people wants to amend the Constitution to allow the legislature to enact legislation that they believe consistent with secularism, human dignity, or democracy. So, for example, suppose a constitutional court holds that a law allowing the wearing of head coverings in certain venues is inconsistent with secularism, or that a law making it easier for women to obtain abortions is inconsistent with human dignity, or that laws regulating campaign finance are inconsistent with democracy. Each of these positions, while reasonable specifications of the standards, is also contestable within the discourses of secularism, human dignity, and democracy. So, making standard-like countermajoritarian provisions unamendable might disempower a majority that believes its specification of those standards is better than the court’s.8 As I just noted, we rarely find rule-like countermajoritarian human-rights related provisions made unamendable. It is not clear to me why this is so. Perhaps constitution drafters think that the particular rule-like provisions on offer are too specific to be included in a constitution. Or perhaps they think that the chances are quite slim that a supermajority will actually favour legislation inconsistent with that position, so that it is enough to include the provision in the Constitution, without making it unamendable. Or perhaps they simply cannot agree on the specification of a rule-like principle that captures the concerns that animate other constitution drafters who insulate standard-like principles such secularism, human dignity, and democracy from amendment. The upshot, though, is to make the difficulty Doyle finds in the

8 Of course we can also imagine unreasonable majority specifications of the standards. My view, though, is that were a majority to favour an unreasonable specification, the nation is probably in such deep trouble anyway that constitutional provisions are unlikely to be enough to save it. Here I echo Learned Hand’s famous comment: ‘a society so riven that the spirit of moderation is gone, no court can save’. Learned Hand, ‘The Contribution of an Independent Judiciary to Civilization’, Address at the 250th Anniversary of the Supreme Judicial Court of Massachusetts (21 November 1942), in L Hand, The Spirit of Liberty: Papers and Addresses of Learned Hand (Irving Dilliard ed. 1952) 118.

Comment on Doyle’s Constraints on Constitutional Amendment Powers 103 use of unamendability relevant in connection with countermajoritarian constraints: the problem associated with specification in connection with standard-like provisions, coupled with the near-absence of rule-like provisions, means that even countermajoritarian unamendability constraints are difficult to justify. Finally, Doyle’s position that making what he calls ‘foundational’ provisions unamendable is clearly correct when taken on its own terms, but may be undermined by expanding our view of constitutional theory to include the possibility of constitutional replacement. Foundational provisions deal with such matters as national sovereignty and the integrity of the national territory. That such provisions should be amendable can be seen quite clearly if we imagine a nation whose constitution provides, unamendably, that the national territory must be preserved. Suppose that the nation faces a military adversary that clearly could invade and seize some province. Instead of resisting militarily and suffering serious human and material costs, the nation’s leaders simply cede the province to the adversary. It would be a very badly designed constitution that made such an action unlawful by, for example, making the leaders’ actions a ground for removing them from office. A somewhat more realistic example involves the creation of a European federation whose component units would not have sovereignty in the traditional sense. Assume (unrealistically at the moment, and likely unrealistic for some time to come) that the people of what are now Europe’s nations come to think that becoming members of a federation that absorbs their nation’s sovereignty is desirable. Judicial enforcement of an unamendable foundational provision that denies them the ability to act on that belief seems an especially substantial denial of the people’s right to govern themselves.9 Now, though, I must retreat—as, in some sense, must Doyle. The reason is that even judicially enforceable unamendable foundational provisions can be changed in a manner entirely consistent with constitutional theory. ‘All’ that is required is that the Constitution containing such provisions be replaced. And such a replacement can occur through means authorised by the Constitution itself. A constituent assembly can be convened, for example, and it can propose a new constitution that defines the national territory in a way that excludes the ceded province. Even more, as I have argued here and elsewhere, constitutional theory must leave some space for the possibility of real constitutional revolutions, whether violent or peaceful. Further, though more controversially, as I have suggested above, it would be a good idea—would improve constitutional theory—were it possible to speak of pro tanto constitutional revolutions. But, of course, accommodating constitutional replacement in any of these ways lowers the theoretical stakes of the claim that the strongest form of unamendability denies the people the power to govern itself. As Doyle’s essay suggests, and as is confirmed by the other essays in this collection, the topic of constitutional unamendability offers the opportunity for rich explorations of constitutional theory. Almost every participant in those explorations acknowledges that the practical stakes are relatively low, because ­unamendable

9  This makes quite problematic the hints from the German Constitutional Court that the German Basic Law prevents Germany from becoming part of a truly federal Europe.

104  Mark Tushnet ­ rovisions rarely obstruct serious movements to change the Constitution10 and p because even those courts that have asserted the general power to strike down constitutional amendments as unconstitutional have rarely done so. I have suggested in these comments that the theoretical stakes might be low as well, because unamendabilty can be overcome by express or pro tanto constitutional replacement.

10  This point must be qualified by the possibility that the emergence of such movements is inhibited by the very existence of the provisions.

5 Constituting the Amendment Power A Framework for Comparative Amendment Law THOMAZ PEREIRA

I. INTRODUCTION

T

HE DISTINCTION BETWEEN constituent and constituted power is as old as modern constitutions.1 It is intrinsic to the very concept of constitutionalism. Since, if the Constitution constitutes, it is therefore necessary that there is something constituting, and something that is constituted by it. This fundamental distinction relates to the two essential innovations brought out by the emergence of constitutional democracies. First, that a new political order could be the product of a foundational political act. Secondly, that this foundational political act, for being substantively different from other political decisions, could establish higher-order legal norms regulating the production of all other legal norms. The deep connection between these two revolutionary ideas was already clear to Sieyès, and is at the basis of his canonical elaboration of the distinction between c­ onstituent power (‘pouvoir constituant’) and constituted power (‘pouvoir ­constitué’).2 It is particularly important to bear in mind that, although Sieyès ‘in the field of theory, had no peer among the men of the French Revolution’,3 this was not the product of a legal scholar preoccupied with minute abstract distinctions. On the contrary, this was a fundamental political question to revolutionaries involved with the founding of a new political order. This is illustrated by the very fact that Sieyès’ often quoted statement that ‘a constitution assumes before anything else a constituent

1  By ‘modern constitutions’, I mean the conception of constitutions as the product of a fundamental political act—developed theoretically and in practice by the revolutionaries involved with the founding of the United States of America (1776–1789) and the French Revolution (1789–1799). In this sense, it is opposed to the earlier descriptive use of ‘constitution’ as referring to the rules regulating public power. For a discussion of the different meanings of constitutionalism, see Ch H McIlwain, Constitutionalism Ancient and Modern (Cornell University, 1940). 2  E Sieyès, Qu'est-ce que le Tiers-État? (first published 1789, PUF, 1982). 3  H Arendt, On Revolution (Penguin, 2006) 154.

106  Thomaz Pereira power’, was made in 1789 before the Constitutional Committee of the French National Assembly4—and not in an obscure academic seminar. It is this real and still very current dimension of an old conceptual discussion that I wish to emphasise; a dimension in which ‘form’ followed from ‘substance’, and not the other way around. In this sense, for the distinction between constituent and constituted power to be able to carry any weight, it must be based on a substantive differentiation. And, in regard to constitutional democracies in which the legitimating principle of the political order is popular sovereignty, what is at stake is the relationship between constituent and constituted powers and this foundational principle. In these cases, the underlying assumption is that the constituent power is legitimated by its identity with the ‘will of the people’. This identity legitimates the institutional order produced by its exercise, including an amendment procedure regulating the terms in which the higher-order norms establishing this institutional order can be revised. But if this procedure is supposed to channel future exercises of popular sovereignty, it can inevitably be criticised for its (in)ability to do so. It is in this context that ‘the truly radical implication of the notion of amendment is more than a little frightening’.5 On the one hand, the idea that the amendment process institutionalises and legitimates revolution6 raises the fear that the process might be used to legitimate constitutional transformations that do not possess deep, broad and decisive popular support. On the other hand, there is the fear that the required procedure might be a conservative tool to limit the will of the people from bringing about legitimate constitutional transformations. These two mirror-­ problems of ‘false negatives’ and ‘false positives’ are intrinsic to any exclusively formal system.7 In either case, what is needed is an independent base from where to criticise the procedure. In this way, because the amendment power lives in this space of continuous tension between form and substance, constitutional scholars have mostly been occupied with: (i) justifying adherence to form, independently of substance;8 (ii) providing a way to identify the presence of substantive requirements, independently of form;9 or (iii) defending the existence of explicit and/or implicit limits to the amending power.10 In this chapter, I intend to do none of these three. Instead,

4  ‘Une constitution suppose avant tout un pouvoir constituant’. E Sieyès, Préliminaire de la Constitution—Reconnaissance et Exposition Raisonnée Des Droits de l’Homme et du Citoyen (Lu les 20 et 21 Juillet 1789, au Comité de Constitution) 19. 5  S Levinson, ‘Introduction: Imperfection and Amendability’ in S Levinson (ed), Responding to Imperfection: The Theory and Practice of Constitutional Amendment (Princeton University, 1995) 4. 6  G Wood, The Creation of the American Republic: 1776-1787 (UNC, 1969) 314–15. 7  B Ackerman, ‘Higher Lawmaking’ in Sanford Levinson (ed), Responding to Imperfection: The Theory and Practice of Constitutional Amendment (Princeton University, 1995) 85. 8  For a defense of a wholly procedural understanding of amendment rules, see J R Vile, ‘The Case against Implicit Limits on the Constitutional Amending Process’ in S Levinson (ed), Responding to Imperfection: The Theory and Practice of Constitutional Amendment (Princeton University, 1995) 191. 9  For a theory a defense of an anti-formalist understanding of amendment rules, see B Ackerman, We the People: Foundations (Harvard University, 1991). 10  See D Conrad, ‘Constituent Power, Amendment and Basic Structure of the Constitution: A Critical Reconsideration’ in J Lütt and M P Singh (eds), Zwischen den Traditionen: Probleme des Verfassungsrechts und der Rechtskultur in Indien und Pakistan (Franz Steiner, 1999) 87; W F Murphy, ‘Merlin’s

Constituting the Amendment Power 107 I believe we can achieve fundamental insights from two separate, but interconnected, undertakings: First, by going back to Sieyès and understanding the substantive legal-political problems he was struggling with, we can gain renewed insight into the relevance of the formal distinction between constituent and constituted power for political orders that have their origin in a revolutionary foundational act. Secondly, I argue that, because not all constitutional orders are created equal, an adequate understanding of the constituent/constituted power divide depends, not on a choice between different abstract constitutional theories, but on the political history and legal culture of different constitutional systems. Because of the deep connection between constituent power and popular sovereignty, the relevance of the latter for the sustained legitimacy of a constitutional order will necessarily determine the relevance of the former. Moreover, even in regard to constitutional orders that are based on the idea of popular sovereignty, constitutional systems can be profoundly different in the ways in which ‘the people’ is (un)able to speak after the founding moment is over. This will lead us to the conclusion, in which I will argue that comparative constitutional law can benefit from these two fundamental insights, which can illuminate our understanding of the deep relationship between the way different constitutional systems regulate the amendment power, and the political history and constitutional culture of these different systems. A.  Sieyès’ Two Different Problems Sieyès made himself famous with the brochure ‘Qu’est-ce que le Tiers Etat?’11 (What is the Third Estate?), which became a sensation among the French revolutionaries of his time and the de facto charter of the representatives of the Third Estate in the Estates-General meeting.12 Nowadays, Sieyès remains famous among constitutional scholars mostly for his conceptual differentiation between constituent and constituted power, elaborated in his writings and in speeches before the French National Assembly.13 These ‘two Sieyès’ are one and the same. Although Sieyès’ conceptual differentiation remains, something has been lost. The abstract concepts were preserved, but the historical political context that generated them is usually missing. His concepts were developed to serve specific needs. His treatment of the constituent power was connected with the specific context of

Memory: The Past and Future Imperfect of the Once and Future Polity’ in S Levinson (ed), Responding to Imperfection: The Theory and Practice of Constitutional Amendment (Princeton University, 1995) 163. 11 

Sieyès (n 2). published in January 1789, it required four editions during this same year. See J Tulard, ‘Préface’ in Sieyès (n 2). 13  See E Sieyès, Préliminaire de la Constitution—Reconnaissance et Exposition Raisonnée Des Droits de l’Homme et du Citoyen. (Lu les 20 et 21 Juillet 1789, au Comité de Constitution) 19. See also P Bastid (ed), Les Discours de Sieyès dans les Débats Constitutionnels de l’An III (2 et 18 Thermidor) (Hachette, 1939). 12  Originally

108  Thomaz Pereira attempting to found a new constitutional order through a revolutionary popular movement. It is my belief that much can be gained by reconnecting these two concepts with their historical origins. More specifically, Sieyès had two different problems that he needed solved in order to successfully found a new constitutional order, and elaborating a conception of the constituent and constituted powers was part of the solution. He had more luck with one, than with the other—but then again, success only seems easy after having been achieved. i.  Constituting the Constituent Power Sieyès’ first problem was: how to establish a legitimate source to enact a constitution for France? Leaving aside tradition and the monarch’s god given powers, there was need for something on which the legitimacy of the new order could be based. Sieyès’ project was to defend the idea that ‘the Nation’ was sovereign and, therefore, held the power to ‘constitute’. In other words, Sieyès’ first problem was constituting the constituent power itself. With this in mind, Sieyès conceptualised the Nation as a unity in opposition to the fragmented notion behind the Estates-General—a political institution supposed to represent the French society that had just been assembled for the first time since 1614—an organ based on an older notion of representation that was inadequate for Sieyès’ purposes. In the Estates-General original configuration, each class (‘Estate’) that constituted the French society of the time was supposed to be represented as a group, voting ‘as one’ on the issues put before it. This is a model of representation that is similar to the English one, in which the House of Commons was supposed to represent the commoners and the House of Lords the nobility, and sovereignty was understood to be held by ‘the king in Parliament’, that is, the collective will of commoners, nobles and the monarch.14 This is already a big step from understanding that the monarch is the sovereign, but it is still very far from our modern notion of popular representation. A notion that was being born at that very time, and had Sieyès as one of its midwives. It was in this context, and against this model, that Sieyes wrote ‘Qu’est-ce que le Tiers Etat?’. The point was clear from its thundering beginning: 1  What is the Third Estate? Everything. 2  What has it been until now in the political order? Nothing. 3  What does it ask? To become something.15

To be ‘everything’ was to be the legitimate holder of sovereign power. A sovereign power that, in being exercised to constitute the new constitutional order, unlimited by constituted powers, was to become the constituent power. In this sense, the challenge at this time was constituting the constituent power itself. This required a subject to exercise it, which, for Sieyès, was the pre-political ‘Nation’.16

14 

MJC Vile, Constitutionalism and the Separation of Powers (Oxford University Press, 1967). E Sieyès, What is the Third Estate? (Praeger, 1964). this, as in other things, Sieyès was influenced by his reading of Rousseau: ‘before examining the act by which a people elects a king, it would be well to examine the act by which a people is a people. For 15 

16  In

Constituting the Amendment Power 109 The move was a radical one. The Third Estate was transformed from one class, among three that constituted the French society, into ‘une nation complète’17—a complete Nation all on its own.18 The practical demands were at first simply that the representatives of the Third Estate be actual members of this class, that there should be as many of them as of the other two Estates, and that the vote should be by representative—and not by Estate. But these pragmatic demands were based on the idea that the Third Estate represents the entire Nation and that the Nation, this pre-political all-powerful being, is the sovereign and, therefore, the holder of the constituent power—unlimited by constituted ones. This is perfectly captured by the eventual transformation of the Estates-General in the National Assembly, which would become the Constituent Assembly. At this moment the problem for Sieyès was ‘how people can grow into “the people”’—how a plurality can be transformed into a unity.19 The important thing to notice is that the constituent power was not an imagined abstraction that needed to be conceived to legitimate the extant constitutional order. On the contrary, the constituent power was based on the actual reality of popular sovereignty manifestations, and being invoked to justify the unlimited power of the Nation to constitute a new political order. Because ‘the modern doctrine of popular sovereignty coincides with the conceptual advent of constituent power’ and they are ‘co-original and coeval’,20 it is easy to simply conflate the two, but they are not the same. After constituting ‘the Nation’, and putting an end to the representation by ‘Estates—the materialization of the denial of one ‘sovereign people’, of a ‘we’ that, in being a unity could speak with ‘one voice’—, the problem became one of understanding the difference between when the representative institutions constituted by the new constitutional order speak as the bearer of ‘constituent power’, and when they don’t. This would be Sieyès second problem. ii.  Limiting the Constituent Power If Sieyès’ first problem was constituting ‘the Nation’ as having the constituent power to establish a new constitution. Sieyès’ second problem was constituting a constitution that would have authority to limit the National Assembly. Sieyès, with the famous distinction between constituent and constituted power, and by linking the constituent power with the unlimited pre-political Nation,

this act, being necessarily prior to the other, is the true foundation of society’. J-J Rousseau, ‘The Social Contract’ in V Gourevitch (ed), The Social Contract’ and Other Later Political Writings (Cambridge University Press, 1997) 49. 17 

Sieyès (n 2) 28. ibid, 28–32. 19  L Corrias, The Passivity of Law: Competence and Constitution in the European Court of Justice (Springer, 2011). For a critique of Sieyès conception of the nation a unified whole, see H Arendt, On Revolution (Penguin, 2006) 154–56. 20  A Kalyvas, ‘Constituent Power’ (Political Concepts: A Critical Lexicon) accessed 1 July 2016. 18 

110  Thomaz Pereira ‘seemingly solved (…) the problem of the legitimacy of the new power, the pouvoir constitué’,21 but his ‘solution for the perplexities of foundation (…) had not resulted and could not result in the establishment of a republic in the sense of “an empire of laws and not of men”’.22 Hannah Arendt develops a whole theory on why this was so,23 but independently of following her philosophical interpretation of the problems with French Revolution, the fact is that, although the revolutionaries were able to ‘liberate’ themselves from the old political powers, they were unable to ‘found’ a stable constitutional order24—as the revolutionary history of France, with constitution being followed upon constitution, easily attests. After successfully establishing the Nation as having the absolute power to legitimate the Constituent Assembly itself, so that it could enact a constitution, what was missing was a basis for the enacted Constitution to limit the power of the National Assembly. On the one hand, the National Assembly, having been constituted by the Constitution, derived its very existence from this source. On the other hand, if the National Assembly was understood as the legitimate representative of the Nation—the prepolitical absolute power that legitimated the very Constituent Assembly—how could it be limited in any way by a previously enacted law that was not the product of a higher authority, since there was none higher than it? Nevertheless, this difficulty was not based on deficiency in theory. It was based on specific historical-political facts. The reason behind ‘the great and fateful misfortune’ that ‘none of the [French] constituent assemblies could command enough authority to lay down the law of the land’,25 was that none of the Constituent Assemblies had been able to substantially differentiate themselves from the National Assembly that preceded it, and the National Assembly that it eventually constituted. There was no substantive basis for the people to feel a deeper attachment to the acts of the Constituent Assembly, as opposed to the acts of the constituted powers, since there was no reason to assume that one of them spoke for the Nation, and the other did not. Or, in technical terms, with no substantive basis to ground the higher authority of the Constitution, it was impossible to claim that ‘lex superior derogat legi inferiori’ (superior norms suppress inferior ones), but only that ‘lex posterior derogat legi anteriori’ (later norms suppress earlier ones). Although the French had the undeniable power to establish a new constitution, what was lacking was authority. If the Constitution was to serve as higher-law, capable of binding the government, there was a need for a higher-authority to establish it. Sieyès successfully established the Nation as the highest possible authority. What he and his fellow French revolutionaries were unable to do was to credibly differentiate the Constituent Assembly from the constituted National Assembly on the basis of this very authority. 21 

H Arendt, On Revolution (Penguin, 2006) 154. ibid 155. 23 ibid. 24  The difference between ‘rebellion’, as having its end on ‘liberation’, and ‘revolution’, as having its end on ‘the foundation of freedom’, is also Arendt’s. See Arendt (n 21) 132–45. 25  Arendt (n 21) 156. 22 

Constituting the Amendment Power 111 Sieyès was aware of this problem and he did try to resolve it through a distinction between ordinary representatives and extraordinary ones. The first are restrained by the forms and conditions imposed by the constitutional order. The second, extraordinary, are deputised by the Nation to—within the limits of their mandates—decide upon the constitutional forms themselves. As such, the extraordinary representatives are, according to Sieyès, like the Nation, independent from all constitutional forms, limited only by the specificity and provisory nature of their mandate.26 The conceptual distinction was there, but there was no actual institutional basis to support such a differentiation. In such a world, where the extraordinary is unable to distinguish itself from the ordinary, an unextraordinary constitution is likewise unable to differentiate itself from (extra)ordinary laws. iii.  Lessons from Failure and Success ‘The question of constituent power is, first, about whether power can turn into authority.’27 As such, going back to Sieyès and discussing the substantive legal-­ political problems he was struggling with allows us to understand that, for constitutional orders that have their origin in a revolutionary foundational act, the relevance of the formal distinction between constituent and constituted power depends on the capacity of the constituent power to establish its extraordinarity. It is only such a substantive differentiation from the ordinarily constituted power that can both legitimate the constituent power’s authority to found a new constitutional order and the authority of the Constitution itself to stand above the constituted powers. Sieyès was successful in his defense of the legitimacy of the constituent power, as a manifestation of the all-powerful pre-political Nation, to disregard existing constituted powers and found a new Constitutional order. But, as Andreas Kalyvas precisely states, ‘[t]he fundamental problem is not only how to differentiate clearly between the constituting power and the constituted power but also how to preserve both terms of the equation without sacrificing one to the other.28 For those not ready to abdicate the legitimating authority of the constituent power, the question is: how is one to preserve this concept and found a stable democratic constitutional order? In this, much can be learned from the failure of the French Revolution and the success of the American one.29 Much has been written on the causes why the Unites States revolutionaries were able to succeed in founding a Constitutional democratic republic. Geographic, social and economic explanations abound, but this is not what interests us here. The question is not what was the material basis that allowed the constitutional order to survive, but what, in the act of constituting the constitutional order itself, allowed the

26 

Sieyès (n 2) 70–72. D Dyzenhaus, ‘The Politics of the Question of Constituent Power’ in M Loughlin and N Walker, The Paradox of Constitutionalism: Constituent Power and Constitutional Form (Oxford University Press, 2008) 129. 28  A Kalyvas, Democracy and the Politics of the Extraordinary: Max Weber, Carl Schmitt, and ­Hannah Arendt (Cambridge University, 2008) 175. 29  This was Arendt’s basic project in writing On Revolution (n 21). 27 

112  Thomaz Pereira Americans to both clearly differentiate between constituting and constituted power and to preserve this difference without sacrificing one concept to the other? On the one hand, the difference was made clear not by political rhetoric, but by the actual extraordinarity of the constituent process. Although many of the delegates to the constitutional convention possessed the charismatic leadership to purport to speak for the people, this is not what happened. The fact is that they did have, as Sieyès had theorised, a mandate that was both specific and provisory in nature. But this was not all, after having achieved a specific draft they put it to a vote by other representatives, who also had specific and temporary mandates to vote on the proposed Constitution, an election held among the highest level of public debate about the specificities of this document. On the other hand, the difference was preserved by the Article V amendment procedure itself. Emulating the unconventional procedure that led to the enactment of the Constitution, it preserved a space for the extraordinary that served to both continuously legitimate the original foundational act and to differentiate it from the ordinary acts of Congress. And it is only in this sense that ‘[p]arliamentary sovereignty finds in the constituent power its own impossibility’ and is ‘exposed as a usurpation of the constituent power by a constituted power’.30 This was done by a procedural solution that differentiated acts ‘by the government’ from acts ‘by the people’, bestowing a special authority to the constitutional order itself. However, this was only successful because there was a substantive reality for the procedure to capture in the first place.31 In other words, the procedure was successful not because it was able to invent a popular sovereign that was not there, but only because it was able to credibly channel the ‘will of the people’ in a way that inevitably differentiated it from the day-to-day politics of the constituted power.32 In this ways, it was only by limiting itself that the constituent power was able to establish an enduring constitutional order. What is important to bear in mind is that the transformation from popular sovereign into constituent power depends on this very ability, because a failure to institute a Constitution ‘amounts to the inability of the constituent power to realize itself’.33 This can only be done through a transformation from power into authority, because only authority can bestow a higher-law nature to the enacted Constitution when confronted by other acts by constituted powers that also claim to speak for ‘the people’. It is only the substantive credibility of the extraordinary representational character of the founding act that can confer such an authority, and it is only by emulating this same extraordinary character that the amendment procedure can serve as a continuous legitimating element of the constitutional order itself, while having the legitimacy to change what was enacted by the constituent power.

30  A Kalyvas, ‘Popular Sovereignty, Democracy, and the Constituent Power’ (2005) 12 Constellations 223, 229. 31  Wood (n 6). 32  For a detailed account of the history, philosophy and institutional development of the two-track US Constitutional system, see Ackerman (n 9). 33  Kalyvas (n 30) 234.

Constituting the Amendment Power 113 In the context of constitutional democracies founded upon the legitimating principle of popular sovereignty, there is an undeniable ‘tense and ambivalent relation’ between the constituent power and the constitutional order it founds, since ‘the sovereign moves uneasily inside and outside the Constitution, escaping, sometimes resisting, its total absorption by the instituted reality’.34 Taming this extraordinary power requires a deep understanding of the possibilities and the limits of constitutional design. In this regard, nothing captures this relationship as perfectly as the regulation of the power to amend. In constituting the amendment power the constituent power establishes the fundamental terms upon which the constitutional order will develop. Because of this, in understanding the terms in which this choice can be made, one gains invaluable insight into the relationship between form and substance when it comes to constitutions. In this way, a framework regarding the regulation of amendment power can provide a typology upon which the whole field of comparative constitutional law can be grounded on. B.  Constituting the Amendment Power ‘Does constituent power disappear at the moment it turns into authority or does it hang around, threatening to disrupt or destroy its creation?’35 This is the fundamental question that an analysis of the nature of the amendment power must answer. It is this sometimes baffling question that is at the basis of the recurring discussions regarding if the amendment power should be understood as constituent or constituted power. The point of the previous discussion on the historical-political context in which this differentiation was originally developed is to make clear that this is not a question that can have a single one-size-fits-all answer. Exactly because constitutional orders are profoundly different in regard to what legitimates their authority, the nature of the amendment power is most likely going to mirror this fundamental difference. There is, first, a basic mistake in assuming that the relationship between popular sovereignty and constitutional democracy is the same in every constitutional context. And, secondly, a fundamental mistake in assuming that two concepts that were developed in the context of a truly popular revolutionary scenario can easily serve to explain the relationship between the authority of the constitutional system and the institutions constituted by it in every situation. Because not all constitutional orders are created equal, an adequate understanding of the constituent/constituted power divide depends, not on a choice between different abstract constitutional theories, but on the political history and legal culture of different constitutional systems. Thus, the constituent/constituted power divide puts two different issues at the table: (i) how to differentiate clearly between them? and (ii) how to preserve both

34  35 

Kalyvas (n 30) 227. Dyzenhaus (n 27) 129.

114  Thomaz Pereira concepts without collapsing one into the other?36 As such, what is at stake for a theory of the amendment power is: how in fact the amendment procedure relates to them both? In this regard, the deep connection between constituent power and popular sovereignty means that the relevance of the latter for the sustained legitimacy of a constitutional order will necessarily determine the relevance of the former. And even in regard to constitutional orders that are based on the idea of popular sovereignty, constitutional systems can be profoundly different in the ways in which the constituent power is differentiated from and preserved within a constituted constitutional system, which can be designed in ways that allow and ways that don’t allow for ‘the people’ to legally speak after the founding moment is over. It is this real dimension of an old conceptual discussion that I wish to emphasise— a dimension in which ‘form’ follows from substance, and not the other way around. For the distinction between constituent and constituted power to be relevant, it must be based on a substantive differentiation and, in the context of constitutional democracies, because the legitimating principle of the political order is popular sovereignty, what is at stake when discussing the nature and limits of the amendment power is the relationship between constituent and constituted power and ‘the will of the people’. Regarding this relationship between constituent power and popular sovereignty, there are two options available: (i)

(ii)

The first, skeptical of the substantive existence of such a thing as ‘the will of the people’, denies the substantive identity between constituent power and popular sovereignty. As such, it must: (a) justify its power to establish higherorder legal norms on some other legitimating principle, or (b) deny the foundational political act and, therefore, the Constitution, any authority to bind a democratically elected government. The second, not only believes in the potential substantive existence of ‘the will of the people’, but also recognises the substantive identity between constituent power and the popular will. As such, it has no problem justifying its power to establish higher-order legal norms, but must elaborate on the relationship between the constituted power and popular sovereignty.

In order to do this, there are three options available: either the constituted power, despite being constituted by a manifestation of popular sovereignty: (a) is not able to represent it in any situation, or (b) is able to represent it at all times, or (c) is able to represent it in some situations but not in others. In the first of these three hypotheses, the popular will is conceived as a pre-­ constitutional and extra-constitutional power. It exists and substantively legitimates the current political system. However, after its original (and originating) manifestation it has no legal channel to manifest itself in the political arena. In this way, the products of its foundational act are forever preserved in the extant constitutional order, and can only be revised by an illegal/revolutionary act.

36 

Kalyvas (n 28) 175.

Constituting the Amendment Power 115 In the second, the popular will never retires from the scene. After establishing a political system, it continues to manifest itself in every political act. In this extraordinary world, there is no such a thing as ordinary political acts. In this political system, there is no hierarchy between the foundational political act and any other political act. The only criteria to differentiate one from the other is temporality. In the third hypothesis, the popular will after establishing the constitutional order preserves certain ways for it to manifest itself again. In this regard, the tricky part is to differentiate the ordinary from the extraordinary, and procedure and substance are two ways to do it. In determining this, the nature of the amendment procedure is fundamental. Is it the case that some acts are ‘the will of the people’ and others are not, or is it the case that some of them are thinner manifestations of the same will? In one situation, elected representatives are not ‘representatives of the will of the people’ if they do not have a special mandate specifically granted to exercise this power. In the other, they are always ‘representatives of the will of the people’, but they only speak ‘for the people’ through certain procedures—a difference normally caught by the ordinary/amendment law divide. i.  People Who? References to ‘the people’ as the authority upon which the constitutional order is constructed have become commonplace in a world in which constitutional democracies abound.37 Nevertheless it is a mistake to assume that even among systems legitimated by this principle, ‘the people’ has the same concrete legitimating force in all constitutional orders. There is a major divide between authors that believe that ‘the sovereign people’ has a concrete reality38 and authors that deny it, describing constitutional systems as elite constructions.39 The point of this article is that this is not an either/or type of discussion. There is a basic problem with attempting to discuss this issue in abstract as if there was only one possible answer. In reality, not all constitutional systems are founded and legitimated upon the same principle. Different constitutional cultures are the product of different political-historical developments—something evident to comparative constitutional law scholars. In this way, for those who deny popular sovereignty the capacity of legitimating a constitutional order, the only options left are either to deny constitutional law any

37  For a discussion on the political meaning of ‘the people’ in traditional and modern constitutions, see JP McCormick, ‘People and Elites in Republican Constitutions, Traditional and Modern’ in Loughlin and Walker (n 27). 38  For an analysis of the reality of popular sovereignty in the historical context of the American Revolution, see Wood (n 6). For a reconstruction of the American Constitutional History on the basis of the reality of ‘we the people’, see Ackerman (n 9). 39  For an indictment of the founding fathers of the Unites States Constitution and their elitist interests, see Ch A Beard, An Economic Interpretation of the Constitution of the United States (first published 1913, Free Press, 1965). For a contemporary discussion of the relationship between the rise of constitutionalism and the interests of economic, political, and legal elites, see R Hirschl, Towards Juristocracy (Harvard University, 2007).

116  Thomaz Pereira higher authority over normal legislative acts, or to base its authority in some other principle, which in secular contexts tends to be some form of liberal political philosophy. Undeniably, any of these two can happen as a historical reality. As such, in one of them the special authority—and maybe even the existence of any constituent power—is denied, in the other the constituent power is made irrelevant in as much as the legitimating principle is not some form of legitimately exercised power, but the authority of a philosophical conception as such, as materialised by the democratic constitutional order. In a more concrete way, if we stick to the traditional canonical examples, the relationship between constitutional law and popular sovereignty is completely different when it comes to constitutional systems as diverse as the United States, England, and Germany. In each of these orders, both the constitutional founding and the specific institutional design of the constituted system are different in their relationship with ‘the people’. As such, while in England the people is understood to legitimate the systems more through their inaction than through their actions—by electing governments with a mandate to change the traditional institutions with ordinary legislative acts—in the United States the people are understood to have concretely legitimated the system though their revolutionary founding. On a completely different note, the German constitutional system is averse to references to the will of the people, holding ‘populism’ as a very bad word. Because of this, there is a problem with conceptualisations of the relationship between constituent and constituted powers that disregard the fact that ‘it is, therefore, no accident that the constitutional discourse of the Federal Republic adopted a thoroughly formalist approach to the issue [of the] constituent power’.40 In this way, in the first two, ‘the people’ is invoked as a legitimating principle, but in the first as an abstraction and only in the second as a concrete historical agent. Meanwhile, in the third, ‘the people’ is not even invoked as the legitimating subject of the constitutional order. As the foundational political act, the Constitution is simultaneously an expression of politics limiting law, and a legal document that limits politics and, because of this, there is a deep connection between the foundations of political order and the nature of constitutional law.41 This is the source of constitutional theory’s focus on the perceived tension between constitutionalism and democracy42—a conflict that

40  Chr Möllers, ‘We are (afraid of) the people: Constituent Power in German Constitutionalism’ in Loughlin and Walker (n 27) 104–05. 41 P Kahn, ‘Reason and Will in the Origins of American Constitutionalism’ (1989) 98 Yale LJ 449, 449. 42  A Bickel, The Least Dangerous Branch (Bobbs-Merrill, 1962); J H Ely, Democracy and Distrust: A Theory of Judicial Review (Harvard University, 1980); R Dworkin, Law’s Empire (Harvard ­University, 1986); B Ackerman, ‘Constitutional Politics/Constitutional Law’ (1990) 99 Yale LJ 453; Ackerman (n 9); B. Ackerman, We the People: Transformations (Harvard University, 1998); ‘The Living Constitution’ (2007) 120 Harv L Rev 1737; J Waldron, Law and Disagreement (Clarendon, 1999); F Michelman, Brennan and Democracy (Princeton University, 1999); C Sunstein, Designing Democracy: What Constitutions Do (Oxford University Press, 2001); J Habermas, ‘Constitutional Democracy: A Paradoxical Union of Contradictory Principles?’ (2001) 29 Pol Theory 766.

Constituting the Amendment Power 117 has captured the way different constitutional systems are portrayed by comparative constitutional law. This has led constitutional theory to often depict constitutional systems, first as manifestations of abstract concepts, and only second as concrete ways in which politics is structured in different democracies. In this way, following our previous example, England, the United States and Germany have been transformed into holding places for parliamentary sovereignty, democratic constitutionalism, and constitutional supremacy, respectively; and all other constitutional democracies tend to be understood as representations of each one of these models. What should be kept in mind—and the history of the American and French Revolution can help us do it—is that the question of the concrete existence of an act of the popular sovereign as the legitimating agent of a constitutional order is not a question of theory, but a question of history, culture and politics. What philosophy can do is discuss the feasibility of this concept as an adequate grantor of legitimacy, in abstract, and the potential problems of adopting this solution. It is not my intention to defend the exclusive nature of popular sovereignty as the sole grantor of legitimacy to Constitutional democracies. On the contrary, what this article argues is that the way in which we conceive of the nature of the constituent power is fundamentally shaped by this historical reality. In any case, the legal regulation of the power to amend is a window into the relationship between ‘the people’ and the constituted powers and, in those constitutional systems that are based on the principle of popular sovereignty, it is inevitable that the nature and limits of the amending power are going to depend on the way this relationship is conceived. ii.  People When? Putting aside constitutional systems that do not base their legitimacy in a concrete exercise of popular sovereignty, the relationship between ‘the will of the people’ and amendment regulations becomes even more concrete. In this regard, what is at stake is not a conflict over the existence or not of a ‘popular sovereign’, but in what situations it is possible to establish that ‘the people has spoken’. As already stated, there are essentially three ways in which this relationship can develop: a.  ‘The People’ has Left the House This is the most common way in which the relationship between popular sovereignty and constituted power is usually conceived. According to this, although the foundational political act that generated and still legitimates the constitutional authority is based on an act of ‘the people’, this is a unique form of manifestation that has no equivalent within the constituted order. This narrative is common both in cases in which there was actually such an original popular act, and in situations in which despite the historical inaccuracy of this version the origin of the constitutional order has been successfully mythologised in such a way.

118  Thomaz Pereira In any case, what is relevant for our purposes is that ‘the people’ has left the house. There is no institutional way that has been left open, formally or informally, that is understood to allow for the people to express itself once again on the fundamental choices regarding the political order. From a theoretical perspective, this is a particularly difficult set up to be legitimated by constitutional theory. While this is not the only version of constitutional design that refrains the people from deciding on certain fundamental elements of the constitutional order, there is a basic paradox in both legitimating the constitutional order on the principle of popular sovereignty, and not allowing the popular sovereign to speak ever again. This is a different situation from the one discussed above, in which the limits on the popular sovereign are based on an alternative account—for example liberal political philosophy—for the legitimacy of the constitutional order. This conceptualisation of the legitimacy of the constitutional order and the nature and limits of the constituent power can take the form of a narrative that is almost exclusively semantic. According to this, the point is not that the popular sovereign cannot express itself ever again. It is only that, in the event that this happens, this would mark the beginning of a new constitutional order—irrespective of the transformation having been formally conceptualised in this way or not. In this sense, this narrative is capable of differentiating the constituent power from the constituted power in a strict way, but it is only capable of preserving this distinction by sacrificing any possibility for ‘the people’ to speak again within the constituted legal order. In the end, it is as if the only possible way to preserve it is to sustain a topological criteria according to which acts by the people that precede the enactment of the constitutional order may have a constituent character, but substantive similar acts that happen after such a date must either conform to the constituted structures or be deemed illegal. b.  ‘We’ are Always Open On the opposite side of the same spectrum, another way of conceptualising the relationship between popular sovereignty and constituted powers is by considering that there is nothing extraordinarily different between the popular representation that legitimated the foundation of the constitutional order and subsequent acts by representatives from the people. According to this, in a representative democracy the people is always able to speak and there is no fundamental difference in the authority of the constituent assembly and the constituted legislative. In this world, the people speaks all of the time, and what should be considered as binding is whatever was said last. This is a world of ‘parliamentary sovereignty’, and it could be argued that it amounts to an abdication to the specific element that distinguishes modern constitutional democracies: the very higher authority of the Constitution. As such, there is a basic problem in differentiating constituent from constituted powers, which end up collapsing into each other. One must bear in mind that, in this situation, it is not the case that there is a refusal of popular sovereignty as the fundamental principle upon which the constitutional order is built—substituted by tradition, or charisma, or rationality, for

Constituting the Amendment Power 119 example—, but a denial of the difference between the foundational act and subsequent ones by the people. In this world, the problem of amendment does not even pose itself as a real one. Since there is a denial of the extraordinary element of constitutional legislation, there is no need to try to develop procedural and/or substantive ways to differentiate them from ordinary legislation. In this way, the deep connection between the constituent/constituted powers divide and the amendment regulation is made clear by the fact that, once we conflate the two, the difference between amendment power and ordinary legislation also vanishes from the scene. c.  Follow the Yellow Brick Road Between these two extremes there is an intermediate situation. One that allows for popular sovereignty to legally speak within a constitutional system, but strives to differentiate this kind of acts from other acts by elected representatives. There are fundamentally three ways to understand this difference: In one situation, the elected representatives are always ‘representatives of the will of the people’, but they only speak ‘for the people’ through certain procedures (for example, ordinary law vs. amendment law). In another, they are only ‘representatives of the will of the people’ in a strong sense—meaning a sense that grants them popular sovereignty—if they have a specially granted mandate for this. According to this second conceptualisation, the difference is not between normal legislation and amendment, but between legislation as a whole (including ordinary law and amendment law) and, on the other side, constitutional conventions. In yet a third way, the difference is not between ordinary and extraordinary representation, but between representation, on one side, and acts of direct democracy on the other. In other words, the people only speaks when it speaks for itself, be it through a plebiscite or through a referendum. On this account, extraordinary procedures leave ways open for the popular sovereign to manifest itself within the constitutional order. This is not a solution without any contestations, and one could criticise it for preserving the difference between constituent and constituted powers only to sacrifice the first to the constituted procedures that regulate it. Another criticism comes from those that cannot accept the idea of a legal procedure that allows for the exercise of unlimited power within the constituted order. Some would argue that this is not compatible with legal logic and/or with our understanding of constitutional democracy. It is true that not all systems have such a broad understanding of the amendment power, but in developing a typology from the ground up, the objective is exactly to reject one-size-fits-all answers to the relationship between constituent power and amendment procedure. Limits to the amendment power are now fairly common in modern constitutional systems—either explicit ones, made crystal clear in the constitutional text, or implied ones, constructed in the jurisprudence of the courts.

120  Thomaz Pereira However, this does not mean that the amendment power is not, and cannot, be conceptualised in certain systems as an unlimited one. In the framework that we have described so far, in which different constitutional systems can be based in different legitimating principles, and in which even the ones that are based on the idea of popular sovereignty can have different configurations, it is a mistake to assume that the fact that certain constitutional orders behave in a way, means that all must conform to the same legitimating narrative. There is yet another account in which the issue of limits to the constituent power is moot. According to this, in fact it is impossible for the popular sovereign not to have a procedure through which it could unlimitedly express itself ever again within the constitutional order. Irrespectively of any limitations to the formal amendment power, because the original constituent authority must have manifested itself through the performance of certain acts to demonstrate the reality of its popular legitimacy—if indeed it had it—then whatever rites were originally performed can always be performed again. In other words, if the Constitution was constituted through a procedure, this procedure can always be invoked to channel the constituent power once again. That government might choose to depart from constitutional law and follow the path of illegality is a fact of life, but this is not what was just described. According to this last narrative, in a world based on the principle of popular sovereignty, it is unconceivable for the constituent power to block the road in which it travelled from being taken ever again. II. CONCLUSION

Comparative constitutional law can benefit from a historically and politically contextualised understanding of the constituent power/constituted power divide. If we bear in mind that these concepts are the product of an attempt to found a constitutional order through a revolutionary popular movement, it is possible to comprehend the deep connection between popular sovereignty and this concept. As such, attempts to appropriate this distinction without consideration for the real possibility of manifestations of popular sovereignty are flawed from the start. Likewise, it is misguided to try to use these conceptual tools to understand the relationship between the constituting authority and the authority of the constituted order in systems that are not legitimated by an appeal to popular sovereignty. Once we realise this deep connection, a different insight can be gained on the theory of constitutional amendment powers. Because of the deep connection between the concepts of constituent and constituted powers and the nature and limits of amendment law, understanding the limits of a one-size-fits-all theory of the constituent power is required for developing a context specific typology of amendment law. In this way, by leaving aside the idea that all constitutional democracies are the same, for being constitutional and for being democracies, and understanding the fundamental differences between constitutional systems in the way that they differentiate the constitutional authority from legal authority and preserve this difference in its political culture through different institutional designs, it is possible to develop a working framework for constitutional amendment law.

6 Sieyès: The Spirit of Constitutional Democracy? LUISA FERNANDA GARCÍA LÓPEZ*

E

MMANUEL JOSEPH SIEYÈS laid the foundations of public law, giving true support to the concept of representation, not only through the election of representatives, but also through the political freedom inherent in the declaration of national sovereignty. Sieyès conceived of constituent power as having three elements: the Tiers État (Third Estate), the National Assembly and the National Constituent Assembly. Constituted power was represented by the legislative power, which could amend the Constitution, and the jury constitutionnaire (constitutional jury), whose role it was to protect the content of the Constitution. Although he was known to be hostile toward the notion of democracy, Sieyès became the ‘mole of the Revolution’ and representative government, which he ardently defended, became central to a type of democracy in which sovereignty and constitution prevailed. I. INTRODUCTION

The great contributions of French Abbot Emmanuel Joseph Sieyès to modern constitutionalism, and the changes that disrupted the status quo of 18th century France marked the beginning of the modern era. Sieyès1 was a fundamental actor in the French Revolution since the coup d’état of 17 June 1789, and he led the political

*  This article is a response to comments on the paper, ‘Constituting the amendment power: A framework for comparative amendment law’ by Thomaz Pereira for the BC-IACL workshop on Comparative Constitutional Amendment (Boston College Law School’s Clough Center for the Study of Constitutional Democracy, 15 May 2015). I want to thank Richard Albert for inviting me to participate and promote an academic debate. I would also like to thank Claudia Garcia, Gregory Dietrich, Alexandra Garcia, Esteban Cano, Sarah Miller, and Kevin Vu for their invaluable help and advice in translating this paper. 1  It is relevant to refer to the description made by Pasquale Pasquino in Sieyès et l'invention de la Constitution en France, a famous expert who has had all of his writings preserved in The French National Archives. Pasquino made an exhaustive study of all of his documents, and his works made a major contribution to modern constitutionalism. Most of the documents were unpublished and did not have a particular significance when they were written for two reasons. The documents were anti-monarchist and anti-Jacobin; as such, they were not made widely available at that time. Pasquino wrote a very eloquent description of his personality and his appearance that explained, to a certain extent, the reason why he was recognised as the inventor of the Revolution of 1789 and the promoter of the Constitution of 1791. At the same time, Pasquino did not share the extremist ideas about the Jacobins, and less still those about

122  Luisa Fernanda García López process since the Estates General (États Généraux), an assembly comprised of representatives from among the nobility, clergy and the Tiers État (Third Estate), was convoked. Sieyès would become one of the victors in the battle against aristocratic and clerical prerogatives, thereby eradicating royal prerogatives. Sieyès’ works served as the basis from which later discussions on the relationship between national sovereignty and representative government—the central tenets of the Republic—developed. In 1789, the concept of nation had a double connotation; it was understood as both a social body and a legal subject.2 Sieyès understood the nation as a social body in 18th century French society—the Tiers État. He did not ascribe to Rousseau’s abstract notion of ‘the people’, which included all of society. For Sieyès, the nation comprised those participating in the production of goods and services in private and public capacities. This included most segments of the population, including peasants, workers, artisans, and traders as well as those serving in liberal professions and governmental activities, such as public servants, teachers, and managers; however, this definition excluded the aristocracy, who were seen as a non-working, non-producing class. Thus, only working members of the population, or the Tiers État,3 were considered part of the nation. The nation, as a subject of law, referred to those who held sovereignty, namely, the Tiers État.4 The nation held the power, and this was written in the Constitution, which defined the structure of the state and conferred competence to its organs to manifest the national will. This represented a transformation in the conceptualisation of sovereignty for two reasons. First, it marked the beginning of institutionalised public law based on general will. Secondly, supreme will was declared in both original (constituent) and derived (constituted) power. This distinction between constituent and constituted power characterises the modern state.5 The power of the nation is given in the Constitution, which is written by the constituent power. This founding power in turn declares who, as the constituted power, can amend the Constitution. Transferring sovereignty to the nation was the first contribution of the Revolution to modern public law. The 18th century was a time of change aimed at dismantling traditional societal yokes and obligations. One movement sought to detach the ideals of individual and collective happiness from any religious concept and the Catholic faith.6 The spirit of the 18th century was also one of progress as manifested in the Revolution. Resistance

Robespierre, because he wanted to preserve the ideals of the Revolution. Despite being less cautious, Sieyès escaped from exile and the guillotine, perhaps because he was not very eloquent. He stated that his orations in the Assembly were very sporadic, and he was a scholar who wrote documents with vast amounts of legal and political content, which were confusing for the majority of his colleagues. His nickname was the ‘mole of the Revolution’. Pasquino included in his work some of Sieyès’ original summaries, preserved in The National Archives, and described with the reference 284 AP. 2 

P Pasquino, Sieyes et l’invention de la constitution en France (Odile Jacob, 1998) 61. Pasquino (n 2) 62. 4 ibid. 5  E Maulin, ‘L’invention des principes’ in S Caudal (ed), Les principes du droit (Economica, 2008) 23–47. 6  Taking up against Saint-Just who declared that with the Revolution, a new idea of happiness had arrived that spread throughout Europe. 3 

Sieyès: The Spirit of Constitutional Democracy 123 against the association between the king and the nation grew along with the realisation that the will of the king was not always in harmony with that of the nation. These factors contributed to the decline of the idea of divine connotation, which refers to the indivisible relationship between the king and the divine order. In addition, the social hierarchy was not to the advantage of the bourgeoisie, an emergent class below the clergy and nobility. In 1789, amidst this setting of fomenting tensions, Abbot Sieyès wrote a pamphlet entitled Qu’est-ce que c’est le Tiers État? (What is the Third Estate?). Its purpose was to put an end to hierarchical society based on orders or privileges. The pamphlet urged people to proclaim freedom from the ancient regime. Within months, the French Revolution broke out, and the coup d’état carried out by the Tiers État brought about new institutions, replacing the Estates General with a National Assembly, which later became the Constituent National Assembly, and giving rise to the notions of representation, constituent power, and constituted power.7 The definition of representation in Sieyès’ writings was constantly shifting throughout the revolutionary period (1789–1795); however, a strong resistance against authoritarianism, whether perpetrated by the monarchy or the people, consistently marked his definition of representation. Sieyès was also preoccupied by the concept of liberalism based on two fundamental notions: the Constitution and the separation of powers. For Sieyès, the Constitution was a pact. Sieyès’ claim that the nation holds constituent power differs from the mid-nineteenth century notion that the nation is bound by a contract, whereby several political actors set the limits of civil liberties and the exercise of power. Such a contract existed between the king and the people in a constitutional monarchy. For Sieyès, the nation refers to an active citizenship. In contrast to the realist conception of the Jacobins, Sieyès notion of sovereignty is based on a representative system—the republique (republic). It is also based on the citizenship–labour relationship. Because labour is the basis for belonging to the Tiers État, it becomes the social foundation of the nation, such that it is understood that non-engagement in labour is an exclusion principle in society and state. Popular sovereignty and elections are the principles legitimising authority. In 1792, Condorcet, a follower of Sieyès’ legacy, stated that sovereignty is the authority of the people; it can legitimately be dismissed and reinstated under other forms. Another type of sovereignty derives from the original and indissoluble contract between the monarchy and the people and can only be dissolved if the terms of this contract are violated by the monarchy. The notion of monarchy was being gradually relegated by the principles of constituent power and national sovereignty. Although some European countries, between 1789 and 1848, attempted to maintain monarchy in the form of constitutional monarchy and parliamentary systems, this was not the case for Sieyès’ France.8

7  J-J Chevallier and G Conac, Histoire Des Institutions et Des Régimes Politiques de La France de 1789 À Nos Jours (8th edn, Dalloz, 1981) 4, 11. 8  Pasquino (n 2) 130-45.

124  Luisa Fernanda García López Representative government was the creation of Sieyès.9 It refers to elected representatives’ power to define the roles and competencies of the governing bodies of the community and is manifested through the constituent power, namely, the Constitution, which stands above the law and restrains the legislative power. According to Sieyès, constitutive powers comprise a combination of legislative, executive and judicial powers, where the legislature stands above the other two. The protection of liberty by liberal constitutionalism can only be assured through the separation of powers and a constitution that guards individual freedoms. As mentioned above, Sieyès believed that effective separation of powers requires a hierarchical arrangement, where the legislature is sovereign while the executive and judicial branches are subordinated. The legislature generally does not act against the people; nevertheless, in absence of some counterweight to legislative power, constitutional control is justified as it can protect civil liberties when the stability of fundamental laws is not certain. In this sense, the jury constitutionnaire (constitutional jury) acts as a constitutional control of ordinary law. Although the legislature is a sovereign power, it cannot rule over the Constitution; therefore, constitutional control is exercised by a body in charge of preserving and periodically revising the Constitution. In other words, constitutional control is a mechanism for protecting the Constitution from any potential threats by the legislature. According to Sieyès’ theory, representation10 is a relative term. The concept of representation can lose validity if representatives do not manifest the people’s true will. In such cases, democracy and representative government can seem like an unachievable ideal.11 In fact, representative democracy involves representation through the electoral process as it pertains not only to the interests of individuals, but also to the general will, which would be carried out by representatives who battle to defend the rights of their constituents in the political field.12 This chapter focuses on Sieyès’ contributions, the origins of representative government from the Tiers État perspective and the concrete manifestation of representation in the Constitution of 1791, which laid the foundations for contemporary democracies (II). This chapter also explores the limits of representation from a ­constituent power and constituted power perspective, as well as the predominance of the law as a manifestation of the supreme will, which acts to preserve the Constitution through a mechanism of judicial review (III). II.  TOWARDS A REPRESENTATIVE DEMOCRACY

Representation is inherent to the French Revolution and stems from debates surrounding royal legitimacy (that is, the right of the monarchy to rule), and the 9  E-J Sieyès, Quelques Idées de Constitution Applicables À La Ville de Paris En Juillet 1789 (Baudouin, 1789). 10 M-H Caitucoli-Wirth, ‘La vertu des institutions : l’héritage méconnu de Sieyès et de Constant’ (2012) 1(16) Histoire@Politique 121. 11  In this sense, D Rousseau, ‘La démocratie continue. Espace public et juge constitutionnel’ (2011) 96 Le Débat 73. 12 In this sense, D Rousseau, ‘Les transformations du droit constitutionnel sous la Ve République’ (1998) 5-6 RDP 1780.

Sieyès: The Spirit of Constitutional Democracy 125 legitimacy of the Constituent National Assembly. Representation already existed in the pre-Revolution regime in the form the Estates General, where the three orders ruled: the clergy, the nobility, and the Tiers État, who voted as a unique body. The revolutionary activities of the Tiers État began on 4 August 1789, with the abolition of the orders’ societies, bodies, and privileges, and brought about a united nation while preserving to a certain extent the inheritance of the ancient regime. Representation under the Estates General system included delegates who were governed by an ‘imperative mandate’ and voted collectively, and who represented the clergy, the nobility and the Tiers État. The opening session of the Estates General on 5 May 1789, was convoked to decide on the allocation of seats and votes among the social classes represented in the Estates General, and to define whether members of the Tiers État would vote individually or collectively. The latter was a central issue because, on the one hand, if Tiers État voted as an order, they would represent a minority against the nobility and the clergy; on the other hand, if they members of the Tiers État voted individually, they would represent a majority against the Nobility and the Clergy. Louis XVI never clarified this issue, which created much discontent among the Tiers État, adding to the tension among the orders of the Estates General. Thus, representation in the Constitution of 1791 was focused on the Tiers État, and, more specifically, on the concept of active and productive citizenship, as well as the right to vote. Sieyès conceived of representation within a framework of work division, where individuals performed specific functions while working toward achieving common objectives. According to Abbot Sieyès, this collective homogeneity represents citizenship.13 A.  From the Tiers État to the National Constituent Assembly The spirit of the century was concretely manifested in the Tiers État. It was composed of members of the nobility and clergy, who represented a minority. The majority of the Tiers État comprised members of bourgeoisie. Generally, the rule of representation in the Tiers État was two bourgeois for one noble and one member of the clergy. No decision was made between 5 May and 9 June, but on 10 June, the clergy and nobility were asked to join the Tiers État, following Sieyès’ example as a clergyman and elected member of the Tiers État. Given their refusal to do so, on 17 June, the Tiers État met without the presence of the clergy and the nobility. During this meeting, the majority of members, including Sieyès, voted to form the National Assembly.14 This was effectively a coup d’état that further legitimised the Tiers État, which became the National Assembly after taking the oath of the Jeu de Paume (Tennis Court Oath) on 20 June 1789. Members of the Tiers État swore to create a constitution and remain united until the new Constitution of 1791 was proclaimed. The National Assembly took decisions that, even if they were undermined by the

13  14 

Pasquino (n 2) 40. Chevallier and Conac (n 7) 12, 26.

126  Luisa Fernanda García López exclusion of the clergy and nobility, were revolutionary and ultimately contributed to the series of events that ended with the final King’s capitulation on 14 July with the storming of the Bastille. Qu’est-ce qu’est le Tiers État?15,16 is one of the most relevant works on equality, privilege suppression and orders. Other seminal works that contributed to understanding of modern democracy and constitutionalism include Montesquieu’s Esprit des Lois, and Rousseau’s Du Contrat Social. The first committee of the Constitution created a short-lived government; it was reformulated in the first Legislative Assembly of 1791–1792. The Constitution of 1791 materialised Sieyès’ ideas on political representation. It outlined a rigid separation of powers, and excluded ‘passive’ citizens from participating in elections.17 Primary electors needed to be 21 years or older, have resided in the commune for at least one year, and they could not be domestic employees. Second electors needed to be at least 25 years old, and to swear to uphold freedom and equality, and to die in defense of freedom.18 The Constitution of 1791 also stated that the nation would replace the king, and that the political ideology of the Revolution would be founded on the Declaration of Human Rights and Citizens of 1789, which, in Article 3, reads: The principle of any sovereignty is based on the Nation. No corporation or estate, no individual can exert authority that does not emanate from sovereignty.

This is the basic principle of power that was born with the Revolution as a product of the Tiers État and its role in transferring power from the King to the People.19 Briefly, the principle holds that those who hold power must exert it on behalf of the people.20 The exercise of power is an emanation of the people’s sovereignty. However, given that those who exert the power only do so temporary, it is unacceptable that they exercise this power for self-serving purposes. In the same way, the Article 2, Title 2, of the Constitution of 1791 declares: The Nation, and only from the Nation that all the powers emanate from, cannot only exert the powers by delegation. The French Constitution is representative, and representatives are the legislative body and so the King.

This article is in line with Sieyès’ notion of political representation where the people will continue to have the right to declare their will, and would simply delegate the exercise of power.21 In summary, the idea of representation in the Constitution of

15  Sieyès (n 9) and Préliminaire de La Constitution Françoise : Reconnoissance et exposition raisonnée des droits de l’Homme & du Citoyen (Baudouin, impr de l’Assemblée nationale, 1789). 16  P Pasquino, ‘Bases de l’ordre ou série raisonnée de quelques idées fondamentales de l’etat social et politique (An III)’, and ‘La république contre la Ré-Totale ’ in P Pasquino, Sieyes et l’invention de la constitution en France (Odile Jacob, 1998). 17  Chevallier and Conac (n 7) 55. 18 ibid. 19  EG de Enterría, La lengua de los derechos: la formación del derecho público europeo tras la Revolución Francesa (1st edn, Alianza, 1994) 102, 108. 20  This refers to pouvoir commettant or tailor-made power, mentioned by Sieyès. 21  Enterría (n 19) 105.

Sieyès: The Spirit of Constitutional Democracy 127 1791 does not centre on the election of representatives but on the ideas of political freedom, and national sovereignty. This aspect would be modified in the Constitution of 1793. On 10 August 1792, elections were held to convoke a new convention, a constituent power with an executive council composed of six secretaries. These elections occurred in an atmosphere of violence, and they set off the Commune de Paris, an insurrectional force that fought against the nobility, the clergy, and the monarchy. The Commune de Paris stopped publication of all the royalist newspapers, allowing patriotic and left wing newspapers to be disseminated. The Commune de Paris generated two dynamics in the Constituent Assembly; on the one hand, there were those who felt apathy or fear following the elections, and on the other hand, there were growing tensions between girondins and montagnards. The Constitution of 1793 differed from that of 1791; it was more democratic and less liberal because sovereignty existed already among the people. Additionally, universal suffrage had been proclaimed in the 1793 Constitution, allowing for popular and direct voting. People would vote directly for their delegates, and they would be referred to as citizens and not as a nation. The Constitution of 179322 appeared as the first republican constitution, and was the first document of the republican tradition ‘it shone as a lighthouse and sparkled as a sacred document embodying essential dreams of democracy, unrealizable in 1793’.23 The Constitution of 1793 established a democratic regime that conferred a limited role to representatives. However, this regime never came into effect.24 B. From Citizenship to Constituent Power: The Foundation of Political Representation In 1789, when the Constitution was being drafted in France, there was a debate in the National Constituent Assembly about the system of governance that would be adopted in the Constitution of 1791. The speech given by Sieyès on 7 September 1789, was about the ‘government of the people’. This important document of the political theory discusses the notion of commercial society with a way of an appropriated government (that is, a government without provileges).25 According to Sieyès, the concept of political representation in modern society involved an association between the division of labor and representation. Modern society was in effect synonymous with commercial society, where consumption and production (that is, wealth) were substitutes for happiness, which was the fundamental value for previous

22  The Constitution of 1793 was more in line with Rousseau’s thoughts; Sieyès’ ideas would appear once again in the new constitutional debate of year III in the National Convention of 1795. 23  Chevallier and Conac (n 7) 61. 24 B Basdevant and J Gaudemet, Introduction Historique Au Droit (XIIIe–XXe Siècle), 3rd edn (LGDJ, 2010) 327, 378. 25  E Sieyes, ‘Representation et Elections’ (1791) 284 AP 4 (doss, 12), in this sense also see Pasquino (n 16) 36.

128  Luisa Fernanda García López generations and societies.26 Sieyès developed and justified his political and constitutional theory of a representative government in accordance with this line of reasoning. Pasquino’s27 affirmations based on his research between 1770 and 1785 relied on the principle of work division. In the Constitution Committee’s report on 2 October 1789, he commented on the new organisation in France, and stated that all faculties needed to be channeled into useful works to obtain a better product with less effort and cost. As a result, there was a separation of labour and responsibilities, which led to an increase in wealth and the perfection of human industry.28 Carrying out political responsibilities became a profession with broad objectives such as pursuing common interests and the improving the social state. Thus, political union can only be achieved through a representative constitution.29 When one assumes a f­unction through division of labour, one assumes the place of others in the implementation of a common objective; this is what Sieyès calls the ‘space of citizenship’.30 This definition of citizenship recognises the principle of government for the people, the identity of rulers and ruled people, and representatives and represented people.31 Sieyès refers in general terms to three topics in his writings: division of labour, politics as a profession, and a commercial society, all of which are considered with a frame of reference to political representation.32 Within a constitutionally representative system, citizens elect those who can deal with public issues requiring care and vigilance, and the government performs its function. Pure democracy as per Sieyès’ vision refers to a government that does not rely on representation, which is a crucial point that ideologically separates Sieyès from Rousseau. Representative government can also be defended from the perspectives of ‘tailor-made’33 power (pouvoir commettant), and law predominance, where the people are the sovereign and only communicate through their representatives. For Sieyès, Rousseau’s definition of direct democracy applied only to a primitive system wherein the people that lacked wealth were not represented; therefore, he devised this new definition of direct democracy. Tailor-made power was evident, given that the people could choose and delegate responsibilities to others who would exercise and advance their rights. Thus, to establish a representative democracy, a constitution is needed, one that includes a constituted power with exclusive authority to amend the Constitution.34 In interpreting Sieyès, Pasquino stated that tailor-made power was reflected in people’s influence over their representatives, and it was the activity of citizens who really had

26 

Sieyes (n 25); in this sense also see Pasquino (n 16) 52 Pasquino (n 16). 28  Pasquino (n 2) 38, 39. 29  This can be found in the comments of the report about the Committee of the Constitution relating to the new organisation of France, Versailles, Baudouin, 34, 35, quoted by ibid. 30  Pasquino (n 2) 40. 31 ibid. 32  ibid 43. 33 Sieyès referred to pouvoir commettant in which the most appropriate translation is tailor-made power as the one that was in charge of guaranteeing the content of the Constitution through voting. If tailor-made power referred to the people who legitimised and authorised through voting, then the elected representatives could write the Constitution, and create law. 34 Sieyès, Préliminaire de La Constitution Françoise (n 15) 36 quoted by Pasquino (n 2) 47. 27 

Sieyès: The Spirit of Constitutional Democracy 129 an active role.35 In contrast, constituent power and constituted power were always exerted by the representatives. The Constitution of 1791 distinguished between the representatives who were members of parliament and served the king, and those who wielded executive power,36 including agents and public delegates, and who did not have any type of representation.37 This difference between people’s representatives and these other agents was firmly established and codified in French Public Law and the Constitution of 1791.38 Indeed, the agents were not representatives but they carried out some of the duties that only the law conferred, whereas parliament was the institution that represented the nation.39 Sieyès’ described parliament as follows: ‘A common body represented by the same legislature is when the Nation’s will must express through its members, and each one of the members live in a regime of total freedom’.40 The Constitution must explicitly state when and how the constituted power is permitted to amend the Constitution. Conversely, it should be codified that ordinary representatives in charge of ordinary legislation cannot modify the Constitution. All rights belonged to the nation, and only the nation could confer special powers or set the limits for ordinary representatives in matters affecting the Constitution. The modern notion of representation and the concept of constituent power were the main and, arguably, most important products of the Revolution; as such, they must promote the rule of law. III.  TOWARDS A CONSTITUTIONAL DEMOCRACY

A representative elected government41 resulted in a society of free citizens operating under the rule of law.42 Additionally, the Assembly proclaimed a constituent power that defined written rules within the framework of exercising the power of the nation. The exercise of power by representatives was subordinate to the citizens’ authorisation during elections to comply with the law. The constituted powers were considered legitimate if they complied with the law.

35 The declaration of power is particularly mentioned in Préliminaire de la Constitution Française (n 15) quoted by Pasquino (n 2) 46. 36 M Troper, ‘La notion d’exécution dans les constitutions révolutionnaires’ 2012 Revue Française d’Histoire des Idées Politiques 299, 310. 37  Enterría (n 19) 105, 108. 38  R Carré De Malberg, La Loi, expression de La Volonté Générale: Étude Sur Le Concept de La Loi Dans La Constitution de 1875, 2 edn (Librairie du Recueil Sirey, 1984) 20, 21. 39  Enterría (n 19) 105, 108. 40 ibid. 41  M Gauchet, ‘L’héritage jacobin et le problème de la représentation’ (2001) Le Débat 32. Marcel Gauchet declared that the Revolution did not question the monarchy. The Royal power had a certain legitimacy, and the will of revolutionaries consisted of the ability to perpetuate the transmission of powers. The King should remain as a head of the executive power, and therefore, the executive power ought to be stronger as it had to serve the Nation, but a regulation had to exist. 42  Related to this topic, see M Raymonde, ‘“Démocratie représentative” ou “République démocratique”: de la querelle des mots (République) à la querelle des anciens et des modernes’ (2001) 325 Annales Historiques de la Révolution Française 1.

130  Luisa Fernanda García López It is important to emphasise that the debate about representation was no longer critical following the proclamation of the Constitution of 1791. The Constitution was finally defined through the election of citizens who did not have a mandate of unlimited or sovereign power. It was a transfer or delegation of power to govern, and the objective was public service, and representatives and those they represented were a homogeneous unit. Additionally, this representation was concentrated in the legislative function, operating within well-defined limits.43 Constituent power in fact represented a limit of powers, and differed from constituted power. Additionally, it is important to grasp how the jury constitutionnaire was a new institution that limited the exercise power, defended the Constitution, and was served as a constituted power. A.  The Sovereign People The division of powers is determinant to establish the difference between constituent power and constituted power, because a nation that delegates the constituent power is exposed to lose its freedom and to change the state without its consent. The Nation must keep its prerogatives to primary assemblies, but it can entrust its special delegates with the duty to present projects of improvement that both time and experience made it necessary.44

Considering the nation as an organised unit, the nation holds constituent power and could delegate to its representatives the duty of writing a constitution that outlines the legislative functions and constituted power. The exercise of the constituent power needed to be clearly defined, and to be delegated to primary assemblies and, in special cases, to the constituted power. Primary assemblies were at the base of the political pyramid, and they consisted of citizens who had the will and the capacity to participate.45 The constituted power, as a separate entity, was at the pyramid’s peak. Legislative power was identified as a supreme power; this was likely a residual effect of the legal centrism of 1789. Additionally, the legislative power was above the Head of State, the Senate, and the Chamber of Delegates, and it functions were corresponded with those of the three branches. Sieyès declared that executive power could only be exerted by representatives of the people; because he had inherited his position, the king could not exert this function.46 This difference between constituent power and constituted power was not very clear, neither in Rousseau’s theories on direct democracy, where the legislative power and the constituent power were not distinguished, nor in Hobbes’ theory of absolute representation, which does not conceive the people as holding the constituent power

43 

Pasquino (n 2) 52. Sieyès and R Máiz Suárez, ‘Bases Del Orden Social O Algunas Ideas Fundamentales Del Estado Social Y Político Año III’, Escritos y discursos de la Revolución (Centro de Estudios Constitucionales, 1974) 284. 45  A Tyrsenko, ‘L’ordre politique chez Sieyès en l’an III’ (2000) 319 Annales historiques de la Révolution française 27. 46 ibid. 44  EJ

Sieyès: The Spirit of Constitutional Democracy 131 because they represent the unity. According to Sieyès, this difference is clear because of the role of the Constitution. In fact, constituent power was articulated around the notion of tailor-made power; constituent power was only exerted by members of the nation, and it consisted of the right to vote and the right of the people to choose their representatives. The act of writing the Constitution belonged to the representatives who were elected. The reasons that justified this distinction were the same as those for founding the representative government. These reasons upheld the principle of the division of labour, and the exercise of politics as a profession.47 Thus, the nation and the representatives were inseparable, and the will of the nation could only be interpreted through its representatives who ‘would be as the Nation’s mouth’.48 Only the elected representatives could decide what form of political unity to be adopted. Finally, Sieyès’ constitutional doctrine recognised the separation of powers, extending Montesquieu’s ideas by emphasising the superiority of the legislative power relative to other powers. Pasquino’s ideas considered the hierarchy of constitutional norms over prevailing ordinary norm, and the practice of judicial review was established as a check on legislative power; thus, the idea of the jury constitutionnaire emerged. Legislative duties were delegated to elected representatives and not to a hereditary monarch. For Sieyès, this aspect constituted a lack of the monarch’s legitimacy in carrying out the law.49 B.  The Sovereign Constitution In the model of the Constitution of 1795,50 Sieyès proposed combining constituent power as it was conceived in 1789 as a base of popular sovereignty, and judicial review to guarantee the conformity of laws with the Constitution. The purpose of this combination was to protect the citizens’ rights in the eyes of the law. The jury constitutionnaire was the achievement of the constituent power and the constituted power for two reasons. First, it protected the sovereign’s will, which is the constituent power when it acted as a court of cassation of the constitutional order, and it denounced unconstitutional legislative acts. Secondly, the jury constitutionnaire improved the Constitution through proposals. The reform acts of the Constitution allowed a chance to amend the Constitution, and the jury constitutionnaire became a constituted power that could update the Constitution according to current needs

47 

Pasquino (n 2) 53,72. ibid 64. 49  The notion of royal veto consisted of giving the King the faculty to participate in writing laws. If the King did not approve a law, he would not penalise it, and this would give power to the executive against the legislative. Sieyès was opposed to this absolute or relative veto, and declared that this veto was a contradiction of the separation of powers as it was against the general will. Leaders did not need to participate in the creation of laws, and the executive did not need to be part of the creation of laws. The right to stop its enactment (through the royal veto) consisted of making the law; however, facing the contradiction with Mounier, they managed to vote the suspension veto and not the absolute veto. 50  The speeches from 2 and 18 Thermidor year III (20 July and 5 August 1795). 48 

132  Luisa Fernanda García López without permanently appealing to the constituent assemblies or conventions initiated with the Revolution.51 Fioravanti affirmed the difference between the constituent power at the head of the primary assemblies (that is, the sovereign people), and the amendment power of the jury constitutionnaire, which was considered a constituted power.52 This was one of Sieyès great innovations; he initially presented it as a representative and limitless government in the Constitution of 179153 against a representative power with marked and defined limits of representation, presented in the proposal for the Constitution of 1795, and the speech of year III.54 Regarding the judicial review, the jury constitutionnaire55 referred to a guardian medium of the Constitution. Sieyès referred to this medium as a mechanism of balance among powers that precedes judicial review through an independent jury comprising members of the three powers. An equivalent is the Constitutional Court system in the United States. Kelsen would implement a similar system in the 1920s in the Austrian Constitution. Proponents of popular sovereignty in the French system were extremely reluctant to implement a non-elected body to control the constitutionality of laws. Raymond Carré de Malberg56 questioned the legitimacy of the constitutional judge to control the constitutionality of laws and pointed to Sieyès and the role of the jury constitutionnaire. The constitutional judge acted as a guardian of the Constitution and the norms of sovereignty. Thus, Georges Vedel declared that the constitutional judge was like a ‘train controller’57 who would not stop the people from acting, but he would direct them along the right route with respect to obeying the law.58 He referred to a jury of 108 members that would be renewed annually at the same time as the legislative assembly (II). Each third was elected by the jury constitutionnaire out of 250 members transferring from a legislative chamber to another (III). The sessions of the jury constitutionnaire were not public (V), and they were used to announce violations or offenses to the Constitution perpetrated by members of voting assemblies, primary assemblies, and the Court of Cassation. Once these condemnations were presented by the Aged Council, the Five Hundred Council or by

51 M Fioravanti, ‘Sieyès et le jury constitutionnaire: perspectives historico-juridiques’ (2007) 349 Annales Historiques de la Révolution Française 87. 52 ibid. 53  See M Troper and L Jaume, ‘Le projet de Constitution européenne et la question du pouvoir constituant’ (2003) 13 Cités 101. 54  The control of the laws’ constitutionality proposed in year III was rejected in the Constitution of 1795, and it was adopted in the Constitution of 1799, year VIII, but with a content more political than jurisdictional. This did not work very well either. It was only in 1958 with the Vth Republic that the project of the Jury Constitutionnaire was materialised in the Constitutional Council as a neutral organ designed to preserve the content of the Constitution. 55  The literal translation would be constitutional jury; however, the expression in French jury constitutionnaire would be used throughout the article. The expression in French would be ‘aiguilleur’ translated as train controller or radar controller. 56  R Carré De Malberg, Contribution À La Théorie Générale de l’Etat, 2 edn (Sirey, 1962). 57  The expression in French would be ‘aiguilleur’ translated as train controller or radar controller. 58  G Vedel, ‘Le Conseil Constitutionnel, Gardien Du Droit Positif Ou Défenseur de La Transcendance Des Droits de L’homme’ (1988) 45 Pouvoirs: Revue Française d’Etudes Constitutionnelles et Politiques 149.

Sieyès: The Spirit of Constitutional Democracy 133 citizens as individuals (VI), the jury constitutionnaire would deliver the sentences (VII) and the acts declared unconstitutional by the jury were null and non-existent (VIII). If the acts declared unconstitutional were acts revealing a certain responsibility, the jury would denounce against competent courts (IX), and first of all the jury constitutionnaire would deal with subjects related to the constitutional act,59 and the improvement of human rights (X). Every ten years, and at the beginning of year VIII of the Republic, and the twelfth year of the Revolution, the jury constitutionnaire studied once again all the judgments found in the archives, and wrote the ‘proposal notebook’ to improve the constitutional act to communicate to the Aged Council and the Five Hundred Council, and to spread it en masse. This communication needed to be done three months before the annual session of the primary assemblies (XI), and after reading the proposal notebook, so that the assemblies could grant whether or not the power of deliberating to the Aged Council. If the majority of primary assemblies denied it, the notebook was considered as non-approved, and the proposals could only be reproduced before the following ten years would expire. If the majority of primary assemblies accepted it, the constituent power would be delegated to the Aged Council that could only deliberate about the proposals without any possible reform (XII). The sessions of the Aged Council were exclusively designated to exert the constituent power. Additionally, the sessions could not exceed more than twelve, and no more than two per decade (XIII). Each one of the constituent power sessions were solemnly registered and saved in the archives of the jury constitutionnaire.60 IV. CONCLUSION

The repercussions of the revolutionary period on the legal world resulted in the construction of a new social order. Three constitutions were drafted during the revolutionary period, and shared a common denominator in the form of sovereignty and the law cult. They existed under the premise that ‘[t]he law is the expression of the sovereign will’,61 and together formed the basis of contemporary law. The Revolutionaries, and particularly Sieyès, were in charge of balancing the primacy of law with two imperatives: the separation of fully autonomous and independent powers, and the pre-eminence of the legislative power over the executive and judicial powers. In the same way, the premise of legal equality implies the end of a society of privileges and the triumph of individualism, where the law considers cases on an individual and not collective basis. The architect of legal equality was Sieyès. The Constitution of 1791 declared that sovereignty belonged to the nation, and active tax-paying citizens had political and voting. The definition of sovereignty and voting

59 

Here Sieyès called it ‘constitutional act’ which is the Constitution. documents on the Jury constitutionnaire are XVII documents, and the writer has translated some of the documents in Pasquino Pasquale. 61 Se le reconoce un carácter superior a las tres Declaraciones de Derechos de 1789 (Art 6), 1793 (Art 4) y 1795 (Art 6) al igual que las tres constituciones del período revolucionario. 60  The

134  Luisa Fernanda García López rights were part of Sieyès’ work that conceived of the nation as an enterprise where only those who contributed to the collective welfare could participate. Citizens were represented in only one chamber because two chambers were thought to lead to local bias or aristocratic groups. The Constitution of 1795 tried to protect some elements of the Constitution of 1789, avoiding demonstrations of the popular government that had led to the Terror’s Age of 1793 by adopting some of Montesquieu’s postulates, evading Rousseau’s postulates, and renouncing to the postulates of the Constitution of 1791. The Constitution of 1795 brought a postulate where Sovereignty consisted in the Citizens’ Universality, without distinction between actives and passives. Citizen in the conception of 1795 referred to anyone who paid taxes. In the same way, laws were still the demonstration of general will expressed through the majority of citizens and representatives.62 The legislative power had two chambers for fear of the abuses that could emerge from both the legislative and the executive powers, and the legislative power had a professional body (a type of minister acting as the King’s advisers). In the Constitution of 1795, a strict separation of powers was predicted. In the three revolutionary Constitutions, the legalism was triumphant, and r­emembered that a perfect law was written by a group of citizens representing the State. It was about a true legal revolution63 where laws prevailed as a source of law typical of the Enlightenment, and a new social order was built.64 Sieyès laid the foundations of the contemporary democracy where the essential values of equality, representation, and participation come together. These values originated the Revolution of 1789 that created a representative government through vote. In spite of rejecting the notion of democracy, Sieyès contributed to the notion of representative democracy, as a determinant element to conclude the Revolution of 1793. Additionally, Sieyès ideas gave rise to the concept of constitutional democracy, and the jury constitutionnaire set the stage for another revolution. The judicial revolution65 and the reinterpretation of the separation of powers66 gave the judges a determinant power to reinterpret the notion of democracy. The constitutional democracy consisted of judges who were turned into fundamental actors of the social order, and where they became the main guides of the Constitution’s meaning. The jury constitutionnaire was a great record of the Judges’ Revolution67 that was only materialised in 20th century Europe, and that Maurice Hauriou had announced. The legislative power had the support of sovereignty that was almost irreconcilable with

62 

Declaración de los derechos y deberes del ciudadano de 1795, Art 6. legal revolution refers to a positive phenomenon that intervenes for the Constitution. It is a modification of the fundamental norm that moves away from the previous Constitution. In this sense, see J-Ph Derosier, ‘Qu’est-ce qu’une révolution juridique? Le point de vue de la théorie générale du droit’ (2015) 2(102) Revue Française de Droit Constitutionnel 391. 64  Basdevant and Gaudemet (n 24) 327, 378. 65  See in this sense BA Ackerman, La Constitución Viviente (Marcial Pons, 2011) 15, 39. 66  See in this sense LF García López, ‘El Juez Y El Precedente: Hacia Una Reinterpretación de La Separación de Poderes’ (2014) 128 Universitas 79. 67  See in this sense R Albert, ‘The Next Constitutional Revolution’ (2010) 88 U Det Mercy L Rev 707. 63  The

Sieyès: The Spirit of Constitutional Democracy 135 the National Constitution’s principles. Therefore, judicial review by judges was to guarantee the rights against the Parliament’s excesses that defended public interests and freedom, as long as the public opinion would support them.68 Having contributed to the conceptual development of representative democracy or constitutional democracy, political freedom and national sovereignty were Sieyès’ great legacy.

68 

See in this sense M Hauriou, Précis de Droit Constitutionnel, 2 edn (Sirey, 1929) 287.

136 

7 Revolutionary Reform in Venezuela Electoral Rules and Historical Narratives in the Creation of the 1999 Constitution JOSHUA BRAVER

B

OTH ‘PROPONENTS’ AND ‘critics’ of constitutional revolution share Carl Schmitt’s definition of it in which a people completely overthrow and rupture with the past in order to create a new constitution. Critics cite Venezuela as evidence that constitutional revolutions facilitate authoritarianism. I turn away from Carl Schmitt and towards Hannah Arendt to retell the Venezuelan story and recover an alternative vision of new beginnings called ‘revolutionary reform’. Like Schmitt, Arendt and revolutionary reformists celebrate momentous breaks, but they argue that they are still partial and best justified as necessary to realise the potential of old ideals. In Venezuela, Chávez’s achievement of first past the post electoral rules for the constituent assembly was the turning point that allowed him to eliminate the opposition through the purging of all other institutions. I argue that Chávez’s rule embodied a radical vision of revolution in which exclusion is necessary to completely rupture with all past Venezuelan History. Lacking its own vision of revolution, the Supreme Court and Congress struggled to contest the content of the electoral rule. I develop an alternative revolutionary reformist narrative of Venezuelan history that might have supported a more inclusive electoral rule, such as proportional representation. I. INTRODUCTION

What is a constitutional revolution, and how is it different than an amendment? Traditionally, a strict dichotomy is drawn between the two terms. In a revolution, a people completely overthrow the old constitution and found a new one. By contrast, government institutions carry out an amendment, and the change is partial.1 This strict division serves the agenda of both ‘radicals’ who favour constitutional

1 See, eg, J Colon-Rios, Weak Constitutionalism: Democratic Legitimacy and the Question of ­Constituent Power (Routledge, 2013) 91–96, 148–49; GJ Jacobsohn, Constitutional Identity (Harvard

138  Joshua Braver revolutions and ‘critics’ who condemn it. ‘Radicals’ invoke revolution against amendment to justify the purging of the opposition from all institutions and often the centralisation of power in their own hands.2 ‘Critics’ then sound the alarm and point to the action of radical actors as evidence that revolutionary constitutionmaking is reckless and dangerous.3 Although they disagree about the potential of constitutional revolution, both sides agree on its definition. Their consensus buries the possibility of a revolution that recognises the people’s freedom and right to institute their own foundations, but avoids the despotism and instability accompanying the invocation of a complete break with the past. The central theorist for the distinction between amendment and revolution is Carl Schmitt.4 The fixation on Schmitt has blinded both ‘critics’ and ‘radicals’ to a more complex understanding of individual cases of constitution-making and stymied the potential to elude the destructive polarisation between revolution and amendment. I turn to Hannah Arendt to develop the alternative of ‘revolutionary reform’. Unlike radicals, Hannah Arendt and revolutionary reformists seek not to end the story of the regime, but to begin a new chapter in its history. The break is momentous, but still partial and it is justified as necessary to realise the potential of old ideals. I use the idea of ‘unconventional adaptation’ to explain the institutional expression of revolutionary reform’s simultaneous drawing upon and breaking with the old regime.5 Even as many of the specific rules of the old institutions are bent, reinterpreted or broken, those institutions are still ‘unconventionally adapted’ to regulate and give shape to the creation of the new constitution. I develop the idea of ‘revolutionary reform’ through retelling the story of the creation of the 1999 Venezuelan Constitution. The existing literature either celebrates6 or condemns7 the process as an example of radical constitutional revolution. Theorists University Press, 2010) 44–45; Y Roznai, ‘Amendment Power, Constituent Power, and Popular Sovereignty’, in this volume; C Schmitt, Constitutional Theory (C Scmitt, Chr Thornhill and J Seitzer (ed)) (Duke University Press, 2008) 132, 150–53. 2 See R Prada Alcoreza, Horizontes de La Asamblea Constituyente (Ediciones Yachaywasi, 2006) 48–56; A Negri, Insurgencies: Constituent Power and the Modern State (University of Minnesota Press, 2009); R Viciano Pastor and R Martínez Dalmau, Cambio Político y Proceso Constituyente en Venezuela (Tirant Lo Blanch, 2001). 3  See A Arato, ‘Conventions, Constituent Assemblies, and Round Tables: Models, principles and elements of democratic constitution-making’, (2012) 1 Global Constitutionalism 173; D Landau, ‘Constitution-Making Gone Wrong’, (2012) 64 Ala L Rev 923; W Partlett, The Dangers of Constitution-Making, (2012) 38 Brook J Int’l L 193. 4 See, eg, Colon-Rios (n 1) 135; Jacobsohn (n 1) 44–5; A Arato, ‘Multi-Track Constitutionalism Beyond Carl Schmitt’, (2011) 18 Constellations 324; O Doyle, ‘Constraints of Constitutional Amendment Powers’, in this volume; Roznai (n 1). 5  Bruce Ackerman coined both the concept of ‘Revolutionary Reform’ and ‘Unconventional Adaptation’ to analyse the creation of the US Constitution. My account builds on his. See B Ackerman, We the People, Vol 2: Transformations (Harvard University Press, 2000) 11–13. However, Ackerman does not connect Arendt’s concept of constituent power to his later idea of revolutionary reform, and the idea is in conflict with his work on constitution-making outside the US; A Arato, ‘Forms of constitution making and theories of Democracy’, (1995) 17 Cardozo L Rev 191, 205–19. 6  See R Combellas, Poder Constituente (2000); Pastor and Dalmau (n 2); Colon-Rios (n 1). 7  See AR Brewer-Carías, Dismantling Democracy in Venezuela: The Chávez Authoritarian Experiment (Cambridge University Press, 2010); AR Brewer-Carías, Poder Constituyente Originario y Asamblea Nacional Constituyente; L Hernández Carmago, El Proceso Constituyente Venezolano (Editorial ­Académica Española 2008); Landau (n 3).

Revolutionary Reform in Venezuela 139 misleadingly focus on the Supreme Court’s January decision to, in the name of the people, approve of a referendum on whether to elect a constitutional assembly that circumvented the old constitution’s amendment rule. This was the turning point that empowered Chávez to pursue true revolution by purging the opposition from all levels of government. Arendt’s work redirects our attention from the decision to hold a referendum to the key choice of first past the post-election rules for the constituent assembly. This enabled Chávez to stack the assembly with supporters and to claim a mandate for a radical revolution. Here, the Court and Congress failed to fight back because they lacked theoretical tools to draw upon. They were trapped between radical revolutionaries’ call for a complete break and ‘critics’ who supported continuity through the traditional amendment procedure. I develop a revolutionary reformist vision of Venezuelan history that could have supported more inclusive electoral rules. The narrative supports a pluralistic process to create a constitution that solicits the input of a multiplicity of constituted powers, including the Court, Congress, as well as city and state governments. II.  CARL SCHMITT’S UNFORTUNATE VICTORY OVER HANNAH ARENDT IN THE ANALYSIS OF POPULAR CONSTITUTION-MAKING

Both Carl Schmitt and Hannah Arendt shared the fundamental insight that the people have the power to initiate extraordinary new beginnings in the forming of a new constitution. However, they disagreed over the powers and identity of the people. For Schmitt, this power was absolute and outside the current constitution, while Arendt argued that it was limited and unconventionally derived its authority from pre-existing institutions.8 Schmitt’s work has had far more influence in the analysis of acts of constitution-making. Articles have applied Schmitt, sometimes sympathetically, to understand constitution-making in the US,9 Egypt,10 Russia,11 Chile, Spain,12 Colombia, ­Venezuela,13 Kenya14 and the European Union.15 The highest constitutional courts of Colombia and Peru have cited him favourably,16 the German Constitutional

8 

Schmitt (n 1); H Arendt, On Revolution (1992). See, eg, P Kahn, Political Theology (Columbia University Press, 2012). 10  Landau (n 3). 11  Partlett (n 3). 12  R Cristi, ‘Metaphysics of Constituent Power: Schmitt and the Genesis of Chile’s 1980 Constitution’ (1999) 21 Cardozo L Rev 1749. 13  Colón-Ríos (n 1). 14 R Stacey, ‘Constituent power and Carl Schmitt’s theory of constitution in Kenya’s constitutionmaking process’ (2011) 9 Int’l J Const L 587. 15  J Muller, ‘Carl Schmitt and the Constitution of Europe’, (1999) 21 Cardozo L Rev 1777; L Vinx, ‘The incoherence of strong popular sovereignty’, (2013) 11 Int’l J Const L 101 (discussing the Lisbon decision); JHH Weiler, ‘Does Europe Need a Constitution? Demos, Telos, and the German Maastricht Decision’, (1995) 1 Eur L J 219. 16  See Colón-Ríos (n 1). 9 

140  Joshua Braver Court may have wrestled with him secretly,17 and he was a decisive influence on constitutional advisors to Augusto Pinochet18 and Hugo Chávez in the creation of their countries’ respective constitutions.19 Arendt’s understanding of the constituent power is far more complex and ambivalent than Schmitt’s. Schmitt gives total sovereign power to the people without hesitation or limitation. The people exist outside of the constituted order and have the power to create a totally new one by any process it sees fit.20 Hence, Schmitt concludes that the people may ignore the old constitution’s amendment rule and choose their own method of authorship and ratification.21 It is on this dramatic point, whether the old amendment rule limits the creation of a new constitution, that articles analysing actual instances of constitution-making most often cite Schmitt.22 It addresses the biggest initial obstacle for movements to create a new constitution as old amendment rules often require burdensome supermajorities and require the approval of legislators who have a vested interest in preserving the traditional bases of their power. While Arendt never directly addresses this legal issue, in stirring words, Carl Schmitt answers with an unequivocal yes to the people’s right to create a new constitution. Once the people realise they have the right to overthrow the old constitution and create a new one, how do they exercise it? What happens the morning after liberation? It is much easier to agree that the people have the right to make their own constitution than to answer the question of how this process should take place. How do you settle disagreement about the identity of the people, such as how to elect individuals to the constituent assembly, and what vote threshold is necessary in the assembly for approval of the draft constitution? If the people have the right to create a completely new regime, what is the role of the substantial minority that oppose this effort? Are they part of the people, and to what extent should they be included in the process of creating a new constitution? How do you prevent the

17  See Vinx (n 15) 114 (discussing Schmitt’s influence on the Court’s Lisbon decision); Weiler (n 15) 223 (arguing that the Maastricht decision’s citation to Herman Heller is a secret allusion to Schmitt). Schmitt loomed in the background during many of the early debates about the role of the Court. Schmitt’s disciple, Ernst-Wolfgang Böckenförde, became one of the most important justices of the Court and a prominent intellectual of public law. J-W Müller, A Dangerous Mind: Carl Schmitt in Post-War European Thought (Yale University Press, 2003) 63–75. 18  See Cristi (n 12) 1763–75. 19  Ricardo Combellas was the leading Venezuelan adviser in constitutional reform attempts for more than a decade and one of Chávez’s most important advisers. He was deeply influenced by Schmitt. Combellas (n 6); Lawyer, historian, and later a member of the constituent assembly Jorge Olivarría called attention to Combellas’ citation of Schmitt in an editorial justifying a sovereign constituent assembly. Olivarria’s paragraph was titled, ‘Tell me Whom You Quote … And I Will Tell You how You Think’; J Olivarría, ‘The Darkest Hour’ El Nacional (25 April 1999). In jail, Chávez read and adopted Antonio Negri’s theory of constituent power, which was deeply influenced by Schmitt. See H Chávez, Understanding the Venezuelan Revolution (Monthly Review Press, 2005) 41. Schmitt also influenced the new Venezuelan Supreme Tribunal. See N Figueroa, A Critique of Populist Jurisprudence: Courts, Democracy, and Constitutional Change in Colombia and Venezuela, 145 (unpublished Ph.D. dissertation, New School of Social Research, 2016) (on file with the author). 20  Schmitt (n 1) 150–53. 21  ibid 137. 22  See (n 4).

Revolutionary Reform in Venezuela 141 new constitution-making process from degenerating into mob rule or charismatic ­dictatorship? These are the questions that confront popular constitution-makers, and Schmitt offers few or no answers.23 Like Schmitt, Arendt also wants to empower the people to realise their own freedom to initiate extraordinary new beginnings, but she places many principled limitations on that power. These limitations are the product of an ‘unraveling’ of the constituent power’s dilemmas, perplexities, and paradoxes.24 By addressing these issues, Arendt helps us arrive at a constituent power that realises the people’s political freedom without sacrificing at its alter both stability and the rule of law. III.  HANNAH ARENDT’S REVOLUTIONARY REFORM

A.  The Dangerous Freedom of the People For Arendt, constitution-making is the highest exercise of political freedom. In a true revolution, the people break free from the constraints of history or tradition to determine their own political destiny. It is the novelty of revolution that fascinates Arendt, that it cannot be predicted or traced back to antecedent causes, but rather exhibits the human capacity and power to generate its own terms of political existence. It radically breaks with the past to inaugurate a new future. The people deliberate and act together to begin ‘anew’, to begin ‘an entirely new story, a story never known or told before’.25 While Arendt celebrates the people’s freedom, she also fears it. Unlike Schmitt, she reminds the readers that after a revolution, ‘constitutional government, if it came into existence at all, had a tendency to be swept away by the revolutionary movement which had brought it into power. Not constitution, the end product and also the end of revolutions, but revolutionary dictatorships, designed to drive on and intensify the revolutionary movement, have thus far been the more familiar outcome of modern revolution—unless the revolution was defeated and succeeded by some kind of restoration’.26 All too often, triumph gives way to disappointment, and the revolution eats its own children. Arendt explains that this is the direct consequence of positing the people as the ultimate authority and then defining them as above and unbound by any law. Revolutions create a crisis in authority. Before, a constitution and its institutions were sanctioned and hallowed by tradition, custom, and religion. Secularisation and revolution revealed these justifications ‘as facile expedients and subterfuges’ that ‘had lost its anchorage, its beginning and principle’. To restore legitimacy to the constitutional order, the people took up the place of God/King as the original source of 23  Indeed, Schmitt is often cited to understand the establishment of dictatorships. See, eg, Arato (n 3) 187–88; Cristi (n 12). 24  A Kalyvas, Democracy and the Politics of the Extraordinary: Max Weber, Carl Schmitt, and Hannah Arendt (Cambridge University Press, 2009) 195. I draw heavily from Kalyvas’ comparison between Arendt and Schmitt, but I depart with his benign reading of the latter. 25  Arendt (n 8) 28. 26  ibid 158–59.

142  Joshua Braver authority.27 Since the people are the source of legality for the entire system, it itself cannot be produced or controlled by law. Although Arendt locates this definition in the thought of Emanuel Sieyes and Robespierre, Carl Schmitt also shares and even further radicalised this definition of the people.28 Furthermore, not only were the people outside the pre-existing political order, but superior to it as well. To justify the overthrowing of the old institutions, they must not be bound by them. Their authority must be greater than that of ordinary law so that they may violate it to create an entirely new legal order. Arendt quotes Sieyes that ‘it would be ridiculous to assume that the nation is bound by the formalities or by the constitution to which it has subjected its mandatories’.29 As Sieyes stated, the nation’s ‘will is always legal. It is the law itself’.30 For Arendt, this understanding of the people posed two great dangers. First, new institutions are unstable because their creators lack legal authorisation for their actions. They are ‘themselves unconstitutional’.31 This is the problem of the authorisation of new beginnings. Law provides a stable and objective source for locating authority. However, once we locate authority with a people that is outside law and all institutions, the identity of the people is continuously contestable. Each time a group claims to represent the people, others may object that they represent the people. Each time an institution tries to make a law or a constitution, groups will claim a right to resist or even seize the power of the state because they act on behalf of or are the people. Arendt argued that no institution is able to find a solid foundation on the ‘will of a multitude’ because it is ‘ever-changing by definition, and that a structure built on it as its foundation is built on quicksand’. Thus, revolutions often degenerate into chaos or mob rule. Exhaustion with this chaos in turn leads to the rise of a dictator who promises to restore order. He too will claim to embody the people. The second danger is the potential tyranny of an unlimited and arbitrary power. The rule of law limits the power of the state. The government can only act pursuant to and in the execution of a law. Inherent to the rule of law are protections against arbitrary executive action, such as detention and unfair trials. However, in order to authorise normal law, the people must be above and unconstrained by it. Hence, the people are an unlimited power. They may jail, imprison and execute at will. Whoever then can sustain a claim to act on behalf of the people, whether it is an assembly or an individual, may break the law and consolidate all power in his person. Arendt notes that ‘Napoleon Bonaparte was only the first in a long series of

27  ibid 162; H Arendt, ‘What is Authority?’ in H Arendt, Between Past and Future (Penguin Classics, 2006) 91. 28 Some speculate that Arendt is using Sieyes as a placeholder for Schmitt. See, eg, Kalyvas (n 24) 194–95. 29  Arendt (n 8)161. 30 E Sieyes, ‘What is the Third Estate?’ in M Sonenscher (ed), Political Writings (Hackett Publishing Company, 2003) 136. For Sieyes thought on amendment, Th Pereira, ‘Constituting the Amendment Power’, in this volume. 31  Arendt (n 8) 184.

Revolutionary Reform in Venezuela 143 national statesman, who to the applause of a whole nation, could declare: “I am the pouvoir constituent”’.32 B.  Renewal and Revolutionary Reform Arendt strives to preserve the sacred freedom of the people to determine their own existence, but stave off the degeneration into lawless mob rule or dictatorship. Rather than a total break with the past, Arendt reconceives of revolution as a radical renewal of the foundations of a polity. Building on the term Bruce Ackerman coined, I explain this idea as ‘revolutionary reform’.33 For Arendt, revolutionary ruptures should be partial and not absolute. The power of the people ‘to begin something new’ ‘does not mean that it is ever permitted to start abovo, to create ex nihilo’.34 The idea of a total break reflects the Judeo-­ Christian idea of a god who created the world from nothingness that marks the beginning of a completely new time. It is impossible for human beings to exercise these demi-urgic powers and the attempt to totally uproot the body politic requires high levels of coercion, violence and destruction.35 Humans act within the boundaries of a pre-existing world. They are not determined by it, but they act against it as a background and may never completely break free from it. Revolutions are still ruptures and ‘miracles’ but on a human scale.36 Rather than total breaks, revolutions should strive to radically renew their foundations, ‘to see constitutions and foundations in terms of restoration and reestablishment’. They should ‘bind themselves back to a beginning’. The legitimacy of the current regime is in crisis because it has strayed from its foundations. The revolution should harken back to its past to justify the break with the present. They seek not to end the story of the regime, but to begin a new chapter in its history. The break is still radical, but it preserves continuity as well because it is necessary for the greater realisation of past ideals. A revolutionary is ‘a founder and preserver at the same time’.37 What does this look like institutionally? Arendt provides a basic framework. Rather than acting outside all institutions, the people ‘derive its general authority … entirely from subordinate institutions’. She illustrates this phenomenon by contrasting the French and American revolutions. French revolutionaries could not derive their authority from their old feudal institutions because their premises denied the existence of a people. Thus, ‘the rupture between king and parliament indeed threw the whole French nation into a “state of nature”; it dissolved automatically the political structure of the country as well as the bonds among its inhabitants, which had rested not on mutual promise but on the various privileges accorded to each order and estate of society’.38 32 ibid. 33 

See (n 5). H Arendt, ‘Lying in Politics’ in H Arendt, Crises of the Republic (Harvest Books, 1972) 5. 35  Arendt (n 8) 112. 36  See B Ackerman, ‘Revolution on a Human Scale’ (1999) 108 Yale L J 2279. 37  Arendt (n 8) 203. 38  ibid 180. 34 

144  Joshua Braver The ‘great good fortune of the American Revolution was that the people of the colonies, prior to their conflict with England, were organised in self-governing bodies, that the revolution … did not throw them into a state of nature’. Hence, the people existed through and were organised by pre-existing institutions, and they acted through them to create new constitutions. The townships elected the framers of state constitutions, and the framers of the American Constitution were tied to individual states. Without the American states, ‘the founders would have met immediately the perplexities of their French colleagues; they would have lost their pouvoir constituant’.39 Old institutions were ‘used as a foothold to secure the new beginning’.40 Arendt’s genius is to break from Schmitt’s false glorification and deification of the people as outside any institution. I supplement her account with an institutional framework to express this new conception of constituent power, or the people’s power to create a new constitution. C.  Extrapolating from Arendt: Unconventional Adaptation Although Arendt’s basic framework is correct, it is also incomplete. If not interpreted carefully, Arendt might be misread as supporting John Calhoun’s state compact theory in which the US federal government draws its authority from the states rather from the people. Indeed, the states greatly influenced and organised the ratification of the American Constitution. But to ensure that authority derived from the people, the ratification process also violated state constitutions and the Articles of Confederation. Furthermore, the drafting in the Philadelphia convention disregarded the instructions of state delegates in the Continental Congress by choosing to create a new constitution rather than merely reforming the Articles.41 How should we understand this paradoxical drawing upon and violation of state authority? I borrow and build on Bruce Ackerman’s idea of ‘unconventional adaptation’ of old institutions.42 It captures Arendt’s ideal of a new beginning as a radical, but incomplete break with the present. There is continuity because revolutions use old institutions, but their rules are bent, reinterpreted or even broken to create a new constitution. The action is illegal, but not lawless. The constitution-making process violates specific rules, but does not degenerate into a contest of strength or force. It is an extra-legal process justified and constrained by the renewal of principles of the old regime. Old institutions are unconventionally adapted to frame a collaborative process to create a new constitution.

39 

ibid 165–66. Kalyvas (n 24) 227. 41  cf Ackerman (n 5) 33–68. 42  ibid. For a more in-depth examination of the role of unconventional adaptation in Venezuela, see Joshua Braver, ‘Hannah Arendt in Venezuela: The Supreme Court Battles Hugo Chávez over the Creation of the 1999 Constitution’ (1 July 2016) 14(3) International Journal of Constitutional Law 555. 40 

Revolutionary Reform in Venezuela 145 IV.  HUGO CHÁVEZ’S RADICAL AND ORIGINAL CONSTITUENT POWER

In December of 1998, riding a wave of popular discontent with the political system, Hugo Chávez was elected president of Venezuela with 56.20 per cent of the vote, the most decisive presidential victory since 1947, the first year of electoral democracy in 20th century Venezuela.43 The central promise of his campaign was to hold elections for a constituent assembly to write a new constitution. Venezuelans blamed the 1961 Constitution for the country’s economic woes. The Constitution was widely believed to have divided political power between elitist parties to the exclusion and detriment of the common citizen.44 Chávez believed that only a constituent assembly could create the radical break necessary for a more ‘participatory and social democracy’. In his inauguration ceremony Chávez swore, ‘before God, before the homeland and before the people, upon this moribund constitution, that I will complete and propel the democratic transformations necessary for the Republic, a new Magna Carta adequate to the new times’.45 The problem was that the 1961 Constitution’s amendment method did not permit an assembly. The Constitution specified two exclusive methods of formal constitutional change; either amendment or ‘general reform’. The latter referred to the radical change or perhaps even the creation of a new constitution through a multi-step process that included the supermajority vote of a joint sitting of the two houses of the legislature and subsequent approval through a popular referendum.46 Furthermore, although Chávez won a landslide in the December 1998 election, the legislative elections occurred a month before, and Chávez lacked the votes in Congress necessary for general reform.47 He claimed a right however to call the referendum on whether to have a constituent assembly because as stated in Article 4 of the 1961 Constitution, ‘sovereignty resides in the people’ and therefore they have the right to create a new constitution.48 But Chávez’s argument was even more radical. He claimed that the people not only had the right to elect a constituent assembly to write a new constitution, but also that the assembly would have the right to rupture completely with the past by overthrowing the existing institutions and directly rule the country. In Latin America, this power is called ‘the original constituent power’ and is exercised after the near complete destruction of the old order through a civil war or violent uprising. Chávez and his constitutional advisers maintained that even though there was no violent

43  M Coppedge, ‘Venezuela: Popular Sovereignty vs. Liberal Democracy’ in JI Domingez and M Shifter (ed) Constructing Democratic Governance in Latin America (Johns Hopkins University Press, 2003) 167. 44  See D Hellinger, ‘Political Overview: The Breakdown of Puntofijismo and the Rise of Chavismo’ in S Ellner and D Hellinger (eds) Venezuela Politics in the Chávez Era (Lynne Rienner Publishers, 2004) 31. 45 H Chávez, Inauguration Speech, in la construcción del socialismo del siglo XXI : discursos del comandante supremo ante la Asamblea Nacional, 1999–2001 (Vol 1, 2013). 46  See Venez Const, 1961, Arts 245, 246. 47  See Coppedge (n 43) 178. 48  Chávez (n 19) 32; H Chávez, ‘Presentacion de Hugo Chávez Ante La Academia de Ciencias Políticas y Sociales’ in AR Brewer-Carías, Asamblea Constituyente y Proceso Constituyente 1999. Colección Tratado de Derecho Constitucional, Tomo VI (Fundación Editorial Juridica Venezolana, 2014) 76–77.

146  Joshua Braver revolution, there was or would be a peaceful one. Chávez argued that the revolution, the ‘end … of the paradigm of liberal democracy’ and the ‘radical change, total, of a model of society, of politics, economics, etc’. required an original constituent power that could purge Congress and other power bases of the old guard. And if Chávez wasn’t allowed to give this ‘wild’ force of the constituent power a ‘peaceful channel, Venezuela would likely arrive ‘at new violent facts’ and ‘events much more tragic than what has passed’.49 On 2 February 1999, the day of his inauguration, and as his first official act of government, Chávez issue a decree calling for a referendum with two separate questions on the ballot: 1. Whether to hold elections for a constituent assembly to draft a new constitution, and 2. Whether to authorise Chávez to unilaterally decide the method of election for the assembly. If the voters said yes, the process would culminate in another referendum in which they would vote on whether or not to approve the Constitution drafted by the constituent assembly. The referendum decree was challenged in the Supreme Court of Venezuela, and a struggle ensued between Chávez and the Court that spanned ten cases and nearly nine months.50 Chávez ultimately prevailed. The Court authorised the referendum on whether to have elections to a constituent assembly. However, they ruled that the second question, as formulated, was unconstitutional. The second question should offer voters the opportunity to approve or disapprove of a specific electoral rule for the constituent assembly.51 Chávez complied, and put on the referendum ballot a first past the post method of aggregating votes for future elections to the constituent assembly, and the Court accepted the electoral rule. The referendum was held on 25 April 1999, and voters approved both the proposition to have a constituent assembly and Chávez’s chosen electoral rule. In the elections to the Constituent Assembly, Chávez’s chosen electoral rule converted his 65.8 per cent of the vote into 93.1 per cent of the seats in the constituent assembly.52 Despite repeated Supreme Court rulings to the contrary53 and with Chávez’s support, the Constituent Assembly declared itself sovereign. As testament to the Assembly’s sovereignty and to prove that his powers derived from the Assembly rather than the ‘moribund constitution’ of 1961, Chávez asked the assembly to ratify him his as president.54 On 9 August, the Constituent Assembly voted unanimously with three abstentions to select Hugo Chávez as the president,55 and he was re-sworn in on 11 August.56 Logically, if they had the power to ratify Chávez’s presidency, they too possessed the

49 

Chávez (n 48) at 76–7. See Combellas (n 6) 32. See Brewer-Carías, Dismantling, (n 7) 52–9; Landau (n 3) 939–49. 51 Sentencia del 19 de enero de 1999, Sala Político-Administrativa, MP Humberto J La Roche ­(approving the referendum); Sentencia del 18 de marzo de 1999, Sala Político-Administrative, MP Hermes Harting (reiterating the approval of the referendum, but requiring Chávez to specify the electoral rule). 52  Coppedge (n 43) 187; R Segura and A M Bejarano, ‘¡Ni unaasambleamás sin nosotros! Exclusion, Inclusion, and the Politics of Constitution-Making in the Andes’ (2004) 11 Constellations 217, 230. 53  See, eg, Sentencia 13 de abril de 1999, Sala Político-Administrativea, MP Hermes Harting; Sentencia del 21 de Julio de 1999, Sala Polítco-Administrativa. MP Hildegard Rondón de Sanso. 54  H Chávez, Opening Speech to the Assembly (n 45) 35 (repeatedly declaring assembly to be ‘most sovereign’ at his speech at the opening of the assembly). 55  Gaceta Constituyente (Diario de Debates), Agosto–Septiembre 1999, No 5, 9 August 1999. 56  Gaceta Constituyente (Diario de Debates), Agosto–Septiembre 1999, No 6, 11 August 1999. 50 

Revolutionary Reform in Venezuela 147 power to fire the holders and limits the powers of other offices. The next day, on the basis of its sovereign power, the Constituent Assembly declared a state of emergency ‘declaring the reorganization of all the organs of public power’ and authorised itself to ‘take, execute and order the measures dealing with the competence of the public powers of the state … that are necessary and indispensable to overcome the situation of emergency confronting the nation’. This was the beginning of a process which developed in August and reached its climax after ratification of the Constitution in December in which the Assembly sweeping and effectively purged, stacked, or asserted effective control over the Congress, the Judiciary, congress, state legislative assemblies, mayors, municipal councils and labour unions.57 The opposition has not only been excluded from the process of drafting the Constitution, but also from all other possible positions of power at every level of government. The Constituent Assembly, stacked with Chávez allies, became a despotic and all-powerful actor that illegitimated usurped the constituted powers. For almost a year, starting from the mid-August 1999 decrees until the first round of elections under the new Constitution on 30 July 2000, the Constituent Assembly and its appointed allies ruled and dominated the country. Similar to the French, Jacobin Convention of 1793, the Constituent Assembly justified this claim on the basis that it was sovereign since the people, as embodied in the March referendum and elections to the constituent assembly, had risen up against and overthrown the old regime.58 This is Hannah Arendt’s greatest fear: the claim to sovereign authority based upon embodiment of the people. As the Supreme Court repeatedly pointed out, sovereignty lies with the people, and not with the assembly. Indeed, no one institution or individual can embody the people, but only represent it. Hence, any one body’s claim to sovereignty is always a usurpation because of the very logic of representation itself.59 The Assembly was not identical with people. Rather, just like Congress and state governments, the Assembly represented the people, and for that reason could only have limited and delegated powers. In the March referendum and elections, as stated on the ballot, the people delegated to the assembly the sole task of drafting a constitution. The Constituent Assembly had exceeded this mandate. Furthermore, the process created the conditions for near complete executive dominance over government. Elections are meaningful, frequent, and contested events in Venezuela with relatively accurate tabulation of votes.60 What are lacking are any meaningful checks on presidential power during the term of office. To be clear, this is not the fault of the Constitution itself. While it did strengthen the presidency by permitting one consecutive re-election and extending the term from four to six years, it also provided for a presidential recall, an independent judiciary, and the protection

57 Brewer-Carias,

Dismantling (n 7) 57–60; Pastor and Dalmau (n 2) 150–1; Landau (n 3) 947. Luis Milquena, president of the constituent assembly’s opening speech. Gaceta Constituyente ­(Diario de Debates), Agosto–Septiembre 1999, No 5, 9 August 1999. 59  See B Garsten, ‘Representative government and popular sovereignty’ in I Shapiro and SC Stokes (eds), Political Representation (Cambridge University Press, 2010) 90. 60 See, eg, Carter Center, Final Report on Venezuela’s April 2013 Elections (2013); Human Rights Watch, Countries at the Cross Roads: Venezuela (2011). 58 

148  Joshua Braver of a generous list of human rights.61 However, assembly’s liquidation of the opposition gave Chávez near complete dominance over the process of implementation of and transition to the new constitution. This allowed him to manipulate the system to undermine the powers and independence of Congress and the judiciary.62 There are also serious, though heavily contested, accusations that Chávez abused his power to seriously hinder, though not make it impossible, for the opposition to compete in elections through various forms of blacklisting, harassment, and ‘legal’ persecution for speech and broadcasting violations, illegitimate use of state resources to support incumbents in elections and occasional arrest or exile.63 Are ‘critics’ of the constituent power right that by authorising the initial referendum, the Court enabled Chávez to establish a semi-authoritarian regime in ­Venezuela? Should it have upheld the Supremacy of the 1961 Constitution and struck down the central plank of Chávez’s electoral campaign to call a constituent assembly to write a new constitution? V.  THE TURNING POINT: THE ELECTORAL RULES FOR THE CONSTITUENT ASSEMBLY

For both ‘radical proponents’ and ‘critics’ of revolutionary constitution-making, the Court’s January approval of the referendum was the turning point that enabled Chávez’s victory. While ‘radicals’ celebrate and anthologise the Supreme Court’s approval of the referendum, ‘critics’ maintain that it was an egregious error that led to its own destruction. For Nicholas Figueroa, in the first referendum decision, the Court ‘shot itself in the foot’ and ‘provided Chávez’s quest for the consolidation of a new political hegemony with a very suitable doctrinal basis’.64 For Landau ‘once [the Court] abandoned the formal limits in the constitutional text, they were left

61  Landau (n 3) 941, fn 66; But see A-B Carias, Constitutional Law in Venezuela (Kluwer Law International, 2012). 62  Coppedge (n 43) 17; Landau (n 3) 940–41, 949. 63 Human Rights Watch, ‘A Decade Under Chávez’ (2008); Letter, More than 100 Latin American Experts Question Human Rights Watch’s Venezuela Report (2008), accessed 1 August 2016. See also S Levitsky and J Loxton, ‘Populism and Competitive Authoritarianism in the Andes’ (2013) 20 Democratization 107, 108–25; DJ Myers, ‘Venezuela: ­Delegative Democracy or Electoral Autocracy’, in J Domínguez and M Shifter, Constructing Democratic Governance in Latin America (Johns Hopkins University Press, 2008) 318. 64 Figueroa (n 19) 145. I am in debt to Figuero’s work. Figeuroa fits uneasily in the critics’ camp because he expresses some openess to developing an alternative definition of the constituent power. Nonetheless, like critics, his focus is on its dangers, not its potential. In his case studies, the openness manifests itself most clearly in the chapter on Colombia in which he supports the legally questionable process up until the Supreme Court strikes down the content limitations on the assembly as a violation of the constituent power. At this point, he objects. Revolutionary reform would be suspicious of such restrictions, and I would argue that they belong to the logic of the Andrew Arato’s post-sovereign model, a model that is hostile to illegality and the ideas of the people and revolution. Indeed, Andrew Arato argues that the model is the best lens through which to understand the success of the Colombian process, and Figueroa characterises it as an ‘eclectic mixture of the two paradigms’, of the post-sovereign model and Schmittian original and lawless constituent power, commending the former’s influence and condemning the latter’s. Andrew Arato, Adventures of the Constituent Power (forthcoming, 2017) (on file with the author); Figueroa (n 19) 118.

Revolutionary Reform in Venezuela 149 with a vague doctrine that was difficult to enforce’.65 And writing shortly after the referendum decision, Allan Brewer-Carías subtitled his article, ‘How the Guardian of the Constitution opened the road to its own violation and to its own extinction’.66 For ‘critics’, once the Court opened up the pandora’s box of revolution, it lacked any means to control Chávez. In contrast, I argue that the turning point was not the Court’s decision to have a referendum, but it and Congress’ insufficiently firm intervention concerning the construction of the electoral rules for the constituent assembly. They should have fought against Chávez’s choice of a first past the post electoral rule that allowed Chávez to stack the Assembly with his allies. The problem was not the invocation of revolution itself, but the particular radical vision of revolution that was enacted. A.  The Difficulty of Fighting Against the Referendum The Court and Congress could not and should not have stopped the referendum. The legitimacy of the old constitution and even the judiciary was at a breaking point. By contrast, Hugo Chávez was at the height of his popularity, even earning the vote of the bourgeoisie, and had made the creation of a new constitution the central promise of his presidential campaign. His election and the creation of the Constitution would be the result of the ‘channeling’ of the people through peaceful means, but he repeatedly warned that if the constituent power were not ‘given space, it would occupy it’. Without a new channel Venezuela would likely arrive ‘at new violent facts’ and ‘events much more tragic than what has passed’.67 If the Court had ruled against Chávez, he would have purged the Court and incurred little political cost. Even if the Court could have stopped Chávez, it’s not clear that this would have been normatively desirable. Radical constitutional reform had been on the agenda long before Chávez based his movement and his campaign on it. There was near universal consensus that neoliberalism, centralisation, corruption and economic crises had shattered the old system of government. Congress had twice made significant efforts to initiate the Congress-centered ‘general reform’ amendment procedure used for greater changes in the Constitution. These efforts at reform failed miserably because Congress lacked the will and legitimacy to overcome even the most tepid opposition.68 After Chávez’s electoral victory, an elected constitutional assembly could accomplish what Congress could not: rally support for badly needed constitutional reform to re-democratise and broaden the base of the political system. The Court and Congress’ best path forward was not to deny the people their rightful power, but to

65 

Landau (n 3) 949. See AR Brewer-Caras, ‘La configuracion judicial dcl proceso constituyente o de como cl guardian de la Constitucionabrio cl camino para su violacion y para su propia extincion’ (1999) Revista de Derecho Publico 77–80, Editorial Juridica Vcnezolana, Caracas. 67  Chávez, ‘Presentacion’ (n 48) 76–7. 68  M Kornblith, ‘Legitimacy and the Reform Agenda in Venezuela’ in D Canache and M R Kulisheck (ed), Reinventing Legitimacy: Democracy and Political Change in Venezuela (Praeger, 1998) 3. 66 

150  Joshua Braver participate in the creation of the people themselves, especially through the construction of the electoral rule for the constituent assembly. B.  Ex-Ante Control of the Constituent Assembly through Electoral Rules The Court and Congress faced the classic problem of what Jon Elster calls, ‘constitutional boot-strapping, the process by which a constituent assembly severs its ties with the authorities that have called it into being and arrogates some or all of the powers to itself’. Once the constituent assembly is convoked, it becomes very difficult for the old regime to control or limit it because ‘almost, by definition, the old regime is part of the problem that a constituent assembly has to solve. But if the old regime is flawed, why should the assembly respect its decisions?’69 Since an assembly cannot be controlled once it’s been convoked, the solution is to choose a selection method that achieves the right members. If courts and legislatures cannot hold the assembly accountable ex post, they should focus on the ex ante method of selection. Recall that Chávez chose a ‘first past the post’ election rule that converted his 65.8 per cent of the votes cast into 94 per cent of the seats in the constituent assembly. This overwhelming majority facilitated his ability to purge Congress, the Court and seize control over all other institutions. By contrast, if the assembly had been more proportional, opponents within the constituent Assembly would have fought back. They may have been able to rally a significant number of moderates within Chávez’s diverse party against the claim that the Assembly was sovereign. Both the Venezuelan Supreme Court and Congress ceded almost complete control over the electoral rules to Chávez. Congress remained silent regarding the electoral rule. The Court, at least, forced Chávez to specify the electoral rule on the referendum ballot, but it failed to exert control over the content of the rules. What explains this failure? Both the Court and Congress lacked theoretical tools to draw upon. In Venezuela theorists were either Schmittian revolutionaries or constitutional supremacists, and the Court was sandwiched in between.70 The supremacists led by Allan Brewer Carías denied that the constituent power could justify a violation of the old Constitution’s amending procedure. The Court rightly recognised that neither position was tenable. But lacking any alternatives, at best it could muddle through. This paper is an attempt to provide some of the missing tools necessary to realise a vision of constitutional revolution that permits a significant break with the past without degenerating into semi-authoritarianism. By acknowledging

69  J Elster, ‘Constitutional Bootstrapping in Philadelphia and Paris’ (1992) 14 Cardozo L Rev 549, 564. 70  Combellas divided Venezuelans jurists position into three categories: 1. Those who favour the violation of the old amendment rule; 2. Those who believed the Constitution was supreme; and 3. Those who believed in an amendment to the Constitution to allow for a constituent assembly. Allan-Brewer Carías, who led the opposition to Chávez in the constituent assembly, is in group 3. I collapse groups 2 and 3 because both groups respect the supremacy of the Constitution. R Combellas, ‘El Proceso Constituyente y la Constitución de 1999’, (2003) 30 Politeia 183, 187.

Revolutionary Reform in Venezuela 151 that breaks can only be partial, it permits the inclusion of opponents and old institutions in the process of creating a new constitution. VI.  RADICAL BREAKS AND EXCLUSIONARY MANDATES

Inclusion obstructs radical change. Radicals define the people as the majority and mobilise it to enact a clean and total rupture with a shameful past. But what about the minority, the potential 49 per cent of the population? Radicals always fail to convert large and influential groups to their cause. Sometimes, these groups are associated with the old regime that revolutionaries seek to overthrow. For radicals, negotiation with these moderates or conservatives would water down the revolution by compromising with those who want to preserve parts of the old order. The more power and representation the minority has, the less radical the break with the past. And since that past is thoroughly corrupt, no part should remain.71 Radicals demonise those who disagree to justify their exclusion from the halls of power. Groups who disagree are not an opposition, but parasites, traitors, and enemies. It is not enough to defeat opponents in an election; they must be expelled or destroyed.72 Exclusion is necessary to accomplish a new order. Radicals have a majoritarian vision of the constituent power. By contrast, revolutionary reform can include diverse parties because the break with the past is partial. Those associated with the past are not essentially antithetical to the revolution. Continuity permits some elements of the past to be incorporated in the new regime. The opposition may need to be marginalised, but not excluded. The goal is not to destroy them, but to put them in their place. The opposition may negotiate, but they should realise the modest chips they bring to the table. A new order is coming, but the opposition can manage to find a place in it.73 Underlying the conflict between Hugo Chávez and the Supreme Court over the level of the level of inclusion in the electoral rule were different visions of how radically the new 1999 Constitution would break with the past. Chávez chose a ‘first past the post’ electoral rule to empower the majority to give a mandate to the constituent assembly to break with the oligarchs’ almost completely uninterrupted dominance over Venezuela that began with Spain’s imperial rule. Unfortunately, the Court and the opposition never managed to muster a rival narrative. I develop an alternative, revolutionary reformist narrative of Venezuelan history and the people to support an inclusive electoral rule to the constituent assembly. Like Chávez, it repudiates much of Venezuelan history, including the corruption and neoliberalism of the 80s. But it recognises and renews the promise of the 1961 Constitution, the longest lasting and first truly democratic constitution in Venezuelan history.

71 

cf Alcoreza, (n 2); Negri (n 2). eg, M Méndez-Rivera, ‘Rhetoric of H.R. Chávez, President of Venezuela’ (2006) 116–17, 176 (unpublished Ph.D. dissertation). 73 See, eg, S Schalvezon, El Nacimiento del Estado Plurinacional (CLACSO, Plural editors, 2013) (discussing some accommodation to the minority in the making of the 2009 Bolivian Constitution). 72  See,

152  Joshua Braver A.  Liberal Democracy’s Potential for Revolutionary Reform The history of liberal democracy in Venezuela opened up a space for revolutionary reform. Revolutionary reform struggles in the context of a transition from authoritarian rule to democracy, but may succeed in the transition among democratic regimes. In an authoritarian regime, revolutionary actors have no access to current institutions. Lacking this access, they cannot ‘unconventionally adapt’ and renew them. Instead, they must build narratives and justifications to justify the overthrow of the old order, often by force. The break is not only with the legality of a particular amendment rule, but with the legitimacy of the previous regime as well.74 In these situations, radical ideologies that seek to destroy the past thrive. In the French revolution against absolutism, the Jacobins executed the king and established a new calendar that started with the year one. Robespierre even attempted to create a whole new religion of reason. Vladimir Lenin not only rebelled against the Tsar, but also condemned all of traditional Russian peasant culture as backward. He imported and adapted the foreign ideology of communism to modernise Russia for the 20th century.75 By contrast, Hugo Chávez was elected through normal and constitutional means. He was tied to and part of the old institutions, and therefore had the potential to initiate a process that built on them rather than tried to wipe them out. The foundations of the system were not inherently corrupt. In fact, a pamphlet written to support the sovereignty of the Venezuelan constituent assembly noted, ‘Paradoxically, one our principle tasks is to consolidate the social rights contained in the 1961 Constitution’.76 The whole system was nudging toward revolutionary reform. The process of political decentralisation had strengthened a new generation of political leaders who had renewed local ties between the political parties and their constituents and purged much of their old leadership. Voting reforms had allowed new parties, including Chávez’s, to infiltrate Congress and facilitated Chávez’s rise to the presidency.77 Unlike France in 1789 or Russia in 1917, in 1999 in Venezuela the democratic system had opened itself up to insurgent groups who now held its top offices. The possibility for revolutionary reform was real. The system still needed a jolt to overcome inevitable opposition and Chávez’s charisma might have provided it. Would Chávez

74 

Arato (n 3) 177. narratives do compete against ones that include the past. In France, after the revolution overthrew absolutism, the Constituent Assembly was split between those who advocated English constitutional monarchy and those who were inspired by US state democracy. The factions compromised and created a constitution based on a general will, but that also included a King with absolute veto power. Perhaps, tellingly the inconsistencies greatly strained the stability of the Constitution. See KM Baker, Inventing the French Revolution: Essays on French Political Culture in the Eighteenth Century ­(Cambridge University Press, 1990) 252. In Russia, the ‘populists’ advocated for a regime based on Russia’s peasant culture. O Figes, A People’s Tragedy: The Russian Revolution: 1891–1924 (Penguin, 1998). 76  F Chacón, ABC de la Constituyente (Editorial Laboratorio Educativo, 1999) 8. 77  J Buxton, The Failure of Political Reform in Venezuela (Ashgate Publishing Limited, 2001). 75  Radical

Revolutionary Reform in Venezuela 153 build on these reforms? Could his own historical narrative support revolutionary reform? B.  First Past the Post: The Revolutionary Mandate to Destroy the Past In Venezuela, the turning point was the conflict over the electoral rule for elections to the constituent assembly. Chávez’s choice of first past the post (FPTP) voting reflects, not only a political calculation, but also a specific vision of how the Venezuelan people would relate to their past history. In it, Venezuela would break with almost all of its history to overthrow the ‘oligarchs’ by directly deciding on the content of their constitution through a mandate to the constituent assembly. Chávez’s narrative begins with Simon Bolivar founding the country by liberating it from imperial oppressors. After Bolivar’s brief influence waned, a new group of rich oppressors, the oligarchs, rose to dominate Venezuela to the systematic exclusion of the people. The oligarchs reigned throughout the civil wars between liberal and conservatives in the 19th century, in the military dictatorships of the early 20th century, and also throughout the longest period of stable liberal democracy from 1958 until the Chávez’s election in 1998. Chávez painted the Spanish imperialists, the conservative caudillos and the liberal democrats with the same broad and damning brush. Chávez would now repeat and complete Bolivar’s feat of liberation.78 The elections to the constituent assembly were the instruments by which the people would finally speak. The elections, not the internal negotiation within the assembly, would be the true site of choice for the design of the new constitution. In the July 1999 elections to the constituent assembly, the people would give the constituent assembly a clear mandate to overthrow the liberal, oligarchic institutions of the 1958 Constitution, to break with almost all of past Venezuelan history, and establish a new ‘participatory and social democracy’.79 Let’s understand how FPTP achieves Chávez’s two different aims of exclusion and a mandate. First, since Chávez’s break with the past was nearly total, he sought to exclude all those associated with it. Venezuela had been a ‘partidocracia’ or ‘party democracy’ where the elites of two indistinguishable political parties collaborated to enrich themselves at the expense of the majority of the population. A first past the post electoral rule would exclude from the constituent assembly the party parasites that had for too long preyed upon the majority. FPTP is inherently exclusionary. It awards all of the power to the individual or party that wins the plurality of the votes. The minority receives no representation. In Venezuela, it would ensure that the oligarchs would not hold back the tide of the revolution. The second aim is to empower the people to directly create their constitution through a mandate. The party elites would be excluded from the constitutional assembly, but that is only part of a process to ensure that the people would be the

78  R Gott, Hugo Chávez and the Bolivarian Revolution (Verso Books, 2011) 91–118; M Lopez-Maya, ‘New Avenues for Popular Representation in Venezuela’ in Canache and Kulisheck (ed) (n 68) 93. 79  See Chávez (n 19) 42.

154  Joshua Braver true authors of their constitution. What’s to stop Chávez or some other group from usurping the constitution-making process? Through a mandate, the people themselves would create a new social and participatory democracy. Chávez and the Constituent Assembly would merely carry out their will. Chávez argued, ‘When I say, as a citizen and as a soldier, I am subordinate to the mandate of the Constituent Assembly, it is because I am firmly convinced that the assembly decides that which is cried out for in the streets, that which is begged for in the roads, … that which the people say galloping towards the revolution’.80 The Constituent Assembly would not represent the people, but embody them. As political scientists have established, for a mandate to work voters must have the option of choosing a single cohesive party who will then wield ‘unblocked control of the policy-making process’. Usually, FPTP voting establishes these conditions by facilitating the establishment of two main political parties and giving to the victor a disproportionate share of the seats after the election.81 In other words, since Chávez wanted the constituent assembly to have a mandate to create a constitution that would overthrow the Venezuelan oligarchs, FTPT voting was the necessary electoral rule. Chávez’s preference for FPTP voting was a principled choice consistent with his long held populist and majoritarian vision of democracy and the people. A mandate is difficult to achieve in a PR system. PR disperses power among numerous parties who must negotiate policy decisions. The site of decision is not taken by the people at the election, but by representatives afterward. The representative is not a delegate or mouthpiece for the people, but must have discretion to make independent judgements about how best to represent them in the bargaining process.82 PR was associated with the old Venezuelan party system and the usurpation of the popularly initiated Colombian Constituent Assembly a decade earlier.83 After party elites had become the new oligarchs of the mid and late twentieth century, how could Chávez use an electoral rule that again would place trust in a new set of elites to give power back to the Venezuelan people? Chávez sought a constitutional assembly that would mirror the Venezuelan people. C.  Proportional Representation: Inclusion and the Pluralised People What alternative narrative and vision of the Venezuelan history might have justified a more inclusive electoral rule, such as proportional representation, for electing the constitutional assembly? It would be a narrative of ‘revolutionary reform’ that both built upon the legitimacy of the past and radically broke with the present to fulfill a set of national ideals. Like Chávez, it would start with Simon Bolivar founding the country. It too might note how in the 19th and early 20th century, caudillos and

80 

Chávez (n 54) 83. Powell, Elections as Instruments of Democracy: Majoritarian and Proportional Visions (Yale University Press, 2000) 4–17. 82 ibid. 83  Chávez (n 79) 44. 81  GB

Revolutionary Reform in Venezuela 155 dictators squandered Bolivar’s victory to aggregate power and wealth at the expense of the people. With Chávez, the narrative would celebrate those rebels who fought against this oppression.84 After following Chávez’s narrative for the first 147 years of Venezuelan history, the narrative would diverge at the establishment of the 1961 constitution and liberal democracy in Venezuela. Rather than another chapter in the oligarchic reign, this Constitution opened the pathway for the modernisation and democratisation, establishing universal suffrage for the first time in Venezuelan history and facilitating the longest period of Venezuelan stability and prosperity. The urban poor greatly benefited from these changes. However, the country had lost its way in the late 1980s and early 1990s as oligarchic political and business elites squandered economic gains. Then, to address the deficit they created, political elites institutionalised neo-liberal reforms that eviscerated the social safety net. They had violated the spirit of the 1961 Constitution, and that constitution as written and interpreted, was no longer sufficient to confront the problems of the present. The narrative would call for the people to renew and accelerate the constitutional reforms initiated in the late 90s so that it could be radically rewritten or amended to fulfill its own promise of representative and even participatory democracy.85 Note that the difference between Chávez’s narrative and that of revolutionary reform is not large. For a revolutionary, Chávez’s narrative is surprisingly backward looking. The paradigm social revolutions look forward. As Karl Marx stated, The social revolution of the nineteenth century cannot take its poetry from the past, but only from the future. It cannot start upon its work before it has striped away of all superstition concerning the past. Former revolutions require recollections of past world history to smother their own content. The revolution of the nineteenth century must let the dead bury their dead in order to arrive at its own content. Then the phrase went beyond the content— here the content goes beyond the phrase.86

Chávez speaks of a total and revolutionary break with the past, but the question is which past? With Marx, are you breaking with all of past history? No. Chávez actually speaks of restoring the past, of refounding Bolivar’s Republic. The narratives of both Chávez and revolutionary reform both start with a veneration of Bolivar and then condemn the subsequent 147 years of Venezuelan history. They part ways for 30 years or so starting in the mid-twentieth century, and then converge again on the moment of neo-liberal reforms in the late 1980s and early

84 

See Gott (n 78) 111. This is very similar to the narrative undergirding the French Constitution. The preamble legitimises the document by having the people harkening back to and ‘solemnly proclaiming their attachment to the rights of man and Citizenship’. It then skips the intervening 169 years of history that were filled with the terror, dictatorships, and monarchy to the claim that the 1789 rights of Man were ‘confirmed and complemented by the Preamble to the Constitution of 1946’. Like the Venezuelan narrative I offered, it begins at legitimate origin, skips hundreds of years of dark history, and then builds upon a significant but deeply flawed period of democracy. The US process, by contrast, contains much more continuity. 86  K Marx, The Eighteenth Brumarie of Louis Bonaparte, accessed 1 August 2016. 85 

156  Joshua Braver 1990s. Unlike Robespierre or Lenin, Chávez did not seek to smash all of past ­history. He sought to restore a purer past, to re-found Bolivar’s Venezuela. As Chávez stated in his speech to the constituent assembly, for a ‘conglomerate of persons’, for the ‘horde’ to become a people, they must be ‘janus-faced, janus of the mythology, that has two faces: one looking to the past and the other to the future … looking and ­feeling its glorious common past, but have a common will to the future’.87 This is not to ignore the significant differences between Chávez and the narrative of revolutionary reform, but I emphasise the convergence to show how an alternative vision was viable, suitable and appealing in a time of constitutional crisis. Even as revolutionary reform’s narrative condemns the old party bosses of the traditional political parties, it provides room for the inclusion of a new generation of political leaders that were in the midst of regenerating their parties. In this narrative and with a more inclusive electoral rule, Chávez and his allies would likely have the most control over the constitution-making process, but they would have to give some concessions to and answer the criticisms of the opposition. By contrast, Chávez’s FPTP electoral rule and constituent assembly merely transferred concentrated power from the old political parties into his own hands. VII. CONCLUSION

The dichotomy between constitutional amendment and revolution has distorted and obscured the potential of revolutionary reform. ‘Radicals’’ vision of a total rupture with the past requires the constituent assembly to usurp the people’s rightful sovereignty and centralise it in an assembly. This is a recipe for semi-authoritarianism. But ‘critics’ of revolution are foolhardy to believe that their case studies will banish popular constitution-making from our political vocabulary. It is the fundamental basis of modern constitutionalism and opens up space for needed constitutional change and the rise of new political actors. In political struggles, critics’ pleas will fall on deaf ears. The issue is not whether, but what type of revolution will be at the forefront of future acts of constitution-making. By invoking Hannah Arendt and retelling the story of the creation of the 1999 Venezuelan Constitution, I positioned constituent power as an act of revolutionary reform. It acknowledges popular sovereignty and permits radical breaks, but by calling upon the ideals and adapting the institutions of the past it seeks to channel, harness and constrain the revolution. It prevents any one actor from dominating the process of writing a new constitution and opens up the process of constructing the people to multiple actors. It is no guarantee against the next would-be charismatic dictator, but offers more hope than the alternative extremes of either abolishing revolution or unleashing it without any restraint.

87 

Chávez (n 54) 46.

8 ‘Revolutionary Reform’ and the Seduction of Constitutionalism JULIANO ZAIDEN BENVINDO*

J

OSHUA BRAVER’S INTRIGUING and fascinating paper ‘Revolutionary Reform in Venezuela: Electoral Rules and Historical Narratives in the Creation of the 1999 Constitution’ provides a thorough analysis of the recent c­onstitutional reality in Venezuela. It shows how, had the Venezuelan Supreme Court and Congress directly participated in the definition of the electoral rules for the creation of the 1999 Constitution, a ‘revolutionary reform’ would have been possible. By recalling the non-foundational feature of constitutional democracy, this paper comments on Braver’s article to argue that, whereas constitutionalism has an inherent seduction, it is also limited by the very historically contingent process of social life. As such, the ‘make-beliefs’ constitutionalism holds have to face the reality, and, in this case, a reality of a country historically marked by semi-authoritarian practices. In this complex scenario, the dilemmas and paradoxes of constitutionalism are pushed to their very extremes. This chapter, while agreeing with Braver’s understanding of how Venezuela has lost sight of constitutionalism, sees that any conclusion in this matter must be intertwined with perspectives of economic development. It concludes that Braver’s article, even though seduced by constitutionalism, is an excellent wakeup call for understanding that constitutional democracy is only possible in its very impossibility. I.  A REVOLUTION IN CRISIS: BRAVER’S NARRATIVE OF THE CREATION OF THE 1999 CONSTITUTION IN VENEZUELA

A fascinating debate over the very limits of constitutional democracy in the context of a revolution in crisis—this might properly depict Joshua Braver’s main arguments

*  I am very grateful to the participants of the Workshop on Comparative Constitutional Amendments, held at Boston College in May 2015, and particularly Richard Albert, Xenophon Contiades and Akmene Fotiadou for organising such a fruitful workshop. I am also thankful to the peer reviewers who contributed with very relevant comments. This chapter critically comments on Joshua Braver’s Article Revolutionary Reform in Venezuela: Electoral Rules and Historical Narratives in the Creation of the 1999 Constitution.

158  Juliano Zaiden Benvindo about Venezuela and its creation of the 1999 Constitution. Hannah Arendt and Bruce Ackerman, through the concepts of ‘revolutionary reform’ and ‘unconventional adaptation’, on the one hand, and Carl Schmitt, as the primary source for what Braver calls ‘radicals’,1 on the other hand, are his main theoretical perspectives to investigate the turbulent transition to a new constitutional moment—if we can call it as such2—in Venezuela. At the core of his argument lies the purpose of introducing a new approach to regime transitions which is neither a ‘radical constitutional revolution’3 nor a simple reform of the past constitutional reality. A ‘revolutionary reform’, as he calls it, structured around the idea of ‘unconstitutional adaptation’, is a better response to these transitional moments. Rather than a total break with the past, Braver argues that, in this case, ‘the constitution-making process violates specific rules, but does not degenerate into a contest of strength or force. It is an extra-legal process justified and constrained by the renewal of principles of the old regime’.4 Moreover, it is a process marked by pluralism and collaboration: ‘Old institutions are unconventionally adapted to frame a collaborative process to create a new constitution’.5 Had the Venezuelan institutions, especially the Supreme Court and Congress, furthered mechanisms for such a ‘revolutionary reform’, they would have better protected Venezuela from already foreseeable tendencies towards semiauthoritarian practices. Braver’s argument is that they failed to do so, particularly with respect to the electoral rules, and this ‘empowered Chávez to pursue true revolution by purging the opposition from all levels of government’.6 Beneath this argument, which Braver stresses by placing side by side Carl Schmitt and Hannah Arendt, lies the central debate over the seat of sovereignty. In a certain way, the question ‘who are the people?’ appears in the midst of a controversy over the role of a ‘people’ in the transition to a new constitution in that country, and particularly how institutions used these ‘people’ to strengthen their own positions of power. Structurally speaking, while introducing his concept of ‘revolutionary reform’ as a possible way out to this revolution in crisis, Braver’s paper revolves around the debate over the seat of sovereignty and how institutions—in the case, the Presidency by Hugo Chávez, the Venezuelan Supreme Court and Congress—behaved in such a moment of constitutional transition. In his view, in the end, the ‘radical constitutional revolution’ Venezuela has undergone has unfortunately placed Carl Schmitt as victorious over Hannah Arendt.7 Instead of a ‘revolutionary reform’, Schmitt’s radical definition of the people prevailed, leading thereby to the premise that they are not only ‘outside the pre-existing political order, but superior as well’.8

1  J Braver, ‘Revolutionary Reform in Venezuela: Electoral Rules and Historical Narratives in the Creation of the 1999 Constitution’, in this volume. 2  I provide a critical analysis of the concept of constitutional moments in JZ Benvindo, ‘The Seeds of Change: Popular Protests as Constitutional Moments’ (2015) 99 Marq L Rev 363. 3  Braver (n 1) 138. 4  ibid 144. 5 ibid. 6  ibid 139. 7 ibid. 8  ibid 142.

‘Revolutionary Reform’ and the Seduction of Constitutionalism 159 Braver’s description of the events in Venezuela is very careful and gives relevant insights about the turbulent constitutional reality of that country. It begins with Chávez’s landslide victory in the presidential elections of December 1998 and his already announced purpose of convening a constituent assembly to write a new constitution. Braver provides a careful discussion of the main episodes of the referendum that followed Chávez’s election, especially by focusing on the turbulent relationship between Chávez’s intents and the Venezuelan Supreme Court. Particularly with respect to the definition of the electoral rules in this referendum, the Supreme Court—and also Congress—could have had a more direct participation, creating thereby mechanisms that could have engendered a ‘new beginning’ without, however, a total rupture with Venezuela’s past constitutional reality. His main thesis is that, had the Supreme Court and Congress been successful in instituting an ex ante method of selection of the members of the constituent assembly, Venezuela could have undergone a ‘constituent power as an act of revolutionary reform’.9 However, reality proved differently. Based on the premise that the sovereignty resides in the people, as stated in Article 4 of the 1961 Constitution, the idea of a total revolution that could nonetheless circumvent the institutional constraints and constitutional thresholds of the past triumphed. As Braver says, in this struggle between Chávez and the Supreme Court, ‘Chávez ultimately prevailed’10 by ‘adopting a first past the post method of aggregating votes for future elections to the constituent assembly’.11 From that point onwards, Chávez had the vast majority of the seats in Congress (93.1 per cent) and could thereby strengthen his authority while presumably acting in the name of the people. The associations with Carl Schmitt, whose unfortunate victory over Hannah Arendt and also the idea of a ‘revolutionary reform’ seems evident in this case, are straightforward. In fact, as Braver describes it, one of the first things this constituent assembly did was to declare a state of emergency.12 As a condition for building up the new constitutional and political reality, to declare the exception seems to be a ­no-brainer. It is as if that exception were ‘the condition of possibility of extraordinary sovereign popular intervention, that is, of democratic constitutional founding’.13 As in Schmitt, no one better to embody this ‘democratic constitutional founding’, this ‘sovereign popular intervention’ than Chávez, in a clear identification between the ruler and the ruled. As Braver mentions, ‘the opposition had not only been excluded from the process of drafting the constitution, but also from all other possible positions of power at every level of government’.14 Braver’s words are strong to describe this Constituent Assembly—he says that it ‘stacked with Chávez allies, became a despotic and all-powerful actor that illegitimately usurped the constituted powers’15—and

9 

10 

ibid 156. ibid 146.

11 ibid. 12 

ibid 147. Kalyvas, Democracy and the Politics of the Extraordinary (Cambridge University Press, 2008)

13  A

119. 14  Braver (n 1) 147. 15 ibid.

160  Juliano Zaiden Benvindo the immediate connection, as did Arendt,16 was with the Jacobin Convention of 1793.17 The people, now personified by this Constituent Assembly, could act in fully unbound ways, in a clear contradiction with the premise that ‘the sovereignty lies with the people, and not with the assembly’.18 The question ‘who are the people?’ remains thus central in Braver’s analysis of those events.19 Braver puts forward a different approach to the reasons why the Supreme Court and Congress could have acted differently in this case in order to constrain Chávez to follow some rules of the game. This is where his thesis of ‘revolutionary reform’ directly applies to that reality. Unlike many who place greater responsibility on the Supreme Court’s approval of the referendum as the reason why Chávez could act in a quite unbound way, Braver contends that ‘the turning point was not the Court’s decision to have a referendum, but it and Congress’ insufficiently firm intervention concerning the construction of the electoral rules for the constituent assembly’.20 In fact, he highlights the very impossibility of the Court to block that referendum as it took place, because of: Chávez’s great popularity;21 Chávez’s possibility of purging the Court with little political cost;22 the demand for change, especially after years of an ineffective Congress;23 ‘constitutional boot-strapping’, that is, ‘once the constituent assembly is convoked, it becomes very difficult for the old regime to control or limit it’;24 and (e) the lack of ‘theoretical tools to draw upon’.25 (a) (b) (c) (d)

Therefore, a possible way to create the conditions for a ‘revolutionary reform’ would be—as Braver says—for the Court and Congress to ‘participate in the creation of the people themselves, especially through the construction of the electoral rule for the constituent assembly’.26 By making pluralism a reality instead of the exclusion of almost any opposition, as it happened after the referendum, the so-called revolution could, on the one hand, significantly break with the past, and, on the other hand, preserve mechanisms that would block semi-authoritarian practices.27 A structural but ­partial rupture with the past would provide an environment to strengthen democracy and not the other way around, providing the means to include diverse parties.28 This was the role the Court and Congress could have undertaken. They could and should not have opposed the referendum itself, but could and should have participated in this

16 

H Arendt, On Revolution (Penguin Books, 1990) 239. Braver (n 1) 147. 18 ibid. 19  See Part II. 20  Braver (n 1) 149. 21  ibid 149. 22  ibid 150. 23 ibid. 24 ibid. 25 ibid. 26 ibid. 27 ibid. 28  ibid 151. 17 

‘Revolutionary Reform’ and the Seduction of Constitutionalism 161 movement by ‘unconventionally adapting’ the old institutions to this new reality. As he says, ‘the Court and Congress best path forward was not to deny the people their rightful power, but to participate in the creation of the people themselves’.29 A ‘revolutionary reform’ through ‘unconventional adaptation’ would thereby preserve some constitutional features of the past. In this regard, Braver argues that albeit Venezuela’s historical troublesome reality, the 1961 Constitution should not be despised as the cornerstone of the previous oligarchical and corrupt regime. Rather, the ‘revolutionary reform’ ‘recognises and renews the promise of the 1961 Constitution, the longest lasting and first truly democratic constitution in Venezuelan history’.30 Especially because the regime transition did not emerge from a dictatorship, but from a ‘truly democratic constitution’, and also because the past ‘whole system was nudging toward revolutionary reform’,31 Braver understands that the conditions for an ‘unconventional adaptation’ were on the table. Yet, although the ‘possibility for revolutionary reform was real’,32 the need to combat the enemy identified with that past, as did Chávez, made this move practically impossible. Braver brings forward a very interesting connection between this need and the Venezuelan history characterised by oligarchical and corrupt power. By drawing from the past narrative of Simon Bolivar as the hero against imperialism, Chávez’s Bolivarian Revolution had to battle against those groups that systematically promoted actions to exclude the majority of the people. Chávez would be the contemporary Simon Bolivar in this purpose of setting the people free from ­oppression.33 In a somewhat direct revival of Schmitt’s path towards democracy, that referendum that led to the almost complete annihilation of opposition in the Constituent Assembly would mean, as Braver says, that ‘the oligarchs would not hold back the tide of revolution’34 and the people, through mandate, could ­‘create a new social and participatory democracy’.35 In some ways, ruler and ruled are identified—‘the Constituent Assembly would not represent the people, but embody them’36—as well as the opponent is the enemy to be ostracised. The scenario could have been different, had the Supreme Court and Congress undertaken a more participatory role. As Braver says, a ‘narrative of “revolutionary reform” that both built upon the legitimacy of the past and radically broke with the present to fulfill a set of national ideals’37 could have been possible. Instead of attacking the 1961 Constitution as an oligarchical product, this ‘revolutionary reform’, according to Braver, would see it as the Constitution that ‘opened the pathway for the modernisation and democratisation’38 as well as provided ‘the longest

29 

ibid 149–150. ibid 151. 31  ibid 152. 32 ibid. 33  ibid 153. 34 ibid. 35  ibid 154. 36 ibid. 37 ibid. 38  ibid 155. 30 

162  Juliano Zaiden Benvindo period of Venezuelan stability and prosperity’.39 Although the late 80s and 90s were characterised by the increase of the oligarchical power and elites’ neoliberal reform, Braver’s ‘revolutionary reform’ narrative would lead the people to ‘renew and accelerate the constitutional reforms initiated in the late 90s so that [the 1961 Constitution] could be radically rewritten or amended to fulfill its own promise of representative and even participatory democracy’.40 Braver’s final argument aims to rebuild the past with the historical and new demands of a society in need of change. His narrative, he argues, is not very different from Chávez’s. He also sees in Simon Bolivar a hero to venerate. He also condemns most of the Venezuelan history of exclusion and oppression.41 Their differences are nonetheless in how they interpret those years of true democracy during the approximately 30 years after the promulgation of 1961 Constitution. The ‘revolution reform’ is thereby a radical but partial change. It preserves that moment when Venezuela paved the way for a new future as a means to avoid, as it happened with Chávez, that the new ‘constituent assembly merely transferred concentrated power from the old political parties into his own hands’.42 His conclusion clearly reminds us of the question ‘Who are the people?’. The message he leaves is that whenever we deify the people, whenever we see the people with no boundaries at all, whenever pluralism is replaced by a hegemonic discourse, whenever someone uses a substantive conception of democracy to strengthen his own power, the path to authoritarianism is looming. His argument is straightforward, as well announced in many parts of his text: he believes in the ‘constituent power as an act of revolutionary reform’.43 He, inspired by Arendt, sees that radical breaks should happen only by ‘calling upon the ideals and adapting the institutions of the past it seeks to channel, harness and constrain the revolution’.44 This is needed to ‘prevent any actor from dominating the process of writing a new constitution and opens up the process of constructing the people to multiple actors’.45 But these strong words come naturally with some doubts: would not Braver be placing constitutionalism as the last source of constitutional democracy, as a metaphysical standpoint from where the solution to have ‘more hope than the extremes of either abolishing the people or unleashing it without any restraint’46 would ever emerge? II.  THE DILEMMA OF THE SOVEREIGNTY OF THE PEOPLE

In a brilliant text whose title—Declarations of Independence—echoes the very enigmatic sense of the tense but productive relationship between constitutionalism and

39 ibid. 40 ibid. 41 

ibid 155. ibid 156. 43 ibid. 44 ibid. 45 ibid. 46 ibid. 42 

‘Revolutionary Reform’ and the Seduction of Constitutionalism 163 democracy, Jacques Derrida poses the question: ‘Who signs, and with what so-called proper name, the declarative act that founds an institution?’47 Derrida, through the example of American constitutionalism, is stressing not only the founding act of a new order, but mostly asking who the people are. Briefly, he is connecting the act of an institutional foundation with the democratic premise of the sovereignty of the people. According to him, such a Declaration exceeds the boundaries of the constative: ‘The signature maintains a link with the instituting act, as an act of language and an act of writing, a link that has absolutely nothing of the empirical accident about it’.48 The constative negotiates with the performative, which, in turn, must always remember the signature, the constative: ‘The founding act of an institution— the act as archive as well as the act as performance—must maintain within itself the signature’.49 Derrida’s words are certainly intriguing and call the attention of how circular the act of founding a constitutional reality is. In the end, there is no end—or, rather, in the beginning, there is no origin. There is an insurmountable negotiation between the constative and the performative,50 instead, which is the very condition of legal history. No grounds, but only traces and interactions mediated by language. If we go further and ask what the bases of constitutional democracy are, we will face simply history, violence, and faith. This is what remains beneath this ‘mystical foundation of authority’,51 this ‘simulacrum of the instant’,52 this ‘vibrant act of faith’.53 The lack of any ground is what makes constitutional democracy possible in its very impossibility. The constative and the performative, the history and the promise are what make constitutional democracy this ongoing paradoxical repetition in the very reinvention of a moment, where there are no guarantees and the risk of its disruption is always there. This is also what makes the so-called people a sort of fiction in this very moment. After all, as Seyla Benhabib argues, ‘in a constitutional democracy, there is no final seat of sovereignty’.54 As seen, Braver’s argument of ‘revolutionary reform’ through ‘unconventional adaptation’ leads inevitably to a discussion of the seat of sovereignty. In the end, in the face of a regime transition, the question ‘Who are the people?’ has to be raised, even to understand how to better negotiate the past, the present, and the future. For if a ‘people’ is the final seat of sovereignty, perhaps we are placing a ground, a final call for constitutional democracy. Furthermore, this might bring about the consequence of using these very ‘people’ to engender what David Landau calls ‘abusive constitutionalism’55 in Venezuela. A metaphysical standpoint would then 47  J Derrida, ‘Declarations of Independence’ in J Derrida (ed), Negotiations: Inteventions and Interviews, 1971–2001 (Stanford University Press, 2002) 47. 48  ibid 47. 49  ibid 48. 50  ibid 49. 51 J Derrida, ‘Force of Law: The Mystical Foundation of Authority’ (1990) 11 Cardozo L Rev 920, 939. 52  Derrida (n 47) 51. 53  ibid 52. 54  Sh Benhabib, ‘Democracy and Difference: Reflections on the Metapolitics of Lyotard and Derrida’ in L Thomassen (ed), The Derrida-Habermas Reader (Edinburgh University Press, 2006) 140. 55  D Landau, ‘Abusive Constitutionalism’ (2013) 47 UCDL Rev 189.

164  Juliano Zaiden Benvindo prevail over the negotiation between the constative and the performative, jeopardising thereby what we understand by constitutional democracy. In particular, Chávez would use the ‘people’ as the seat of sovereignty, now reaffirmed by him as the source of his power, and the source for circumventing the constitutional constraints and institutional thresholds limiting his—as a derivative of the people’s—power. The ground—the ‘people’—would then make the constitution a flexible constraint the government could easily abuse, while the argument would still sound very sympathetic for the masses. This is why the concept of ‘revolutionary reform’ should be carefully revisited. The idea of a people building a new reality while dealing with the previous constitutional constraints and institutional thresholds is, after all, at the core of this concept. Braver draws from the opposition between Carl Schmitt and Hannah Arendt, refined by Bruce Ackerman’s concept of ‘unconventional adaptation’, his main thesis.56 But maybe he could have gone further and explored what these ‘people’ really mean for constitutional democracy and thereby for his own argument. An interesting discussion is, for example, the very fictional feature of these people as paradoxically a protection for the real people. As Edmund Morgan, in his book Inventing the People, argues ‘Government requires make-believe (…) Make believe that the people have a voice or make believe that the representatives of the people are the people’.57 This fictional feature of these people, even though needing to bear some resemblance to fact in order to make it viable,58 is thereby a condition for constitutional democracy: ‘the fictional qualities of popular sovereignty sustain rather than threaten the human values associated with it’.59 Or, as Zoran Oklopcic says, we should see these people as ‘imaginative and institutional leftover space, delineated by our conscious refusal to embrace its alternatives’.60 They transcend, therefore, the real people—without forgetting them, though— because these people are indeed many people and can be as such called insofar as constitutionalism fosters pluralism and protects minorities. While not confusing with a certain particular people, sovereignty is affirmed through the very characteristics of constitutionalism: people are as such understood insofar as they free themselves while limiting themselves through constitutionalism. Again, but now with different words, Edmund Morgan is in some ways also emphasising that we cannot place a final ground in the concept of sovereignty. Moreover, there is a more prosaic aspect in the question ‘Who are the people?’. The simple fact that the people’s voice is channeled into institutional mechanisms of representation and deliberation already imposes a limit upon the spectrum of this sovereignty. The people, as the primary constituent power,61 are originally constrained by the very characteristics

56 

See Braver (n 1) 139–144. ES Morgan, Inventing the People (WW Norton & Company, 1988) 13. 58  ibid 14. 59  ibid 15. 60 Z Oklopcic, ‘Constitutional Theory and Cognitive Estrangement: Beyond Revolutions, Amendments and Constitutional Moments’, in this volume, 70. 61  The primary constituent power refers to the people in constitution-making, while the second constituent power relates to the process of constitution-amending. See Y Roznai ‘Amendment Power, Constituent Power, and Popular Sovereignty’, in this volume. 57 

‘Revolutionary Reform’ and the Seduction of Constitutionalism 165 of constitution-making, while, likewise, the very process of constitution-making and constitution-amending (when the secondary constituent power acts) is limited by the very fictional feature of popular sovereignty.62 This double bind functions then as a mutual possibility and impossibility of constitutional democracy.63 The problem resides, however, when this double bind is put in jeopardy. This is where Joshua Braver’s comments on Carl Schmitt makes a lot of sense as an example of a constitutional theory which is anchored to placing a final ground in the concept of sovereignty. As long as, for Schmitt, democracy has as its central concept a people and not the humanity and is defined as the identity between rulers and ruled,64 it follows that ‘democratic substance precedes formal legalism and abstract proceduralism’.65 A substantive concept of democracy, kept away from any fictional or procedural view of the sovereignty of the people as above discussed, effectively surrenders constitutionalism to decisionism.66 The consequence is that whoever has the authority of the last word—and for Schmitt the last word resides in the people in this identity between rulers and ruled—can change the constitution or draft an entirely new constitution according exclusively to his or her (identified with the people’s) will and in an entirely unfettered way. It is extralegal rather than illegal,67 it is when the political68 overcomes the law. Simply speaking, Carl Schmitt is the paramount source whenever we are faced with the use of concepts such as democracy or popular sovereignty to mean what constitutional theory usually aims to avoid. The substantive feature of those concepts in Schmitt’s viewpoint contradicts many of those assumptions constitutional theory has developed, and consequently it works as a wake-up call for the dangers of abusing of such concepts.69 Although, as Joshua Braver mentions, Schmitt might not be the source for understanding the aftermath of a transition to a new constitutional order,70 it is certainly the inspiration for building up a façade of democratic discourse where authoritarianism prevails. The use of the word ‘people’ or ‘sovereignty of the people’ to promote quite unbound changes in constitutionalism paves the way

62 

See Roznai (n 61). L Thomassen, ‘A Bizarre, Even Opaque Practice: Habermas on Constitutionalism and Democracy’ in L Thomassen (ed), The Derrida-Habermas Reader (Edinburgh University Press, 2006) 186. 64  See C Schmitt, Verfassungslehre [Constitutional Theory] (Duncker & Humblot, 2003) 234. 65  Kalyvas (n 13) 117. 66  This argument is related to Schmitt’s statement that the ‘sovereign is who decides on the exception’ [Souveräin ist, wer über den Ausnahmezustand entscheidet] See C Schmitt, Politische Theologie [Political Theology] (Duncker & Humblot, 2009) 11. 67  See Kalyvas (n 13) 118. 68  Carl Schmitt defines the political in a very concrete perspective according to the division between friend and enemy and by directly focusing on the exception rather than the rule. According to him, ‘the concept of friend and enemy are to be understood in their existential and concrete sense but not as symbols or metaphors. It cannot be mixed or weakened by the economy, morals or other concepts as an expression of private feelings and tendencies’. C Schmitt, Der Begriff des Politischen [The Concept of the Political] (Duncker & Humblot, 2009) 27. For Schmitt, therefore, there is no political without the figure of the enemy. As Derrida argues, ‘the disappearance of the enemy would be the death knell of the political as such’. J Derrida, ‘The Politics of Friendship’ (1988) 85 The Journal of Philosophy 632, 684. 69  See Landau (n 55) 189. 70  Braver (n 1) 141. 63  See

166  Juliano Zaiden Benvindo to undermine pluralism and then democracy itself. The ‘make-believe’ of a people constitutional democracy relies on as a protection of the real people is replaced by an abusive and strategic use of this very people. Braver’s distinction between radical transition and ‘revolutionary reform’ lies thereby in the very limits and constraints the people have when establishing a new constitutional reality. Hannah Arendt understands, contra Schmitt, the paradoxical nature of constitutional democracy and, as an admirer of the American ­Revolution,71 certainly saw in James Madison’s institutional design to challenge the threats of factions,72 let alone the many examples in history, a message to be taken seriously. Therefore, in some ways similar to many relevant perspectives of constitutional theory, such as Derrida’s and Morgan’s above, Hannah Arendt also questions the idea of attributing the final seat of sovereignty to the people, as if they were the ‘ultimate authority’73 and, as such, ‘unbound by any law’.74 In the same manner, Arendt sees how using a substantive concept of a people, as if there were a ‘will of multitude’,75 can lead to the disruption of pluralism and, consequently, engender the rise of authoritarianism through the identification of the ruler and ruled. Yet, despite the ‘dangerous freedom of the people’,76 aren’t these people the final seat of sovereignty, after all? Jacques Derrida, as previously introduced, may bring forward a more radical approach to this subject. When we lose sight of the fictional character of the people, and identify rather a particular people as the only source of law, as if they could do whatever they want in this new constitutional moment, there is a false presumption here. For there is no new beginning that is entirely pure or unbound. Even though marked by violence and faith, there is, in every beginning, some repetition; there is a certain ‘call for self-conserving repetition’.77 The Derridian concept of iterability— as the repetition in every new beginning78—shows that, even if the people are willing to completely break with the past, there will be limits on doing that. Their sovereignty is already bound to some extent to constraints that weaken these people as the ‘ultimate authority’. Instead of a final ground, these people are interpreted as part of a process, of a negotiation wherein ‘they do not exist as an entity, the entity does not exist before this declaration, not as such’.79 Derrida’s words clearly contradict the argument that the people are the ‘ultimate authority’, since ultimately ‘the signature invents the signer’.80 This might explain why Arendt, as Braver contends,81 sees no point in sustaining a revolution that totally breaks with the past, arguing instead in

71 

See Arendt (n 16) 215–81. J Madison, ‘The Federalist, 10’ in A Hamilton, J Madison, and J Jay, The Federalist Papers (Oxford University Press, 2008) 49–55. 73  Braver (n 1) 141. 74 ibid. 75  ibid 142. 76  ibid 141. 77  Derrida (n 51) 997. 78  ibid 997. 79  Derrida (n 47) 49 80  ibid 49. 81  Braver (n 1) 143. 72 

‘Revolutionary Reform’ and the Seduction of Constitutionalism 167 favour of a renewal of the foundations of the current regime. This is also why Arendt places greater emphasis on the role of institutions in constitutional transitions while comparing the French and American Revolutions.82 Still, as Joshua Braver clearly argues,83 this focus on the capacity of old institutions to impose some limits upon the people, as it happened in the United States, might be overstressed. Since those limits were violated during the ratification process of the American Constitution, in the end, the power of the people is not as radical as Schmitt sustains, but it is not as limited by the old institutions as Arendt seems to argue, either. Schmitt, therefore, would represent the argument in favour of the people, placing greater emphasis on democracy (in his case, radicalised by the materialisation of the concept of a people). Arendt, in turn, would represent the argument in favour of constitutionalism, stressing the necessary constraints of old institutions upon the people. This unbalance should be solved by bringing another concept, one that more properly understands the complexity of regime transitions. This is where Braver recalls, in an adapted manner, Bruce Ackerman’s concept of ‘unconventional adaptation’ of old institutions84 to coin his argument in favour of a ‘revolutionary reform’. In his words, this ‘unconventional adaptation’ would mean that ‘there is continuity because revolutions use institutions, but their rules are bent, reinterpreted or even broken to create new constitution’.85 This concept may better show that the constituent power is a complex phenomenon, wherein institutions and people negotiate between themselves in an adaptive manner. It might better explain this unbalance between democracy and constitutionalism. Yet, how this adaptation will evolve in a certain context is its Achilles’ heel. Braver understands that the concept of ‘revolutionary reform’ is, at least in the context of contemporary Venezuela, the most suitable resolution to this deadlock, to this unbalance between constitutionalism and democracy. He says that ‘by invoking Hannah Arendt and retelling the story of the creation of the 1999 Venezuelan Constitution, [he] positioned constituent power as an act of revolutionary reform’.86 The question is whether, in the end, Braver is not falling into the risk of placing greater emphasis on constitutionalism, even though recognising the value of popular sovereignty. After all, he writes that his argument ‘acknowledges popular sovereignty and permits radical breaks, but by calling upon the ideals and adapting the institutions of the past it seeks to channel, harness and constrain the revolution’.87 In fact, by defending the need of constraints upon the people, he seeks to defend pluralism: his view would ‘[prevent] any one actor from dominating the process of writing a new constitution and [open] up the process of constructing the people to multiple actors’.88 But, once again, wouldn’t Braver’s viewpoint mean a certain look at this reality somehow expressing Arendt’s excitement about American constitutionalism? And, while doing that, wouldn’t he fall into the risk of placing a final ground in 82 

Arendt (n 16) 155–78. Braver (n 1) 144. 84  ibid 144. 85 ibid. 86  ibid 156. 87 ibid. 88 ibid. 83 

168  Juliano Zaiden Benvindo the negotiation between the constative and the performative, the iteration and the alteration?89 III.  THE SEDUCTION OF CONSTITUTIONALISM: WHEN CONSTITUTIONALISM FACES THE CONTINGENCIES OF SOCIAL LIFE

Constitutionalism is a seductive word. Especially in a context where we can clearly see that constitutionalism is not working properly, where we can interpret it as a form of ‘abusive constitutionalism’,90 ‘stealth authoritarianism’91 and the like, our genuine reaction is to stand up against such practices that may cause it any harm. We defend constitutionalism because we see that democracy is only as such understood as a constitutional democracy. In the negotiation between constitutionalism and democracy, one is not possible without the other. ‘Constitutionalism is precisely not only an enabling but also a necessary condition of democracy (and vice versa).’92 Moreover, there is no possibility of thinking of democracy without direct emphasis on institutions. We could not imagine stable democracies without the institutionalisation of distinct channels of exercising free public speech and deliberation.93 Still, in this interplay of concepts, there is a vast area of uncertainties and fragilities which certainly push constitutional theory—and many of its presumptions—to its very limits. Venezuela’s recent constitutional history is a paradigmatic case to observe how one can subvert this interplay of concepts in the course of authoritarian actions: constitutionalism surrenders to the will of the people as the final seat of sovereignty, now embodied by the executive power. The paradoxical concept of ‘revolutionary reform’ according to which institutions of the past are adapted to the new times in the very moment of constitution-making would thereby give to this people’s ‘ultimate authority’94 a less risky meaning. These people would be interpreted in a pluralistic way, as ‘multiple actors’,95 and not as a concept one single actor could manipulate according to his will. Seductive as it is, Braver’s belief in the strength of constitutionalism—and, especially the past constitutionalism as the cornerstone for his concept of ‘revolutionary reform’—seems quite overstated. Even if tempered with a people who is able enough to further a radical break with the past while preserving some of past ideals and institutions, the very complexity of constitutionalism may contradict some of his optimistic narratives. In fact, the constitutional phenomenon should be interpreted in less normative terms. When faced with the contingencies of social life, constitutionalism, although necessary, may not be enough, especially in a country whose

89 

Thomassen (n 63) 186. Landau (n 55) 189. 91  O Varol, ‘Stealth Authoritarianism’ (2015) 100 Iowa L Rev 1673. 92  Thomassen (n 63) 179. 93  See, eg, Benhabib (n 54) 145. 94  Braver (n 1) 141. 95  ibid 156. 90 

‘Revolutionary Reform’ and the Seduction of Constitutionalism 169 former institutions are not what one could really call democratic.96 Constitutionalism is, after all, a fragile form of collective self-binding precommitment97 even in relative successful democracies, let alone in realities whose constitutional tradition falls short of providing stability and predictability.98 Where institutions have not historically worked side by side with the citizenry, where the constitution has been rather an instrument for keeping untouched privileges and extractive political and economic institutions,99 the idea of constitutionalism as an effective instrument for limiting power and preventing tyranny and private oppression100 has to be assumed in less abstract terms. Constitutionalism, in countries with a past of social oppression and rampant inequality like Venezuela, may not have the necessary strength and binding force to prevent the rise of semiauthoritarian practices, many of them abusively adopting constitutional means for their purposes.101 This is the reason why constitutionalism must be intertwined with some perspectives of economic development,102 as long as it can be—and normally is—used as an argument for preserving the status quo against any effective social and economic change. Therefore, the concept of ‘revolutionary reform’, according to which those old institutions should be shaped for the new regime, constraining somehow whoever is in power, might not be enough and, indeed, might even engender the opposite of any desirable construction of a constitutional democracy.

96  As Julie Buxton argues: ‘Certainly it is the case [to judge Bolivarianism as authoritarian] that if Chávez’s Venezuela is to be judged by the procedural benchmarks of liberal democracy, there is a deficit of checks and balances on government, the rule of law is weak, the military is not apolitical, and executive power is pronounced. But this leads to a number of related considerations. It has never been the case that liberal democracy was consolidated in Venezuela. During the Punto Fijo period, the country had a model of illiberal democracy that delimited participation, restricted access to power, privileged a minority, and politicized all state institutions. The rule of law was historically weak, and corruption and human rights abuses were pronounced. To present the Bolivarian process as some form of democratic regression or authoritarian aberration in this historical context is misleading. It denies the structural legacies of Puntofijismo and negates the progress that has been made in extending social and political inclusion in a historical context characterized by disaffection with political parties, politicians, and institutions.’ J Buxton, ‘Foreword: Venezuela’s Bolivarian Democracy’ in D Smilde and D Hellinger (eds), Venezuela’s Bolivarian Democracy (Duke University Press, 2011) XV. 97  See, eg, J Elster, ‘Don’t Burn Your Bridge Before You Come to It: Some Ambiguities and Complexities of Precommitment’ (2003) 81 Tex L Rev 1758. 98  This does not mean that economic prosperity is a necessary requirement for constitutional democracies, although it may play a relevant role. Examples of rather successful constitutionalisms in contexts of severe inequality and economic difficulties can be found in India and South Africa. The particularity of Venezuela relates above all to the fact that Chávez could control the pace and form of constitutional change in a practically unbound way and without any particular need of negotiating conflicting interests. In such a scenario, the economic background of longstanding exclusion of a large part of the population appeared as a fundamental argument for change, which could be easily embraced by a charismatic leader to sustain his own power. 99  D Acemoglu and JA Robinson, Why Nations Fail: The Origins of Power, Prosperity, and Poverty (Crown Business, 2012) 73. 100  See S Holmes, Passions and Constraint: On the Theory of Liberal Democracy (University of Chicago Press, 1995) 6. 101  Landau (n 55) 189. 102  See, eg, J González-Jácome, ‘On Abusive Constitutionalism: Two Critical Impulses’ (International Journal of Constitutional Law Blog, 11 June 2015) accessed 1 August 2016.

170  Juliano Zaiden Benvindo True, Braver carefully examines the Venezuelan history and acknowledges the dominance of an oligarchical and oppressive regime in its almost entirely past. His ‘revolutionary reform’ would, in this case, ‘start with a veneration of Bolivar and then condemn the subsequent 147 years of Venezuelan history’.103 He even says that the ‘revolutionary reform’ ‘repudiates much of Venezuelan history, including the corruption and neoliberalism of the 80s’.104 But he places great confidence in the ‘promise of the 1961 Constitution, the longest lasting and first truly democratic constitution in Venezuelan history’.105 The constitutional strength of this very past would enable a ‘revolutionary reform’ that could preserve the rules of the game in order to avoid semi-authoritarian practices. Instead of a radical revolution, there would be a renewal and acceleration of the constitutional reforms of the 1990s,106 keeping alive the spirit of 1961 Constitution to promote a more ‘representative and even participatory democracy’.107 Briefly, the ‘revolutionary reform’ demands a strong confidence in constitutionalism and in that the people could and should uphold its premises as their guarantee of a promising future. In the specific case of Venezuela, the spirit of the 1961 Constitution could have lasted, had the Supreme Court and Congress participated in the definition and discussion of the electoral rules of the Constituent Assembly. As with the past democratic constitutionalism, Braver’s ‘revolutionary reform’ also relies on the past institutions and on their capacity to control the rising constitutional transition. Normatively speaking, this is the best-case scenario. Still, constitutionalism, as powerful as it may seem, is only the tip of the iceberg in the midst of multiple contingencies of social life that can even play a greater role in such a regime transition. For instance, his main thesis anchored in the capacity of the Supreme Court and Congress to avoid such an outcome and bring about the ‘revolutionary reform’, had they imposed ex ante constraints on Chávez, seems overstated. Braver’s argument that ex ante constraints would be more efficient than ex post ones, because ex post constraints would be unfeasible in the context of Chávez’s great popularity,108 ‘constitutional boot-strapping’,109 longstanding demand for change,110 among other causes,111 is certainly reasonable. Yet, especially where institutions have historically failed to guarantee the rule of law and are seen as part of a past to be overcome, it does not follow that ex ante constraints would necessarily perform any better. The Supreme Court and Congress, after all, would also be faced with similar dilemma of confronting a highly popular president in a country in severely need of change. As with ex post constraints, they would also be in serious threat of being purged with little political cost.112 In these circumstances, both the Supreme Court and

103 

Braver (n 1) 155. ibid 151. 105 ibid. 106  ibid 155–156. 107  ibid 155. 108  ibid 149. 109  ibid 150. 110 ibid. 111 ibid. 112  ibid 149. 104 

‘Revolutionary Reform’ and the Seduction of Constitutionalism 171 Congress would have to act in a vigorous and courageous way to contest Chávez’s electoral rules with distinct remedial options that may however prove vulnerable. Braver’s confidence in the capacity of past institutions to control the pace and form of constitutional change stems thus from his belief in the ‘promise of the 1961 Constitution’113 that could have lasted as a limit to the exercise of Chávez’s hegemonic power. Beneath this argument lies a visible dichotomy between the so-called past democratic reality114 and the semi-authoritarian order in Chávez’s government. This dichotomy is almost personified in his two main theoretical sources, Hannah Arendt and Carl Schmitt. Hannah Arendt (with Bruce Ackerman) would bring, through the concept of ‘revolutionary reform’ and ‘unconventional adaptation’, the argument that could have made Venezuelan history to evolve differently. Had the Venezuelan Supreme Court and Congress better knowledge and more effective tools to set the conditions for a constitutional transition, imposing thereby constraints upon Chávez, Venezuela might have had a distinct outcome.115 On the other hand, Carl Schmitt would be to some extent the personification of those authoritarian practices that took place in that country. In this case, Braver even argues that Schmitt was a ‘decisive influence on constitutional advisors’116 to Hugo Chávez in the creation of the 1999 Constitution, as well as on the new Venezuelan Supreme Court117 and constitutional literature.118 Schmitt is, as previously argued, normally interpreted as the paramount subversion of concepts such as democracy to legitimise authoritarian practices. However, ­Braver’s emphasis on Schmitt, as an opposition to Hannah Arendt and the concepts of ‘revolutionary reform’ and ‘unconventional adaptation’, may simplify the set of negotiations and variables that are in play in such a context of regime transition. Whether Schmitt, despite his possible influence, indeed served as the theoretical basis for those authoritarian outcomes is nonetheless an open question. And if he did, this may be more accidental than Braver suggests. It is precisely in this matter that we can observe that, albeit his analysis of Venezuelan history to justify why his concept of ‘revolutionary reform’ is ‘viable, suitable and appealing’,119 his argument ultimately revolves around constitutionalism and institutions, not dwelling on more structural causes that made Chávez possible. Yet, unlike Braver’s argument,120 that old constitutional and institutional framework, even if the seeds of a democratic future were there, was also characterised by ‘structural and cultural determinants of exclusion and political illegitimacy’,121 and could not possibly constrain Chávez as Braver suggests.

113 

ibid 151.

114 ibid. 115 

ibid 139. ibid 140. 117  ibid 150. 118 ibid. 119  ibid 156. 120  ibid 151. 121  E Lander, ‘Venezuelan Social Conflict in a Global Context’ in S Ellner and M Tinker Salas (eds), Hugo Chávez and the Decline of an ‘Exceptional Democracy’ (Rowman & Littlefield Publishers, 2007) 24. 116 

172  Juliano Zaiden Benvindo In fact, if we shift focus from constitutional theory to the very contingencies of social life, rather than Schmitt, we could simply argue that Chávez was someone who experienced the very reality of social inequality and exclusion of his country and saw in the idea of bringing to the forefront the historically oppressed popular masses a genuinely reasonable cause for action.122 He nevertheless ended up doing it in an erratic way, as it not rarely occurs in circumstances of combining a charismatic leader, a society marked by extractive economic and political institutions, and weak constitutionalism. Furthermore, in an environment where distinct forms of authoritarianism have been entrenched for centuries, to abuse constitutionalism was nothing new. Instead, distinct forms of abusive constitutionalism have a longer genealogy123 in that region than even Schmitt would ever imagine and it is no wonder that Chávez may have just followed similarly passed paths, but with a much greater visibility. In this regard, Braver might have fallen into the common Manicheanism of interpreting the past Venezuelan regime as if it were quite an exception in Latin ­America.124 Although acknowledging Venezuela’s historical oppression and oligarchical dominance, he depicts Chávez as the responsible for hindering the so-desired ‘revolutionary reform’. Some continuities and ruptures—and in the very practices of authoritarianism—are evident in this transition, whose complexities demand looking at this situation without such a confidence in the strength of constitutionalism. Even though Braver introduces a very interesting analysis of Venezuelan history to coin the concept of ‘revolutionary reform’, some further relevant discussions of structural economic and social development of that region would make his concept more solid and convincing, and possibly show that ‘revolutionary reform’ may better work in the companion of also other premises and categories. IV. CONCLUSION

Braver’s paper leads us to a direct confrontation with many theoretical assumptions regarding constitutionalism. Venezuela, after all, is a perfect scenario for bringing the natural seduction of constitutionalism and its many abstractions and fictions to face reality, without overlooking, though, their performative meaning. The promises of constitutional democracy are continuously challenged by such limits, where the

122  See VM Figueroa, ‘The Bolivarian Government of Hugo Chávez: Democratic Alternative for Latin America?’ (2006) 32 Critical Sociology 187, 199 (arguing that Chávez’s Bolivarian Democracy ‘is a process seeking to combat economic and social exclusion while opening the way to an inclusive democracy’). 123  See, eg, González-Jácome (n 102). 124 Many scholars name this positive understanding of Venezuela as ‘Venezuelan exceptionalism’. According to Ellner and Salas: ‘Notions of Venezuelan exceptionalism influenced the work of political scientists and other scholars. Three basic formulations underpinned exceptionalism writing: that Venezuela was privileged with respect to the rest of Latin America, that it remained free of acute class and racial conflict and cleavages that threatened political stability elsewhere, and that its democratic system and political culture were healthy and solid.’ S Ellner and M Tinker Salas, ‘The Venezuelan Exceptionalism Thesis: Separating Myth from Reality’ in Ellner and Tinker Salas (eds) (n 121) 5.

‘Revolutionary Reform’ and the Seduction of Constitutionalism 173 questions ‘Who are the people?’, ‘What does constitutionalism mean?’ or ‘What do we understand by democracy?’ reveal their insurmountable fragility. For, in the end, even though we aim to see, in every context, those paramount categories of constitutionalism, the fact is that such categories can be only interpreted in the historically contingent process of social life.125 The ‘right to self-critique and perfectibility’126 constitutional democracy holds makes us seduced by its potentiality to bring about a new future, leading us thereby to defend it against its very risks of disruption. Braver calls our attention by bringing such a brilliant examination of how Chávez’s government disrupted many of the ‘make-beliefs’127 of constitutional democracy, but it also reveals the very limitations of constitutional theory and its categories to interpret such a complex reality. A substantive concept of a people, as Braver interprets it in Venezuela, is just the other side of the coin of a country where constitutionalism has been historically subverted and abused, where the double bind of constitutionalism and democracy is continuously put in jeopardy. While we believe in constitutionalism and its institutions (as Braver’s faith in the 1961 Constitution and the Venezuelan Supreme Court, had it better knowledge and tools), and we are normally seduced by constitutionalism, social life reveals its boundaries and fragilities. The ‘non-foundational foundations of law’,128 as a negotiation between the constative and the performative, between constitutionalism and democracy, is permeated by history, violence and faith. Braver’s paper, with its intriguing and dense analysis of such a complex constitutional context, is hence a wake-up call for understanding that constitutional democracy is only possible in its very impossibility.

125  In this regard, see Thomas Pereira contribution to this volume, who clearly points out that concepts such as popular sovereignty cannot merely be interpreted abstractly and should always make reference to the contingencies of real life. As he argues: ‘In a more concrete way, if we stick to the traditional canonical examples, the relationship between constitutional law and popular sovereignty is completely different when it comes to constitutional systems as diverse as the United States, England, and Germany. In each of these orders, both the constitutional founding and the specific institutional designs of the constituted system are different in their relationship with “the people”’; Th Pereira, ‘Constituting the Amendment Power: A Framework for Comparative Amendment Law’, in this volume. 126  J Derrida, Rogues: Two Essays on Reason (Stanford University Press, 2005) 86. 127  Morgan (n 57) 13. 128  G Teubner, ‘Societal Constitutionalism: Alternatives to State-Centred Constitutional Theory?’ in Ch Jöegers et al (eds), Constitutionalism and Transnational Governance (Hart Publishing, 2004) 16.

174 

Part II

The Traditions of Constitutional Amendment

176 

9 Constitutional Sunrise SOFIA RANCHORDÁS*

I. INTRODUCTION

C

ONSTITUTIONS ARE INTERGENERATIONAL contracts between ‘we, the people’ and our future selves. However, some constitutions also aspire to be primarily monologues rather than dialogues between the constituent power and the future generations. They burden future generations by entrenching constitutional values against future decline,1 and limit constitutional change by institutionalising complex constitutional amendment procedures and including eternity clauses.2 The legitimacy of this inter-temporal binding as well as the attempt to shape and protect future generations from themselves has been thoroughly discussed in the literature.3 Contingent constitutional change, that is, the idea that the constituent power could defer the coming into effect of certain constitutional provisions and make them contingent upon the verification of factual or legal conditions (for example, a referendum, the enactment of a law as determined

*  I would like to thank Richard Albert, Jack Balkin, David Grewal, Oran Doyle, Eric Fish, Jurgen Goossens, Nicholas Robinson, Yaniv Roznai, Mark Tushnet, and Ozan Varol, and the participants of the Boston College Law School Workshop on Constitutional Amendments for insightful comments on earlier versions of this chapter. I would also like to extend my gratitude to the Information Society Project and the Niels Stensen Fellowship. 1  M Dorf, ‘The Aspirational Constitution’ (2008) 77 Geo Wash L Rev 1631; N Barber, ‘Why Entrench’ (2016) Int’l J Const L (forthcoming). 2  Y Roznai, ‘Towards a Theory of Unamendability’ (2015) Working Paper 15-12/2015 accessed 1 August 2016. See also Y Roznai, ‘Unamendability and the Genetic Code of the Constitution’ (2015) New York University Public Law and Legal Theory Working Paper 514/2015 accessed 1 August 2016; UK Preuss, ‘The Implications of “Eternity Clauses”: The German Experience’ (2011) 44 Israel L Rev 429; A Friedman, ‘Dead Hand Constitutionalism: The Danger of Eternity Clauses in New Democracies’ (2011) 4 Mexican L Rev 77, 80. 3  AC Desai, ‘What a History of Tax Withholdings Tells US about the Relationship Between Statutes and Constitutional Law’ (2014) 108 Nw U L Rev 859; MJ Klarman, ‘Majoritarian Judicial Review: The Entrenchment Problem’ (1997) 85 Georgetown L J 492, 496; R Dixon and T Ginsburg, ‘Deciding Not to Decide: Deferral in Constitutional Design’ (2011) 9 Int’l J Const L 636, 637; see also J Balkin and R Siegel (eds), The Constitution in 2020 (Oxford University Press, 2009); J Balkin, Constitutional Redemption: Political Faith in an Unjust World (Harvard University Press, 2011).

178  Sofia Ranchordás by the Constitution) has nonetheless remained undertheorised in the literature on comparative constitutional law.4 In the past decade, a growing number of scholars has analysed the relationship between time and law, in particular in the context of constitutional law.5 The legal literature has focused on the complexities of temporary legislation, timing laws and their effects,6 sunsetting judicial opinions,7 comparing temporary and permanent legislation and the incentives behind its adoption,8 lasting legislation,9 and, in the field of constitutional law, constitutional change and amendments by desuetude,10 and temporary constitutions.11 This strand of literature has focused on why, when, and how constitutional and legal dispositions should terminate. However, little has been said about the opposite phenomenon: the commencement of constitutional dispositions and whether sunrise clauses should be allowed to make it contingent upon a future event.12 Sunrise clauses have remained a highly overlooked topic in the literature.13 In this chapter I argue that for example in post-conflict and highly divided societies and in transitional contexts,14 sunrise clauses can be important tools to guarantee that the ‘future does not arrive too early’ and that important, transitional, and aspirational matters that cannot be realistically promised to the people at that time are nevertheless included in the constitutional agenda. As this chapter explains, this was the case of the former Directive 45 of the Indian Constitution that deferred the commencement of the right to education to a later date in the hope that the country would have the financial means to grant Indian citizens free access to education. Sunrise clauses have been employed in the past century in different constitutions throughout the world so as to delay the commencement of some of their provisions 4 In the US constitutional law, sunrise clauses have been discussed primarily by Akhil Amar. See A Reed Amar, America’s Unwritten Constitution: The Precedents and Principles We Live By (Basic Books, 2012) 474-476; America’s Constitution: A Biography (Random House, 2010). See also EB Foley, ‘The Posterity Project: Developing a Method for Long-Term Political Reform’ (2013) 66 Okla L Rev 1; DE Herz-Roiphe and D Singh Grewal, ‘Make Me Democratic, But Not Yet: Sunrise Lawmaking and Democratic Constitutionalism’ (2016) 90 N Y L Rev 1975. 5 Specifically, on time and law, and the effects of ‘temporary de-juridification’, see A Kouroutakis and S Ranchordás, ‘Snoozing Democracy: Sunset Clauses, De-juridification, and Emergencies’ (2016) 25 Minn J Int’l L 29. 6  JE Gersen and E A Posner, ‘Timing Rules and Legal Institutions’ (2007) 121 Harv L Rev 543. 7  N Katyal, ‘Sunsetting Judicial Opinions’ (2004) 79 Notre Dame L Rev 1237. 8 See F Fagan, Law and The Limits of Government: Temporary Versus Permanent Legislation (Edward Elgar, 2013). 9 For a critical approach to temporary legislation see RM Kysar, ‘Lasting Legislation’ (2011) 159 U Pa L Rev 1007; RM Kysar, ‘The Sun also Rises: The Political Economy of Sunset Provisions in the Tax Code’ (2006) 40 (2) Ga L Rev 335. 10 See R Albert, ‘Constitutional Amendment by Constitutional Desuetude’ (2014) 62 Am J Comp L 641. 11  O Varol, ‘Temporary Constitutions’ (2014) 102 Cal L Rev 409. 12  On commencement clauses, see M Gobbi, ‘When to Begin: A Study of New Zealand Commencement Clauses with Regard to those Used in the United Kingdom, Australia, and the United States’ (2010), 31 (3) Stat L Rev 153. 13  In the context of Constitutional Law, see Amar (n 4) 474–76. 14 S Choudhry (ed), Constitutional Design for Divided Societies: Integration or Accommodation? (Oxford University Press, 2008). See also the special 2013 issue of the International Journal of Constitutional Law on Constitutional Transition (vol 11(3)); see S Choudhry, ‘Constitutional Transitions in the Middle East: Introduction’, 11 (3) Int’l J Const L 611.

Constitutional Sunrise 179 to a certain date (automatic sunrise clauses) or make it contingent upon a certain event (contingent sunrise clauses). To illustrate, the Constitution of Georgia of 1995 delayed the coming into effect of some constitutional provisions to a later date. While the Constitution was due to enter into force ‘upon the date of recognizing the powers of the newly elected President and the Parliament of Georgia’ (Article 104[1], the commencement of Articles 49(1) [composition of Parliament] and 58(1) ­[formation of Parliamentary factions] was made dependent upon the enactment of ‘appropriate amendments and additions made to the organic law Election Code of Georgia’ (Article 104–1[1]).15 As this chapter explains, this type of provisions may not abound but it is far from uncommon.16 Although sunrise clauses are relatively unknown in the literature on comparative constitutional law, the idea of delaying the coming into effect of legal provisions or making it contingent upon a determined event is present in different fields of law and it can even be considered to be intrinsic to a legal system. When explicating the use of conditions in contracts, Arthur Corbin reminded us that ‘legal relations are merely mental concepts which are useful in enabling us to foresee the p ­ hysical facts of the future’.17 Along with interpretation, the inclusion of conditions in c­ ontractual relations also promote the foreseeability of future facts since they allow a legal relation to start before all the contractual conditions (for example, payment) are fulfilled. As this chapter explains, sunrise clauses can play a similar role. The explicitly forward-looking nature of sunrise clauses in addition to the imposition of vague conditions for the coming into force of a constitutional provision might nonetheless raise important questions as to their constitutionality and the limits of the constituent power. This chapter makes two contributions to the literature. First, it explores the meaning and functions of sunrise clauses as instruments of constitutional change. Secondly, it discusses their constitutionality in light of two judicial cases from South Africa and Ireland where this question was raised. This chapter relies on the literature on comparative constitutional law and offers a number of examples of different types of sunrise clauses included in different constitutions. While the Constitutions selected for this chapter are not always immediately comparable, these examples help operationalise the concept of ‘sunrise clause’ and show the instrumental diversity of contingent constitutional change.18 This chapter proceeds in three parts. In Section II I define and analyse the concept of ‘sunrise clause’ and provide an overview of different types of constitutional sunrise clauses. I refer not only to the use of sunrise clauses in the constitutional context but also in legislation and long-term contracts, where these provisions appear to

15 According to Art 107(2) of the Constitution of Georgia, Arts 18(2) and (3) of the Constitution would also only enter into force after the respective criminal procedural legislative acts are adopted. 16  Another example is Art 148 of the Constitution of Tunisia (2014) that also delays the commencement of multiple dispositions, making them for example contingent upon ‘the definitive results of the first general elections’, the direct election of the President, the election of the Assembly of Representatives, and the enactment of a number of operationalising statutes. 17  A Corbin, ‘Conditions in the Law of Contract’ (1919) 28 Yale L J 731. 18  For a reflection on the methodology of comparative constitutional law, see R Hirschl, ‘The Question of Case Selection in Comparative Constitutional Law’ (2005) 53 Am J Comp L 125.

180  Sofia Ranchordás be more common. This section offers a broad definition of sunrise clauses in order to encompass a number of sunrise variations, including automatic and contingent sunrise clauses. I also categorise sunrise clauses according to the type of contingency employed to determine their commencement. In this section, I also distinguish between sunrise clauses and by-law clauses. In Section III, I explore the analogy of the Constitution to an intergenerational contract, which is required to achieve a compromise between short-term and longterm commitments by adopting either a backward- or forward-looking approach to constitutional change.19 I suggest that sunrise clauses should be employed to operationalise this forward-looking perspective and consider future contingencies that given the history of the country can be predicted by the constituent power. In Section IV, I analyse a judicial decision of the Irish Supreme Court where the constitutionality of sunrise clauses was challenged. In light of this case, I argue that sunrise clauses are valuable instruments of constitutional change and ‘self-­ amendment’, which find their legitimacy directly in the constituent power. The constitutional intergenerational dialogue does not imply that all provisions should enter into effect at the same time. Rather, the constitutional will formulated in some provisions may remain dormant, ‘sunrising’ at a later date or upon the verification of a certain condition. II.  SUNRISE CLAUSES

Sunrise clauses can be broadly defined as dispositions providing that a constitutional provision only comes into force on a specific date or that its coming into effect is contingent upon the verification of specific conditions.20 Sunrise clauses are a form of legal condition which have been employed in constitutions, primary and ­secondary legislation, and international treaties.21 Sunrise clauses are forward-looking instruments which determine that a provision shall remain dormant for a certain period. Their rationale is determined by the need to take into account conditions that are not yet verified at that time. Although contingent constitutionalism has been studied in the United States to explain the interdependence between constitutional rights and state law,22 the ­concepts of contingent constitutional change and sunrise clauses as instruments of 19 

See A Sajó, ‘Preferred Generations: A Paradox of Restoration’ (1993) 14 Cardozo L Rev 847. See M Freeman, Necessary Evils: Amnesties and The Search for Justice (Cambridge University Press, 2009) 142. 21  For a historical perspective on the use of contingent legislation and contingent provisions in interstate commerce, see JP Comer, Legislative Functions of National Administrative Authorities (Columbia University Press, 1927) 63–70. For some types of treaties, such as humanitarian law conventions (eg, Geneva Convention of 1949), the effective date of the withdrawal from a treaty can be made contingent upon external events, such as the cessation of an existing armed conflict, see LR Helfer, ‘Terminating Treaties’ in D Hollis (ed), The Oxford Guide to Treaties (Oxford University Press, 2014) 634, 642. 22  See WA Logan, ‘Contingent Constitutionalism: State and Local Criminal Laws and the Applicability of Federal Constitutional Rights’ (2009) 51 Wm & Mary L Rev 143; M Manheimmer, ‘The Contingent Fourth Amendment’ (2015) 64 Emory L J 1229. 20 

Constitutional Sunrise 181 constitutional change have remained underexplored.23 Contingent legal instruments have been historically employed in the legislative and constitutional contexts not only in the United States but also in other common law (for example, India, Ireland) and, to a smaller extent, civil law countries (for example, Belgium). In this section, I start by providing an overview of the use of different types of sunrise clauses in ­legislation. I then distinguish between automatic and contingent constitutional sunrise clauses, and explain the difference between these provisions and by-laws. A sunrise clause is a constitutional disposition that determines that a part of that constitution will only come into force after a specific date in the future or its commencement will be contingent upon the verification of specific conditions.24 Sunrise clauses bear thus a strong resemblance to contingent legislation and tie-barring provisions since these three legislative instruments share a common feature: the coming into effect of certain provisions is delayed and it is conditional. Until a future event occurs, the dispositions lie dormant. Although sunrise clauses have not been widely discussed in the comparative constitutional law literature, this does not mean that they have never been analysed by prominent scholars25 or considered by constitutional drafters as an option for accommodating constitutional change. In the literature, Akhil Amar recently suggested that the framers should have adopted this approach for issues such as secession and slavery.26 They did not and thus erred, surrendering the future, ‘giving slavocrats extra political clout in every election in perpetuity’. Amar argues: just as the Constitution allowed the transatlantic slave trade to continue for twenty years but provided that Congress could ban this odious traffic in 1808 and thereafter, so the document should have allowed slave states to get extra credit in the House and electoral college (…) but not thereafter.27

Since in 1789 slavery was already perceived by many as being ‘morally wrong’, Amar contends that the framers should also then have agreed that after 1808 antislavery rules would sunrise.28 While the proposition can only remain at the theoretical level in the case of slavery, the use of sunrise clauses can have a more practical dimension in other issues, such as that of self-determination or socioeconomic rights. From a normative perspective, the promise of the sunrise of the appreciation of selfdetermination rights on a specific date or upon the verification of specific conditions may guarantee enhanced consensus, avoid the deferral of important decisions with constitutional value to the ever-changing legislature, and complement sunset clauses by providing a timed follow-up. 23  As this chapter explains below, in the United States, Akhil Amar has regretted the limited use of sunrise clauses in the Constitution, see Amar, America’s Unwritten Constitution (n 4) 474–76; A Reed Amar and S Levinson, ‘What Do We Talk About When We Talk About the Constitution?’ (2013) 91 Tex L Rev 1119, 1143. 24  See M Freeman, Necessary Evils: Amnesties and The Search for Justice (Cambridge University Press, 2009) 142. 25  See Amar (n 4) at 474–76. 26  ibid 474. 27  Amar and Levinson (n 23) 1143. 28 ibid.

182  Sofia Ranchordás A.  Automatic and Contingent Sunrise Clauses Sunrise clauses can be automatic when their commencement is determined by objective events, that is, only the effective date of a constitutional provision is delayed. In a certain sense, automatic sunrise clauses bear a resemblance to the so-called ­commencement clauses, which define an exceptional date of coming into effect in order to avoid the retroactive application of a new law which would occur if general rules would be applied.29 Contingent sunrise clauses determine that the entering into force of a provision is dependent on a subset of factual or legal conditions. The commencement of a clause can therefore be ‘tied’ to a legal action which requires the intervention of the Parliament or the Executive. This is the case of the enactment of a statute or the results of a referendum organised at a later date (for example, a sunrise clause on the future secession of a province or state). i.  Automatic Sunrise Clauses Automatic sunrise clauses merely delay the commencement of a constitutional provision. This form of sunrise clause determines when a provision constitution should begin in order to avoid its premature coming into effect. In addition, its main goal is to avoid immediate commencement according to general rules on publication and commencement of laws or delay the commencement of a law for a longer period until legal or administrative problems regarding the new act are solved. The difference between this first type of sunrise clause and other forms of sunrise clauses can be at times almost imperceptible as the latter can also grant some flexibility regarding the date of coming into effect of a law.30 The correct timing of the coming into effect of a constitutional provision is relevant since it may determine the success of the operationalisation of the constitutional change. Within this category of sunrise clauses, we can also find sunrise clauses which might be associated with sunset clauses in two different ways. First, we may find constitutional dispositions that sunrise naturally, for example, after the termination of an interim constitution or set of dispositions. Secondly, some clauses may also have a contingent sunrise, if they do not commence within a certain period, the provisions expire. An example of the latter can be found in Section 2 of the Twenty-second Amendment of the US Constitution,31 which reads: This article shall be inoperative unless it shall have been ratified as an amendment to the Constitution by the legislatures of three-fourths of the several states within seven years from the date of its submission to the states by the Congress.

29  M Gobi, ‘When to Begin: A Study of New Zealand Commencement Clauses with Regard to those Used in the United Kingdom, Australia, and the United States’ (2010) 31(3) Stat L Rev 153. 30  ibid 153, 186. 31  See The Constitution of the United States, Amendment XXII. This type of sunrise/sunset clause is also present in other amendments of the Constitution of the United States, see, for example, s 3 of the Twenty-first Amendment which repealed the Eighteenth Amendment.

Constitutional Sunrise 183 ii.  Contingent Sunrise Clauses Contrary to automatic sunrise clauses, contingent sunrise clauses determine that the coming into effect of a provision or constitutional amendment is dependent upon an objective event such as the ratification of a number of states, a subjective condition (for example, in the case of aspirational contingencies as explained in Section III) or the adoption of operationalising legislation. An example of the first type of contingency can be identified in Article XIV of the Constitution of Maryland which determines that: ‘any Constitution, or change, or amendment of the existing Constitution, which may be adopted by such Convention, shall be submitted to the voters of this State, and shall have no effect unless the same shall have been adopted by a majority of the voters voting thereon’. The establishment of dependency relations between dispositions have been present both on legislative and constitutional levels. The practice of making the coming into effect of one law dependent on the enactment of another one has been called tie-barring at the legislative level and has raised a number of legal issues. A similar rationale has been applied on the constitutional level with contingent sunrise clauses. Article 35 of the Belgian Constitution is an example of this type of sunrise clauses, since the transitional disposition included in this article determines that the constitutional provision on communal and regional powers will only enter into effect after a statute listing the exclusive powers of federation is enacted. Article 35 of the Belgian Constitution reads: The federal authority only has competences in the matters that are formally assigned to it by the Constitution and the laws passed by virtue of the Constitution itself. The Communities and the Regions, each in its own field of concern, have competences for the other matters, under the conditions and in the terms ­stipulated by the law. This law must be adopted by a majority as described in Article 4, last paragraph. Transitional provision: The law referred to in the second paragraph determines the date on which this Article comes into force. This date cannot precede the date of the entry into force of the new Article to be inserted in Title III of the Constitution, which determines the competences exclusive to the federal authority. (emphasis added)

Since the subject underlying Article 35 has proven to be highly controversial in ­Belgium, this constitutional provision continues to lie dormant at the time of writing. Contingent sunrise clauses can become problematic when constitutional s­ unrise clauses imply discretionary decisions of the Executive on highly subjective concepts employed as contingencies (for example, the gradual concretisation of rights). In a number of developing countries, the right to education, for example, is presented as an aspirational right or made contingent upon ‘the economic capacity and development’ of the country (for example, Article 41 of the Indian Constitution). One of the challenges of contingent sunrise clauses refers to the identification of the correct condition or the risk that the contingency is too vague to be ever triggered. Sunrise clauses that use vague language and subjective conditions for example in the context of aspirational constitutionalism may therefore be devoid of meaning.

184  Sofia Ranchordás B.  Sunrise Clauses and the Use of Conditions Although sunrise clauses are underanalysed instruments, the use of conditions is inherent to law and it should hence not be overlooked in the context of constitutional law. In this section, I support this claim by analogising constitutions with long-term contracts which often require condition-precedent and condition-subject clauses to meet the concerns of the parties regarding future changes.32 I draw upon a 1996 decision of the South African Constitutional Court which examined the validity of the sunrise clauses included in the KwaZulu-Natal constitution.33 A constitution is a long-term contract which, according to empirical research, typically lasts 19 years,34 which might be shorter than many of our financial and personal contractual obligations. In most long-term contracts, the central contractual object is deemed to endure, but, in order to fit changing circumstances and accommodate foreseeable issues, the parties may agree to include condition-subsequent clauses, in common law, or the so-called ‘resolutory clauses’ in civil law.35 These clauses update and shape the original content of the contract, by determining the termination of certain contractual elements. In the concrete case of constitutions, sunset clauses determine the termination of obsolete constitutional dispositions or guarantee smooth constitutional transitions, for example from dictatorial to democratic regimes. A much less common instrument in the constitutional context is, what in contractual terms one would call ‘condition-precedent’ clauses (or ‘suspensive clauses’ in civil law). These clauses determine that a certain disposition will be inactive until a specific date or its coming into effect will be contingent upon a set of facts or upon the verification of a specific condition.36 These ‘sunrise clauses’ are a form of ‘contingency law’ since they establish that the effects of one disposition will be dependent on the verification of certain conditions. In 1996, the South African Constitutional Court was asked to validate a provincial constitution containing a number of sunrise clauses. In March 1996, the IFP37 announced that the provincial constitution would include a sunrise clause reflecting the provincial powers this party wished to secure for the province of ­KwaZulu-Natal.38 These provisions would allegedly guarantee that this province would b ­ enefit from

32 The contractual terminology was also used by the Irish Supreme Court in the case analysed in ­Section IV. 33  Constitutional Court of South Africa, Certification of the Constitution of the Province of KwaZulu, Case CCT 15/96, 6 September 1996. 34  Z Elkins, T Ginsburg and J Melton, The Endurance of National Constitutions (Cambridge ­University Press, 2009). 35 For the clarification of this terminology, see, eg, B Fauvarque, ‘Position Paper on Conditions, ­Unidroit, Working Group for the Preparation of the Principles of International Commercial Contracts’, UNIDROIT (2007) 10, accessed 1 August 2016. 36  This analogy was also made by the Constitutional Court of South Africa in the assessment of the constitutionality of the provincial constitution of KwaZulu-Natal in 1995 (see below for more i­nformation on this example). 37  Inkatha Freedom Party, still nowadays one of the largest political parties in South Africa. 38  O Malley, ‘Increased Repression and Continued Attack in 1995’ accessed 1 August 2016.

Constitutional Sunrise 185 a higher degree of autonomy after and if the definitive national constitution would come into effect.39 Until then, a number of sections would lie dormant, waiting for that automatic trigger—the enactment of the new ­Constitution—to be pulled. The South African Constitutional Court did not validate this provincial constitution, since these clauses violated the Interim Constitution valid at that time. The Court considered that the provincial constitution was fundamentally flawed on three grounds: usurpation of national powers,40 consistency clauses and ‘­suspensive conditions’.41 The Constitutional Court considered that the ‘suspensive clauses’ could not be qualified as unwritten or inexistent just because they would commence later and were dependent upon a certain condition. Comparing these clauses to contractual ones, the Court argued that: it is well established that in the field of contract an agreement subject to a suspensive condition is already a binding agreement, that its terms are clearly established and that, for example, a provisional creditor may, even before the condition precedent has been fulfilled, institute proceedings to protect such creditor’s provisional right.1[9] But what is clear is that merely to suspend part of the text of a provincial constitution that is inconsistent with the interim Constitution, cannot save the constitution from the consequence of such inconsistency.

The Court added that: A suspended provision is part of the text, and it does not cease to be such simply because its operation is suspended until a future date, or is made contingent upon the happening of a future event. The text of the provincial Constitution is to be evaluated and certified as an integrated whole, because the meaning and effect of one particular clause can be crucially dependent on that of another. If certain clauses of the text come into operation after others, then the fact that certain clauses are inoperative for a period of time may well influence the

39  The negotiation of constitutional principles was particularly complex in the post-apartheid period when the different parties involved diverged as to the objectives and structure of the transition. A problematic case was KwaZulu-Natal, where there was significant resistance to the national government. The Zulu king was at the time struggling to demonstrate his authority and reject central government and its influence. The local legislature began working on a provincial constitution design to maintain a kingdom within the South African republic, establish a constitutional monarchy and guarantee independent control over security forces in the province. See Africa Confidential, vol 37, 5 July 1996, p 64 (‘a series of provincial powers would remain inactive until the national constitution came into force’). According to s 160(1) of the 1993 Constitution of South Africa (the ‘Interim Constitution’), the provincial legislatures were entitled to pass a constitution for their provinces. The provincial constitution had to be certified by the then recently established Constitutional Court of South Africa that had the task of guaranteeing that none of its provisions was inconsistent with the Interim Constitution. In order to pass the test of the Constitutional Court, KwaZulu-Natal decided an ingenious plan which included making the coming into effect of most controversial constitutional provisions contingent upon the enactment of the new and permanent South African Constitution. See Th Eloff, ‘Managing Negotiations: Lessons and Pitfalls 1990–1994’, in B de Villiers and J Sindane (eds), Managing Constitutional Change (HSRC Publishers, 1996) 1, 8. 40  Constitutional Court (South Africa), Certification of the Kwazulu-Natal Constitution (CCT15/96) [1996] ZACC 17; 1996 (11) BCLR 1419; 1996 (4) SA 1098 (6 September 1996) [14]. 41  South Africa is a bi-juridical or a mixed system of law with Roman Dutch civil law and common law, which explains the civil law terminology here.

186  Sofia Ranchordás effect and meaning of those parts of the text which do come into operation immediately upon certification in the absence of the suspended clauses.42

In this case, the inclusion of sunrise clauses by itself was not contested. Rather, due to the lack of other rules to assess its validity, the South African Court solved the controversy by referring to the common use of conditions in contract law and interpreting these sunrise clauses according to the analogy, that is, in light of the ‘entire contract’. In this specific case, more was at stake than an inter-temporal problem. The sunrise clauses under analysis were a disguised attempt to circumvent the Interim Constitution, making the coming into effect of the provincial powers contingent upon the enactment of the definitive constitution, which according to the predictions of that time, would be potentially more favourable to the self-determination of the province than the former. C.  Sunrise Clauses and By-Laws Constitutional dispositions often delegate constitutional decisions to Parliament (‘by-law clauses’) or determine that a certain disposition shall be regulated as ‘determined by law’.43 While the enactment of such ‘law’ may hinder the practical concretisation of a constitutional right, the coming into effect of these dispositions is not at stake. As this chapter explains, sunrise clauses are only effective upon the verification of a certain condition and until that day arrives, they lie dormant. By-law clauses do not delay constitutional decision-making to the future and in particular to future legislatures.44 Instead of a deferral, by-law clauses operationalise delegation. They decide who should have the power to legislate on a certain matter and determine that the legislature is the most competent body to do so. Tom Ginsburg and Rosalind Dixon have argued that ‘by-law clauses’ can avoid the ambiguity of constitutional silences, minimise design costs and error costs in the constitutionalmaking process.45 These clauses can either ‘require the legislature to decide certain constitutional issues in the future, or else explicitly empower the legislature to decide such issues’.46 Ginsburg and Dixon contend that by-law clauses consist in explicit delegations to future decision-makers to decide on a specific constitutional question. However, once we look closely, by-law clauses do not ‘defer the right to decide’ to the legislature, as the title of Ginsburg and Dixon’s work suggests. Rather, they identify actions that must be taken by the legislature through legislation, rather than by the Executive. They delegate the right to decide to the legislature. Contrary to by-law clauses, sunrise clauses make a constitutional decision and defer the coming into effect of the provision. In the case of contingent sunrise clauses, the verification of a condition might rest upon the legislature or the executive government, but also

42  Constitutional Court (South Africa), Certification of the Kwazulu-Natal Constitution (CCT15/96) [1996] ZACC 17; 1996 (11) BCLR 1419; 1996 (4) SA 1098 (6 September 1996) [39] [41]. 43  See Dixon and Ginsburg (n 3). 44  On by-law clauses, see Dixon and Ginsburg (n 3). 45  ibid 639. 46 ibid.

Constitutional Sunrise 187 here, the central constitutional decision has been previously made by the constituent power. III.  CONTINGENT CONSTITUTIONAL CHANGE BETWEEN RETROSPECTIVE AND FORESIGHT

As democracies change over time, either as a result of a succession of gradual or disruptive events, so do constitutions.47 Constitutional change was traditionally thought to operate exclusively according to formal amendment procedures and to be limited by supermajoritarian requirements.48 More recently, the less institutionalised paths of constitutional amendment and the problem of ‘unconstitutional constitutional amendments’ initiated an intense debate in the literature.49 The idea of contingent or delayed constitutional change has nonetheless remained overlooked in the comparative legal literature. Constitutional change is a problem for the legal enforcement of constitutions enacted in the wake of conflicts or other unsettling events.50 On the one hand, constitutional change is antithetic to the most fundamental desire to guarantee constitutional stability and avoid the high costs of constitutional instability at the political, economic and social levels.51 On the other hand, constitutions do not live as long as we would like them to, they ‘do not age well’, and tend to obsolesce ­rapidly.52 ­Unsurprisingly, constitutions are often amended by desuetude, that is, obsolete ­dispositions are silently amended because they are no longer implemented.53 Obsolete dispositions create nonetheless an undesirable feeling of uncertainty, which could open the door to abusive litigation.54 In addition, while there is the 47 For a historical perspective on constitutional and legislative change, see M Schwartzberg, ­ emocracy and Legal Change (Cambridge University Press, 2009); see also B Friedman and S B Smith, D ‘The ­Sedimentary Constitution’ (1998) 147 U Pa L Rev 20. 48  For an overview and comparison of the majority requirements for constitutional amendments in a number of countries, see X Contiades (eds), Engineering Constitutional Change: A Comparative Perspective on Europe, Canada and the USA (Routledge, 2013). 49  On other paths to amend the Constitution beyond Art V of the Constitution of the United States, see generally B Ackerman, We the People: Foundations (Harvard University Press, 1993); more specifically, on constitutional amendment procedures and rules, see R Albert, ‘Amending Constitutional Amendment Rules’ (2015) 13(3) Int’l J Const L 655; ‘Constitutional Amendment by Constitutional Desuetude’ (2014) 62 Am J Comp L 641. There is a vast literature on unconstitutional constitutional amendments, see, eg, A Barak, ‘Unconstitutional Constitutional Amendments’ (2011) 44 Israel L Rev 321; J Mazzone, ‘Unamendments’ (2005) 90 Iowa L Rev 1746; R Albert, ‘Nonconstitutional Amendments’ (2009) 22(1) Canadian Journal of Law & Jurisprudence 5; C Bernal, ‘Unconstitutional constitutional amendments in the case study of Colombia: An analysis of the justification and meaning of the constitutional replacement doctrine’ (2013) 11 Int’l J Const L 339. 50  SM Griffith, ‘The Problem of Constitutional Change’ (1996) 70 Tul L Rev 2121. 51  Ph A Joseph and G R Walker, ‘A Theory of Constitutional Change’ (1987) 7(2) Oxford Journal of Legal Studies 155, 175. 52  A Vermeule, ‘Constitutional Amendments and the Constitutional Common Law’ in RW Bauman and T Kahana (eds), The Least Examined Branch: Legislatures in the Modern State (Cambridge University Press, 2007) 229; On the ‘life expectancy’ of constitutions, see Elkins, Ginsburg and Melton (n 34). 53  See R Albert, ‘Constitutional Amendment by Constitutional Desuetude’ (2014) 62 Am J Comp L 641. 54  See MJ Mitchell, ‘Cleaning up the Closet: Using Sunset Provisions to Clean up Cluttered Criminal Codes’ (2005) 54 Emory L J 1671.

188  Sofia Ranchordás traditional belief that constitutional longevity and stability is particularly important for countries emerging from post-conflicts,55 there seem to be exceptions to the rule, proving that short-lived constitutions (such as the 1997 Thai Constitution) can also produce an important impact in the constitutional future of the country.56 Striking the balance between stability and constitutional change implicates firstly an analysis of the relationship between constitutions and time, secondly, the assessment of the instruments employed in national law to confer a certain degree of stability to constitutions without impeding change; and thirdly, allowing constitutions to take into consideration unpredictable or future circumstances. It is in this last context that the idea of contingent constitutional change as operationalised by sunrise clauses emerges. Constitutions are compromises between the political forces present at the time of the drafting, the desire to shape posterity, and potential contingencies that may occur in the meanwhile. The evolution of democracy might be highly dependent on present and future contingencies.57 These contingencies are not only interpreted in the sense of ‘emergencies’ but rather more broadly as ‘future and sometimes unpredictable circumstances’, which have not been analysed in the context of the constitutional change. The hermeneutics of constitutional change require not only an analysis of past and present facts, but also a predictive theory of constitutionalism, which can tell us ‘what is occurring or, better, what will occur’.58 Traditional constitutionalism is often backward-looking placing considerable emphasis on the drafting moment, expecting the Constitution to become entrenched as it is.59 However, particularly in the contexts of transition, the Constitution should also be allowed to play a predictive role, allowing the constitutional project to deal with potential legitimacy deficits in a more gradual way.60 A theory of constitutional change must thus encompass a backward-looking and a forward-looking approach to this intergenerational approach. In the next subsection I explain why timing clauses such as sunset and sunrise provisions can promote a forward-looking approach to constitutional change. A forward-looking approach to constitutional change can have multiple expressions. Constitutions can use sunset clauses to terminate constitutional provisions or interim constitutions; promote the openness of the constitutional text, for

55 See J Hatchard, ‘Establishing Popular and Durable National Constitutions in Commonwealth Africa’, in M Andenas (ed), The Creation and Amendment of Constitutional Norms (The British Institute of International and Comparative Law, 2000) 1. 56  See T Ginsburg, ‘Constitutional Afterlife: The Continuing Impact of Thailand’s Postpolitical Constitution’ (2009) 7(1) Int’l J Const L 83: Ginsburg describes the context in which the 1997 Interim Thai Constitution was adopted and examines how some of its institutional innovations (such as the introduction of judicial review) were passed on to the Constitution that superseded it. 57  See A Przeworski, ‘Democracy as a Contingent Outcome of Conflicts’ in J Elster and R Slagstad (eds), Constitutionalism and Democracy (Cambridge University Press, 1988) 64–66. 58  Ph A Joseph and GR Walker, ‘A Theory of Constitutional Change’ (1987) 7(2) Oxford Journal of Legal Studies 155, 171. 59 C Dupré and J Yeh, ‘Constitutions and Legitimacy over Time’, in M Tushnet, Th Fleiner and Ch Saunders, Routledge Handbook of Constitutional Law (Routledge, 2013) 45, 52. 60  Dupré and Yeh (n 59) 45, 53.

Constitutional Sunrise 189 example, by employing vague constitutional language,61 indeterminate concepts, ­constitutional silences,62 or as mentioned in Section II by deferring further decisions to the legislator; or by including sunrise clauses. A.  Sunset Clauses Temporality, by determining the automatic termination of constitutions at the end of a fixed period, is meant to be an expression of a forward-looking approach to constitutional change. Thomas Jefferson famously argued in 1789 that: ‘No ­society can make a perpetual constitution, or even a perpetual law. The earth belongs always to the living generation.’63 Jefferson pleaded for temporary constitutions that would reign over one generation and then expire automatically. Sunset clauses can connect our idea of justice with that of ‘our posterity’ by acknowledging that some constitutional dispositions that respond to specific concerns of our generation (for example, ensuring transition from an authoritarian regime to democracy) should not last beyond ‘ourselves’. For example, in January 1776, New Hampshire adopted a simplistic Constitution which might not have been intended to endure but rather to timely reject the colonial charters under which it had been governed.64 In the 1990s, Hungary and South Africa drafted Interim constitutions to respond to transitional problems, and more recently, Thailand also adopted a temporary constitution.65 Sunset clauses66 are dispositions that determine the termination of a law, constitution or parts of it within a beforehand determined period.67 These provisions are

61  On constitutional language, see, eg, F Schauer, ‘An Essay on Constitutional Language’ (1981) 29 UCLA L Rev 803. On constitutional indeterminacy in the Constitution of the United States, see M Klarman, ‘Fidelity, Indeterminacy, and the Problem of Constitutional Evil’ (1997) 65 Fordham L Rev 1739. 62 Constitutional silences are not necessarily cartes blanches for Parliament to legislate on constitutional matters not listed in the Constitution. See DN Hoffman, Our Elusive Constitution: Silences, ­Paradoxes, Priorities (SUNY Press 1997); L Tribe, ‘Toward a Syntax of the Unsaid: Construing the Sounds of Congressional and Constitutional Silence’ (1982) 57 Ind L J 524. 63  S Straub ‘Th Jefferson, Letter to James Madison, September 6, 1789’ (The Federalist Papers Project) , accessed 19 August 2013: this letter referred to the possibility ‘of the representatives of a nation validly engage debts beyond what they may pay in their own time, that is to say, within thirty-four years of the date of the engagement’. 64 J Dinan, ‘American Constitutions: Natural Law and Constitution-Making in the Founding Era’ (American Founding and Constitutionalism) accessed 1 August 2016. 65  For a thorough analysis of temporary constitutions, see Varol (n 11) 409. 66  It is important to distinguish between ‘sunset clauses’ and ‘sunrise clauses’: While sunset clauses determine the termination of a law or some of its dispositions, sunrise clauses, on the contrary, only determine that a law will come into effect later on a certain date. Until that period, the clause will be inactive. See M Freeman, Necessary Evils: Amnesties and the Search for Justice (Cambridge University Press, 2009) 142. 67 On the definition of ‘sunset clause’, see Parliament of the United Kingdom, Glossary, [‘sunset clause’] accessed 1 August 2016. (‘A provision in a Bill that gives it an “expiry date” once it is passed into law. “Sunset clauses” are included in legislation when it is felt that Parliament should have the chance to decide on its merits again after a fixed period.’)

190  Sofia Ranchordás conceived to automatically ‘erase’ a provision or a constitution which is no longer necessary either because it has fulfilled its function or because it is no longer effective. Although there are multiple examples of temporary constitutions (for example, interim constitutions in South Africa, Hungary, Thailand) and temporary constitutional dispositions (for example, state constitutions in the United States), Jefferson’s plea for temporary constitutions was not widely embraced in most countries, where the idea of a temporary constitution appears to be contrary to the dominant view of enduring constitutions. Temporary constitutions seem to have served primarily transitional purposes and not generational ones. Although Jefferson considered temporality as a form of unchaining the future generations from past commitments, sunset clauses also constrain the constituent power, by determining when certain clauses should automatically ‘sunset’ without relying on the future generations’ judgement on whether the disposition should be maintained or not.68 B.  Sunrise Clauses and Aspirational Constitutionalism A forward-looking approach to constitutional change includes an explicit acknowledgement of the need to construct the constitutional project gradually. Constitutions might confer the power to operate this gradualism to legislators (for example, ­Article 242 of the Constitution of Angola) or predict these contingencies in their texts, delaying the coming into effect of certain constitutional provisions. An example of the latter has emerged in the context of aspirational provisions. In post-conflict societies and developing countries, the drafters may resort to sunrise clauses to allow certain rights to sunrise when the economic and political conditions are more auspicious. Even when such sunrise is timed, the sunrise of rights or provisions containing vague language may however imply a great amount of discretion. This was the case of Directive 45 of the Indian Constitution, which read, in the first version of the Constitution:69 Provision for free and compulsory education for children: The State shall endeavor to provide, within a period of ten years from the commencement of this Constitution, for free and compulsory education for all children until they complete the age of fourteen years.

The ‘endeavours’ of this country took however much longer than the initial ten-year period and the right to education was only to sunrise many decades later, with the 2002 constitutional amendment. The potential late sunrise of provisions is thus a risk present in vaguely formulated clauses that in this case might not have given the State sufficient incentives to ‘pull the trigger’ and initiate the sunrise of the provision. Aspirational constitutionalism is a clear form of contingent constitutionalism which tries to bridge the wishes of past and future generations, as well as the different social, economic, and political circumstances they may be confronted with. Aspirational constitutionalism refers to ‘a process of constitution building (…) in 68  Elkins, Ginsburg and Melton (n 34) 13, (‘a sunset clause arguably constrains future generations as much, and perhaps more, than does presumed permanence. Why should citizens artificially abandon a well-functioning constitution simply because it has reached a certain age?’). 69  India amended its Constitution in 2002 to include a right to education.

Constitutional Sunrise 191 which constitutional decision makers understand what they are doing in terms of goals that they want to achieve and aspirations that they want to live up to’.70 These aspirations can be present in the preambles of constitutions, in the choice of institutional design and arrangement of lists of rights.71 Sunrise clauses can also concretise such aspirational goals. This forward-looking approach is often contingent upon economic, social, and political circumstances of a country. The Constitution of South Africa includes a number of forward-looking and aspirational clauses that translate the different gradation of this perspective. This is the case of Article 29: ‘(1) Everyone has the right— (a) to a basic education, including adult basic education; and (b) to further education, which the state, through reasonable measures, must make progressively available and accessible.’ In South Africa, sunset and sunrise clauses appear to have played an important role in conflict resolution both in constitutional and lawmaking areas. These clauses embody the acknowledgement that not ‘all aspirations can be immediately met without destroying the demands made by adversaries and these aspirations might not also be met at the same time or in short-term’.72 In the post-apartheid context, the gap between rich and poor was trying to be solved by the inclusion of numerous socioeconomic rights in the Constitution such as the right of access to education, housing, health care. Although these rights could not be immediately met, using sunset and sunrise clauses ‘could ensure that [such] aspirations do not fall off the national agenda’.73 Sunrise clauses formalise this forward-looking approach by predicting future contingencies (for example, a referendum, enactment of a statute). The inclusion of these dispositions does not make the constitutional project incomplete but in permanent dialogue with the future. Both sunset and sunrise clauses impose this dialogue. Sunrise clauses address however a different side of inter-temporality: They reflect the acknowledgement of the need to decide a constitutional matter by determining determine that society is not yet ready to embrace it or might only be, when and if certain conditions are satisfied. Therefore, the past generation attempts to make predictions about the future generation wishes, the same way sunset clauses assume that our future selves might not wish to be bound by the exceptional powers contained in an interim constitution. IV.  THE CONSTITUTIONALITY OF SUNRISE CLAUSES

This section starts with the discussion of an Irish case involving a less visible form of contingent constitutional change but which, at the resemblance of the mentioned South African decision, also had self-determination roots: Riordan v An Taoiseach. The applicant objected to the fact that the Irish Constitution was only to be amended 70 KL Scheppele, ‘Aspirational and Aversive Constitutionalism: The Case for Studying Cross-­ constitutional Influence through Negative Models’ (2003) 1(2) Int’l J Const L 296, 299. 71  ibid 299. 72 Ch Villa-Vicencio, Walk with Us and Listen: Political Reconciliation in Africa (Georgetown ­University Press, 2009) 85. 73  ibid 85.

192  Sofia Ranchordás once certain other events—in this case, a referendum, had taken place to the ‘satisfaction of the Government’. He contended that the constitutional amendment procedure of Article 46 of the Irish Constitution had been violated. This section also delves into the legitimacy of constitutional sunrise clauses and their ability to operate ‘self-amendments’ when they are not dependent on subjective contingencies or events that imply the intervention of other institutions. A.  Riordan v An Taoiseach Riordan v Taoiseach was a case decided in 1999 by the Irish Supreme Court,74 in the wake of the 1998 Belfast Agreement, which involved three interdependent components: first, an agreement between the political parties in Northern Ireland as to the internal structures for Northern Ireland; secondly, an agreement between the ­Northern Ireland parties and the Irish Government about the establishment of a number of all-Ireland institutions; thirdly, an agreement between the United ­Kingdom and ­Ireland. In this context, the Irish government committed to seeking the repeal of Articles 2 and 3 of the Constitution, which made a territorial claim to Northern Ireland. According to the formal constitutional amendment laid down in Article 46 of the Irish Constitution, a referendum was compulsory to effectuate such an amendment. In this particular case, the success of the amendment was nonetheless dependent on the result of another referendum in Northern Ireland on the same issues.75 First, the IRA claimed that it was the legitimate government for the whole island of Ireland, as it was the successor of the second Dáil (Assembly), the last body to be elected on the basis of an all-Ireland vote. Secondly, since the 1998 Belfast Agreement was a negotiated agreement, simultaneous agreement was required from the different parties. The only exit under these circumstances was to establish a conditional amendment to Articles 2 and 3, relying on the results of the referenda taking place on the same day. These facts triggered the case Riordan v Taoiseach, where the applicant challenged this constitutional amendment, arguing that the procedure of Article 46 had been violated since the amendment was not only dependent on a constitutional referendum, but also on external events, that is, the referendum in Northern Ireland. The two decision-making moments were therefore ‘tie-barred’ (see Section II). As the Supreme Court explained: before any amendment to Articles 2 or 3 can occur, two conditions precedent must be met. They are the coming into effect of the British/Irish Agreement (…) and the declaration of the Government to the effect that the State has become obliged, pursuant to the multi-party agreement, to give effect to the amendment to the Constitution.

The Supreme Court rejected the applicant’s arguments. Previously the High Court had already explained that: The applicant ha[d] failed to understand the problem which confronted the Government in relation to the Belfast Agreement. That agreement imposed reciprocal obligations on 74  Denis Riordan Applicant v An Taoiseach Bertie Ahern, The Government of Ireland, (No 2), 1998 No 213 JR, SC No 202 of 1998. 75  I thank Oran Doyle for the contextual information on this case.

Constitutional Sunrise 193 the various parties to it and each party wished to be reassured that the other parties would carry out their respective obligations. For instance the Irish Government undertook to have Articles 2 and 3 of the Constitution amended but only on the basis that the British Government and the unionist parties to the Agreement would establish the power sharing executive and the cross-border bodies contemplated in the Agreement. On the other hand, the British Government and the Unionist parties did not wish to establish the power sharing executive and the cross-border bodies only to find that the proposal to amend Articles 2 and 3 was defeated in a constitutional referendum.

Both the Supreme Court and the High Court agreed on the validity and necessity of the ‘condition-precedent’, or sunrise clauses. The High Court underlined: The people have a sovereign right to grant or withhold approval to an amendment to the Constitution. There is no reason therefore why they should not, provided the matter is properly placed before them, give their approval subject to a condition. It is quite wrong to suggest that the people have delegated to the Government the right to amend the Constitution. This is not so. The people have consented to an amendment to the Constitution subject to the happening of a particular future event.

In this case, a contingency was necessary for the operation of constitutional change, adding an additional burden to the formal constitutional amendment process. Also, contrary to the South African decision on the sunrise clauses in the Constitution of KwaZulu Natal, the contingency included in the Irish case was not being used to circumvent the Constitution but rather to promote one of its core values: peace in Northern Ireland. As this case demonstrates, the existence of a condition and connected contingencies taking place outside the Irish Constitution was not judged unconstitutional. This decision illustrates however a different dimension of constitutional sunrise clauses since the contingency is not imposed by the Constitution but by an international agreement between Ireland and the United Kingdom. Although this relationship of ‘condition-precedence’ could be contested, the process of constitutional change and the contingency seem to be logically interconnected. This logic interconnection or tie-barring does not only result from the Irish Supreme and High Courts, but it has also been defended, under different circumstances, in US state and federal courts deciding on contingency-legislation and more specifically on tie-barring provisions. According to some states’ constitutions (for example, Florida and Michigan), the validity of one bill may not be made contingent upon the passage of another one. According to the case law of state courts, tie-barring does not violate the mentioned constitutional provision as long as there ‘is a reasonable relationship between the statutes which have been tie-barred to each other’ or a ‘direct and relative interdependence between them’.76 Although this criterion has been developed in 76  See Re Advisory Opinion to the Governor, 239 So 2d 1, 9 (Fla 1970). See also Gaulden v Kirk [1950] 47 So 2d 567. Town of San Mateo City v State ex rel. Landis, 117 Fla. See also the Cargo Brig Aurora regarding the revival of an act by proclamation of the President, and Field v Clark, where the constitutionality of broad contingent legislation was challenged. In Brig Aurora, the Court denied that the President was exercising unwarranted discretion in reviving a law, although the reasons to revive the law were not established in the statute at the time. See Cargo of the Brig Aurora v United States [1813] 11 US 382. However, the enactment of contingent legislation has also been criticised in the United States, particularly when the legislative acts are found to be incomplete and uncertain until another agent acts. For a critique of contingent legislation, see Th M Cooley, Constitutional Limitations (Legal Classics

194  Sofia Ranchordás a different setting, the logic interdependence between the sunrise of a provision and its contingency appear to be a reasonable rule applicable to any contingent sunrise clauses discussed in this chapter. B.  Sunrise Clauses and the Constituent Power Following the Irish case, I explain in this section why the constituent power should be allowed to include sunrise clauses in the constitutional text, deferring the coming into effect of a constitutional provision to the future or requiring future generations to take further actions in order to determine its commencement. While automatic sunrise clauses can be triggered automatically, a contingent sunrise clause imposes a deliberative burden on future generations. By doing so, sunrise clauses distinguish themselves from other timing provisions such as sunset clauses that determine the termination of a clause. Regardless of whether sunrise clauses leave some or no discretion to the Executive or Parliament as to when and how to allow a certain provision to sunrise, it is clear that they voice a constituent desire of constitutional change. Instead of allowing the future generation to make the central decision on such change, sunrise clauses enable the constitutional drafter to include it in the text and only defer its sunrise. If such a constitutional change is operated automatically by simply delaying the effective date of a provision, a self-amendment or an ‘autopoietic amendment’ would occur.77 Such an amendment would operate regardless of existing formal constitutional amendment procedures. This might seem disconnected from reality but it is not. For example, in the case of a constitution enacted after a conflict, the constituent power may foresee that the country will only be prepared to accept or concretise certain provisions at a later date (for example, Articles 167 and 169 of the Weimar Constitution). In this sense, contingent constitutional change is susceptible of materialising the traditional concept of the constituent power as a rational and unlimited power (for a thorough discussion of the concept of constituent power see the chapter by Thomaz Pereira),78 which ‘even after the constitution is adopted does not lose the right to alter its decision’.79 When sunrise clauses are triggered automatically, the constituent power ­sunrises, resulting in a ‘self-amendment’ of the Constitution. This idea can seem at first Library, 1987) 121. See also F Kameny, ‘Are Inseverability Clauses Constitutional?’ (2005) 68 Alb L Rev 957, 1016, citing 16 CJS Constitutional Law ß 166, at 532 (1984): ‘In any case, as a general rule, the enactment of the statute itself may not be made contingent on the action of officers or people; the act must be complete in itself, must be made law by the legislature, and only its effect and operation may be made dependent on the contingency.’ For a more recent study of severability and conditions in constitutional law, see E Fish, ‘Severability as Conditionality’ (2015) 64 Emory L J 1298. 77  I borrow the term from Niklas Luhmann and Gunther Teubner, but not entirely with the meaning developed by these scholars, see G Teubner, Autopoietic Law: A New Approach to Law and Society (Walter de Gruyter, 1988). 78  See E Sieyès, What is the Third Estate? (Hackett Publishing Company, Inc, 2003) 79  See J Colon-Rios, Weak Constitutionalism: Democratic Legitimacy and the Question of Constituent Power (Routledge, 2012) 8. See also M Loughlin, ‘The Concept of Constituent Power’ (2014) 13(2) Eur J Pol Theory 218.

Constitutional Sunrise 195 ­ roblematic as it blurs the concepts of ‘original constituent power’, the sovereign p power to ‘make the Constitution’, and ‘derived constituent power’, the power to amend the Constitution.80 However, the legitimacy of these constitutional sunrise clauses is found in the original constituent power, which comprehends the power to draft the Constitution and determine when it comes into effect. The term ‘self-amendment’ might sound rather imprecise as on the one hand it suggests that sunrise clauses amend the Constitution; and on the other, it is antithetical as the constituent power only speaks once. Can sunrise clauses be qualified as alternative and legitimate instruments of constitutional change? As Akhil Amar explains in the context of sunrise clauses: ‘at first, it might be wondered whether anything truly constitutional can be said about which future amendments should be adopted. After all, we the People are free to adopt just about any amendments we like’.81 The Constitution is a project with unwritten parts, which can be enlightened not only by the legislature but also by the framers. Amar argues that, on a second thought, this enlightenment regarding future amendments might be entirely constitutional depending on the topics. Sunrise clauses can thus be interpreted as devices that shape the future trajectory of the Constitution. Amar suggests, for example, that a future amendment (or sunrise clause) regarding the criminalisation of flag-burning or the restriction of equality rights of same-sex couples should be rejected as such amendments ‘would do violence to the trajectory of the American constitutional project over the past two hundred years’. Amar suggests however different future amendments or sunrise clauses, such as those used by ‘various states in the Founding era to achieve the gradual abolition of slavery itself. Under these rules, existing slaves would not be liberated—but eventually their future children would walk free’.82 Sunrise clauses can thus potentially accommodate constitutional change without requiring predictable constitutional amendments. These clauses can also constrain future constitutional change by impeding the constituted powers to decide on the life of a certain disposition.83 These instruments induce a process of ‘self-amendment’ which is determined by the constituent power in an attempt to avoid the perpetuation of dispositions motivated by transitional concerns or insufficient information.84

80  Referring to the distinction in the French literature between original constituent power, as the power to make the Constitution, and derived constituent power, as the power to enact and amend (‘établir la constitution’) and arguing that the power to amend constitutions is ‘sui generis’ constituent power, see Rosnai, ‘Towards a Theory of Unamendability’ (n 2). On the traditional definitions of original constituent power and derived constituted power in the French literature, see, eg, G Burdeau, Droit constitutionnel (LGDJ 1988) 76–77; K Gözler, Le pouvoir de révision constitutionnelle (Presses Universitaires du ­Septentrion, 1997) 9. 81  Amar (n 4) 451. 82  ibid 446. 83  See House of Lords, Select Committee on the Constitution, The Process of Constitutional Change, 15th Report Session 2010–2012, 21 (‘even without a codified constitution, it is possible to provide for specific procedural requirements within constitutional bills, including such measures as (…) sunrise and sunset clauses, and parliamentary super-majorities’.) 84  Temporary instruments are often used to gather information and overcome the cognitive bias that often characterises the lawmaking process, see JE Gersen, ‘Temporary Legislation’ (2007) 74 U Chi L Rev 247.

196  Sofia Ranchordás A sunrise clause may also provide additional time for society to reach an agreement on controversial topics that the society might not have been prepared to accept at the time of the drafting (for example, the organisation of a referendum on secession or other self-determination rights) and concretise (for example, socioeconomic rights). In this context, it is important to emphasise that, at the resemblance of sunset clauses and transitional provisions, the use of sunrise clauses should also be limited to constitutional provisions that do not consecrate fundamental rights or state institutions necessary for the normal functioning of the society at stake.85 This normative statement, based on the case-law of the German Constitutional Court, implies a casuistic analysis which should consider the political and economic situation of the country. V. CONCLUSION

Time and constitutions are so intimately connected that it is often hard to separate them.86 The intertwining of past, present and future constitutional wills is personified in a constitution that is supposed to live beyond its framers. A constitution implies multiple commitments between the past, present, and future generations, and a more durable constitution is commonly regarded as a shield against the whims of political power, shifting political majorities and crises.87 However, ‘the past cannot fully control the future (…) [and] there are not enough rules in the world adequate to navigate centuries of change and crisis’.88 Sunrise clauses personify the challenges of this inter-generational dialogue but instead of deferring constitutional decisions to the future, they allow the framers to make them. Sunrise clauses allow the constitution-making and amending process to be not only a backward-looking effort in the sense of ‘claiming our constitutional inheritance’ but also a form of ‘making a constitutional donation’ to ‘our posterity’.89 ­Nevertheless, sunrise clauses determine a priori the path to our posterity, by entrusting the constituent power with the responsibility of predicting an effective date for certain constitutional changes as well important contingencies that might c­ onsolidate the enforcement of a constitutional provision.

85 See my previous work S Ranchordás, Constitutional Sunsets and Experimental Legislation: A Comparative Perspective (Edward Elgar, 2014). Deciding on the constitutionality of transitory law, the ­German Constitutional Court decided in 1976 that ‘laws that are indispensable for the legal capacity and [normal] functioning of a state’ and the laws that are required for the concretisation of fundamental rights guarantees (eg, media and broadcasting laws in that specific case) are not compatible with a temporary or transitory nature. See 1 BvR 79/70 of 09.11.1976. 86  Dupré and Yeh (n 59) 45. 87  For a thorough discussion of the value of temporary constitutions as a reality opposed to the traditional perception that constitutions should be rigid and long-enduring, see Varol (n 11) 411 (‘a durable constitution is thought to constrain political majorities in moments of irrational fear or passion. A constitution represents a powerful acknowledgement by a society of its own weaknesses and its ability to fall prey to pernicious majoritarian impulses’). 88  J Balkin, Constitutional Redemption: Political Faith in an Unjust World (Harvard University Press, 2011) 9. 89  Amar (n 4) 477.

Constitutional Sunrise 197 The added value of sunrise clauses when compared to constitutional amendments or by-law clauses resides namely first in the fact that here the constituent power decides to confer a constitutional character to such controversial topics and ensure that they are included in the constitutional agenda. Secondly, instead of delegating further decisions on these issues to the legislature using, for example, by-law clauses, the constituent power decides and only delays the coming into effect of such dispositions to a moment when, for example, all the necessary economic or political conditions will be verified. Sunrise clauses remain a relatively overlooked instrument in the constitutional practice. This chapter has sought to fill this gap, opening the door to future research on both sunrise clauses and contingency constitutionalism. This perspective on contingent constitutional change complements the ‘one-way communication’ or monologue between the past and future generations, and acknowledges a basic fact of life that has permeated most fields of law: in long-term contracts, change is inevitable but constitutional drafters might also want to ensure that the validity of the constitutional contract remains contingent upon it.

198 

10 Constitutional Change and Interest Group Politics Ireland’s Children’s Rights Referendum ORAN DOYLE AND DAVID KENNY*

I. INTRODUCTION

C

ONSTITUTIONS ARE SUBJECT to both formal and informal change. Whereas comparative constitutional law literature on informal change tends to focus on the relationship between judicial processes and specific doctrinal changes, the literature on formal change tends to focus on general questions about the purposes,1 modalities,2 difficulties,3 and limits of constitutional change.4 That literature also explores the role of ‘the people’, perhaps as a bearer of constituent power, in relation to fundamental constitutional change.5 However, the dynamics of formal change, although studied by political scientists, rarely receive the same doctrinally informed scrutiny as does informal constitutional change. In other words, there is little consideration of how the political processes of formal amendment relate to and are affected by the precise change to constitutional law that they seek to make. In this chapter, we respond to that gap in the literature. Ireland’s Children’s Rights Amendment, approved by Referendum in November 2012 and upheld by the Supreme Court in April 2015,6 provides an interesting example of these dynamics for three reasons. First, it involves an under-studied topic: formal constitutional

*  We are grateful to the members of the public law discussion group in Trinity College Dublin and to Dr Conor O’Mahony for their comments on earlier drafts of this paper. 1  See, for instance, R Albert, ‘The Expressive Function of Constitutional Amendment Rules’ (2013) 59 McGill L J 225. 2  See, for instance, Sofia Ranchordás’s contribution to this volume. 3  See, for instance, Derek O’Brien’s contribution to this volume and also R Albert, ‘Constructive Unamendability in Canada and the United States’ (2014) 67 Sup Ct L Rev (2d) 181. 4  See, for instance, the contributions of Yaniv Roznai and Oran Doyle to this volume. 5  See, for instances, the contributions of Juliano Benvindo, Joshua Braver and Thomaz Pereira to this volume. 6 See Jordan v Minister for Children and Youth Affairs [2015] IESC 33. The Supreme Court rejected a challenge claiming the result was influenced by the government’s unconstitutional information campaign, discussed below.

200  Oran Doyle and David Kenny change as a means to amend intricate, judicially created doctrine rather than to make foundational constitutional decisions. Secondly, it provides a further insight into the dynamics of referendum campaigns, currently a significant concern in constitutional politics. Thirdly, it focuses attention on the role of constitutional lawyers and interest groups as both framers and agents of formal constitutional change. We argue that the Children’s Rights Referendum was a failure in deliberative democracy, not because of elite control (a common concern over referendums) but rather because constitutional lawyers and public interest groups combined to offer a false diagnosis of a constitutional malaise.7 This led to calls for reform under the simplistic rubrics of ‘children’s rights’ and ‘the best interests of the child’. Ultimately, this confused analysis of constitutional law allowed civil society groups to support a referendum that made minimal change and largely reproduced what those groups had claimed to be a constitutional malaise. Conversely, the referendum was opposed by some religious/conservative voices, notwithstanding that it largely maintained the constitutional status quo that they purported to favour. There followed a confused and dispiriting referendum campaign, which included a finding by the Supreme Court days before the referendum that the government had funded an unconstitutionally biased information campaign. The result was a (for many) surprisingly narrow approval of 58 per cent: 42 per cent for ‘children’s rights’, with the third lowest ever turnout for a referendum vote.8 We argue that four generalisable lessons should be drawn from this sorry affair. First, where the desire is to change judicially created legal doctrines, only significant change warrants engaging the process of constitutional amendment. Although constitutional changes may be worthwhile for either expressive or functional reasons, we should be prepared to live with sub-optimal constitutional law and not propose trivial changes. Secondly, constitutional lawyers should be wary of legal analysis that lends itself to presentation in terms of political cleavages that exist in society. This can generate calls for totemic change and impede efforts to understand and explain why constitutional change is required. This problem is particularly acute where the constitutional change is intended to alter judgemade rules. Thirdly, much of the academic literature on referendums focuses on the danger of political elites manipulating the populace in order to legitimise their change-projects. The experience of the Children’s Rights Referendum suggests the need to pay more academic attention to the dangers posed by interest groups securing control of the political agenda. Finally, campaigns for constitutional change that

7  For a different, although partially complementary, account of the dynamics of the Children’s Rights Amendment, see CO’Mahony, ‘Falling Short of Expectations: The 2012 Children Amendment, from Drafting to Referendum’ (2016) 31 Irish Political Studies 252. 8  Though this result represented a clear victory for the Yes side, the Referendum had been broadly supported by civil society groups and political parties, and was long desired by many sectors of Irish society. The common expectation was that the Referendum would be almost uncontested and would pass with a far more decisive margin. (Three years later, the far more divisive proposal to allow same-sex marriage was approved by 62% of the people.) The poor quality of the campaign, the confusion around the proposal, and the Supreme Court finding against the government information campaign probably all contributed to this. F Sheahan, ‘Children’s Referendum passed by thin margin of 58pc to 42pc’ The Irish Independent 3 (11 November 2012); H McGee ‘Why the referendum was closer than predicted’ Irish Times (13 November 2012).

Constitutional Change and Interest Group Politics 201 seek to entrench or refine constitutional values should engage directly and clearly with those values; otherwise, the potential for meaningful engagement with voters will be lost. II.  REFERENDUMS AND CONSTITUTIONAL CHANGE

Referendums have a chequered history as instruments of legal change. Gallagher and Uleri note the ‘standard criticism’ that ‘the general public is simply too ill-informed or irresponsible to be trusted to make decisions on complicated questions’.9 For some, the deliberative and democratic credentials of referendums are irretrievably undermined by the mismatch between the purity of the democratic claim (‘the people have spoken’) and the possibility for elite control. Elites can control who votes (either through delineating the territory or the criteria for identifying voters), what is voted on (constructing the question), the flow of information (whether through funding rules, limits on spending or restrictions on free speech), and the amount of time for deliberation.10 The Summer 2015 referendum in Greece was widely perceived as problematic on account of the short campaign and the opacity, perhaps even pointlessness, of the question asked.11 These concerns were substantiated by the fact that the Greek Government subsequently signed up to bailout terms that were far harsher even than those apparently rejected in the referendum.12 Likewise, concerns have been raised over the use of referendums against the backdrop of physical force and secessionist claims.13 Even if these complaints turn on the misuse of referendums, referendums can still be problematic in functioning democracies that seek to avoid elite control. The problem of rational ignorance (given the low likelihood that any one vote will affect the outcome) may be even more acute for referendums that decide issues than for elections of representatives.14 Others accept that these problems may arise, but argue for measures that reduce elite control and increase the possibilities for deliberative engagement. If a referendum process can genuinely be an exercise in deliberative democracy, the authenticity of the people’s voice that emerges is a valuable element of a democracy, particularly on questions of constitutional foundation and constitutional

9 

M Gallagher and PV Uleri, The Referendum Experience in Europe (Macmillan Press, 1996) 240. S Tierney, Constitutional Referendums: The Theory and Practice of Republican Deliberation (Oxford University Press, 2012) 23–24. 11 See X Contiades and A Fotiadou, ‘The Greek Referendum: Unconstitutional and Undemocratic’ (Constitution Making & Constitutional-Change, 7 July 2015) accessed 26 July 2015. 12  S Shuster, ‘Greece Agrees to its Third European Bailout after Marathon Talks’ (Time, 13 July 2015) accessed 26 July 2015. 13  See, for instance, S Tierney, ‘Sovereignty and Crimea: How Referendum Democracy Complicates Constituent Power in Multinational Societies’ (2015) 16 German Law Journal 523. 14  M Setälä, ‘Introduction’ in M Setälä and T Schiller (eds), Referendums and Representative Democracy: Responsiveness, accountability and deliberation (Oxford, 2009) 9, citing A Downs, An Economic Theory of Democracy (Harper, 1957). 10  See

202  Oran Doyle and David Kenny change, not attainable in any other way.15 The Scottish Independence Referendum is generally seen as a positive example of the referendum process.16 Several features of the Irish referendum process protect against the sort of problems that arose in Crimea and Greece. Since a referendum is required for all constitutional change and referendums can, apart from one never utilised exception, only be used for that purpose, there is little scope for elites to use referendums opportunistically or to skew the wording of referendum proposals.17 Although the Government controls the drafting and initiation process, it must secure support for its referendum proposal in both Houses of Parliament where amendments can and have been made in response to public disquiet. Judicial decisions prohibit the Government spending money on one side of the referendum campaign and require fair media coverage.18 An independent Referendum Commission explains referendum proposals to the public and encourages them to vote. Statute requires a campaign period of between 30 and 90 days,19 although Governments generally allow close to the minimum time for referendum campaigns. Although the short campaign period has been criticised in respect of other referendums,20 the Children’s Rights Referendum was formulated in response to demands from civil society interest groups and followed several other proposals for change that had been periodically discussed for over a decade. In short, the Children’s Rights Referendum can fairly be characterised as one where public engagement and deliberation was not crowded out by elite control. Nevertheless, we argue that the Referendum was a failure in deliberative democracy: changes to constitutional doctrines were misleadingly presented through simplified political narratives that reflected a political cleavage in Irish society. Constitutional lawyers and civil society interest groups combined to cause this problem. III.  POLITICAL CLEAVAGES AND NARRATIVES FOR CHANGE

Richard Sinnott identifies a ‘potentially powerful religious-conservative versus secular-liberal cleavage that remains politically subliminal’ underlying politics in Ireland.21 This cleavage manifested itself in constitutional referendums on divorce (1986 and 1995) and abortion (1983, 1992 and 2002) and subsequently in the 2015 marriage referendum. As political parties and interest groups coalesce around cleavages, the support of these groups for a proposal can help voters to reach a decision

15  We take this to be the core argument of Stephen Tierney in his impressive contribution to the debate on the desirability of constitutional referendums. See Tierney (n 10). 16  See, for instance, S Tierney, ‘Reclaiming Politics: Popular Democracy in Britain after the Scottish Referendum’ (2015) 86 The Political Quarterly 226. 17  Setälä (n 14) 5. 18  McKenna v An Taoiseach (No 2) [1995] 2 IR 10 and Coughlan v Broadcasting Complaints Commission [2000] 3 IR 1, respectively. 19  The Referendum Act 1994, s 10. 20  See M Marsh, J Suiter and T Reidy, ‘Report on Reasons Behind Voter Behaviour in the Oireachtas Inquiry Referendum 2011’ (January 2012). 21 R Sinnott, ‘Cleavages, parties and referendums: Relationships between representative and direct democracy in the Republic of Ireland’ (2002) Eur J Pol Research 811, 815.

Constitutional Change and Interest Group Politics 203 that reflects their interests, broadly defined.22 Lupia and Johnston argue, however, that voters can be competent without being fully informed, once they reach the same judgement as they would if fully informed.23 The Children’s Rights Referendum was presented in terms of the cleavage identified by Sinnott.24 In the years leading up to the referendum proposal, the interest groups supporting the proposal were, broadly speaking, on the secular/liberal side of the political spectrum. More starkly, most of the voices raised against the proposal were from a religious and/or conservative background. Mary O’Rourke, a leading politician who chaired the Oireachtas Committee that formulated one of the reform proposals, characterised opponents of the children’s rights referendum as ‘an extreme right-wing element in Ireland which— even though the headlines scream of the shameful abuse of children—will regard any such intrusion as being against the Constitution and contrary to the fundamental rights of the family’.25 According to this narrative for change, conservative forces wanted to maintain a constitutional preference for parental rights over children’s rights. However, this depended on a misunderstanding of the constitutional position. The Constitution, as we shall shortly see, did not grant parents rights over their children, but rather gave parents considerable authority to decide on what was in the best interests of their children and whether to assert their children’s rights. Defenders and opponents of this constitutional status quo could both legitimately claim to respect a child’s best interests and children’s rights; they differed over the appropriate level of parental authority in respect of those best interests and rights.26 Although the pre-amendment Articles 41 and 42 of the Irish Constitution contain one or two references to parental rights, the better characterisation is that they accord a high level of authority to parents in respect of children’s rights. The Family is protected ‘in its constitution and authority’. Parents have a ‘right and duty’ to provide for the education of their children. There is one explicit child’s right: the right to attend a school receiving public money without attending religious instruction. There is one explicit parents’ right: the rights of parents in the matter of religious and moral formation. However, the centrally important provision of Article 42.5 gave a residual power to the State to protect children, framed in terms of the child having rights and parents owing a duty to their children. This is fundamentally inconsistent with a parental rights model. The better reading of the Constitution is therefore that it recognised children’s rights and provided that parents were to be the authoritative decision-makers for children, but with a residual power for the State to interfere in the case of parental failure. It could be argued that this gave too much authority to parents, but it did not protect parental rights in opposition to children’s rights. 22  T Christin, S Hugz and P Sciarini, ‘Interests and information in referendum voting: An analysis of Swiss voters’ (2002) Eur J Pol Research 759, 759. 23  A Lupia and R Johnston, ‘Are Voters to Blame? Voter Competence and Elite Maneuvers in Referendums’ in M Mandelsohn and A Parkin (eds), Referendum Democracy (Palgrave, 2001) 191. 24  O’Mahony supports this analysis. See O’Mahony (n 7) 256–57. 25  M O’Rourke, Just Mary (Gill & McMillan, 2012) 174. 26  There can, in principle, be an objective conception of the child’s best interests. It is simply the case that no model of children’s rights can secure a child’s best interests in all circumstances, because both parents and the state are prone to mistakes. On the general difficulties with a best interests test, see J Herring and C Foster, ‘Welfare means relationality, virtue and altruism’ (2012) 32 Legal Studies 480; S Parker, ‘The Best Interests of the Child—Principles and Problems’ (1994) 8 Int’l J of Law and the Family 26.

204  Oran Doyle and David Kenny Consistent with this analysis, the courts interpreted Article 42.5 as mandating considerable deference to parents when making joint decisions about their children, unless they had failed in their duty. Decisions made by well-meaning parents to refuse medical treatment or procedures would not be overturned, unless serious consequences were immediately likely.27 In custody cases, the courts had ordered the return of a child to her parents who married after an adoption process commenced but before it completed. In an early case, the Court ordered the return against what appeared to be in the best interests of the child.28 In the most recent case, the Court characterised the best interests of the child as being likely served by return to the now married natural parents.29 The voluntary placement for adoption of marital children was not permissible. Furthermore, the adoption of marital children against their parents’ wishes was not legislatively permitted until 1987.30 All of these are evidence of a high degree of parental authority. The question is how they came to be presented as a rejection of ‘children’s rights’, making that phrase of totemic importance to a cleavage between conservative and liberal forces. IV.  CHILDREN’S RIGHTS: FROM LAWYERS TO INTEREST GROUPS AND BACK AGAIN

Academic writing persistently characterised the Constitution as protecting parents’ rights. Writing in the late 1970s, William Duncan commented that the Irish courts had privileged too greatly the rights of parents: ‘The guarantee of rights to a parent who is fit, willing and able to perform his parental duties, has closed the door on even the most compelling arguments for denying those rights in the best interests of the child.’31 Duncan later described the Supreme Court judgment in re JH,32 directing the return of a child to its now married parents, as a ‘powerful blow’ for the rights of married parents. Though ‘couched in the language of children’s rights’, the decision in JH was ‘[i]n reality … an assertion of parental privilege’.33 Alan Shatter argued that the courts’ interpretation of the rights of the family rendered it ‘constitutionally impermissible to regard the welfare of the child as the paramount consideration in any dispute as to its upbringing or custody between parents and third parties’.34 He characterised this as ‘placing such other rights of the legitimate child in a constitutionally inferior or subservient position to the rights of the legitimate child’s parents’.35 Gerard Hogan and Gerry Whyte, editors of the leading treatise on

27 See NWHB v HW [2001] [2001] 3 IR 622, in which the Supreme Court refused to order that a child undergo the PKU test, and Temple Street Hospital v D [2011] IEHC 1, in which the High Court ordered that a child be given a blood transfusion, against the wishes of the parents. 28  M v An Bord Uchtála [1977] IR 287. 29  N v Health Services Executive [2006] IESC 60. 30  re Art 26 and the Adoption (No 2) Bill 1987 [1989] IR 656. 31  W Duncan, ‘------------’ (1978) 2 Dublin University Law Journal 67, 70. 32  [1985] IR 375. 33  Duncan (n 31) 80. 34  A Shatter, Family Law, 4th edn (Wolfhound, 1997) [1.69]. 35  ibid [13.117].

Constitutional Change and Interest Group Politics 205 Irish constitutional law, contended that, in the specific context of custody disputes, the Constitution had ‘engendered an approach which tends to emphasise the rights of adults, on some occasions to the detriment of the rights of the child’.36 The view that the Constitution prioritised parental rights over children’s rights is still prevalent in the work of leading contemporary scholars.37 This analysis was flawed in its characterisation of the problem. There is scant evidence that the courts ever understood the Constitution as a charter for parents’ rights. Even if the academics’ criticism was plausible in the 1970s and 1980s, it was untenable long before the Children’s Rights Referendum. Most explicitly, in N v Health Service Executive Hardiman J in the Supreme Court observed: It would be quite untrue to say that the Constitution puts the rights of parents first and those of children second. It fully acknowledges the ‘natural and imprescriptible rights’ and the human dignity, of children, but equally recognises the inescapable fact that a young child cannot exercise his or her own rights. The Constitution does not prefer parents to children. The preference the Constitution gives is this: it prefers parents to third parties, official or private, priest or social worker, as the enablers and guardians of the child’s rights.38

Nevertheless, the idea that the Constitution protected parents’ rights over children’s rights took hold in the public consciousness. This became linked with the idea that the Constitution precluded the effective protection of children from abuse, whether by their parents, the state, or religious clergy. In 1993, the Kilkenny Incest Report detailed how social services had failed a woman who had been repeatedly raped by her father. Although not identifying any specific way in which the Constitution had impacted negatively on the victim’s situation, the Report nevertheless called for constitutional change on the ground that the Constitution might consciously or unconsciously have been interpreted as giving a higher value to the rights of parents than to the rights of children. This could be addressed by the Constitution containing ‘a specific and overt declaration of the rights of born children’.39 In 1996, the Constitution Review Group similarly recommended the textual protection of specified children’s rights, including those already enumerated by the courts and those protected under the UN Convention on the Rights of the Child, as well as the incorporation of an express obligation to treat the best interests of the child as the paramount consideration in any actions relating to children.40 From the outset, therefore, advocates for reform adopted the incorrect view that the Constitution protected parents’ rights in opposition to children’s rights, loosely

36  G Hogan and G Whyte and JM Kelly: The Irish Constitution, 4th edn (Butterworths, 2003) [7.6.198] citing the Kilkenny Incest Report. cf [7.6.207] et seq. 37 Ursualla Kilkelly argues that ‘children’s rights are ignored or underplayed when they come into conflict with those of their parents’. U Kilkelly, Children’s Rights in Ireland: Law, Policy and Practice (Tottel, 2008) 83. 38  [2006] IESC 60. 39 ‘Kilkenny Incest Investigation’ (Government Publications Office 1993) 96 accessed 16 June 2014. 40  ‘Report of the Constitutional Review Group,’ (Government Publications Office, 1996) 328–9.

206  Oran Doyle and David Kenny blamed this constitutional position for bad outcomes for children, and sought to address the situation by amending the Constitution to protect ‘children’s rights’. They sought a general declaration of children’s rights, the enumeration of specific rights, and the inclusion of a best interests test. For example, Kilkelly and O’Mahony suggested that the Constitution should recognise the independent rights of children and require that a child’s best interests be a primary consideration in any dispute affecting the child.41 However, this proposal would have both under-achieved and over-achieved. Simply enumerating children’s rights would make no difference since it would not affect the question of who had authority to assert the child’s rights. Making a child’s best interests a primary consideration in any dispute affecting the child would require the state (in the form of social workers or the courts) substitute its own judgment for that of the parents whenever there was a dispute, which dispute could itself be triggered by the fact of the state’s disagreement with the parents. The state would thereby become the primary decision-maker for the child. ‘Children’s rights’ grabbed attention but achieved little. The ‘best interests’ test appeared innocuous but achieved too much. This sort of analysis did not address directly and openly the crucial question of how to distribute authority between parents and state; it did, however, provide a catchy and capacious political slogan around which interest groups were able to coalesce, leading the campaign for constitutional change. We do not use the term ‘interest group’ in a pejorative way.42 The groups that lobbied for constitutional change in Ireland exhibit the three characteristics suggested by Beyers et al as defining interest groups: some level of organisation, to distinguish them from simple waves of popular opinion or broad social movements; political interests, or particular policy outcomes sought to be achieved; and informality, insofar as they usually do not run candidates for election or seek public office.43 They included the Children’s Rights Alliance, formed in 1995 to secure the rights of children in Ireland through the full implementation of the Convention on the Rights of the Child; Barnardos, a children’s charity that also engages in campaigning for children on issues such as housing, public health, education, and child protection; and the Irish Society for the Prevention of Cruelty to Children (ISPCC), a charity that, amongst other things, provides support services for children. The interest groups secured the commitment of the main political parties to constitutional reform under the slogan of children’s rights, perhaps because that slogan could accommodate two completely contradictory answers to the question of the state’s responsibility for protecting children. Reflecting this, the various reform

41  U Kilkelly and C O’Mahony, ‘The Proposed Children’s Rights Amendment: Running to Stand Still’ (2007) 2 Irish Journal of Family Law 19. 42  Some prefer using terms such as social movement organisation or civil society organisation to define movements not involved in specific sectoral lobbying. J Beyers, R Eising and W Maloney, ‘Researching Interest Group Politics in Europe and Elsewhere: Much We Study, Little We Know?’ (2008) 31(6) West European Politics 1103, 1110. 43  Beyers et al (n 42) 1106–07. Murphy notes that Irish interests groups have occasionally run candidates for political office. See G Murphy, ‘Interests Groups in the Political Process’ in J Coakley and M Gallagher (eds), Politics in the Republic of Ireland, 5th edn (Routledge, 2010) 328.

Constitutional Change and Interest Group Politics 207 proposals differed wildly in content.44 The Government’s proposal in 2007 was minimalist in character.45 It affirmed the existence of children’s rights but did not guarantee to enforce them and provided no new rights. It made little change to the threshold for state intervention in families and gave only limited application to the best interests test. In contrast, the All-Party Oireacthas (Parliament) Committee proposal in 2010 not only acknowledged and affirmed the rights of children but also pledged the protection and vindication of those rights insofar as practicable.46 It further provided constitutional protection for a range of specific rights, some of a socioeconomic character, and entrenched the best interests principle, thereby substituting court decision-making for parental decision-making. What finally secured a proposal that would actually be put to the people was the revelation of the scale of child abuse. In 2009, the Commission to Inquire into Child Abuse found that physical, emotional and sexual abuse was widespread in residential institutions and industrial schools in Ireland.47 These schools were owned and managed by religious congregations but were funded by the state and generally housed children referred there by state authorities, which the Commission found to have a ‘deferential and submissive’ attitude towards the religious congregations. Complaints made by parents to state authorities were not properly investigated. The same year a statutory Commission concluded that the Archdiocese of Dublin placed its reputation ahead of the protection of children.48 Senior police officers had seen it as their role to report allegations of abuse by priests to the church authorities rather than carry out their own investigation. In October 2010, a report detailed how social services had failed for 15 years properly to address a case of incest, neglect and rape within a family.49 Bad systems, a misplaced willingness to believe that the situation would improve and a tardy and misplaced reaction to an interim court order secured by the mother were all identified as contributing to the way in which social services failed the family. Most of the abuse catalogued in these reports was done with the complicity of the state, often as a result of parents being ignored. Where parents were at fault, there is no suggestion that the Constitution actually impeded the state from taking action to protect the children in question. Notwithstanding that the reform campaign had originated in a claim that the Constitution overly protected parents’ rights, the

44  For a more detailed comparison of these proposal, see D Kenny, ‘The Children’s Rights Referendum: A Short History’ (Trinity College Dublin Conference on the Children’s Rights Amendment, 1 November 2012) accessed 26 July 2016. 45  Twenty-Eighth Amendment to the Constitution Bill 2007. 46  ‘Final Report of the Joint Committee on the Constitutional Amendment on Children’ (Government Publications Office, 2010). 47  ‘The Report of the Commission to Inquire into Child Abuse’, available at accessed 19 March 2013. The following details are taken from the executive summary. 48  ‘Report of the Commission of Investigation into the Catholic Archdiocese of Dublin’, available at accessed 19 March 2013. The following details are taken from the Overview. 49  ‘Report of the Inquiry Team to the Health Service Executive’, available at accessed 21 March 2013.

208  Oran Doyle and David Kenny ‘children’s rights’ slogan was sufficiently capacious to include the contradictory claim that the Constitution gave too much power to state institutions at the expense of parents. This capaciousness also explains how such widely divergent proposals could be formulated under the rubric of ‘children’s rights’, and equally welcomed by ‘children’s rights’ interest-groups. The phrase ‘children’s rights’ had become totemic for the secular-liberal side of the political cleavage. It was an effective totem because it was agnostic as to the core constitutional issue of parental authority and therefore could plausibly be attached to any reform proposal. The Amendment that ultimately passed largely preserves the constitutional status quo. It retains the preference for parental authority in respect of children’s rights. Although replacing Article 42.5, it makes only marginal changes to the threshold at which the state can intervene in families. O’Shea argues, and we agree, that the postamendment threshold could still ‘serve as a significant barrier to State i­ ntervention’.50 It seems to change little, and would not appear to alter the required outcome of the healthcare cases considered above. Article 42A.4 requires that legislation provide that the best interests of the child will be the paramount consideration in proceedings brought by the State to protect child welfare or involving the adoption, guardianship, custody of or access to a child. The legislation must also ensure that the views of the child will be ascertained and given appropriate weight. This use of the ‘best interests’ test is radically different from that in the 2010 proposal. Here ‘best interests’ is the criterion to be employed by the court only where it has already deemed that parents have failed in their duty requiring the court to take their place. Under the 2010 proposal, ‘best interests’ had provided both the threshold for intervention and the criterion to guide intervention. The Amendment allows married parents voluntarily to place their children for adoption. It authorises the legislature to require the application of the best interests test and the giving of weight to the views of the child in adoption, guardianship and custody proceedings. This might allow for a different outcome in those rare but troubling cases, noted above, where unmarried parents marry before an adoption process is completed. These are the only two changes of any significance effected by the Amendment. O’Shea, a strong advocate of constitutional change to improve child protection, characterised the 2012 changes as ‘likely to be minimal’, and only a ‘small step in the right direction’; the rights of the child ‘would seemingly remain inferior and ­subordinate to the protection afforded to marital autonomy’.51 More commonly, however, reform advocates sought to extract from the amendment proposal some constitutional provisions that could be said to amount to significant change. For instance, some pointed to the state’s recognition and affirmation of the natural and imprescriptible rights of all children, and more particularly its commitment to ­‘protect and vindicate’ those rights as far as practicable. Kilkelly and Corbett

50  N O’Shea, ‘Can Ireland’s Constitution Remain Premised on the ‘Inalienable’ Protection of the Marital Family Unit Without Continuing to Fail its International Obligations on the Rights of the Child?’ (2012) Irish Journal of Family Law 86, 90. 51  O’Shea (n 50) 90–92.

Constitutional Change and Interest Group Politics 209 s­ eparately argued that this was a ‘game-changer’,52 making it the State’s responsibility to uphold the rights rather than leaving it within the private confines of the family. This analysis was incorrect, confusing the notion of the state vindicating children’s rights with the notion of the state asserting children’s rights. Under the Amendment, it is only children’s rights as asserted by parents that the State is under an obligation to defend and vindicate. Despite retaining the parental authority model of children’s rights that initially prompted calls for constitutional reform and despite making almost no change to the threshold at which the state can intervene in families, the referendum proposal received practically unanimous support from major children’s rights groups upon its launch.53 The reasons for this are difficult to discern. There is little research on the relationships between and within interests groups, the reasons for their political action, and the effect this might have on the tactics they adopt.54 There is also a significant under-exploration of the role of interest groups in constitutional change, and why interest groups might adopt a strategy focused on constitutional amendment.55 Nevertheless, it is possible to make some assessment of why interest groups supported the referendum proposal. Campaigners and interest groups should be slow to pursue constitutional rights protection unless it is truly necessary. Posner and Landes note that constitutional amendment is ‘costly and time-consuming’.56 Waldron notes that the considerable political difficulties involved give reason to hesitate before seeking constitutional protection of one’s interests.57 There are several reasons why an interest group might feel the need to pursue protection at a constitutional level in spite of these difficulties. First, the interest group might only be able to achieve its end through constitutional change, since the constitutional order entrenches something antithetical to their goal. However, the Children’s Rights Amendment only made two such changes, in the field of adoption; these were not the focal point of the campaign for change, and do not explain the broader aspects of the proposal. Secondly, Posner and Landes

52  U Kilkelly, ‘Yes could be a real game-changer’ The Irish Times (7 November 2012); cf M Corbett, ‘The Children’s Referendum is a Game-Changer for Children’s Rights in Ireland’ (2012) Irish Journal of Family Law 94. 53  C Kenny and M Minihan, ‘Referendum Wording Welcomed’ The Irish Times (19 September 2012). 54  Beyers et al (n 42) 1120. 55  Of the small amount of literature that exists, most of it is found in the Law and Economics movement. The reasons for interest groups to seek constitutional change was briefly discussed by W Landes and R Posner in a landmark paper, ‘The Independent Judiciary in an Interest-Group Perspective’(1975) 18(3) Journal of Law and Economics 875. Other law and economics scholars continued to apply their work to constitutional amendments, particularly at the State-level in the US; WM Crain and RD Tollison, ‘Constitutional Change in an Interest Group Perspective’ (1979) 8 J Legal Stud 165-75 cf DJ Boudreaux and AC Pritchard, ‘Rewriting the Constitution: An Economic Analysis of the Constitutional Amendment Process’ 1993 62 Fordham L Rev 111; D Sutter, ‘Constitutional Politics Within the Interest Group Model’ 1995 6(2) Constitutional Political Economy 127. 56  Landes and Posner (n 55) 892. The extent of the difficulty is greater in the United States, where they were writing, but still true in Ireland. As Dixon puts it, the ‘bargaining cost’ of seeking constitutional amendments is high. R Dixon, ‘Constitutional Amendment Rules: A Comparative Perspective’ in T ­Ginsburg and R Dixon (eds), Comparative Constitutional Law (Elgar 2011) 104. 57  J Waldron ‘A Rights-Based Critique of Constitutional Rights’ (1993) 13 Oxford Journal of Legal Studies 18, 25–26.

210  Oran Doyle and David Kenny argue that a ‘constitutional provision confers more durable protection than is possible by ordinary legislative action’ and would thus be more valuable to an interest group.58 This has the advantage of forcing any interest group pursuing a contrary outcome to operate at the constitutional level in future. It could also answer concerns that a court could dislodge the status quo in the interests of vindicating the reputed constitutional rights of other groups. The Pro Life Amendment Campaign (PLAC), which lobbied successive Irish governments for constitutional outlawing of abortion, was probably motivated in part by such considerations.59 However, this tactic could not have motivated the interest groups to support the children’s rights referendum proposal since it presupposes that significant changes to the status quo were being made in the first place. These first two explanations could make sense of the attitude adopted by interest groups only if they did not realise that the proposed Amendment would change little. Given the legal-academic misunderstanding of the constitutional position on children’s rights, this is quite possible. Other explanations focus not so much on the content of the proposal but on its symbolism. The third explanation for why the interest groups sought constitutional change is that the Constitution—as a statement of national priorities and values—is one of the most desirable places for interest groups to have their values expressed and affirmed. Referendums on constitutional change, in particular, offer the chance for a strong and clear statement of ideals,60 and this could be either a way to strengthen a societal ethic, or be a capstone to a process of building that ethic. But this was not the focus of the children’s rights campaign. Instead, the interest groups concentrated on specific injustices and apparent failings in the constitutional order, and the need for their rectification. Perhaps there was a dissonance between the narrative around the need for change and the reality of what most campaigners wanted: a statement of constitutional values rather than substantive constitutional alteration. If so, this emphasis on specific constitutional failings was perhaps a mistake: making a statement of values through constitutional change is a risky business, and one can easily fail to achieve this purpose even if a desired amendment is ratified. Clear articulation of the values being endorsed is important.61 It would have been advisable to put this symbolic goal front and centre, and articulate clearly the change in values that was sought, if this were the core reason for change. The closely related fourth explanation is that the constitutional amendment would represent a political victory for the interest groups, providing a strong statement of the power of the lobby, and the extent to which its views are shared by politicians

58 

Landes and Posner (n 55) 892. Girvin, ‘Social Change and Moral Politics: the Irish Constitutional Referendum 1983’ (1986) 34 Political Studies 61, 68; 73–74. cf M Gallagher, ‘The Constitution and the Judiciary’, in Coakley and Gallagher (eds) (n 43) 86. 60  Of course, other methods of amendment besides referendum have expressive functions; see R Albert, ‘The Expressive Function of Constitutional Amendment Rules’ (2013) 59(2) McGill L J 225. 61  Dixon argues that even nominal success in changing a constitution cannot guarantee that a desired shift in norms and values will take place. She cites the example of the ‘race power’ amendment to the Australian constitution in the 1960s, and hypothesises that an unclear set of priorities about what message to send, along with an equivocal government campaign, contributed significantly to that failure. R Dixon, ‘Amending Constituting Identity’ (2012) 33 Cardozo L Rev 1847. 59 B

Constitutional Change and Interest Group Politics 211 and the people.62 This might be seen as worthwhile for intrinsic reasons or for secondary instrumental reasons, such as the dissuasion of opposition.63 This explanation would characterise the push for constitutional change chiefly as an assertion of political power by the liberal/secular side of the political cleavage identified by Sinnott above. The magnitude of the political victory depends only on the perceived significance of the change, rather than the actual significance, which might explain the groups’ ambivalence as to the content of the proposal. Indeed, abandoning substantive changes could be tactically useful in mollifying potential opposition to this flexing of political muscle. Somewhat related to this explanation, O’Mahony suggests that the interest groups supported the proposal in order to ‘take what you can get’. It was clear that this was the end-point of a lengthy process and the interest groups had already received a philanthropic donation of €1.7m to fund their campaign.64 They may therefore have been pre-committed to support whatever the Government proposed. In all likelihood, some combination of these factors was at play. After many years, a Government had finally published its wording and set a date for holding a referendum claiming to protect children’s rights. Any qualms or uncertainties about the wording of a complex and poorly understood legal change might have been overridden by this achievement, which would have been seen as a potential for a symbolic statement of values, or a status-affirming victory for those who had campaigned so long for a children’s rights referendum. This would not attribute cynical motives to those campaigning for change: they might have thought that the problems they identified demanded change, but lost sight of these substantive changes and became fixated on rhetorical change, or the appearance of change, instead. Completion of the process, and winning the battle, may have become more important than making particular changes. If a final proposal would make some modest improvement in the constitutional position of children, or was not actively deleterious to that position, then supporting it would make sense regardless of the precise content. V.  THE REFERENDUM CAMPAIGN

Whatever the explanation, the attitude of the interest groups made it considerably more difficult for the referendum campaign to function in support of a genuine exercise of deliberative democracy. An amendment that largely reproduced the constitutional status quo was being presented, through the rubric of children’s rights, as addressing a significant constitutional and societal malaise. This made it almost

62  Schlag describes this as rights becoming end goals of a particular political project. P Schlag, ‘Rights in the Postmodern Condition’ in A Sarat and TR Kearns, Legal Rights: Historical and Philosophical Perspectives (University of Michigan Press, 1996) 294. 63  Girvin (n 59) 68, suggests that the campaign for a referendum on abortion has to be seen partially in this light—‘as a riposte to the secularizing tendencies which had appeared so strong throughout the 1970s’. 64  O’Mahony (n 7) 266.

212  Oran Doyle and David Kenny impossible for people to discern precisely what they were being asked to vote on.65 Tierney, although generally sympathetic to referendums, queries whether a referendum is suitable to determine an issue that is so complicated that a reasonably well informed voter is unlikely to understand it.66 The Children’s Rights Referendum provides an example of this. First, there was a disproportion between the length of the amendment proposal and the significance of the changes. Any explanation faced the difficult task of elaborating why so much text achieved so little. Secondly, the calls for reform that animated the children’s rights movement were based on a fundamental misunderstanding of the Constitution as a charter for parents’ rights. This meant that there was no coherent account of how the Constitution should be amended, against which the Government’s actual proposal could be measured for effectiveness. Thirdly, these two problems combined in an unusual way. Both sides could claim to protect children’s rights, and the proposal made little or no change to what most people would have thought it was intended to address: the state’s power to intervene to protect children. As noted above, voters can be competent without being fully informed, once they reach the same judgement as they would if fully informed. In this regard, cues from political parties and interest groups are essential.67 Christin, Hug and Sciarini note theoretical models that suggest endorsements by political parties and interest groups can give important information about the issues at stake in a referendum. This is particularly the case since parties and interest groups seldom have any incentive to misinform voters about their own preferences.68 The Children’s Rights Referendum campaign did not reliably provide these cues to voters, however. All political parties lined up in favour of the proposal, which was strongly endorsed by the interest groups that had sought the Referendum.69 This likely led many voters to believe incorrectly that the proposal took significant steps to improve the ways in which children were treated by the Constitution. Several opponents of previous children’s rights proposals also cautiously welcomed the proposal, apparently satisfied that the proposal did not undermine parental authority within families.70 The people were therefore presented with most existing protagonists to the debate either supporting or not opposing the proposal. This elite consensus was the likely cause of three features of the subsequent campaign. First, proponents of the referendum suggested that the broad support from political

65 A survey after the referendum established that 41% of people felt that they did not understand the referendum either at all or particularly well. Referendum Commission’s Report on the Referendum on the 31st Amendment of the Constitution (Children) Bill 2012, available at accessed 27 July 2016. This level of declared misunderstanding is not unusual for recent referendums. 66  Tierney (n 10) 228. 67  A Lupia and R Johnston, ‘Are Voters to Blame? Voter Competence and Elite Maneuvers in Referendums’ in M Mandelsohn and A Parkin (eds), Referendum Democracy (Palgrave, 2001) 191. 68  T Christin, S Hugz and P Sciarini, ‘Interests and information in referendum voting: An analysis of Swiss voters’ (2002) European Journal of Political Research 759. 69  C Kenny and M Minihan, ‘Referendum Wording Welcomed’ The Irish Times (19 September 2012). 70  M Minihan, ‘Who Stands Where on the Children’s Amendment’ The Irish Times (19 September 2012).

Constitutional Change and Interest Group Politics 213 parties and advocacy groups could lead to difficulties in starting a proper debate on the issues.71 Secondly, it led the press coverage to suggest that the government would ‘secure a comfortable majority’ for the proposal because of this support.72 This may have engendered apathy, the public disengaging from an issue that seemed settled in advance. Thirdly, the elite consensus meant that the people were not being asked to check governmental power but instead to legitimise a conclusion already settled on by the elites in political parties, possibly causing further disengagement from the issues.73 Though opposition to the amendment did emerge, the small number of prominent No campaigners, the diversity of their views on the issue, and lack of central organisation made fruitful debate between the Yes and No campaigns difficult. As one source put it to the Irish Times, the problem was that the No campaign was ‘so small and so much on the margins of society’.74 All of these features reduced the capacity of the campaign to be a valuable exercise of deliberative democracy. Overall, the arguments advanced for and against the proposal bore little relationship to the actual proposal. The Yes campaign dealt largely in generalities, emphasising the proper place of children in a caring society and the opportunity to right the wrongs of the past. The No Campaign responded with claims that the proposal would undermine the rights of parents. In turn, the Yes Campaign sought to counter these claims, maintaining that the proposal would not undermine parents. This in turn raised the question of whether the referendum would achieve anything. At the broadest level of generality, the Yes campaign focused on the assertion that the amendment would protect children in a manner more profound than the Constitution did previously. This was put in several different ways. The framing of the proposal as the ‘children’s rights referendum’ implied that constitutional rights of children needed alteration or improvement. Campaigners for the proposal suggested that it would right the wrongs of the past,75 or offer a better level of protection for children generally.76 On other occasions, it was framed as a referendum on whether children should be both seen and heard.77 In context, this argument seemed to be more about children being valued in society rather than a specific reference to the voice of the child in judicial proceedings. These sweeping, platitudinous statements were often accompanied in campaign literature and posters with images of smiling

71 

A Carr, ‘50/50 broadcast rule “problematic”’ The Irish Times (24 September 2012). de Bréadún and M Minihan, ‘Broad Welcome for wording of proposed children’s amendment’ The Irish Times (20 September 2012). 73  See B Kissane, ‘From People’s Veto to Instruments of Elite Consensus: The Referendum Experience in Ireland’ in Setälä and Schiller (n 14) 31. This problem is accentuated where civil society groups join that elite consensus. 74  D de Bréadún, ‘RTÉ executives to outline referendum coverage plan’ The Irish Times (26 September 2012). 75  O O’Leary, ‘Today with Pat Kenny’, RTE Radio 1 (10 October 2012); F Fitzgerald, ‘Speech at Publication of wording of the Children’s Referendum’, 19 September 2012, available at accessed 31 January 2015. 76  Fitzgerald (n 75); F Fitzgerald, ‘TV3 Children’s Referendum Debate’, 31 October 2012. 77 The Labour Party’s poster campaign in favour of the referendum bore this legend; see also A Balbirnie, ‘Amendment will enhance child rights, secure those of parents’ The Irish Times (1 October 2012). 72  D

214  Oran Doyle and David Kenny children or, in one case, the image of a somewhat scared-looking child clutching a teddy bear.78 The umbrella group campaigning on behalf of various children’s organisations was called ‘Yes for Children’. The implicit suggestion was that those in favour of children and child-protection should simply vote Yes. The dominant meme of the Campaign was the incoherent position that ‘children’s rights’ was the way to secure a child’s best interests, simultaneously protecting children from their parents and from the state. This amounted to little more than sloganeering. Tierney notes the danger that ‘unscrupulous elites can hijack the referendum with simplistic campaign slogans, appealing to populist sentiments which ignore the complexity of the issues involved’.79 In this referendum, however, the emptiness of the slogan was mirrored by the emptiness of the referendum proposal itself. The chief allegation of many on the No side was that the amendment was in fact about the state taking children away from parents, resulting in more frequent state intervention in the family after less serious parental failures.80 This was often framed as the proposal giving new rights to children that would clash with parental rights.81 It was also argued that the best interests of the child would lead to the overriding of the wishes and interests of the parents in most or all cases where the state disagreed with them.82 These arguments relied on precisely the same misconception of children’s rights as held by the Yes campaign; they merely differed in normative direction. They exaggerated the scope of the changes proposed, and were sometimes put across in an alarmist fashion.83 In response, Yes campaigners relied frequently on the fact that Article 41 was not being altered, and thus the rights of parents would remain unaffected.84 As one commentator put it, since Article 41 remained intact, the amendment would ‘not diminish parent’s rights; it will enhance children’s rights’.85 Lacking any clear understanding of the constitutional position, this debate between the two sides simply could not be resolved. Instead, it was a case of unverifiable claim and counter-claim, almost calculated to reduce public understanding. The reassurance about parental authority provided by the Yes campaign invited the charge that the referendum would achieve little.86 It was frequently said by Yes campaigners that the passage of the referendum would mean that children were

78  Available at accessed 31 January 2015. 79  See Tierney (n 10) 262. 80  See, for instance, K Sinnott ‘In Your Child’s Best Interests’, Alive!, April 2012; J Waters, ‘Amendment a threat to rights of parents’ The Irish Times (5 October 2012). 81  K Sinnott, ‘State will back kids against parents,’ Alive!, June 2012; J Waters, ‘Say “No” to children’s amendment to protect society’ The Irish Times (2 November 2012). 82  K Sinnott, ‘This is the End of the Family,’ Alive!, May 2012; Waters (n 80). 83  H O’Connell, ‘Sinnott accused of engaging in ‘hysterical hyperbole’ over referendum’ The Journal (28 September 2012) accessed 31 January 2015. 84  See, amongst others, C McGuniness, ‘Vote on children’s rights a statement of our values’ The Irish Times (9 October, 2012); Fitzgerald (n 75). 85  Balbirnie (n 77) 86  O’Shea (n 50) 91 described the lack of change in Article as ‘glaringly avoid[ing] the elephant in the room’, and suggested that it resulted in ‘an incoherent family policy’. Critics of this sort typically suggested the amendment was either redundant or insufficiently radical, while not calling for a No vote. See V Brown, ‘Children’s referendum deficient in many ways’ The Irish Times (26 September 2012).

Constitutional Change and Interest Group Politics 215 given rights in the Constitution for the first time.87 However, it was pointed out— accurately—that the Constitution already protected the rights of children as implied personal rights.88 When challenged on this point, several variations on this argument were used to defend the novelty and necessity of the proposal. First, it was said that this was the first time children would receive independent or express recognition,89 or recognition ‘as citizens in their own right’.90 Sometimes, there were vague suggestions that there would be some marked difference between the sort of personal rights protected by the Constitution before the amendment and the sorts of rights protected after its passage.91 This was not the case, however, and, for the most part, this collapsed into the second variation, the rhetorical effect of change: the Irish people would show the value placed on child welfare by making a statement about children in the Constitution.92 Against this, even some supporters of the change pointed out that this rhetorical effect would seem hollow and of small comfort if legislation to enact changes and resource allocation for children were not similarly prioritised.93 The third variation was to claim that the amendment was needed to give rights to all children equally, and prevent discrimination based on marital status.94 However, this result had already been achieved in respect of constitutional rights by judicial interpretation over 30 years previously in G v An Bord Uchtála while legislative discrimination against non-marital children was a thing of the past. A fourth version stressed the more specific changes that would be made by the proposal, such as the provision for the voluntary adoption of marital children, and provision for the voice of the child to be heard in judicial proceedings.95 The adoption changes had some significance but did not speak to the necessity and usefulness of the broader aspects of the amendment, and hearing the voice of the child could have been achieved by legislative change. Fifthly, emphasis was sometimes placed on change in the threshold for State intervention. Some argued that earlier and more

87  See for example Fitzgerald (n 75): ‘Sub-article 1 will provide, for the first time, a strong affirmation of each individual child’s inherent rights’. 88  H O’Flaherty, ‘We don’t need a referendum to protect our children's rights’ The Irish Independent (11 September 2012); V Browne (moderator), TV3 Children’s Referendum Debate, 31 October 2012. 89  Fitzgerald, TV3 Children’s Referendum (n 76). 90 The Taoiseach, Enda Kenny, framed the proposal this way; de Bréadún and Minihan (n 72). cf O’Rourke (n 25) 174. 91  When challenged on this point by Vincent Browne, moderator of the TV3 debate, the Minister for Children and Youth Affairs said that when words are put in the Constitution, they have a meaning, and therefore it would make a difference and have consequences. She was not specific as to what these were expected to be. TV3 Children’s Referendum Debate, 31 October 2012. 92  Minister Frances Fitzgerald was keen to frame this amendment as a reflection of the values of the Irish people. She stated in the Second Stage Dáil Debate on the 2012 Amendment Bill that ‘[t]he question facing us is simple. Do we believe that the way children were treated in this State represents what we believe to be the values, morals, and ethics of the Irish people?’ Dáil Éireann Debate, Thirty First Amendment to the Constitution Bill, Second Stage, Vol 773 No 5. cf McGuniness (n 84). 93  See, for instance, K Holland, ‘Lonergan says real test will be putting in place necessary resources’ The Irish Times (20 September 2012). 94  Fitzgerald (n 75); Balbirnie (n 77); the Taoiseach, Enda Kenny, was also reported as making this point: Holland (n 93). 95  McGuniness (n 84); Fitzgerald, TV3 Children’s Referendum (n 76).

216  Oran Doyle and David Kenny effective intervention would be possible.96 Others stressed that it would allow for the child’s best interests to be put first.97 However, these arguments rarely offered detailed suggestions of how these changes would play out in practice and, based on our analysis of the referendum proposal above, were not supported by the actual text. In political terms, arguments of this type again left the proposal open to the No campaign’s allegations that the amendment would subordinate the interests of parents and drastically increase the scope of state intervention, a point strongly denied by the Yes side. Many have attributed the difficulty of referendum campaigns in Ireland to the McKenna principles,98 which prohibit the Government from spending public money advocating a yes vote in a referendum. There has been political concern, whenever referendum proposals were rejected, that the Government’s role was unduly restricted by the McKenna decision, engendering a lack of public awareness, leading to proposals being rejected by the people on a precautionary basis. Academics and practising lawyers have also expressed concern about the exclusion of the Government from referendum campaigns.99 For the children’s rights referendum, the Government decided that it should provide its own ‘public information campaign’, in addition to that provided by the impartial Referendum Commission. The Supreme Court declared this campaign unconstitutional in McCrystal v Minister for Children and Youth Affairs, shortly before polling day.100 That the Government felt the need to fund its own information campaign and the Supreme Court’s conclusion that the information campaign was unbalanced both indicate the difficulties that the Government faced in campaigning to persuade people to vote yes. However, even with the benefit of its unconstitutional campaign, the vote carried by a less than impressive margin of 58 per cent: 42 per cent. The difficult campaign stemmed far more from the confusion over the proposal and what it would achieve than it did from any restrictions imposed by the McKenna principles. VI. CONCLUSION

The failure of Ireland’s Children’s Rights Amendment as an exercise in deliberative democracy cannot be attributed to elite control. Rather, it illustrates more generally the difficulties of achieving formal constitutional change to judicial doctrines,

96 ibid. 97 

Balbirnie (n 77). McKenna v An Taoiseach (No 2) [1995] 2 IR 10. 99  G Barrett, ‘Building a Swiss Chalet in an Irish Legal Landscape? Referendums on European Union Treaties in Ireland & the Impact of Supreme Court Jurisprudence’ (2009) 5 Eur Const L Rev 32; B Ruane, ‘The Doherty Case and Issues Regarding the Provision of Information and Funding for C ­ onstitutional Referenda’ in E Carolan (ed), The Constitution of Ireland: Perspectives and Prospects (Bloomsbury ­Professional Press, 2012) 239. 100  [2012] IESC 53. The Supreme Court subsequently rejected a challenge to the referendum on this basis. Jordan v Minister for Children and Youth Affairs [2015] IESC 33. 98 

Constitutional Change and Interest Group Politics 217 particularly through a referendum process, and offers several lessons about how to approach this type of constitutional change. A flawed analysis of the Constitution generated the appealing slogan of children’s rights. This became totemic for the secular-liberal side of the political cleavage, with people lining up to support or oppose it depending on where they stood in respect of that cleavage, rather than their attitude to the underlying constitutional issues. The difficulty of amending the Constitution in order to change judicial doctrine contributed to this failure. Law—particularly judge-made law—is complicated. Formulating new laws to adjust judge-made law is even more complicated. The law rarely, if ever, perfectly addresses the social problem that it is intended to solve, and using constitutional amendment to tinker with judicial interpretations is extremely difficult.101 These difficulties are accentuated where the constitutional legislator is the people. Although formalised consultation processes can enhance the ability of the general population to grapple with constitutional problems (and the Irish Referendum Commission makes a significant contribution in this regard), there are limits to the ability of people to reach considered conclusions where the very meaning of judicial doctrines is itself hotly contested. This is particularly so where the incentive of political actors is not to educate the public on constitutional law, but rather to win the referendum campaign. If the need for and effect of changes cannot be explained to the people in a straightforward way, it is questionable whether constitutional change is worth seeking given the uncertainty about its ultimate efficacy. A lesson we might draw from this is that, sometimes at least, we should be prepared to live with constitutional laws that we regard as sub-optimal rather than engaging in long and protracted campaigns for changes that ultimately achieve little. To be clear, we are not suggesting that it is improper or illegitimate for legislatures or the people to seek to overturn or nuance judicial interpretations of the Constitution. It is rather that when the desired changes require alteration of minute points of judge-made constitutional doctrine, this may be very difficult in practice, and thus may—on balance—not be worth pursuing.102 The problems of public understanding were considerably exacerbated by the misunderstanding of the Constitution’s position on children’s rights that took hold in Irish civic society. By the time of the referendum, it was too firmly embedded to be overcome. The proper place for legal academics in the constitutional reform process is open for debate, but this particular referendum—dealing as it did with a complex area of constitutional law, which required broad understanding of the

101 Dixon is more sanguine than some about the possibility of changing judge made law through formal amendment; Dixon (n 61). Ackerman similarly believes that any desired change in constitutional law can be brought about using formal amendment rules. B Ackerman, ‘Transformative Appointments’ (1987–1988) 101 Harv L Rev 1164, 1180. We do not argue that textual amendments are irrelevant or entirely without use in changing judicial interpretations. We only contend that this is often difficult to achieve, and there may be no way to ensure one’s textual changes will achieve the desired result. This is not an argument in favour of informal change, however, which may suffer from similar difficulties. 102  Different considerations pertain where the change is to the detail of clearly posited rules in the constitutional text, such as the referendum in Ireland in 2011 to allow reductions in judicial salaries. In those circumstances, the challenge of public explanation is far smaller.

218  Oran Doyle and David Kenny constitutional framework and constitutional case law to grasp—surely needed expert input. Despite efforts made by many constitutional lawyers—including the authors of this article—to weigh in on aspects of this issue in the media, this failed to clarify issues in the public mind. Rather, it likely fed a narrative of claim and counter-claim about the effects of the referendum, further clouding public understanding. A second lesson is for legal academics to be careful in their analysis of areas of constitutional law where change is being debated or considered, while being resolute and timely in correcting misunderstandings, and not overestimating their ability to make a helpful contribution to public debate during the campaign proper. The political cleavage was a crucial component in determining the attitude of interest groups to the referendum proposal. They were too quick to support a proposal that, even adopting the flawed antithesis between parents’ and children’s rights, made very little change. A fundamentally misleading signal was thus sent to voters about what the proposal would achieve. In reality, the fact that the proposal was a compromise between competing interest groups—trying to satisfy children’s rights campaigners while not alienating conservative interests—meant that it was difficult to sell convincingly to the public. In trying to please both sides, the government’s proposal ended up confusing people as to what, if anything, was actually being achieved. A third lesson we might draw from this experience is that more attention needs to be paid to the role of interest groups in advocating and promoting constitutional change, and the motivations for those groups seeking change on a constitutional level. It may be that the true purpose of many of those who supported the Children’s Rights Amendment was to effect a symbolic change in the way that children are viewed and valued. There can be no objection to constitutional amendments advancing symbolic aims or seeking to enshrine values in a constitution; constitutions are important sites of national symbolism. However, for symbolic change to happen, it must be understood, and meaningfully deliberated upon. Political actors must be prepared to engage in real debates about values. Instead of framing the amendment as being about symbolism, identity, and values, campaigners for the children’s rights referendum focused on the minutiae of alleged practical changes that were often elusive. Even when symbolism was discussed, it was dwarfed by the focus on micro-management of legal outcomes promised by the referendum. Instead of embracing the fact that issues underlying the referendum were part of a significant and hotly contested debate about values, the value reductionism that was seen in the referendum campaign—that we should vote Yes if we were in favour of Children— avoided any meaningful engagement with exactly what values we should be trying to prioritise and defend. The final lesson of this experience is that when constitutional changes speak to matters of societal values, they should be debated as such, with an open acknowledgement that not all possible interpretations of those values can be foreseen. Without this, that which is perhaps most important will be talked around and lost in the minutiae of policy debate, and the potential for referendums to engage voters with what the Constitution means, and should mean, will be lost.

11 Amendment-Metrics The Good, the Bad and the Frequently Amended Constitution XENOPHON CONTIADES AND ALKMENE FOTIADOU*

I.  INTRODUCTION: DOES THE FREQUENCY OF AMENDMENT RELATE TO CONSTITUTIONAL QUALITY?

C

ONSTITUTIONAL CHANGE IS an enigmatic process. How a constitution changes and why it changes is not always obvious. By contrast, it is easy to find out how frequently a constitution undergoes formal change, that is, its amendment rate. This is perhaps the reason why often much is read into the amendment rate of constitutions, which is correlated to a series of reasons and results. The trend of building theories based on constitutional amendment rates thrives in the field of empirical constitutionalism. Frequency of amendment is considered to signify things about the Constitution per se, often seen merely as a number detached from the historical, political and cultural context in which formal amendment takes place. Within this tendency a continually recurring endeavour focuses on linking the rigidity level of the Constitution to its word length and correlating both to the economy. The latest addition to this literature goes as far as to argue that long constitutions have higher amendment rates and are thus bad constitutions. ‘Bad’ is an indeed harsh epithet to use for a constitution. Can frequent amendment be used to detect bad constitutions? Does a definition of a bad constitution exist and can the amendment rate serve as an indicator of constitutional quality? Does use of strong and absolute characterisations, such as ‘good’ and ‘bad’, contribute anything to the conceptualisation of constitutions? Does the use of such strong and absolute characterisations have anything to contribute to the

*  The authors would like to thank James E Fleming for his insightful response to this paper. We owe debt for insightful comments to Richard Albert, with whom we co-organised the IACL-BC Workshop on Comparative Constitutional Amendment, which was held at Boston College Law School on 15 May 2015 and to the participants at this workshop Mark Tushnet, Yaniv Roznai, Zoran Oklopcic, Oran Doyle, Juliano Zaiden Benvindo, Thomaz Pereira, Joshua Braver, Luisa Fernanda López, Sofia Ranchordás, Jean-Philippe Derosier, and Jurgen Goossens. Thanks also go to the anonymus peer reviewer for many invaluable remarks.

220  Xenophon Contiades and Alkmene Fotiadou conceptualisation of constitutions? Aged documents that defy formal change, but are subject to constant informal modification, do not necessarily qualify as better constitutions than the ones amended according to their own terms by the constitutionally designated amenders. The need for informal change may be viewed as suggestive of shortcomings in the constitutional text. Ultimately, it is flawed to make any assumption about constitutions and constitutional change without making the correlation between formal and informal change. This paper shall explore unexpected and precarious uses of the constitutional rigidity level. In doing so, it shall reveal that there is more in the misuse of constitutional amendment rates than meets the eye, and that the stakes may be high when the conclusions drawn aim at offering criteria to constitutional designers. The first part probes why the longstanding question about the economic effects of constitutions appears to have acquired renewed importance and examines the use of empirical methodology and the persuasive capacity of metrics in explaining constitutional change and in setting constitutional design aims. The second part addresses the assertion that long constitutions harm the economy and are thus bad constitutions, as their frequent amendments suggest, pinning down the fallacies underlying such approaches. Different correlations are explored casting doubts on the neutrality of the empirical finding that long constitutions are bad and it is explained why frequent amendments cannot be used as an indicator of bad constitutional quality. The third and final part explains why cutting out words from the Constitution is not a neutral task and explores criteria of constitutional quality. II.  CONSTITUTIONAL LENGTH AND THE ECONOMY: AN UNEXPECTED RELATIONSHIP

A.  Using Metrics to Evaluate the Quality of Constitutions The relationship between constitutions and the economy has long puzzled constitutional literature and economic scholarship. The global financial crisis brought to light the economic effects of constitutions, but also the impact of economic shocks on constitutions. In that context the way constitutions interact with the economy acquires particular importance because, in case the constitutional text has the power to impact the economy of a country, constitutional designers would be compelled to add a novel criterion to constitution-writing. The assertion that longer constitutions relate to lower levels of GDP per capita and higher corruption, which inescapably makes them bad constitutions, could be the ultimate leximetric approach to constitutional texts.1 The most recent elaboration of this theory is put forth in a paper authored by George Tsebelis and D ­ ominic

1 RD Cooter and T Ginsbourg, ‘Leximetrics: Why the Same Laws are Longer in Some Countries than Others’ (Social Science Research Network, 20 October 2003) accessed 3 April 2015.

Amendment-Metrics 221 Nardi based on empirical evidence from (OECD) countries.2 Constitutional ­‘badness’ according to this approach is confirmed by frequent amendments taking place despite stringent formal amendment rules. The determining factors for characterising constitutions as good or bad, understood mainly as having positive or negative impact on the economy respectively, are the length in words of a constitution and its rigidity level. A lengthy, difficult to formally amend constitution is a bad constitution, which affects negatively the economy and the corruption level.3 The question here is whether this also suggests that long constitutions, which are considered bad constitutions, have an important role in both the way an economic crisis breaks out and in the way it unfolds. During the recent global financial crisis we witnessed how different constitutions responded to it. A series of interacting parameters determine not only constitutional reactions towards a financial crisis, but also the possibility of a constitution operating as a resilience building mechanism.4 In case constitutional word-length had indeed the potential to influence the economy, a simple constitutional antidote would be available. Still, the wide range of constitutional reactions towards the crisis shows that there is no general answer to the question of how constitutions respond to financial crises5 and no explanation based solely on measurable data can be given to the relationship between constitutions and the financial crisis. The use of metrics in constitutional law is not a novel phenomenon.6 Ever since the 1970s political scientists have systematically used empirical methodology to explain constitutional issues. We have now reached a point where empirical constitutionalism has developed into a distinct way of understanding constitutional law based on empirical analysis and the use of metrics.7 The unprecedented accessibility of data on the content of constitutions and the translation of the world’s constitutions in English following the launch of ‘constitute’ is a turning point. There is a potentiality that the facilitation of quantitative analysis shall allow a newly-bred comparative empiricism to take the lead in constitutional law, casting a shadow over the importance of qualitative criteria and normative analysis. Constitutional amendment and levels of constitutional rigidity have been repeatedly subjected to empirical research. Although as a rule empirical research is being conducted with regard to the small ‘c’ constitution,8 a bulk of empirical work has been done aiming to correlate constitutional length and formal amendment rules

2 G Tsebelis and D Nardi, ‘A Long Constitution is a (Positively) Bad Constitution: Evidence from OECD Countries’ (2014) British Journal of Political Science 1; A Montenegro, ‘Constitutional design and economic performance’ (1995) 6 Constitutional Political Economy 161; S Voigt, ‘Explaining Constitutional Garrulity’ (2009) 29 Int’l Rev L & Econ 290. 3  Tsebelis and Nardi (n 2) 25. 4  X Contiades and A Fotiadou, ‘On Resilience of Constitutions: What Makes Constitutions Resistant to External Shocks?’ (2015) 9 ICL Journal 3. 5  X Contiades and A Fotiadou, ‘How Constitutions Reacted to the Financial Crisis’ in X Contiades (ed), Constitutions in the Global Financial Crisis: A Comparative Analysis (Ashgate, 2013) 9. 6  L Epstein and AD Martin, An Introduction to Empirical Legal Research (Oxford University Press, 2014) 352; D Law, ‘Constitutions’ in P Cane and H Kritzer (eds), The Oxford Handbook of Empirical Legal Research (Oxford University Press, 2012) 379, 398. 7 See K Wittington, ‘Constitutionalism’ in GA Caldeira, RD Kelemen and KE Whittington (eds), The Oxford Handbook of Law and Politics (Oxford University Press, 2008) 292–4. 8  Law (n 6) 379.

222  Xenophon Contiades and Alkmene Fotiadou with amendment rates, endurance and durability.9 Among the most interesting ­findings one of those studies has brought forth is the assertion that the longer a ­constitution is, the higher its amendment rate is, and vice versa, shorter constitutions have lower amendment rates.10 In addition, the higher the amendment rate is, the more difficult it gets to distinguish the Constitution from ordinary legislation, that is, frequent amendments undermine the conceptualisation of the Constitution as the higher law.11 Furthermore, important work has focused on the determinants of constitutional endurance. Factors such as the stringency of formal amendment rules, inclusiveness, detail, length, federalism and history have been scrutinised to explain what makes constitutions endure. It has been suggested that more detailed constitutions last longer because more people have stakes in them.12 Details seem to bond more people to the Constitution. Nonetheless, the question of what constitutes a good constitution cannot be answered by reference to its endurance. A long life does not necessarily mean a good or productive life.13 Thus, the quest of pinning down the features that make a constitution good continues. Within the ongoing dialogue about the correlation between constitutional length and GDP per capita Tsebelis and Nardi appear to introduce a new perspective, that of constitutional quality. Their analysis rests on two key empirical observations: first, long constitutions are more difficult to amend and secondly long constitutions are the most frequently amended. Tsebelis and Nardi offer a stunning explanation to this finding. Longer constitutions are amended more often than shorter ones, despite their high level of rigidity, because they are ‘bad constitutions’. The reason why they are bad is that they are restrictive and thus impose a burden on society.14 Thus, ­‘frequent revisions indicate that a constitution is not simply garrulous, but also imposes significant negative costs on society’.15 Amendment is viewed as a corrective function aimed at remedying shortcomings. Amendment is linked to flaws, whereas the amendment of rigid constitutions is perceived as an indicator of the seriousness of these flaws. What’s more, metrics are used to correlate the length of the Constitution to lower GDP per capita and high levels of corruption. Lower GDP per capita is caused by restrictions that deter economic growth. The explanation offered is simple. The longer the Constitution is, the more the restrictions it includes are, while stringent amendment rules render the amendment or removal of such restrictions strenuous.

9  See D Lutz, ‘Toward a Theory of Constitutional Amendment’ (1994) The American Political Science Review 355; see also A Lorenz, ‘How to Measure Constitutional Rigidity: Four Concepts and Two Alternatives’ (2005) Journal of Theoretical Politics 339; BE Rasch/R Congleton, ‘Constitutional Amendment Procedures’ in BE Rasch and R Congleton (ed), Democratic Constitutional Design and Public Policy: Analysis and Evidence (The MIT Press. 2006, Vol 1) 372. 10  Lutz (n 9) 357. 11  ibid 359. 12 Z Elkins, T Ginsburg and J Melton, The Endurance of National Constitutions, (Cambridge ­University Press, 2009) 97. 13 A Hutchinson and J Colon-Rios, ‘Democracy and Constitutional Change’, (2011) 58 Theoria: A Journal of Social and Political Theory 43; see also the classical Jefferson/Madison disagreement (Th. Jefferson. Letter to J Madison, 6 September 1789). 14  Tsebelis and Nardi (n 2) 11. 15 ibid.

Amendment-Metrics 223 The link between constitutional length and corruption, although definite, is less straightforward because it might be a case of bidirectorial causation.16 It is considered possible that constitutional length could imply the attempt of the framers to control corruption through more detailed regulation.17 Still, it is also asserted that longer constitutional texts, by putting in place a wide range of restrictions, encourage rent seeking, since this creates more opportunities for such behaviour. The Constitution, and even more specifically its word length, is considered to be a determining factor of several economic, social and cultural phenomena. It is the Constitution that shapes the political culture and not vice versa, it is the word length of the constitutional text that influences GDP and not vice versa.18 Besides the obvious one-dimensional view of how culture is shaped, which evades questions about the interactions, actors and processes that create shared constitutional cultures, a more vexing question must be dealt with.19 Is it just length that matters or is there is an underlying link between form and substance beneath the ‘long means bad’ assertion? B.  Persuasion through Numbers: Dubious Explanations Numbers in their enormous symbolic capacity appear value-neutral.20 Cutting down the number of words appears to be a neutral goal for constitutional design. Metrics are presented as an infusion of value-free data in the constitutional design dialogue. However, when focus is placed on the specific provisions that should lose constitutional status the neutrality of the word-trimming project is severely tested. For example, all provisions that impose restrictions on economic activity are considered particularly harmful by Tsebelis and Nardi.21 This syllogism would at some point have to follow some consecutive steps. First of all, the pursuit of economic growth would have to be set as the aim of constitutional amendment. Once this was decided, and prioritised over, (or reconciled with) other possible goals of a particular constitutional amendment, reflection on the means to achieve it would have to follow. In case the goal pursued would be the elimination of provisions that potentially harm the economy, it would probably be a good idea to begin by assessing the content of the suspect provisions, considering also the reason they were given constitutional status. Although it is obvious that provisions restricting economic activity or restraining the state do add words to the text (as do all provisions), it is less obvious why they

16 

Law (n 6) 388. X Contiades and A Fotiadou, ‘Does Constitutional length matter?’ (in Greek) (2014) 4 Eφημερίδα Δ ιοικητικού Δικαίου [Administrative Law Newpaper] 419. 18  According to Tsebelis and Nardi (n 2) ‘It is entirely possible that constitutions influence cultures rather than vice versa … we believe constitutional length affects incomes not vice versa’. 19 On how constitutional culture is shaped see J Mazzone, ‘The Creation of a Constitutional Culture’ (2004) 40 Tulsa L Rev 671; H Vorländer, ‘What is “Constitutional Culture”’ in S Hensel, U Bock, K Dircksen and H-U Thamer (eds), Constitutional Cultures (Cambridge Scholars Publishing, 2012) 21. 20  P Legrand, ‘Econocentrism’ (2009) 59 2 U Torondo L J 215. 21  Tsebelis and Nardi (n 2) 11–3. 17 

224  Xenophon Contiades and Alkmene Fotiadou have an inevitable negative impact on GDP per capita or encourage corruption. In that sense, it is not clear why the word-count argumentation takes priority over the endeavour to assess the normative content of each specific provision explaining whether and how it is responsible for malfunctions and shortcomings of the state apparatus and the economy. Or instead, under the empirical cloak lie purely ideological choices put forth as neutral at a time of economic malaise? Characterising the task at hand is helpful. It has been argued that lawyers specialise in persuasion, whereas social and natural scientists aim to make inferences.22 Here empirical data seems to be used as a tool of persuasion rather than inference, since its use aims at setting goals for constitutional design. Numbers being ‘symbols of precision, accuracy and objectivity’,23 boost the persuasive power of constitutional design proposals hiding ‘the difficult choices that go into a count’.24 Methodological choices matter greatly. Can a constitution be characterised as a bad one, harmful to the state and the economy on the basis of metrics and without a profound analysis of its content? Does harm to the economy, even if proved, suffice or even matter to characterise a constitution as ‘bad’? It is noteworthy that the notion of a bad constitution is not commonly used by constitutional scholars: ‘bad’ is indeed a heavy epithet when used to describe a constitution. Is the explanatory theory put forth to back up the conclusion that long constitutions are bad constitutions sound, or does it entail certain logical leaps? Are there alternative readings to the empirical evidence, or any missing variables? Is an underestimation of ‘all others things equal’ possible?25 Furthermore, there is a clear possibility that a cumulative effect is created, reproducing findings of empirical studies and absolving them from doubts. If the finding that long constitutions are bad is used without questioning neither the methodology used, nor the correlations made, it may easily reproduce faulty impressions.26 And what does it mean, in terms of comparative constitutional law, to ignore the existence of striking exceptions to what is asserted? Are examples of lengthy constitutions that work well and operate in countries with high GDP per capita to be simply approached through metrics, or should their characteristics be examined more closely in search of the reason behind the exceptions? The Constitution of Austria ranks among the lengthier constitutions in the world, its amending formula is quite flexible, its amendment rate particularly high, while Austria has a flourishing economy and the GDP per capita is even higher than the German one. The Constitution of Portugal is very long, has a stringent formula,

22 

See L Epstein and Gary King, ‘The rules of inference’ (2002) 69 University of Chicago Law Review

23 

Deborah A Stone, Policy Paradox and Political Reason (HarperCollins, 1988) 136–37.

1,7.

24 Id.

25  On the problems of the use of empirical methodology, see X Contiades/A Fotiadou, ‘The Determinants of Constitutional Amendability: Amendment models or amendment culture’ (2016) 12(1) Eur Const Law Rev 192. 26 For example Hélène Landemore exploring the notion of a good constitution with regard to the Icelandic constitution-writing experiment reproduces the idea that empirical evidence suggests long ­ ­constitutions are worse than short ones. See H Landemore, ‘What is a Good Constitution’ in T Ginsburg and AZ Huq, Assessing Constitutional Performance (Cambridge University Press, 2016) 74.

Amendment-Metrics 225 a very low amendment rate and the GDP per capita in Portugal is relatively low. The German Constitution, which is four times longer than the US Constitution, has a flexible amendment formula, high amendment rate and the GDP per capita is high. The US Constitution is very short and has an extremely demanding amendment formula, while France has a short Constitution and a high amendment rate. Greece has a long Constitution (slightly longer than the German Constitution), a demanding amending formula and a relatively low amendment rate compared to Austria, France and Germany. There is a long list of departures from the scheme elaborated by Tsebelis and Nardi, suggesting that longer constitutions are more difficult to amend, are amended more often, and cause lower GDP per capita and higher levels of corruption. As a result it is difficult to trace a group of countries, where all the dimensions of the Tsebelis-Nardi working hypothesis can be detected in order to study their shared characteristics. Quantitative analysis without reference to qualitative characteristics may prove to be misleading with regard to understanding possible interactions between constitutional features, constitutional change and the economy. More so, when the aim is to draw conclusions about constitutions. The maxim that long constitutions are bad suggests that criteria for labeling a constitution as a bad one exist, and it is crucial to decide whether such criteria can be purely empirical. Furthermore, even where the low GDP and corruption plus a long-difficult to amend-frequently amended-constitution chain can be detected without any link missing, explanations to that vary and may lead to quite different conclusions that those linking word length with amendment rate and poverty. Viewed from the aspect of constitutional law there are several points that merit some attention. For example, it is asserted that ‘the number of provisions actually amended is less relevant than the fact that a supermajority of the country agrees to an amendment in the first place’.27 Hence, extensive revisions are regarded as equally important for drawing conclusions as minor interventions to the text or the amendment of a few provisions.28 All formal changes are treated as of equal importance regardless of their purpose, content, or scope. Beyond the distinction between super-amendments and regular amendments the significance of each amendment relies on multiple factors. The amendment of the Spanish Constitution through a fast track procedure in order to adopt the balanced budget rule and the public debt cut, as a response to the financial crisis, is a characteristic example. The importance of this amendment can be evaluated only in light of its particularities: the aim of the amendment was to convey a strong message to the international markets, the Constitution of Spain is notoriously rigid, and the amendment was made possible through the consensus between the major political parties.29

27 

Tsebelis and Nardi (n 2) 6. would mean, for example, that we must treat in the same way the 2001 revision of the Greek Constitution, which was a sweeping revision with wide consensus, with the 2008 revision that made minor alterations to very few articles and the continuous incremental change of the German and Austrian constitutional texts. 29  AR Robledo, ‘The Spanish Constitution in the turmoil of the global financial crisis’ in Contiades (ed), (n 5) 154–7. 28  This

226  Xenophon Contiades and Alkmene Fotiadou Still, the first question that needs to be addressed is the reason why long constitutions have higher amendment rates. The explanation offered by Tsebelis and Nardi is simple: long constitutions are frequently amended because they are bad. At this point metrics give way and a value judgement is inserted to explain the longstanding empirical finding that long constitutions have higher amendment rates. To begin with, rival explanations appear to be based on more sound reasoning. It has been argued that long constitutions are more frequently amended because they entail more analytical provisions that require constant formal updating, which cannot be effected by way of informal change.30 In other words, the wider the scope of a constitution is, the more adaptation to changing circumstances it needs, since it contains more rule-like norms than standards. Secondly, constitutional length is explicable through a series of parameters such as age, legal culture, inclusiveness, the existence of ethnic and other minorities, and the political and constitutional negotiations that contributed towards the adoption of more analytical rules limiting government and protecting fundamental rights.31 These explanations are not overlooked by Tsebelis and Nardi. They assert however that another parameter is far more crucial: low GDP per capita and high levels of corruption are the result of constitutional length. This happens because long constitutions being bad constitutions cause poverty and encourage corruption. The flaw of lengthy constitutions from this viewpoint is that they impose too many restrictions on the state and on the society. The people, according to this approach, view constitutions as stray jackets. Making constitutions shorter allows the people to escape those constitutional stray jackets. Oddly enough, this approach is constitution-centric since it furbishes the Constitution with exceptional powers beyond its traditional functions (as perceived by constitutional scholarship). The Constitution underlies a series of causalities: the (lengthy) Constitution causes poverty and corruption, the Constitution shapes the political culture (and not the other way round), the Constitution exerts remarkable influence on the economy due to its word length. It is not clear however why those one-dimensional closed causalities should be accepted without questioning whether there exists a third factor that accounts for explanatory variables.32 The omission of very important variables may compromise the result. It is not clear why the two-fold relationships that lie at the heart of this approach are not impacted or determined by other parameters. For example, the possibility that the political culture could be a determinant factor for both the length of the Constitution and the low GDP in a specific country is ruled out for no apparent reason.

30  Lutz (n 9) 237–4, 355; See also J Ferejohn, ‘The Politics of Imperfection: The Amendment of Constitutions’ (1997) 22 Law and Social Inquiry 501; R Dixon, ‘Constitutional Amendment Rules: A Comparative Perspective’ in T Ginsburg and R Dixon (eds), Comparative Constitutional Law (Edward Elgar Publishing, 2011) 96. 31  T Ginsburg, ‘Constitutional Specificity, Unwritten Understandings and Constitutional Agreement’ in A Sajo and R Uitz (eds), Constitutional Topography: Values and Constitutions, (Eleven International Publishing, 2010) 109; Montenegro, (n 2) 161. 32  G King, RO Keohane, S Verba, Designing Social Inquiry: Scientific Inference in Qualitative Research (Princeton University Press, 1994); J Elster, Explaining Social Behavior: More Nuts and Bolts for the Social Sciences (Cambridge University Press, 2007).

Amendment-Metrics 227 Likewise, it would be worth exploring the working hypothesis that constitutional length may in certain countries result from poverty, due to the enshrinement of economic and social rights catalogues in the constitutional text aimed to enhance social protection and combat poverty.33 Inclusion, which is particularly important in poor countries since it generates constitutional faith, is another aspect that may account for constitutional length.34 Corruption and distrust can also provide the explanation for the large(er) number of words, since detailed provisions are deemed necessary to ensure enforceability. Conclusions drawn on the basis of the correlation between constitutional word length and GDP per capita risk ignoring or downplaying other variables, while they may indicate that the ‘post hoc ergo propter hoc’ fallacy has been committed. Still, the weakest link in the ‘long constitutions causing poverty argumentation’ lies elsewhere. A misconception about what constitutional amendment is and how it works undermines the whole venture of arguing in favour of small constitutions. Constitutional amendment is depicted as nothing more than a corrective mechanism at work when overwhelming majorities detect shortcomings in the Constitution: majorities suffer from these shortcomings and thus they amend the Constitution in order to fix problems.35 It is true that formal amendment is used to correct problems in the constitutional text, when these are held responsible for malfunctions. Nonetheless, this is merely one of the functions served by constitutional amendment. Formal amendments allow constitutions to evolve along with an evolving reality. Constitutional maintenance and updating through the amendment process ensure that constitutions are efficient and workable. Most importantly, constitutional amendment enables people to revisit their constitution through the exercise of constituted amending power. Constitutions draw legitimacy through this process, which in turn allows them to perform their functions better. Constitutional amendments may also take place to adjust the constitutional provisions to novel demands set by the participation in supra-national entities such as the EU, to constitutionalise to already accomplished informal changes, or to respond to constitutional jurisprudence in case a dialogical model of communication between law-maker and judge is in place. The constitutional amendment moment is a supreme democratic moment, when the rules of the game can be reconsidered, to the extent that material limits and eternity clauses so allow. In recent years, there is a growing tendency to involve the people in constitutional amendment through processes enhancing deliberation and popular participation. This trend envisages constitutional amendment as a participatory process aimed at creating a bridge between constituent and constituted power.

33 

H Zacher, Sozialpolitik und Verfassung [Social Policy and Constitution] (CF Müller, 1980). Elkins, Ginsburg, and Melton (n 12) 97–9. 35 ‘… the very attempt to amend the constitution indicates that the existing constitution had—in the opinion of overwhelming majorities in the country—serious shortcomings, and that overwhelming ­majorities understood and suffered from these shortcomings. This is a fundamental point of our ­argument: frequent revisions indicate that a constitution is not simply garrulous, but also imposes s­ ignificant ­negative costs on society’. 34 

228  Xenophon Contiades and Alkmene Fotiadou The fragile balance between constitutionalism and democracy is constantly r­eassessed through constitutional amendment, which is an ongoing attempt to reconcile the two. Furthermore, amendments change the way the people, leg­ islators and judges think about the issue at stake. The judicial toolkit changes through ­constitutional amendment and so does the way judicial doctrine is shaped.36 ­Constitutional amendment serves a matrix of functions that lie at the heart of each constitutional order. To adopt a presumption of malfunction or harm correlated to constitutional amendment, through which to understand formal change ignores or by-passes not only the functions, but also the raison d’être of delegated amending power.37 Simply put, constitutions are not amended because there is something wrong with them, and the amendment rate does not suggest that a constitution is bad or harmful. The presumption of badness to account for the need to amend the Constitution is not only a misleading, but also a potentially dangerous way to describe constitutional change. III.  POOR COUNTRIES, LENGTHY CONSTITUTIONS AND HIGH AMENDMENT RATES: IN SEARCH OF AN EXPLICABLE CORRELATION

A.  Correlating History and Political Culture to Constitutional Length The empirical study of the correlation between constitutional length and GDP per capita is not new. In 1995 Alvaro Montenegro associated long constitutions with low GDP per capita.38 Stefan Voigt has been studying the determinants of constitutional length since 2009,39 while he has recently focused with Christian Bjørnskov on the economic effects of constitutional garrulity.40 These studies indeed probe the link between the number of words of the Constitution and GDP per capita. Nonetheless, they explore multi-faceted explanatory schemes and place emphasis on the impact of culture and trust. Tsebelis and Nardi are in continuity with this stream of empirical studies of constitutional length but also in contrast to them as they succinctly pin down the culprit for low GDP and support the idea that a constitution can cause poverty due to its number of words. This claim is intriguing, since it sets a challenge to explore how complex and multi-parameter constitution making and constitutional change are in reality. Furthermore, it raises the question of how necessary is understanding constitutional amendment before using amendment rates as a variable. Which in turn sets issues with regard to the empirical study of ­constitutional amendment.

36 See HK Gerken, ‘The right to vote: Is the Amendment Game Worth the Candle?’ (2015) 1 23 Wm & Mary Bill Rts J 12. 37  See BD Denning and JR Vile, ‘The Relevance of Constitutional Amendments: A Response to David Strauss’ (2002) 77 Tul L Rev 247; R Albert, ‘Constitutional Handcuffs’ (2010) 42 Ariz St L J 664. 38  Montenegro (n 2) 161–9. 39  Voigt (n 2) 290–303. 40  Chr Bjørnskov and S Voigt, ‘Constitutional verbosity and social trust’ (2014) 161 Public Choice 91–112.

Amendment-Metrics 229 A first clue may be sought in history. The early constitutions, among which lie some that have endured and are still alive after few amendments and revisions, were as a rule short constitutions. It is noteworthy that several among them did not include catalogues of fundamental rights. In the early years of constitutionalism, a dominant approach of the Constitution and political legitimacy dictated that the constitutional text should be nothing more than a codification of basic procedural rules, aimed at guaranteeing that the political power is rationalised and subjected to ground rules for resolving political conflicts allowing decision-making.41 The Constitution thus appears to be neutral, neither restricting nor imposing any directions with regard to the content of political decisions.42 Through this prism the Constitution is an ‘instrument of government’ which delineates and organises state power without interfering with political, economic and social conflicts. The distinction between the procedural and the substantive conceptions of governance rests on these premises.43 Gradually, however, since the beginning of the 20th century the tendency to enshrine values within constitutional texts became prevalent. Constitutional documents started to include much more than procedural dictates: fundamental values, directions to the lawmaker and limitations to public policies became part of the text. This teleogical, value-based approach of the Constitution results in lengthier constitutional texts.44 Enshrining robust rights may entail delineating in detail exactly what rights embody.45 Still, the mantra of brevity-abstractness-generality as assets of the Constitution is being constantly repeated in constitutional scholarship.46 Tsebelis and Nardi adhere to the constitutional brevity ideal with a twist. They claim that longer constitutions cause poverty, based on the link between constitutional word length, low GDP per capita and high levels of corruption. We shall attempt here to explore whether a different explanatory theory for the above linkage may be put forth. To begin with, constitutions that have endured since the 18th and 19th century are short, brevity being a characteristic of the era they were enacted. In addition, their amendment formulae are as a rule stringent, the most striking example being that of Article V of the US Constitution, which continuously triggers scholarly discussion about the need for its modification. The question is why did those constitutions endure? Is it because they are good constitutions? The emblematic US Constitution has not escaped severe criticism with regard to its content and the need for its

41  See B-O Bryde,’Verfassungsentwicklung’ [‘Constitutional Evolution]’ (1982) Nomos 17; D Grimm, ‘Verfassungsfunktion und Grundgesetzreform’ [‘Constitutional function and constitutional reform’] in D Grimm (ed), Die Zukunft der Verfassung [The Future of the Constitution] (Suhrkamp, 1991) 313. 42  H Heller, Staatslehre [Theory of the State] (Mohr, 1930); W Hennis, Verfassung und Verfassungswirklichkeit [Constitution and Constitutional Reality] (Mohr, 1968) 372, 374. 43  G Fox and G Nolte, ‘Intolerant Democracies’ (1995) Harv Int’l L J 1. 44  A Hollerbach, ‘Ideologie und Verfassung’ [‘Ideology and Constitution’] in W Maihofer (ed), Ideologie und Recht [Ideology and Law] (Suhrkamp, 1969). 45  Gerken distinguishes two stages in the constitutional enshrinement of rights: passing the amendment and enforcing it. The two are interrelated since the more abstract the amendment is due to the difficulty of passing a more precise wording, the more laborious the enforcement stage becomes. See Gerken (n 36). 46  See AM Samaha, ‘Levels of Generality, Constitutional Comedy and Legal Design’ (2013) 5 U Ill L Rev 1733.

230  Xenophon Contiades and Alkmene Fotiadou amendment is constantly addressed.47 Whittington and McLean’s image of the US Constitution as a zombie shambling ‘onward, completely out of its own context and barely coherent, and yet refusing to die’,48 sums up the dark side of endurance as persistence. Moreover, it is difficult to envisage the US Constitution without reference to the Supreme Court jurisprudence. Critics target a variety of constitutional malfunctions and problems in the US Constitution, while the same is true in other countries with aged documents like Norway and the Netherlands. Can it be said nonetheless, that these long-lived constitutions have been the catalyst to the high standards of living or the low level of corruption in these countries? To explain the constitutional longevity of these documents perhaps we should look for something more than merely signs of good constitutional design, resulting from the choice of optimal rules for organising the polity. Instead, we must focus on the absence of regime changes or internal and external conflicts, that is, the type of major crises that often lead to constitutional replacement. Another crucial parameter is whether mechanisms of informal constitutional change are available, allowing the Constitution to remain relevant through the passage of time.49 Countries that still have in place constitutions enacted in the era of constitutional brevity are countries with longstanding political and institutional stability, which did not experience the type of traumatic events that trigger the exercise of constituent power. These are the exact conditions that provide fertile ground for the economy to flourish. Constitutions thus may have played a role in economic development in those countries, but not a crucial one, that would justify linking directly their brevity with high GDP per capita. The brevity of those emblematic archetypal constitutions has been imprinted in the collective understanding of the definition of constitutions as a virtue. The ability of the people to understand and even memorise the Constitution, or parts of it, is indeed a legitimacy-building factor. Comprehensibility and accessibility of the constitutional text can also be attributed to linguistic characteristics; simple phrases fraught with symbolic meaning are more appealing to citizens than legal language. The implications of constitutional language not only relate to its interpretation, but also to constitutional legitimacy. Nonetheless, in particular contexts detailed constitutions, that resemble more to ordinary law, may be prove to be more functional. A working constitution generates legitimacy even if it is not easily memorisable. In any case, memorability of the text relates to the symbolic function of the ­Constitution and not to its economic impact. A different category of constitutions that requires attention consists of constitutions adopted more recently but are nonetheless relatively short—not as short though as the abovementioned aged constitutional documents. The historical context of their enactment and legal tradition are vital factors for understanding their

47 KE Whittington and J McLean, ‘It’s Alive! The Persistence of the Constitution’, (2002) 11(2) The Good Society 8; JR Vile, Re-Framers (ABC-CLIO 2014). 48  ibid at 8. 49  See X Contiades and A Fotiadou, ‘Models of Constitutional Change’ in X Contiades (ed), Engineering Constitutional Change. A Comparative Perspective on Europe, Canada and the USA (Routledge, 2012) 417.

Amendment-Metrics 231 length. Often such constitutions can be correlated to a consensual culture that allows political co-operation and compromises, while it also creates the ground for strong credible institutions, state efficiency, ethnic homogeneity and effective justice systems.50 These features on one hand help economic growth, while on the other create no need for detailed, lengthy constitutional regulation. Trust in institutions, the legislature and the judiciary is reflected in the Constitution, since enforceability is not pursued through detail. Here again higher GDP per capita cannot be attributed to the Constitution’s length, but to the institutional, political and social context. Laconic constitutional language is attributed to the same reasons as high GDP per capita, it does not increase it. Turning towards the countries with lengthier constitutions in order to explore common characteristics, what can be detected apart from the impact of legal tradition and history is often political polarisation, state inefficiency, clientelism, corruption, lack of confidence in the justice system, citizens’ distrust of political institutions, minority issues and difficulty in achieving social peace. These problems are not caused by the Constitutions. Contrarily, constitutions try to address malfunctions by more detailed provisions, covering a wider scope of issues. This results in more words. Such attempts may prove to be too ambitious. It is also possible that detailed constitutional regulation becomes part of the problem it is supposed to address. Still, nothing suggests that constitutions cause poverty. Constitutional maximalism is a symptom of the political, social and institutional context, which causes higher level of corruption and lower GDP per capita. Let’ s test this working hypothesis on two examples Tsebelis and Nardi use to support their explanatory scheme for the link between constitutional word length, amendment rate, GDP per capita and corruption. Iceland has a short constitution, which has gone through few revisions, although the amendment procedure is far from stringent. By contrast, Mexico has the lengthier constitution among OECD members. The Constitution of Mexico is far longer than the Icelandic Constitution, is revised on an almost yearly basis since its enactment, and has a stringent, demanding amendment formula. It is noteworthy that Mexico was the first country to constitutionally enshrine a detailed catalogue of socioeconomic rights. Tsebelis and Nardi consider Iceland and Mexico characteristic examples of the link between length and amendment rate.51 At first glance the two examples seem also suitable for proving the theory about the constitutional length and GDP per capita being correlated. However, there are striking differences between the two countries with regard to political history and culture. These differences are the determinant factor for their respective wealth and poverty and for constitutional amendment rates. Governmental and political instability, inefficient state mechanisms, wars and internal crises characterise Mexico. Adversely, political stability, efficient administrative mechanisms, trust in institutions, and absence of internal and external crises marked the landscape in Iceland, until 2009 when despite the brevity of the Constitution a

50  See the characteristics of the pragmatic model of constitutional change in Contiades and Fotiadou (n 49) 417. 445. See also Montenegro (n 2) 163. 51  Tsebelis and Nardi (n 2) 7–8.

232  Xenophon Contiades and Alkmene Fotiadou major economic crisis burst out, followed by a constitutional crisis.52 Can Iceland’s prosperity and Mexico’s poverty be attributed to their Constitutions? Had Iceland 50 years ago adopted the text of the Mexican Constitution and vice versa, would their history and level of prosperity be different? B.  When do Constitutions Change? The ways of constitutional change cannot be comprehended through the use of metrics. Empirical analysis offers insight and information on parts of the versatile constitutional amendment landscape. Nonetheless, it appears ill-suited to provide a holistic and coherent explanatory theory. The indices used by empirical constitutional scholarship to pin down the rigidity level of constitutions do not allow, as a rule, the whole picture of constitutional change to emerge. This happens because such indices fail to grasp the causes and the unexpected interactions that determine how constitutional change unfolds.53 The traditional divide between rigid and flexible constitutions based on the stringency of the amendment formula, (including the stages provided for and the involvement of veto players in the process) is also not suitable for understanding the whole matrix of parameters that impact constitutional change. Such parameters allow some very old constitutions to go on living, despite having complex and stringent formal amendment rules that hinder consensual amendment. Political and financial crises impact different constitutions and the way they change differently.54 To hold the Constitution responsible for the poverty induced by a financial crisis would entail a specific conceptualisation of the Constitution and its functions. Formal amendment rules interact with the functions of the Constitution in multiple ways. The use of metrics alone, without reference to constitutional culture and history cannot explain such interactions. Amendment rates may signify many things. The assertion that the missing link, which explains that the correlation between rigidity level and amendment rate is constitutional length55 is questionable. Can we rely on a single, isolated factor to understand patterns of change? The level of constitutional rigidity is determined by a number of independent variables, and does not depend exclusively on their word length nor even on the formal amendment rules in place. Stringent, complex and demanding amendment formulas often sidetrack constitutional change, which is channeled through informal legislative or judicial routes.56 And even that is not a given, but depends on country-specific historical

52  Tsebelis and Nardi (n 2) use evidence from constituteproject.org up to 2006, that is, prior to the global financial crisis. 53  D Oliver and C Fusaro, ‘Changing Constitutions: Comparative Analysis’ in D Oliver and C Fusaro (eds), How Constitutions Change (Hart, 2011) 381. 54  Contiades and Fotiadou (n 5) 51. 55  Tsebelis and Nardi (n 2) 6. 56  D Strauss, ‘Commentary: The Irrelevance of Constitutional Amendments’ (2001) 114 Harv L Rev 1457; KM Sullivan, ‘What’s Wrong with Constitutional Amendments’ in L Seidman and V Sloan (eds), Great and Extraordinary Occasions: Developing Guidelines for Constitutional Change (The Century Foundation Press, 1999) 39.

Amendment-Metrics 233 and cultural factors. What’s more, the amendment formula does not operate in a vacuum, since the degree of difficulty of constitutional change depends on various factors such as the electoral system, judicial identity, and political culture. Having a consensual culture or adversely a culture of distrust exerts immense influence on the way mechanisms of constitutional amendment operate. To be more precise, majoritarian electoral systems not only produce single party governments but also facilitate constitutional amendment due to the seat allocation, since there is no need for the co-operation or consensus of the opposition parties. The enhanced majorities set by amendment formulas are easily achievable and are thus less of a barrier to amendments, which renders the rigidity level significantly lower. By contrast, in proportional electoral systems it may be more difficult to achieve the necessary majorities. Differences between consensus and majoritarian democracies are also crucial with regard to the rigidity level of the Constitution.57 Where consensual decision-making is part of the political and constitutional culture, different patterns apply to the way constitutional change is channeled. This multiplicity of interacting factors has been analysed by constitutional scholarship. It has been argued that the concept of multiple ‘constitutional rigidities’ is much more suitable to express the matrix of interacting parameters that account for the rigidity level of different constitutions, than the vague notion of rigidity.58 This set of parameters comprises of distinct indices of the rigidity level, including several that cannot be approached empirically, that is, they cannot be measured. The multi-faceted approach of the routes of constitutional change suggests that one-dimensional understandings of the reasons behind constitutional change are quite problematic. The Tsebelis and Nardi approach is characteristic: it attributes the amendment of lengthy constitutions to flaws, that is, it rests on the assumption that though amendment overwhelming majorities and political actors acknowledge that the Constitution has serious shortcomings. This view finds an inherent link between flaws, shortcomings and amendment. At the heart of this approach lies an etymological rather than constitutional use of the word amendment. Amendments mend, they fix problems rather than change the Constitution for various reasons. What’s more, any attempt to effect constitutional change relies on different forms of consensus. How long-lasting and broad is the consensus required to pass a specific amendment varies. Therefore it is neither measurable, nor dependent on formal amendment rules. The reasons why an endeavour to pass a particular constitutional amendment may succeed or fail are vastly unrelated to the quality of provisions targeted. Constitutional change engineering cannot be explained on the basis of quantitative empiricism. Path dependence and bidirectorial causation complicate analysis and may easily compromise the results.59 An even greater danger exists that a third factor may be the actual cause explaining the correlation between two variables.

57 A Lijphart, Democracies: Patterns of Majoritarian and Consensus Government in Twenty-one Countries (Yale University Press, 1984). 58  See Contiades and Fotiadou (n 49) 417, 458. 59  Law (n 6) 336.

234  Xenophon Contiades and Alkmene Fotiadou Causation is an elusive concept. Poverty may cause the constitutional lawmaker to enshrine in the Constitution a list of social rights, but it is also possible that both poverty and constitutional maximalism correlated to wordiness are induced by political instability or malfunctions in the state apparatus. Judicial culture is also important. Where judges play an important role in constitutional policy-making through interpretation, constitutional detail may aim at narrowing the scope of judicial intervention.60 Rule-like details often aim at taming the actors of constitutional change. Therefore, in case the lawmaker or the judge claim the leading role in informal constitutional change, the constitutional legislator has strong motivation to restrain them. This endeavour to tame the drivers of informal constitutional change may work either in combination with the attempt to also tame the constitutional legislator through a stringent formula, or alternatively with the choice of a relaxed formula. Detail as a manner of constitution-writing can be thus attributed to a variety of reasons. It can be a holistic constitutional design choice, or it may have a specific target and scope. As a conscious choice aimed at constraining political and constitutional actors, constitutional detail may be used to serve one of the fundamental constitutional functions: the limitation of government through legal rules, that is, constitutionalism. The explanation of these choices is context-based, and can only be found in the political and social environment and the constitutional culture in which constitutions are designed and operate. Multiple variables exist, establishing causations is quite precarious and amendment-metrics shed light on parts of the landscape but are ill-suited to provide the whole picture. Metrics-based causal inference is a dubious (if not a dangerous) way to approach constitutional phenomena whose nature is complex, comprising of parts that can be measured but also of parts that cannot. C. The Greek Example: Testing the Neutrality Amendment Proposals for a Shorter Constitution In 2001 the number of words of the Greek constitutional text was increased by 25 per cent and ten years later the GDP per capita also dropped by 25 per cent approximately. Could there be a deeper causal relationship between the two? ­Tsebelis and Nardi suggest that the Greek Constitution should be rendered shorter as soon as possible. It has to be kept in mind that the whole concept of long and bad constitutions is correlated to poverty, which furbishes it with extra significance in a financial crisis struck country desperately striving to bounce back. In case the theory attributing poverty to constitutional word length stands up to scrutiny, Greece should make it a top priority to revise its Constitution in order to shorten it. The Greek Constitution is indeed lengthy and has a stringent amending formula. Relaxing the amending formula, which is very demanding, would open up the possibility of a quick constitution-lift by way of shortening it. The first step towards a better constitution would therefore target the stringency level of the formal amendment 60 

Contiades and Fotiadou (n 49) 417, 422.

Amendment-Metrics 235 rules and the second would entail cutting down words. Thus, it would be possible to render a ‘bad’ constitution ‘good’ by cutting down its number of words. Then nice things would start happening in the economy. The suggestion made, however, is not a general call for brevity. Specific examples of redundant provisions of the Greek Constitution that should be cut away as soon as possible are given. The suggestions made in line with this approach are illuminating. A characteristic provision considered unnecessary is the provision constitutionally enshrining proportionality. Proportionality had acquired constitutional status in 2001 (Article 25, para 1 Gr Const). The doctrine had already been jurisprudentially derived from the rule of law principle in a landmark case of the Supreme Administrative Court in 1984, and had been used ever since for reviewing restrictions imposed on constitutional rights. The purpose of its explicit adoption in the constitutional text was to render its use more rigorous, full and uniform, and to ensure that it would be applied not only by the judiciary, but by the legislature and the executive as well. Indeed, following the constitutionalisation of proportionality its application became more rigorous, while throughout the financial crisis it turned out to be one of the basic tools for the constitutional review of the crisis legislation. Tsebelis and Nardi, based on the evidence that it was already in use, find its constitutionalisation unnecessary and thus a good candidate for losing constitutional status. It is quite obvious that although enshrining proportionality in the Constitution is not a prerequisite for its use, since it is an internationally accepted doctrine, explicitly taking it out of the Constitution is open to various interpretations. Is it indeed an action recognising that constitutional status adds nothing to its application, which will continue unfettered, or is it a signal of rejection? Can’t it be taken to imply that alternative methods of review should be used? It is hard to imagine ousting proportionality from the constitutional text without repercussions. Furthermore, unless someone is a critic of proportionality it is hard to argue that the explicit addition of proportionality in a constitution renders it worse. In a more context specific analysis, the fact that most constitutional judgements addressing constitutional rights infringements due to the financial crisis use proportionality to conduct very delicate balancing acts, an amendment that would cause proportionality to lose constitutional status would open up a variety of issues that would not even exist had not proportionality acquired constitutional protection in the first place. A second suggestion of how to shorten the Greek Constitution is by letting go of social rights. Social rights are described as non-justiciable aspirational goals. The right to work is used as an example of a non-enforceable goal, which has not aided the countries that constitutionally protect it to totally abolish unemployment. Beyond the one-sided choice to ignore the large bulk of social rights theory,61 it is clear that the core idea of social rights being useless is far from neutral. To consider social rights as an unnecessary addition of words, one must adhere to a very 61  X Contiades and A Fotiadou, ‘Social Rights in the Age of Proportionality: Global Economic Crisis and Constitutional Litigation’ (2012) 10(3) Int’l J Const L 660; C Fabre, Social rights under the constitution (Oxford University Press, 2000); KG Young, Constituting Economic and Social Rights (Oxford University Press, 2014).

236  Xenophon Contiades and Alkmene Fotiadou precise view of non-justiciability. What’s more, this approach poses the question of what would be the case if less words were used to protect social rights, through the enshrinement of the welfare state doctrine as is the case in Germany.62 Would this be less harmful to the Constitution? In a like manner, Article 16 of the Greek Constitution, which prohibits private universities, is described as an obstacle to the amelioration of the quality of higher education through competition, and more importantly as responsible for making the Greek economy less competitive. Consequently, proposed cuts exit the realm of addressing word length expressed by numbers and enter the area of political choices and constitutional theory. There are many arguments in favour of depriving the above provisions of constitutional status, none of which is however numerical. Adversely, even if numerical criteria were applied, the selection of the redundant constitutional provisions would still entail some degree of substantive choices. Examples of constitutional provisions that add words aimed at combating corruption are those establishing public servants’ permanency, and those notorious for its verbosity, detail and incompatibility with EU-law provision of Greek Constitution establishing that ‘the capacity of owner, partner, major shareholder or managing director of a media enterprise is incompatible with the capacity of owner, partner, major shareholder or managing director of an enterprise that undertakes towards the public sector to perform works or to supply goods or services and extended the prohibition over all types of intermediary persons, such as spouses, relatives, or financially dependent persons’. These provisions may add words to the Constitution and in the second case indeed unnecessary ones, still the length added does not cause corruption but is a clear attempt to address it. The above examples can be discussed as bad features of a bad constitution, good features of a bad constitution or good features of a good constitution. Nonetheless, this discussion entails constitutional analysis and value judgements. Nothing suggests that their basic feature is that they add words, and thus have a negative impact on the income of Greek citizens. IV.  IS THERE SUCH THING AS THE ‘IDEAL CONSTITUTION’?

A recurring question for constitutional comparatists is whether there is an ideal amendment formula, which strikes the perfect balance between constitutionalism and democracy giving the correct dosage of continuity and change. A remarkably wide variety of amendment formulas consisting of different blends of procedural and material limits63 exists. This implies that the use of different amendment mechanisms is dictated by the indiosyncratic features of each constitutional order and the particularities of the constitutional document, whose change formal amendment 62 K Lange, ‘Soziale Grundrechte in der deutschen Verfassungsentwicklung und in den derzeitigen Länderverfassungen’ [‘Social Rights in the German Constitutional Evolution and the state Constitutions’] in E-W Böckenförde, J Jekewitz and Th Ramm (eds), Soziale Grundrechte [Social Rights] (Müller, 1981) 49–60; J Lücke, ‘Soziale Grundrechte als Staatszielbestimmungen und Gesetzgebungsaufträge’ [‘Social Rights as Constitutional State Objectives and Constitutional Orders’] (1982) 107 AÖR 15–23, 19. 63  Contiades and Fotiadou (n 49) 417, 431.

Amendment-Metrics 237 rules are built to control. Still, the quest for perfecting amendment procedures is ongoing.64 Seeking the optimal way to structure formal amendment mechanisms generates novel criteria for channeling change. The answers seem to lie in finding the match between mechanisms of constitutional change and the environment in which it is expected to unfold, rather than working on creating a ‘one size fits all’ formula. Likewise, the perfect constitution does not exist. Each polity has its own requirements with regard to the delimitation and allocation of the state power. Nonetheless, criteria for constitutional design can be developed. In creating rules for constitutionwriting normative, conceptual and empirical approaches of the Constitution have a role to play, each offering different conceptual lenses. By contrast, the attempt to make generalisations with regard to the optimal content of a constitution is a dangerous path, leading to a slippery slope. What is remarkable about the Tsebelis and Nardi approach is the attempt to put forth a substantive criterion to assess the quality of constitutions. It is asserted that long constitutions are bad constitutions.65 A clear standard is thus set, dictating a specific aim to constitution-writing. A good constitution is necessarily short. Achieving shortness is a permanent goal. Even in case the circumstances of constitutionwriting resulted in a long, comprehensive and inclusive document,66 the ‘brevity’ criterion dictates that ways to shorten it must be sought. Two things follow this approach. First, it is deemed possible that constitutional design and the evaluation of constitutions can be done without reference to their specific historical, political, institutional and social context. Thus, it is considered feasible to draft and evaluate constitutions in laboratory conditions. Secondly, taking things out of the Constitution is treated as similar to not putting things in the Constitution. Which is not the case. To begin with, whether a constitution is good or bad cannot be evaluated abstractly. Less so, using length as the evaluation criterion. Whether a constitution can be described as a good one does not depend on its word number. The criterion for assessing whether a constitution is good is whether it organises the state power and regulates social co-existence in a given polity legitimately, efficiently and in a workable way. What matters is setting up the right rules for the right game. How wide the scope of the Constitution is, the array of topic it covers, and the degree of constitutional detail can only be explained in reference to the real-world conditions of constitution making and constitutional enforcement. Constitutional legitimacy is an equally important criterion for evaluating a constitution. Whether a constitution inspires faith or fidelity is highly relevant. ­Constitutions are neither designed nor enforced in an imaginary world, and cannot be evaluated in abstracto. Constitutions perform specific functions, practical as well as symbolic, and they can only be judged in light of their ability to perform these functions. Details, which short text proponents often describe as garrulity,

64 S Levinson (ed), Responding to Imperfection (Princeton University Press, 1995); R Albert, ‘The Structure of Constitutional Amendment Rules’ (2014) 49 Wake Forest L Rev 913. 65  Tsebelis and Nardi (n 2) 66  Elkins, Ginsburg, and Melton (n 12) 97.

238  Xenophon Contiades and Alkmene Fotiadou limit ­government power and governmental discretion due to distrust in government, in the judiciary or in the lawmaker, or distrust between the political forces. Constitutionalism in a context of political distrust may be pursued though comprehensive, detailed constitutional provisions. From this aspect, Tsebelis and Nardi do have a point in suggesting that lengthier Constitutions impose more constraints than short ones.67 However, this is not by definition wrong. The constitutional legislator is attempting through the use of detail to address specific issues. If this results in designing workable detailed provisions that can accomplish their goal then those provisions can be characterised as good ones. The criterion is the ability to attain the stated goal, not the number of words involved. A ‘good’ constitution is comprised of ‘good’ provisions, suitable for accomplishing their purpose. It is not obvious why a short constitution, comprised of abstract provisions that leave wide interpretative discretion to the lawmaker and the judge, is the best choice. Furthermore, such deferrals68 to the future may be made by choice, or the maybe merely the best available solution due to negotiation costs.69 The abstract wording of a provision may imply trust in those who will interpretatively bring it to life, or it may result from distrust that does not allow consent on a more precise form, but it is still deemed best to entrench a vague guarantee and hope for the best. This is a strategy that may lead to a future amendment, constitutionalising a more robust (and wordy) version in line with post-entrenchment interpretation of the original version. The only way to evaluate a constitutional provision is within its working environment and through the study of its history and application. Adhering to a divide between good and bad constitutions based on metrics entails a severe danger. Directing all efforts towards abstract design aims distracts from the task of perfecting the functions of a specific constitution through value-driven interpretation70 and pinning down malfunctions caused by interpretation. The very distinction between good and bad constitutions, although emblematic, may be potentially misleading. By describing a constitution as ‘bad’, a divisive taxonomy is created between good and bad constitutions. Not only are the criteria for such a distinction context-dependent and elusive, they also reproduce a mythical conceptualisation of constitutions corresponding to abstract and absolute virtues and vices. As it is becoming increasingly obvious that no default form of constitutionalism exists, since distinct varieties of constitutionalism that share common characteristics but still have considerable differentiations are detectable, it is challenging to make the hypothesis that under each type different features are required of the Constitution. Using simple and clear terms has certain advantages. The notion of the bad ­constitution is strong enough to attract attention, while also characterising a constitution as ‘good’ connotes much more than a functioning constitution. Nonetheless, 67 

Tsebelis and Nardi (n 2). R Dixon and T Ginsburg, ‘Deciding Not to Decide: Deferral in Constitutional Design’ (2011) 9 Int’l J Const L 636. 69  Gerken (n 36). 70  JE Fleming, ‘Fidelity, Change, and the Good Constitution’ (2014) 62 Am J Comp L (forthcoming) (Social Science Research Network, 23 July 2014) accessed 3 April 2015. 68 

Amendment-Metrics 239 a working constitution capable of performing efficiently constitutional functions is not only the key element for a functioning legal order, but it is also a tangible notion subject to precise and clear criteria. The oversimplification stemming from the use of the dichotomy between good and bad constitution overlooks the intricacy of constitutional functions. A pro-development constitutional frame is not necessarily efficient in protecting constitutional rights. Even the criterion of constitutional abstraction interpreted as ‘less restrictive’ leaves open the question: ‘less restrictive to whom?’ Different models of constitutional change exist, and the drivers of change are not the same in each model. Therefore, constitutional detail may be a way to bind the legislature, but also may restrain the judiciary from interpreting the Constitution setting obstacles to the other branches of government. The idea that constitutional brevity is a permanent goal claims to be a neutral proposal, based on the assumption that small constitutions are better, which sets a permanent amendment target of reducing the size of the Constitution. This target would be neutral only if the only criterion would be the word count. In case other criteria are introduced, then we enter the realm of constitutional and political choices. But can any attempt to shorten the Constitution remain neutral? Every word in the Constitution counts. Once a word enters the Constitution it acquires a special normative and symbolic weight. Eliminating words from the Constitution is not a linguistic exercise. Every word deleted from the Constitution, just as every word added to it, has normative content. Removal conveys a message to those who interpret the Constitution, quite possibly a message of disapproval which has to be respected. With the possible exception of genuine constitutional desuetude,71 where removal confirms disuse, removal of words equals removal of normative content. The removal of words from the Constitution impacts the whole matrix of constitutional interpretation. Regardless of interpretative choices, words and meanings in the constitutional text interact and are intepretatively interrelated, so eliminating words from the Constitution cannot but have open-ended and probably severe consequences. Cutting out words is therefore very different from not including them in the constitutional text during its enactment, or by way of amendment. Characterising words as redundant or unnecessary and trimming them away from the Constitution in the name of brevity is fraught with unexpected consequences. V. CONCLUSION

The notion of the good constitution is inescapably context-sensitive. It depends on the ability of a constitution to perform its functions in a given environment. Thus, it depends on deciding the functions required from a constitution and the analysis of the environment in which they are performed. Functions may vary and so do environments. In longstanding democracies different constitutional functions take priority than in post-conflict or more turbulent environments. Models of constitutional change interact with constitutional functions.

71 

R Albert, ‘Constitutional Amendment by Constitutional Desuetude’ (2014) 62 Am J Comp L 641.

240  Xenophon Contiades and Alkmene Fotiadou Amendment-metrics are a growing enterprise. They are a useful tool but may also prove to be a dangerous one. Empirical approaches offer new insight and information on constitutional change. Nonetheless, they cannot substitute the normative choices made each time the amending power is exercised. Furthermore, amendmentmetrics provide no answers with regard to constitutional quality, neither about the Constitution as a whole nor about distinct constitutional provisions. Good constitutions change and so do bad constitutions. And sometimes the line between a good constitution and a bad one is difficult to draw. Whether an amendment ameliorates or worsens the Constitution is a distinct issue and can only be judged in a context specific manner, in light of the matrix of constitutional functions. The use of amendment-metrics cannot turn constitutional change into a value-free neutral exercise, nor can they disguise it as one.

12 Comment on Amendment-Metrics: The Good, the Bad and the Frequently Amended Constitution JAMES E FLEMING*

I.  INTRODUCTION: DOES THE FREQUENCY OF AMENDMENT RELATE TO CONSTITUTIONAL QUALITY?

I

AM GRATEFUL for the opportunity to comment on Xenophon Contiades and Alkmene Fotiadou’s instructive paper, ‘Amendment-Metrics: The Good, the Bad and the Frequently Amended Constitution’.1 They open by asking, ‘Does the frequency of amendment relate to constitutional quality?’. In particular, they ask, ‘Can frequent amendment be used to detect bad constitutions?’. They observe that some empiricists, enamored of the project of ‘leximetrics’, have seized upon ‘amendment rate’ as an empirical datum on which to build theories. For those empiricists, a frequently amended constitution is a bad constitution. Contiades and Fotiadou dispute this claim, especially in so far as its proponents may ‘offer[] criteria to constitutional designers’. More worrisome still, Contiades and Fotiadou observe, some of these empiricists have sought to link the rigidity level of a constitution to its word length and correlate both to the economy. Such scholars assert that long constitutions harm the economy and thus are bad constitutions.2 Contiades and Fotiadou undertake to ‘pin[] down the fallacies underlying such approaches’. In particular, they seek to ‘cast[] doubts on the neutrality of the ­empirical finding that long constitutions are bad’ and to explain ‘why frequent amendment cannot be used as an indicator of bad constitutional quality’. F ­ urthermore, they

*  I prepared this comment for the BC-IACL Workshop on Comparative Constitutional Amendment, held by the Boston College Law School’s Clough Center for the Study of Constitutional Democracy, 15 May 2015. I wish to thank Richard Albert for inviting me to participate. I also want to express my admiration for and appreciation of all that Richard together with Vlad Perdu do through the Clough Center to promote important scholarship and to enrich the greater Boston intellectual community. Thanks finally to my research assistant, Mike DiMaio, for helpful comments as well as to the anonymous reviewer. 1  X Contiades and A Fotiadou, ‘Amendment-Metrics: The Good, the Bad and the Frequently Amended Constitution’, in this volume. 2  Contiades and Fotiadou (n 1) 219–20.

242  James E Fleming challenge the claim of a recent paper, by George Tsebelis and Dominic Nardi, based on empirical evidence from Organization for Economic Co-operation and Development (OECD) countries.3 Tsebelis and Nardi argue that ‘a lengthy, difficult to formally amend constitution is a bad constitution, which affects negatively the economy and the corruption level’.4 My first reaction to Contiades and Fotiadou’s summary of Tsebelis and Nardi’s arguments was bewilderment and incredulity concerning those arguments. In the abstract, it sounds absurd to suggest that a long constitution, as such, is bad, or that a frequently amended constitution, as such, is bad. Furthermore, in the abstract, such an analysis certainly invites a ‘what’s the world coming to’ response—we might wonder whether Tsebelis and Nardi were writing a parody of leximetrics (or amendment-metrics) analysis—just as I sometimes have wondered whether certain law and economics scholars were writing parodies of law and economics scholarship. Contiades and Fotiadou clear-headedly expose some problems with and fallacies in the use of leximetrics or, as they put it, ‘amendment-metrics’. I share their skepticism concerning whether this ‘newly bred comparative empiricism’ can ‘take the lead in constitutional law, casting a shadow over the importance of qualitative criteria and normative analysis’.5 But, on further reflection, I wonder whether Tsebelis and Nardi may be onto something important, at least at a general level. Moreover, I would be less hesitant than Contiades and Fotiadou are about offering criteria for a bad ­constitution or a good constitution. And I would be more open than they are to the possibility that a long, frequently amended constitution may, in certain circumstances, be a bad constitution (or, at any rate, not a good constitution). II.  THE NOTION OF A BAD CONSTITUTION

Contiades and Fotiadou observe that ‘numbers … appear value-neutral’ and that ‘metrics are presented as an infusion of value-free data in the constitutional design dialogue’.6 Fair enough as a criticism of leximetrics scholarship in general. But I would caution them here concerning their critique of Tsebelis and Nardi. Just because Tsebelis and Nardi use numbers does not necessarily mean that they claim to be value-neutral. In fact, it seems that they are not, and do not aspire to be, valueneutral. They do not seem to be putting an ‘empirical cloak’ over ‘purely ideological choices’.7 To the contrary, they seem to be making an avowedly normative argument that it is bad to restrict a government in a detailed way concerning economic arrangements and its ability to respond to economic challenges and crises. Furthermore, they seem to be making the overtly normative argument that it is bad when

3  ibid 2 (criticising G Tsebelis and D Nardi, ‘A Long Constitution is a (Positively) Bad Constitution: Evidence from OECD Countries’ (2016) 46 British Journal of Political Science 457 (available on CJO 2014 doi: 10.1017/S0007123414000441). 4  Contiades and Fotiadou (n 1) 221. 5 ibid. 6  ibid 223. 7  ibid 224.

Comment on Amendment-Metrics 243 government, before addressing an immediate and pressing economic crisis, has to secure an amendment to the constitution, especially through onerous procedures.8 If this is what Tsebelis and Nardi are saying, then there may be something to their argument. At this point, Contiades and Fotiadou ask, ‘Can a constitution be characterized as a bad one, harmful to the state and the economy on the basis of metrics and without a profound analysis of its content?’.9 They suggest that it cannot, and they are surely right in doing so. But again, it seems that Tsebelis and Nardi are making precisely an analysis of a constitution’s content rather than just tallying the metrics: They are criticising certain OECD constitutions for saddling their governments with burdensome, detailed provisions that make it difficult to respond to economic crises. Pressing the point, Contiades and Fotiadou ask, ‘Does harm to the economy, even if proved, suffice or even matter to characterize a constitution as “bad”?’10 Since they put ‘bad’ in skeptical quotation marks, they imply that a constitution’s contributing to harm to the economy does not entail that it is a bad constitution. I would argue that contributing to harm to the economy—at least in a sustained or catastrophic way—certainly does matter to characterising a constitution as bad. For example, if the people of the United States were to amend the US Constitution specifically to adopt Justice Clarence Thomas’s narrow view of Congress’s power to regulate interstate commerce—taking us back to understandings that were repudiated during the Great Depression—that would make the Constitution a bad constitution.11 For doing so would render the Congress incapable of addressing problems arising in a 21st Century national and international economy. I will return to this point below. Contiades and Fotiadou continue: ‘It is noteworthy that the notion of a bad constitution is not commonly used by constitutional scholars’.12 They seem taken aback by Tsebelis and Nardi’s overtly normative judgements about bad constitutions and good constitutions. They find it odd to speak of constitutions being bad. Why would this be so? Now, I am a normative constitutional theorist who primarily writes about US constitutional law. In my corner of the world, we certainly do make normative constitutional judgements about good and bad constitutions: good and bad provisions, good and bad decisions interpreting provisions, good and bad approaches to interpretation, and the like. For example, John McGinnis and Michael Rappaport, in Originalism and the Good Constitution, argue that a supermajoritarian constitution like the US Constitution, which is difficult to amend, is a good constitution, provided that we interpret it according to original methods originalism (generally, discovering and applying the original meaning using the original methods that the founders used and accepted as legitimate).13 In criticism, I argue that original methods

8  9 

See Tsebelis and Nardi (n 3) 13–15. Contiades and Fotiadou (n 1) 224.

10 ibid.

11 See United States v Morrison 529 US 598, 627 (2000) (Thomas J, concurring); National Federation of Independent Business v Sebelius [2012] 132 S. Ct. 2566, 2677 (Thomas J, dissenting). 12  Contiades and Fotiadou (n 1) 7. 13  J McGinnis and M Rappaport, Originalism and the Good Constitution (Harvard University Press, 2013).

244  James E Fleming originalism promotes a bad constitution that would not be worthy of our fidelity. Instead, I argue that what Ronald Dworkin called a moral reading, or what I call a Constitution-perfecting theory, promotes a good constitution: or at least promotes interpreting an imperfect Constitution so as to make it the best it can be.14 Moreover, in US constitutional law and political science at the present time, people openly argue that the US Constitution is bad because it has caused or contributed to political dysfunction; indeed, there is a widespread discourse of constitutional failure.15 And the discourse about dysfunction is not just functional talk; it is normative talk expressing judgements that the arrangements that are not functioning are bad (democratically unjustifiable). Of course, a constitution may be bad for any number of reasons, for example, because: (1) it is dysfunctional; (2) it lacks popular support, or (3) it impedes economic development. It may even be bad for all of these ­reasons combined. Or these factors may conflict among themselves. Instead of speaking reductively of bad constitutions, it would be useful to distinguish these various ways of being a bad constitution. Still, we should not lose sight of the fact that in doing so we would be developing criteria for bad constitutions. Therefore, I would resist the claim that we cannot make judgements about good and bad constitutions. Perhaps Contiades and Fotiadou are not troubled by normative judgements as such—after all, they say we have to engage in normative analysis—but instead by normative judgements cloaked as value-neutral empirical findings. But here, again, I question whether Tsebelis and Nardi claim to make only value-neutral empirical findings. To the contrary, they seem to be making overtly normative judgements—that, in the context of OECD countries, long constitutions and frequently amended constitutions are (or at least are more likely to be) bad constitutions. Contiades and Fotiadou certainly are right that ‘history and political culture ­[correlate] to constitutional length’. Length is not simply an abstract matter of good or bad. They also surely are right that ‘the multi-faceted approach of the routes of constitutional change suggest that one-dimensional understandings of the reasons behind constitutional change are quite problematic’. They further contend: ‘Each polity has its own requirements with regard to the delimitation and allocation of the state power’. To be sure, one size does not fit all. They also state: ‘Constitutions perform specific functions, practical as well as symbolic, and they can only be judged in light of their ability to perform these functions’. Finally, our authors say ‘the notion of a good constitution is inescapably context-sensitive. It all depends on the ability of a constitution to perform its functions in a given environment’.16 As an outsider—a US normative constitutional theorist who is not primarily a comparative constitutional law scholar—let me offer several related speculations

14  R Dworkin, Freedom’s Law: The Moral Reading of the American Constitution (Harvard University Press, 1996); JE Fleming, Securing Constitutional Democracy: The Case of Autonomy (University of Chicago Press, 2006). 15  See ‘Symposium: America’s Political Dysfunction: Constitutional Connections, Causes, and Cures’ (2014) 94 BUL Rev 575; JE Fleming, Fidelity to Our Imperfect Constitution: For Moral Readings and Against Originalisms (Oxford University Press, 2015) 167–70. 16  Contiades and Fotiadou (n 1) 239.

Comment on Amendment-Metrics 245 concerning why Contiades and Fotiadou find it odd to speak of constitutions being bad. Perhaps comparative constitutional law, as a field, tends toward relativism, historicism, or functionalism—or at least the idea that, to understand a constitution, we have to view it in its context, in light of its history, and in relation to its functions. And perhaps comparative constitutional law scholars shy away from big normative pronouncements about bad constitutions and good constitutions because they are relativist, historicist, and functionalist. Here I fear that a comparative constitutional law relativism, historicism, or functionalism (as reflected in the passages quoted in the previous paragraph) has blinded our authors to the desirability of having normative criteria for what is a good constitution and what is a bad constitution. ‘Nonetheless’, Contiades and Fotiadou argue, ‘criteria for constitutional design can be developed’. They explain: ‘In creating rules for constitution-writing, normative, conceptual and empirical approaches of the constitution have a role to play, each offering different conceptual lenses’. So far so good. But then they say: ‘By contrast, the attempt to make generalizations with regard to the optimal content of a constitution is a dangerous path, leading to a slippery slope’. Next, they write: ‘What is remarkable about the Tsebelis and Nardi approach is the attempt to put forth a substantive criterion to assess the quality of constitutions’.17 I just don’t see the dangerous path, I don’t get the slippery slope, and I don’t see what is remarkable about putting forward a substantive criterion to assess the quality of constitutions. This would seem like a dangerous path or a slippery slope only to someone who is overly relativist, historicist, and functionalist. (I am doubly confused when our authors go from criticising Tsebelis and Nardi for pretending to be value-neutral to criticising them for putting forth a substantive criterion for assessing the quality of constitutions.) III.  CRITERIA FOR A GOOD CONSTITUTION

I shall make some observations, growing out of what I shall call the Marshallian tradition in US constitutional law, bearing on the questions of length, amendment rate, and criteria for a good constitution. In Marbury v Madison (1803), Chief ­Justice Marshall wrote: A constitution—that is, ‘The exercise of [the] original right [to establish a constitution]’—‘is a very great exertion; nor can it, nor ought it, to be frequently repeated’.18 I, like Marshall, conceive the establishment of a constitution as ‘a very great exertion’ not to be frequently repeated. For I, like Marshall, conceive a constitution as a charter of abstract principles, powers, and ends, not a code of detailed rules.19 In McCulloch v Maryland (1819), Chief Justice Marshall further argued that the ‘nature’ of a constitution ‘requires that only its great outlines should be marked’, and that it not ‘partake of the prolixity of a legal code’.20 In other words, a constitution 17 

ibid 237. Marbury v Madison [1803] 5 US 137, 176. Fleming (n 15) 10, 37, 62. 20  McCulloch v Maryland [1819] 17 US 316, 407. 18  19 

246  James E Fleming should be a ‘great outline’, not a detailed legal code. He also added, famously: ‘We must never forget that it is a constitution we are expounding’. Well, what is a constitution? Again, it is a great outline, not a detailed legal code. He explained that a constitution is ‘intended to endure for ages to come, and, consequently, to be adapted to the various crises of human affairs’.21 The Marshallian tradition entails several criteria for a good constitution. First, as a very great exertion, it ought not to be frequently repeated. This desideratum applies not only to the making of a constitution in the first place, but also to its amendment. Secondly, as a great outline, a constitution ought not to be interpreted to be, or amended to become, a detailed legal code. Thirdly, if a constitution is designed to endure for ages to come and to be adapted to the various crises of human affairs, that constitution ought not to make it difficult for the government to respond to crises (including economic crises). But I must acknowledge that there is a competing Jeffersonian tradition in US constitutional law. The Jeffersonian argument prescribes letting the people frame a new constitution from time to time, even if we think that the present constitution has not failed but, to the contrary, is successful or at least adequate. Doing so would be an exercise not only in responsible constitutional self-government but also in maintaining the people’s capacities to make and reform constitutions. In recent years, scholars like Sanford Levinson, who has criticised Article V for making it so difficult to amend the US Constitution, have represented this Jeffersonian tradition in US constitutional law.22 In support of Article V, I would make two points. First, I would give two cheers for Article V in a defensive sense, for it has protected the Constitution and its citizens against the recent rash of ‘amendmentitis’ (a term that Kathleen Sullivan has used).23 Numerous illiberal and ill-conceived amendments that would erode basic liberties or limit important powers have been introduced in Congress in recent years: the Flag Burning Amendment, the Balanced Budget Amendment, the Parental Rights Amendment, the Religious Freedom Amendment, the Human Life Amendment, and the Federal Marriage Amendment, to name a few. Fortunately, none has been approved by the two-thirds vote of both houses of Congress required by Article V to submit a proposal to the states for consideration for ratification. Thus, Article V’s requirements have protected the Constitution and its citizens from such measures. Secondly, there is much to be said for Article V in an affirmative sense. As Lawrence Sager has cogently argued, the obduracy of Article V to ready and easy amendment of the Constitution has encouraged and fostered broad interpretation of the Constitution’s rights-protecting and power-conferring provisions.24 It has underscored the character of the Constitution as a ‘great outline’ or charter of ‘majestic

21 

ibid 415. Levinson, Our Undemocratic Constitution: Where the Constitution Goes Wrong (and How We the People Can Correct It) (Oxford University Press, 2006). 23  KM Sullivan, ‘Constitutional Amendmentitis’ 6(23) The American Prospect 20 (Fall 1995). I have developed these two points in Fleming (n 15) 54. 24  LG Sager, Justice in Plainclothes: A Theory of American Constitutional Practice (Yale University Press, 2004) 164, 183–87, 213–19. 22  S

Comment on Amendment-Metrics 247 generalities’25—abstract principles, general powers, and general frameworks and structures—as opposed to a code of relatively specific original meanings (as original expected applications). Thus, Article V has underwritten approaches to constitutional interpretation like those of Ronald Dworkin’s moral reading, Sager’s justiceseeking constitutionalism, and my own Constitution-perfecting approach.26 My argument in support of Article V’s onerous procedure for amendment, I hasten to add, applies in the first instance to US constitutional practice—and it applies in the context of a relatively brief constitution (a great outline, not a prolix legal code). If you will, I go halfway with Tsebelis and Nardi—(1) in favour of a brief constitution (and against a long constitution)—but (2) in favour of a constitution that is hard to amend. And I make that argument because I think that such arrangements make it more likely that we can promote a good constitution through the ways we build it out over time in light of experience and in pursuit of the aspiration to interpret a constitution so as to make it the best it can be, to mitigate its imperfections.27 Finally, in the grand Marshallian tradition of conceiving a constitution as a great outline or general framework, Jack Balkin, in his book, Living Originalism, has propounded three criteria for a good constitutional theory. These might also serve as three criteria for a good constitution. Balkin argues that a good theory should conceive the Constitution as being capable of serving as ‘basic law’, ‘higher law’, and ‘our law’.28 He argues that conventional, ‘non-living’ originalisms fail abysmally on all of these criteria, particularly the second and third. Originalisms fail to show why we should respect the Constitution as ‘higher law’ (as an expression of worthy aspirations) and why we should affirm it as ‘our law’ (as distinguished from viewing it as an authoritarian imposition by people who are long dead and gone). His abstract ‘living originalism’—which conceives the Constitution as a framework of commitments to be built out and redeemed over time—fares better on these criteria. What are the implications of Balkin’s analysis for good constitutional design? First, a good constitution should be capable of serving as basic law: As such, it should focus on fundamentals, not be a detailed legal code, and it should be hard to amend. Secondly, it should also be capable of serving as higher law, as an expression of worthy aspirations: Again, it should commit us to abstract principles, frameworks, and ends, not be a detailed legal code. Finally, a good constitution should be capable of serving as ‘our law’. The people in the here and now should be able to affirm it as their law, rather than conceiving it as the law of the fathers who long ago made our decisions for us and ordered us to follow their commands and expectations.29 All other things being equal, a long constitution that is frequently amended might

25 I allude to Justice Jackson’s famous formulation in West Virginia v Barnette [1943] 319 US 624, 639. 26  Dworkin (n 14); Sager (n 24); Fleming (n 14); SA Barber and JE Fleming, Constitutional Interpretation: The Basic Questions (Oxford University Press, 2007); Fleming (n 15). 27  Fleming (n 15) 62–63. 28  JM Balkin, Living Originalism (Harvard University Press, 2011) 59–73. 29 To be sure, a constitution with any easy amendment rule might seem superior on the third ­criterion—‘our law’—to one with a difficult amendment rule. And this criterion raises further questions about who ‘our’ is.

248  James E Fleming have more difficulty satisfying these criteria for a good constitution than would a ­Marshallian great outline that is hard to amend and that is designed to endure for ages to come. (I hasten to make clear that I am not prescribing universal criteria for all constitutions based on a sample size of one (the US)! My point instead is twofold: (1) to show that it is perfectly intelligible to sketch criteria for a good constitution; and (2) to illustrate that, from the standpoint of the Marshallian tradition, some arguments like those made by Tsebelis and Nardi do not seem as far-fetched as Contiades and Fotiadou suggest.) IV.  THE PURPOSES OF AMENDMENT

Contiades and Fotiadou also object to Tsebelis and Nardi’s evident view of constitutional amendment ‘as a corrective mechanism aimed at remedying shortcomings’, correcting ‘flaws’, or fixing problems.30 I do not understand what is objectionable about such formulations. I always suppose that the main reason for adopting a formal amendment is—in terms of the title of Levinson’s book on the theory and practice of constitutional amendment—‘responding to imperfection’.31 Again, I work within the Marshallian tradition. Why would a people amend a constitution if not to respond to an imperfection, or to attempt to make the Constitution better in some way? To be sure, I can imagine a Jeffersonian saying, a people should amend the Constitution every generation simply to make it their own constitution, without regard for whether the new constitution or amendments adopted are good or whether the old constitution or provisions repudiated were bad. Within the ­Jeffersonian ­tradition, we might argue that a self-governing people living under a constitution need to exercise their muscles for self-government from time to time: just to keep their capacities for constitutional self-government toned. Here the people might amend the Constitution for a reason other than to correct a flaw: they might do so in order to make it, in Balkin’s formulation, ‘our law’. In a similar vein, Sotirios Barber, in his recent book, Constitutional Failure, argues that the most important constitutional virtue of a self-governing people is the capacity to reform a constitution.32 But Barber contemplates that this important capacity to reform a constitution should be exercised precisely in making constitutional reforms to avert constitutional failure and, short of that, to correct flaws or at least to improve the Constitution. What purposes besides correcting shortcomings or responding to imperfections might amending a constitution serve? Our authors acknowledge: ‘[I]t is true that formal amendment is used to correct problems in the constitutional text, when these are held responsible for malfunctions’. ‘Nonetheless’, they contend, ‘this is merely one of the functions served by constitutional amendment’. They list several other

30 

Contiades and Fotiadou (n 1) 227, 233. S Levinson (ed), Responding to Imperfection: The Theory and Practice of Constitutional ­Amendment (Princeton University Press, 1995). 32  SA Barber, Constitutional Failure (University Press of Kansas, 2014) 111. 31 

Comment on Amendment-Metrics 249 functions, but I would interpret all of them as responding to imperfections or as correcting shortcomings. —— ‘Formal amendments allow constitutions to evolve along with an evolving reality’. As I see it, these amendments respond to the imperfection that the Constitution did not correspond to that evolving reality. —— ‘Constitutional maintenance and updating through the amendment process ensure that constitutions are efficient and workable’. That is, these amendments address inefficiency and unworkability to ensure that the Constitution is efficient and workable! —— ‘Most importantly, constitutional amendment enables people to revisit their constitution through the exercise of constituted amending power. Constitutions draw legitimacy through this process, which in turn allows them to perform their functions better’.33 Here our authors track Balkin’s third criterion for a good constitution: that the people conceive it as ‘our law’. Even here, it seems that the people are amending the Constitution to address a problem of legitimacy and, in our authors’ own words, to allow the Constitution to perform its functions better. That sounds to me like amendment is responding to imperfections or correcting shortcomings. V.  THE FALLACY OF CONFUSING CORRELATION WITH CAUSATION

Finally, I believe Contiades and Fotiadou are generally on the mark in pointing out the risks of confusing correlation with causation—in particular, of committing the ‘after this, therefore on account of this’ fallacy when it comes to analyzing the length of constitutions and the difficulty and frequency of amendment. They are certainly right to argue, for example, that it is preposterous to claim that a long constitution, as such, or a frequently amended constitution, as such, would cause poverty. They are certainly right to observe in general that the relation, if there is one, is instead correlation; for example, the same factors that contribute to poverty might contribute to having a long constitution and a frequently amended constitution.34 Contiades and Fotiadou charge Tsebelis and Nardi with committing this fallacy in claiming that ‘longer constitutions cause poverty’.35 Whatever may be the case with Tsebelis and Nardi’s particular argument, I want to generalise the analysis and suggest that there may be something to the line of argument that detailed constitutional provisions regarding the economy—worse yet, detailed constitutional provisions that are difficult to amend—make it more difficult for a government to respond 33 Contiades and Fotiadou (n 1) 227. For insightful analysis of the expressive function of formal c­ onstitutional amendment rules, see R Albert, ‘The Expressive Function of Constitutional Amendment Rules’ (2013) 59 McGill Law Journal 225–81. 34  Contiades and Fotiadou (n 1) 225–27. 35  ibid 229–31.

250  James E Fleming to economic problems and crises. First of all, these detailed provisions themselves restrict government’s freedom of action in responding to the problem or crisis. That sounds like causation to me: in the sense that the restrictions might exacerbate economic harm. Secondly, stringent amendment rules make it difficult to amend the detailed provisions that restrict government’s freedom of action, thus making it even more difficult for government to respond to the problem or crisis.36 Again, that sounds like causation to me: that the stringency might contribute to economic harm. Here, I may expose my parochialism as a US constitutional law scholar. In US constitutional law, aside from libertarians, scholars and judges generally believe that it is bad to interpret the Constitution as imposing thick limitations on governmental regulation of the economy that restricts government’s authority to respond to economic problems or crises. The ghost of Lochner v New York haunts US constitutional law: Judges and scholars vow never to go back to those bad old days when the Supreme Court interpreted the Constitution to embody a libertarian economic theory and accordingly invalidated regulations of the economy, including economic programs during the crisis of the Great Depression.37 (Aside from the libertarians who argue that Lochner in fact was rightly decided and should be revived.)38 For this reason, there may be good grounds for believing that a constitution with detailed restrictions relating to the economy is—for that reason—a bad constitution. And for believing that such a constitution that is difficult to amend is—for the more compelling reason—a bad constitution. And for concluding that such a constitution might cause (in the sense of contributing to or exacerbating), not merely be correlated with, economic harm. To move beyond Tsebelis and Nardi’s particular analysis and invocations of the ghost of Lochner, let me generalise and hypothesise a modest claim about causation, not merely correlation. Let us imagine circumstances in which a country has a very long constitution—one that partakes of the ‘prolixity of a legal code’, rather than the brevity of a great outline, to recall Marshall’s formulation. And let us imagine further that the country’s constitution is very difficult to amend. Finally, let us imagine that the country faces an economic crisis that the political officials feel obligated to address through immediate action. It seems that, in such circumstances, a long constitution, with detailed provisions concerning the economy, including detailed limitations upon governmental action, would be a bad constitution. In these circumstances, it would be better to have a brief constitution—a ‘great outline’—that could more readily be ‘adapted to the various crises of human affairs’, to recall Marshall again. And I daresay that in these circumstances onerous procedures for amendment might make it a worse constitution. It would not be the length as such that would make it bad—as if editing out some words at random would make it better. Nor would it be the onerous requirements for amendment standing alone that

36  I acknowledge that the situation can be exactly the opposite. It may be that a constitution is specific precisely in order to limit the courts’ discretion in interpreting limitations on governmental authority. I am grateful to the anonymous reviewer for raising this possibility. 37  See Fleming (n 14) 7–8, 27–28, 41–43, 55 (analysing [1905] Lochner v New York 198 US 45). 38 See, eg, D Bernstein, Rehabilitating Lochner: Defending Individual Rights against Progressive Reform (University of Chicago Press, 2011).

Comment on Amendment-Metrics 251 would make it bad. It would be the combination of detailed provisions and onerous requirements constraining governmental action that would make it bad. VI. CONCLUSION

Deeper still, it would reflect poorly on the country that it had designed such a constitution in the first place: that it purported to resolve so many economic matters in advance, as a matter of basic law. And that it made it so hard to address economic problems or crises. A practice of frequently amending such a detailed, obdurate constitution might make for an even worse constitution—incapable of serving as basic law, higher law, and our law. On these criteria, a good constitution should be brief rather than prolix, and it should be infrequently amended.

252 

13 Constituting ‘the People’ The Paradoxical Place of the Formal Amendment Procedure in Australian Constitutionalism LAEL K WEIS

I. INTRODUCTION

O

N ONE PROMINENT strand of normative theorising about constitutional amendment, an amendment to a constitution is an episode of higher-­ lawmaking that emanates from ‘the people’ who are so constituted. I will refer to such views as ‘popular sovereignty-based’ theories of constitutional amendment. Here ‘constitutional amendment’ means a change in the substantive content of those fundamental legal norms and principles that govern the validity of ordinary (non-constitutional) law.1 The normative criterion that popular sovereignty-based theories offer for justifying changes of this kind requires identifying how the popular will in its constitutional law-making capacity is manifest. Popular sovereigntybased theories differ, however, in the way that criterion is understood. First, there are theorists who locate the relevant popular will external to a constitution’s formal (ie, entrenched) amendment procedure. Secondly, there are theorists who locate the relevant popular will internal to the formal amendment procedure—and, indeed, as constituted by it. I will refer to these two distinct approaches as ‘popular constitutionalist’ and ‘formalist’, respectively. Until recently, the constitutional theory literature has privileged the popular constitutionalist approach to identifying the popular will in its constitutional law-­making capacity. Such theories describe how constitutional amendment occurs outside of the formal amendment process. This typically involves explaining how values immanent in social movements and other political or cultural developments, mediated through judicial interpretation, lead to the creation of new constitutional legal norms.2

1  In other words, the use of the formal amendment procedure is not decisive. Thus a formal amendment to the text of a constitution that merely codifies a well-established norm (for instance) would not qualify as an ‘amendment’ in this sense. 2  This type of view is most famously associated with the influential work of Bruce Ackerman, see, eg, We the People: Foundations (Harvard University Press, 1995); We the People: Transformations (Harvard University Press, 2001).

254  Lael K Weis In recent years, however, important comparative work on constitutional amendment has demonstrated why the popular constitutionalist approach is dissatisfying.3 In the first place, its emphasis on manifestations of popular will outside of the formal amendment process reflects parochial concerns based on American constitutional practice. The US Constitution’s amendment procedure is generally regarded as too difficult to be an effective method of constitutional change. Accordingly, the defence of constitutional amendment effected by judicial interpretation has been an ongoing scholarly preoccupation. In the second place—and as a result of this ­preoccupation—the approach fails to engage with important questions concerning the design of amendment procedures and practices. As a result, theorists from the popular constitutionalist camp have frequently overlooked the possibility contended for by theorists from the formalist camp: ­ namely, that the formal amendment procedure itself may be better understood as the site for the emergence of ‘the people’ in their constitutional law-making capacity. The objective of this chapter is to critically examine the alternative normative understanding of constitutional amendment that the formalist approach offers by considering its application in the Australian context. Doing so will demonstrate some important attractions of the formalist approach, but it will also reveal some important limitations. As we shall see, the Australian Constitution’s formal amendment procedure occupies a paradoxical place within Australian constitutionalism, which is captured by the role that it has played in the High Court of Australia’s interpretive practice. On the one hand, it is critical for understanding the sense in which the ­Australian people are the source of sovereignty that the formal amendment procedure be understood as the exclusive, or at least most important, site for the emergence of the popular will in its constitutional law-making capacity. This is evident in the High Court’s reticence to draw on contemporary social values as a source of constitutional meaning. On the other hand, the particulars of Australian constitutional culture and amendment practice simultaneously make it seem implausible to regard the formal amendment procedure as the site of popular sovereignty. This is evident in the High Court’s scepticism about the amendment process as a source of constitutional meaning, and the Court’s privileging of the Constitution’s text over manifestations of the popular will. Examining Australian constitutional practice thus helps identify a key site of tension for popular sovereignty-based theories of amendment that adopt the formalist approach to identifying the popular will in its constitutional law-making capacity. Namely, does the normative understanding of the amendment procedure contended for by formalists presuppose an already constituted people wherein sovereignty resides? Or can the amendment procedure itself constitute the people as sovereign?

3  This line of critique has been developed in the recent work of Joel Colón-Ríos, who argues that the normative authority of a constitution turns on whether its formal amendment procedure preserves the constituent power: see Weak Constitutionalism: Democratic legitimacy and the question of constituent power (Routledge, 2012); ‘The Counter-Majoritarian Difficulty and the Road Not Taken: Democratizing Amendment Rules’ (2012) 25 Canadian Journal of Law and Jurisprudence 53.

Constituting ‘the People’ 255 A.  Australia’s Amendment Procedure in Context The proposition that a constitution’s formal amendment procedure is the locus of popular sovereignty may seem like a non-starter from a perspective informed by American constitutional practice. The conventional wisdom is that the procedure prescribed by US Constitution Article V is not an effective method of amendment because it is too difficult to use, even when the popular will demands constitutional change.4 But things look different where a constitution prescribes an amendment procedure that is both reasonably practicable, in the sense that it is neither overly onerous to propose or to approve an amendment, and reasonably democratic, in the sense that it is designed to engage the body politic in matters concerning fundamental norms and values. Under these threshold conditions, might we plausibly regard the amendment process as constitutive of the popular will in its constitutional lawmaking capacity? Australia provides an ideal case for exploring the parameters of an affirmative answer to this question. The Australian Constitution prescribes an amendment procedure that meets the two threshold conditions just described. Moreover, as we shall see, the acceptance of the proposition that the amendment procedure is the locus of popular sovereignty plays an important role in Australian constitutionalism. This section describes the procedure for amending the Australian Constitution and provides context needed for the analysis in the following section. Set out in section 128, the Australian Constitution’s amendment procedure requires that constitutional amendments be proposed by Parliament, and be approved by popular referendum. The proposed amendment must first pass both houses of Parliament by a majority, or twice in one house. Once successfully proposed, the amendment is then submitted to the electors qualified to vote in federal elections for approval. If the proposed amendment is approved by a majority of the electors in the majority of the states, and by an overall majority of all electors, then it becomes law. Although not constitutionally mandated, Australia has had a system of compulsory voting in place for constitutional referendums since 1924.5 This guarantees that the voter turnout for referendums (as for all elections) is very high: typically in excess of 90 per cent of qualified electors.6 Section 128 is thus both reasonably practicable and reasonably democratic in the sense just described. The double-majority decision rules for proposing and adopting amendments are not nearly as onerous as the supermajority decision rules found elsewhere. Moreover, the requirement to use a popular referendum as the method for approval is noteworthy, particularly for a constitution of considerable age. Indeed, although not without critics, today the use of popular referendums is typically

4 This conventional wisdom, too, has been the subject of challenge in recent years: see V Jackson ‘The (Myth of Un-)Amendability of the US Constitution and the Democratic Component of Constitutionalism’ (2015) 13 Int’l J Const L 575. 5  Commonwealth Electoral Act 1918 (Cth) s 128A. Compulsory voting has been in place for general elections since 1918. 6  Australian Electoral Commission, ‘Compulsory Voting in Australia’ (16 January 2006) accessed 3 January 2017.

256  Lael K Weis thought to be among the most important tools available for ‘democratising’ constitutional law-making processes. The requirement for approval of proposed amendments by popular referendum thus appears to signal a commitment on behalf of the Constitution’s framers to the notion that constitutional law-making must engage the popular will. Despite providing an attractive mechanism for keeping what is now a very old Constitution up to date, however, section 128 has not lived up to its promise. The Constitution is now over 115 years old and there have only been eight successful referendums. The amendment rate is about as low as the amendment rate for the US Constitution, but there have been far fewer significant changes to the Australian Constitution. Two of the eight successful referendums codified existing practices, and thus do not strictly qualify as constitutional amendments.7 The last successful referendum was held in 1977 and made only very minimal changes to Australia’s constitutional arrangements.8 The lack of success in using section 128 to amend the Constitution is not, and should not be mistaken for, an endorsement of current constitutional arrangements. The Australian Constitution is widely regarded as out of step with contemporary social needs and values, which reflects both its age and its infrequent amendment. For example, it contains provisions that expressly confer legislative power on the Commonwealth (ie, federal) Parliament to make racially discriminatory laws, and that tacitly accept that persons may be excluded from a state’s voting franchise on the basis of race.9 It is also one of the few contemporary constitutions that lacks a bill of rights, a feature that many find lamentable.10 In recent years there has been a clamouring for a Constitution that better reflects Australian national identity and contemporary social values.11 Outside of the formal amendment process, there have been several judicial interpretations of constitutional provisions that arguably amount to amendments to the Constitution.12 For example, since 1992 the High Court of Australia has recognised 7 These are the 1928 referendum (State Debts), and the 1946 referendum (Social Services). See G William and D Hume, People Power: The History and Future of the Referendum in Australia (UNSW Press, 2010) 200–01. 8 One of the three successful proposals adopted at referendum in 1977 (Senate Casual Vacancies) also qualifies as a mere codification of existing practice and thus also does not strictly qualify as a ­constitutional amendment: ibid 157–58, 200–01. The other two successful proposals set a mandatory retirement age for federal judges at age 70 (Retirement of Judges), and gave electors in Territories the right to vote in referendums (Referendums). 9 Constitution s 51(xxvi), s 25. Both provisions are the subject matter of current debates about ­constitutional amendment involving the formal recognition of indigenous persons: see Final Report of the Joint Select Committee on Constitutional Recognition of Aboriginal and Torres Strait Islander Peoples (June 2015) accessed 1 August 2016; Law Council of Australia, ‘Constitutional Recognition of Indigenous Australians’ Discussion Paper (31 August 2011) accessed 1 August 2016. 10  See M McKenna ‘The Need for a New Preamble to the Australian Constitution and/or a Bill of Rights’ (1997) Parliamentary Research Paper 12/1996-97. 14–24 accessed 1 August 2016. 11 See E Arcioni and A Stone, ‘The small brown bird: Values and aspirations in the Australian ­Constitution’ (2016) 14 Int’l J Const L 60, 76–79. Arcioni and Stone identify current debates about becoming a republic and about the constitutional recognition of indigenous people. 12  Most informal amendment has been through expansive interpretations of Commonwealth l­ egislative power at the expense of the states, resulting in a degree of vertical imbalance in the Australian federation that was clearly not contemplated by the Constitution’s framers.

Constituting ‘the People’ 257 a limited, implied guarantee of freedom of expression in political matters.13 This development has endured notwithstanding the fact that the Constitution contains no such express guarantee and was designed to leave the protection of rights to Parliament. Nevertheless, developments of this kind are exceptional. Constitutional amendment outside of the formal amendment procedure has been heavily constrained by two key factors: first, the limited aspirations and contents of the Australian Constitution, and secondly, the High Court’s extremely formalistic approach to its interpretation. Both factors are critical for understanding the place of section 128 in Australian constitutionalism, the topic explored in the following section. I describe them briefly here by way of background. First of all, in comparative terms, the Australian Constitution is an unusual founding document. This is partly a function of its origins and history. It was enacted in 1900 as an Act of British Parliament,14 and Australia has pursued a gradual, incremental path to independent nationhood.15 Sovereignty didn’t come to reside in Australia until 1986, when the authority of the British Parliament to legislate on behalf of Australia was finally terminated by enactment of the Australia Acts.16 The Constitution’s substantive content also makes it an unusual founding document. The Constitution is best described as a pragmatic charter of government: concerned primarily with establishing the institutional structure and powers of the federal government (known as ‘the Commonwealth of Australia’), and with the division of powers between the Commonwealth and the States. In addition to lacking a bill of rights, it is devoid of the symbolism and reference to fundamental values that one commonly associates with founding documents. The Australian Constitution is a highly legalistic, technical document that has been invariably described as ‘a prosaic document expressed in lawyer's language’17 and ‘inaccessible’ to ordinary citizens (and sometimes even to lawyers).18 The second critical factor that constrains the prospect of constitutional amendment outside of the formal amendment procedure is related to the first. This concerns the High Court’s established method of constitutional interpretation. Domestically referred to as ‘legalism’, that method is not substantially different from the textualist originalism prescribed by the late Justice Scalia of the US Supreme Court.19 The High Court predominantly relies on ordinary methods of statutory interpretation to discern the meaning of constitutional provisions: the focal point is the Constitution’s 13  Australian Capital Television Pty Ltd v Commonwealth [1992] 177 CLR 106; Nationwide News Pty Ltd v Wills [1992] 177 CLR 1. Irrespective of whether the Court’s recognition of the implied freedom is regarded as legitimate, many would regard the result as a ‘constitutional amendment’ in the sense used here: see, eg, J Goldsworthy, ‘Constitutional Cultures, Democracy, and Unwritten Principles’ (2012) U Ill L Rev 683, 704. 14  Commonwealth of Australia Constitution Act 1901 (UK). 15 See China Ocean Shipping Co v South Australia [1979] 145 CLR 172, 183 (Barwick CJ) (‘The historical movement of Australia to the status of a fully independent nation has been both gradual and, to a degree, imperceptible’); Ch Saunders, The Constitution of Australia: A Contextual Analysis (Hart, 2011) 19–30. 16  Australia Acts 1986 (Cth, UK). 17  Sir AF Mason, ‘The Australian Constitution in Retrospect and Prospect’ in R French, G Lindell and Ch Saunders (eds), Reflections on the Australian Constitution (The Federation Press, 2003) 8. 18 The Hon Justice R Sackville, ‘The 2003 Term: The Inaccessible Constitution’ (2003) 27 UNSW L J 66. 19  See J Goldsworthy, ‘Originalism in Constitutional Interpretation’ (1997) 25 Fed L Rev 1.

258  Lael K Weis text and structure,20 read in light of the objective understanding of its drafters.21 At the same time, the Court has consistently rejected ‘living constitutionalism’, or the view that the meaning of constitutional provisions can evolve over time in response to contemporary social values.22 As I have argued elsewhere,23 the ­mainstream status of this method of interpretation in Australian constitutional practice is best understood as a reflection of the unusual status of the Constitution as a founding document, as just described. Owing to its character and its origins, the Australian Constitution is not a constitution ‘for the people’ in any socially profound sense, but rather a technical legal document that can plausibly be understood as a statute in all relevant respects, differing from ordinary statutes only in its subject matter and the process of its revision or repeal. B.  The Place of the Formal Amendment Procedure in Australian Constitutionalism With this background in place, I now turn to consider the place of the formal amendment procedure in Australian constitutionalism. My objective here is to examine a core set of tensions that attend the status of section 128 as the locus of Australian popular sovereignty and which reveal a paradox. On the one hand, the notion that section 128 carries special normative weight as the locus of the popular will in its constitutional law-making capacity is critical because it resolves (or at least avoids directly confronting) uncertainty about the authority of the Constitution. On the other hand, there are reasons to doubt that section 128 can in fact serve as the locus of Australian popular sovereignty, due to well-established features of amendment practice and constitutional culture. The paradox that attends this set of tensions, I shall suggest, is importantly captured by the role that the amendment procedure plays in the High Court’s interpretive practice. This analysis will ultimately reveal some important reasons for preferring the formalist approach to identifying the popular will in its constitutional law-making capacity to the popular constitutionalist approach. It will also reveal some limitations of the formalist approach. These are considered in the final section. i.  Section 128 as the locus of Australian Popular Sovereignty Despite its infrequent use, the formal amendment procedure occupies an important place in Australian constitutionalism. It is critical to Australian constitutionalism 20  That is, it is an approach committed to ‘textualism’: the view that a written constitution, like any other written law, is nothing more than its text, including presumptions and implications that follow from text and structure. 21  That is, it is an approach committed to ‘anti-intentionalism’: the view that the relevant source of a written constitution’s meaning is the public meaning of the text at the time of its enactment, and not the subjective intentions or expectations of its drafters. 22  It is an approach that is committed, in other words, to what the literature refers to as the ­‘semantic fixation’ thesis. This is the thesis that the language used in a written constitution continues to mean what it meant at the time of enactment: see LB Solum, ‘Semantic Originalism’ (2008) Illinois Public Law Research Paper 07-24/2008 accessed 1 August 2016. 23  LK Weis, ‘What does comparativism tell us about originalism?’ (2013) 11 Int’l J Const L 842.

Constituting ‘the People’ 259 that the formal amendment procedure be understood as the most important site— arguably, even the exclusive site—for the emergence of the popular will in its constitutional law-making capacity. This has to do with the uncertain status of Australian popular sovereignty. It is generally thought to be important to Australia’s contemporary status as an independent and sovereign nation that the source of the Constitution’s authority resides with the Australian people. However such a proposition is complicated by the fact that the Constitution is an Act of the British Parliament. As noted above, the British Parliament had the power to legislate on behalf of Australia until 1986. This means that the legal authority of the Australian Constitution initially resided in its status as an Act of that Parliament, and the principles of parliamentary sovereignty that attend such Acts.24 Thus celebrated Australian jurist Sir Owen Dixon once contrasted the Australian Constitution to the US Constitution in the following terms: ‘It is not a supreme law purporting to obtain its force from the direct expression of a people’s inherent authority to constitute a government. It is a statute of the British Parliament enacted in the exercise of its legal sovereignty over the law everywhere in the King’s domains.’25 It is important not to overstate this point. The picture that Dixon presents is complicated by the fact that the Constitution was drafted by representatives from the colonies in a convention and adopted by eligible electors in a popular referendum. Its Preamble refers to ‘the people’ of the various then-existing colonies who ‘have agreed to unite in one indissoluble Federal Commonwealth’. Nevertheless, the modest aims of federation make it difficult to displace the view that the authority of the Constitution came to reside with the Australian people only sometime after the founding. Nor does ongoing popular acceptance of the Constitution, as evidenced by its considerable endurance, offer a satisfactory explanation for this shift. There are well-known difficulties with grounding legal authority in tacit consent. Moreover, to have any plausibility such an argument would require the existence of a popular constitutional culture. As discussed below, this is something that Australia clearly lacks. Thus while it is now accepted that Australia is an independent and sovereign nation, due to the lack of a central founding moment that served to ‘constitute’ the Australian people, the status of popular sovereignty as the source of constitutional authority remains a matter of some uncertainty and ambivalence.26 It can be accepted that eliminating the power of the British Parliament to legislate on behalf of Australia vested Australia with sovereignty. But who, exactly, are the sovereign ‘Australian people’? How and when were they constituted as such? In the absence of answers to these questions it is unclear how it can be claimed that the legal authority of the Constitution is vested in the sovereign Australian people.

24  See G Lindell, ‘Why is the Constitution Binding? The Reasons in 1900 and Now, and the Effect of Independence’ (1986) 16 Fed L Rev 29, 32–33. 25  O Dixon, ‘The Law and the Constitution’ (1935) 51 Law Quarterly Review 590, 597. 26  See R French, ‘The Constitution and the People’ in French, Lindell and Saunders (eds) (n 17), 60–61; Lindell (n 24) 37–43; Saunders (n 15) 59–65.

260  Lael K Weis The amendment procedure set out in section 128 is thought to resolve this ­puzzle, or at least to permit it to be circumvented, by vesting the Australian people with the sole authority to amend the Constitution. This view is aptly captured by former High Court Justice Michael McHugh in his judgment in McGinty v Western ­Australia,27 a case that required interpreting the constitutional phrase ‘chosen by the people’.28 After noting that Australian popular sovereignty is a somewhat uncertain and contested notion, his Honour observed that: in a country governed by a rigid Constitution … ultimate sovereignty resides in the body which made and can amend the Constitution. On that view, the sovereignty of Australia originally resided in the United Kingdom Parliament. Since the Australia Act 1986 (UK), however, the sovereignty of the Australian nation has ceased to reside in the Imperial Parliament and has become embedded in the Australian people. Only the people can now change the Constitution. They are the sovereign.29

In other words, it is the fact that the Australian people—and only the Australian people—now have the power to amend the Constitution that vests them with sovereignty. Implicit in this understanding is the further premise that the mere existence of a constitutional amendment procedure is not sufficient. It must be plausible to describe the amendment procedure as vesting the power of constitutional amendment in the Australian people. The requirement that proposed amendments be approved by popular referendum is generally regarded as significant in this respect. For instance, the referendum requirement has been described as ‘a clear expression of the sovereignty of the Australian people’ that ‘has come to supersede the sovereignty of the United Kingdom in explaining why Australia’s Constitution is authoritative’.30 The significance of this understanding of section 128 can be appreciated by observing the role it plays in constitutional interpretation. The notion that section 128 is the locus of Australian popular sovereignty figures prominently in debates about interpretive method, and is evident both in the High Court’s interpretive practice and in scholarly commentary. When called upon to defend the High Court’s preferred formalist method of constitutional interpretation, section 128 is routinely invoked as a foundational reason why it is illegitimate for judges to update the meaning of constitutional provisions by reference to contemporary social values. Al-Kateb v Godwin,31 a notorious case concerning the government’s powers of detention, provides a good illustration. Under the Migration Act 1958 (Cth), unlawful migrants are required to remain in immigration detention until one of two events occurs: either they are granted a visa allowing them to remain in Australia, or they request to be returned to their country of origin. The plaintiff, an unlawful migrant to Australia, faced the prospect of indefinite detention after his visa application was rejected because he was a stateless person (a possibility not contemplated by 27 

McGinty v Western Australia [1996] 186 CLR 140. s 7, s 24 (relevantly providing that members of the Senate and the House of ­Representatives, respectively, be ‘directly chosen by the people’). 29  ibid 230. 30  Williams and Hume (n 7) 18 (emphasis added). 31  Al-Kateb v Godwin [2004] 219 CLR 562. 28 Constitution

Constituting ‘the People’ 261 the legislation). The High Court was urged to imply a constitutional limitation of ­substantive fairness—effectively, a right to due process—qualifying the government’s powers of detention. A majority rejected this, holding that although there was good evidence of public support for an implied right of this kind, contemporary social values are an impermissible source of constitutional meaning. The majority’s refusal to update the meaning of the Constitution in light of contemporary social values provoked a lively dissent by Justice Kirby. Justice Kirby is one of the High Court’s most famous dissenting voices and has consistently advocated for a progressive approach to constitutional interpretation both on and off the bench.32 In Al-Kateb, he criticised the majority for failing to give effect to a manifest change in popular will—as evidenced in particular by norms in international law— and he held that judges both can and ought to imply new legal norms from the text of the Constitution on that basis.33 Justice McHugh defended the majority’s approach against this set of criticisms. He conceded that interpreting the Constitution to find an implied right of due process would likely reflect the preferences of ‘a great many’ contemporary Australians.34 However, he insisted, new legal norms are ‘not to be inserted into our Constitution by judicial decisions’ but ‘by persuading the people to amend the Constitution’.35 His Honour therefore rejected Justice Kirby’s contention that ‘the Constitution provides both for formal amendment and judicial reinterpretation’,36 and that ­ both methods of amendment are equally legitimate. In Justice McHugh’s view, the latter amounts to the naked judicial usurpation of the popular will, ‘amending the Constitution … without the authority of the people acting under s 128 of the Constitution’.37 Commentators who defend the High Court’s formalist approach to constitutional interpretation have also relied upon the notion that section 128 is the locus of popular sovereignty. Critics of interpretive formalism hold that it is sometimes necessary for judges to update the meaning of constitutional provisions in light of contemporary social values in order to give effect to the principle of popular sovereignty. To resist progressive interpretation when the popular will has manifestly changed, they argue, is to displace popular sovereignty with a principle whereby the present generation is governed by the ‘dead hand’ of the past. From this perspective, the formal amendment procedure is both an unnecessary and undesirable constraint on the manifestation of the popular will for constitutional law-making purposes (at least some of the time). 32  For exemplary statements of interpretive method in his Honour’s judicial writings, see Al-Kateb ibid, 624 [175]; Grain Pool of WA v Commonwealth [2000] 202 CLR 479 522-25 [110]–[118]; Kartinyeri v Commonwealth [1998] 195 CLR 337, 417–18 [166]; Newcrest Mining (WA) Ltd v Commonwealth [1997] 190 CLR 513, 657–58; Re Wakim; Ex parte McNally (1999) 198 CLR 511, 599–600 [186]. For the best extra-judicial statement of his Honour’s views on constitutional interpretation, see M Kirby, ‘Constitutional Interpretation and Original Intent: A Form of Ancestor Worship’ (2000) 24 Melbourne U L Rev 1. 33  Al-Kateb (n 31) 625–26 [180]. 34  ibid 568 [48]. 35  ibid 595 [73]. 36  ibid 625 [178]. 37  ibid 589 [62], 592 [68].

262  Lael K Weis In response to this line of criticism, a core contention of defenders of formalist approaches to interpretation is that a constitution’s formal amendment procedure is not a constraint on popular sovereignty, but rather constitutive of it. This is a recurrent theme in the scholarly work of Professor Jeffrey Goldsworthy, one of ­Australia’s most prominent constitutional theorists, and a leading figure internationally in the defence of a ‘modest originalism’.38 As Professor Goldsworthy explains, ‘[t]hose who are restricted by the exclusivity of [the amendment] procedure are government officials … who might otherwise have had the power to alter the constitution themselves’.39 Thus in his view, the ‘strongest normative argument for originalism … appeals to … “[t]he people” … who have exclusive legal authority to change their constitution’ and ‘[t]he most powerful objection to non-­originalism is that it usurps their power’.40 In other words, because the formal amendment ­procedure is the locus of popular sovereignty, constitutional law-making outside of that procedure—whether by judges or other government officials—is a usurpation of popular sovereignty.41 ii.  Section 128 as the Site of Ordinary Politics So far we have seen that section 128 occupies a prominent place in Australian constitutionalism as the locus of popular sovereignty. I now turn to a core set of difficulties that beset that contention, making it unclear how section 128 can in fact plausibly be understood as the site for the emergence of the popular will in its constitutional law-making capacity. As I shall suggest, once the particulars of Australian amendment practice and constitutional culture are taken into consideration, it would seem that section 128 is at best a site of ordinary politics—and not even a particularly inspiring picture of this. There are two sets of considerations that weigh against the notion that the formal amendment procedure is the locus of Australian popular sovereignty. The first set of considerations goes to practice. The concern here is that the amendment process has been captured by parliamentary politics, driven by the partisan objectives of political parties and the ambitions of politicians. Bipartisan support for proposed amendments is rare, and most referendum campaigns reflect less an attempt to engage ‘the people’ in their higher law-making capacity, than simply an extension of the partisan debates between the Government of the day and the Opposition. As a result, there is no good reason to think that the politics of amendment are constitutional politics in any robust sense.

38 See Goldsworthy (n 19) 38–9; J Goldsworthy ‘Interpreting the Constitution in its Second C ­ entury’ (2000) 24 Melbourne U L Rev 677, 683–87; J Goldsworthy, ‘The Case for Originalism’ in G Huscroft and B W Miller (eds), The Challenge of Originalism: Theories of Constitutional Interpretation (Cambridge University Press, 2011) 51–57. 39  Goldsworthy, ‘The Case of Originalism’ (n 38) 57. 40 ibid. 41 I develop these points about the role of amendment in the defence of formalist approaches to ­interpretation in detail in LK Weis, ‘Constitutional Amendment Rules and Interpretive Fidelity to ­Democracy’ (2014) 38 Melbourne U L Rev 240.

Constituting ‘the People’ 263 Commentators have identified several features of amendment practice that account for this state of affairs, but there are two general themes. The first is the method of proposing amendments. Parliamentary initiation increases the likelihood that proposed amendments will reflect the aims and ambitions of the governing party, or at least be viewed as such.42 The second theme is the method of engaging the public leading up to the referendum. The Yes and No cases are typically run by the Government and the Opposition, respectively, and driven by short-term political agendas.43 The focus ‘is on winning the argument at all costs, not on informing people’ which ‘fosters partisanship, exaggeration and misinformation’.44 Pathologies inherent in both features of amendment practice are exacerbated when proposals concern technical or ‘lawyerly’ topics, as they often have. Proposals on such topics are highly susceptible to partisan manipulation because their significance is difficult for ordinary citizens to grasp.45 But even proposals on topics that command great public interest, where citizens are capable of independent reflection and evaluation, frequently fare little better. They too solicit suspicion of hidden, partisan agendas and fail to produce the kind of deliberation that would warrant describing the amendment process as the locus of popular sovereignty. The 1988 Referendum provides a good example. Four proposals were put forward by the Labor Government, led by Prime Minister Bob Hawke: 1.  2.  3.  4. 

the imposition of four year terms for both houses of Parliament; the entrenchment of a ‘one vote, one value’ principle; the recognition of local government; and the extension of rights and freedoms guaranteed at the Commonwealth level to the State level.

All four proposals had been topics of discussion by the Constitutional Reform Commission, which was undertaking an extensive review at the time.46 The proposals thus had the capacity to overcome the difficulties associated with parliamentary initiation.47 The proposals also had the capacity to command bipartisan support. Indeed, several changes would have advantaged the parties forming the Opposition (a Liberal-National Party coalition) and disadvantaged the Labor Party.48 Polling conducted at the outset suggested that at least two thirds of the public supported the proposals.49

42 Williams and Hume (n 7) 38, 246–47. Many suggestions for reform in this area focus on how ­arliament formulates proposed amendments, prior to formally submitting them for referendum, P ­recommending the use of conventions convened for this purpose or other non-partisan bodies. 43  C Saunders, ‘The Parliament as Partner: A Century of Constitutional Review’ (2000) Parliamentary Research Paper 3/2000-01, 27–28 accessed 1 August 2016; Williams and Hume (n 7) 63–4. 44  Williams and Hume (n 7) 261. 45  Saunders (n 43) 28. 46  Final Report of the Constitutional Commission (Australian Government Publishing Service, 1988). 47  E Campbell, ‘Southey Memorial Lecture 1988: Changing the Constitution—Past and Future’ (1989) 17 Melbourne U L Rev 1, 7–8. Campbell notes that all but the first proposal ‘were substantially ones emanating from the Constitutional Commission’. 48  Williams and Hume (n 7) 170. 49  The Weekend Australian (Canberra, 3–4 September 1988) 2.

264  Lael K Weis Despite this promise for constitutional reform, the referendum was an unprecedented failure, with three of the proposals rejected by the largest margins ever. This failure was the result of a powerful No campaign waged by the Opposition on political rather than ideological grounds. As George Williams and David Hume describe in their recent study of referendums in Australia, the Opposition’s decision to oppose the proposals was a strategic move by then Liberal Party leader John Howard in order to advance his political career.50 Howard was concerned about retaining his leadership, as his party had recently lost an election and the polls showed that he was far less popular than Prime Minister Hawke. He ‘identified that, if the 1988 referendums failed, this might signal a turning point in the Coalition’s fortunes’.51 The No campaign seized on critical errors made by the Government in rushing the amendment process. The proposals had been put forward before the Commission’s Final Report was ready, and the Government did not effectively explain or defend the proposals to the public.52 This allowed the No campaign to caricature the proposals as power-grabbing by the Labor Government.53 Advertisements warned that ‘Canberra wants you under its thumb’, and that a Yes vote would ‘let Hawke win his blank cheque referendum!’54 There was little discussion of the substance or merits of the proposed changes.55 Pre-referendum polling showed that nearly half of voters didn’t understand the proposals, or even know what they were about.56 In short, as this example helps illustrate, the formal amendment process as it is currently practiced does not appear to provide suitable conditions for constitutional law-making. Indeed, it is doubtful whether it even provides suitable conditions for ordinary law-making. This makes it difficult to contend that section 128 is the locus of popular sovereignty. A second consideration that creates difficulties for this proposition goes to Australian constitutional culture, and in particular the uncertain ‘constitution’ of the Australian people by the Constitution. The concern here is that even if the amendment process could be reformed to create a site for constitutional politics, it is unclear that ‘the people’ would rise to the occasion given the absence of a popular constitutional culture. The Constitution plays a limited role in Australian public life. As leading commentators have observed, ‘[t]o an extent that would surprise many outside observers, the Australian Constitution is not understood to be a repository of shared values, is not thought to contain fundamental principles to which the citizenry agree or aspire and does not frame public debate’.57 There are several reasons for this. To begin with, ordinary Australians notoriously have poor knowledge about the Constitution, and even low-levels of awareness that Australia has a written constitution.58 Recent civics education studies indicate that

50 

Williams and Hume (n 7) 172–74. ibid 174. 52  Campbell (n 47) 8–9; Williams and Hume (n 7) 168–69. 53  Williams and Hume (n 7) 174–75. 54  ibid 176. 55  Campbell (n 47) 9. 56  Williams and Hume (n 7) 178. 57  Arcioni and Stone (n 10) 60. 58  Civics Expert Group, Whereas the People: Civics and Citizenship Education (1994). 51 

Constituting ‘the People’ 265 there are both low levels of constitutional knowledge among school students and low levels of teacher confidence with constitutional law topics.59 These studies also indicate that students leave school with only the most basic understanding of constitutional change: although most know that ‘a referendum’ is a process used to amend the Constitution, only 3 per cent understand the significance and implications of amendment.60 Although this goes some way toward explaining Australia’s lack of a popular constitutional culture, there is a deeper explanation. This is based on features of the Constitution itself, which have been discussed above. Two features bear noting: first, the absence of a central founding moment defining fundamental Australian values, and secondly, the formalistic character of the Constitution. These features suggest that even if there were higher-levels of public knowledge about the Constitution, it would still be unlikely that members of the public would deliberate about matters concerning fundamental legal norms and values in constitutional terms. Indeed perhaps the reason that civics lessons on constitutional law fall short is that there is too little to inspire interest in the first place: the Constitution is too disconnected from Australians’ self-understanding as a people.61 Cumulatively, the foregoing considerations suggest that the formal amendment procedure does not in general provide Australians with a site for constitutional law-making, and that Australians are not well-composed for that task in any case. Whereas the former points primarily to features of the current practice of constitutional amendment, the latter points to the failure of the Constitution to ‘constitute’ the Australian people in any robust sense. As a result, it is unclear whether section 128 can in fact or even in principle serve as the locus of popular sovereignty—­notwithstanding the prominent place that notion plays in Australian constitutionalism. This uncertainty about the status of section 128, too, is reflected in the High Court’s interpretive practice. The flipside of the normative understanding of the formal amendment process discussed in the previous section is the Court’s lukewarm (and at times, even hostile) treatment of referendums in constitutional interpretation. The Court’s refusal to draw inferences about constitutional meaning from referendums casts doubt on the view that section 128 is the locus of popular sovereignty, or at least seriously qualifies it. Sovereignty appears to reside in the text of the Constitution—not in the Australian people. To illustrate, we can consider the following two examples. The first is drawn from the Work Choices case,62 where the Court was invited to draw inferences from failed referendums in interpreting the scope of the Commonwealth’s power to regulate corporations under the ‘corporations power’.63 The Commonwealth Parliament had 59  S Mellor, K Kennedy and L Greenwood, Citizenship and democracy: Australian students’ k ­ nowledge and beliefs: the IEA Civic Education Study of fourteen year olds (ACER, 2002) xviii, 4, 73, 113, 114 (Table 7.10), 115, 122, 151 (Table BS.2) accessed 1 August 2016. 60  J Fraillon et al, National Assessment Program: Civics and Citizenship Years 6 and 10 Report 2013 (ACER, 2014) 43 accessed 1 August 2016. 61  See Sackville (n 18) 84. 62  New South Wales v Commonwealth [2006] 229 CLR 1 (‘Work Choices’). 63  Constitution s 51(xx).

266  Lael K Weis relied on that head of legislative power to make controversial changes to industrial relations law in areas previously thought to be within the sole legislative competency of the States. In its submissions, the State of Queensland argued that a series of three failed referendums to expand Commonwealth legislative power in order to permit the enactment of legislation of this kind showed that such legislation was beyond the scope of the un-amended corporations power.64 In a joint judgment, a majority of the Court noted that there are ‘insuperable difficulties in arguing from the failure of a proposal for constitutional amendment to any conclusion about the Constitution’s meaning’.65 The majority further observed that the ordinary politics character of referendums casts doubt on the notion they are the site of the manifestation of the popular will in its constitutional law-making capacity. They noted in particular the impact of ‘party politics’ on amendment practice, warning that it is ‘altogether too simple to treat … rejections [of proposed amendments] as the informed choice of electors’.66 Finally, in its knock-down blow to Queensland’s submissions on this point, the majority held that by conferring power upon the federal judiciary to determine constitutional matters,67 the Constitution makes the High Court—and not the Australian people—the final arbiter of constitutional meaning, and that this conferral of judicial power has primacy over the amendment procedure prescribed by section 128.68 In rejecting failed referendums as a source of constitutional meaning, the High Court’s analysis in the Work Choices case is striking in its hostility to section 128 as the locus of popular sovereignty. But perhaps not much can be drawn from this example: after all, the proposed amendments at issue were largely technical changes and thus (arguably) less plausible candidates for the emergence of the popular will in its constitutional law-making capacity in any event. Moreover it is unclear whether rejected proposals ought to be afforded the same weight as approved proposals, particularly given the infrequency of the latter. A second example concerning a successful referendum on a topic with broad popular interest will help confirm this sceptical approach to referendums. This example concerns the most successful referendum in Australian history, which was held in 1967 on the topic of Aboriginal people. The proposed amendment did two things. Firstly, it deleted section 127, which had provided that: ‘[i]n reckoning the numbers of the people of the Commonwealth, or of a State … aboriginal natives shall not be counted’. Secondly, it gave the Commonwealth power to enact ‘special laws’ on behalf of Aboriginal people under the ‘races power’, section 51(xxvi). That section originally provided that: The Parliament shall, subject to this Constitution, have power to make laws … with respect to: … the people of any race, other than the aboriginal race in any State, for whom it is deemed necessary to make special laws.

64 

Work Choices (n 62) 99–100 [125]–[130]. ibid 100 [131]. 66  ibid 101 [132]. 67  Constitution s 76(i). 68  Work Choices (n 62) 101 [134]. 65 

Constituting ‘the People’ 267 The 1967 amendment struck out the phrase ‘other than the aboriginal race in any State’, so that section 51(xxvi) now provides that: The Parliament shall, subject to this Constitution, have power to make laws … with respect to: … the people of any race for whom it is deemed necessary to make special laws.

The 1967 amendment thereby extended the Commonwealth’s power to enact ‘special laws’ to Aboriginal people. The objective of the proposed amendment was ‘[r]ooting out … constitutionalised discrimination’ against Aboriginal people.69 However, whereas section 127 was a blatantly racist provision—effectively denying the status of Aboriginal people as members of the body politic—the races power was (and is) facially neutral as to whether it permits racially discriminatory laws. There is no doubt that the framers of the Constitution intended it to have this operation. There is also no doubt that the amenders of the provision intended the power to extend only to laws benefitting Aboriginal people, and not to racially discriminatory laws. But the provision remains facially neutral. A question therefore arises whether the 1967 amendment overrides the original understanding of the races power, or whether it leaves that understanding intact. If the authority of the Constitution is grounded in popular sovereignty and the locus of popular sovereignty is section 128, then the circumstances of the amendment process seem relevant. The 1967 referendum is regarded as among the most significant moments in Australian constitutional history, and ‘an important part of changing Australia’s conception of itself as a nation’.70 The proposal passed with an unprecedented 90.8 per cent of the vote (with a turnout rate of 93.8 per cent).71 This success is credited both to the decades-long grass-roots campaign for Aboriginal rights that paved the way for the referendum,72 and to the bipartisan campaign, which had support ‘from all quarters’.73 The 1967 referendum thus overcame the limitations of amendment practice and constitutional culture described above, resulting in a genuine moment of constitutional (as opposed to ordinary) politics. Thus for many the proposed amendments contemplated not simply a change in how the census is calculated and a change in the Commonwealth’s power to enact ‘special laws’, but a reconfiguration of the Australian body politic. Nearly 50 years later, the referendum has been described as ‘a nation-building exercise, a symbolic coming together’, which even if imperfectly achieved in practice ‘stands for something very important’.74 The High Court considered the significance of the referendum for the interpretive question posed by the facial neutrality of the races power in 1998 in the case ­Kartinyeri v Commonwealth.75 Kartinyeri concerned the validity of a discriminatory

69 

Williams and Hume (n 7) 141.

70 ibid.

71 Australian

Bureau of Statistics, ‘The 1967 Aborigines Referendum’ Year Book Australia 2004 (Cat No 1301.0, 27 February 2004) accessed 1 August 2016. 72  Winterton and Hume (n 7) 142. 73  ibid 144–45. 74  L Behrendt, ‘The 1967 Referendum: 40 Years On’ (2007) 11 AILR 12, 14, 15. 75  (1998) 195 CLR 337.

268  Lael K Weis law, made pursuant to the amended races power that singled out a particular group of Aboriginal people in order to deny them a benefit that they otherwise would have enjoyed under standing legislation. Only four of the five sitting judges addressed the question.76 Of those four judges, only Justice Kirby—in another vigorous dissenting judgment—held that the amenders’ understanding should prevail. In his Honour’s view, the determinative consideration was that the 1967 referendum was an unequivocal expression of the sovereign will of the Australian people to eliminate constitutionalised discrimination against Aboriginal people. The remaining three judges largely accepted, or at least did not dispute, this characterisation of the 1967 referendum. However they rejected its relevance as a source of constitutional meaning. Irrespective of the significance of the 1967 referendum as a rare manifestation of the popular will in its constitutional law-making capacity, they held that the resulting amendment did not override the original understanding of the races power due to the minimal change made to the text of the Constitution. Justice Gaudron reasoned that the amendment was a ‘bare deletion of an exception or limitation on power’, and that ‘whatever the intention of those voting in the 1967 referendum … the consequence of an amendment of that kind is to augment power’.77 Similarly, Justices Gummow and Hayne reasoned that if the amendment had intended to restrict the Commonwealth’s power to enact ‘special laws’, then the text of the amendment would have included a reference to ‘benefits’.78 Implicit in both sets of reasons is a strong commitment to interpretive formalism, to an extent that arguably displaces the notion that popular sovereignty, located within the amendment procedure, is the controlling principle in understanding the authority of the Constitution. As Justices Gummow and Hayne insisted in their reasons, ‘it is the constitutional text which must always be controlling’, not the ­‘aspiration [of the amenders] … to advance the situation of persons of the ­Aboriginal race’.79 Kartinyeri thus confirms the position taken in Works Choices: sovereignty appears to reside in the text of the Constitution, as interpreted (ultimately) by the High Court, and not in the Australian people. II.  EVALUATION AND CONCLUSION

This chapter has examined an application of the less-studied ‘formalist’ approach to identifying the relevant popular will for the purpose of popular sovereignty-based constitutional amendment theory. By holding that the locus of popular sovereignty lies within the formal amendment procedure, the formalist approach presents an important but less well-understood alternative to popular constitutionalist approaches. As we have seen, the formalist approach has great attraction in the Australian context given the structure of the Australian Constitution’s formal amendment procedure 76 Two judges, Chief Justice Brennan and Justice McHugh, declined to address the constitutional ­ uestion, deciding the matter on the basis of implied repeal and the doctrine of parliamentary supremacy, q ibid 357 [17]. 77  ibid 363 [32]. 78  ibid 383 [94]. 79  ibid 382 [91].

Constituting ‘the People’ 269 and the strength of Australia’s formalist interpretive tradition. Yet, in considering the application of the approach to Australian constitutional practice, we have seen that the Constitution’s formal amendment procedure occupies a paradoxical place. On the one hand, it is critical that section 128 is understood as the locus of popular sovereignty, providing the exclusive site for the manifestation of the popular will in its constitutional law-making capacity. On the other hand, it is simultaneously ­difficult to see how section 128 fulfils this normative function attributed to it. I conclude by considering broader lessons for constitutional amendment theory that can be taken from this analysis. My suggestion is that understanding the paradoxical place of the formal amendment procedure in Australian constitutionalism helps us better evaluate the formalist alternative. In particular, the analysis provided in this chapter helps us appreciate an important attraction of that approach. It also assists us in identifying an important limitation, which merits further inquiry and investigation. I begin with the attraction of the view relative to its main competitor. A key difficulty with the popular constitutionalist approach, which locates the popular will external to the formal amendment procedure, is that it often either ignores the amendment process, or else seems too fatalistic about its possibilities. To tease out the relative advantage of viewing the popular will as internal to the formal amendment procedure, we can observe that a core problem that constitutional amendment resolves is the special problem presented by old, outdated constitutions (‘special’ because endurance and stability are more common problems). The authority of a written constitution has traditionally been understood as derived, in some way, from the people it governs. However, constitutions that succeed in providing an enduring and stable source of fundamental legal norms and values over time can become out of date, reflecting the needs and values of the past rather than the present. When this happens, constitutions can face a crisis of authority. Formal amendment procedures are designed in part to respond to this problem. It is well known, however, that they do not always succeed: sometimes because they are too difficult (the conventional wisdom about Article V of the US Constitution), but sometimes because they are ineffective for other reasons (as is the case with section 128 of the Australian Constitution). In the absence of an effective ­amendment procedure, the primary alternative method of keeping a constitution up to date is progressive judicial interpretation. Yet, most constitutional theorists accept that judicial interpretation is rarely (if ever) a first-best means of constitutional amendment—even if they think that such practices are sometimes justified. If we accept some version of the view that a constitution’s authority is derived from the people it governs, then progressive judicial interpretation is an unattractive solution because it makes judges, not the people, the agents of constitutional change. Accepting that progressive judicial interpretation is generally not regarded as a first-best solution, it seems odd—and, indeed, problematic—for a theory of constitutional amendment to give normative priority to a conception of ‘the people’ in their constitutional law-making capacity that is external to the formal amendment process, and that requires the mediation of judicial interpretation. Such an approach seems too fatalistic, or worse: a self-fulfilling prophecy. The core preoccupation with justifying instances of judicial amendment in constitutional scholarship arguably

270  Lael K Weis bears this out.80 As Professor Goldsworthy has argued in response to those who advocate a more progressive approach to constitutional interpretation in Australia, ‘The point of including s 128 in the Constitution was to enable subsequent generations to change any aspects of it that might seem to them undesirable. What has happened to the spirit of democracy that we should hang back and expect judges to do our work for us?’81 Australian constitutional practice thus helps demonstrate an important attraction of the formalist approach. By holding that the formal amendment procedure is the exclusive, or at least most important, site of the popular will in its constitutional law-making capacity, that approach demands greater attention to critically evaluating and reforming the amendment process. Australians cannot afford to succumb to fatalism about constitutional amendment precisely because section 128 is positioned as the locus of popular sovereignty, precluding judicial interpretation as a solution to the problems posed by an old, outdated constitution. As a result, there is a robust and diverse body of Australian scholarship, including research papers commissioned by the government, that aims at diagnosing the reasons why the amendment process has not fared as expected, and that has proposed various ways that amendment practice can be reformed to better engage the public in matters of constitutional law.82 This includes former High Court judges who have invoked section 128 to defend a formalist interpretive stance,83 as well as scholars and other advocates of the High Court’s formalist approach to interpretation.84 At the same time, however, Australian constitutional practice also reveals an important limitation to the notion that the formal amendment procedure is the locus of popular sovereignty. In order to appreciate this contribution to our evaluation of the formalist alternative, it is first necessary to make an observation about the usual criticisms of this approach. Standard criticisms are informed by a perspective where the written constitution has a more certain status as a founding document or plays a more significant role in popular culture (or both). From that perspective, the

80 

See Colón-Ríos ‘Counter-Majoritarian’ (n 3). Goldsworthy, ‘Interpreting the Constitution’ (n 38) 704. eg, Campbell (n 47); R McGarvie, ‘The Wisdom of Hindsight—The 1999 Republican ­Referendum—Lessons for the Future’ (2001) 3 U Notre Dame Aust L Rev 11; G Orr, ‘Administration of Referenda and Plebiscites’ (2000) 11 Pub L Rev 118; Saunders (n 43); JM Williams, ‘The Constitutional Amendment Process: Poetry for the Ages’ in HP Lee and P A Gerangelos (eds), Constitutional Advancement in a Frozen Continent: Essays in Honour of George Winterton (Federation Press, 2009) 1; W ­ illiams and Hume (n 7). It is noteworthy that at the recent National Law Reform Conference many of the papers presented on constitutional law reform topics contained proposals for improving the amendment process (Conference at the Australian National University, 14–15 April 2016) accessed 1 August 2016. 83  For instance, after his retirement from the High Court, Justice McHugh, the author of the formalist judgment in Al-Kateb described above, went on to advocate that the Constitution be amended to include a bill of rights in order to prevent future ‘tragic’ cases, and made suggestions for how this might be done: see, eg, M McHugh, ‘The Need for Agitators—The Risk of Stagnation’ (Speech delivered at the Sydney University Public Law Forum, Sydney University, 12 October 2005) . 84 See J Goldsworthy, ‘A Role for States in Initiating Referendums’ in Upholding the Australian ­Constitution: Proceedings of the Eighth Conference of the Samuel Griffith Society (Samuel Griffith ­Society, 1997) 39. 81 

82 See,

Constituting ‘the People’ 271 ­ roblem with the formalist approach to identifying the emergence of the popular p will in its constitutional law-making capacity is that it takes an overly limited view of the relevant popular will, failing to appreciate the significance of social movements and other developments outside of the formal amendment process on the development of constitutional norms. As a threshold matter, I have suggested that for the normative understanding that the formalist approach attributes to the formal amendment procedure to have any plausibility, the procedure in question must be both reasonably practicable and reasonably democratic. But the analysis provided in this Chapter suggests that while these are necessary ingredients, they are not sufficient. In the Australian context, the difficulty is that the formalist approach demands too much from the formal amendment procedure. It requires thinking of section 128 as constituting ‘the people’, when the nature of Australian constitutional arrangements and constitutional culture otherwise cast doubt on the notion that the authority of the Constitution resides in popular sovereignty. This is analogous to a set of observations made by Sujit Choudhry in the C ­ anadian context.85 Reflecting on the Quebec secession crisis, Choudhry observes that an amendment procedure cannot achieve constitutional settlement where the ­‘settlement’ achieved by the constitution in question is otherwise contestable.86 So much was the case with the place of Quebec in the Canadian federation, and as a result, Choudhry argues, the amendment procedure prescribed by the Canadian Constitution Act 1982 could not be regarded as a neutral site for the manifestation of the popular will in its constitutional law-making capacity. Similarly, reflecting on the Australian context, here we might observe that an amendment procedure also cannot achieve constitutional settlement where the constitution in question is otherwise incapable of being described as ‘constitutive’ of the people. While the Australian Constitution may achieve settlement of some kind, it is unclear that it achieves settlement of a constitutive kind. As a result of the ambivalence of the ‘constitution’ achieved by the Australian Constitution, its amendment rule is at best an indeterminate site for the manifestation of the popular will in its constitutional law-making capacity. The Australian case thus highlights a different kind of problem for the formalist approach than those usually cited. Namely, it suggests that there are limitations to thinking of a constitution’s amendment procedure as the locus of popular sovereignty in absence of some independent notion of the sovereign ‘people’ constituted by that constitution. The plausibility of regarding the formal amendment procedure as the locus of popular sovereignty appears to turn on other background conditions, such as constitutional culture, which do not obtain in Australia given the unusual character of the Australian Constitution as a founding document. At the same time,

85 S Choudhry, ‘Ackerman’s higher lawmaking in comparative constitutional perspective: Constitutional moments as constitutional failures?’ (2008) 6 Int’l J Const L 193. 86  ibid 221. The question before the Court in the Secession Reference Case was whether Quebec could unilaterally secede from Canada, and whether secession could be obtained without a prior constitutional amendment. The Court held that if a ‘clear majority’ of Quebec’s population supported secession at referendum, Canada would be obligated to enter negotiations: Reference re Secession of Quebec [1998] 2 SCR 217.

272  Lael K Weis however, the success of Australia’s 1967 referendum suggests that moments of constitutional law-making instantiated through the formal amendment process can be achieved—even if they are constrained by the relevant background conditions—if the process is conducted in a way that is designed to address inadequacies in those background conditions. Important future research thus lies in exploring the possibilities and parameters of designing amendment practice to perform a ‘supplemental’ constitutive function in circumstances where, as in Australia, the notion of a constituted ‘people’ cannot be taken for granted.

14 Hard Amendment Cases in Canada KATE GLOVER

I

N PATRIATION AND Amendment of the Constitution of Canada, Peter Oliver writes that moments of constitutional amendment usually reveal much about a country’s political culture because they ‘expose the stress spots and irregularities’ in a country’s social and political architecture.1 That these moments offer insight into a country’s identity is not unexpected. As FR Scott reminds us, ‘[c]hanging a constitution confronts a society with the most important choices, for in the ­constitution will be found the philosophical principles and rules which largely determine the relations of the individual and of cultural groups to one another and to the state’.2 But, the amending procedures that facilitate the moments of change tend to be ‘unforthcoming’, disclosing little about a country’s constitutional anxieties and aspirations.3 And yet, in Canada, ‘such is not and was never likely to be the case’.4 Rather, Canada’s amending procedure lays bare some of the most sensitive and divisive issues in the country’s history, exposing worries and preoccupations that have long haunted and animated Canadian constitutional life. As Oliver writes, the ‘Canadian formula is probably the most complex in the world, and many would agree that this is a fair reflection of where we have come to in our country's history’.5 The Canadian amending procedure is set out in Part V of the Constitution Act 19826 (‘Part V’) and is revealing in part because of its form. It contains multiple amending procedures that collectively prescribe which orders of government, in what numbers, must consent to which amendments, in what circumstances.7

1 P Oliver, Patriation and Amendment of the Constitution of Canada (Ph.D. thesis, University of Oxford, 1992) [unpublished] at 180. 2  FR Scott, Essays on the Constitution: Aspects of Canadian law and politics (University of Toronto Press, 1977) ix. 3  Oliver (n 1) 180. 4 ibid. 5  P Oliver, ‘Canada, Quebec, and Constitutional Amendment’ (1999) 49 UTLJ 519 at 520. 6  Constitution Act 1982, Part V, being Schedule B to the UK Canada Act 1982, c 11 (Constitution Act 1982). 7  The provisions of Part V can be divided into two groups. One group—ss 38(1)–(3), 41, 42, 43, 44, 45, 47(1)—prescribes the consensus required for entrenching a formal constitutional amendment. The other group—ss 38(4), 40, 46, 47(2), 48 and 49—deals with the logistics of the amendment process, including provincial compensation and timelines. In any particular case, determining which procedure

274  Kate Glover The levels of consent are set according to a general rule and a series of exceptions. In the normal course, the exceptions are triggered by the subject matter of a proposed amendment. For instance, any amendment to the Constitution of Canada in relation to the use of the English or French language requires the unanimous consent of the houses of Parliament and the legislative assemblies of the provinces.8 In contrast, Parliament alone can amend the Constitution in relation to the executive government and the House of Commons.9 By listing subject matters and matching them with varying levels of legislative consent, Part V points to the issues that have been particularly troubling to Canada’s national psyche over the course of history. These issues include not only the use of the English or French language, but also, for example, the powers of the Senate, the method of selecting of senators, the composition of the Supreme Court of Canada and the design of the amending procedure itself. In addition, the structure of Part V highlights a set of questions about institutional authority and relationships that Canadian constitutional life grapples with. By ­tethering power over amendment to broadly described and historically contentious subject matters, the design of Part V all but guarantees that the courts will often play a meaningful role in the operation of the amending procedures in Canada. This active role for the courts in Canadian constitutional amendment has come to fruition, as several disputes over the meaning and obligations of Part V have already been litigated.10 These cases, the Senate Reform Reference and the Supreme Court Act Reference in particular, have advanced the law and resolved many uncertainties. This said, many questions about the application of Part V remain. This chapter aims to address some of the queries about Part V that have lingered since patriation. It develops foundational principles governing the application of Part V, as extracted from a careful reading of the text and jurisprudence. Further, it showcases the operation and merit of these principles by applying them to a case study, namely to longstanding debates about how to reform the Supreme Court of Canada. Supreme Court reform is an apt example when inquiring into the foundations and traditions of constitutional amendment in Canada for several reasons. Both the Court and its composition are expressly listed in Part V, thereby suggesting the deep legal and political questions about national identity and institutional legitimacy that are implicated in the design and operation of the Court. Moreover, proposals for constitutional amendment on issues related to the Supreme Court have been prominent on the agenda of mega-constitutional reform since the 1950s,11 although the applies depends on the subject matter and scope of the proposed amendment. The amending procedures set out in ss 38, 41, 42 and 44 apply to amendments ‘in relation to’ a list of ‘matters’, while the procedures set out in ss 43 and 45 apply to amendments of particular scope. Section 43 applies to amendments to the Constitution of Canada that applies to one or more, but not all, provinces. Section 45 applies to ­amendments to the Constitution of a province. 8 

Constitution Act 1982, s 41(c). Constitution Act 1982, s 44. 10  See, eg, Reference re Senate Reform, 2013 SCC 32; Reference re Supreme Court Act, ss 5 and 6, 2013 SCC 21; Hogan v Newfoundland (AG), 2000 NFCA 12. 11  For a chronology of these proposals, see J Aiello, ‘The Supreme Court of Canada: A Chronology of Change’ in N Verrelli (ed), The Democratic Dilemma: Reforming Canada’s Supreme Court (McGillQueen’s University Press, 2013) 277. See also E Crandall, ‘DIY 101: The Constitutional Entrenchment of the Supreme Court of Canada’ in E Macfarlane (ed), Constitutional Amendment in Canada (University of Toronto Press, 2016) 211 at 212–14, 215–19. 9 

Hard Amendment Cases in Canada 275 history of proposals to reform the Court through statutory and executive channels has been long, with the success of those proposals mixed.12 Further today, while the Senate Reform Reference and the Supreme Court Act Reference have answered some questions about the interpretation of Part V in the context of institutional reform, the case of Court reform exposes lingering uncertainties. This chapter confronts these uncertainties in three parts. In Part I, it sets out the existing rules of constitutional amendment in Canada and how they apply to Supreme Court reform. This part explains why Court reform is an archetypal example of a ‘hard amendment case’ and why we must become comfortable with such hard cases in Canada. Part II then identifies two new principles that should inform the interpretation and application of Part V. These considerations deal with the analytical approach to assessing reform proposals, and the types of change that trigger Part V. The final part of this chapter applies these considerations to a specific example, namely mandatory bilingualism for judges of the Court. It establishes that it is unlikely that Parliament alone could lawfully enact a mandatory bilingualism requirement for judges of the Supreme Court. Such a requirement would amount to a constitutional amendment in relation to the Court’s composition and therefore require the unanimous consent of Parliament and the provincial legislatures. I.  PART V AND HARD AMENDMENT CASES

A.  The Logic of Part V Part V contains multiple amending procedures. The general procedure (section 38(1)) provides that an amendment to the Constitution of Canada requires the consent of the houses of Parliament and the legislative assemblies of at least two-thirds of the provinces representing 50 per cent of the population. This ‘7/50 rule’ applies to amendments that do not fall within any of the other procedures, as well as to amendments in relation to matters that are expressly listed in section 42(1). The listed matters include the powers of the Senate, the method of selecting senators, and the Supreme Court of Canada. Amendments in relation to the office of the Queen, the use of the English or French language, and the composition of the Supreme Court require the unanimous consent of the houses of Parliament and the provincial legislatures (section 41). Part V also contains a ‘special arrangement’ procedure (section 43). It provides that an amendment to any provision of the Constitution of Canada that applies to one or more, but not all, provinces requires the consent of the houses of Parliament and the legislative assembly of the provinces to which the amendment applies. Finally, sections 44 and 45 provide for unilateral amending powers. Section 44 provides that Parliament alone can, with some exceptions, amend the Constitution of Canada in relation to the executive, the Senate, and the House of Commons.

12  See Aiello, ibid. For current arguments on the merits and shortcomings of modern reform proposals, see, eg, the contributions to Verrelli, ibid.

276  Kate Glover Section 45 provides that, subject to section 41, the legislature of a province can exclusively make laws amending the Constitution of the province. There is an internal logic to Part V. The centrepiece is the general ‘7/50’ rule. Sections 41, 43, 44 and 45 identify exceptions and specify the conditions in which the unanimous, bilateral and unilateral amending procedures apply. Of particular note is the line that divides the multilateral and unilateral procedures, a line drawn between that which engages both federal and provincial interests (and therefore calls for multilateral action) and that which engages the interests of only one order of government (and is therefore the proper authority of that order alone). The general amending rule, and indeed all of the multilateral procedures, are animated by the principle that ‘substantial provincial consent must be obtained for constitutional change that engages provincial interests’.13 Unilateral powers in relation to constitutional and institutional reform reflect the same principle, emphasising the need to navigate the demands of sovereignty and the limits of autonomy in a cooperatively federal constitutional order. This limit on the capacity for unilateral change ‘reflects the principle that Parliament and the provinces are equal stakeholders in the Canadian constitutional design. Neither level of government acting alone can alter the fundamental nature and role of the institutions provided for in the ­Constitution’.14 Thus, the houses of Parliament and the provincial legislative assemblies can act unilaterally to maintain and even alter these central institutions, but their unilateral authority ends where alterations to the institutions’ fundamental nature and role begin.15   Turning to a practical example, the operative provisions for reform of the Supreme Court are sections 42(1)(d) and 41(d). Pursuant to section 42(1)(d), an amendment to the Constitution of Canada in relation to the Supreme Court must be made in accordance with the 7/50 rule. Section 42(1)(d) is subject to section 41(d), which provides that an amendment to the Constitution of Canada in relation to the composition of the Supreme Court must have the consent of the Senate, the House, and the legislative assembly of each province. Also important is section 101 of the Constitution Act 1867, which provides: The Parliament of Canada may, notwithstanding anything in this Act, from Time to Time provide for the Constitution, Maintenance, and Organization of a General Court of Appeal for Canada, and for the Establishment of any additional Courts for the better Administration of the Laws of Canada.16

Pursuant to section 101, Parliament can unilaterally enact ‘routine amendments necessary for the continued maintenance of the Supreme Court’, as long as those amendments do not alter the ‘constitutionally protected features of the Court’.17

13 

Senate Reform Reference (n 10) para 34. ibid, para 48. 15 ibid. 16  Constitution Act 1867 (UK) 30 & 31 Vict c 3, s 101, reprinted in RSC 1985, Appendix II, No 5 (Constitution Act 1867). 17  Supreme Court Act Reference (n 10) para 101. 14 

Hard Amendment Cases in Canada 277 B.  Hard Amendment Cases Whenever a proposal for reform is assessed against the demands of Part V, the first issue is always one of application: Does the proposal trigger Part V? The answer to this question turns on whether the proposal is an ‘amendment to the Constitution of Canada’ within the meaning of Part V.18 If yes, Part V applies and the issue becomes which specific amending procedure is triggered. If no, the proposal can be enacted through ordinary legislative channels. A proposal is an ‘amendment to the Constitution of Canada’ within the meaning of Part V if it alters an entrenched part of the Constitution of Canada. The preliminary question is: What is entrenched in the Constitution? The most straightforward part of the answer deals with the constitutional text. As a general proposition, the text of the Constitution Acts 1867 and 1982 is entrenched. Section 52(2) provides that the ‘Constitution of Canada includes (a) the Canada Act 1982, including [the Constitution Act 1982]; (b) the Acts and orders referred to in the schedule; and (c) any amendment to any Act or order referred to in paragraph (a) or (b)’. The schedule lists dozens of Acts and orders, including the Constitution Act 1867 and the Statute of Westminster 1931, although the Supreme Court Act is not listed. Section 52(3) of the Constitution Act 1982 provides that ‘[a]mendments to the Constitution of Canada shall be made only in accordance with the authority contained in the Constitution of Canada’. The easiest amendment cases are therefore those involving proposals that modify the words of a Constitution Act or order.19 The text of the Acts and orders is entrenched and Part V will generally apply to any change thereto.20 Providing a more complete response to the entrenchment question is complicated by the nature of the Constitution of Canada, the entrenched parts of which are not exhausted by text.21 Rather, the Constitution has written and unwritten dimensions, encompassing ‘the global system of rules and principles which govern the exercise of constitutional authority in the whole and in every part of the Canadian state’.22 These rules and principles are found in ‘an understanding of the constitutional text itself, the historical context, and previous judicial interpretations of constitutional meaning’.23 Moreover, the Constitution has a shape, an ‘internal architecture’ that binds its elements to each other, and aspires to a vision of government and public life that animates the Constitution as a whole.24 Thus, the harder amendment cases are those involving proposals that do not expressly alter the text of the Constitution, but which bear on its structure, its assumptions, or its meaning.

18  With the exception of s 45, each of the individual procedures set out in Part V starts with a version of ‘An amendment to the Constitution of Canada in relation to the following matters may be made by…’. See, eg, Constitution Act 1982, ss 38(1), 41, 42(1), 43, and 44. See also Oliver (n 5) at 575–83. 19  These Acts and orders are listed in s 52(2) of the Constitution Act 1982 and the Schedule thereto. 20  Constitution Act 1982, s 52(2), (3). 21  Reference re Secession of Quebec [1998] 2 SCR 217 para 32. 22  Reference re Resolution to amend the Constitution [1981] 1 SCR 753 at 874. 23  Secession Reference (n 21) para 32. 24  OPSEU v Ontario (AG), [1987] 2 SCR 2 at 57.

278  Kate Glover For the most part, reform of the Supreme Court falls within this category of harder cases. On the one hand, Part V provides that changes to the Constitution of Canada in relation to the Supreme Court and its composition are subject to the multilateral amending procedures. And yet, on the other, the Supreme Court is not mentioned elsewhere in the constitutional texts beyond Parliament’s general authority to establish and maintain a ‘general court of appeal for Canada’. The Court is established wholly by statute and the history of official Court reform has unfolded through the legislative process.25 What, then, about the Court is entrenched and therefore protected by Part V? According to the majority of the Supreme Court in the Supreme Court Act Reference, the existence of the Court is entrenched, though this was not always the case. The majority reasoned that the Court gained constitutional status ‘as a result of its evolution into the final general court of appeal for Canada, with jurisdiction to hear appeals concerning all the laws of Canada and the provinces, including the Constitution’.26 This status, the Court held, was then ‘confirmed in the Constitution Act 1982, which made modifications of the Court’s composition and other essential features subject to stringent amending procedures’.27 These ‘other essential features’ include, at a minimum, ‘the Court’s jurisdiction as the final general court of appeal for Canada, including in matters of constitutional interpretation’, as well as the Court’s independence’.28 In other words, the Supreme Court Act Reference establishes that the existence and essential features of the Court are entrenched in the Constitution of Canada and subject to change only by virtue of Part V. However, it offers little guidance on the scope of those essential features, what other features might count as essential, and what constitutes a ‘qualitative difference’ to the fundamental nature and role of the Court that triggers Part V. The next part of this chapter sets out two principles that assist in resolving these uncertainties. II.  PRINCIPLES OF APPLICATION

The aim of this chapter is to clarify when and why the formal amending procedures set out in Part V apply. It presents this clarification using the specific example of Court reform. As we know, the answers to these ‘when’ and ‘why’ questions are not obvious because proposals for Court reform tend to raise hard cases. In thinking through the routes of formal reform available to Parliament, the question about application is, in essence, a question about the scope of Parliamentary authority to act unilaterally under section 101 of the Constitution Act 1867 in relation to the scope of the multilateral requirements set by Part V. The multilateral obligations set by Part V kick in as the limits of Parliament’s unilateral authority are exceeded. Those limits are drawn by that which ‘change[s] the constitutionally protected features of the Court’.29 More specifically, they are drawn by that which brings about 25 

Aiello (n 11). Supreme Court Act Reference (n 10) para 95. 27 ibid. 28  ibid, para 94. 29  ibid, para 101. 26 

Hard Amendment Cases in Canada 279 ‘substantive change’30 or makes a ‘qualitative difference’ to the architecture of the Constitution of Canada.31 These types of changes trigger Part V. The two principles set out below help to alleviate some of the challenges in hard amendment cases. As we will see, these principles demonstrate the importance of understanding an institution’s internal design and its place in the grander constitutional order when assessing whether Part V applies to any particular proposal for reform. A.  Supporting Characteristics and Qualitative Assessments i.  The Principle Public institutions are created as one part of a much larger constitutional order. Within that order, the reform of one institution can have residual effects on other institutions, as well as on the overall order itself. Similarly, each feature of a single institution is one part of that institution’s overall design. Some features of an institution will be ‘essential’.32 These are the characteristics necessary for an institution to fulfill its primary constitutional role. Other features will be supporting or secondary. These are the characteristics that operationalise the institution’s essential features. Given these connections within an institution’s design, the alteration of one institutional feature can have residual effects on other features, as well as on the institution as a whole. In the constitutional amendment context, the alteration of a non-entrenched feature of an institution can trigger an amendment to an entrenched constitutional feature. In Canada’s constitutional order, a change to a non-entrenched feature (that is, a change to a supporting, rather than an essential, characteristic within an institution’s design) will reach the threshold of a constitutional amendment triggering the multilateral procedures of Part V only if it makes a ‘qualitative difference’33 or ‘substantive change’34 to the entrenched dimensions of the Constitution of Canada. Part V must be capable of capturing these cascading effects of change. To ensure it is able to do so, the analysis of any reform proposal must attend to substance over form, and both the purposes and effects of the proposal. Moreover, the analysis must attend to the interlocking nature of institutional design, which inevitably links essential roles and characteristics of an institution to many more supporting, operationalising features. In short, determining whether Part V is triggered in any particular case calls for a qualitative assessment of the proposal in question. ii.  The Principle in Action The importance of qualitative assessments of reform proposals is highlighted by an example from the context of Senate reform. One of the Senate’s essential features is 30 

ibid, para 106. Senate Reform Reference (n 10) para 80. 32 See Re: Authority of Parliament in relation to the Upper House, [1980] 1 SCR 54 at 78; Supreme Court Act Reference (n 10) paras 19, 74, 76, 90, 94. 33  Senate Reform Reference (n 10) para 80. 34  Supreme Court Act Reference (n 10) para 105. 31 

280  Kate Glover its independence. This independence is not expressly mentioned in the Constitution Act 1867, but rather is implemented by a set of supporting characteristics, including long tenure for senators. In 1867, senators were appointed for life. This was changed in 1965, when life tenure was replaced with mandatory retirement at age 75.35 This change was not an amendment requiring multilateral agreement under convention or law at the time, and would not be one requiring multilateral consensus under Part V today. Replacing life tenure with a term ending at 75 would not alter the Senate’s capacity for independent legislative review. Retiring senators would be unlikely to seek re-appointment or other employment when finishing their term. Thus, upon a qualitative assessment of the proposal’s effects, we see that this modification to a secondary characteristic (term of tenure) would not impact an essential, entrenched feature of the Senate’s fundamental nature and role (independence). It is a change that amounts to housekeeping that does not impact federal-provincial relations in a meaningful way. It is, therefore, within Parliament’s unilateral authority. And yet, a proposal to lower the retirement age to 50 or to fix the length of Senate terms leads to a different analysis. ‘The Senate is a core component of the Canadian federal structure of government’, the Court wrote in the Senate Reform Reference.36 Its fundamental nature and role is be a ‘complementary legislative body of sober second thought’.37 The current length of senatorial terms is ‘directly linked to this conception of the Senate’.38 The long term, ‘roughly the duration of [senators’] active professional lives’, aims to cultivate and preserve the independence of senators in their legislative review.39 In assessing the constitutional impact of a proposal to significantly reduce or fix the length of Senate terms, the relevant question is whether fixed terms would alter the fundamental nature or role of the Senate and, as a consequence, ‘engage the interests’ of the provinces. A qualitative assessment of the effect of such a proposal establishes that there would be a change to the Senate’s fundamental nature and role. At some point, a reduced, fixed term could impair the Senate’s capacity for independent second thought. ‘Fixed terms provide a weaker security of tenure’, the Court contended in the Reference. ‘They imply a finite time in office and necessarily offer a lesser degree of protection from the potential consequences of freely speaking one’s mind on the legislative proposals of the House of Commons’.40 As a result, a proposal to implement fixed terms of senatorial tenure falls outside the scope of Parliament’s unilateral amending authority and can only be achieved by virtue of the 7/50 formula. These examples from the realm of Senate reform show how the modification of a secondary characteristic (eg, tenure) can have effects that cascade into the entrenched features of the Constitution (eg, independence). These cascading effects must be accounted for in the application of Part V. Without this capture, political actors would be able to formally—and yet indirectly—amend the Constitution without the

35 

Constitution Act 1965, 14 Eliz II, c 4, Part I (Can). Senate Reform Reference (n 10) para 77. 37  ibid, para 79. 38 ibid. 39 ibid. 40  ibid, para 80. 36 

Hard Amendment Cases in Canada 281 consent of the interested legislative actors. The method of ensuring this capture is a straightforward, qualitative assessment of proposals, with a view to their effects on the Constitution as a whole, architecture and all. iii.  The Case of Supreme Court Reform How do these considerations apply in the context of Court reform? The easier cases of Court reform are those dealing with direct and explicit changes to the fundamental role of the Court or its essential characteristics. These are easier because by directly altering the architecture of the Constitution, these proposals constitute ‘amendments to the Constitution of Canada’ and trigger Part V. An example would be a proposal to remove the Court’s jurisdiction over private law cases (as has been suggested in past proposals to establish the Supreme Court as a constitutional court). This dramatic change to the generality of the Court’s jurisdiction and its role in overseeing the development of a coherent legal system would strip the Court of its capacity to fulfill its fundamental constitutional role as the final general appellate Court for the country.41 This directly affects the interests of the provinces and would, therefore, be an amendment to the Constitution of Canada within the meaning of Part V. The 7/50 procedure would apply by virtue of section 42(1)(d). Another example of an easier case would be a proposal to amend section 101 of the Constitution Act 1867, revoking Parliament’s authority to create the Supreme Court. This would, in effect, dismantle the Court. Such a proposal would trigger Part V because it amounts to a change to an operative part of the constitutional text. Yet imagine, instead, that Parliament repealed the Supreme Court Act, again dismantling the Court. This action would not alter the written text of the Constitution and yet it must still constitute an amendment to the Constitution of Canada within the meaning of Part V. Abolition of the Court would profoundly renovate the architecture of the Canadian constitution, dismantling one of its ‘fundamental premises’.42 Repealing the Supreme Court Act would, in effect, constitute a qualitative change to the entrenched constitution. To explore the harder cases of change to secondary characteristics, let’s look again at the ‘composition of the Supreme Court of Canada’ in section 41(d) as an example. The Supreme Court Act Reference establishes that the composition of the Court, as imagined and protected by the Constitution, encompasses a set of eligibility criteria provided for in the Supreme Court Act.43 When will Part V apply to proposals that modify or add to these criteria? First, the general principles confirm that not all modifications to the statutory eligibility criteria will trigger Part V.44 Rather, a reform proposal will constitute an ‘amendment to the Constitution of Canada in relation to the composition of the Supreme Court of Canada’ only if it makes a ‘substantive change’ to that which is

41 

Supreme Court Act Reference (n 10) para 95. ibid, para 89. ibid, para 104. 44  See Supreme Court Act Reference (n 10) para 115, per Moldaver J. 42  43 

282  Kate Glover currently entrenched.45 A change to a secondary characteristic can—but will not always—meet that threshold. Consider a practical example. Section 6 of the Supreme Court Act provides that at least three judges of the Court must be drawn from the judges and advocates of Quebec. Section 6 ‘reflects the Court’s bijural character and represents the key to the historic bargain that created the Court in the first place’.46 As the majority in the Supreme Court Act Reference explained, ‘the guarantee that one third of the Court’s judges would be chosen from Quebec ensured that civil law expertise and that Quebec’s legal traditions would be represented on the Court and that the confidence of Quebec in the Court would be enhanced’.47 Accordingly, a legislative proposal that would make a qualitative difference to the representation of Quebec on the Court would trigger the unanimity procedure. An easy case would be a proposal to amend section 6 to reduce the minimum number of judges from Quebec from three to two. Such a proposal would alter the entrenched content of the Constitution in relation to the composition of the Supreme Court — the proportion of judges drawn from Quebec would be diminished and section 41(d)—the unanimity procedure—would clearly apply. A trickier case would be a proposal to amend section 5 of the Supreme Court Act to allow for the appointment of advocates of at least nine years standing at the bar of a province instead of the current requirement of ten years standing. Such a proposal would change the language of section 5. If it is true that all eligibility criteria fall within ‘composition of the Supreme Court’ in section 41(d) and that sections 4(1), 5 and 6 of the Supreme Court Act codify the eligibility criteria, when the proposed change to section 5 would seemingly satisfy the ‘amendment to the Constitution of Canada’ threshold with ease. But a closer reading of the jurisprudence indicates that what is entrenched in section 5 is not the number ‘10’ but rather what it signifies, namely that the Court’s bench is to be comprised of judges with a certain level of expertise, maturity, judgment, and independence of thought. Recall the ‘qualitative difference’ threshold noted above. At some point along the timeline of membership at the bar, a change in the requisite tenure period meaningfully alters the status quo. But it does not follow that all such changes have a meaningful effect. Again, a qualitative assessment is required. Both a reduction from the ten-year threshold to a one year minimum and an increase from ten years to 25 years48 would be a meaningful change to current understandings of the ­expertise, maturity, judgement, and independence of thought needed for the Court to function competently and legitimately. A change in numerical thresholds set out in the ­Constitution has the potential to be transformative and these proposals would, therefore, fall within the scope of section 41(d) and require unanimous consent in order to be implemented.

45 

Supreme Court Act Reference (n 10) para 105. ibid, para 104. 47 ibid. 48 Cyr contemplates this example in H Cyr, ‘The Bungling of Justice Nadon’s Appointment to the Supreme Court of Canada’ (2014) 67 SCLR (2d) 73 at 106. 46 

Hard Amendment Cases in Canada 283 Yet, a shift from ten to nine years of bar membership does not signal a substantive change to the constitutionally-protected dimensions of the Court. There is no magical transformation associated with crossing the ten-year membership mark. The ten-year threshold reflects a policy choice, driven by the underlying goal of ensuring institutional legitimacy and competency. Without clear evidence to the contrary, there is no empirical or conceptual reason to conclude that a shift from ten to nine years would have an effect on the Court’s capacity to fulfill its constitutional role or the composition of the Court’s bench. It is a matter of legislative policy to determine what the requisite length of membership should be, within the bounds of what bears directly and meaningfully on the constitutionally-entrenched dimensions of the Court.49 Accordingly, a proposal changing the ten-year requirement in section 5 of the Supreme Court Act to a nine-year requirement would be within Parliament’s unilateral legislative authority, notwithstanding the change to the statutory text. Again, the substance rather than the form of legislative proposals and their effect on the Constitution is the touchstone of the analysis.50 B.  Enhancements and Alterations of Architectural Interests i.  The Principle A second general principle of Part V provides that a proposal for reform that implements, enhances, or engages a structural principle or interest of the Constitution, which on its own might not trigger Part V, will be an amendment to the Constitution of Canada if it also has a transformative effect on the architecture of the Constitution. This second principle is a species of the first as it relies on qualitative assessments of the effects of proposed action in its application. However, it differs from the first principle discussed above because it directs our attention to the types of change that can flow from proposals rather than to the character of the analytical approach that Part V demands. ii.  The Principle in Action In what ways is this second principle relevant? Some examples will help answer this question. First, a legislative enactment that is declarative of existing constitutional requirements but has no transformative effect on the entrenched dimensions of the Constitution will likely not constitute an amendment to the Constitution of Canada; no change has been made. Such was the case with one of the amendments to the Supreme Court Act considered in the Supreme Court Act Reference. At the time of the Reference, Parliament had enacted legislation adding section 5.1 to the Supreme Court Act. Section 5.1 provided: ‘For greater certainty, for the purpose of section 5, 49  This is similar to the Supreme Court’s analysis of proposals to change the length of senatorial tenure in the Senate Reform Reference (n 10) para 82. The difference in the Senate context was that the impugned proposals did not merely alter the length of senatorial tenure. They represented a conceptual change from an (effectively) life-long term to fixed term tenure. 50  Senate Reform Reference (n 10) para 52.

284  Kate Glover a person may be appointed a judge if, at any time, they were a barrister or advocate of at least 10 years standing at the bar of a province’. The majority concluded that enacting section 5.1 was within Parliament’s unilateral legislative authority because it was ‘redundant’ rather than transformative.51 Section 5.1 was an enactment in relation to the composition of the Supreme Court of Canada, a matter contemplated by section 41(d). However, section 41(d) was not triggered because section 5.1­ simply stated the law as it already stood. Such legislative restatements, which address—but do not change—constitutionally entrenched interests (like the composition of the Supreme Court), will not trigger Part V. Secondly, turning to the heart of the principle, a legislative enactment that implements, expresses or engages a structural constitutional principle will constitute an amendment to the Constitution of Canada for the purposes of Part V if it has a transformative effect on other dimensions of the constitutional architecture. This principle is of particular relevance when contemplating the application of Part V to legislation that is constitutional or quasi-constitutional, in the sense that it ‘bears on an organ of government’,52 and that is intended to enhance or modernise a central institution or enhance the public expression of constitutional principles. Such legislation, like, for example, the Official Languages Act,53 the Canadian Multiculturalism Act,54 and the Referendum Act,55 could be a valid exercise of Parliament’s legislative authority under section 91 of the Constitution Act 1867 and section 44 of the Constitution Act 1982, and represent opportunities for ‘constitutional experimentation and innovation’ without triggering the amending procedure.56 However, in some circumstances, such enhancing legislation will have transformative effects on entrenched dimensions of the Constitution. For example, in the context of Senate reform, the Attorney General of Canada argued that the government’s proposals to establish advisory elections for selecting senators did not amend the Constitution of Canada but rather advanced a core constitutional principle, democracy. Selecting senators by virtue of a process that relied on the democratic process would, it was argued, enhance the democratic legitimacy of the Senate. The Supreme Court rightly rejected this submission. While an elective scheme would advance the democratic mandate of the Senate, it would also have a transformative effect on the architecture of the Constitution. Enhancing the democratic mandate of the Senate would ‘weaken the Senate’s role of sober second thought and would give it the democratic legitimacy to systematically block the House of Commons, contrary to its constitutional design’.57 This transformative effect on the architecture of the Constitution rendered the election proposals an

51 

Supreme Court Act Reference (n 10) para 106. OPSEU v Ontario (n 24). Official Languages Act, RSC 1985, c 31. 54  Canadian Multiculturalism Act, RSC 1985, c 24 (4th Supp.). 55  Referendum Act, SC 1992, c. 30. 56  Warren J Newman, ‘Constitutional Amendment by Legislation’ in Macfarlane (ed), (n 11) 117–21. 57  Supreme Court Act Reference (n 10) para 60. On the Official Languages Act, see also Jones v AG of New Brunswick [1975] 2 SCR 182. 52  53 

Hard Amendment Cases in Canada 285 amendment to the Constitution, despite their enhanced expression of the democratic principle.58 Before turning to a final example that deals with reform in relation to the Supreme Court, we should pause to notice that the examples dealing with Senate reform reinforce the importance of analysing both the purposes and effects of a proposal when determining whether Part V is engaged, and to consider them qualitatively. Examining legislative purpose alone risks seeing only the intended (or articulated) goals of a proposal, which might very well enhance constitutional considerations, like democratic legitimacy or the rule of law. It may be that the altering and impairing effects of the proposal are revealed only by a qualitative assessment of the proposal’s effects. Moreover, both of the above examples dealing with Senate reform—advisory elections and tenure—also establish that the relevant effects can be either internal to the institution being reformed, or external to that institution and found in the relationships, design, or power structures of the institutions around it. III.  THE CASE OF MANDATORY BILINGUALISM AT THE COURT

With these principles of application in place, we can now consider a final example of Court reform, that of a legislated bilingualism requirement for judges of the Supreme Court. At present, there is no such requirement, nor has there ever been.59 The Supreme Court is exempt from the duty resting on every other federal court in Canada to ensure that every judge who hears a case is able to understand the language chosen by the parties for the proceedings, whether English and/or French, without the assistance of an interpreter.60 And the private members bills that have been introduced in recent years to formalise a bilingualism requirement for the Supreme Court have died on the order paper,61 but not before sparking a polarised debate. The most effective way to implement a statutory bilingual requirement would be to amend the Supreme Court Act or the Judges Act. Removing the exemption in the Official Languages Act, while important for parties, would not ensure that all Supreme Court judges be bilingual, as only a quorum of the Court would have to be functionally bilingual in order to comply with the Act.62 That said, regardless of the effectiveness of a legislative solution, Parliament likely cannot unilaterally legislate a mandatory bilingual requirement for Supreme Court judges; a constitutional ­amendment is required.

58 

Supreme Court Act Reference (n 10) para 60. On the linguistic competency of judges generally, see Société des Acadiens v Association of Parents, [1986] 1 SCR 549; R v Beaulac, [1999] 1 SCR 786. 60  Official Languages Act, s 16(1). 61  See, eg, Bill 203, An Act to amend the Supreme Court Act (understanding the official languages), 1st Sess, 42nd Parl (2015) (Introduction and First Reading in the House of Commons) and Bill 208, An Act to amend the Supreme Court Act (understanding the official languages), 2nd Sess, 41st Parl (Defeated at 2nd Reading). 62  Pursuant to s 25 of the Supreme Court Act, RSC 1985, c S-26, ‘[a]ny five of the judges of the Court shall constitute a quorum and may lawfully hold the Court’. 59 

286  Kate Glover As explained above, in any amendment analysis, the first question is always: is the proposal an ‘amendment to the Constitution of Canada’ within the meaning of Part V? A legislated French-English bilingualism requirement would change the existing eligibility criteria for judges of the Supreme Court. The issue becomes whether such a requirement is a change to an entrenched part of the Constitution. The answer to this question is contested. One argument is that the Supreme Court Act Reference confirmed that the eligibility criteria set out in sections 4(1), 5 and 6 of the Supreme Court Act are entrenched in the Constitution of Canada and that any change to those criteria—whether by addition, deletion, or modification—amounts to an amendment in relation to the composition of the Supreme Court.63 Section 4(1) provides that the Court comprises the Chief Justice of Canada and eight puisne judges. Section 5 provides, ‘Any person may be appointed a judge who is or has been a judge of a superior court of a province or a barrister or advocate of at least ten years standing at the bar of a province’. And section 6 guarantees the representation of Quebec on the Court. It establishes that at least three of the judges shall be appointed from among the judges of the Quebec Court of Appeal or Superior Court or from among the advocates of Quebec.64 On this view, a statutory bilingualism requirement would be a new eligibility criterion for appointment to the Court and therefore can be implemented only with the consent of Parliament and the provincial legislatures under the unanimity procedure set out in section 41(d) of the Constitution Act 1982. One opposing position starts by defining the scope of section 41(d). The claim is that ‘composition of the Supreme Court of Canada’ in section 41(d) protects a ‘core’ of matters related to composition from unilateral change.65 This core includes some, but not all, criteria used to determine who is eligible to be appointed to the Supreme Court. The challenge is then to find the boundaries of the core. The literature offers two possible maps of the core of section 41(d), both of which stake the boundaries according to a purposive interpretation of ‘composition’. The first claim is grounded in framers’ intent and the fundamental character of the Court. The argument is that ‘composition’ in section 41(d) is intended to protect the representation of Quebec on the Court, as well as the Court’s role as the guardian of the Constitution against

63  See, eg, E Macfarlane, ‘The Uncertain Future of Senate Reform’ in Macfarlane (ed), (n 11) 242. See also L Sirota, ‘The Comprehension of “Composition”’ (Double Aspect, 16 May 2016), accessed 1 August 2016; P Daly, ‘Administering Constitutional Change: the Case of Bilingual Supreme Court Judges’ (Administrative Law Matters, 16 November 2015) accessed 1 August 2016. 64  Section 6.1 of the Supreme Court Act was declared to be unconstitutional in the Supreme Court Act Reference (n 10) paras 104–6. 65  S Grammond, ‘Can Parliament enact a requirement that Supreme Court judges be bilingual?’ (Administrative Law Matters, 13 May 2016) accessed 1 August 2016. See also Cyr (n 48) at 104–9. The amending procedures both permit amendments when the procedural requirements have been followed and protect the constitutional status quo from alteration when the applicable procedures have not been met: see WJ Newman, ‘Living with the Amending Procedures: Prospects for Future Constitutional Reform in Canada’ (2007) 37 Sup Ct L Rev (2d) 383 at 386.

Hard Amendment Cases in Canada 287 court-packing and abolition, from unilateral reform. Eligibility criteria fall within the scope of section 41(d), but only to the extent that they are directed to the representation of Quebec and the Court’s fundamental role. Grammond outlines this view in a short essay.66 He concludes that this approach offers a defence to Parliament for unilateral action. A legislated bilingualism requirement would not affect the protected core of composition; Quebec’s seats on the bench and the Court’s position remain untouched. As a result, section 41(d) is not triggered. Cyr offers a different way to define the scope of section 41(d), which turns on the extent to which the criteria for appointment to the Court are constitutionally protected.67 Cyr contends that sections 4(1), 5, and 6 of the Supreme Court Act set a minimum threshold for eligibility. Any alteration to the specific requirements of these sections could not be altered without triggering section 41(d), as it would, in effect, amount to a reform of the constitutional text.68 However, Cyr contends, Parliament could unilaterally add criteria to the minimum, as long as the additional qualifications ‘refer to the actual capacities required to carry out the highest judicial functions of the land’.69 Here, Cyr distinguishes between ‘[c]apacities related to the actual conduct of judicial proceedings’, such as literacy and mental competence, and capacities associated with political standards of legitimacy, such as citizenship and respect for the rule of law.70 He argues that the former would be readily justified as lawful unilateral action because they are ‘directly connected to the act of judging’.71 In this sense, they are implied by the nature of the judicial office. The latter could also be justified as unilateral additions by Parliament, as long as they do not interfere with the ‘ability of the regions to be adequately represented’.72 On this analysis, Parliament is authorised to establish a statutory bilingualism requirement for judges of the Supreme Court. The requirement would not ‘rais[e] the level of qualification already specified’ in sections 4(1), 5 or 6 and bilingualism is an issue of competence, bearing on the judge’s capacity to perform the duties of his or her office.73 It is, therefore, directly connected to the act of judging and a justified unilateral enhancement of the constitutional minimum for eligibility. This is a form of the claim, consistent with Newman’s account of constitutional legislation,74 that a bilingualism requirement would implement and enhance existing constitutional interests, without upsetting the fundamental nature and role of the Court or the basic structure of the Constitution, and therefore falls within Parliament’s unilateral legislative authority. Indeed, this argument could be extended to encompass a claim that the bilingualism requirement enhances the institutional expression of Canada’s

66 

Grammond, ibid. Cyr (n 48) 104–09. 68  ibid 106. 69  ibid 105 [emphasis in original]. 70  ibid 105. 71 ibid. 72  ibid 105–06. 73  ibid 106. 74  See Newman (n 56). 67 

288  Kate Glover constitutional ­commitment to its two official languages and is therefore an exercise in constitutional implementation rather than amendment.75 The given accounts differ in their interpretation of the Supreme Court Act ­Reference and ‘composition of the Supreme Court of Canada’ in section 41(d) of the Constitution Act 1982. Yet none of the claims is fully persuasive. One analysis reads the Reference and ‘composition’ too broadly, the others, too narrowly. The mapping of the boundaries around section 41(d) of the Constitution Act 1982 differs from the boundary-setting exercises outlined in the literature and described above. While the approaches outlined by Grammond and Cyr are properly qualitative in their assessments of what constitutes an amendment to the Constitution of Canada within the meaning of Part V, those two mappings of section 41(d) are too narrow. The scope of section 41(d) proposed in Grammond’s essay is too narrow, failing to account for the principles of legitimacy and competence that inform the Court’s constitutional character and go beyond concerns of abolition and Courtpacking. The boundaries contemplated in Cyr’s account are also too constraining. They allow for unilateral reform on matters of competency and legitimacy, the very issues that the Supreme Court Act Reference indicates are guarded against federal incursions by virtue of constitutional evolution and the text of Part V. The approach to understanding the constitutional status of the Court and the scope of Part V offered in this chapter draws insight from constitutional structure and principle. It is consistent with constitutional jurisprudence and the institutional architecture imagined within Canada’s constitutional order. In the Senate Reform Reference, the Court held that the Part V amending procedures apply to substantive changes to the architecture of the Constitution.76 This is a logical consequence of the nature of Canadian constitutionalism, which understands the Constitution of Canada to be more than textual provisions. The architecture of the ­Constitution captures the ‘structure of government that [the constitution] seeks to implement’.77 It includes the institutions and institutional relationships that are necessary to achieve the Constitution’s goals of governance, as well as the principles and aspirations that those institutions and relationships are intended to realise and on which they are built. Understanding Canada’s constitution therefore requires attention to its systematic and holistic qualities, and to the demands and insights derived from the ‘assumptions that underlie the text and the manner in which the constitutional provisions are intended to interact with one another’.78 The Senate Reform Reference establishes that the Senate is embedded within the structure of Canadian constitutionalism as a ‘complementary legislative body of sober second thought’.79 Legislative proposals that would make a ‘qualitative difference’ to this ‘fundamental nature and role’ of the Senate constitution an amendment to the Constitution and can be implemented only by virtue of the Part V amending

75 This point is alluded to in E Macfarlane, ‘Conclusion: The Future of Canadian Constitutional Amendment’ in Macfarlane (ed) (n 11). 76  Senate Reform Reference (n 10) para 27. 77  ibid, para 26. 78 ibid. 79  ibid, paras 52, 54–63.

Hard Amendment Cases in Canada 289 procedure, regardless of whether such proposals contemplate alterations of the existing text of the Constitution Acts.80 By altering this fundamental role of the Senate, the proposals engage the interests of both Parliament and the provinces; multilateral amendment is thus required.81 This approach relies on substantive understandings of Canadian constitutionalism and qualitative assessments of the purpose and effect of legislative proposals. On this approach, appointing senators following consultative elections amounts to an amendment to the Constitution of Canada despite no textual change, because senators endorsed by a popular mandate would undermine the Senate’s responsibility for independent review of policy, legislative proposals, and issues of public concern. Moreover, it would offer a principled justification for an expanded mandate of senatorial blocking of the House agenda.82 The implications of this reasoning for the institutional design of the Court are considered above. Here, we are interested in its relevance to the specific issue of eligibility criteria for judges of the Supreme Court. Let’s return to the question that started this section: Is Parliament authorised to unilaterally impose a statutory bilingualism requirement for judges of the Supreme Court? On this point, the accounts of section 41(d) offered by Grammond and Cyr adopt a qualitative approach but lead to the wrong result. Further, the broader readings of the Supreme Court Act Reference reach the correct outcome but they are too formalistic in their approach. Requiring French-English bilingualism for judges of the Supreme Court would enhance the Court’s legitimacy within a constitutional culture that is officially bilingual and the Court’s competence as the final court of appeal. As noted above, this might be taken as legislation that ‘enhances, rather than transforms’ the Court or ‘implements and advances structural constitutional principles’ such that it is ‘constitutional in character without being constitution in status’.83 Such a characterisation would, under Newman’s analysis, render the legislation a valid exercise of Parliamentary authority under either section 91 or 101 of the Constitution Act 1867, akin to the Official Languages Act or the Multiculturalism Act.84 But of ultimate importance for determining whether Part V is triggered in the case of a legislative bilingualism requirement is whether it would bring about a qualitative change to the composition of the Court and a qualitative assessment of the mandatory bilingualism proposals seems to suggest that the effects of such a proposal would be transformative to the composition of the Court within the meaning of Part V. This conclusion would not, as others have suggested, simply be a consequence of the proposal’s addition to the existing eligibility criteria. Rather it would be a consequence of the proposal’s substantive effect on the conception of composition of the Court, as set out in section 41(d) of the Constitution Act 1982. A legislated ­bilingualism requirement would introduce a new marker of identity that all judges of the Court must possess. By formally prioritising that marker over

80 

ibid, paras 26, 52, 80. ibid, paras 75, 78, 82. 82  ibid, paras 60–3. 83  See Newman (n 56) 117. 84 ibid. 81 

290  Kate Glover others, a ­bilingualism requirement would seem to impose a new constraint on the flexibility that currently attaches to the exercise of composing the Supreme Court bench. While this flexibility has failed in most ways to realise a bench that can be said to be representative, formalising a requirement that all judges must be bilingual diminishes this flexibility by adding another small measure of homogeneity to a constitutional conception of composition that does not currently contemplate it. The composition of the Court will always have individual and institutional dimensions. Some qualities, such as membership in the legal profession, are justifiable requirements for all members of the Court. Other qualities, such as the guaranteed seats for judges from Quebec, are directed at the institutional dimensions of composition, ensuring that the legal traditions and values of Quebec are properly represented on the bench as a whole. The current constitutional configuration of composition is open as to how other communities and qualities will be represented. It seems that formally privileging one such measure, and at the same time potentially limiting diversity and the representation of other constitutionally significant values and aspirations within the composition of the Court as a whole, represents a qualitative change to the current entrenched constitutional configuration governing ‘composition’ of the Court. If this is the case, a formal bilingualism requirement can be lawfully implemented only by virtue of the unanimity procedure set out in section 41(d) of the Constitution Act 1982. Addressing the amendment question at stake in the bilingualism debate is timely given current legislative proposals and executive policy. But it is also a case study that strikes at the foundations and traditions of constitutional amendment in Canada more broadly. It does important analytical work for thinking through the meaning and application of Part V in both the general and particular case. On the issue of interpreting Canada’s amending procedure, the bilingualism debate reveals gaps in the doctrine of Part V and the difficulties in determining both what is entrenched in the Constitution of Canada and what amounts to an amendment.85 That said, a discussion of the procedural question at stake in the bilingualism debate should not eclipse continued reflection on the appeal and desirability of pursuing a bilingualism requirement for judges of the Supreme Court. Such a

85  On the issue of bilingualism as an eligibility criterion, a general practice of appointing bilingual judges to the Court by political actors might not trigger the amending formula, notwithstanding the effect of legislation achieving pursuing the same goal. Uncertainty arises when a general practice is codified in policy, but not legislative, form. For instance, does a written policy indicating that the Prime Minister will only appoint bilingual candidates trigger the amending formula? This informal method of action is different from the enactment of ‘organic’ constitutional statutes contemplated by WJ Newman, ‘Putting One’s Faith in a Higher Power: Supreme Law, the Senate Reform Reference, Legislative Authority and the Amending Procedures’ (2015) 34(2) National J Const L 99, 112–13, which are ways of testing out policy initiatives before pursuing the formalities required for amendment. Moreover, this is not a general claim that practices can never constitute a constitutional amendment. One possibility of a practice that could, over time, amount to an ‘unconstitutional constitutional amendment’ is former Prime Minister Harper’s refusal to appoint senators. On unconstitutional constitutional amendments in Canada generally, see R Albert, ‘The Theory and Doctrine of Unconstitutional Constitutional Amendment in Canada’ (2015) 41 Queen’s L J 153.

Hard Amendment Cases in Canada 291 requirement would respond to an important and longstanding call to appreciate the undeniable linguistic competencies that attach to the role of a Supreme Court judge. The ­significance of attending to this appreciation of competencies has constitutional grounding in two of the legal traditions of the country, in the official languages of Canada, and in Canadian constitutionalism’s deep commitments to minority and language rights. That said, there is a concern that crystallising a French-English bilingualism requirement that does not allow for any measure of exception may in effect be at odds with the ideals of diversity, representation, and legal and substantive pluralism that characterise Canadian constitutionalism. We must be attentive to the effect of this prioritisation of bilingualism. Does prioritisation hinder the representation of other communities? Does it exacerbate marginalisation? Who does this requirement privilege? Continued reflection on the merits of a bilingualism requirement must include a more robust analysis of how mandatory bilingualism reconciles with representation and diversity on the Supreme Court bench more generally. IV. CONCLUSION

This chapter has raised considerations that should guide the interpretation and application of the formal amending procedures in Canada both generally and in the specific case of Supreme Court reform. It focuses on formal processes of reform, exploring boundaries and opportunities of the amending procedure set out in Part V of the Constitution Act 1982. This focus on Part V should not, however, distract from the reality that Part V has not been—and never will be—the primary way by which the Constitution—or the Court—changes. There are many other important routes of change, each governed by a logic of procedure, argumentation, and analysis, some deliberately invoked in the pursuit of Court reform and others not. We find these routes of change by looking to custom, executive policy, legislation, judicial decision, constitutional culture, civic action, intergovernmental agreement, internal procedures, and so on. This capacious understanding of constitutional change is the beauty and the burden of the richness of Canada’s constitutional order. The discussion in this chapter highlights how lessons about the boundaries of Part V are also lessons in what falls on and pushes up against the other side of these lines and that these matters and processes provide the context in which Part V should be interpreted and applied. For example, discerning the scope of the entrenched constitution stakes boundaries around what types of change warrant the protection and opportunities that come with multilateral consensus. Similarly, setting the boundaries around what counts as an ‘amendment’ within the meaning of Part V bears on what counts as legitimate instances of change by alternate means. Further, interpreting the reach of the multilateral amending procedures is simultaneously an exercise in interpreting the grants of unilateral authority to amend the Constitution. Ultimately, in discerning the boundaries of Part V and applying Part V in ­particular cases, we are forced to identify whose interests are put at stake by a proposal for reform. This question implicates another inquiry: whose interests ­ could possibly count? That is, what are the relevant constituencies to choose from?

292  Kate Glover As the law ­currently stands, Part V is concerned only with the interests of the ­central and ­ provincial legislatures. And yet the role of pluralism and multijuralism in ­Canadian constitutionalism compel us to ask what other interests might be relevant and should, as a matter of constitutional expression, be considered when thinking through the merits of reform and the processes of implementation. Such is a q ­ uestion that will ultimately shake—but necessarily so—the foundations and traditions of constitutional amendment in Canada.

15 Formal Amendment Rules and Constitutional Endurance The Strange Case of the Commonwealth Caribbean DEREK O’BRIEN

I. INTRODUCTION*

W

RITING IN THE early 1960s about the wave of decolonisation that was occurring across the British Empire, the distinguished constitutional scholar, Kenneth Wheare, predicted that these former colonies, which had attained their independence by means of an Act of the imperial Parliament, would very quickly move to replace their independence constitutions with ‘autochthonous’ constitutions enacted by their own sovereign parliaments. This prediction has proved largely accurate in the case of many of Britain’s former colonies, in Africa for example, which were replaced in their entirety almost before the ink was dry on the paper, but it has proved to be almost entirely inaccurate in the case of Britain’s former colonies in the Commonwealth Caribbean. With the exception of Guyana and Trinidad and Tobago, all of the remaining countries in the region retain the Constitutions that were bestowed upon them by means of an Act of the British Parliament upon their independence. In some cases, for example Jamaica, the independence Constitution has endured for over 50 years, and Jamaica is followed closely by Barbados where the independence Constitution has endured for 49 years. The Constitutions of these two countries have thus already exceeded the average lifespan for democratic constitutions, which is 42 years,1 and there are others which will soon exceed this benchmark.2 Elkins et al have famously argued that one of the key factors in explaining a constitution’s endurance is its flexibility; that is, the inclusion of formal amendment procedures which allow the Constitution to adapt to meet changing political, social and economic circumstances. According to Elkins et al this flexibility is crucial to

* 

This chapter was current as of 1 June 2016. Z Elkins, T Ginsburg, and J Melton, The Endurance of National Constitutions (Cambridge ­University Press, 2009) 32. 2  For example, The Bahamas, which gained independence in 1973. 1 

294  Derek O’Brien constitutional endurance because ‘it offers an ongoing and inclusive alternative to wholesale replacement of the constitution’, thus helping to maintain the core elements of the constitutional bargain while all the time generating ‘vital constitutional politics’.3 Rigid constitutions, by contrast, because they do not allow for the readjustment of the constitutional bargain from time to time are much less resilient and much more likely to be replaced in their entirety.4 While not wishing entirely to disavow this theory, I propose to show in this chapter that the longevity of the independence constitutions of the Commonwealth Caribbean does not fit comfortably within this account of constitutional endurance for at least two reasons. The first is that the Constitutions of a number of the eastern Caribbean islands are extremely rigid, and include some of the most onerous referendum requirements to be found anywhere in the Commonwealth. Yet, as I will demonstrate, they have proven to be no less resilient than the Constitutions of neighbouring countries which are less rigid. Their endurance is all the more remarkable in view of the profound disconnect between the procedures that were followed at the time of their creation by the so-called constituent power, Britain, and the constraints that the constituent power placed on the constituted power in order to amend the Constitution.5 The second reason is that even in those countries with relatively flexible constitutions, which permit constitutional amendment by means of a simple majority in a referendum or a special legislative majority, constitutional reform has still proved to be remarkably difficult to achieve. There have thus been relatively few amendments to these ostensibly flexible constitutions, and where they have occurred they have tended to be conservative, ‘directed toward perfecting the system by preserving its essence intact’.6 I propose, therefore, to offer an alternative explanation for constitutional endurance in the region, which sheds a different light on the rationale for the inclusion of formal amendment rules in the context of the decolonisation of the region, and which is informed by Ginsburg and Melton’s theory of ‘amendment culture’.7 This involves recognising that within any constitutional system there is ‘a borderline level of resistance to formal constitutional change’, which can ‘vary according to the political weight attached to the value of entrenchment’.8 Amendment culture thus encompasses ‘a set of attitudes about the desirability of constitutional amendment, which exists independently of the substantive issue under consideration’.9 I will argue that this is a particularly helpful analytical tool in the context of the Commonwealth Caribbean because it suggests that there may be a contextual explanation

3 

Elkins et al (n 1) 82.

4 ibid.

5 As discussed by Yaniv Roznai, ‘Amendment Power, Constituent Power, and Popular Sovereignty: Linking Unamendability and Amendment Procedures’, in this volume. 6  P Sutton, ‘Democracy in the Commonwealth Caribbean’ (1999) 6(1) Democratization 67–86, at 69. 7 T Ginsburg and J Melton, ‘Does the Constitutional Amendment Rule Matter at All? Amendment Cultures and the Challenges of Measuring Amendment Difficulty’ (2014) Public Law and Legal Theory Working Paper 682/2014 accessed 1 August 2016. 8 ibid. 9  Ginsburg and Melton (n 7) 12.

Formal Amendment Rules and Constitutional Endurance 295 for the failure of the region’s governments to effect constitutional reform, which does not depend exclusively on the existence of particular institutional obstacles, and which, while subject to some local variation, also functions at a regional level. I intend, however, to go further than merely asserting the existence of such a culture by identifying the factors that have contributed to the emergence of such a culture across the region. The paper is in two parts. In Part I, I will discuss the rationale for the inclusion of formal amendment rules in Commonwealth Caribbean constitutions, which had very little to do with promoting flexibility and everything to do with securing the survival of the region’s political leaders in the post-independent state as well as, in some cases, preserving in perpetuity the system of government inherited from the former colonial power. In Part II, I will explore the impact of these formal amendment rules upon post-independence constitutional reform in the region and their interaction with the region’s amendment culture. In conclusion I will argue that while the experience of the Commonwealth Caribbean may not invalidate Elkins at al’s theory about constitutional design and the importance of flexibility it does demonstrate that a theory which is based on constitutional design alone is not sufficiently rich to capture the other forces that may contribute to constitutional endurance. Particularly for postcolonial societies, such as those in the Commonwealth Caribbean, it is necessary to take account of the experience of colonial rule and how this informed both the institutional structure for amending the Constitution as well as attitudes towards constitutional amendment in the postcolonial era. II.  THE RATIONALE FOR THE INCLUSION OF FORMAL AMENDMENT RULES IN COMMONWEALTH CARIBBEAN CONSTITUTIONS

Though there is some variation as between Commonwealth Caribbean constitutions in terms of the provisions that they entrench and the degree to which those provisions are entrenched, the two most common amendment rules to be found in these constitutions are special legislative majorities and referendum requirements. In most countries special legislative majorities are necessary, but not sufficient, to effect constitutional amendment since, regardless of the size of the special legislative majority, the amendment still needs to be approved by a prescribed majority of citizens voting in a referendum. A.  Special Legislative Majorities In some cases a special legislative majority involves no more than an absolute ­majority of all the members of the House, whether or not all the members of the House are present when the vote is taken.10 In most cases, however, the requirement is more demanding; involving a two-thirds or even a three-quarters parliamentary majority

10 

See, for example, Barbados, s 49(4) and Jamaica, s 49(b).

296  Derek O’Brien in order to approve a constitutional amendment, depending upon the importance of the constitutional provision to be amended. This majority can take different forms. In countries, such as Antigua, Belize and St Lucia, it is two-thirds or three-quarters of all the members of the House of Representatives only; thereby allowing the elected House to bypass the Senate, which is wholly nominated.11 In those countries with unicameral parliaments, such as Dominica, Guyana and St Vincent, it is two-thirds of all the elected members;12 thus again allowing the elected element to bypass the nominated element. In the Bahamas, Barbados, and Jamaica, however, it is two-thirds of all members of both Houses of Parliament.13 In Trinidad and Tobago too, for the more deeply entrenched provisions, it is a three-quarters majority of all members of the House of Representatives and a two thirds majority of all members of the Senate.14 The standard rationale for the inclusion of a procedure for amending any constitution is the need for constitutions to adapt to changes in society. Over an extended period of time a constitution may no longer adequately meet the needs of the society for which it was designed or may embody principles that have long since been rejected by that society.15 In such circumstances it will be necessary to have some means of amending the Constitution. However, amendment rules that insist on special procedures, such as special legislative majorities, arguably detract from this objective by making the Constitution more difficult to amend, and opens amendment rules to the criticism that they are counter-majoritarian and, therefore, undemocratic.16 This criticism is usually countered by two arguments. First, that the broad consensus which is required to achieve a special legislative majority expands the class of persons whose interests are taken into account before a constitution can be amended, thereby making the process, arguably, even more democratic than that required for enacting ordinary legislation.17 Secondly, that because the support of the opposition is usually required to satisfy a special legislative majority such a provision creates ‘a climate or environment of deliberation’ about the content of the proposed amendment and forces those advocating the amendment to advance arguments ‘based on general and abstract principle while avoiding narrow partisan or sectoral interests’.18 This should, in turn, ensure that the amendment becomes part of the Constitution ‘with a near conclusive presumption of legitimacy’.19 These may be compelling arguments for the inclusion of formal amendment rules, such as the requirement for a special legislative majority, but they were certainly not articulated at the time the region’s independence constitutions were being drafted.

11  Antigua, s 47(2), though the Senate can delay the Bill for one parliamentary session (s 55); Belize, s 69(3) and (4); and St Lucia, s 41(3) to (5). 12  Dominica, s 42(2); Guyana, Art 164(2); and St Vincent, s 38(2). 13  Bahamas, Article 54(2); Barbados, s 49(2); Jamaica, s 49(40(a). 14  Trinidad and Tobago, s 54(3)(i) and (ii). 15  R Ku, ‘Consensus of the Governed: The Legitimacy of Constitutional Change’ (1995) 64 Fordham L Rev 535, 542. 16  CL Eisgruber, Constitutional Self-Government (Harvard University Press, 2001) 10–25. 17  LG Sager, ‘The Birth Logic of a Democratic Constitution’ in J Ferejohn et al (eds), Constitutional Culture and Democratic Rule (Cambridge University Press, 2001) 131–33. 18  J Ferejohn and L Sager, ‘Commitment and Constitutionalism’ (2003) 81 U Tex L Rev 1929, 1957. 19 BP Denning and JR Vile, ‘The Relevance of Constitutional Amendments: A Response to David Strauss’ (2002) 77 Tul L Rev 247–82.

Formal Amendment Rules and Constitutional Endurance 297 Instead, the main reason for including a requirement for a special legislative majority, and thus making the Constitution more difficult to amend, was to secure the survival of the region’s political leaders in the post-independent state. Parkinson thus argues that in Jamaica, where the negotiations surrounding independence took place against the backdrop of an upcoming election, the result of which was uncertain, with the two main parties—the People’s National Party and the Jamaica Labour Party—anxious about the rise of the minority People’s Progressive Party, the two main parties sought to secure their dominance by making provision for a two-party system and enshrining the office of the Leader of the Opposition within the independence Constitution.20 This was then ‘locked-in’ by the inclusion of an entrenchment mechanism, which ensured that provisions, such as those establishing the Leader of the Opposition, could not be altered without a two-thirds majority of both Houses of Parliament.21 Far from being concerned to facilitate constitutional amendment in the future, or to deepen and enrich the debate that should precede constitutional amendment, the inclusion of a requirement for special legislative majorities was thus motivated primarily by considerations of realpolitick. One of the foremost scholars of the new Commonwealth constitutions that emerged in the wave of decolonisation that took place in the late 1950s and early 1960s, SA de Smith, has also written of the level of distrust that existed amongst political leaders of Britain’s former colonies who feared that they might be the ­losers in the independence stakes.22 In de Smith’s view it was this distrust that led directly to the inclusion of rigid amendment procedures in these new Commonwealth constitutions: [I]n drafting a new constitution for a new state it may be unrealistic to begin with presumptions in favour of brevity and flexibility. Often it will be politically impossible to obtain general agreement on a new constitution unless it is both lengthy and rigid. Those who don’t expect to find themselves in power on Independence Day may well be profoundly distrustful of the majority party and its leaders, and the price of their acquiescence in the new order is therefore likely to be a somewhat cumbersome constitutional machine with built-in resistances against subsequent modifications.23

The most rigid amendment procedures to be found in Commonwealth Caribbean constitutions are the referendum requirements, some of which are especially o ­ nerous, but, as well see below, their inclusion also owes much to the British Government’s distrust of the region’s independence leaders. B.  Referendum Requirements The Constitutions of all but three countries in the region—Barbados, Trinidad and Tobago, and Belize—stipulate that various constitutional provisions require, in 20  C Parkinson, Bills of Rights and Decolonization: The Emergence of Domestic Human Rights Instruments in Britain’s Overseas Territories (Oxford University Press, 2007). 21  Jamaica Constitution, s 49. 22  A distinguished constitutional scholar and legal adviser to the British Government. 23  SA de Smith, The New Commonwealth and its Constitutions (Stevens and Sons, 1964).

298  Derek O’Brien a­ ddition to the approval of a special legislative majority, the approval of a prescribed majority of citizens voting in a referendum before they can be amended. In the three countries that do not include a referendum requirement it was adjudged that this additional barrier to constitutional amendment was unnecessary. As the Wooding Commission, which a decade after independence had been charged with reviewing and recommending reforms to Trinidad and Tobago’s independence Constitution, explained in response to a proposal to include a referendum requirement in the revised Constitution: We disagree. In our view, a referendum is not a particularly accurate method of determining the state of public opinion on issues of constitutional reform. Under a system of party politics it is quite probable that many an answer given will not be an answer on the merits of the question asked, but will merely reflect loyalty to what is known to be the party’s view, lest defeat of the party on the issue submitted should result in a consequences too undesirable to be permitted. Accordingly, we reject the idea of a referendum as the final prerequisite for amending the Constitution.24

Everywhere else, however, provision was made for certain constitutional amendments to require the additional approval of voters in a referendum. In most cases a simple majority of the electors voting in the referendum is all that is needed.25 However, in three countries—Antigua,26 St Vincent and the Grenadines,27 and ­Grenada28—approval by a two-thirds majority in a referendum is necessary.29 In the Federation of St Kitts and Nevis also, a two-thirds majority of the voters in Nevis is required in order for Nevis to secede from the Federation.30 To appreciate the rationale for the inclusion of such onerous referendum requirements it is necessary first to understand the decolonisation process in the eastern Caribbean and the extent to which this was informed by the British Government’s fears of what might happen after independence. C.  Constitution-Making in the Eastern Caribbean Immediately prior to independence the islands of the eastern Caribbean—Antigua and Barbuda, Dominica, Grenada, St Kitts and Nevis, St Lucia and St Vincent and the Grenadines—enjoyed ‘Associated State’ status under the West Indies Act 1967 (WIA 1967). As well as bestowing full internal self-government upon these countries,

24 Constitution Commission of Trinidad and Tobago, ‘Report of the Constitution Commission (22 January 1974) 430–1. Available at accessed 1 August 2016. 25  The Bahamas, Dominica, Guyana, Jamaica, St Kitts and St Lucia. 26  Section 47(5)(c). 27  Section 38(30(b). 28  Section 39(5)(c). 29  In St Kitts and Nevis, a two-thirds majority of the electors in Nevis in a referendum is required if Nevis wishes to secede from the Federation of St Kitts and Nevis. Section 113 of the Constitution. 30  In Jamaica too, a two-thirds or three-fifths majority may be required in a referendum, but this is only if the amendment has previously been rejected by the Senate.

Formal Amendment Rules and Constitutional Endurance 299 section 10 of the WIA 1967 made provision for the termination of their status as Associated States prior to independence. This differed depending upon whether it was the British Government or the Associated State that wished to terminate the association. In the first case, the association could be terminated by a relatively straightforward procedure, which required only that the British Government give the Associated State concerned six months’ notice of its intention to terminate the status of association, followed by an Order-in-Council of Her Majesty approved by a resolution of each House of the British Parliament.31 In the second case, a Bill providing for termination required the support of not less than two-thirds of all the elected members of the legislature of the Associated State and no less than two-thirds of the votes cast in a referendum.32 The WIA thus offered a mechanism through which the will of the people could be expressed by holding a referendum prior to termination of the association with Britain. Deciding which of the available routes to independence should be followed proved, however, to be a highly contentious matter in a number of the Associated States. This was because in many cases opposition groups, even if they supported independence in principle, wanted a referendum to be held prior to termination of the association, whereas the British Government preferred to proceed by a simple Order-in-Council. One of the most notorious examples of disagreement between opposition groups and the British Government occurred in Grenada, which at the time was governed by the Grenada United Labour Party, led by Eric Gairy, who had once been memorably described in a Colonial Office briefing as ‘a man who is almost too bad to be true’.33 Gairy had previously been excluded from Grenada’s Legislative Council as a result of electoral malpractice during the 1957 elections, and in June 1962 had been dismissed from office as Chief Minister by the British Government as a result of a report of the Commission of Enquiry into the Control of Public Expenditure in Grenada.34 Though Gairy was returned to office, following his party’s victory in the 1967 elections, his authoritarian style of leadership had not changed, and in the years leading up to independence he embarked on what Mawby has described as ‘a policy of repression in which any challenges to his personal rule were interpreted as a threat to political order in Grenada’.35 It was during this period that Gairy established a special secret police force made up principally of ex-convicts,36 known locally as the ‘Mongoose Gang’. Their role, which has been compared to that of Haiti’s Tonton Macoutes, was to intimidate opposition groups, such as the ­emerging New Jewel Movement (NJM): a coalition of Marxists, Black Power activists and

31 

WIA 1967, s 10(2). 1967, s 10(1). There would also have to be an interval of not less than 90 days between the introduction of the Bill and its second reading. 33  TNA: CO 1031/5218, Personality Notes for the Windwards Constitutional Talks, 18 April 1966. Quoted by S Mawby, Ordering Independence: The End of Empire in the Anglophone Caribbean, 1947–69 (Palgrave Macmillan, 2012) 216. 34 The report found that Gairy had been implicated in ‘the deliberate and systematic violation of financial regulations, the browbeating of public servants’ and ‘the illegal purchase with public monies of luxury items’. For the full report see Parliamentary papers 1961–1962: Cmnd.1735. 35  Mawby (n 33) 215–22. 36 ibid. 32  WIA

300  Derek O’Brien liberal reformists. NJM leaders were subject to brutal attacks by Gairy’s Mongoose Gang, which left them with serious head injuries, broken jaws and teeth,37 and which culminated in the in the killing, in January 1974, of one of the leaders of the NJM, Rupert Bishop.38 Notwithstanding Gairy’s appalling record of authoritarianism and of brutal political repression, the British Government was happy to terminate Grenada’s status of association by means of a simple Order-in-Council, rather than accede to the opposition’s request for a referendum to be held. While it may have been the most notorious, Grenada was not the only example of an eastern Caribbean island where there was profound disagreement about whether or not a referendum should be held prior to the termination of association with Britain. In St Lucia, tensions between the governing and opposition parties over this issue were of such concern to the British Government that the possibility of having a Royal Navy Ship close at hand as a precautionary measure was actively considered.39 Nevertheless, the British Government was determined to proceed to independence without a referendum on the basis that: Any attempt by the British Government to delay a decision on independence would be to go against the wishes of an elected Government with universal adult suffrage and be seen locally as support for the St Lucia opposition. This would cause dismay among other Commonwealth Caribbean Governments and could have a serious effect on the progress of the three remaining Associated States to independence.40

The essence of the British Government’s thinking on the utility of referendums in this context is perfectly encapsulated in a report to the Minister of State for Foreign and Commonwealth Affairs in connection with the Dominican Government’s request for independence. Responding to demands by opposition groups for a referendum to be held in Dominica prior to the termination of its association with the United Kingdom, the report’s author summarily dismissed the suggestion on the basis that: ‘It [a referendum] seemed to me a rather blunt instrument to use to decide complex constitutional issues.’41 The author of the report was, instead, supportive of the Dominican Government’s decision not to hold a referendum that would require the approval of a two-thirds majority of its citizens: ‘Few governments would be confident of obtaining such a massive plurality on any public issue.’42 In conclusion, the author of the report candidly admitted that: The British Government’s consistent policy [has] been to be guided by what seemed to ­British Ministers to be the wishes of the majority of the territory. I do not think that further evidence of popular opinion can reasonably be demanded.43

37 

HC Deb 11 December 1973, vol 866 cc331–61. Bishop’s son, Maurice Bishop, went on to lead the coup which eventually removed Gairy from power in 1979. 39  Report of St Lucia Constitutional Reform Commission 2011, p 62, unpublished. On file with author. 40 ibid. 41  Dominica, Termination of Association: A Report to the Minister of State for Foreign and Commonwealth Affairs by RN Posnett. Cmnd. 7279, para 39. 42 ibid. 43 ibid. 38  Rupert

Formal Amendment Rules and Constitutional Endurance 301 Notwithstanding the British Government’s scepticism about the value of referendums as a measure of the expression of the will of the people, referendum requirements were included in the independence Constitutions of all six Associated States. As Tierney argues, a referendum requirement may be justified in plural societies where they are necessary specifically to protect the interests of particular minorities and the consent of the specific minorities should be part of any legitimate consensus on the issue at hand.44 Such a justification did not, however, apply in the case of these Associated States, which were not in any sense plural societies. The inclusion of a referendum requirement in their case arose instead from the British ­Government’s deep distrust of the region’s independence leaders; especially Gairy in Grenada, Vere Bird in Antigua, and Ebenezer Joshua in St Vincent and the Grenadines, all of whom had fallen foul of the Colonial Office at different times on their country’s journey to independence.45 It is thus no coincidence that the independence Constitutions of these three countries contain some of the most heavily entrenched constitutional provisions to be found anywhere in the Commonwealth, let alone the Caribbean, requiring ­two-thirds approval in a referendum before they can be amended.46 Such onerous referendum requirements are self-evidently counter-majoritarian. They permit a relatively small percentage of the population to veto constitutional reform. As Oran Doyle observes in his chapter, ‘Constraints on Constitutional Amendment Powers’, constitutional devices that place a polity’s democratic structure beyond the reach of contemporary democratic majorities are, democratically, deeply suspect. This did not, however, appear to concern the British Government at the time. As the report on the termination of association of Dominica made clear, the British Government was perfectly aware that the possibility of any government being able to secure a two-thirds majority in a referendum on constitutional reform was vanishingly remote. The inclusion of such a requirement was thus clearly intended to place the system of government inherited upon independence beyond the reach of a future democratic majority of the citizens of these islands, thereby guaranteeing the preservation of the Westminster model whatever happened in the post-independence era. III.  POST-INDEPENDENCE CONSTITUTIONAL REFORM AND AMENDMENT CULTURE

In this section I wish, first, to examine the impact of the two main ­categories of ­ formal amendment rules—special legislative majorities and referendum ­requirements—upon post-independence constitutional reform. I will then proceed to explore the i­nteraction of these amendment rules with the region’s amendment culture.

44 S Tierney, Constitutional Referendums: The Theory and Practice of Republican Deliberation (Oxford University Press, 2014) 271. 45  Mawby (n 33) 215-31. 46  St Vincent and the Grenadines Constitution, s 38; Grenada Constitution, s 39; and Antigua and Barbuda Constitution, s 47.

302  Derek O’Brien A.  Special Legislative Majorities It is an indisputable fact that where constitutional reform has occurred in the region, with the exception of Guyana’s ‘socialist’ Constitution of 1980 (which is discussed in more detail below), it has occurred where there has been no referendum requirement to satisfy. It is no coincidence, therefore, that two of the most constitutionally active countries in the region have been Trinidad and Tobago and Belize, both of which have constitutions that do not include a referendum requirement. Trinidad and Tobago was thus able in 1976 to move from a constitutional monarchy to a republic by means of a two-thirds legislative majority. This was no obstacle to Eric Williams and his Peoples National Movement, which had led the country into independence, and which had consistently won at least two-thirds of the available seats in parliament in each of the post-independence elections up until 1986. In Belize there were no less than eight amendments to the Constitution in the decade between 2001 and 2011.47 Again this was possible because in each of the general elections from 1998 up until 2008 the winning party had won over three-quarters of the available seats in parliament and was, therefore, comfortably able to satisfy the special legislative majorities required by sectio 69 to amend the Constitution. It is also no coincidence that in the case of two of the other most constitutionally active countries—Guyana and Jamaica—the amendments which have been enacted have been to those provisions in each Constitution which were not subject to a referendum requirement. Thus, in 1970, Guyana was able to replace the Queen as head of state with a ceremonial President by a simple legislative majority because express provision had been made for such an amendment in the independence Constitution.48 In Jamaica too, the single most important constitutional reform that has taken place since independence—the replacement of the Bill of Rights included in its independence Constitution with a new Charter of Fundamental Rights and Freedoms—was possible because it could be implemented by means of a two-thirds majority of both Houses of Parliament and did not require approval in a referendum. B. Referendums There have, to date, been four constitutionally mandated referendums in the postindependence era: in Guyana, Nevis, the Bahamas, and St Vincent and the Grenadines. In this section I will first explore the local factors which were particular to each and how they contributed to the outcome before proceeding in the following section to discuss the wider amendment culture in the region and how this too contributed to the outcome. I will begin with what was, undoubtedly, the most controversial referendum to have taken place in the region since Jamaica voted to withdraw from the West Indies Federation in 1961, and that is Guyana’s referendum in 1978, which paved the way for Guyana to become a ‘socialist Republic’. 47  GP Smith, ‘Constitutionalism in Belize: Lessons for the Commonwealth Caribbean’, University West Indies, Faculty Workshop Series 2008–9, 2008. Unpublished. On file with the author. 48  Guyana Independence Constitution 1966, Art 73(5).

Formal Amendment Rules and Constitutional Endurance 303 i. Guyana In the period between independence and the establishment of a ‘socialist’ C ­ onstitution in 1980 there were a number of formal amendments to the Guyanese Constitution, none of which required a referendum to be held. For example, in 1970 the Queen was replaced as Head of State and the right of appeal to the JCPC was abolished.49 However, the constitutional reforms that could be achieved by means of a legislative majority alone were not enough to satisfy the Peoples National Congress (PNC), who argued that a wholesale reform of the 1966 Constitution was needed in order to achieve their political objective of transforming Guyana into a ‘Socialist C ­ ooperative Republic’. The first step towards achieving this objective was to introduce a Bill in the National Assembly to amend Article 73 of the Constitution, the effect of which would be to remove the referendum requirement for the constitutional reforms that the Government was contemplating. Thus, whilst the Bill to amend Article 73 would itself have to be ratified by a referendum, once the Bill had been enacted large parts of the Constitution could be amended by a two-thirds majority of the National Assembly. Since the PNC had hijacked the administration of elections shortly after its victory in the 1968 elections, and had secured 37 out of the 53 available seats in the National Assembly at the elections in 1973, which were themselves widely believed to have been rigged, achieving such a legislative majority presented no obstacle to the PNC. This would mean that in future the PNC would be able to amend large parts of the Constitution at their will. The referendum that preceded the amendment of Article 73 of the Constitution has been described by Lutchman as ‘one of the bitterest political and constitutional controversies in the history of Guyana’.50 It also confirms Tierney’s observation that the egregious manipulation of referendums tends to occur most often in countries where there is already an established pattern of electoral misconduct.51 During the campaign the Government conducted an ‘extensive and aggressive propaganda exercise’.52 This was made possible because, post-independence, the PNC had promulgated the idea that all institutions under public control, such as the media, should be mobilised in support of the socialist cause as embodied by the PNC.53 Opposition groups were not allowed to place advertisements in the state-owned media and pressure was brought to bear on those sections of the media that were not under direct governmental ownership; for example, cutting off the supply of newsprint or the termination of advertising contracts with the Government.54 Those who were opposed to the Government’s plans to amend Article 73 experienced difficulty in securing permission to hold their meetings. Even where permission was granted the meetings were disrupted by PNC supporters in the

49 

Guyana Independence Constitution 1966, Art 73(5). RW James and HA Lutchman, Law and the Political Environment in Guyana (Institute of Development Studies University of Guyana, 1984) 61. 51  Tierney (n 44) 102. 52  James and Lutchman (n 50) 64. 53  ibid 65. 54  ibid 70. 50 

304  Derek O’Brien presence of police officers who did not seek to intervene.55 The collective actions of the Government resulted in a decision by opposition groups to boycott the referendum altogether and the Government, unsurprisingly, succeeded in winning an outright majority in the referendum. Contrary to the PNC’s claims that a majority of the registered voters had taken part in the referendum and had overwhelmingly voted to approve the amendment of Article 73, Lutchman argues that the results of the referendum ‘must rank as among the most corrupt results ever in an election type exercise’.56 Certainly, there was widespread suspicion about the results, based on the fact that a ‘derisively minuscule’ proportion of the population had actually voted in the referendum.57 In protest, opposition groups refused to take part in the Constituent Assembly that was subsequently established by the Government and charged with the task of drafting the new Constitution. The result was that the draft Constitution that was approved by the Constituent Assembly was substantially the same in form and substance as the draft that had been submitted to the Constituent Assembly by the PNC.58 In the subsequent elections in 1980, which should have taken place in 1978, but were postponed to allow the referendum to take place, the PNC increased its majority still further, winning 41 out of the 53 sets in the National Assembly. ii. Nevis St Kitts and Nevis, which achieved independence in 1983, is the world’s smallest federation. Though the two islands had first been linked when a party of British colonists from St Kitts landed on Nevis in 1628, political relations between the islands have been tense ever since the British Government decided in 1882 to unite the islands (together with the island of Anguilla) into one administrative unit. This tension is reflected in section 113 of the independence Constitution, which expressly provides for Nevis to secede from the federation if this is approved by a two-thirds majority of all the members of the Nevis Assembly and two-thirds of Nevisians in a referendum. For the first decade of independence the secession issue remained latent. However, it resurfaced in 1996. The catalyst on this occasion was the federal government’s plans to regulate offshore financial services within the federation.59 The Nevis administration contended that the federal government’s real objective was to place the financial and business sectors of Nevis under its control and direction, the effect of which would have been to undermine the constitutional and legislative authority of Nevis at the same time as destroying its economy.60 The Concerned Citizens Movement (CCM) having won a majority in the 1997 elections to the Nevis Assembly in which the CCM had pledged to hold a referendum on secession, the Premier 55 ibid. 56 

ibid 71.

58 

ibid 75. Sir F Phillips, Commonwealth Caribbean Constitutional Law (Cavendish Publishing Limited, 2002)

57 ibid. 59 

137. 60 ibid.

Formal Amendment Rules and Constitutional Endurance 305 of Nevis, Vance Amory, immediately invoked section 113 of the Constitution and tabled a Separation Bill. The Bill was approved by all five members of the Nevis Assembly, and thus easily satisfied the two-thirds majority required by the Constitution. However, in the subsequent referendum, which was held on 10 August 1998, ‘unaccountably and to the surprise of nearly all observers’, the requisite two-thirds of the electorate did not vote in favour of secession. On what was a relatively low turnout only 61.7 per cent of those who took part in the referendum voted in favour of secession.61 Two possible reasons for the Government’s failure to win the necessary two-thirds majority in the referendum have been suggested. The first was the lobbying of the electorate by political leaders elsewhere in the region who were haunted by the spectre of secession in their own multi-island states: for example, Antigua and Barbuda and Trinidad and Tobago. The second, according to Midgett, was the absence of a meaningful political movement that could realise the goal of secession.62 Nevisians may have been ardent separatists but, in the final analysis, an insufficient number of Nevisians were ‘willing to take the political step of re-creating themselves as a nation-state’.63 iii.  The Bahamas The Constitution of the Bahamas has remained virtually unaltered in the 40 odd years of its existence. Prior to 2002 there had been some minor constitutional amendments, such as changing the definition of the financial year, none of which required a referendum. However, amending the Constitution in this way is very much the exception to the rule, since 104 of the 137 Articles of the Constitution require approval in a referendum before they can be amended.64 Accordingly, in 2002, when the Government wished to make a number of reforms to the Constitution, having achieved the necessary two-thirds majorities in both Houses of Parliament, it still needed to obtain majority approval in a referendum. The reforms that were proposed included: the removal of discriminatory provisions relating to women under the citizenship section; the inclusion of gender as a prohibited ground of discrimination; the establishment of a Parliamentary Commissioner; the transfer of the Attorney General’s powers relating to criminal proceedings to a newly established Director of Public Prosecutions; putting the Teaching Services Commission on an equal footing with other Public Service Commissions; the establishment of a Boundaries Commission to carry out the functions previously exercised by the Constituencies Review Commission; and an extension of the retirement age for senior judges. Though progressive, these reforms were, essentially, very

61  D Midgett, ‘Pepper and bones: the secessionist impulse in Nevis’ (2004) 78 New West Indian Guide 43, at 61. 62  ibid 67. 63 ibid. 64 ‘Report of Constitutional Review Commission of the Bahamas 2013’, para 12.1. Available at accessed 1 August 2016.

306  Derek O’Brien modest, leaving the basic structure of government virtually unchanged. Nevertheless, the Government still failed to secure the approval of a majority of the voters in the referendum. The Constitutional Review Commission, which reported in 2013, suggested that there were a number of interlocking reasons why the Government failed to win popular approval for its package of reforms. These included: the contamination of the referendum by other political controversies; the imminence of a general election; ambivalent feelings about the gender-equality aspects of the citizen-related proposals; the complexity of the Constitutional Amendment Bills; and, lastly, a lack of public education about the proposed amendments.65 iv.  St Vincent and the Grenadines St Vincent and the Grenadine’s 2009 referendum on constitutional reform offers the perfect antithesis to Guyana’s 1978 referendum, both in terms of its legitimacy and in terms of the efforts that were made by the Government to engage civil society in the reform process. This process began in 2003 with the honouring of a manifesto commitment by the United Labour Party to establish, on a statutory basis, a Constitution Review Commission (CRC), which was charged with consulting widely before making recommendations for the reform of the Constitution. The CRC was composed of 25 people from a cross-section of the political, social and legal community (including the diaspora). The process of consultation was thorough, with various publications being produced along the way, outlining the options available and summarising the level of progress achieved. There was also a significant effort to encourage input from across Vincentian society and the wider diaspora.66 The CRC’s report was published in 2005, and again in revised form in 2006. It was comprehensive and recommended a number of major reforms to the ­country’s Constitution. These included: the removal of the British monarch as head of state and her replacement with a ceremonial president; an increase in the number of elected members of the National Assembly from 15 to 17; the addition of seven ‘civil society’ senators; limiting the Prime Minister to two terms in office and reducing his role in bureaucratic appointments; capping the number of ministers to 13 to ensure that the cabinet does not have a majority in the legislature; and establishing a National Advisory Council of Elders (NACE) which would provide ‘the conscience of the nation’. Once the draft Constitution had been finalised 15,000 copies were printed and distributed cheaply to ensure that most people in the country—of approximately 30,000 households—would have access to it. In addition there were numerous town hall meetings where the Drafting Committee, along with members of the Constitutional Reform Steering Committee—a slimmed-down version of the CRC—and Government, discussed the process with the wider population. These meetings were

65 

ibid, para 8.4. following account of the referendum is drawn from ML Bishop, ‘Slaying the “Westmonster” in the Caribbean? Constitutional Reform in St Vincent and the Grenadines’ (2011) 13 British Journal of Politics and International Relations 420–37. 66  The

Formal Amendment Rules and Constitutional Endurance 307 broadcast live on national radio, continuing the pattern of television and radio coverage that had accompanied many of the parliamentary debates on the subject. All of this public deliberation and consultation was not enough, however, to persuade a sufficient majority of the electorate (two thirds) of the need for constitutional reform, and the Constitutional Reform Bill was, ultimately, rejected by 55 per cent of voters in the referendum. Bishop has suggested a number of reasons for the Government’s failure to win the referendum, such as the length of the draft Constitution, which was a massive tome, running to 160 pages, ‘with endless clauses and clarifications used to codify the myriad new institutions’, unlikely to enthuse a population that had already been subjected to six years of public debate about constitutional reform. Bishop also points out that the draft Constitution was very much a watered-down version of the much more radical reforms recommended by the CRC. There was no limit on the number of terms of office that a Prime Minister could serve, there was no room for the seven civil society’s senators, and no place for NACE. Ultimately, while the draft Constitution included some significant reforms it was not sufficiently ‘uplifting’ to transcend party politics. The Prime Minister of St Vincent, Ralph Gonsalves, while apportioning much of the blame for the failure of his government’s efforts to implement constitutional reform, was also highly critical of the two-thirds referendum requirement in his country’s independence Constitution. According to Prime Minister Gonsalves, this was a relic of colonialism that had been introduced by the British Government ‘to counter the type of authoritarianism that had been the hallmark of Eric Gairy in Grenada’ and ‘to neuter radical approaches to governance in the early post-­independence era’.67 Indeed, at a recent conference in Jamaica, Prime Minister Gonsalves revealed that he never realistically expected that his government would succeed in achieving a two-thirds majority in the referendum. Nevertheless, he pressed ahead with the referendum, hoping that it would be possible to secure a simple majority vote. This, he believed, would have provided him with a sufficient democratic mandate to invoke the constituent power by inviting the British Government to amend St Vincent’s Independence Constitution in the same way that it had been created, that is, by an Order-in-Council of Her Majesty.68 C.  Amendment Culture It is clear that there is a degree of correlation between constitutional design and constitutional amendment as measured by reference to the frequency of constitutional amendment in those countries where the only requirement is a special legislative majority and those countries where, in addition to a special legislative majority, a

67 On Constitutional Reform in St Vincent and the Grenadines, Statement to Parliament, 28 May 2009. Available at http://www.hansardsvg.com/styled/downloads-4/files/28th%20May%202009.pdf. Accessed 22 December 2016. 68  Note on file with the author.

308  Derek O’Brien majority of voters in a referendum is required. It is also clear, however, that, viewed in the round, the formal amendment rules have not served to fulfil the function envisaged by Elkins et al of promoting flexibility by allowing these constitutions to adapt to meet changing political, social and economic circumstances. More often than not, they have simply preserved the status quo ante, thereby fulfilling the role intended for them by the region’s political leaders at the time of independence and by the British Government. This does not, however, answer the underlying question of why so many governments have been unable to satisfy the amendment requirements of their constitutions. In the case of those countries with very onerous referendum requirements, it is possible to argue that a two-thirds majority represented an impossibly high threshold for any government to satisfy, as demonstrated by the failure of the Government of Nevis to persuade a two-thirds majority of its citizens to vote for secession. However, in the case of St Vincent and the Grenadines, the Government was unable to persuade even a simple majority of its citizens to vote for its package of reforms. The Government was, therefore, in no different position to the Government of the Bahamas, which also failed to persuade a simple majority of its citizens to vote for what were essentially very modest constitutional reforms. There were, of course, as we have seen in both cases, local factors at play. However, the lack of political consensus generally around constitutional reform and the inability of governments even to obtain a simple majority in a referendum process which they, effectively, control points to the existence of a wider amendment culture in the region, which transcends the institutional obstacles presented by formal amendment rules, and which exists ‘independently of the substantive issue under consideration’,69 But what factors have contributed to the emergence of such a culture in the region? One factor, which has been mentioned by scholars, such as Sutton, is the deeply conservative nature of the political culture of the region, which derives in part from the small size of almost all of the countries in the region (small states tending to be quite conservative), but also from the history of slavery and assimilation in the region.70 As Sutton explains, as ‘free people of colour’ emerged as elements of a middle class in the years following the abolition of slavery their value system was shaped by the metropolitan connection and the colonial elite who represented it in the individual colony.71 This was further buttressed by the brief tutelage of the region’s political leaders in the operation of the so-called Westminster model of government in the decade or so leading up to independence.72 This meant that by the time of independence local political leaders were very well versed in the system of government embodied in their independence constitutions. As Norman Manley, the Premier of Jamaica at the time its independence Constitution was being negotiated, explained: I make no apology for the fact that we did not attempt to embark upon any original or novel exercise for constitution-building. We had a system which we understood; we had 69 

Ginsburg and Melton (n 7) 12. Sutton (n 6) 69. 71 ibid. 72  JI Dominguez, ‘The Caribbean Question: Why has Liberal Democracy (Surprisingly) Flourished?’ in JI Dominguez et al (eds), Democracy in the Caribbean (John Hopkins University Press, 1993) 57. 70 

Formal Amendment Rules and Constitutional Endurance 309 been operating it for many years with sense. It’s a system which has endured in other countries for generations successfully. It is a system which is consistent with the sort of ideals we have in this country, and it was not difficult to decide that we would follow that familiar system with those modifications which we thought the circumstances of Independence deserved.

Within the region there was not, therefore, the sense of a lack of autochthony which Wheare had predicted would compel these newly independent countries very quickly to replace their constitutions with something ‘homegrown’. There was not the appetite for constitutional amendment which existed in Britain’s former colonies in Africa, such as Ghana and Tanzania. Instead, there was a popular attachment to the text of these constitutions which has not only survived independence, but which appears actually to have grown in the intervening period. A second factor is the political tribalism and adversarialism, which has become so engrained in the region’s political culture as a result of the ‘winner takes all’ nature of the ‘first past the post electoral system’, which is the electoral system of choice across the region.73 Where constitutional reform has taken place, as in Trinidad and Tobago and in Belize, it has occurred during a period when the Government has had a sufficiently large legislative majority that it has been able to press ahead with reform without the need for the support of the opposition. Absent an in-built majority, governments have found it almost impossible to build a political consensus around constitutional reform. In Jamaica, for example, it proved impossible for the People’s National Party when in power to garner the requisite two-thirds legislative majority to amend the Constitution in order to replace the Judicial Committee of the Privy Council (JCPC) with the Caribbean Court of Justice (CCJ) as the country’s final appellate court, even though the amendment enjoyed considerable popular support.74 In Trinidad and Tobago too, an attempt by the Government in 2011 to amend the Constitution to preclude constitutional challenge to the implementation of the death penalty by prisoners based on the grounds of delay or inhuman or degrading prison conditions failed because the Government was unable to achieve the 31 votes in the House of Representatives that were needed to amend the Constitution, even though the death penalty is extremely popular with voters in Trinidad and Tobago.75 While it is true that in Jamaica the Government was able to build the necessary political consensus to achieve the two-thirds majority of both Houses of Parliament required to replace its Bill of Rights with a new Charter of Fundamental Rights and Freedoms, it took nearly two decades, and a number of compromises with regard to the contents of the Charter, in order to achieve this consensus.76 Bishop also notes that the failure of the Government of St Vincent and Grenadines to win even a nominal majority in the referendum on constitutional referendum was

73  With the exception of Guyana, where a party list proportional representation system is deployed. See D O’Brien, The Constitutional Systems of the Commonwealth Caribbean (Hart Publishing, 2014) 73. 74  D O’Brien, ‘The Caribbean Court of Justice and its Appellate Jurisdiction: A Difficult Birth’ (2006) Pub L 344–63. 75  Constitution (Amendment) Capital Offences Bill 2011. 76  See D O’Brien and S Wheatle, ‘Post-Independence Constitutional Reform in the Commonwealth Caribbean and a New Charter of Fundamental Rights and Freedoms for Jamaica’ (2012) Pub L 683–702.

310  Derek O’Brien in no small part due to the fact that the bipartisanship, which had accompanied the creation of the CRC and which had been maintained for the first few years of public consultation, had completely dissipated by 2007, with the opposition New Democratic Party withdrawing from the entire process.77 Here, as elsewhere in the region, the levels of political tribalism and adversarialism generated by a ‘winner takes all’ culture created an environment of distrust which was not conducive to consensus-building around constitutional reform. As Contiades and Fotiadou argue in Chapter 11, ‘Amendment-Metrics: The Good, the Bad, and Frequently Amended Constitution’, a culture of distrust exerts an immense influence on the way that the mechanisms of constitutional amendment operate. A third factor which has contributed to amendment culture in a number of countries is the institutional structure surrounding the amendment process. As Ginsburg and Melton have argued, it is possible that a low threshold for amending the Constitution can lead to the exercise of caution when constitutional reform is being proposed precisely because it is too easy.78 Evidence of just such a phenomenon can be found in the emergence of an implied referendum requirement amongst those countries in the region with constitutions that do include an express referendum requirement: Barbados, Belize, and Trinidad and Tobago. In Barbados, for example, following the recommendation of the Forde Commission, in 1998, that Barbados should become a parliamentary republic, with a President replacing the Queen as Head of State,79 it would have been perfectly possible for the Government to have acted upon this proposal and to have amended the Constitution by means of a two-thirds majority in both Houses of Parliament. In both the 1999 and 2003 elections the Barbados Labour Party had won with large majorities. These translated into 26 of the 28 seats available in the 1999 elections and 25 out of the out of the 28 seats available in the 2003 elections. Notwithstanding its overwhelming electoral mandate, the Government decided, however, that such a fundamental constitutional reform needed to be put to the vote of the people in a referendum. A Referendum Bill was, accordingly, introduced in parliament in 2000, but for various reasons was not enacted until 2005. Though the Referendum Act 2005 did not itself fix a date for the referendum to be held, the Government decided that to save costs it should coincide with the date of the next general election in 2008. In the event, however, the referendum was not held in 2008 and has been deferred by successive governments ever since. In Trinidad and Tobago too it would be perfectly possible, from a strictly legal perspective, to amend the Constitution by means of a two-thirds legislative majority in order to abolish the right of appeal to the JCPC and to replace it with a right of appeal to the CCJ.80 However, the Prime Minister, Kamla Persad Bissessar, has recently announced that before any such amendment can take place it will have to be approved by the people in a referendum, in accordance with the recommendation of

77 

Bishop (n 66). Ginsburg and Melton (n 7) 14. Report of the Constitution Review Commission Barbados 1998. Unpublished. On file with author. 80  Constitution Trinidad and Tobago, s 54. 78  79 

Formal Amendment Rules and Constitutional Endurance 311 the 2013 Report of the Constitution Reform Commission (CRC).81 Though it is true that because of the strained relationship between the Government and the Opposition it would have been difficult for the Government to achieve the two-thirds legislative majority necessary to implement this amendment, the acceptance by the Prime Minister of the need for a referendum, where none is required by the Constitution, still represents a remarkable acknowledgment of implied constraints on the amending power of the constitutional legislator. The CRC has even gone one stage further by recommending that not only this reform, but also a host of other constitutional reforms which it was recommending, should only be introduced following majority approval in a referendum, though this is not strictly required by the Constitution.82 Finally, in Belize, which has been one of the most constitutionally active countries in the post-independence era, the Government eventually came unstuck in its efforts to reform the Constitution in 2008 when attempting to introduce legislation to remove the protection afforded by section 17(1) of the Constitution to the owners of: [P]etroleum minerals and accompanying substances, in whatever physical state located on or under the territory of Belize … the entire property and control over which are exclusively vested, and shall be deemed always to have been so vested, in the Government of Belize.83 In proceedings brought by a group of disgruntled landowners, Bowen v Attorney General,84 the proposed amendment was struck down, even though it had been approved by the special three-quarters majority required by section 69 of the Constitution, on the ground that it offended the ‘basic structure’ of the Constitution. According to the Chief Justice Conteh, section 69 was no more than a ‘procedural handbook’, whereas the Constitution is ‘the embodiment and immanation of the people of Belize as a whole: it is their constitution’.85 It followed, therefore, that a fundamental reform of the Constitution, such as the disapplication of the fundamental rights guarantee contained in section 17(1), could only be achieved by means of a referendum, even if this was not expressly required by the Constitution. Following the judgment in Bowen, the Government tried for a second time to amend section 17 of the Constitution by simultaneously amending section 69 of the Constitution to provide that once a Bill had received the approval of the special legislative majority required by the Constitution the constitutionality of the Bill could not be reviewed by the courts. Once again, however, the Government’s attempt to effect such a fundamental reform of the Constitution without securing the approval of the people in a referendum was struck down by the Supreme Court.86

81 ‘Report Trinidad and Tobago Constitution Reform Commission’, 27 December 2013, para 277 http://www.constitutionnet.org/files/national_consultaion_on_constitutional_reform.pdf. Accessed 22 December 2016. 82  ibid, para 33. 83  Belize Constitution (Sixth Amendment) Bill 2008. 84  BZ 2009 SC 2. Unreported. Available at . 85  This was because the Belize Act 1981 passed by the UK Parliament expressly recognised Belize’s selfgoverning status and the right of the Belizean Parliament to amend the Order-in-Council issued by Her majesty, which it did in September 1981. 86  British Caribbean Bank Ltd v Attorney General Belize, Claim No 597 of 2011. Unreported. Available at .

312  Derek O’Brien In all three countries then a norm is clearly emerging at the political/judicial level, which regards fundamental constitutional reform that has not been mandated by the majority of citizens in a referendum as democratically unacceptable, even if it appears to be permitted by the constitutional text. The emergence of such a norm points to a special political weight being accorded to the value of entrenchment in these countries and to an amendment culture which is responsive to the Constitution’s institutional structure. IV. CONCLUSION

As our survey of constitutional reform in the region indicates, there is only very limited evidence to support the theory that formal amendment rules contribute to constitutional endurance by allowing constitutions to be amended rather than being replaced in their entirety. As we have seen, there has been very little amendment of the region’s constitutions in the post-independence era. Indeed, there is plentiful evidence that in a number of cases the formal amendment rules included in the region’s constitutions have instead acted as a barrier to constitutional amendment. This is especially true of those countries with constitutions which include a referendum requirement. As we have seen, only one government, the PNC in Guyana, has so far succeeded in obtaining the support of the requisite majority of its electors in a referendum for its constitutional reform programme, and it is widely suspected that the PNC only managed to achieve this by manipulating all aspects of the referendum process. To this extent, at least, they have functioned in exactly the way they were designed to function by the region’s political leaders at the time of independence and by the British Government. It is thus with some justification that the Prime Minister of St Vincent and the Grenadines has complained that the very onerous referendum requirements included in his country’s Constitution were ‘a colonial anachronism’, obliging him to consider taking the extraordinary step of inviting the former colonial ruler, retrospectively, to amend his country’s Constitution. To paraphrase Joshua Braver in Chapter 7, ‘Hannah Arendt in Venezuela: The Supreme Court Battles Hugo Chavez Over the Creation of the 1999 Constitution’, the citizens of these countries were born into a constitutional history and narrative that was not of their choosing and from which it is nearly impossible for them to break free. It would be wrong, however, to conclude that referendum requirements are the only reason for the glacial pace of constitutional reform in the region. As we have seen, regard must also be had to the amendment culture abroad in the region, which is indicative of an underlying level of resistance to constitutional reform. This suggests an attachment to the text of the region’s independence constitutions that defies the predictions of scholars that they would soon be replaced after independence and is in marked contrast to the lack of attachment to their founding texts manifested by the citizens of Britain’s former colonies in Africa, in countries such as Tanzania and Ghana, which replaced their constitutions very shortly after independence. There are a number of factors that have contributed to the emergence of such a culture, including the experience of colonial rule in the Commonwealth Caribbean, which included

Formal Amendment Rules and Constitutional Endurance 313 an extended period of self-government prior to independence, and a degree of political tribalism which is inimical to building a consensus around constitutional reform. There is also the emergence of a norm in response to the threat of constitutional amendment by special legislative majority in those countries with constitutions that do not include a referendum requirement, which insists that fundamental constitutional reform demands the legitimacy that can only be conferred by a referendum. It is impossible to quantify with any degree of exactitude the strength of the amendment culture within a single country, let alone across an entire region. However, we should soon have a better appreciation of the strength of the region’s amendment culture as three constitutional referendums are scheduled to take place in the region within the next year: in Grenada, St Lucia and the Bahamas. In each case the referendum has been preceded by a Constitutional Review Commission, which has consulted widely and deliberated at length.87 A failure by the government to obtain a majority in any of these referendums could only add weight to the impression that, regardless of the substance of the amendments being proposed, there is a pervasive amendment culture at work across the region which is resistant to constitutional reform.

87  See, eg, the description of the consultation process outlined in the Report of St Lucia Constitutional Reform Commission, 1–18 March 2011, unpublished. On file with author.

314 

16 The French People’s Role in Amending the Constitution A French Constitutional Analysis from a Pure Legal Perspective JEAN-PHILIPPE DEROSIER

Un peuple a toujours le droit de revoir, de réformer et de changer sa Constitution. Une génération ne peut assujettir à ses lois les générations futures.1

F

RANCE IS A democracy, a modern democracy even, defined as an efficient and stable system, based on the universality of citizens.2 This is the result of more than 200 years of History, from the French Revolution in 1789. Proof, if at all necessary, lies today within the first articles of the current Constitution. Article 1: ‘France shall be an indivisible, secular, democratic and social Republic’. Article 2: ‘The principle of the Republic shall be: government of the people, by the people and for the people’. Article 3: ‘National sovereignty shall vest in the people, who shall exercise it through their representatives and by means of referendum’. This seems to place the French people at a very central, and even a fundamental, position: the people might very well be the source of all power. If this is undoubtedly the case, one needs to appreciate what this fundamental position does entail from a legal perspective. And this requires us to clearly establish the methodological scheme I will be using to question the people’s position as regards the amending of the French Constitution. As ‘legal perspective’ I understand a pure legal analysis, free from any non-legal arguments, such as political or sociological ones. It would certainly be possible and of great interest to include political and/or sociological aspects in this analysis. But it would not be possible in that case to identify if the people’s position and role are the result of a legal provision or that

1  Art 28 of the Declaration of human and civic rights of 23 June 1793: ‘A people has always the right to reconsider, to amend and to change its Constitution. A generation cannot subjugate to its laws the coming generations’. 2  To this definition one could add the elements provided by Professor Carcassonne who defines the modern democracy as a system in which ‘the governed people choose effectively the governing actors, the governing actors have effectively the means to govern and they are effectively responsible before the ones who are governed’, G Carcassonne, La Constitution. Introduite et commentée, 11th edn (Seuil, 2013) 21.

316  Jean-Philippe Derosier of a social or of a political will (or perhaps of all three). On the contrary, focusing on a study of the Law—and only of the Law—enables us to determine on which legal procedures and provisions the people can or cannot act. It is then necessary to introduce a distinction between the ‘Is-Sphere’ (Sein), that concerns the material facts, and the ‘Ought-Sphere’ (Sollen), that concerns the legal norms: this allows us to make the distinction between what ‘is’, in fact, from what ‘ought to be’, in Law. A norm is the translation of an act of will, with which it is commanded, empowered, permitted, derogated or prohibited.3 A norm expresses the will of the person stating it and it rules the behaviour of the persons for whom it is intended, according to its author’s will: ‘the individual who commands, permits, or authorizes wills; the man to whom the command, permission, or authorization is directed ought to’.4 A norm is then an ‘Ought’ (Sollen) and not an ‘Is’ (Sein): it is expressing what must be, a behaviour, that must (or can) be followed, but it does not guarantee that this behaviour is (or will be) followed indeed.5 Therefore, it has to be differentiated from the act of will itself: the norm is only its legal translation, the act of will being not a legal phenomenon (it does not belong to the ‘Ought-Sphere’) but a factual one (it belongs to the ‘Is-Sphere’).6 The hierarchical organisation of the legal norms (the ‘Norms’ hierarchy theory’) is based on this fundamental distinction between the ‘Ought-Sphere’ and the ‘Is-Sphere’. If the Law, as an ‘Ought’, is clearly different from an ‘Is’, it can only exist according to such an ‘Ought’ and not according to an ‘Is’. In other words, a norm is of a legal nature only if it is legally valid, and it can only be so if it fulfils the conditions of its validity which are stated by one or many other legal norms: a fact cannot be the basis of a legal norm.7 That means that ‘law regulates its own creation and application’8 and that only a legal norm can be the basis of another legal norm. This leads to a recursive structure of the legal order,9 explained by the Norms’ hierarchy theory.10 This hierarchy and its inherent recursive structure lead to the identification of a first norm, which forms the basis of the validity of all others. It is possible to consider that 3  See the general definition and function of the norm, H Kelsen, Pure Theory of Law (Max Knight tr of the 2nd edn, University of California Press, 1967) 3–23; H Kelsen, General Theory of Norms (Michael Hartney tr, Oxford University Press, 1991) 1–12. 4 Kelsen, Pure Theory (n 3) 5. 5  On this distinction between Sein and Sollen (Is and Ought), see Kelsen, Pure Theory (n 3) 4–10; Kelsen, Theory of Norms (n 3) 9–15 and 58–86; H Kelsen, Hauptprobleme der Staatsrechtslehre entwickelt aus der Lehre vom Rechtssatze, 2nd edn (JCB Mohr, 1923) 1 at 33. 6 Kelsen, Pure Theory (n 3) 5. 7  ‘The reason for the validity of a norm can only be the validity of another norm’, Kelsen, Pure Theory (n 3) 193. 8 Kelsen, Pure Theory (n 3) 71. 9 O Pfersmann, ‘La production des normes: production normative et hiérarchie des norms’ in D Chagnollaud and M Troper (ed) Traité international de droit constitutionnel, Tome 2: Distribution des pouvoirs (Dalloz, 2012) 494 at 496. 10  ‘Since a legal norm is valid because it is created in a way determined by another legal norm, the latter is the reason of validity of the former. The relation between the norm regulating the creation of another norm and this other norm may be presented as a relationship of super- and sub-ordination, which is a spatial figure of speech. The norm determining he creation of another norm is the superior, the norm created according to this regulation, the inferior norm. The legal order […] is therefore not a system of norms coordinated to each other, standing, so to speak, side by side on the same level, but a hierarchy of different levels of norms’, H Kelsen, General Theory of Law and State (Anders Wedberg tr, Harvard University Press, 1946), 124.

The French People’s Role in Amending the Constitution 317 a norm is valid because it has been produced according to the ­provisions ­contained in another norm that is deemed metaphorically superior to that very norm.11 This superior norm itself is valid because it was similarly produced according to the provisions stated in yet another norm, also deemed metaphorically s­ uperior. This leads to an infinite regress; it now has to stop: this infinite regress ends with the supreme norm, that is, a norm which validity cannot be explained with a reference to another superior norm. This norm is the legal basis of the legal order: it is the basis (directly or indirectly) for the validity of all norms that belong to this legal order, and these norms belong to this legal order—and not another one—because they have this superior norm—and not another one—as legal basis. In that sense, this norm gives the legal order its unity: ‘an “order” is a system of norms whose unity is constituted by the fact that they all have the same reason for their validity’.12 In positive law, this supreme norm is the Constitution. This has two main consequences. On the one hand, its particularity is to be valid only on its sole ground and not on the ground of another (superior) norm. That means the Constitution is also the legal basis of its amending norms. On the other hand, the act of will that leads to the establishment of this first and supreme norm of the legal order, is the one of whom wants to establish this legal order, that is, the sovereign: and in a democracy, this sovereign is the people. This is why the people’s position in a democracy such as France, as confirmed by the above-mentioned first articles of the Constitution, is fundamental. It is fundamental and absolute, at least in appearance, because as sovereign, the people are also the ones expressing the act of will to any constitutional amendment and modification: in other words, what the people want, the people do. But is this also the case from a strict and pure legal perspective? Actually not, since in that case the legal people can only legally do what the Law enables them to do: this position is then limited and relative. I.  THE POSITION OF THE PEOPLE: FUNDAMENTAL AND ABSOLUTE, IN APPEARANCE

According to the French Constitution, its origin and history, its first articles and its amending procedure, the people’s position seems to be very fundamental and absolute. It is the people who desired and founded this Fifth Republic’s Constitution (1958), it is in the people’s name that any constitutional amendment is valid, it is the people who are able to decide anything regarding the Constitution. A.  The People’s Fundamental Desire for the Present Constitution As stated by Article 3 of the Constitution, the people are the sovereign. The current Constitution of the Fifth Republic, 4 October 1958 is valid on the ground of the 11 Kelsen,

Pure Theory (n 3) 193–94. 31. It has also been qualified of ‘original norm’, see Y Laurans, Recherches sur la catégorie juridique de constitution et son adaptation aux mutations du droit contemporain (Université de Nancy II: Thèse, dact., 2009) 106 at 108. 12  ibid,

318  Jean-Philippe Derosier people’s will as expressed through the referendum of 28 September 1958. There are two different hypotheses about this ‘new’ Constitution: it is either perceived as a result of a legal revolution, or it is presented as an amendment of the former one (the Constitution of the Fourth Republic, 27 October 1946). The first perception is due to the change of Constitution: from the Fourth Republic to the Fifth Republic, from the Constitution of 27 October 1946 to the one of 4 October 1958, from a parliamentary derived into an assembly regime to a so-called ‘semi-presidential’ regime—actually better named ‘rationalised parliamentary’ regime. In addition to this change, politicians and authors argued, in 1958 and after, that there was a breach into the constitutional continuity between the two Constitutions, constitutive of this legal revolution. Indeed, according to the principle, or at least to the maxim delegata potesta non potest delegari, it has been argued that the amending procedure was unlawful. The latter was ruled by Article 90 of the former Constitution. The sovereign people (Article 3 of the former Constitution also) had delegated their power to the Parliament, and mainly the National Assembly. On the basis of Article 90, the Parliament has adopted the statute of 3 June 1958,13 transferring the power to amend the Constitution to the Government: this sub-delegation (from the Parliament to the Government) of a delegated (from the people to the Parliament) power (to amend the Constitution) seemed to be unlawful, especially in the context of 1958, very similar to the one of 1940.14 But it was not: first, the constitutional statute of 3 June 1958 had been legally adopted, and second, if its unique article read ‘notwithstanding its Article 90’s provisions, the Constitution will be amended by the on 1 June 1958 invested Government’, the last paragraph required to submit the project to the people via referendum. The constitutional statute of 3 June 1958 was adopted on the basis of Article 90’s provisions. According to the latter, an amendment to the Constitution required: (a) a National Assembly’s resolution deciding on the amendment and giving its object (Article 90, paras 2 and 3); (b) a confirmation of this resolution by the National Assembly or the Council of the Republic (the second House) (Article 90, para 4); (c) the adoption of the amendment by the Parliament (Article 90, para 5); (d) its ratification by referendum, except if the National Assembly adopted it with a 2/3rd majority, or if the two Houses adopted it with a 3/5th majority (Article 90, para 6); and (e) the amendment to be finally promulgated by the President of the Republic (Article 90, para 7). Requirements of points (a) and (b) were met by a resolution passed by the National Assembly on 24 May 1955 and confirmed by the Council of the Republic on 19 July 1955. It stated: ‘Articles 17, 49, 50, 51 and 90 of the Constitution will be amended’.

13  Loi constitutionnelle du 3 Juin 1958, portant dérogation transitoire aux dispositions de l’article 90 de la Constitution, JORF du 4 Juin 1958, 5326. 14  In both cases, there was, during the war (in 1958, France is in a civil war with Algeria), a sub-delegation of the amending power to the Government whose Chief is a former soldier who distinguished himself and ‘saved’ the country during the previous war (Philippe Pétain, 1940 and Charles de Gaulle, 1958).

The French People’s Role in Amending the Constitution 319 Even if three years had swiftly passed between the Council of the Republic’s confirmation and the governmental Bill for a constitutional statute (1 June 1958), the resolution was not obsolete. Hence, the National Assembly could regularly pass the Bill (2 June 1958) and the Council of the Republic could adopt the same text, according to the ordinary legislative procedure (3 June 1958): point (c) was fulfilled as well. There was no need for a popular ratification (referendum, point (d)) since the two Houses had enacted the Bill with a 3/5th majority. The promulgation by President René Coty was thus regular, and the constitutional statute is not unlawful—from a formal point of view. As from the material point of view, this constitutional statute is not unlawful either. It stated that the Constitution shall be amended by the Government, according to the following procedure: (1)  (2)  (3)  (4) 

Bill’s elaboration by the Government, with the opinion of an advisory committee, adoption by the Council of State, ratification by referendum, promulgation as a constitutional amending statute by the President of the Republic.

All these procedural steps were followed between June and October 1958. In particular, the referendum took place on 28 September 1958 and the ‘Yes’ vote was victorious by 82.6 per cent. Thanks to the latter, the principle of the people as the ultimate titular of the amending power was respected: the Government was only the organ preparing the draft, in the name of the people, according to an empowerment grounded in the constitutional statute of 3 June 1958, enacted by the Parliament (representing the people). In other words, there was no sub-delegation of a delegated power. The people were the sovereign, and as such titular of the amending power under the Fourth Republic’s Constitution; it is the people who amended this Constitution establishing the Fifth Republic’s Constitution (which, actually, was only at that time the amended Fourth Republic’s Constitution); the people remained, according to the ‘new’ Constitution, the sovereign, and as such titular of the amending power under the Fifth Republic’s Constitution as well. B.  The People’s Central Role in the Constitution’s Amendment Procedure Under the present Constitution, the fundamental desire of the people has been shifted to a central role as regards the Constitution’s amending procedure, this according to Article 89: the people are the source of any power, thus especially of the Constitution itself (1958) and of any amendment. As always, the current amending procedure has three steps: initiative, elaboration, ratification. The initiative belongs either to the President of the Republic or to the members of Parliament, both elected by the people (although in an indirect way as for the members of the Senate, the Parliament’s second House) (Article 89, para 1). The elaboration is incumbent upon the two Houses of the Parliament, separately, which have to adopt a common text (Article 89, para 2). The ratification is done via referendum (Article 89, para 2) or, in case of the President of the Republic’s initiative if he so decides, by the Congress with a 3/5th majority, that is, the Parliament’s two Houses together.

320  Jean-Philippe Derosier There are therefore four actors (President of the Republic, National Assembly, Senate, people) and a director (the Government, and in particular the Prime Minister and the Minister for Justice).15 To have a final amendment, it is required to attain three of the four possible consensuses, including those of the two Houses that are mandatory. That means the Parliament and the people can amend the Constitution without the President’s agreement (although this never happened), or that the President and the Parliament can amend it without the people’s direct agreement (this was the case in 21 of the 24 amendments of the Fifth Republic’s Constitution).16 But even in this case, the people’s position remains central because the amendment is done in the people’s name. On the one hand, Article 3 of the Constitution states that ‘the national sovereignty shall vest in the people’, adding that the latter ‘shall exercise it through their representatives and by means of referendum’: the people’s representatives, embodied by the President (directly elected by the people), the Deputies (directly elected by the people), and the Senators (indirectly elected by the people), are fully empowered to exercise the sovereignty in the name of the people. On the other hand, Article 89 ensures that there are always at least three different means of representing the people, and two different powers to agree on the amendment. That way, one representative or one power only cannot exercise the sovereignty alone. This requirement of a common agreement guarantees there can be no usurpation of sovereignty. However, this procedure gives to both Houses of the Parliament a veto right: if one of the two does not support the amendment the latter cannot be amended, even if the three other actors (namely the President, the other House and the people) agree on it. This might be quite problematic from a democratic point of view, especially if the unsupportive House is the Senate, that is the actor that is the least directly linked to the people. Article 89 gives no solution in itself to this blockade, which has occurred many times since the beginning of the Fifth Republic.17 But to avoid such a blockade, President de Gaulle decided in 1962 to amend the Constitution on the basis of Article 11. C.  The People’s Absolute will Reflected in the Constitution The basic principle of popular sovereignty seems to be that the people can always amend or change the Constitution (especially the one that they adopted originally). This is the theory of the people’s constituent power, as developed in France since 1789: if the constituent power belongs to the people, which is actually the case in a

15 

See Carcassonne (n 2) 394. Amongst the 24 amendments of the Fifth Republic’s Constitution, 21 were ratified by the ­Congress, one was ratified by the people (according to the Art 89 procedure, but following a presidential i­nitiative), one was adopted on the basis of former Art 85 (concerning the Community, ie the former French ­Colonies), and one was adopted on the basis of Art 11 (see below). 17 Amongst others, we can mention the Bills of July 1984 (on the extension of the referendum’s domain), of March 1990 and of November 1992 (on constitutional review and on the organisation of public authorities). 16 

The French People’s Role in Amending the Constitution 321 democracy, the latter can do whatever they want with the Constitution. According to this theory, the constituent power would be ‘an initial, primary and independent power’18 since it is meant to be the origin of the Constitution and thus at the basis of the legally organised society. In that sense, ‘being the State’s supreme power, the constituent power cannot be bounded, even by itself’.19 In light of the above, President of the Republic General de Gaulle decided to amend the Constitution in 1962 on the basis of its Article 11, allowing him to directly petition the people by referendum and without soliciting the Parliament. General de Gaulle wanted to amend the electoral process for President of the Republic, shifting from an indirect to a direct ballot. But the Parliament was not in agreement with this change as it considered it would lead to the reinforcement of the President’s power, and to the proportional loss of its own. However, de Gaulle, knowing that he would be blocked by the Parliament, decided to bypass the latter by resorting to Article 11 and not to Article 89. He argued, first, that the people, being sovereign (Article 3), had the competence to amend the Constitution, at any time. He added, secondly, that this Article 3 was giving the people the competence to act via referendum. He concluded, third, that even if there was a specific way to amend the Constitution through the Parliament (Article 89), another article of the Constitution (Article 11) gave the President of the Republic the competence to submit a referendum to the people on any Bill concerning the organisation of public authorities.20 The 28 October 1962 referendum was successful with 62.25 per cent voting ‘Yes’. However, the Senate’s President, as so empowered by Article 61 of the Constitution, questioned the Conseil constitutionnel21 about the legality of this constitutional amendment. In its decision, the Conseil decided that it was not competent to examine a norm that had been directly adopted by the people. It considered that Article 61 of the Constitution proclaimed its competence, but without any reference to the nature of the norms it might come to examine, thus without specifying if the latter were all of a legislative nature or, on the contrary, were only those ones enacted by the Parliament. But it concluded that the Constitution had made it a regulatory organ of the public authorities’ activities and that it was therefore not competent when it came to norms directly adopted by the people—since those represent a direct expression of the people’s sovereignty.22 With this decision, the Conseil Constitutionnel confirmed that the people are sovereign and have an absolute power over the Constitution: since the people spoke, none has the capacity to interfere, even on the ground that they have acted the wrong (unlawful) way. But is this position the fundamental and absolute position of the people, as confirmed by the Law itself? No, it is not.

18  G Burdeau, Essai d’une théorie de la révision des lois constitutionnelles en droit positif français (Thèse pour le doctorat en droit 1930) 74. 19  G Vedel, Manuel élémentaire de droit constitutionnel (Sirey, 1949), 117. 20  See Charles de Gaulle’s speech broadcasted on both radio and television, 20 September 1962. 21  The French constitutional Court or, at least, the competent organ for the constitutional review of statutes. 22  Decision n. 62-20 DC, 6 November 1962; confirmed for an ordinary statute (not a constitutional amendment) by decision n. 92-313 DC, 23 September 1992.

322  Jean-Philippe Derosier II.  THE POSITION OF THE PEOPLE: LIMITED AND RELATIVE, IN LAW

Even if, according to the present Constitution and its first articles, the people’s ­position seems to be fundamental and absolute, as confirmed by the Conseil ­constitutionnel and by the 1962 precedent at least, this position is, from a pure legal perspective, limited and relative. This means the people can only act accordingly to the Constitution: their action is limited by law and necessarily relative to the legal prescription that enables them to act. This leads to draw a basic distinction between the social and the legal people and to the conclusion that the 1962 precedent constituted a legal revolution, one that probably could not happen again. A.  Social People and Legal People The legal order is a ‘coercive order of human behaviour’.23 When a norm commands, empowers, permits or derogates,24 its goal is to limit its addressees’ action, setting the limits within which this action may occur. If someone acts outside these limits, he/she might be sanctioned (the sanction may also be provided by the legal system).25 In other words, whatever its specific object is (to command, to empower, to permit or to derogate), a norm is necessarily a limitative phenomenon. On the other hand, the Law is specific since ‘it regulates its own creation and application’,26 meaning that ‘one legal norm determines the way in which another norm is created, and also, to some extent, the contents of that norm’.27 This is precisely how a norm is of a legal nature: it is because it was produced according to the rules provided in another legal norm, and belongs therefore to the legal order, that it is (legally) valid. Thus, the legal limitation does not only concern the human behaviour but the Law as well: the production of norms is also limited. Consequently, if the constituent power is apprehended as the power that is legally organised for the production of constitutional norms, it is not possible to argue that it is absolute, that is, unlimited: indeed, it is necessarily limited by the norms empowering and ruling its action. Similarly, if it is argued that this (legal) power belongs to the sovereign, then it means the latter is legally organised (and limited) and, since its acts are based on its constituent power, that it does so and has to do so within the limits positive law (and the Constitution) provides and imposes.28 This is the reason why the original constituent power cannot be a legal concept, but only a political or a social one. If it has a sovereign competence, understood as an absolute and unlimited power to do anything, this competence is not a legal one either. The original constituent power as such cannot be considered as a legally 23 Kelsen,

Pure Theory (n 3) 33. Those are the essential functions of the norm, according to Kelsen, ibid. 23–24; and Kelsen, Theory of Norms (n 3) 1. 25 Kelsen, Pure Theory (n 3) 33. 26  ibid., 71. 27  Kelsen (n 10) 124. 28 See O Beaud, ‘Maastricht et la théorie constitutionnelle (1ère partie). La nécessaire et inévitable distinction entre le pouvoir constituant et le pouvoir de révision constitutionnelle’ (1993) 39 Les Petites Affiches 14. 24 

The French People’s Role in Amending the Constitution 323 organised and limited power. Its goal is to create the Constitution, which will be founding a new legal order. But before this norm exists (ie, is valid), the legal order so constituted does not exist either. And if this one does not exist, the constituent power cannot have a (legal) existence in that legal order. The original constituent power, as so qualified, is thus nothing else than a single sociological or political fact, not a legal action (legal from the point of view of the legal order this constituent power might constitute). In the case of a democracy, this constituent power belongs to the people, although in that respect the people are not the legal people, but the social or political people. The latter are not bounded by any norms nor can they produce any. On the contrary, the legal people are the ones who have the power to produce norms, especially those that amend the Constitution, according to the procedure enshrined in the Constitution itself. But it is then a constituted power, not a constituent one, that is, a power organised and limited by the Constitution. The confusion is easy in French, the doctrine using more the terms of ‘derived constituent power’ [pouvoir constituant dérivé] and less those of ‘amending power’ [pouvoir de révision].29 This leads to the conclusion that a valid constitutional amendment can only be produced accordingly to the valid amending procedure, as stated in the Constitution itself. In France, this is ruled by Article 89 that gives to the people the power to ratify the amendment, except if the President decides to submit the ratification to the Congress. At the constitutional level (amending power), the people are then a legal people, but only on the basis of Article 89 and according to the provisions stated in this article. If the people act on the basis of another article, they turn into the social or political people and cannot produce a legal norm, that is, a norm legally connected to the present Constitution. In doing so, a legal revolution occurs: this is what happened in 1962. B.  The 1962 Legal Revolution A legal revolution constitutes a breach into the constitutional continuity: it occurs when ‘the constitutional modification leads to a constitutional breach’.30 It corresponds to the entry into force of an unlawful norm.31 This actually looks like an oxymoron and requires an explanation.

29  When using the concepts of ‘original constituent power’ and ‘derived constituent power’, French authors confronted the latter to those of ‘constituted powers’, ie those that are provided and organised by the Constitution; they hence considered the first concepts as not ‘constituted’ because they create the Constitution. The distinction between constituent power and constituted power is traditionally due to Abbey Sieyès, the French Revolutionary, when the distinction between original constituent power and derived constituent power is traditionally due to Roger Bonnard (R Bonnard, Les actes constitutionnels de 1940 [LGDJ 1942] 5). But some authors before the latter, like Georges Burdeau, already introduced a distinction between constituent power and amending power (Burdeau [n 18] 38–39). 30 H Kelsen, Allgemeine Staatslehre (Springer, 1925) 249; see also A J Merkl, ‘Das Problem der Rechtskontinuität und die Forderung des einheitlichen rechtlichen Weltbildes’ (1926) 5 ZöR 513. 31  This definition is similar to the one of other authors, Kelsen foremost amongst them: ‘a revolution […] is every non-legitimate change of this Constitution or its replacement by an other Constitution. […] Decisive is only that the valid Constitution has been changed or replaced in a manner not prescribed by the Constitution valid until then’, Kelsen, Pure Theory (n 3) 209. See also Ph-I André-Vincent,

324  Jean-Philippe Derosier A norm is valid as soon as it meets the requirements (of validity) contained in another norm: the same applies to constitutional amending norms, but with the (fundamental) exception that the Constitution contains its own amending process. The entry into force of a constitutional norm may represent a revolutionary phenomenon when that norm is considered valid although it has not met the required conditions for its validity. In this case, it cannot be valid from the point of view of the current positive legal order since the latter provides for conditions of validity that were not fulfilled by that very norm; yet, it is valid because considered so. That means the validity of the new constitutional norm cannot be based on the legal order that was in force until now. Its validity should hence be either denied, and the norm does not exist and has even never existed (it is not ‘born’ because of a lack of one or more conditions of validity), or accepted ex nihilo, a new validity without any normative basis. Its validity is then based on a new legal order, without any legal link to the former one it has substituted: a legal revolution occurred. Despite Charles de Gaulle’s arguments, Article 11 of the French Constitution cannot be used to amend the Constitution. The main reason is the Lex specialis generalibus derogat principle: when a specific rule exists, it might substitute the general rule. The French Constitution provides different procedures for a (national) referendum: one lies in Article 11, another in Article 89.32 The former deals with ‘any Government Bill which deals with the organization of the public authorities, or with reforms relating to the economic, social or environmental policy of the Nation, and to the public services contributing thereto, or which provides for authorization to ratify a treaty which, although not contrary to the Constitution, would affect the functioning of the institutions’. Let’s call this the general rule: if the President of the Republic wants to pass a Bill via referendum, Article 11 applies. The latter deals with ‘the amendment [of the Constitution that] shall take effect after approval by referendum’. Let’s call that the specific rule: to amend the Constitution, and if the President wants to pass the amendment via referendum, Article 89—and not 11—applies. When, for a determined (restricted) object, a specific procedure is provided, it has to be followed instead of the general procedure that was provided for a general object. Amongst the French existing procedures for (national) referendum, one applies to ‘any Government Bill’ (on determined topics, amongst them, the organisation of public authorities) and another is specifically dedicated to the ‘amendments of the Constitution’. Both allow the people to make a direct decision via referendum but, whereas the first one contains no other requirement than a submission by the President (on recommendation of the Government), the second requires to have an

Les révolutions et le droit (LGDJ, coll. Bibliothèque de philosophie du droit, 1974) 91; G Liet-Veaux, La continuité du droit interne. Essai d’une théorie juridique des révolutions (Librairie du Recueil Sirey, Thèse, Rennes, 1942) 43–109; S Rohatyn, ‘Die juristische Theorie der Revolution’, (1929–1930) IV Revue internationale de la théorie du droit. Internationale Zeitschrift für die Theorie des Rechtes 193–227, sp. 210–27; E Cartier, La transition constitutionnelle en France (1940–1945). La reconstruction révolutionnaire d’un ordre juridique ‘républicain’ (LGDJ, coll. Bibliothèque constitutionnelle et de ­ ­science politique, 2005) 529 seq. 32 

There is a third one at Art 88-5, for the accession of a new State to the European Union’s membership.

The French People’s Role in Amending the Constitution 325 identical text adopted by the two Houses of the Parliament. The ‘organisation of public authorities’, as mentioned in Article 11, is necessarily an organisation of subconstitutional level: for the constitutional level, Article 89 must be used. Consequently, the constitutional amendment based on Article 11 of the Constitution cannot generate a validly produced amending norm of the Constitution that is currently in force. Its validity has to be either denied, or admitted but with the validity of an entirely new Constitution, ex nihilo, and hence as a legal revolution. This is precisely what happened in 1962. The Constitution that had been in force since 4 October 1958 (of the Fifth Republic), which in fact was the Constitution in force since 27 October 1946 (of the Fourth Republic) and amended in 1958, was actually not amended in 1962: it ‘disappeared’, that is, it lost its validity, due to a legal revolution. Because of that revolution, a ‘new’ Constitution, quite identical to the former one although differing in Article 6, can then be valid. The use of Article 11 generated a breach in constitutional continuity.33 The people who answered ‘Yes’ in 1962 are not the same people who had given a similar answer in 1958. In 1958, it was the legal people who answered: their action, will and response were based on a legal norm (namely, the constitutional statute of 3 June 1958). In 1962, it was the social, or political, people who did: their action was not based on a legal norm since ­Article 11 was not applicable. In its 62-20 DC decision, the Conseil constitutionnel did not uphold the possibility to use Article 11 for a constitutional amendment. On the contrary, it clearly decided it was not competent to examine whether the people were acting accurately or not, whether they were acting lawfully or not, and whether the Constitution enabled this referendum or not. And that it could not do of course since it was petitioned after the legal revolution had occurred. For sure, it could have considered that the amendment was not valid and hence avoid the constitutional breach and legal revolution. But the sovereign had given its opinion, and even if it was not a legal opinion, it was at least one that stated: ‘Yes, I want this legal revolution’. The Conseil constitutionnel had no legitimacy to say the contrary. In this respect, the will of the social or political people is absolute indeed, but not that of the legal people. Today, such a use of Article 11 could no longer be possible. C.  The Amending Procedure of Article 11 Questioned After 1962, Article 11 was never used for constitutional matters again, except in 1969, by General de Gaulle once again. But this time, the ‘No’ was victorious, ­President de Gaulle dismissed, and Article 11 was from then on only used for ordinary Bills. Today, its use remains possible since a legal revolution is something that operates outside of the legal limits. Nevertheless, we could apprehend the case law of the Conseil constitutionnel as an indication it would not allow it anymore. Does this mean the Conseil would repel an amendment adopted under Article 11 referendum? Probably not: since the people would give their sovereign opinion, the Conseil would

33  See O Pfersmann, in L Favoreu et al., Droit constitutionnel (Dalloz, coll. ‘Précis’, 14th edn, 2012) 103–06. On those constitutional transition’s questions, see also Cartier (n 31) 11–17.

326  Jean-Philippe Derosier not allow itself to disagree.34 But things would go a different way if the Conseil examined the use of Article 11 before the people answer. In 2000, for a referendum based on Article 89 of the Constitution, the Conseil constitutionnel accepted for the first time to examine the referendum’s preparatory acts, before the referendum itself happened.35 The Conseil is competent to ‘ensure the proper conduct of referendum proceedings’, according to Article 60 of the Constitution. But it normally has to examine the applications it receives after having proclaimed the referendum’s results. The Conseil justified this decision on the basis of its general role to monitor referendum proceedings, and because of the risk to compromise that control if it had waited for the results. Differently put, if the applications suggest the possible cancellation of the referendum proceedings, the Conseil constitutionnel has to examine them straight away as it will not be competent to do so afterwards. The Conseil confirmed this decision in 2005, on the occasion of a referendum based on Article 11.36 The political context of the 2000 decision was very important: the President had dissolved the National Assembly in 1997 and lost the following elections, there was then a so-called ‘cohabitation’ between Jacques Chirac, President of the Republic (Conservative, Right wing) and Lionel Jospin, Prime Minister (Socialist, Left wing). There were hence tensions at the top of the executive branch. One could consider that the Conseil constitutionnel indicated then that if Chirac wanted to use Article 11 to amend the Constitution against Parliament’s will, the Conseil would be competent to examine the issue—and would probably not allow it to happen. If Article 11 is not the right constitutional basis to amend the Constitution via referendum, the decree that calls for a vote does not have any legal basis. Such a decree is part of the referendum’s preparatory acts, thus something that the Conseil constitutionnel may now examine, and, in that case, probably cancel. This might prevent another legal revolution from happening, and ensure the people have a legal role, not a social nor a political one. But at the same time, it strengthens the veto-right of the Senate in the case of an amendment this House does not support—even though the amendment is supported by all other three actors (the President, the National Assembly, and the people) … The social or political people can also make the Law evolve, outside of the legal limits! But sometimes only.

34 

As already mentioned, it was the same decision even for ordinary statutes, see decision 92-313 DC. Decision 2000-21 REF, 25 July 2000. 36  2005-31 REF, 24 March 2005. 35 

17 The Implication of Conflation of Normal and ‘Constitutional Politics’ on Constitutional Change in Africa DUNCAN OKUBASU I. INTRODUCTION One of the major causes of political and constitutional instability during Africa’s first three decades of independence was caused by the ease with which post-independence leaders subverted constitutionalism by regularly amending constitutions to suit selfish political agendas.1

T

HE FOREGOING STATEMENT by Fombad captures a phenomenon that continues to frustrate democratic consolidation in Africa even after the post 1990 political liberalisation gains associated with the so called ‘third wave of democratisation’. The perception that the problem of consolidation of constitutional rule in Africa was associated with ease of amendments of post-independence constitutions influenced post 1990 constitutional reforms that imposed and entrenched procedural and substantive limits on the power to change constitutions.2 Such limits were intended to ensure that constitutions do not change easily as normal legislations do and indeed, one of the distinctions between a constitution as a higher law and ordinary legislation is that it should be more cumbersome to alter than normal legislation.3 Unfortunately, the response to that concern—entrenchment of formal rules of change—does not appeal to have yielded the desired result because the arrangements of the state in relation to power creation, use and distribution are, in reality, still altered in most post 1990 African states: for the same reason they were changed before the 1990s.4 The underlying concern before 1990 was how to shield

1  CM Fombad, ‘Some Perspectives on Durability and Change Under Modern African Constitutions’ (2013) 11(2) Int’l J Const L 382–413. 2 ibid. 3  See M Tushnet, ‘Constitution’ in M Rosenfeld and A Sajó (eds), The Oxford Handbook of Comparative Constitutional Law (Oxford University Press, 2012). 4  See HK Prempeh, ‘Africa’s “Constitutionalism Revival”: False Start or New Dawn?’ (2007) 5 Int’l J Const L 469.

328  Duncan Okubasu constitutions from ‘abusive constitutionalism’5 yet even after entrenchment of rules of change, use of ­constitutional change as an instrument to consolidate political power and subvert constitutionalism and democracy still persists.6 This contribution attempts to explain the persistence of this spectacle. In reflecting on what appears to be the underlying concern for constitutional change processes in Africa, this chapter deduces that constitutional changes occur out of the scope of the amendment clauses to the extent that: (a) the effective constitution (including the formal rules of change) is often radically at variance with the formal ­constitution, and (b) changes that occur in the name of formal amendment clause largely defeat the very significance of the amendment clause. If Ackerman’s theory of constitutional changes out of the contemplation of the amendment clause is used to judge constitutional additions and subtractions in Africa, an inference that there is an unhealthy fusion of normal and constitutional politics easily crystallises and the result on constitutionalism is unkind. Part of Ackerman’s theory is that there is a distinction between the decision by the government and the decision by the people and that these two must not be confused.7 Whereas, as shall be demonstrated, there is a functional distinction between what these genres of politics are intended to achieve, it is difficult to discern the distinctiveness of their outcome by examining constitutional change processes in most African countries. Constitutional developments in Africa, which strictly defeat the purport of rules of change, can be said to be fraught with inappropriate use of ‘normal politics’ to secure constitutional changes and a concomitant use of ‘constitutional politics’ to achieve what should be achieved by normal politics. This scenario had, until recently, incubated an unhealthy relationship between constitutions and politics in Africa and indeed impacts worryingly on the stability of constitutions. This chapter illustrates this coalescence and its effect by highlighting the interaction between political elites and citizens well as the interaction between political elites and constitutional change processes by making reference to happenings in various African countries. Political elites denote persons in power or those seeking power. It first revisits Ackerman’s decision making models as the framework of assessing and illustrating constitutional change that does not occur within the contemplation of the amendment clauses. Since the rules of change are intended to open up a constitution for deliberation on how it should be changed, changes that occur though the formal process yet deflate the essence of the procedural and substantive limits—permitting abusive constitutionalism—are deemed in this contribution to be occurring outside the amendment clause. Secondly, it demonstrates how political elites see in constitutional change processes an opportunity for institutional capture well as for consolidation and retention of state power rather than for creating a level playing ground for political competition hence making a constitution a ­predator on

5  The term has been used by Landau to mean ‘the use of mechanisms of constitutional change to erode the democratic order’. See D Landau, ‘Abusive Constitutionalism’ (2013) UC Davis L Rev 189. 6  See HK Prempeh, ‘Presidential Power in Comparative Perspective: The Puzzling Persistence of Imperial Presidency in Post-Authoritarian Africa’ (2007) 35(4) Hastings Const L Q 761. 7  B Ackerman, We the People: Foundations (Harvard University Press, 1991) 3–33.

The Implication of Conflation of Normal and ‘Constitutional Politics’ 329 rather than a promoter of democracy. It achieves this objective by locating the place of partisan politics in momentous constitutional developments. Thirdly, it illustrates the relationship between political regimes and constitutional change by linking ­constitutional change processes to political survival. It then contends that the consequence of this amalgamation affects constitutional stability not only because of the real and perceived illegitimate outcome of constitutional development processes but also because it augments unattractiveness of constitutions in conditioning normal politics. In the end, it suggests a solution. II.  DECISION MAKING IN A POLITY

Bruce Ackerman’s postulation of decision making in a polity presents two prisms with which to assess decision making.8 There are decisions made by the people (American people) and those made by (their) government.9 He presents the decision by the people as being uncommon occurring only under certain preconditions.10 In a functional democracy—ideally—it is supposed to precondition decision making by government. The basic inference from this postulation is that if a constitution is a higher law that binds the government which in turn can exercise legitimate authority over the people that are presumed to be the fountain of all powers, then its alteration must be by the people and not by government and it must be orchestrated by and be a product of high-minded substantive deliberations.11 The circumstances under which decisions by the people are made are what I delineate as constitutional politics and I adopt the following reflection of Ackerman— though it concerns America as a functioning democracy—as explaining a setting that constitutional politics should incubate:12 a political movement must, first, convince an extraordinary number of its fellow citizens to take its proposed initiative with a seriousness that they do not normally accord to politics; second, allow opponents a fair opportunity to organize their own forces; third, convince a majority of Americans to support transformative initiatives as their merits.13

If thus a decision is supposed to be made by the people, then it should be presented in a manner as to elicit ‘seriousness’ from an extra-ordinary number of citizens— implying their involvement. Also, it must not sideline those who are likely to oppose it, implying a chance to the opponents to rally against it on the basis of merits and—I would add—to result into a greater satisfaction not only for the good of the ‘people’ making the decision but also to the minorities and the posterity.

8  See Ackerman (n 7); B Ackerman ‘Constitutional Politics/Constitutional Law’ (1989) Faculty Scholarship Series. Paper 140, p 461. 9 ibid. 10 ibid. 11  See R Bellamy and J Schönlau, ‘The Normality of Constitutional Politics: An Analysis of the Drafting of the EU Charter of Fundamental Rights’ (2010) 11(3) Constellations: An International Journal of Critical and Democratic Theory 412 (in the context of EU). 12  Ackerman, ‘Constitutional Politics’ (n 8). 13  ibid 461.

330  Duncan Okubasu Normal politics on the other hand are about decision making by government. It, as can be deduced from Bellamy and Schönlau, should focus on policy considerations and promotion and balancing of sectorial interests.14 It should be the one that ensures phenomena such as unemployment, poverty and similar grievances are addressed to the taste of the popular majority who in turn can select those who make those decisions. Decisions by government are important but should not necessarily be as high minded affairs as decisions by the people. Ackerman also highlights the circumstances of making of this strand of decisions in a functional democracy thus: Decisions made by the government occur daily, also under special constitutional conditions. Most important, key decision makers must be held accountable at the ballot box for their performance; moreover, a structural effort is made to encourage them to deliberate seriously about the public interest and to constrain efforts by narrow but well-organized interests to use government to oppress especially vulnerable or poorly organized groups.15

Constitutional politics, thus, relate to decisions by the people, while normal politics relate to decisions by the government—the latter includes the decision to select the persons making decisions as government. The healthy relationship between constitutional and normal politics is that normal politics should be preconditioned by constitutional politics and not the other way round. The process of securing a consensus about how and by whom the decisions by government are to be made including the process of selecting individuals to make the decisions (elections) should be a product of ‘seriousness’ and ‘high deliberations’ that involve every fraction of society and should therefore be preconditioned by constitutional politics. Hence what appeals as normal politics should not usurp the functionality of constitutional politics just the same way, constitutional politics should not concern itself with day-to-day decisions in a polity: the government should not hold a national conference and a referendum each time it wants repair a bridge! Constitutional politics, according to Colomer should result in widespread satisfaction among political actors and a rational consensus on common values and procedures.16 It should be accommodative of wide interests and concern.17 Conceivably, the individual’s claim over society and the state is established and secured by constitutional politics while through normal politics a platform is ­provided for securing the claim of the majority society over the individual and minorities— and these are attributes of a good constitution.18 If stability is to be achieved in a polity, changes to the Constitution must be midwifed through ­constitutional politics and the essence of entrenchment of formal amendment rules is to provide an insurance against the infiltration of normal politics in processes resulting to constitutional change. The reasoning model that can be borrowed from Ackerman vouches for substantive rather than none or merely formal involvement of the people in

14 

Bellamy and Schönlau (n 11). Ackerman (n 8). 16  RJ Lipkin, Constitutional Revolutions: Pragmatism and the Role of Judicial Review in American Constitutionalism (Duke University Press, 2000) 45. 17  ibid 45. 18  See, Sir J Laws ‘The Good Constitution’ (2012) 71(3) Camb LJ 567–82. 15 

The Implication of Conflation of Normal and ‘Constitutional Politics’ 331 constitutional change processes.19 This has manifested differently in some African countries as the next session demonstrates. III.  EXPRESSION OF NORMAL POLITICS IN CONSTITUTIONAL POLITICS

The relationships between political and constitutional regimes in most African states reveal that there is an untoward conflation of these genres of politics. Untoward in the sense that there is a failure of normal politics and an attendant use of c­ onstitutional politics in normal politics. The failure of normal politics is occasioned by frustration of electoral democracy through a regression into electoral authoritarianism, meaning that the decision to choose decision making agents in the government is hijacked and the ballot cannot be used to hold accountable decision makers for inept decisions relating to normal politics.20 Constitutional politics are also appropriated by the ruling elites such that a constitution is important as formally legitimising a regime and also for thwarting threats towards the regime. The countenance of this malady is that: (a) constitutional change processes are initiated and controlled by the ruling administration, (b) such changes help entrench political powers of the ruling parties, (c) they are made in the name of the people though without their meaningful involvement, and (d) are seen as also providing political grievance for opposition groups that want to ascend into political power—­ without having to resort to unconstitutional measures. In sum, a constitution is presented by political elites, either in government or opposition, as the cause of the problem that should ordinarily be addressed by normal politics. When this happens, the process of changing a constitution is trivialised just as that of electing political leaders and both of them are similar because in reality they are made by political elites. The foregoing pre-condition results in a scenario where first, the connection between political regimes and constitutional lifespan and durability is apparent in most African countries. Save for cases such as Ghana, constitutions in Africa are seen to endure to the extent that they serve only one political regime or players. South Africa’s 1996 Constitution has not been substituted, debatably, because it has served only the African National Congress (ANC) party since its promulgation. It is not, however, surprising that the same ANC has on myriad occasions called for its replacement noting at one point that it ‘may have been appropriate for a political transition, but it has proven inadequate and even inappropriate for a social and economic transformation phase’.21 This is not because the 1996 Constitution has not

19  See, eg, C Klein and A Sajó, ‘Constitution-Making: Process and Substance’ in Rosenfeld and Sajó (eds) (n 3) 419. 20  A Schedler, ‘The Logic of Electoral Authoritarianism’ in A Schedler (eds), Electoral ­Authoritarianism: The Dynamics of Unfree Competition (Lynne Rienner Publishers, 2006) 1–4. 21  See City Press, ‘ANC Wants A new Constitution’ (3 March 2012) accessed 1 April 2014. See also L Mazimbuko, ‘Why is Zuma So Eager to Change Our Constitution?’ (Politics Web, 9 January 2014) accessed 1 April 2015.

332  Duncan Okubasu constricted the rules of constitutional change, structured an effective ­government or failed to protect and promote rights; but it is perhaps because its attractive ­formal values can be used to defeat ANC dominance. Uganda has had one constitution since 1995. Yoweri Museveni has also been Uganda’s president since then.22 The ­Constitution of Rwanda, adopted in 2003 has served only Mr Paul Kagame’s regime.23 Chama Cha Mapinduzi in Tanzania has operated under one constitution since Tanzania’s independence (though there are proposals for a new constitution)24 and even in Kenya and Senegal where other political parties have come into leadership, there is only change of political party names and not actual change of players. Kibaki and Uhuru, Kenya’s third and fourth presidents, who have served under the 2010 Constitution, were members and officials in the dominant post-independence Kenya’s political party, the Kenya African National Union (KANU).25 In fact Uhuru Kenyatta was first nominated to be a presidential candidate by long-serving ­President Moi of KANU.26 President MackySall of Senegal has an identical political history; he defected to form the Alliance for the Republic having been a faithful of the Senegalise Democratic Party.27 Secondly, decisions by elites in government manifest in constitutional alteration processes more so in constitution making. As hinted, the processes are trivialised because the outcome aims at seeking ordinary partisan assurances. This trivialisation is achieved through essentialising constitutional change in political party’s manifestos, which arguably distracts the attention of the populace. Though implicit, the manifesto of South Africa’s Economic Freedom Fighter (EFF) led by Julius Malema that promises a struggle for ‘democratic ownership and control of key factors of production by our people … as a Marxist Leninist fanonian organization’28 sees the solution to the problems bedeviling South Africa in constitutional change.29 Most political parties in Kenya in 2007, including those in opposition, such as the Orange Democratic Movement (ODM) had, as one of their manifestos, the delivery of a new constitution.30 The political manifesto for Zimbabwe’s opposition party, Movement for Democratic Change (MDC) does not have a constitutional substitution promise but an implementation agenda, understandably because of the explanation in the manifesto that change of the Lancaster constitution was one of their founding

22  IDEA, ‘Uganda: Museveni to run for another term after 30 years in power’ (Constitutional Net, 28 July 2015) accessed 31 July 2015. 23  J Kiplagat ‘Raila asks presidents Kagame, Museveni not to cling to power’ (Daily Nation, 3 September 2015) accessed 31 July 2015. 24 M O’Gorman ‘Why the CCM Won’t Lose: The Roots of Single Party Dominance in Tanzania’ (2009) 30(2) Journal of Contemporary African Studies 313–33. 25 See M Kibaki—Contemporary Black Bibliography, available at: accessed 31 July 2015. 26  See Ch Honsby, Kenya: A History since Independence (IB Tauris & Co Ltd, 2012) 688. 27  A Osei, Party-Voter Linkage in Africa (VS Verlag für Sozialwissenschaften, 2012). 28  See EFF Elections Manifesto, accessed 1 April 2014. 29 ibid. 30  Honsby (n 26) 755.

The Implication of Conflation of Normal and ‘Constitutional Politics’ 333 objectives31 together with the fact that its President Morgan Tsvangirai was the ­Chairperson of the National Constitutional Assembly at its initiation, also contained in its manifesto.32 Constitutional change intentions are also exhibited in Uganda’s opposition party Forum for Democratic Change (FDC) manifesto 2010–2016. UDF sought to be elected so that it can ‘honour the views of Ugandans as expressed in the Odoki and Ssempebwa Commission Reports, hold dialogue, agree and implement a viable federal system of governance across Uganda where power and resources shall be shared equitably between the regions and the central government’.33 ­Constitutional change to entrench land ownership is explicitly listed as part of U ­ ganda’s opposition party intentions.34 Nigeria’s opposition party led by Muhammadu Buhari that won the 2015 Nigeria’s general election had indicated that it planned to ‘… initiate action to amend the Nigerian Constitution with a view to devolving powers, duties, and responsibilities to states in order to entrench true Federalism and the Federal spirit …’ and to ‘amend the Constitution to remove immunity from prosecution for elected officers in criminal cases … amend the Constitution to require Local governments to publish their meeting minutes, service performance data, and items of spending over N10M …’.35 As a consequence, constitution-making processes in most African countries are commandeered as typical political strategies. Dominant political parties appropriate the process and use it to entrench party or individual ideologies. Ndulo notes in this regard that the use of constitutional commissions has been ineffective because the ruling parties manipulate the process which results in the imposition of deserved constitutional models by the ruling elites.36 Uganda’s constitution-making process was described as ‘undemocratic’ because of the role played by the ruling party. Tripp laments regarding it that ‘the undemocratic characteristic of 1995 Constitution can be traced in the undemocratic aspects of constitution making’.37 In this respect, political elites see in constitution-making processes, first an opportunity to pre-empt constitution making demands particularly by the opposition groups, and also a chance to entrench their status in the polity.38 It is unsurprising against such a setting that literature on constitution making in Africa decries exclusion and interference by dominant parties.39 A 2003 attempt by Kenya to have a new

31  See MDC Political Manifesto, at accessed 1 April 2014. 32 ibid. 33 See FDCPC Manifesto, 27 available at accessed 1 April 2015. 34 ibid. 35 See M Buhari, ‘My Manifesto and Vision for Nigeria’ (Abusidiqu, 13 December 2014) accessed 1 April 2015. 36  M Ndulo ‘Constitution Making Process in Africa: Assessing both the Process and Content’ (2001) Cornell Law Faculty Publications Paper 57, 15. 37  A Tripp, ‘The Politics of Constitution Making in Uganda’ accessed 1 April 2015. 38  Concerning interests and passion, see generally, J Elster ‘Constitution-Making and Violence’ (2012) 4(1) Journal of Legal Analysis 7. 39  See, eg, JF Wapakhabulo, ‘Uganda’s Experience in Constitution Making’ (2001) available at accessed 12 August 2014; Ndulo (n 36).

334  Duncan Okubasu ­ onstitution was thwarted after the then president, Mwai Kibaki, was uncomfortC able with some power sharing arrangements in a draft constitution produced in 2003 known as the ‘Bomas Draft’. Zimbabwe’s ZANU PF is also accused of hijacking the ­constitution-making process in Zimbabwe. Dzinesa writes about Zimbabwe’s constitution-­making process that: while the government invited members from a cross-section of society in an effort to give the commission a semblance of being fully representative, the majority of the commission’s 400 members were ZANU-PF members or supporters—for example, all 150 members of parliament were included (of which only three belonged to other political parties). Although some reputable academics and civil.40

Likewise, Tanzania’s Chama Cha Mapinduzi (CCM) has been accused of taking over the constitution-making process in Tanzania.41 This means that the ruling parties in Africa see in the constitution-making processes, opportunities to entrench their preferred models that can help them retain their political power. Thirdly, democracy supporting rights are made a subject of easy abridgment. Though most, if not all constitutions in Africa, contain a catalogue of rights intended to facilitate electoral democracy such as expression, assembly, thought and conscience and liberty, they contain limitation clauses, vaguely framed, that allow for the criminalisation of exercise of those rights on the basis of vague conceptions such as national ‘security’, ‘interest’ or ‘health’. What therefore should be preserved to strengthen democracy as a core value of the state is abridged using normal politics through ­normal legislation such as those that criminalise ‘insult to president’.42 The significance of those rights—the effective content of rights—is changed through legislation. Because of the vague formulations, the effective constitution is changed through ­qualification of democracy supporting rights via normal politics.43 This would be directed at suppressing opposition. Fombad captures the resultant scene thus: Opposition political parties, long considered to be an essential structural feature of modern liberal democracy are barely tolerated on the continent. The large numbers of opposition parties in Africa have often degenerated into narrow ethnic opportunistic alliances. Fractious and diverse, many of the opposition parties, even in countries like Botswana, Africa’s ‘best example of successful multiparty democracy’ spend their time squabbling and in most cases pose more competition for each other than for the ruling parties.44

40  GA Dzinesa, Zimbabwe’s Constitution Reform Process: Challenges and Prospects (2012), available at http://dspace.africaportal.org/jspui/bitstream/123456789/33658/1/IJR%20Zimbabwe%20Constitutional%20Reform%20OP%20WEB.pdf?1 accessed 1 April 2014. 41 See, eg, P Nyanje ‘Party interest threatens Tanzania’s Constitution Making’ (The East African, 8 March 2014). 42 See T Rhodes, ‘A bid to rid Africa of criminal defamation, sedition laws’ (2010), accessed 31 July 2015. 43 International Service for Human Rights, ‘Statement Under Item 7, In Response to The Activity Report of the Special Rapporteur on Freedom of Expression And Access To Information In Africa’ accessed 31 July 2015. 44  Ch Fombad, ‘Constitutional Reforms and Constitutions in Africa: Reflections and Some Current Challenges and Future Prospects’ (2011) 5. Available at accessed 31 July 2015.

The Implication of Conflation of Normal and ‘Constitutional Politics’ 335 The other area of fusion of normal politics in constitutional affairs has been on presidential ‘term limits’. Griner contextualises the essence of terms limits in democratic consolidation noting that ‘with term limits, transitions take place as a natural course of events’ and hence ‘politics ceases to be viewed as a zero-sum game. Ruling parties are able to cultivate new leadership which can carry on the successful policies of their former leaders, but also correct for past missteps’.45 Despite the third wave of democratisation, amendment of constitutions to eliminate presidential term limits has been a keen concern of ruling elites in a number of countries. The latest effort in this respect was Burundi’s amendment of the Constitution to allow President Nkurunziza to run for another term after expiry of the prescribed two terms.46 Daniel Vancovsky writes on constitutional amendments to extend term limits as follows: Seven presidents, most of them long-serving leaders of their countries, secured constitutional amendments that allowed them to stand for a third term in office, and all seven won subsequent re-elections. These were presidents Blaise Compaore of Burkina Faso, Idriss Deby of Chad, Omar Bongo of Gabon, Lansana Conte of Guinea, Sam Nujoma of Namibia, Gnassingbe eyadema of Togo and Yoweri Museveni of Uganda ….47

In most of these endeavours, the common denominator is substantive exclusion of the people as there are no high minded merit based deliberations on the proposal to amend a constitution. In the context of Kenya, it is only the 2010 Constitution that appears to be a product of valuable participation of ‘we the people’. The 1963, 1964 or 1969 constitutional change processes that were ruinous to democratic rights and processes were an exclusive enterprise of the then dominant political party, KANU.48 The 1963 Constitution was negotiated in Lancaster with KANU and KADU and none of the amendments to it were made through popular participation—­referendum— including the significant one that transmuted Kenya into a republic and abolished the office of the prime minister replacing it with that of the president.49 South Africa has had four considerably different constitutions since 1960.50 These were in the 1961, 1983, the interim one of 1993 and the Final Constitution of 1996. The politics resulting in the 1961 and 1983 Constitutions were certainly not an affair of ‘we the people’ but exclusionary towards a significant part of the polity, the black population.51 The aim of the pre-1993 constitutional developments was to entrench and

45  S Griner, ‘Term Limits can Check Corruption and Promote Political Accountability’ (2008), available at accessed 25 June 2016. 46  B Kabumba, ‘Did Constitutional Court misinterpret provisions of law on term limits?’ (The East African, 16 May 2015) accessed 31 July 2015. 47  D Vancovski, ‘Presidential Term Limits in Africa’ (Africa Files, 20 January 2008) accessed 31 July 2015. 48 See HWO Okoth-Ogendo, ‘The Politics of Constitutional Change in Kenya since Independence’ (1968) 71 The African Affairs 17; G Kamau Kuria, ‘Confronting Dictatorship in Kenya’ (1991) 2(4) Journal of Democracy 119. 49  G Maingi, ‘The Kenyan Constitutional Reform Process: A Case Study on the Work of FIDA Kenya in Securing Women’s Rights’ accessed 31 July 2015. 50  See IDEA, ConstitutionNet, ‘Constitutional History of South Africa’ accessed 31 July 2015. 51 ibid.

336  Duncan Okubasu ensure racial segregation and had to purposefully disregard the majority black population that would obviously have thwarted the amendments had they been involved. It is the involvement of the entire South African populace in the post 1993 constitutional developments that saw the creation of what is acclaimed to be a progressive constitution. The next segment attempts an assessment of how this relationship has implicated on stability. IV.  REVISITING EFFECTIVENESS AND STABILITY

If significant constitutional processes are meant to be harvests of constitutional and not normal politics, how are constitutions to be effective and relevant in Africa? Some theorists think that they cannot. Daniel Elazar delimited some constitutional models as ‘tempered’ political ideals linking them to the third world in an effort at construing them as being vulnerable to change.52 These constitutions ‘are an expression of what citizens believe it should be’ and ‘the basic authority which enables power holders to rule with a measure of legitimacy’.53 To him, there is a mixture of regime realities and constitutional aspirations without any serious expectation that the polity or regime will achieve the constitutional ideals.54 Though Elazar does not make explicit reference to African constitutions his characterisation aptly describes some of them and their odyssey. They are more of political ideals than framers of governments, protectors of liberty and regulators of constitutions of change. In the same vein, Ogendo examines African constitutions in a scenario he describes as a paradox and concludes that there is a ‘commitment to the idea of constitutions, and a rejection of the classical notion of constitutionalism’.55 Though his musings focus on the pre-1990 epoch, they are of relevance even today. Political elites are seen to accept that there should be a constitution, and rules of change in the constitution, but decline to use it as the validating source of their actions. In many African countries there is an acknowledgement that the state cannot exist without some basic document establishing it equally with a dismissal of what it is meant to do. Political elites justify whatever they do or fail to do, particularly when it is of dubious legality using the formal constitution. For instance, because constitutions make use of the word ‘people’ in their preambles, in effect political elites in Africa distort the concept and make reference to it to justify their refusal to cede power at the expiry of their term.56 They then amend the Constitution and present themselves for

52 D Elazar ‘Constitution Making: The Pre-eminent Political Act’ accessed 2 April 2014. 53 ibid. 54 ibid. 55  HWO Okoth-Ogendo ‘Constitutions without Constitutionalism: Reflections on an African Political Paradox’ in IG Shivji, State and Constitutionalism: African Debate on Democracy (Southern African Political Economy Series, 1991) 3. 56  Rwanda’s Paul Kagame has been cited as saying that he did not want a third term but it was the people to decide. See The New Times, ‘Rwanda—Kagame says he doesn’t want third term but will let people decide’ available at accessed 1 April 2015.

The Implication of Conflation of Normal and ‘Constitutional Politics’ 337 elections arguing that ‘the people’ have decided. In reality, it is not the people that have decided, since the concept of the people as is used in constitutional law, does not mean popular majority. In fact, one of the constraining aspects of constitutions is to prevail upon the popular majority. Some African leaders reign for as long as 20 years, not because the constitution allows them to but because of skillful manipulation and use of constitutional concepts like ‘sovereignty of the people’.57 This scenario arguably precipitates constitutional instability because, first, a­ constitution becomes unattractive to the popular majority well as the anti-­ constitutional forces especially the political opposition—sometimes from its very inception. It is unattractive because if a constitution is a merchandise of a process that is not inclusive of interests, instability is created at one level as the excluded or un-accommodated entities have no reward to support it. Because also of some of the inappropriate means of power retention through constitutional amendments, social and political actors fail to support a constitution and prefer extra constitutional mechanisms to attain power and in turn alter a constitution to accommodate their interests and segregate others. Instability at the second level comes as a result of absence of genuine constitutional enforcement by institutions. A constitution that is designed through normal politics and which consequently executes a system of values or model that is preferred by one social or political group is most likely unable to intermediate between social forces and political institutions.58 Conceivably, institutions do not genuinely believe in enforcing a constitution except to the extent that its provisions provide a textual framework for subverting opposition or legitimising autocratic rule. If social forces permeate and pervade political decision making institutions, they are unlikely to pursue pluralistic ends and often devote state power to sectarian interests. Such kinds of institutions are selective in protecting rights and liberties and in asserting or enforcing constitutional authority. The constitutional order thus becomes unattractive to the majority who can easily be mobilised against it. The net result of government decisions in constitutional outcomes is that they reflect only the wishes of the political elites and not demands of the whole population. Such set-ups are in turn prone to demands from opposition for their abolition, demands that are at times attended to by constitutional amendments or revision. This results in a society being in a continuous state of constitutional modifications—the ruling class enacting a new constitution to pre-empt constitutional substitution demands and amending it to entrench political power-using majorities. In the end this conundrum results in the irrelevance of constitutions both to the ruling elites and to the popular majority and the next segment briefly examines that phenomena.

57  Some of these leaders include Robert Mugabe of Zimbabwe, Paul Binya of Cameroon, Do Santos of Angola, Museveni of Uganda, just but to mention a few. See BM Dulani, ‘Personal Rule and Presidential Term in Africa’ PhD Thesis, Submitted at Michigan State University, accessed 1 April 2015. 58  On the role of institutions (hence constitutions to intermediate social forces and political institutions, see S Huntington, Political Order and Social Change (Yale University Press, 1968) 8.

338  Duncan Okubasu V.  CONSTITUTIONAL IRRELEVANCE

Other than formally constituting a people into a polity,59 a functional constitution should be understood to play the important role of establishment or recognition and of control.60 Scholars describe constitutional function as constitutive, attributive and regulative.61 Classically, there are three mostly classified functions of a constitution,62 the first being establishment of government, the second being the protection of rights and the third being the prescription on rules of change. A functional constitution must have an effect in society in the sense that it forms a basis for consideration of legality or otherwise of state actions and should be the actual basis for power creation, use and distribution, including popular sentiments on governance. A constitution is thus relevant if it actually frames or officially recognises an effective government. If power is created and distributed by extra-­constitutional sources it becomes irrelevant to the extent that those extra-­ constitutional authorities will be responsible for power use, control and distribution. If a constitution is to frame an effective government, it should provide a framework for shielding institutions from capture and furtherance of sectarian interest.63 In such cases, the mechanisms of checks and balances are weakened as other arms of government are subservient to the capturing arm(s).64 A power imbalance would also most likely result in deflating the essence of stabilising institutions such as judicial review. When normal politics formulate institutional models that are preferred by a social or political fraction, it is more likely that those institutions be obsolete to a new political regime. Conceivably, this is one possible explanation

59  On constitutions see E Young, ‘The Constitution Outside the Constitution’ (2007) 117(408) Yale L J 408. 60  Generally see, Y Habesse and C Penalli, ‘Constitutions’ in M Tushnet et al, Routledge Handbook of Constitutional Law (Routledge, 2013); Other scholars describe this functionality as attributive, constitutive and regulative. Leonard Baselink notes thus: ‘Constitutions as we know them in Europe seem to have three functions; they are constitutive in establishing or officially recognising institutions of society, attributive in empowering these institutions and regulative that is to say, regulating and recognising institutions of society’. See. L FM Besselink, ‘The Notion and Nature of the European Constitution after the Reform Treaty’ (18 January 2008) accessed 1 April 2015. 61  Ogendo notes in this regard that constitutional law is concerned not with ‘abstract norms but with the creation, distribution, exercise, legitimization effects and reproduction of power’. See Okoth-Ogendo (n 55). 62  See UK Preuss ‘The Implication of “Eternity Clauses”: The German Experience’ (2011) 44 Israel L Rev 429. Gardner, though making reference to state constitutions, sums up this role noting that: ‘constitutions serve three main functions: they authorise and empower the state government to achieve the public good; they grant the state power sufficient to resist abuses of national authority by the national government; and they control the state power thus granted by establishing institutions of governmental self-restraint so that the state government does not become an undue threat to its own people. Any state constitution thus strikes a balance between empowerment and restraint of state government…’. See JA Gardner, ‘The Regulatory Role of State Constitutional Structural Constraints in Presidential Elections’ (2001) 29 Florida State University Law Review 625. 63  For justification see J Madison, Federalist Paper No 51 accessed 1 April 2015. 64  See JC Ville, Constitutionalism and the Separation of Powers, 2nd edn (Liberty Fund, 1998).

The Implication of Conflation of Normal and ‘Constitutional Politics’ 339 for constitutional instability in some African countries.65 Most constitutions created power, yet failed to either constrict the exercise of that power to what secures public good and justice or structured government in a manner that rendered the judiciary subservient to the executive and legislature.66 A relevant constitution is also one that is expected to substantively secure commitment to its rules of change. Predominantly, constitutions directly control formal but not informal change.67 Since constitutions generally lack provisions that render legitimate other provisions when others are impugned, contempt for any of its provisions arguably signifies contempt for it in its entirety. This is because in theory, constitutions, unlike normal legislation, function when they are considered as legitimately binding to the main political actors.68 It is possibly safe to conclude that blatant disregard to constitutions generally in some parts of Africa is a result of disdain of their rules of change.69 In the end, there is a lack of loyalty on the part of the political elites which leads to instability because it also provides precedent and justification for anti-constitutional forces to resort to extra-constitutional means of constitutional alterations. It is for this reason that impetuous circumstances initiate the process of constitution making in most African countries. This is not however a peculiar problem to Africa. Jon Elster suggests generally that constitution-making processes follow or occur in the context of violence as follows:70 We should not be surprised that constitution-making goes together with violence. According to a cliche´ to which I have unfortunately contributed (Elster 1984, ch. II.7), constitutions are typically written in a calm and reflective moment that enables sober and public-spirited framers to design institutions that will prevent the interests and passions of future actors from acting against the general interest. The reality is different.71

Readily available evidence exists to show the prevalent role of cataclysmic events in constitutional substitution. In Africa, constitutional change, prior to the 1990s in most countries has indeed been inaugurated by set-ups such as regime change, military subjugations, and civil wars.72 Coup d’état’s have been very common initiators

65  In the context of Kenya, see MW Mutua, ‘Justice Under Siege: The Rule of Law and Judicial Subservience in Kenya’ (2001) 23 Human Rights Quarterly 96‑118, accessed 31 March 2015. 66  Okoth-Ogendo (n 55) 3. 67  On informal constitutional change, See S Griffin, ‘Constituent Power and Constitutional Change in America’s Constitutionalism’. accessed 20 March 2015; See also J Arato, ‘A Theory of Informal ­Constitutional Change in International Organizations’ (International Journal of Constitution Law Blog, 23 January 2013) accessed 1 April 2015. 68 For a contrary view (on reverence to textuality), see R Albert, ‘The Cult of Constitutionalism’ (2012) 39 Fla St U L Rev 374. 69  Generally see Okoth-Ogendo (n 55). 70  J Elster ‘Constitution-Making and Violence’ (2012) 4(1) Journal of Legal Analysis 7. 71 ibid. 72 See A Aguda, ‘The Military and Constitutionalism in Africa: The Nigerian Experience’ in Shivji (n 55) 203.

340  Duncan Okubasu of constitutional change in most African countries and have not disappeared from the African political landscape even in the period 1990–2010.73 In recent times, and despite supra-national efforts to condemn unconstitutional change of governments in Africa, it is ubiquitous to see anti-constitutional forces resort to anti-­ constitutional means to alter constitutional orders and examples in this regard are too many to cite. In 2009 there was a successful coup d’état in Madagascar.74 There have also been attempts at regime and constitutional change through use of force in Central African Republic (CAR), Guinea-Bissau, Cote d'Ivoire, Burundi and Sao Tome & Principe.75 Other recent incidents were evidenced in Arabic speaking countries, resulting in constitutional change.76 Even those that have had a fairly peaceful constitutional transitional process have experienced some violence of some sort, for instance Kenya. The process of making the 2010 Constitution, though facilitated by an inbuilt replacement clause, was invigorated by the 2007–2008 post-election violence.77 Anti-constitutional forces in societies resort to these mechanism because they attach no significance on rules of change as being able to open up the Constitution for useful deliberation. VI. CONCLUSION

This chapter has attempted to explain the repercussion of conflation of what I delimit as ‘normal’ and ‘constitutional’ politics on constitutional change—as an apparent failure of the dualist project. It attributes the unpleasant constitutional odyssey associated with amendment and substitution to improper use of normal politics to secure substantive constitutional change. Africa would easily provide a good illustration of the problems associated with constitutional change (of the effective constitution) in the absence of constitutional moments. Since normal politics are supposed to attract casual attention from the citizens, the use of normal politics to devise constitutional change arguably leads to casual attention to constitutional developments and sociologically illegitimate constitutional processes and outcomes that do not attract wider societal acceptance: ultimately resulting in a sort of vicious circle of instability. To remedy this condition, this contribution recommends that constitutional design in Africa should envisage the possibility of diminishing the place of normal politics in significant constitutional change, especially constitutional amendment and replacement. Institutional autonomy should be essentialised as a design feature to guard against capture of political institutions and the success of judicial review— through judicial empowerment—as a means of perfecting both formal and informal

73  P Tesragiorgis and W Wahiu, frica: Constitution Building and Coup Making IDEA’ (openDemocracy, 28 April 2009) accessed 12 August 2014. 74 ibid. 75 ibid. 76  Generally see M Lynch, The Arab Uprising: The Unfinished Revolution of the New Middles East (Public Affairs, 2013). 77 D Okubasu ‘The Implications of Constitutional Flexibility on the Lifespan of the Kenyan ­Constitution’ (2014) 24 Comparative Constitutional and Administrative Law Quarterly 38.

The Implication of Conflation of Normal and ‘Constitutional Politics’ 341 constitutional change mechanism should be of overarching concern for constitutional architects.78 The latter concern is vital—not just because it can help in vetting formal means of change but also because most momentous changes to the effective constitution occur informally yet their legality can be challenged—by ordinary citizens, opposition or social forces—through judicial review. In this way, constitutional processes are more likely to be acceptable and legitimate which might in turn lead to more stable and enduring constitutions given the relationship that has been said to exist between inclusive constitutional developments and constitutional stability and endurance.79

78 See T Ginsburg, Judicial Review in New Democracies: Constitutional Courts in Asian Cases ­(Cambridge University Press, 2003) 6, 7–8. 79  D Anckar, ‘Notes on Constitutional Endurance’ (2014) 5 Beijing Law Review 80–88, 81.

342 

18 Direct Democracy and Constitutional Change in the US Institutional Learning from State Laboratories JURGEN GOOSSENS1

I. INTRODUCTION It is one of the happy incidents 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. Justice Brandeis (dissenting opinion), New State Ice Co v Liebmann2

Traditionally, research of constitutional law scholars in the US mainly focuses on the federal Constitution, thereby often neglecting the rich source of constitutional tradition on the state level. Nevertheless, states may serve as laboratories, yielding useful insights in the successes and failures of their experiments. It is the aim of this essay to fill a gap in the literature by thoroughly analysing the constitutional amendment procedures in all 50 state constitutions. Article V of the federal Constitution maps out four different paths to amendment.3 Regarding the initiative to propose amendments, a two-thirds vote in the House and the Senate is required, or two-thirds of the state legislatures can call a national proposing convention. Thereafter, Congress has the power to choose the mode of ratification: approval by three-fourths of the state legislatures, or by conventions in three-fourths of the states. As a result, the 13 least populous states—together

1  The author would like to thank the organisers and participants of the Works-In-Progress Symposium at Yale Law School (10 April 2013), the World Congress of Constitutional Law in Oslo (18 June 2014), the Annual Conference of the Cambridge Journal of International and Comparative Law (8 May 2015), and the IACL-BC Workshop on Comparative Constitutional Amendment at Boston College Law School for granting him the opportunity to present and discuss drafts of the paper. Moreover, the author would like to express special gratitude for the stimulating discussions with Akhil Amar, Bruce Ackerman and Dieter Grimm. Finally, the author would like to thank Richard Albert for inviting him to publish early findings on the I-CONnect blog. 2  New State Ice Co v Liebmann [1932] 285 US 262, 311. 3  See below Section II(A).

344  Jurgen Goossens merely representing approximately 4.4 per cent of the total US population4—could in theory veto any amendment in the ratification stage. The equal representation of states in the Senate also gives small states a disproportional amount of power in the proposing phase of the amendment process, while Article V requires unanimity to alter the equal suffrage clause. The four paths to constitutional change in Article V only involve legislatures. The federal constitutional amendment procedure does not provide any form of direct democracy, such as an initiative petition by citizens or a popular referendum, and has not been altered since its adoption 228 years ago. Article V champions federalism via involvement of federal and state legislatures over direct involvement of the People. At least one could argue that the People are more directly involved at the state level than at the federal level. In 1787 when the Constitution was enacted, the People identified more with their state level rather than with their nation. Nevertheless, it remains an open question whether Article V should be read as the exclusive way to alter the federal Constitution. Based on the principle of popular sovereignty one could argue that the People have an inalienable right to alter or abolish the Constitution that they, themselves, have ordained and established.5 In contrast to Article V of the federal Constitution, the amendment procedures in the state constitutions have undergone numerous alterations and improvements. State tradition regarding amendment and revision procedures has in principle become more flexible than the onerous federal approach and is mainly characterised by direct democracy and majoritarian voting rules. The analysis of the 50 state constitutional amendment processes could serve as a starting point for a better informed normative debate about altering the amendment procedures in the states and on the federal level. If one reflects on the possible implementation of state experiments on the federal level in the future, one has to take into account the different features of the federal context, such as a higher likelihood of factionalism. II.  ARTICLE V OF THE US CONSTITUTION

A.  The Four Classic Paths to Amendment The first part of Article V of the Constitution provides the following paths for formal amendment of the Constitution: ‘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.’ In other words, the amendment procedure can be initiated through the federal or state legislative level. Either a two-thirds vote of both Houses of Congress is 4  5 

See below (n 10). AR Amar, America’s Constitution: A Biography (Random House, 2005) 292.

Direct Democracy and Constitutional Change in the US 345 ­ ecessary to propose an amendment or two-thirds of the state legislatures can oblige n Congress to call a special convention to propose amendments. Hitherto, not a single constitutional amendment has been adopted via the latter option.6 Thereafter, the proposed amendment needs to be ratified by three-quarters of the states for approval. Congress may decide whether the states can act through their legislatures or via special ratifying conventions. There are no time limits to this procedure, although it would make sense to impose them.7 In an important empirical study, Donald Lutz analysed the constitutions of 32 countries as well as the 50 US states, which led to the remarkable conclusion that ‘the [federal] US Constitution is unusually, and probably excessively difficult to amend. The United States should move either to the strategy of using a referendum, in which case its amendment rate may well triple, or else reduce the number of states required for amendment ratification to two-thirds (from three-fourths) …’.8 Indeed, Article V does not provide any form of direct democracy, such as initiative petitions or referenda. It is still reminiscent of the initial fear of Southern slavocratic leaders for a popular vote.9 Moreover, the equal representation of states in the Senate gives small states substantially more power in the proposal stage of amendments than states with a large population. In addition, due to the required three-fourths supermajority of the states to approve any amendment, simple majorities of the delegates at state conventions or members of the state legislatures in the 13 least populous states, approximately only representing 2.22 per cent of the US population,10 could potentially block any amendment in the ratification stage. Moreover, calling a national proposing convention does not seem to be a preferred or feasible option for the states, so that a two-thirds vote of both chambers of Congress is usually necessary. Consequently, Congress has de facto always a veto power, which makes it unlikely to pass any amendment limiting the power of Congress or increasing the power of the state level.11

6  MB Rappaport, ‘Reforming Article V: The Problems Created by the National Convention Method & How to Fix Them’ (2010) 96 Va L Rev 1509, 1512–3 (invoking the uncertainty of a runaway convention in order to explain why states find it unattractive to call a national proposing convention). 7  MC Hanlon, Note, ‘The Need for a General Time Limit on Ratification of Proposed Constitutional Amendments’ (2000) 16 J L & Politics 663; see also M Kalfus, Comment, ‘Time Limits on the Ratification of Constitutional. Amendments Violate Article V’ (1999) 66 U Chi L Rev 437. 8  DS Lutz, ‘Toward a Theory of Constitutional Amendment’ in S Levinson (ed), Responding to Imperfection: The Theory and Practice of Constitutional Amendment (Princeton University Press, 1995) 237, 265. See also S Levinson, Our Undemocratic Constitution (Oxford University Press, 2006) 21 and 160–5. 9  cf Amar (n 5) 298. 10  Based on the data of the census of 2010, available at accessed 1 August 2016. In according to the census of 2010, the 13 states with the smallest population number together have 13,725,340 citizens, which approximately represents 4.44% of the total US population (308,745,538). It should, however, be noted that principally a simple majority of the delegates in a state convention or of the members of the state legislature, thus approximately representing 2.22% of the US population, are sufficient to ratify an amendment. See P Suber, ‘Population Changes and Constitutional Amendments: Federalism Versus Democracy’ (1987) 20 U Mich J L Reform 409 (denouncing the three-fourths ratification requirement and noting how population changes de facto altered the amendment process of Art V). 11  Rappaport (n 6) 1511–13.

346  Jurgen Goossens In conclusion, the current amendment process of Article V champions federalism over direct involvement of the People. Nevertheless, one could invoke the principle of popular sovereignty to defend a non-exclusive reading of Article V. Hereafter, the debate about an exclusive versus a non-exclusive reading of Article V will briefly be analysed. B.  Exclusive Reading of Article V A majority of legal scholars support an exclusive reading of the amendment procedure in Article V of the federal Constitution in order to formally amend the Constitution.12 According to this view, one can only rely on this formal procedure to alter the Constitution. Moreover, it is widely acknowledged that in order to alter the amendment process of Article V itself one must first comply with the current amendment procedure in Article V. Nevertheless, based on the principle of popular sovereignty one could argue that the People have an inalienable right to alter or abolish13 the Constitution that they, themselves, have ordained and established.14 This argument could be invoked to oppose a strict exclusive reading of the amendment procedure in Article V. C.  Non-Exclusive Reading of Article V Contrary to the current mainstream reading of Article V as the exclusive way to formally amend the Constitution, one could argue that Article V supplements the inalienable right of the People to alter the rules governing them. Admittedly, based on an intra-textual argument, one could try to counter this argument by invoking that exclusivity is also implied in other provisions of the Constitution without being explicitly stipulated. For instance, the Article III roster of cases and controversies limits the federal court’s jurisdiction to the nine listed lawsuits without mentioning ‘only’.15

12  See, eg, DR Dow, ‘The Plain meaning of Article V’ in S Levinson (ed), Responding to Imperfection: The Theory and Practice of Constitutional Amendment (Princeton University Press, 1995) 117; HP Monaghan, ‘We the Peoples, Original Understanding, and Constitutional Amendment’ (1996) 96 Colum L Rev 121, 127; LB Orfield, The Amending of the Federal Constitution (University of Michigan Press, 1942) 39; LH Tribe, ‘Taking Text and Structure Seriously: Reflections on Free-Form Method’ in Constitutional Interpretation’ (1995) 108 Harv L Rev 1221, 1233; JR Vile, ‘Legally Amending the United States Constitution: The Exclusivity of Article V’s Mechanisms’ (1991) 21 Cumb L Rev 271. 13  Some might argue that the People only retain the inalienable right to entirely abolish the Constitution, as opposed to just alter it. I reject this view, inter alia, because abolishing the entire constitution and replacing it with the same constitution with minor changes is de facto the same as an alteration. 14  Amar (n 5) 292; Y Roznai, ‘Amendment Power, Constituent Power, and Popular Sovereignty: Linking Unamendability and Amendment Procedures’, in this volume (arguing that the Constitution cannot restrict the primary constituent power); see also Th Pereira, ‘Constituting the Amendment Power: A Framework for Comparative Amendment Law’ in this volume (analysing the relationship between popular sovereignty and amendment power). 15  Amar (n 5) 295.

Direct Democracy and Constitutional Change in the US 347 Nonetheless, three important arguments can be invoked to reject an exclusive reading of Article V: 1.  popular sovereignty was the bedrock principle for the establishment of the ­Constitution and could also be invoked in favour of direct involvement of the People to alter the Constitution; 2.  the Framers intentionally amended state constitutions without respecting ­several state amendment procedures, which required a non-exclusive reading of those Article V counterparts; 3.  there has always been a strong constitutional tradition in the states of organising constitutional conventions outside the state constitutional provisions analogous to Article V.16 At the time of drafting Article V of the new US Constitution, the first state Constitutions did not constitute useful examples for drafting a provision about the constitutional amendment procedure.17 The first state Constitutions of Virginia (1776), North Carolina (1776) and New York (1777) did not encompass any explicit amendment process, while Connecticut and Rhode Island still lived under the old Crown Charters which also did not comprise any amendment procedure. New ­Jersey’s Constitution of 1776 and South Carolina’s Constitution of 1778 principally granted the state legislature the power to amend the Constitution by ordinary law-making.18,19 Delaware required a five-sevenths vote of the state legislature and Maryland demanded a legislative vote in two consecutive sessions with an election in between.20 However, all these state constitutions were not established through an act of direct popular involvement and did not regulate how the People themselves could have a say in the amendment process. Georgia’s Constitution of 1777 provided citizens with the right to petition for calling a proposing convention, but remained silent about the next step.21 Hence, those ten state constitutions by no means constituted valuable examples for drafting an amendment process to alter the federal Constitution which was uniquely ordained by the People. Only the Constitutions of Pennsylvania (1776), Massachusetts (1780) and New Hampshire (1776) were established by more direct democracy.22 They provided that an amendment process might be initiated at a fixed date or at fixed time intervals by an institution other than the legislature. However, those three documents remained silent about amendment through other means or at other times.23

16 

See below Section III(C). (n 5) 287–88 (explaining in more detail how the states were able to amend their first constitutions). 18  NJ Const of 1776, Art XXIII. A member of New Jersey’s legislator only had to swear in his oath to not alter certain provisions of the Constitution. 19  SC Const of 1778, Art XLIV. 20  Del Const of 1776, Art 30; Md Const of 1776, Art LIX. 21  Ga Const of 1777, Art LXIII. 22  Pa Const of 1776, s 47; Mass Const of 1780, part II, ch VI, Art X; NH Const of 1776, part II. 23  Amar (n 5) 288–9. 17 Amar

348  Jurgen Goossens The lack of useful examples at the state level establishing a detailed amendment process, which provided involvement of the People, was one of the reasons why the Framers did not include direct democracy in Article V. Consequently, the federal Constitution does not address the following fundamental and intriguing questions. Could citizens take the initiative to call a proposing convention without petitioning of the States? Could citizens petition to propose a specific amendment? Could a proposed amendment be ratified by popular vote in a referendum? In this regard, James Wilson argued in 1787 that ‘[t]he people may change the Constitution whenever and however they please. This is a right of which no positive institution can ever deprive them’.24 Besides options that favour more direct democracy, one could also pose the following two questions. Should Congress on its own be capable of calling a proposing convention without petitioning by the states? And, should the state legislatures be allowed to petition a ballot measure proposing a specific amendment? An analysis of the constitutional tradition on the state level reveals that many states have already amended their constitution or adopted new constitutions via procedures involving the People, which were not provided in the constitutional amendment process of their state constitution. In 1787, state governments had initially sent their delegates to the Constitutional Convention in Philadelphia to amend the Articles of Confederation. There was originally no intention to abolish those Articles entirely. Article XIII of the Articles of Confederation stated that any amendment required the consent of all 13 state legislatures. This unanimity requirement, however, turned out to be de facto ­unattainable.25 Rhode Island even refused to send delegates to the Constitutional Convention.26 As a result, the delegates started drafting a constitution that did not require ratification by all 13 state legislatures. Pursuant to Article VII of the ­Constitution, approval by conventions in at least nine states was sufficient for the Constitution to enter into effect for the ratifying states. In other words, the amendment procedure in the Articles of Confederation was simply set aside.27 By seeking ratification of the Constitution through conventions in at least nine states, the delegates to Philadelphia intentionally sidestepped the required unanimity in Article XIII of the Articles of Confederation. Moreover, though often overlooked, the Constitution—being the Supreme Law28—also altered existing state constitutions and sidestepped the amendment procedures of the seven state constitutions encompassing such a procedure.29 Although the Constitution of Pennsylvania, Massachusetts and New Hampshire at that time prescribed a scheduled amendment procedure at a fixed date (1795 in Massachusetts) or at fixed time intervals (every seven years in New Hampshire and Pennsylvania), the Philadelphia delegates did not wait for those scheduled times to seek ratification of the Constitution. Moreover, four other 24  J Elliot, The Debates in the Several State Conventions on the Adoption of the Federal Constitution (1836) Vol II, 432. 25  See The Federalist No 43 (James Madison). 26  Amar (n 5) 285–86. 27  See, eg, BA Ackerman, ‘The Storrs Lectures: Discovering the Constitution’ (1984) 93 Yale LJ 1013, 1017–23. 28  See the supremacy clause in Art VI US Const. 29  Amar (n 5) 289.

Direct Democracy and Constitutional Change in the US 349 state constitutions granted the power for amending the state constitution to the state legislature and thus not to state conventions. In New Jersey and South Carolina, the Constitution could be altered by ordinary law-making, in Delaware by a two-thirds vote of the state legislature, and in Maryland by a legislative vote in two consecutive sessions with an election in between.30 Consequently, those conscious acts of the Framers obviously relied on a non-exclusive reading of several state amendment procedures, which are counterparts of Article V of the federal Constitution.31 Moreover, Pennsylvania has already organised five proposing conventions, although its state constitution does not explicitly provide the option of calling a constitutional convention. Similarly, Massachusetts’ Constitution does not mention the option of a constitutional convention, even though four conventions have already been held.32 Bruce Ackerman wrote that before the Civil War at 16 occasions state legislatures have not interpreted the silence in their state constitutions as a reason for exclusivity.33 Moreover, Roger Hoar listed 34 conventions that have been held before 1917 in states of which the constitutions did not prescribe the option to call a convention.34 Akhil Amar argues that Article V should be read as the exclusive way for the federal and state legislators to propose and ratify amendments without involvement of the People, thereby expressing the distrust of the People against ‘imperfect representatives’.35 According to Amar, however, this viewpoint implies that Article V does not constrain the inalienable right of the People itself to alter the Constitution via a national referendum.36 Even if one defends such a view, one should in my opinion first attempt to adopt an explicit amendment to Article V in order to add more direct democracy to the amendment process. In the book Our Undemocratic Constitution, Sanford Levinson convincingly advocates that the Constitution should be revised. However, arguing that the mechanism of Article V is too onerous, he provocatively (and hypothetically) proposes to organise a national referendum on the question: ‘Shall Congress call a national convention empowered to consider the adequacy of the Constitution and, if thought necessary, to draft a new constitution that, upon completion, will be submitted to the electorate for its approval or disapproval by majority vote? …’.37 Undoubtedly, the organisation of a proposing and ratifying referendum would have to rely on a non-exclusive reading of Article V in order to have binding force.38 30 

See above, n 18. See also Amar (n 5) 289. 32  See below Section III(C). 33  BA Ackerman, We The People: Transformations (Harvard University Press, 1998) 80. 34  RS Hoar, Constitutional Conventions: Their Nature, Powers, and Limitations (Little, Brown, and Company, 1917) 39–40. 35  Amar (n 5) 296; AR Amar, ‘Philadelphia Revisited: Amending the Constitution Outside Article V’ (1988) 55 U Chi L Rev 1043, 1054–55 and 1069. 36  Amar (n 35) 1056–60; AR Amar, ‘The Consent of the Governed: Constitutional Amendment Outside Article V’ (1994) 94 Colum L Rev 457, 489–94 (developing an intra-textual argument based on the Preamble, the First, Ninth, and Tenth Amendments); ibid (1994) at 457 and 481–7. Contra Vile (n 12) 271. 37 Levinson, Undemocratic Constitution (n 8) 11–2. 38  ibid 177 (endorsing Akhil R Amar’s non-exclusive reading of Article V and the legality of a ratifying national referendum by simple majority vote). 31 

350  Jurgen Goossens In We The People: Foundations, Bruce Ackerman puts forth his enlightening concept of ‘constitutional moments’. During constitutional moments, citizens deliberately produce higher law-making that better reflects the will of the People than the acts of majoritarian institutions during normal politics, which for instance happened during the constitutional conventions at the Founding.39 Ackerman also endorses a non-exclusive reading of Article V and acknowledges that successful constitutional transformations or amendments can take place outside the formal amendment process. He believes that this for example happened during the New Deal, when democrats publicly and self-consciously sought a new constitutional solution to expand federal regulatory power outside the traditional amendment process of Article V. This took place under the lead of President Roosevelt, relying on the appointment of new judges in the Supreme Court.40 III.  STATE LABORATORIES Federalism enables a people to try experiments in legislation and administration which could not be safely tried in a large centralised 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. Viscount Bryce, The American Commonwealth (1888)41

Friedrich von Hayek introduced the insight that competition is a learning ­process.42 Dividing a country in several jurisdictions with its own legislative powers enables them to learn from each other’s successes and failures. Wallace Oates called this ‘institutional learning’ or ‘laboratory federalism’.43 Decentralisation enables the simultaneous occurrence of multiple policy experiments, such as different amendment processes.44 Competition between legislators may lead to innovation, which could generate improved institutional systems and legal rules. Decentralised jurisdictions could learn from each other and the federal level could also beneficially learn from state experiments. Thus, it is useful—though often neglected—to analyse the amendment procedures in the state constitutions, which have undergone numerous improvements.45

39 

BA Ackerman, We The People: Foundations (Harvard University Press, 1991) 185–86. 50–52. See also DA Strauss, The Living Constitution (Oxford University Press, 2010) 121–23 (referring to the New Deal and McCulloch v Maryland as ‘extratextual amendments’). 41  VJ Bryce, The American Commonwealth, Vol 1 (Liberty Fund, 1995) 312 (1888). 42  FA von Hayek, ‘Competition as a Discovery Procedure’ in FA von Hayek (ed), New Studies in Philosophy, Politics, Economics, and the History of Ideas (Routledge, 1978) 179–90. 43  WE Oates, ‘An Essay on Fiscal Federalism, Journal of Economic Literature’ (1999) Journal of Economic Literature 1120, 1123. 44  KS Strumpf, ‘Does Government Decentralization Increase Policy Innovation?’ (2002) 4 Journal of Public Economic Theory 207, 208. 45  See JJ Dinan, The American State Constitutional Tradition (University Press of Kansas, 2006) 29–63 (analysing the debates in the state conventions to explain why the more flexible state amendment processes differs from the onerous federal process). 40  ibid

Direct Democracy and Constitutional Change in the US 351 Each state has its own constitution, constitutional tradition, and distinct constitutional amendment procedure.46 There have been almost 150 state constitutions, which have been amended approximately 12,000 times.47 It is widely acknowledged that the state approach concerning amendment and revision is more flexible that the onerous federal approach.48 This contribution will analyse the different paths to constitutional amendment in the Article V counterparts of the current 50 state constitutions. An analysis of all 50 state constitutional amendment processes may serve as a starting point for a better informed normative debate about the amendment procedures on the federal and state level. In the near future, it is advisable that comprehensive empirical research would be carried out with regard to the actual practice and efficacy of the existing constitutional amendment procedures. Generally, one can distinguish four different paths to proposing a constitutional amendment or revision in the states: (1)  (2)  (3)  (4) 

proposal of the state legislature; popular initiative; calling a constitutional convention; and establishing a constitutional commission.

The ratification of proposed amendments and revisions usually takes place through a referendum. A.  Amendment by Proposal of the State Legislature Currently, 49 state constitutions allow the state legislature to propose constitutional amendments which are then brought before the voters. In Delaware, however, the state legislature on its own can amend the Constitution in two consecutive sessions without popular vote.49 Hereafter, it will be examined how this constitutional amendment process differs from state to state, whereby we will mainly focus on three aspects. The following questions will be answered: 1.  whether a proposed amendment needs to be approved by the state legislature during one or two legislative sessions, and whether a simple majority or supermajority vote of the legislature is required; 2.  whether a proposed amendment can be put on the ballot during a general and/ or special election; and 3.  whether the popular vote requires a simple majority or supermajority of the voters to ratify the amendment.

46  See generally GE Connor and Chr W Hammons, The Constitutionalism of American States (University of Missouri Press, 2008). 47  JJ Wallis, ‘NBER/University of Maryland State Constitution Project’ accessed 10 April 2015. 48  Dinan (n 45) 30–31. 49  See G Benjamin, ‘Constitutional Amendment and Revision’ in GA Tarr and RF Williams (eds), State Constitutions For The Twenty-First Century (State University of New York Press, 2006) Vol 3, 177, 185.

352  Jurgen Goossens i.  One or Two Legislative Sessions—Supermajority vs Simple Majority Vote In 35 states, an affirmative vote of the legislature during only one session is sufficient to propose an amendment. Ten states50 out of those 35 require a simple majority vote of the state legislature. Among those ten states, a remark should be made about the situation in Oklahoma and New Mexico. Oklahoma’s constitution prescribes a simple majority vote, unless the legislature puts the amendment on a special ballot in which case a two-thirds vote is required.51 Moreover, in New Mexico an amendment proposed by the state legislature that would restrict the rights created by ­Section 1 or Section 3 of Article VII or Section 8 and Section 10 of Article XII requires a threefourths vote of the state legislature in order to be put on the ballot.52 Nine out of those 35 states require a 60 per cent supermajority vote to bring a proposed amendment before the voters.53 Among those nine states, Nebraska’s constitution imposes a supermajority of 80 per cent in order to submit a proposed amendment to a special election.54 In 16 states,55 a two-thirds vote in the state legislature is required to propose an amendment. Besides those 35 states, the constitutional amendment procedure in two other states should be mentioned. South Carolina’s constitution prescribes that an amendment requires a two-thirds vote in one legislative session in order to bring it before the voters. However, after the approval of the voters, it returns to the legislature for a second, simple majority vote that must take place ‘after the election and before another’.56 In Pennsylvania, an amendment can only be brought before the voters after one session in case of a major emergency declared by the legislature by a two-thirds vote. Otherwise, an amendment proposal needs to be approved in two consecutive legislative sessions by a simple majority vote.57 In twelve states (including the aforementioned states South-Carolina and Pennsylvania), the state constitution requires votes in two consecutive sessions of the state legislature to propose amendments.58 Eight states require a simple majority vote in both sessions before seeking popular ratification.59 Although Delaware’s constitution also requires consideration in two consecutive sessions, the proposed amendment does not have to be brought before the voters to seek ratification.60 The other three states require different voting thresholds in the first and second session. As above-mentioned, in South Carolina a two-thirds vote of the state legislature in the

50  Arizona, Arkansas, Minnesota, Missouri, New Mexico, North Dakota, Oklahoma, Oregon, Rhode Island, South-Dakota. 51  Okla Const, Art XXIV, para 1. 52  NM Const, Art 19, para 1. 53  Alabama, Florida, Illinois, Kentucky, Maryland, Nebraska, New Hampshire, North Carolina, Ohio. 54  Neb Const, Art XVI, para 1. 55 Alaska, California, Colorado, Georgia, Idaho, Kansas, Louisiana, Maine, Michigan, Mississippi, Montana, Texas, Utah, Washington, West Virginia, Wyoming. 56  SC Const, Art XVI, para 1. 57  Pa Const, Art XI, para 1. 58 Delaware, Indiana, Iowa, Massachusetts, Nevada, New York, Pennsylvania, South-Carolina, ­Tennessee, Vermont, Virginia, Wisconsin. 59  Indiana, Iowa, Massachusetts, Nevada, New York, Pennsylvania, Virginia, Wisconsin. 60  Del Const, Art XVI, para 1.

Direct Democracy and Constitutional Change in the US 353 first session refers a proposed amendment to the ballot and, after the popular vote, a simple majority vote of the state legislature is required to enter the amendment into force. Article XI, Section 3 of Tennessee’s Constitution uniquely requires a supermajority of two-thirds in the second session, which is different from the simple majority required in the first session. In Vermont, a two-thirds majority in the Senate and a simple majority in the House of Representatives are required in the first session, while a majority in both Houses is sufficient in the second session.61 In four states, a proposed amendment requires a simple majority in two legislative sessions or a supermajority vote in one session. The state legislature of Connecticut can choose to approve the proposal of an amendment by a supermajority vote of three-fourths in one session or a simple majority in two consecutive sessions.62 In Hawaii, the state legislature can approve a proposed amendment by a two-thirds vote in one session or a simple majority in two successive sessions of the state legislature.63 In New Jersey, a supermajority vote of 60 per cent in one session or a simple majority in two consecutive sessions is required to propose an amendment.64 As aforementioned, a simple majority vote in two successive sessions of the state legislature can bring a proposed amendment before the voters in Pennsylvania, but if a ‘major emergency threatens or is about to threaten the Commonwealth’ a two-thirds vote in the legislature can refer a proposed amendment to the ballot in one session.65 Finally, Oregon’s constitution uniquely regulates how a proposal to wholly or partially revise the Constitution can be brought before the voters outside a constitutional convention, namely with a two-thirds vote of the state legislature instead of the simple majority vote that is required to propose specific constitutional amendments.66 ii.  Special vs General Election In 20 states, amendments proposed by the state legislature can only trigger a popular vote in a general election.67 In addition to those 20 states, in Tennessee the popular vote on proposed amendments must specifically take place during the general election in a year that one can elect the governor.68 The option between a special or general election is made possible by 25 state constitutions.69 Article XIV, Section 2 of West Virginia’s constitution imposes the following unique condition regarding

61 

Vt Const, S 72. Conn Const, Art XII. Haw Const, Art XVII, para 3. 64  NJ Const, Art IX. 65  Pa Const, Art XI, para 1. 66  Or Const, Art XVII, para 2. 67 Alaska, Arkansas, Colorado, Connecticut, Georgia, Hawaii, Idaho, Illinois, Indiana, Kentucky, Maine, Maryland, Minnesota, Montana, New Jersey, Rhode Island, South Carolina, Utah, Washington, Wyoming. 68  Tenn Const, Art XI, para 3. 69  Alabama, Arizona, California, Florida, Oregon, Pennsylvania, Iowa, Kansas, Louisiana, ­Michigan, Mississippi, Missouri, Nebraska, Nevada, New Mexico, New York, North Carolina, Ohio, Oklahoma, Oregon, Pennsylvania, Texas, Virginia, West Virginia, Wisconsin. In order to submit a proposed amendment to a special election, Nebraska’s constitution imposes a supermajority of 80%. Similarly, O ­ klahoma’s constitution requires a two-thirds vote of the state legislature to bring a proposed amendment before the voters at a special election. 62  63 

354  Jurgen Goossens proposed amendments at a special election: ‘Whenever one or more amendments are submitted at a special election, no other question, issue or matter shall be voted upon at such special election.’ Finally, the state constitutions of New Hampshire, North Dakota, South Dakota, and Vermont do not mention whether the state legislature can bring an amendment before the voters at a general and/or special election. iii.  Size of Popular Vote After an affirmative vote of the state legislature to propose an amendment, a simple majority vote of the citizenry is required in 40 states in order to approve the proposed amendment.70 In addition, a proposed amendment demands a simple majority vote by the citizens of Louisiana. However, if the amendment affects more than five parishes or five municipalities in Louisiana, it requires approval by a majority vote state-wide and a majority vote in the respective parishes or municipalities that it affects.71 Moreover, Article XIV, Section 1 of the Maryland Constitution provides that certain constitutional amendments may apply to only one county or to the City of Baltimore. In that case, the proposed amendment must be approved by a majority vote state-wide and in the respective county or in Baltimore. In general, a simple majority of citizens voting on the proposed amendment is required, but in Minnesota and Wyoming a majority of the voters in the election is needed for approval of the proposed amendment. Three states require a double majority vote by the citizens to ratify proposed amendments. This generally means that besides a simple majority of the votes cast on a particular proposal, a certain percentage of all the voters in the election is also required. This is necessary even if not all the voters in the election cast a vote for the proposed amendment. Consequently, this makes amending more difficult as it is likely to happen that less people vote on the proposed amendment than those who vote on the candidates in the election. Article XVII, Sections 2 and 3 of Hawaii’s Constitution outlines that a proposed amendment is approved in two cases. Firstly, an approval is achieved by a majority vote on the specific question if this majority also represents at least 50 per cent of all the votes generally cast in the election. Secondly, an approval is obtained in a special election by a majority of all the votes on the question if this majority represents at least 30 per cent of all the registered voters in the state at the time of the election. In Nebraska, a proposed amendment requires a majority vote on the amendment, in addition to at least 35 per cent of those voting in the election for any office.72 Tennessee’s Constitution demands a majority of those voting on the proposed amendment in addition to a majority of all the citizens of the state voting for governor.73 Besides those three states, in Utah only a vote of

70  Alabama, Alaska, Arizona, Arkansas, California, Colorado, Connecticut, Georgia, Idaho, Indiana, Iowa, Kansas, Kentucky, Maine, Massachusetts, Michigan, Minnesota, Mississippi, Missouri, ­Montana, Nevada, New Jersey, New Mexico, New York, North Carolina, North Dakota, Ohio, Oklahoma, O ­ regon, Pennsylvania, Rhode Island, South Carolina, South Dakota, Texas, Vermont, Virginia, Washington, West Virginia, Wisconsin, Wyoming. 71  La Const, Art XIII, para 1. 72  Neb Const, Art XVI, para 1. 73  Tenn Const, Art XI, para 3.

Direct Democracy and Constitutional Change in the US 355 at least a majority of the electors of the state, which are voting at the next general election after the proposal of the amendment, is required to approve a proposed amendment.74 In Illinois an amendment proposed by the state legislature requires a simple majority of those casting a ballot for any office in that election or a supermajority vote of 60 per cent of the people voting on the question.75 Since 2006, Article XI, Section 5 of Florida’s Constitution requires a supermajority vote of 60 per cent of the people voting on the question for approval of a proposed amendment. In addition, a proposal to introduce a new State tax or fee in Florida via amendment of the state constitution requires the approval by two-thirds of the voters.76 Article 100 of New Hampshire’s Constitution requires a positive vote by two-thirds of the voters voting on the amendment. Finally, as aforementioned the state legislature of Delaware can uniquely amend the Constitution on its own in two consecutive sessions without popular vote. In conclusion, only Florida, New Hampshire, and Delaware substantially depart from the state tradition of direct democracy and majoritarian voting rules. B.  Amendment by Popular Initiative One could advocate that citizens should be given the initiative to propose amendments via petitioning.77 In 18 states, voters are not only allowed to approve proposed amendment, but can also themselves initiate a proposal to amend the Constitution.78 Those states allow citizens to propose amendments through petitions that must be signed by a certain percentage of voters, ranging from three to 15 per cent of designated voting groups that vary in each constitutional provision.79 Nine states also have requirements about the territorial distribution of those signatures.80 When the required amount of voters signed the petitions, the proposed amendment is brought before the voters for ratification.

74 

Utah Const, Art XXIII, para 1. Ill Const, Art XIV, para 1. 76  Fla Const, Art XI, para 7. 77  See Orfield (n 12) 177–80 (discussing proposals advocating to allow a proposal of amendments by popular initiative on the federal level and referring to constitutional practice in the states). 78  Arizona, Arkansas, California, Colorado, Florida, Illinois, Massachusetts, Mississippi, Michigan, Missouri, Montana, Nebraska, Nevada, North Dakota, Ohio, Oklahoma, Oregon, and South Dakota. See KK DuVivier, ‘By Going Wrong All Things Come Right: Using Alternative Initiatives to Improve Citizen Lawmaking’ (1995) 63 U Cin L Rev 1185. 79  See Ariz Const, Art XXI, para 1; Cal Const, Art II, para 9; Colo Const, Art V, para 1; Fla Const, Art XI, para 3; Ill Const, Art XIV, para 3; Mass Const, Art XLVIII, The Initiative, IV, para 2; Miss Const, Art XV, para 273; Mich Const, Art XII, para 2; Mo Const, Art III, para 50; Mont, Art XIV, para 2; Neb Const, Art III, para 2; Nev Const, Art 19, para 2; ND Const, Art III, para 9; Ohio Const, Art II, section 1a: Okla Const, Art V, para 2; Or Const, Art IV, para 1; SD Const, Art XXIII, para 1. 80  Ark Const, Art 5, para 1; Fla Const, Art XI, para 3; Mass Const, Art XLVIII, General Provisions, II; Miss Const, Art XV, para 273; Mo Const, Art III, para 50; Mont, Art XIV, para 2; Neb Const, Art III, para 2; Nev Const, Art 19, para 2; Ohio Const, Art II, s 1a. 75 

356  Jurgen Goossens In 13 out of the 18 states, a simple majority of the voters is sufficient to ratify the proposed amendment.81 In addition, three states require a certain percentage of votes cast in the election in addition to a simple majority voting yes on the particular amendment. Massachusetts requires a simple majority of the voters voting on the particular amendment in addition to at least 30 per cent of the total number of ballots in the state election.82 Mississippi demands a simple majority of the votes on the particular amendment and at least 40 per cent of the total votes cast at the election.83 Nebraska requires a simple majority of the votes cast on the particular amendment and not less than 35 per cent of the total vote cast at the election at which the proposed amendment was submitted.84 In Florida, a supermajority vote of 60 per cent is needed for approval.85 Finally, in Illinois a ratification of a proposed amendment can be achieved by 60 per cent of the people voting on the amendment or a majority of those voting in the election.86 Nevertheless, in several states the requirements for this form of direct democracy are so strict that it has only been successfully used on rare occasions. Especially in Illinois, Massachusetts and Mississippi the gate to successful popular initiative has traditionally been extremely difficult to open, inter alia due to the following burdensome conditions: (1)  I n Illinois, popular initiative is limited to amendments regarding structural and procedural subjects contained in Article IV of Illinois’ Constitution. Moreover, a proposed amendment needs to be approved by 60 per cent of the people voting on the amendment or a majority of those voting in the election.87 (2)  In Massachusetts, initiative petitions are explicitly prohibited for numerous constitutional provisions.88 Furthermore, a proposed amendment by initiative petition can itself be amended a priori, that is, before the popular vote, by a 75 per cent vote of the state legislature in joint session. Additionally, an affirmative vote of at least 25 per cent in a joint meeting of the state legislature in two sessions is required before it can be put on the ballot. The voters can approve the initiative amendment or the legislative substitute by at least 30 per cent of the total number of ballots in the state election in addition to a majority of the voters voting on the particular amendment.89 (3)  In Mississippi, an initiative petition to amend the Constitution must be signed by at least 12 per cent of the votes cast for all candidates for Governor in the last gubernatorial election and this must be accomplished within twelve

81  Arizona, Arkansas, California, Colorado, Michigan, Missouri, Montana, Nevada*, North Dakota, Ohio, Oklahoma, Oregon, and South Dakota. * Voters in Nevada need to approve constitutional amendments by initiative with a simple ­majority vote in two separate elections in order to become part of the state constitution (Nev Const, Art 19, para 2). 82  Mass Const, Art XLVII, IV, paras 3–5. 83  Miss Const, Art XV, para 273. 84  Neb Const, Art III, para 4. 85  Fla Const, Art XI, para 5. 86  Ill Const, Art XIV, para 3. 87  Ill Const, Art XIV, para 3. 88  Mass Const, Art XLVII, II, para 2. 89  Mass Const, Art XLVII, IV, paras 3–5.

Direct Democracy and Constitutional Change in the US 357 months.90 Moreover, it cannot relate inter alia to the Bill of Rights and there are tough conditions regarding territorial distribution of the signatures. The signatures of the qualified electors from any congressional district may not exceed one-fifth of the total number of signatures required to qualify an initiative petition for placement upon the ballot, otherwise the signatures in a district that are in excess of one-fifth of the total number of signatures shall not be considered. Additionally, the initiative must mention the amount and source of revenue that is required for its implementation. If the initiative requires a reduction of government revenue or a reallocation of funding from currently funded programs, the text of the initiative must identify which program funding must be reduced or eliminated for its implementation. Finally, an approval or amendment of a constitutional initiative requires a majority vote of each house of the Legislature before it can be put on the ballot. In order to approve the original initiative or the proposed amendment approved by the legislature, a simple majority of the votes on the particular amendment and at least 40 per cent of the total votes cast at the election are required for ratification. Over time, requirements have also been made substantially more stringent in ­Florida, Montana, Nebraska, Nevada, Oklahoma, and Oregon. For instance, in 2006 an amendment to Article XI, Section 5 of Florida’s Constitution raised the percentage of voters required to approve a proposed amendment on the ballot from a simple majority to a supermajority vote of 60 per cent. In 1994, the Supreme Court of Nebraska ruled in Duggan v Beermann91 that a part of Article III, para 4 of Nevada’s Constitution has been implicitly repealed by an amendment to Article III, para 2 in 1988. Section 4 provides that the required number of signatures to put an amendment on the ballot is 10 per cent of the people who voted for governor in the most recent gubernatorial election, while section 2 has replaced this by 10 per cent of the state’s registered voters. Voters in Nevada need to approve constitutional amendments by initiative in two separate elections in order to become part of the state constitution.92 In Oklahoma, 15 per cent of the legal voters that cast a vote at the last general election for the Office of Governor are required in order to propose constitutional amendments by petition.93 Article IV, para 1b of Oregon’s Constitution is the only provision in a state constitution (ie, not in legislation) explicitly stating that it is unlawful to pay or receive money or other thing of value based on the number of signatures obtained for an initiative. C.  Constitutional Convention A proposing constitutional convention at the state level is a gathering of elected delegates with the aim of proposing major revisions and amendments to the e­ xisting

90 

Miss Const, Art XV, § 273. Duggan v Beermann, [1994] 245 Neb. 907, 515 NW 2d 788. Nev Const, Art 19, para 2. 93  Okla Const, Art V, para 2. 91  92 

358  Jurgen Goossens state constitution or drafting an entirely new constitution. In 40 states it is now explicitly regulated how such a constitutional convention can be called: (1)  F  irstly, 14 states constitutions provide a ballot measure that is automatically put on the ballot asking the voters whether a convention should be convened. This respectively takes place every ten years in five states,94 16 years in Michigan95 and 20 years in eight states.96 In twelve out of those 14 states, a simple majority vote of the People is sufficient to approve the question and Oklahoma’s constitution does not specify which majority is required.97 In contrast, Illinois demands a supermajority vote of 60 per cent of those voting on the question or a majority of those who cast a ballot for any office in the election.98 (2)  Secondly, 27 state legislatures can autonomously decide to place such a question on the ballot to seek ratification by the voters in addition to or instead of an automatic ballot question. There are inter alia divergences with regard to: (a) the percentage of votes in the state legislature required to put the question on the ballot: a simple majority vote in ten states,99 a 60 per cent vote in two states,100 and a two-thirds vote in fifteen states;101 (b) the number of legislative sessions in which the members of the legislature must vote in favour of the measure in order to put it on the ballot: one session in 26 states102 and two sessions in Kentucky; (c) the percentage of citizens required to vote on the ballot in order to approve the question asking whether to organize a proposing convention: a simple majority in 26 states,103 and in Illinois a 60 per cent supermajority of those

94 

Alaska, Hawaii, Iowa, New Hampshire, Rhode Island. Michigan, Art XII S 2 of the state Constitution stipulates ‘At the general election to be held in the year 1978, and in each 16th year thereafter and at such times as may be provided by law, the question of a general revision of the constitution shall be submitted to the electors of the state. If a majority of the electors voting on the question decide in favour of a convention for such purpose, at an election to be held not later than six months after the proposal was certified as approved, …’ (italics added). 96  Connecticut, Illinois, Maryland, Missouri, Montana, New York, Ohio, Oklahoma. 97  Okla, Art XXIV, para 2. 98  Ill Const, Art XIV para 1. 99  Alabama, Hawaii, Iowa, Kentucky, New Hampshire, New York, Rhode Island, Tennessee, West Virginia, Wisconsin. 100  Illinois, Nebraska. 101 California, Colorado, Delaware, Idaho, Kansas, Minnesota, Montana, Nevada, New Mexico, North Carolina, Ohio, South-Carolina, Utah, Washington, Wyoming. 102 Alabama, California, Colorado, Delaware, Hawaii, Idaho, Illinois, Iowa, Kansas, Minnesota, Montana, Nebraska, Nevada, New Hampshire, New Mexico, New York, North Carolina, Ohio, Rhode Island, South Carolina, Tennessee, Utah, Washington, West Virginia, Wisconsin, Wyoming. 103  Alabama, California, Colorado, Delaware, Hawaii, Idaho, Iowa, Kansas, Kentucky*, Minnesota, Montana, Nebraska**, New Hampshire, New Mexico, New York, Nevada***, North Carolina, Ohio, Rhode Island, South Carolina, Tennessee, Utah, Washington, West Virginia, Wisconsin, Wyoming. * In Kentucky, an approval requires a simple majority voting on the proposition in favour for calling a convention and the total number of votes cast for the calling of the Convention needs to be equal to one-fourth of the number of qualified voters who voted at the last preceding general election in this State (KyConst, S 258). ** In Nebraska, a simple majority of the electors voting on the proposition is only sufficient if the votes cast in favour of calling a convention shall not be less than 35% of the total votes cast at the election (Neb Const,Art XVI para 2). 95  In

Direct Democracy and Constitutional Change in the US 359 voting on the question or a majority of those who cast a ballot for any office in the election.104 In addition to an automatic ballot question, Alaska’s Constitution also provides the option for the state legislature to call a constitutional convention, though without further specifying the required percentage of the legislature’s vote, the number of legislative sessions or the percentage of voting citizens required to approve the question.105 (3)  Thirdly, in six states the legislature can call a convention by a supermajority vote without the need of approval of the voters in order to convene the convention.106Additionally, four states107 allow citizens to take the initiative to petition for putting a question on the ballot whether the People want to call a constitutional convention. In those cases, a simple majority vote of the voting citizens is sufficient to approve the call for a convention.108 Besides those 40 states, Article XVIII Section 1 of Oregon’s Constitution states that ‘[n]o convention shall be called to amend or propose amendments to this Constitution, or to propose a new Constitution, unless the law providing for such convention shall first be approved by the people on a referendum vote at a regular general election’. Moreover, Article XXI of Arizona’s Constitution provides that the state legislature cannot call a convention ‘unless laws providing for such Convention shall first be approved by the people on a Referendum vote at a regular or special election …’ This analysis reveals that several states allow multiple methods to propose a call for a convention. In seven states a constitutional convention can be called both through an automatic ballot referral and via action of the state legislature seeking ratification of the voters by putting a question on the ballot.109 In addition to those two options, Montana’s constitution also allows popular initiative to put a question whether to call a convention on the ballot. In Connecticut, an automatic ballot question and a supermajority vote of the state legislature without approval of the voters are the two available paths to call a convention. South Dakota’s Constitution provides the latter option in addition to popular initiative. Besides these states, 32 states only allow one of those options. In addition to the 42 states analysed above, Pennsylvania already organised five conventions, although its state constitution has never stipulated the rules how to call *** Article XVI, Section 2 of Nevada’s Constitution proclaims that a majority of the electors voting at the election need to vote in favour of calling a Convention in order to approve it. Moreover, it states that in determining what constitutes a majority of the electors voting at the election, reference has to be made to the highest number of votes cast at such election for the candidates for any office or on any question. 104 

Ill Const, Art XIV, para 1. Alaska Const, Art XIII, para 2. 106  Connecticut, Georgia, Louisiana, Maine, South Dakota, Virginia. 107  Fla Const, Art XI, para 4; Mont Const, Art XIV, para 2; ND Const, Art III, para 1; SD Const, Art XXIII, para 2. 108  Akhil Amar relies on a non-exclusive reading of Article V of the federal Constitution to defend that US Congress would be obliged to call a proposing convention if a simple majority of American voters so petitioned. Amar (n 35) 1065 (supplementary invoking the First Amendment’s right of the people to petition the Government). 109  Hawaii, Illinois, Iowa, New Hampshire, New York, Ohio, Rhode Island. 105 

360  Jurgen Goossens a constitutional convention. The last convention was held in 1968 during which the current constitution was adopted. Consequently, it seems to be considered as a legal tradition in Pennsylvania that the state legislature can vote to put a call for a constitutional convention on the ballot. Similarly, Massachusetts’ Constitution does not mention the option of a constitutional convention, but four conventions have already been held, the most recent one in 1917–1919. Finally, Vermont’s Constitution does not provide for constitutional conventions, although in 1969 the state legislature voted to put an advisory question to the ballot whether to organise a ­convention. Nevertheless, a majority of the voters answered the question n ­ egatively.110 Arkansas, Indiana, Mississippi, New Jersey, and Texas also do not have a constitutional provision about calling a constitutional convention. Voters principally choose the delegates to the convention. As those delegates are only chosen for the purpose of the convention, a convention is considered to result in more direct democracy than a vote of the state legislature whose members are chosen for a bulk of tasks. The state constitution or the legislator usually prescribe the way in which the convention itself is organised. Afterwards, the proposal of major revisions or the draft of a new constitution by a convention is in principle brought before the voters for ratification, even if it is not explicitly prescribed by the state constitution. Out of the 40 states in which the Constitution explicitly provides how a constitutional convention can be called, 24 states111 explicitly require a simple majority vote of the people for ratification, two states112 require a 60 per cent vote, and New Hampshire requires a two-thirds vote. In the other 13 states,113 the state constitution does not explicitly mention which majority of the voters needs to vote yes in order to ratify the revisions proposed by a proposing constitutional convention. D.  Constitutional Commission Constitutional commissions are in principle established by the (state) legislature for the purpose of recommending constitutional amendments. Such a commission is generally not authorised to put its proposals directly on the ballot. For instance, since amendment 4 (1996), Article XI, para 1 of New Mexico’s Constitution authorises the establishment of an independent commission that can propose amendments to the state legislature for its review. 110  Office of the Vermont Secretary of State, ‘Statewide Referendum. 1969: Constitutional Convention Referenda’, 1, 2 (Vermont Secretary of State, 6 July 2006) accessed 10 April 2015. 111 California, Colorado, Connecticut, Georgia, Hawaii*, Idaho, Illinois, Iowa, Kansas, Louisiana, Maryland, Michigan, Nebraska, New Mexico, New York, North Carolina, North Dakota, Oklahoma, Rhode Island, South Dakota, Tennessee, Utah, Virginia, Wisconsin. * In Hawaii, an approval of a proposed amendment or revision can be achieved by a majority vote of at least 50% of all the votes generally cast in the election, or by at least 30% of all the registered voters in the state at the time of the election in case of a special election (Haw Const,Art XVII,para 2). 112  Minnesota and Florida. 113  Alabama, Alaska, Delaware, Kentucky, Maine, Montana, Missouri, Nevada, Ohio, South Carolina, West-Virginia, Washington, Wyoming.

Direct Democracy and Constitutional Change in the US 361 Florida is the only state in which the Constitution uniquely establishes two constitutional commissions that can directly bring proposed constitutional amendments before the voters without approval of the state legislature.114 The voters can approve those proposed measures with a 60 per cent majority.115 Firstly, the Constitution Revision Commission convenes every 20 years.116 Such a Constitution Revision Commission convened for the first time in 1977. In 2017, such a commission will convene for the third time. Secondly, also once every 20 years, the Florida Taxation and Budget Reform Commission convenes with the authority to directly propose constitutional amendments to the voters regarding taxation and the state budgetary process if two-thirds of the full commission, that is, 18 out of the 25 members, approve it.117 This commission convened for the first time in 2007. In 2008, only one recommendation of the 2007–08 Taxation and Budget Reform Commission made it to the ballot and it was rejected by the voters. IV.  ALTERATIONS TO THE FEDERAL AMENDMENT PROCEDURE

In light of the tradition of state constitutionalism as analysed in the previous chapter, reform proposals will be developed with regard to the federal amendment procedure. First, the proposals will focus on the possibility of introducing initiative petitions and constitutional referenda. Secondly, it will be illustrated that a popular vote is almost inseparably linked with a simple majority voting rule. It is the main aim of this section to revive the important debate about changing Article V of the US Constitution.118 A.  Initiative Petitioning Currently, two-thirds of both chambers of Congress can directly propose amendments to the federal Constitution. One could advocate that citizens should also be given the initiative to propose amendments via petitioning. This proposal would contribute to more direct democracy. Regarding the concrete implementation of this

114  See B Buzzett and SJ Uhlfelder, ‘Constitution Revision Commission: A Retrospective and Prospective Sketch’ The Florida Bar Journal 4 (April 1997) 22, 22; Dinan (n 45) 300, n 7. 115  Fla Const, Art XI, para 5. 116  Fla Const, Art XI, para 2. 117  Fla Const, Art XI, para 6. 118  See, eg, J W Burgess, Political Science and Comparative Constitutional Law: Vol 1 Sovereignty and Liberty (Ginn & Company, 1902) 137, 151–52 (repudiating ‘the artificially excessive majorities’ and suggesting a proposal of amendments by two successive sessions of Congress via simple majority vote and a ratification by a simple majority of state legislatures); WS Livingston, Federalism and Constitutional Change (Clarendon Press, 1956) 248–53 (proposing several proposals to reform the amending procedure); Orfield (n 12) 168–221 (noting numerous proposals made in the 1920s and 1930s to improve Art V); JR Vile, The Constitutional Amending Process in American Political Thought (Praeger, 1992) 137–56 (discussing the views of Progressive Era commentators on the amending procedure); SM Griffin, ‘The Nominee is … Article V’ (1995) 12 Const Comment 171; T Lynch, ‘Amending Article V to Make the Constitutional Amendment Process Itself less Onerous’ (2011) 78 Tenn L Rev 823; Rappaport, n 6) 1509.

362  Jurgen Goossens reform, one could borrow useful insights from numerous state examples regarding petitioning for amendments of state legislation and constitutions. In 18 states voters are allowed to initiate a proposed amendment of the state constitution through petitioning. Those petitions must be signed by a certain percentage of voters, ranging from three to 15 per cent of designated voting groups which differ from state to state. In most states, the threshold is eight or ten per cent of the total votes cast for Governor in the last gubernatorial election. Moreover, nine states have a territorial distribution requirement for the signatures. Consequently, a reform of Article V in order to provide the possibility of petitioning for proposing federal amendments could for example be carried out by imposing a requirement of eight or ten per cent of the total votes cast for Governor in the last gubernatorial election in at least one half of the states. However, this is only one possible proposal. Other proposals are imaginable to implement such a reform for which inspiration can be found in state constitutional tradition. Nevertheless, one has to be wary to impose too strict requirements. In some states the requirements for petitioning are so strict that this option has almost never been successfully used. One could also advocate to introduce more direct democracy by entrusting the initiative to call a proposing national convention to the People. Currently, four states explicitly allow citizens to take the initiative to petition for putting a question on the ballot asking whether the People want to call a proposing constitutional convention to amend the state constitution. This is not an overwhelming number of states, but they may serve as examples for introducing popular initiative to call a national proposing convention. Although Akhil Amar relies on a non-exclusive reading of Article V to defend that Congress would be obliged to call a proposing convention if a simple majority of American voters so petitioned,119 an explicit amendment of Article V seems favourable in this regard. A threshold lower than a majority of the voters also seems recommendable, as it would be more in line with constitutional tradition in the states. First, Florida requires that such a petition is signed by 15 per cent of the electors in each of one half of the congressional districts and of the state as a whole based on the last preceding election of presidential electors.120 Although in Florida more signatures are required to call a proposing national convention than to propose an amendment,121 a simple majority is sufficient to approve the question on the ballot. The latter is less than the required supermajority of 60 per cent on the ballot for a popular initiative proposing an amendment. Secondly, in Montana the petition has to be signed by 10 per cent of the qualified electors of the state, including 10 per cent of the qualified electors in each of two-fifths of the legislative districts.122 Thirdly, Article III, Section 1 of North Dakota’s Constitution allows for calling a

119 

Amar (n 35) 1065. Fla Const, Art XI, para 4. 121  Fla Const, Art XI, para 3: ‘Signed by a number of electors in each of one half of the congressional districts of the state, and of the state as a whole, equal to 8% of the votes cast in each of such districts respectively and in the state as a whole in the last preceding election in which presidential electors were chosen.’ 122  Mont Const, Art XIV, para 2. 120 

Direct Democracy and Constitutional Change in the US 363 constitutional convention without mapping out the rules of the game. Fourthly, in South Dakota the call for a proposing constitutional convention may be initiated by a petition signed by 10 per cent of the qualified voters that cast a vote for Governor in the last gubernatorial election.123 B.  Constitutional Referenda Currently, Congress can choose the required manner and threshold for ratification: approval by three-fourths of the state legislatures or by conventions in three-fourths of the states. Nevertheless, one could consider adding the option of a national convention to ratify proposed amendments. However, the current option of ratifying state conventions seems more advisable, as it brings the debate closer to the People and raises the chance that the interests of the constituents are taken into account. Nonetheless, one should consider opting for more direct democracy with regard to the final approval of constitutional amendments, namely a ratifying referendum.124 In this regard, one can invoke the Preamble of the US Constitution, stating ‘We, the People’ instead of ‘We, the States’, and one can rely on constitutional tradition in the states in order to introduce popular referenda in the federal amendment process. Popular referenda were unknown in 1787, but times have substantially changed.125 Some people may insist on involvement of the state level by state legislatures or state conventions as an essential part of federalism. Nonetheless, it is doubtful that the federalism argument can deprive the People of the right to alter themselves what they have established and ordained. Ratification by referendum is undeniably an indispensable part of constitutional tradition on the state level. All states, except Delaware, require a popular vote after the proposal of constitutional amendments by the state legislature.126 Moreover, only Florida (in principle a 60 per cent vote and two-thirds for a new State tax or fee) and New Hampshire (two-thirds) substantially depart from the state tradition of a majoritarian voting rule for constitutional referenda by requiring a supermajority vote. In 18 states, voters can themselves initiate a proposal to amend the Constitution via petitioning. Once the percentage or number of voters required to sign the petition is achieved, the proposed amendment has to be put on the ballot in all those 18 states for ratification. In 13 states a simple majority of the voters is sufficient. In addition, three states require a certain percentage of votes cast in the election in addition to a simple majority voting yes on the particular amendment. Only Florida and Illinois127 require a supermajority of 60 per cent voting on the proposed amendment. 123 

SD Const, Art XXIII, para 2. Orfield (n 12) 192–203; ibid at 192: ‘The chief proposal for the alteration of the amending process to receive serious consideration in the past two decades has been that for a referendum.’ 125  Orfield (n 12) 192. 126  See above Section III(A)(iii) for more details, including the relevant states and their state constitutional provisions. 127  In Illinois, an approval of a proposed amendment can be obtained by 60 per cent of the people ­voting on the amendment or a majority of those voting in the election (Ill Const, Art XIV, para 3). 124  See

364  Jurgen Goossens Finally, 24 out of the 40 states where the Constitution explicitly provides how a constitutional convention can be called explicitly require a simple majority vote for ratification of proposed amendments, two states require a 60 per cent vote, and New Hampshire requires a two-thirds vote. The other 13 state constitutions do not explicitly regulate which majority of the voters is required. Most of those 13 state constitutions only mention that the proposed amendments ought to be put on the ballot. In the other states, those amendments are in principle brought before the voters even if this is not explicitly required by the state constitution. Bruce Ackerman has argued to explicitly amend Article V of the US Constitution in order to introduce the option of a national referendum. He proposed the ­following amendment process. A President in his/her second term may propose amendments to the US Congress. Then, if two-thirds of both Houses approve the proposal, it requires 60 per cent of the voters in two successive Presidential elections to ratify the amendment.128 This procedure might, however, be too onerous. The constitutional tradition in the States might constitute an argument for introducing popular ratification of proposed amendments with a simple majority of those voting—that is, not of all who are eligible or registered to vote129—in one election.130 Despite Akhil Amar’s defence of a non-exclusive reading of Article V in favour of the right of a simple majority of the People to amend the Constitution, it is recommendable to explicitly amend Article V in order to include direct democracy.131 I disagree with Henry Monaghan’s argument that federalism and the role of the states—as inter alia embedded in James Madison’s notable quote in Federalist No. 39 that ‘[the Constitution is] neither wholly national nor wholly federal’132—should exclude direct democracy via national referenda.133 One could ensure a balanced territorial distribution of the required ratification across the states in order to avoid an unjustified veto power of the smallest states or an unjustified prevalence of the most populous states. One could for example require ratification by a majority in three-fourths of the states or by a number of states representing at least three-fourths of the US voters in the preceding national election.

128  Ackerman (n 39) 54–55; Ph J Weiser, Note, ‘Ackerman’s Proposal for Popular Constitutional Lawmaking: Can It Realize His Aspirations for Dualist Democracy?’ (1993) 68 NYUL Rev 907 (criticising Ackerman’s proposal and advocating for deliberation). 129  Amar (n 35) 1064, n 78. 130  Ackerman would probably not accept this as a threshold for a constitutional moment of higher law-making, because he would probably not consider a vote of the People during one election as a sustained moment of higher law-making. 131  See Amar (n 5) 295–97. 132  The Federalist No 39, at 246 (James Madison) (Clinton Rossiter edn, 1961): ‘If we try the Constitution by its last relation to the authority by which amendments are to be made, we find it neither wholly national nor wholly federal. Were it wholly national, the supreme and ultimate authority would reside in the majority of the people of the Union; and this authority would be competent at all times … to alter or abolish its established government. Were it wholly federal, on the other hand, the concurrence of each State in the Union would be essential to every alteration that would be binding on all.’ 133 Monaghan (n 12) 121 (criticising Akhil R Amar by invoking federalist papers of Madison and Hamilton about the role of the states).

Direct Democracy and Constitutional Change in the US 365 C.  Simple Majority for Popular Votes As shown above, direct democracy in the form of referenda is inherently part of state tradition. Therefore, the latter could inspire to add this option to the federal amendment process and thereby give the Constitution back to the People. Consequently, the question has to be posed whether a simple majority vote would be sufficient or a supermajority vote should be required. Constitutional tradition in the states constitutes an important argument in favour of the former option. It is relevant to look at the threshold required by the state constitutions for the popular vote with regard to: (1)  approval of amendments proposed by the state legislature; (2)  ratification of amendments proposed by popular initiative; (3)  approval of the question asking if a constitutional convention should be organised; and (4)  ratification of amendments proposed by a constitutional convention:134 (1) After an affirmative vote by a state legislature to propose an amendment, the proposal is referred to the ballot in 49 states. State tradition overwhelmingly ­ requires simple majorities for the popular ratifying vote. Only Florida (in principle a 60 per cent vote and two-thirds for a new State tax or fee) and New Hampshire (two-thirds) require a supermajority vote. (2) Eighteen states allow popular initiative of their citizens to propose amendments via petitions. In 13 out of those 18 states a simple majority of the voters is sufficient to approve the proposed amendment. Moreover, three states require a simple majority vote on the proposed amendment and a certain percentage of votes cast in the election. Only two states require a supermajority. Florida requires a supermajority vote of 60 per cent and in Illinois ratification can be achieved by 60 per cent of the votes on the amendment or a majority of voters in the election. Again, constitutional tradition in the states favours simple majority voting rules. (3) 40 state constitutions explicitly prescribe how a constitutional convention can be called. Three different options can be distinguished: an automatic ballot question, initiative by the state legislature, and popular initiative. Firstly, state constitutions in 14 states provide an automatic ballot question at regular time intervals. In 12 out of those 14 states, a simple majority vote of the People is sufficient to approve the question and Oklahoma’s constitution does not mention which majority is required. Only Illinois demands a supermajority vote of 60 per cent of those voting on the question or a majority of those who cast a ballot for any office in the election.135 Secondly, 26 out of the 27 states in which the state legislature can autonomously put a question on the ballot for calling a constitutional convention, require

134  See above Sections III(A)(iii), (B) and (C) for more details, including the relevant states and their state constitutional provisions. 135  Ill Const, Art XIV, para 1.

366  Jurgen Goossens a simple majority vote by the People for ratification. Only Illinois requires a 60 per cent supermajority of the votes on the question or a majority of the voters who cast a ballot for any office in the election. Finally, in four states the question can be put on the ballot through petitioning. In each of those four states, a simple majority vote can approve the call for a convention. (4) 24 out of the 40 state constitutions explicitly prescribing how a constitutional convention can be called explicitly require a simple majority vote for ratification of proposed amendments, while only three states require a supermajority vote. ­Minnesota and Florida require a 60 per cent vote, while New Hampshire requires a two-thirds vote. The other 13 states constitutions do not explicitly prescribe which majority of voters is required to vote yes for ratification. In conclusion, a popular vote is almost inseparably linked with a simple majority voting rule. Akhil Amar has been vigorously defending the right of a simple majority of the People to alter the Constitution.136 Although this may scare some people,137 many US citizens are already subjected to a majoritarian voting rule for ­amendments to their state constitution. Moreover, on the federal level there is a ‘safety in ­numbers’. In contrast, one state is more likely to be dominated by a ­majoritarian view which may entail negative consequences.138 Finally, not only a tyranny of a ­majority, but also a tyranny of a minority can arise, which may block positive and reasonable reforms favoured by a strong majority. V. CONCLUSION

In 1787, there was a lack of useful examples at the state level providing a detailed amendment procedure which provided direct involvement of the People. Moreover, the federal amendment process has never been improved since its adoption 228 years ago. Constitutional tradition in the states has, however, evolved and has generated an extremely rich, though often ignored, source of (experiments with) amendment procedures. For instance, popular referenda were unknown in 1787, but times have substantially changed. Presently, state constitutional tradition clearly champions direct democracy and majoritarian voting rules. The analysis of the 50 state constitutional amendment processes in this contribution could serve as a starting point for a better informed normative debate about constitutional amendment procedures on the state and the federal level. If one reflects on the possible implementation of state experiments on the federal level, one has to take into account the different features of the federal context, such as a higher likelihood of factionalism. Nevertheless, state tradition could be an important argument

136  Amar (n 5) 10–13 and 295–97 (invoking several important speeches, essays and ratification texts at the Founding of, inter alia, James Wilson, Publius, Alexander Hamilton, and James Iredell). 137  Amar (n 35) 1096: ‘the right of a current majority to amend our Constitution may scare you. To be honest, it scares me a little too. However, … the alternative seems even scarier’. 138  Amar (n 35) 1097 (invoking the Federalist Papers).

Direct Democracy and Constitutional Change in the US 367 in favour of adding more direct democracy to the federal constitutional amendment process through the introduction of initiative petitioning and ratifying referenda. A thorough analysis of state constitutionalism shows that a popular vote is almost inseparably linked with a simple majority voting rule. If one would opt for the introduction of a ratifying referendum at the federal level, it is recommendable to ensure a balanced territorial distribution of the required ratifiying vote across the states in order to avoid an unjustified veto power of the smallest states or an unjustified prevalence of the most populous states. VI.  ATTACHMENT: METHODS FOR CONSTITUTIONAL AMENDMENT PROVIDED BY THE STATE CONSTITUTIONS Proposal by State Legislature

Popular Initiative

Constitutional Convention

ALABAMA

X

ALASKA

X

ARIZONA

X

X

ARKANSAS

X

X

CALIFORNIA

X

X

X

COLORADO

X

X

X

CONNECTICUT

X

X

DELAWARE

X

X

FLORIDA

X

GEORGIA

X

X

HAWAII

X

X

IDAHO

X

X

ILLINOIS

X

INDIANA

X

IOWA

X

X

KANSAS

X

X

KENTUCKY

X

X

LOUISIANA

X

X

MAINE

X

X

MARYLAND

X

X

MASSACHUSETTS

X

X

MICHIGAN

X

X

MINNESOTA

X

MISSISSIPPI

X

Constitutional Commission

X X

X

X

X

X

X

X

X X

X (continued)

368  Jurgen Goossens Proposal by State Legislature

Popular Initiative

Constitutional Convention

MISSOURI

X

X

X

MONTANA

X

X

X

NEBRASKA

X

X

X

NEVADA

X

X

X

NEW HAMPSHIRE

X

NEW JERSEY

X

NEW MEXICO

X

X

NEW YORK

X

X

NORTH CAROLINA

X

NORTH DAKOTA

X

X

X

OHIO

X

X

X

OKLAHOMA

X

X

X

OREGON

X

X

X

PENNSYLVANIA

X

RHODE ISLAND

X

SOUTH CAROLINA

X

SOUTH DAKOTA

X

TENNESSEE

X

TEXAS

X

UTAH

X

VERMONT

X

VIRGINIA

X

X

WASHINGTON

X

X

WEST VIRGINIA

X

X

WISCONSIN

X

X

WYOMING

X

X

X

X

X X X

X X X

Constitutional Commission

Conclusion The Emergence of Comparative Constitutional Amendment as a New Discipline Towards a Paradigm Shift XENOPHON CONTIADES AND ALKMENE FOTIADOU*

Begin with an individual, and before you know it you find that you have created a type: begin with a type, and you find that you have created—nothing. F Scott Fitzgerald

I.  A PARADIGM SHIFT IN COMPARATIVE CONSTITUTIONAL CHANGE

A

PARADIGM SHIFT has recently occurred in the study of constitutional change. A fundamental change in the basic concepts and the methodology through which constitutional amendment is approached is becoming conspicuous. A few years back, constitutional change was understood at the national level, mainly through the study of the amendment formula and the rigidity level it aimed to establish. The growth of comparative constitutional law, the focus placed on constitutional change encompassing both formal and informal change, the application of quantitative research methodologies, and the experimental attempts to enhance popular participation in formal constitutional amendment triggered by the financial crisis mark a move away from the traditional paradigm. The shift in the study of comparative constitutional amendment has the characteristics of a scientific revolution as described by Thomas Kuhn.1 Constitutions are understood through the way they change and constitutional change is understood

*  The authors would like to thank the participants in the Comparative Constitutional Amendment workshop held in May 2015 in Boston. The exchange of ideas during this workshop has been an inspiration to us. 1  Th S Kuhn, The Structure of Scientific Revolutions, 4th edn (University of Chicago Press, 2012).

370  Xenophon Contiades and Alkmene Fotiadou through different perceptions of the Constitution. New questions have emerged and older ones are revisited through novel conceptual lenses: —— Do formal amendment rules matter or not? —— Can the routes of constitutional change be pre-determined through constitutional design? —— Does constitutional rigidity have a precise meaning? —— What are the restraints imposed on the amending power and how are they enforceable? Are all types of restraints permissible? —— What is the relationship between the constituent power and the amending power? —— How does the notion of ‘the people’ relate to constituent power and to constituted power? —— How do constitutional revolutions relate to the divide constituent/constituted power? —— How do constitutions relate to the passage of time? How does constitutional change occur in the face of crises? —— What methodology is best suited for approaching constitutional change? Empirical, comparative, conceptual, culturalist approaches compete or may also complement one another in the study of constitutional amendment. —— Which are the functions of constitutional amendment and which are the functions of formal amendment rules? —— Is the assessment of constitutional quality and performance possible and under what criteria? Are transnational or idiosyncratic criteria applicable for such an endeavour? Constitutions are much like houses. Houses can be bought, inherited, or built according to one's needs, but also rent as a permanent or temporary solution. Houses are turned into homes through sentiment. Homes respond to the basic needs of the inhabitants, they provide shelter and protection, while they may also satisfy more complicated requirements. Different types of houses are built for different climates, nonetheless some basic features can be found in all houses, whereas some design elements have gradually become universal. Constant maintenance is important for homes, but also structural renovations may prove necessary when the needs of the inhabitants change. Refurbishing or even small improvements may prove useful in making home life more pleasant. The creation of a bond between a home and its inhabitants is a complex process. How one came to live in a house may be important, pre-existing ties may exist, or inhabitants may have taken part in the original architectural design. Functionality of the home is equally important, inhabitants may easily love a home that responds to their needs, while they may feel discomfort when it fails to do so, or it creates problems. A sense of permanence is important with regard to homes. Moving home is always a traumatic experience. Still, every house is a work in process. It requires care and maintenance to remain functional, repairs when necessary and adaptation to the changing needs of the inhabitants. What has to be kept in mind is that not all innovations are successful, and who decides on what needs to be improved is also crucial, while there may be conflicting views or needs to be addressed.

Conclusion 371 Amendability has become a basic feature of constitutions, signifying but also building constitutional identity. How (un)amendable a constitution is, is a focal point for understanding not only how it was conceptualised by its framers, but also the way it is applied in practice, competing theories of constitutional interpretation, and the role of political and constitutional actors involved in constitutional change.2 The amendability factor plays a key role in understanding why, how and by whom constitutions change. The route toward the paradigm shift has been a ‘there and back’ journey. The focus on formal amendment rules gave way, and attention turned toward informal constitutional change, followed in turn by the comeback of the study of formal amendment rules, enriched by the scholarship that questioned their importance.3 Empirical and conceptual analysis of constitutional amendment are at a point of converging to the recognition of the importance of cultural elements that shape the channels of formal and informal constitutional change.4 Part of the unfolding paradigm shift is the combination of competing methodologies to approach the phenomenon of constitutional change. Universalist and culturalist approaches, normative and empirical analysis are gradually becoming compatible and used in conjunction. II.  THE AMENDABILITY FACTOR: UNDERSTANDING CONSTITUTIONS THROUGH THEIR CHANGE

Formal amendment rules prescribe the desired level of constitutional amendability. Mechanisms aimed at controlling the relationship of the Constitution to the passage of time and delineating the role of the actors involved in constitutional change are set up through a combination of procedural and material limits. Amendment formulas encapsulate how constitutional change is envisaged by the constitutional designer, which is a good starting point for understanding how close to the envisaged modus of constitutional amendment constitutional change is in reality.5 The distance or the proximity of the prescribed amendability level with the actual amount and quality of constitutional change that occurs depends on institutional and factual hindrances and facilitations to constitutional change. The motto ‘stability through change’ is multifaceted and depends on what changes and what remains

2  X Contiades and A Fotiadou, ‘Models of Constitutional Change’ in X Contiades (ed), Engineering Constitutional Change. A Comparative Perspective on Europe, Canada and the USA (Routledge, 2013) 417–68. 3  D Strauss, ‘The Irrelevance of Constitutional Amendments’ (2001) 114 Harv L Rev 1457, 1487; BP Denning and JR Vile, ‘The Relevance of Constitutional Amendments: A Response to David Strauss’ (2002) 77 Tul L Rev 247; R Albert, ‘The Structure of Constitutional Amendment Rules’ (2014) 49 Wake Forest L Rev 913. 4  See, eg, D Oliver and C Fusaro (eds), How Constitutions Change: A Comparative Study (Hart Publishing, 2011); T Ginsburg and J Melton, ‘Does the Constitutional Amendment Rule Matter at All? Amendment Cultures and the Challenges of Measuring Amendment Difficulty’ (2015) 13(3) Int’l J Const L 686; X Contiades and A Fotiadou, ‘The Determinants of Constitutional Amendability: Amendment models or amendment culture’ (2016) 1 EuConst 192. 5  Contiades and Fotiadou (n 2) 418.

372  Xenophon Contiades and Alkmene Fotiadou the same, and on how and why change happens. Is there such thing as a fixed amendability level? Is amendability in the eye of the beholder? Vicky Jackson recently described the unamendability of the US Constitution as a myth.6 Jackson challenges the prevailing perception of the US amendment formula. According to this view, difficulty of formal change may be exaggerated. This puts forth the idea that the level of constitutional rigidity may turn into a self-fulfilling prophecy. Political and constitutional actors regard a specific level of rigidity as a given and tend to reproduce it. The relevancy of formal amendment rules to the way constitutions change has been constantly questioned. At the one side of the spectrum lies the idea that the starting point for understanding how a constitution changes is the amendment formula. At the far end of the spectrum is the belief that formal amendment rules have no impact at all on constitutional change. In between lie understandings of constitutional change based on both the amendment formula and the parameters that impact its application. A further question is whether other functions served by formal amendment rules are correlated to the competence of amendment rules to channel constitutional change. In case, for instance, formal amendment rules are irrelevant to constitutional change, does this impact the way those rules distinguish the Constitution as superior to other legislation? Can formal amendment rules perform functions beyond the regulation of constitutional change regardless of their failure to impact how such change unravels? In case this syllogism is taken one step further it could mean that constitutional designers should give up the aspiration to control how constitutions change through the design of amendment rules. It would be nonetheless paradoxical for designers to continue drafting formal amendment rules only for the sake of the remaining functions after excluding the regulation of constitutional amendment. The variety of amendment formulas in different constitutional orders is astonishing. A wide range of different combinations of procedural and material limits indicate that amendment rules are a topical constitutional matter resistant to transplants. Nonetheless, it is also the case that often the amendment formulas in place in distinct constitutional orders may have profound similarities and yet correspond to completely different profiles of constitutional change. Thus, although constitutional change is constantly subject to comparative assessment, becoming thus a distinct feature of comparative constitutional law, there is a paradoxical relationship between the idiosyncratical features of constitutional amendment and the study of comparative constitutional change. Approaching the conundrum of comparative constitutional change leads to a better understanding of distinct conceptualisations of the Constitution. The amendability factor revisits the rigidity/flexibility dipole, incorporating the interaction between the amendment formula and external sources facilitating or hindering constitutional change. At the two ends of the spectrum of control over constitutional change lie opposite beliefs. Constitutional change may be viewed as totally resistant to being shaped through constitutional design. Adversely

6  VC Jackson, ‘The (myth of un)amendability of the US Constitution and the democratic component of constitutionalism’ (2015) 13(3) Int’l J Const L 575.

Conclusion 373 formal amendment rules are understood as the basic mechanism of constitutional change dictating the rigidity level of the Constitution. The very need for change is also a matter of controversy linked to the functions of the Constitution. Beyond the Jefferson—Madison dilemma lie questions about the inevitability of change and the different forms and shapes it takes. What amounts to constitutional change and whether such change is a process of perpetual amelioration are important questions that appear to be perpetually reproduced and rephrased. The terminology used is illuminating in itself: perfection and imperfection, functionality and dysfunctionality, endurance and resilience correspond to different notions of the Constitution and its functions. Different terminology is used to express degrees of affection towards the Constitution, as opposed to subtle criticism. There is an air of nobility in finding the Constitution imperfect, which suggests constitutional faith and devotion.7 There is wisdom in the acceptance of imperfection, including constitutional imperfection. However, all kinds of amendment aimed at perfecting the Constitution involves a conceptualisation of constitutions as subject to perpetual amelioration. The ­constitution-perfecting task has an honourable goal, yet it is a fundamentally interpretative strategy targeted to judge-made informal change.8 Judges are well suited for giving life to abstract notions through moral principles. Even when moral principles are not involved in the pursuit of the desired judicial outcome, constitutional battles can be fought through primarily legal weaponry. Things are different in constitutional orders where the primary route of constitutional change is formal amendment, or where a dialogical model between judge and constitutional legislation is prevalent.9 Constitutional amelioration is a goal dependent on multiple variables. To begin with, formal amendment processes involve political elites and the electorate. Passing an amendment requires compromises that impact on the final outcome. The text itself is a product of such compromises and may be influenced in various directions. Compromises may lead to a more abstract text, which may prove easier to pass. Adversely, compromises may lead to a lengthier text if bargaining entails adding extra parameters.10 Improvement from the aspect of constitutional and political choices is anything but value-neutral. A different aspect of quality is the degree to which a given amendment achieves its stated goal, that is, the result of formal constitutional change

7  JE Fleming, Fidelity to Our Imperfect Constitution: For Moral Readings and Against Originalisms (Oxford University Press, 2015). 8  JE Fleming, ‘Comment on Amendment-Metrics: The Good, the Bad and the Frequently Amended Constitution’, in this volume. 9  See R Dixon and A Stone, ‘Constitutional Amendment and Political Constitutionalism: A philosophical and comparative reflection’ in D Dyzenhaus and M Thorburn (eds), Philosophical Foundations of Constitutional Law (Oxford University Press, 2016) 95. Dixon and Stone consider one of the basic functions served by amendment procedures that they ‘provide a means for overriding judicial interpretations of existing provisions of the constitution’. From a comparative aspect this function is basic where the judge is a lead actor in constitutional change and where a dilogical communication between judge and constitutional legislator is part of the constitutional culture. 10  HK Gerken, ‘The Right to Vote: Is the Amendment Game Worth the Candle?’ (2014) 23 Wm & Mary Bill Rts J 11.

374  Xenophon Contiades and Alkmene Fotiadou corresponds to the purpose sought. Formal amendment rules may foster such change or may render it difficult to achieve. It is through this lens that the substantive content of constitutional change can be evaluated. Does the content of a specific amendment impact the process, having the power to disturb established patterns? And if so, to which degree? A difficult to change constitution may demonstrate flexibility by allowing an amendment which is deemed important.11 III.  NEW TAXONOMIES OF CONSTITUTIONAL CHANGE

The possibility of ‘unconstitutional’ amendments has renewed interest in the boundaries between constituent and constituted power.12 What are the prerequisites of revolutionary constitutional change, that is, the criteria for discerning this particular kind of change from a plain violation of the Constitution by an unconstitutional amendment? How relevant is the success of a successful constitutional transformation to discern awakenings of the constituent power as opposed to misuses of the constituted power?13 Although it is possible to elaborate criteria based on the legitimacy of the attempt, demanding not only constituting but also continuing legitimacy,14 they offer little help in real time, that is at the moment when the contested episode unfolds. Deferral until a future time of the evaluation of whether non-adherence to the amending formula constitutes a democratic revolution or violates it could erode the functions of the formal amendment rules. A paradox exists in the trend to revisit conceptualisations of constituent power. On one hand there is a yeaning in constitutional theory to consider constituent power a perpetual possibility, while on the other the notion of the constituent power is being deconstructed.15 Non-derived constituent power retains a revolutionary tint, and the reluctance to let go of the idea of revolutions as a means of constitutional change16 expresses the agony to preserve a permanent link between the people and the Constitution. Such narratives are based on memories of constitution-making linked to the national state of the 19th century. Ongoing constitution-making dictates observing

11  For example in Spain where the Constitution is extremely difficult to change the balanced budget rule was passed with remarkable speed as a response to the financial crisis. See AR Robledo, ‘The Spanish Constitution in the Turmoil of the Global Financial Crisis’ in X Contiades (ed), Constitutions in the Global Financial Crisis: A Comparative Analysis (Ashgate, 2013), 141–65, 154–57. 12 Y Roznai, ‘Unconstitutional Constitutional Amendments: A Study of The Nature and Limits of Constitutional Amendment Powers’ (DJur thesis, Department of Law of the London School of Economics and Political Science, 2014) accessed 1 September 2016. 13  See M Tushnet, ‘Peasants with Pitchforks, and Toilers with Twitter: Constitutional Revolutions and the Constituent Power’ (2015) 13(3) Int’l J Const L 639; J Cormarek, ‘Constitutional revolutions and the constituent power: a reply to Mark Tushnet’ (2015) 13(4) Int’l J Const L 1054. 14  R Albert, ‘Democratic Revolutions’ (2011) (unpublished manuscript) (on file with author). 15  See Z Oklopcic, ‘Constitutional Theory and Cognitive Estrangement: Beyond Revolutions, Amendments and Constitutional Moments’, in this volume. 16 See A Hutchinson and J Colón-Ríos, ‘Democracy and Revolutions: An Enduring Relationship?’ (2012) 89(3) Denver University Law Review 593.

Conclusion 375 the exercise of constituent power in progress, which brings forth departures from the normative conceptualisation of the pouvoir constituant.17 The notion of constitutionalism at the transnational and post-national levels is being redefined and reinvented. A multiplicity of constituent powers18 is replacing the pre-existing monolithic conceptualisation of constituent power. A variety of constitutionalisms is gradually becoming acceptable, leading to novel taxonomies and rendering reference to one exclusive notion of constitutionalism with a fixed content difficult, if not impossible.19 The study of the derived amending power suggests the notion of constitutional rigidity must also be replaced by the notion of multiple rigidities.20 This passage from one-dimensional notions to multiplicities is a symptom of the unfolding paradigm shift. Distinct schools of thought have emerged in the study of constitutional change. Distinct in their approaches, they share one common feature, which is focus of the research on constitutional change. The process, nature and content of change are central in the study of the Constitution. Constitutional change is treated as a magic mirror able to reveal the true face of the Constitution. Within this trend rival approaches exist. A first major question relates to the limits of comparability. Beyond the universalist-culturalist divide and the dilemmas it sets in general in the study of constitutional law there is a question with regard to the prerequisites of comparability.21 Can the study of constitutional change be conducted in a like manner in established democracies, transitional situations, and authoritarian regimes? Is constitutionalism a minimum prerequisite, or does the new trend of distinguishing between different forms of constitutionalism22 allow cross-cultural analysis? A second, and to a certain degree interrelated issue, has to do with the growing use of quantitative empirical methodology in the study of constitutional amendment, especially from a comparative aspect.23 Three problems emerge: (a) it is not self-evident that empirical methodology is always suitable for explaining comparative constitutional amendment, (b) it remains open whether empirical methodology is rival or complementary with more traditional methods of constitutional analysis and how this can be achieved, and (c) thematic-specific problems related to empirical methodology could compromise results in specific areas of study. The underlying question is whether there is not much constitutional designers can put under control24 or whether they can improve their results by performing an impact assessment of the amending formula under construction or reform. Interestingly enough,

17 N Krisch, ‘Pouvoir constituant and pouvoir irrirant in the postnational order’ (2016) 14 Int’l J Const L 657; M Kumm, ‘Constituent power: Constitutional constitutionalism and post-positivist law’ (2016) 14 Int’l J Const L 697. 18  Krisch (n 17) 662. 19  M Tushnet, ‘Varieties of constitutionalism’ (2016) 14 Int’l J Const L 1. 20  Cotiades and Fotiadou (n 2) 458. 21  See R Hirschl, Comparative Matters (Oxford University Press, 2014) 194. 22  Tushnet (n 19). 23  D Law, ‘Constitutions’ in P Cane and H Kritzer (eds), The Oxford Handbook of Empirical Legal Research (Oxford University Press, 2012) 379–98. 24  Ginsburg and Melton (n 4); Contiades and Fotiadou (n 4) 192; L Epstein and A D Martin, An Introduction to Empirical Legal Research (Oxford University Press, 2014).

376  Xenophon Contiades and Alkmene Fotiadou empirical scholarship has started to question the classic countermajoritarian problem as non-existent, suggesting new understandings of judicial review and politics.25 Although it would be risky to totally dismiss the countermajoritarian dilemma and all constitutional scholarship that goes with it as irrelevant, still attention must be paid to detecting who the drivers of constitutional change are. Similar comparability problems are obvious in attempts to create a taxonomy for constitutional change. What are the minimum shared characteristics and commonalities that allow categorisation, is an open question. Detecting diverse sources of amendment difficulty is important in order to discern patterns of constitutional change based on the matrix of parameters that affect it and construct models accordingly. Such taxonomies can only be successfully created if all factors that impact constitutional change are taken into account. In case one or several important parameters are left out, the whole image may be distorted. Institutional and factual drivers and barriers to constitutional change are in intermittent interaction. The rapport between them, being a dynamic relationship, is in constant flux, thus requiring continuous monitoring. An effective taxonomy depends on understanding this particularity of constitutional change. Even more important is that any attempt at intentionally inducing change in the channels of constitutional amendment (including altering the formal amendment rules) requires taking this fluidity into consideration. Constitutional orders may gradually or abruptly move to a different model, hybrid combinations exist, while modelisations of constitutional change are polymorphic tools that can be shifted to provide better insight into the complexity of constitutional amendment.26 Still, understanding constitutional change through the classification in various types and models is one of the basic tools approaching comparative constitutional amendment.27 Underlying comparability are questions about culture. How culture-dependent is constitutional change? According to Jack Balkin, a constitutional culture features many subcultures and may always produce a new equilibrium and a ‘new constitutional common sense’.28 Being dynamic it is comprised of different views that change over time, and compete for dominance. The use of culture is crucial to explain constitutional amendment. Nonetheless the way culture may be used in the study of constitutional change is not predetermined. Empiricists who use culture as a variable seek to find ways to measure culture. Conceptual analysis of constitutional change entails understanding formal amendment rules in relation to the context in which such rules operate. Whether subcultures exist and how they impact on formal amendment also remains open. Judicial culture, referendum culture, a dialogical

25 

See Law (n 23) 385–86. Contiades and Fotiadou (n 2). 27  M Rosenfeld, The Identity of the Constitutional Subject: Selfhood, Citizenship, Culture and Community (Routledge, 2009); R Dixon, ‘Amending Constitutional Identity’ (2012) 33 Cardozo L Rev 1847; Contiades and Fotiadou (n 2). 28 See J Balkin, Constitutional Redemption: Political faith in an unjust world (Harvard University Press, 2011) 178. 26 

Conclusion 377 culture characterised by the communication between lawmaker and judge, the political culture, and so on, impact constitutional change in various ways. Microcultures interact, while it has also been suggested that it is possible to pin down distinct amendment cultures.29 A different trend relates to the predictability of constitutional change. The study of constitutional change is thought to enable predictions.30 Constitutional predictions however are a slippery road. Does the study of amendment formulas allow predicting how change shall unravel? Constitutional change is dependent on extraconstitutional developments. It may be induced as a response to crises, it may result from unexpected changes in the political landscape, or it may express changes in the equilibrium between the drivers of change. Can a thorough understanding of the routes of formal and informal constitutional change in a given legal order ensure the predictability of constitutional change? This question is self-reflective, probing the reasons for concentrating on the study of constitutional change. The feasibility of predictions about constitutional change is not the same with the endeavour to engage in impact assessment of formal amendment rules. How the amendment formula will behave in future scenarios and even under stress is distinct from predicting constitutional change. IV.  SYMBOLISM AND FUNCTIONALITY OF FORMAL AMENDMENT RULES

Formal amendment rules are multi-functional.31 They are about values as much as they are about process. They set up mechanisms of formal constitutional change and convey meanings about what the Constitution is. Michael Rosenfeld has explored the idea that the identity of the constitutional subject may be formed through formal amendment procedures.32 Inherent in formal amendment rules is the memory of the constituent moment. A semantic memory of the constituent episode, but also an implicit memory without awareness, is embedded in the wording of each amendment formula. Aimed at truthful, representative conceptualisations of the constituent moment formal amendment rules create a bridge between the past and the present. In that sense amendment rules are crucial in creating constitutional narratives, shared representations of the constitutional myth. Dependent on the use of eternity clauses and levels of rigidity, amendment formulas determine how quasi-religious or mundane a constitution is, and may render it an object of worship. Worship is distinct from other feelings toward the Constitution. Bonding with the Constitution may also depend on functionality and the ability of a constitution to deliver the goods. Formal amendment rules may often seek to simulate the consensus as well as the importance of the original constituent moment—rendering the

29 

It is also suggested that these cultures may be measurable. See Ginsburg and Melton (n 4) 686–713. Fusaro and D Oliver, ‘Towards a Theory of Constitutional Change’ in D Oliver and C Fusaro (eds), How Constitutions Change: A Comparative Study (Hart Publishing, 2011) 431–32. 31  R Albert, ‘The expressive function of constitutional amendment rules’ (2013) 59 McGill L J 225. 32  Rosenfeld (n 27) 1; Dixon (n 27) 1847. 30  C

378  Xenophon Contiades and Alkmene Fotiadou amendment process more ritualistic. Complexity of the amendment formula may correspond to a representation of the amending power as a reflection of the constituent power. However the two powers, although perpetually correlated, serve different functions. It is the more down to earth mundane functions of the amending power that allow the text of the Constitution to adapt to the changes of an evolving world enhancing constitutional resilience.33 The symbolic function of formal amendment rules is not incompatible with other functions, but it is not necessarily harmonious either. The question is whether it undermines the efficacity of the amendment formula in channeling formal constitutional change. Amendment formulas create mechanisms of formal amendment through the combination of procedural and material limits, and the allocation of roles to the players involved in constitutional change. The goal of achieving some level of rigidity is selfevident in entrenchment. Constitutional rigidity however may also be a side-effect, the unexpected consequence of dysfunctionalities caused by faulty mechanisms. The achievement of stability through change is the holy grail of formal amendment rules. Detached from practicability concerns, rigidity may become a smug notion aimed at safeguarding the myth of the Constitution, while constitutional change follows informal paths. Although there is a growing feeling that constitutions have entered a demystification era,34 entrenchment and rigidity are still more central to the conceptualisation of the Constitution than functionality. Understanding formal amendment rules consists of three steps. The first step entails pinning down the aim of the amendment formula, detecting the intended level of rigidity pursued, as well as the equilibrium between constitutionalism and democracy sought. The second step entails identifying the means used to achieve this aim, that is, the exact combination of eternity clauses, tiered procedures, super-majorities and enhanced majorities, veto-players, temporal limitations including mandatory time lapses, etc. The third step entails placing the aims of the amendment formula and the means it lays out to achieve them within the political and constitutional context in which they operate.35 These three steps are interrelated: the aim of the amendment rules may be read in light of their context, whereas the mechanism created can be evaluated in connection to the purpose and/or to their working environment. The working environment, within which the mechanisms put in place by formal amendment rules operate, is a mosaic composed of combinations of constitutional culture and identity, political ethos and practices, and extra-constitutional identities linked to history, politics, religion, economic factors etc.36 Within a changing environment the role of players

33  X Contiades and A Fotiadou, ‘On Resilience of Constitutions. What Makes Constitutions Resistant to External Shocks?’ (2015) 9 ICL-Journal 3. 34  Contiades and Fotiadou (n 2) 463; M Vesteeg and E Zackin, ‘Constitutions Un-entrenched: Toward an Alternative Theory of Constitutional Design’ (2016) Am Polit Sci Rev (forthcoming) accessed 1 September 2016. 35  R Albert, ‘Temporal limitations in Constitutional Amendment’ (2016) 21 Review of Constitutional Studies (forthcoming). 36  Contiades and Fotiadou (n 2) 435–441.

Conclusion 379 may change as well. Transformations of the political system followed by power shifts affect the drivers of constitutional change. What emerges is that there is no such thing as one-dimensional constitutional rigidity. Multiple institutional and factual sources of rigidity exist, and their interaction results in the way a constitution responds to demands for change.37 Reasons for change are themselves multi-faceted. The need for constitutional adaptation and continual renewal of democratic legitimacy dictate the constant revisiting of the Constitution, unexpected shocks and crises may bring calls for change, flaws in constitutional design may require correction and political changes may trigger constitutional amendments while political reasons that lie outside the realm of constitutional necessity may also result in constitutional change.38 Attitudes toward constitutional change are shaped to a considerable extent through the application of formal amendment rules. Experiments in constitutional design are not a likely option where formal hurdles render constitutional amendment difficult to happen. Constitutional disuse may also appear more easily where removing obsolete or dysfunctional provisions is too difficult. Drivers and barriers to constitutional amendment besides the way constitutions respond to the passage of time, also determine their reactions to multiple stimuli. Formal amendment rules are the basic institutional rigidity, that is, institutionally embedded way of increasing amendment difficulty.39 They create formal amendment mechanisms whose operation is influenced by external factors, while they coexist with mechanisms of informal change. Formal amendment rules are thus the basic ingredient of amendment practice, but cannot produce result by themselves. To uncodify constitutional change a three-partite relationship must be analysed: mechanisms, actors and constitutional culture determine how change takes place in a particular context and as a reaction to particular stimuli. Unravelling the puzzle of comparative constitutional amendment through the study of its multiple facets creates a challenge to revisit stereotypes about constitutional entrenchment, rigidity and longevity. Recent scholarship on constitutional amendment detects a shift in the way contemporary constitutions draw legitimacy and exert their appeal. Efficiency becomes more crucial than mystification, constitutional sanctity is being replaced by constitutional efficacity. Empirical findings seem to verify this phenomenon. Entrenchment gives way, no longer being a basic feature of democratic constitutions that are subject to frequent amendment, as the ability to respond to changing circumstances acquires more value.40 V.  CONSCIOUS CONSTITUTIONAL DESIGN: DRAFTING AMENDMENT RULES

Conscious constitutional design takes place when drafters make deliberate choices having awareness of the environment in which constitutional provisions are expected 37 ibid.

38  See DM Okubasu, ‘The Implication of Conflation of Normal and Constitutional Politics on Constitutional Change in Africa’, in this volume. 39  Contiades and Fotiadou (n 4) 464. 40  Vesteeg and Zackin (n 34) 1.

380  Xenophon Contiades and Alkmene Fotiadou to operate. Formal amendment rules can be the result of conscious constitutional design when drafters intentionally create mechanisms of constitutional change aiming to enable specific functions, fully aware of the legal cultural and political environment. Conscious constitutional design consists of two tiers: intension and awareness. An even more sensitive demand for constitutional designers is the amendment of formal amendment rules. Are amendment formulas open to improvement or correction after they have been tested in practice, if they are not themselves explicitly entrenched? Designers drafting formal amendment rules choose from a toolkit containing a wide variety of procedural and material requirements that may be combined in various ways. Through their combination drafters create mechanisms of varying complexity, which interact with their context producing different degrees of rigidity. Beyond accomplishing the desired rigidity level, that is the combination of stability and change aimed at by the amendment formula, formal amendment rules also serve other functions, such as symbolic functions, which may also be intentionally pursued. The growing study of comparative constitutional change can help choose from that toolkit by providing knowledge about how distinct mechanisms have operated in specific environments and how they impacted on constitutional change. Conscious design does not instantly result in successful design. Much depends on the context, and the environment in which constitutional change occurs is fluid. This has led to the argument that there is probably not much the designers of formal amendment rules can put under their control.41 A different approach embraces volatility by placing emphasis on constitutional resilience, that is, the ability of the Constitution to withstand external shocks.42 By creating mechanisms of constitutional change amendment formulas are constitutional resilience mechanisms, because they allow constitutions to adapt in the face of unexpected crises. Conscious design does not underestimate cultural parameters that shape constitutional change nor the susceptibility of such change to the influence of extra-­ constitutional factors including unexpected crises. By contrast, conscious design places emphasis on cultural factors. Taking into consideration constitutional and political history, constitutional, political and judicial culture is a fundamental part of conscious design. Culture-consciousness helps avoid unsuccessful transplants, opens up the possibility of successful ones and helps avoid constitutional disuse by ­minimising the possibility of unworkable amendment rules. The three steps of understanding constitutional change, that is, pinning down its aims, exploring the context in which amendment mechanisms shall operate and choosing appropriate mechanisms, can be used as a practical guide to conscious design of amendment formulas. Aim-setting rests on the realisation that entrenchment is distinct from rigidity. Entrenchment aims to create a degree of rigidity, which is dependent upon extra-formula factors. Approaching constitutional change as a

41  42 

Ginsburg and Melton (n 4). Contiades and Fotiadou (n 33).

Conclusion 381 system of interactions43 dictates that drafting amendment rules aims at creating levers for change, that may operate in different ways under different circumstances, rather than formulas that can produce the same result every time they are put to use. Secondary aims are also important. Amendment rules are a multi-functional tool, whose multiple utilities define constitutional meaning. The interaction of these rules with the whole constitutional text is an important consideration—drafting of constitutional provisions is necessarily different under formal amendment rules that aim for stringency. Trial and error must be ruled out, phenomena of constitutional desuetude44 may occur more often, and the degree of compromise for affording constitutional protection to novel rights or principles may be heightened. The costs of entrenchment are also thus dependent upon formal amendment rules. As part of conscious constitutional design of formal amendment rules numerous characteristics of the constitutional document are revisited. Constitutional length and the degree of detail of constitutional provisions are explored with regard to their interaction with the manner in which formal constitutional change is materialised. It has been suggested that constitutional detail is per se a form of entrenchment.45 The aims of formal amendment rules are achievable when the mechanisms created to channel constitutional change and to tame the actors involved match the constitutional and political context. The conscious design of all constitutional provisions entails awareness of the available constitutional change mechanisms. Another aspect of consciousness is the growing awareness that formal amendment rules embody the link between constituent and constituted power. Who changes the Constitution is a permanent underlying question. Amendment formulas on one hand respond to an abstract and elusive concept of the people, which lies at the heart of the permanent constituent/constituted power dipole, while on the other hand regulate the behaviour of competing actors of constitutional change. The correlation between the remote transcendental constituent power46 and the tangible protagonists of constitutional change produces a perpetual interplay between the imagery and the reality of constitution-writing. Could it be that the sovereign is he who decides on constitutional change? In modern constitutional states it is difficult to discern the sovereign, not at least in the absolute, indivisible notion conceived by classical theory of sovereignty. Sovereignty appears fragmented, allocated by the Constitution between multiple actors. Still, sovereignty has an intrinsic tendency towards concentration into the hands of one sole actor, apparent despite its institutional allocation. Fragmentation of sovereignty does not correspond to an image of multiple actors taking multiple simultaneous decisions on the same issue, but

43  A Vermeule, The System of the Constitution (Oxford University Press, 2011); Jenna Bedmar, ‘Constitutional Systems Theory: A Research Agenda motivated by Vermeule, the System of the Constitution and Epstein, design for liberty’ (2012) 48 Tulsa L Rev 325. 44  R Albert, ‘Constitutional Amendment by Constitutional Desuetude’ (2014) 62 Am J Comp L 641. 45  Vesteeg and Zackin (n 34). 46  See A Kalyvas, ‘Popular Sovereignty, Democracy, and the Constituent Power’ (2005) 12(2) Constellations 223.

382  Xenophon Contiades and Alkmene Fotiadou signifies instead that there is no sole actor who gets to decide on all issues.47 Constitutional change is an ideal scene for battles over sovereignty to be fought, taking the form of competition over the final say in constitutional change. Conscious design is even more crucial when amending amendment formulas. Although the design of formal amendment rules during the enactment of a constitution is considered crucial, the issue of amending formal amendment rules is less explored.48 To begin with, more stress has been placed on why and how to entrench such rules than on the possibility of improving them after they have been tested through application, correcting them in case they prove dysfunctional, or updating them in case their working environment has changed. Although the dangers that political actors may seek to exploit the possibility of changing formal amendment rules must be acknowledged,49 it is also true that amending them may also prove invaluable for reclaiming constitutional change from players who take advantage of shortcomings in the formula to channel way outside it, or to hinder necessary constitutional change. Awareness of the interaction between formal amendment rules and other institutional hindrances to constitutional change is important not only for the original design of formal amending mechanisms but also for their amendment. Changes effected in the electoral system, in the jurisdiction of the courts, or in particular elements of the system of government may affect the mechanisms of constitutional change creating extra hurdles or removing the ones in place. Such changes may be deliberately pursued to impact the channels of change, or occur for reasons unrelated to constitutional amendment, but nonetheless create important side-effects. In the former case legislative change is used to effect informal change to the amending rules while their text remains the same, whereas in the latter case it is crucial to monitor unexpected changes in the operation of the formula. Either way, recognising such institutional rigidities provides an alternative path to refurbish formal amendment rules. A different aspect of conscious design relates to the intension to draft the amendment rules in such ways as to deliberately enable them to perform additional functions. Richard Albert has compiled a list of such functions which, beyond the ­facilitation of improving or correcting the text of the Constitution, includes distinguishing the Constitution from ordinary law, pre-committing political actors to certain values, pacifying constitutional change, and creating public awareness and expressing values.50 These functions are culture-dependant. For example, the pacification of constitutional change is more valuable to novel democracies, the expression of values is linked to the constitutional history of each legal order, etc.

47 

Contiades and Fotiadou (n 2) 421. Rasch, ‘Rigidity in Constitutional Amendment Procedures’ in E Smith (ed) The Constitution as an Instrument of Change (Forlag, 2003) 77; BE Rasch and R Congleton, ‘Amendment Procedures and Constitutional Stability’ in R Congleton and B Swedenborg (eds), Democratic Constitutional Design and Public Policy: Analysis and Design (The MIT Press, 2006) 319; DS Lutz, Principles of Constitutional Design (Cambridge University Press, 2006); I Lorenz, ‘How to Measure Constitutional Rigidity. Four Concepts and Two Alternatives’, (2005) 17 Journal of Theoretical Politics 339. 49  R Albert, ‘Amending Constitutional Amendment Rules’ (2015) 13 Int’l J Const L 655. 50  Albert (n 31) 230–31. 48  BE

Conclusion 383 Part of the paradigm shift is the recognition of the multiple functions of formal amendment rules as being equally important as the creation of amending mechanisms. As a result, the design of formal amendment rules includes the aim of facilitating specific functions. Most importantly perhaps, design-consciousness with regard to any constitutional provision means taking into account the prerequisites for its future transformation. VI.  LEGITIMACY AND CONSTITUTIONAL CHANGE

Part of the paradigm shift is the autonomisation of legitimacy of the constituted power from the constituted power. The question is whether the amendment power is legitimised in different ways than the primordial constituent power.51 In longstanding democracies with an embedded culture of constitutionalism the amending power gradually becomes authonomised from the mother constituent power, and does not draw its legitimacy solely from its connection to it. Other sources of legitimacy gradually become crucial to the Constitution. The legitimacy of constitutional change is distinct from the legitimacy of the constituent moment. Thus, besides the original constitutional identity building moment, the torch is passed to the use of delegated amending power to retain, reinforce and refurbish constitutional legitimacy. Such legitimacy is constitutional culture-dependent, following the development and consolidation of the constitutional and political culture. It may be linked to the adherence by formal amendment rules, or to the legitimacy of the actors involved in formal and informal constitutional change, or to the strengthening of the bond between the people and the Constitution through participatory procedures,52 or to the efficacy53 of the Constitution ensured by constant updating and adaptation. In other words, constitutional amendment may draw legitimacy from the procedures followed, the actors involved, or the substantive content of the amendment. The flip side of the legitimacy of change is the legitimacy of the status quo. As Rosalind Dixon and Adrienne Stone suggest, amendment procedures by providing an ‘ongoing popular participation’, ‘confer legitimacy not only on changes to the Constitution’ but ‘also to the status quo’.54 The legitimacy of unchanged constitutional text rests on the possibility of its change. A distinction is thus made between the will to preserve constitutional content and the unavailability of formal ways to alter it.

51  M Loughlin, ‘The concept of constituent power’ (2014) 13(2) European Journal of Political Theory 218; Joel Colón-Ríos, ‘Five Conceptions of Constituent Power’ (2014) 130 Law Quarterly Review 306; J Widner and X Contiades, ‘Constitution- writing Process’ in M Tushnet, Th Fleiner and Ch Saunders (eds), Routledge Handbook of Constitutional Law (Routledge, 2013) 57–69. 52  X Contiades and A Fotiadou, ‘The people as Amenders of the Constitution’ in X Contiades and A Fotiadou (eds), Participatory Constitutional Change: The people as Amenders of the Constitution (Routledge, 2017) 9–27. 53  Z Elkins, T Ginsburg, J Melton, ‘Time and Constitutional Efficacy’ in T Ginsburg and A Z Huq, Assessing Constitutional Performance Constitutions (Cambridge University Press, 2016) 233. 54  Dixon and Stone (n 9) 95.

384  Xenophon Contiades and Alkmene Fotiadou Focus on the content of constitutional change as a source of its legitimacy brings forth the possibility to evaluate constitutions. Legitimacy may stem from functionality ensured or enhanced through constitutional amendments. Constitutional assessment may differ with regard to constitutional expectations, that vary in different constitutional orders. Still, the very idea of assessing constitutions prioritises a demystified conception of constitutions, that draw legitimacy from efficacy. Constitutional assessment relates constitutions to reality, placing focus on their actual performance.55 Criteria vary and also change over time.56 The challenge is to link the criteria for assessing constitutions to the criteria for assessing constitutional change. Part of the overall legitimacy of the Constitution is the legitimacy of the mechanisms of change set out by the amendment formula. No ideal or archetypal formula exists, as the wide variety in the design of amending formulas suggests. Nonetheless, criteria for assessing amendment mechanisms do exist and are context specific, that is, they stem from the environment in which each amending formula is required to operate and how well it responds to the need of imposing discipline on the actors of constitutional change. Formal amendment rules ‘set up mechanisms that endeavour to tame constitutional actors and encapsulate the relationship between the constitution and passage of time’.57 Amending formulas thus contain mechanisms aimed to control how constitutional actors behave with regard to formal constitutional amendment. Those mechanisms consist of material and procedural limits that connect to produce the elusive equilibrium between stability and change. Formal amendment rules interact with external factors, established patterns of negotiation, as well as with the existence of veto players. Choices between pursuing stability or flexibility do not have to be onesided; combinations are also possible. Reform procedures can be blended with adaptation procedures.58 The legitimacy of each separate procedure followed, and also how the legitimacy of future constitutional changes is to be ensured, is an important consideration, crucial to the conscious design of amending mechanisms. If Richard Fallon’s taxonomy of constitutional legitimacy distinguishing between legal, sociological and moral legitimacy59 is applied to pin down how constitutional change derives legitimacy, two sources of legitimacy emerge: process based and content based legitimacy. Process based legitimacy corresponds to the way formal amendment rules are applied by the actors of constitutional change, a source of legitimacy which is more complex in case choices exist within the amendment formula. Content based legitimacy depends on the substantive changes pursued. Process based and content based legitimacy entail multiple and even overlapping layers of legal, social and moral legitimacy. These are subject to an additional factor, which

55 

Contiades and Fotiadou (n 2) 464; Elkins, Ginsburg and Melton (n 53). Carey, ‘Does it Matter how a Constitution is created’ in Z Barany and RG Moser (eds), Is Democracy Exportable? (Cambridge University Press, 2009) 155; Fleming (n 7). 57  Contiades and Fotiadou (n 2) 431. 58 N Behnke and A Benz, ‘The Politics of Constitutional Change between reform and Evolution’ (2009) 39 Publius 213–40. 59  RH Fallon, Jr , ‘Legitimacy and the Constitution’ (2005) 118(6) Harv L Rev 1787; Albert (n 35). 56 JM

Conclusion 385 is the legitimacy of the Constitution under amendment. The stronger the legitimacy of the Constitution is, the higher the prerequisites for the legitimacy of its change are. The legitimacy of a constitutional amendment is also subject to an external factor: the response of the international community to a particular amendment may influence how it is perceived by the citizenry. Legitimacy of constitutional change differs from that of constitutional enactment because the result must pass the additional test of comparison to the amended text. Popular participation generates legitimacy for political choices resulting in amended articles, but this is not a given when the text is rendered more dysfunctional than its previous version. Formal constitutional change may draw legitimacy from popular participation. Distinct from the legitimacy conferred on constitution-making by participation and inclusiveness, participatory constitutional change as part of the amendment formula depends on amendment mechanisms. Apart from the obvious danger of rendering the amendment formula more complex, augmenting thus the rigidity level and as a result alienating the people from the Constitution instead of enhancing their faith in it, other pitfalls exist.60 Mark Tushnet suggests that during adoption it is possible that through malfunctions of the comment process with regard to the draft ‘popular participation may undermine the very legitimacy it is supposed to generate’.61 This danger is even more acute in constitutional amendment. The exercise of delegated amendment power entails bargaining and compromises that may even result in worsening the pre-existing constitutional text. Legitimacy of an amendment is thus dependent on fragile interactions between the designated actors of constitutional change. In case civic participation in constitutional change takes the form of constitutional referendums a dual goal must be met. The prerequisites of democratic, deliberative referendums must be ensured by the amendment rules, while at the same time the requirements of efficient amendment mechanisms able to achieve the required rigidity level must be met.62 The enhancement of direct public participation in constitutional change is a distinct trend in constitutional design. On the one hand participatory constitution-writing experiments, such as the processes engaged in Iceland and Ireland in the aftermath of the financial crisis, created a climate of innovation with regard to the potential of civic participation. On the other hand, recent referendum practice has highlighted some inherent problems of direct participation. Controversial referendums such as the anti-migrant Hungarian referendum63 and the Grexit referendum64 demonstrated the dark side of elite manipulation. Furthermore, focus was placed on the repercussions of the use of referendums following the Brexit

60 

Contiades and Fotiadou (n 52). M Tushnet, ‘Constitution-making: An Introduction’ (2013) Tex L Rev 1983, 1999. 62  See Contiades and Fotiadou (n 52) 25. 63  Z Szente, ‘The Controversial Anti-Migrant Referendum in Hungary is Invalid’ (Constitution-Making & Constitutional Change, 11 October 2016) accessed 1 September 2016. 64 X Contiades and A Fotiadou, ‘The Greek Referendum: Unconstitutional and Undemocratic’ (Constitutional Change through Euro Crisis Law, 7 July 205) accessed 1 September 2016. 61 

386  Xenophon Contiades and Alkmene Fotiadou referendum, which by being lawful created scepticism about the very use of referendums.65 The above referendums were not constitutional referendums, nonetheless their outcome had the clear potential of triggering enormous informal constitutional change and even creating the need for formal change. Both innovative processes and controversial ones are closely connected to the legitimacy of the resulting change, intensifying popular participation in constitutional change. Civic participation is becoming an important source of legitimacy, affecting jurisdictions where such augmented participation is not required by the amendment rules. Still, procedural weaknesses, flaws or malfunctions have the power to delegitimise lawful procedures and the constitutional changes they result in. The introduction and function of popular participation in a constitutional order depends on whether a referendum culture is in place and on the existing profile of constitutional change. In recent years constitutions have faced challenges and external pressures that created new expectations from them. Besides the dichotomy of constitutional success and constitutional failures, constitutional resilience is a distinct dimension of constitutional endurance. Constitutions are expected to respond to external shocks and bounce back after disasters that may affect core constitutional functions, such as the balance of powers and the protection of fundamental rights. The ability of a constitution not only to withstand a severe shock, but also to enable the legal order whose ground rules it sets out to seek recovery within the constraints of these ground rules defines constitutional resilience. ‘A resilient constitution, when faced by a severe crisis, can respond without losing its normativity and allows democracy and the rule of law also to demonstrate resilience. This is a two-pronged definition, because it entails the ability of a constitution to withstand shocks but also its capacity to enable the legal order, whose ground rules it sets out, to seek recovery within the constraints of these ground rules.’66 Constitutional resilience is distinct from constitutional endurance, as it signifies the ability of a constitution to navigate through hardships, retaining its core purpose.67 Endurance signifies durability of the constitutional text through time. Resilience requires more criteria to be met: preservation of a constitution’s core purpose, and integrity through hardships and resistance to shocks. A resilient constitution can absorb shocks and withstand disasters, which is different from mere textual longevity. Constitutional resilience building is a matter of constitutional design, while it also offers criteria for assessing constitutions. The ability of a constitution to preserve its core functions throughout a severe crisis and support or enhance the ability of the constitutional order to handle disaster is a feature attributable to a good

65 S Tierney, ‘Was the Brexit Referendum Democratic?’ (UK Constitutional Law Association, 25 July 2016) accessed 1 September 2016. 66  Contiades and Fotiadou (n 33) 3. 67 ibid and ‘The Resilient Constitution: Lessons from the Financial Crisis’ in A Herwig and M Simoncini (eds), Law and the Management of Disasters. The Challenge of Resilience (Routledge, 2017) 189–207.

Conclusion 387 constitution.68 A constitution, by sailing smoothly through a severe crisis or by bouncing back after it, renews its legitimacy. VII.  CONCLUSION: TOWARD A HOLISTIC CONCEPTUALISATION OF COMPARATIVE CONSTITUTIONAL AMENDMENT

Amending power is no longer perceived as the poor relation of the constituent power. In established democracies there is a focus shift toward the study of the delegated amending power. Even during the exercise of constituent power in the enactment of new constitutions, the importance of how future amendments shall take place has become apparent. Conscious constitutional design entails putting in place formal amendment rules compatible with the political and constitutional culture and drafting constitutional provisions aware of the difficulty level of their amendment. Rigidity is no longer perceived as a fixed characteristic of constitutions, but as a compound of multiple interacting institutional and factual drivers and barriers to constitutional change. Conscious design is guided by awareness of such interactions, uses them deliberately in the design process, or attempts to monitor them through intentional interventions. Constitutional longevity as a constitutional quality is currently being revisited. Qualitative criteria are being elaborated as part of a new dialogue on the notion of a good constitution and the possibility of assessing constitutions. The newly developed concept of resilience adds new criteria for the evaluation of constitutions, which are useful for constitutional design. Resilience thinking dictates taking into consideration constitutional preparedness against shocks and crises. Constitutional resilience is linked to a good constitutional life despite obstacles and not simply to a long life.69 In that sense constitutional resilience is something more than constitutional endurance as it refers to the capacity of a constitution to navigate smoothly through hardships.70 The criteria for constitutional assessment are also constitution-writing tools. ­Further criteria stem from the focus on unconstitutional constitutional a­ mendments.71 To begin with, constraints on amendment power are questioned and viewed as requiring justification. This has led to elaborate taxonomies of the constituent/constituted dichotomy and of the constraints on amendment power.72 The underlying question

68  X Contiades and A Fotiadou, ‘How Constitutions Reacted to the Financial Crisis’ in X Contiades (ed), Constitutions in the Global Financial Crisis: A Comparative Analysis (Ashgate, 2013) 9–59. 69  A Hutchinson and J Colón-Ríos, ‘Democracy and constitutional change’ (2011) 58 Theoria: A Journal of Social and Political Theory 43, 62. 70  Contiades and Fotiadou (n 33). 71  Y Roznai, ‘Amendment Power, Constituent Power, and Popular Sovereignty: Linking Unamendability and Amendment Procedures’, in this volume. See V Jackson, ‘Unconstitutional Constitutional Amendments: A Window into Constitutional Theory and Transnational Constitutionalism’ in M. Bäuerle et al. (eds), Demokratie-Perspektiven: Festschrift für Brun-Otto Bryde zum 70. Geburtstag (Möhr ­Siebeck, 2013) 47. 72 See O Doyle, ‘Constraints on Constitutional Amendment Powers’, in this volume; M Tushnet, ‘Comment on Doyle’s Constraints on Constitutional Amendment Powers, in this volume.

388  Xenophon Contiades and Alkmene Fotiadou remains whether formal amendment rules draw legitimacy through a simulation of the primordial constitution-making moment. Should amendment power be always captive to that original set up? The degree or nature of the change pursued often remains linked to the participation prerequisites set by the paradigm of the enactment.73 Nonetheless, it is far from clear that it is feasible or even desirable to linger on the notion of a constituent power that sleeps with one eye open. Comparative study of formal amendment rules suggests that such simulations may end up in adverse consequences. The more focus is placed on the future function of amendment rules, the better they serve the proximity of the citizens to the Constitution. Thus, novel conceptualisations of the constituent power are linked to the quest of fostering an enduring relationship between the people and the Constitution. A clearly discernible trend, which underlies the paradigm shift in the study of comparative constitutional change is the passage from unequivocal conception of fundamental notions to pluralities characterised by variety. Constitutionalisms instead of constitutionalism, constituent powers instead of power, constitutional rigidities instead of rigidity. Identified by modifiers constitutionalism, constituent power and constitutional rigidity address multi-faceted phenomena, acquire accuracy and add precision to abstract notions.74 For example, the distinction between authoritarian and liberal constitutionalism, the recognition of regional constitutionalisms, or the distinction between institutional and factual rigidities, that is, distinguishing between institutionally embedded barriers to constitutional change and rigidities that stem from the political, legal and social culture and the constitutional ethos, allow taxonomies that facilitate constitutional theory to break free from preconceived notions and informs conscious constitutional design. The emergence of this proliferation and fragmentation of notions results in a more coherent and holistic conceptualisation of comparative constitutional change.

73  74 

Doyle (n 72). Tushnet (n 19).

Index NB—Page numbers in bold relate to information in tables and graphs abolition of slavery, 13, 308 constraint on amendment powers, 84, 86, 90, 91 sunrise clauses, 181, 195 abusive constitutionalism, 2, 168–69, 172, 187–88 Africa, 18, 327–28 Venezuela, 163–64 Ackerman, Bruce, 328 Art. V US Constitution, 364 constitutional moments, 54, 60–62, 350 Secession Reference (Quebec) compared, 62–63 decision-making in a polity, 328, 329–30 revolutionary reform, 143, 158, 171 unconventional adaptation, 144, 158, 164, 167, 171 Africa, 327–29, 340–41 abusive constitutionalism, 18, 327–28 alteration, predominance of, 332–34 constitutional lifespans, 331–32 instability, 18, 336–37 irrelevance of constitutions, 338–40 frequently-amended constitutions, 18, 336–37 ‘normal’ politics, failure of, 331 political and constitutional regimes: relationship between, 331–32 political elites, 328 abuse of power, 328–29, 332–33 rights, 334 term limits, 335 alteration, 3 Canadian Constitution: enhancements and structural alterations, 283–85 predominance in Africa, 332–34 amendability: basic feature of constitutions, as, 371 formal amendment rules, 371–72 culture, 376–77 predictability, 377 relevancy, 372 variety, 372–73 inevitability of change, 373 see also unamendability amendment rates, see frequency of amendment amendment rules, 4–7 Australia, 17–18 limitations, 257–58 ‘reasonably democratic’, 255

‘reasonably practicable’, 255 s.128 of the Constitution, 255–57 Canada, 18, 273–76, 291–92 Caribbean, 18 constitutional design, 307–08 constitutional endurance, 293–95 rationale for inclusion in constitutions, 295–301 referenda requirements, 297–98 special legislative majorities, 295–97 constitutional rules distinguished, 12–14 formal importance, 12 functional importance, 12 other constitutional rules compared, 12–13 temporally-bound amendment rules, 4 see also constitutional design; formal amendment rules amendment thresholds: Australia: ‘reasonably democratic’, 255 ‘reasonably practicable’, 255 South Africa, 13 USA, 4–5 Arendt, Hannah, 16, 139 constituent power, 139–40 empowerment of the people, 141, 167 freedom of the people, 141–43 renewal and reform compared, 143–44 revolutionary reform, 110, 138, 141–44, 156, 158–60, 167, 171 revolutionary renewal, 143–44 rule of law, 142–43 Schmitt compared, 139–41, 158–59, 164, 166, 171 unconventional adaptation, 144, 158–60, 164 Australian Constitution: amendment procedure: failure of s.128, 256–57 limitations, 257–58 ‘reasonably democratic’, 255 ‘reasonably practicable’, 255 s.128, 255 amendment thresholds: ‘reasonably democratic’, 255 ‘reasonably practicable’, 255 constitutional texts, 12, 17–18 formal amendment rules, 17–18, 255–56, 268–72 formalism, 258–62

390  Index judicial interpretation, 256–57 limitations of amendment procedure: Constitutional limitations, 257 judicial constitutional interpretation, 257–58 popular sovereignty, 17, 258–62 referenda, 262–68 s.128 of the Constitution, 255–56 failure of, 256–57 ordinary politics, 262–68 popular sovereignty, 258–62 authority: constituent power, 5–6, 111, 113–14 autopoietic amendment, see self-amendment bad constitutions, 17, 219–20, 237–39, 242–45, 250 see also constitutional quality; good constitutions Bahamas, see Caribbean basic structure doctrine, 6–7 constraints on amendment powers, 85–86, 88, 90, 91, 93–94 best practice, see good constitutions bills of rights: Australia, 256, 257 counter-majoritarian values, 101 Jamaica, 302 South Africa, 9 Trinidad, 309 Brexit referendum, 10–11, 385–86 Canadian Constitution: constitutional texts, 12, 18 legislative and judicial conflict, 18 constraints on amendment powers, 83–84 enhancements and structural alterations to the Constitution, 283–85 formal amendment rules, 18, 273–76, 291–92 hard amendment cases, 277–78 legislated bilingualism, 285–91 Part V amendment procedures, 273–75 general procedure, 275 special arrangement procedure, 275 unilateral amending powers, 275–76 qualitative assessment of reform proposals, 279 independence of Senate, 279–81 Supreme Court reform, 281–83 Secession Reference (Quebec), 63 Supreme Court reform, 274–75 legislated bilingualism, 285–91 Part V provisions, 276, 278–83 Caribbean, 312–13 constitution-making, 298–301 cultural considerations, 307–08 historical considerations, 308–09 institutional structure and the amendment process, 310–12 political tribalism and adversarialism, 309–10

formal amendment rules, 18 constitutional design, 307–08 constitutional endurance, 293–95 rationale for inclusion in constitutions, 295–301 referenda requirements, 297–98 special legislative majorities, 295–97 post-independence constitutional reform, 302 referenda requirements, 297–98 Bahamas, 305–06 Guyana, 303–04 Nevis, 304–05 post-independence constitutional reform, 302–07 St Vincent and the Grenadines, 306–07 special legislative majorities, 295–97, 302 centralisation of power, 137–38, 149 see also abusive constitutionalism Chavez, Hugo, see Venezuela checks and balances: entrenchment of formal amendment rules, 4–5 judicial review of amendments, 46–48 weakening of, 338 Children’s Rights Amendment (Ireland), 17, 199–201, 204–11, 216–18 referendum campaign, 211–16 civil law systems, 181 conceptual justification, 78 resolutory clauses, 184 see also French Constitution; German Constitution; Greek Constitution; Venezuelan Constitution codified constitutions, 269–71 see also individual countries cognitive estrangement, 52 abandoning people’s personhood, 52–53, 54–55 anthropomorphising the constitution, 53 estrangement prevention, 53–54 Dworkin, 57–59 Holmes, 56–59 Pettit, 56–59 personifying constitutional order, 53 revolution/amendment binary, 52–54 sovereign peoplehood, 55, 70–71 Colón-Riós, Joel, 52, 60–61 common law, 181 condition-subsequent clauses, 184 see also Australian Constitution; Canadian Constitution; Indian Constitution; Ireland; United Kingdom; US Constitution comparative constitutional amendment, 387–88 defined, 3 schools of thought, 375–76 study of, 369–71 condition precedent, see sunrise clauses conscious constitutional design, see constitutional design condition-subsequent clauses, 184

Index 391 consensus, 66–67 majoritarian electoral systems, 233–34 referenda, 301 consent, 17–18, 55, 61, 238 Canada, 273–76 referenda, 301 sovereign people, 130 tacit consent, 259 consequences of constitutional amendment: frequently-amended constitutions, 11, 241–42 infrequently-amended constitutions, 12 instability, 11, 241–42 constituted power: constituent power distinguished, 105–07 unamendability, 26 constituent assemblies, 29, 35, 41, 98–99, 103 France, 109–11 Guyana, 304 Russian, 39 Sieyès, 121, 127–28, 131–32 special constituent assemblies, 37–40, 43–45, 48 Venezuela, 16, 139–40, 145–46, 156, 159–62 electoral rules, 148–51, 170 first-past-the-post system, 153–54, 159 National Constituent Assembly, 137, 146–48 constituent power theory: constituted power distinguished, 105–07 constitutionalisation of constituent power, 105–06 Costa Rica, 6 Sieyes, 108–09 constituent and constituted powers, 4, 15, 105–07 basic structure doctrine, 6 Colombia, 6–7 conceptually unconstitutional constitutional amendments, 78–79 India, 6 Sieyès, 5–6 constituting the constituent power, 108–09 limiting the constituent power, 109–11 Tiers Etat, 107–11 legitimacy, 111–13 popular sovereignty, 114–17 will of the people, 117–20 new conceptualisations, 374–75 origins, 15–16 ‘the people’, 115–17 will of the people, 117–20 popular sovereignty and: 114–17 power and authority, 111, 113–14 unamendability: constituted power, 26 primary constituent power, 24–25 secondary constituent power, 25 unconstitutional constitutional amendments, 374

constitutional amendment powers, see primary constituent power; secondary constituent power; spectrum theory constitutional brevity, 229–31, 235, 236–39 constitutional conventions, 347 proposing constitutional amendments, 351, 357–60 constitutional design: conscious constitutional design, 379–83 constitutional endurance, 293–95 constraints on amendment powers, 81 criteria, 237–39, 245, 295 good constitutions: Jeffersonian tradition, 246 Marshallian tradition, 245–48 see also formal amendment rules; good constitutions; unamendability constitutional endurance, see stability constitutional flexibility, 13, 18, 307–08 amendability, 372–73, 384 constitutional endurance, 293–94 sunrise clauses, 182 see also constitutional rigidity constitutional legitimacy, 2, 111–13, 237–38, 383–87 amendment thresholds, 4–5 foundational constraints on amendment powers, 94 popular sovereignty, 114–17 popularly-legitimated constitutional amendment, 2 Philadelphia Convention, 5 referenda, use of, 10–11 USA, 4–5 will of the people, 117–20 see also constitutional quality; good constitutions constitutional quality, 17, 370 constitutional longevity, 387 evaluation, 220–23 frequency of amendment, 219–20, 241–42 see also good constitutions constitutional rigidity, 13, 18, 370, 375, 377–79, 385, 387–88 amendability, 372–73 amendment rate, 232 majoritarian electoral systems, 233 selective rigidity mechanism, 40 word length, 219–20, 221–22, 241 constitutional supremacy, 117, 148, 150–51 constitutional texts: importance of: Australia, 12, 17–18 Canada, 12 see also length of constitutions constitutionalism, 3–4, 29, 49, 98, 101 abusive constitutionalism, 163–64, 168, 327–28 Australia: formal amendment procedure, 253–54, 258–68, 269

392  Index popular sovereignty, 255, 258–62 s.128, 257, 258–68 authoritarian constitutionalism, 388 aspirational constitutionalism: sunrise clauses, 190–91 Canada, 288–89, 291 contingency constitutionalism, 197 constituent and constituted power, 15, 52, 105 democracy, 116–17, 126, 162–63 foundational constitutionalism, 53–59, 65–68, 66 liberal constitutionalism, 80, 124 authoritarian constitutionalism distinguished, 388 moral constitutionalism, 75–77 new conceptualisations, 375 political distrust, 238, 247 revolution/amendment binary, 53–61, 66 rule of law, 65 sovereignty of the people, 162–68 USA: state constitutionalism, 361, 367, 367–68 value of constitutionalism, 74 moral value of constitutionalism, 75–77, 80–81 Venezuela, 16, 157, 167–73 abusive constitutionalism, 163–64, 168 weak constitutionalism, 52, 54, 60–61, 66 see also Sieyès, Emmanuel Joseph, constituent power theory; constituted power constitutions: contractual nature, 177–78, 184 protection of fundamental rights, 86 purpose of, 86 counter-majoritarian purpose, 86 foundational purpose, 86 majoritarian purpose, 86 protection of fundamental rights, 86 see also under individual headings constraints on amendment powers, 14, 15, 73–74, 90, 95 abolition of slavery, 84, 86, 90, 91 Australia: Constitutional limitations, 257 judicial constitutional interpretation, 257–58 concept of constraint, 81 conceptually unconstitutional constitutional amendments, 77–80 content constraints, 83 counter-majoritarian content-constraint, 86 foundational content constraint, 86 majoritarian content constraint, 86 counter-majoritarian constraints, 86–87, 94–95 court-constraint on amendment powers, 98 basic structure doctrine, 85 democracy, 102–03 distribution of power and justiciability, 87–88

explicit unamendability, 24–26, 40, 46–47, 82, 85, 106, 119 foundational constraints, 86, 94 legitimacy, 94 Germany, 90, 90–91 Honduras, 83, 86, 90, 92 human dignity, 102–03 Iceland, 90 implicit unamendability, 25–26, 40, 46–47 India, 86, 90, 90, 91 Ireland, 90 judicial constraints, 82, 98 Australia, 257–58 justification for, 82–83, 89–90 contextual factors, 91–92 counter-majoritarian constraints, 94–95 fairness, 93 foundational constraints, 94 levels of justification, 89–91, 90 majoritarian constraints, 92–94 legislator-created constraints, 84–85 majoritarian constraints: basic structure doctrine, 86 content constraint, 86, 89 deliberation requirements, 89 double approval requirements, 89 justification, 92–94 process constraints, 86–87, 89 morally unconstitutional constitutional amendments, 75–77 obligations: constitutional obligations, 81–82 moral obligations, 81 process and content constraints combined, 83–84 Canada, 83–84 USA, 83 process constraints: counter-majoritarian process constraint, 87 dual-approved provisions, 83 foundational process constraint, 86 majoritarian process constraint, 86–87 temporal limitations, 83 rule constraints, 102 content constraints, 84 process constraints, 84 standard constraints, 102 content constraints, 84 process constraints, 84 types of constraint, 89–91, 90, 97 content constraints, 83 counter-majoritarian process constraint, 86–87 court-created, 84–85 foundational process constraint, 86–87 legislator-created, 84–85 majoritarian process constraint, 86–87 process and content constraints combined, 83–84 process constraints, 83

Index 393 rule constraints, 84 standard constraints, 84 unamendable provisions: amendment of, 99–100 counter-majoritarian values, 101–03 democracy, 102–03 federalism, 100, 102 human dignity, 102–03 secularism, 100, 102 unconstitutional constitutional amendments doctrine, 80–81 conceptual approach, 74, 75, 77–80 moral approach, 74, 75–77 positive approach, 74–75 USA, 83, 86, 90, 91 values served by constraints: counter-majoritarian, 87 foundational, 86 majoritarian, 86–87 see also distribution of power; obligations; unconstitutional constitutional amendments contingency law, see sunrise clauses continuity, 33, 34, 79, 101, 139, 143, 144, 151, 318 breach, 323–25 unconventional adaptation, 167 corruption, 57, 58, 149, 151–52, 170 abusive constitutionalism, 2, 18, 187–88, 327–28 frequency of amendment, 220–28, 231–32, 236 coups d’état, 339–40 courts, see judicial interpretation; judicial review of amendments custom, 141–42, 291 decentralisation, 152, 350 demanding amendment powers, 24, 42, 44–45, 47–49 spectrum theory, 37, 38–40 see also primary amendment powers democracy: unamendability, 7, 28–29, 102–03 Derrida, Jacques, 162–63, 166–67 direct democracy, 19, 44 estrangement prevention, 53–54 Rousseau, 128–29, 130–31 US Constitution, 19, 344 constitutional referenda, 363–64 initiative petitioning, 361–63 simple majority for popular votes, 365–66 see also US Constitution distribution of power: court constraints, 88 majoritarian content constraints, 89 majoritarian process constraints, 89 different generations, 87 justiciability, 87 standard constraints, 87

legislator constraints: content-constraints, 87 process-constraints, 88 rule-constraints, 87 standard constraints, 87 majoritarian content constraints, 89 majoritarian process constraints: deliberation requirements, 89 double approval requirements, 89 rule-constraints: non-justiciable, 87–88 standard constraints: justiciable, 87 non-justiciable, 87 Dworkin, Ronald, 69–70 estrangement prevention, 57–59 good constitutions: moral reading, 244, 247 law as integrity, 57–59 personification of groups, 57 political community, 57–58 economics: constitutional length and its impact, 220–28 electoral rules, see voting rules emergencies, 10, 83, 93, 147, 159, 188, 352–53 eternity clauses, 177, 227, 377–78 explicit unamendability, 24–26, 40, 46–47, 82, 85, 106, 119 federalism, 7, 61, 99, 100, 350–51 Art. V US Constitution, 344–46 constitutional referenda, 363–64 initiative petitioning, 361–63 unamendability, 7, 100, 102 see also direct democracy first-past-the-post system (Venezuela), 137, 146, 149, 151 exclusion and mandate, 153–54 flexibility, 13–14, 18, 182, 289–90, 293–95, 308, 374, 384 see also constitutional rigidity formal amendment rules, 371–72 conscious constitutional design, 379–83 functionality, 377–79 memory of the constituent moment, 377 multifunctional nature, 377 relevancy, 372 variety, 372–73 formal methods of amendment, 2 Australia, 17–18, 255–56, 268–72 Canada, 18, 273–76, 291–92 development of multiple tracks for, 6 see also under individual countries; formal amendment rules formalism: Australia, 17–18, 255–56, 268–72 see also popular constitutionalism

394  Index forms of constitutional amendment, see formal methods of amendment; informal methods of amendment foundational provisions, 103 French Constitution: amendment procedure, 319–20 will of the people, 320–21 Art. 11 of the Constitution, 324–26 citizenship, 123, 127–29 Constitution, 123, 133–34, 315 analysis from a legal perspective, 316–17 Art. 11, 324–26 citizenship, 123 representative government, 124 constitutional continuity, 323–25 jury constitutionnaire, 121, 124, 130, 131–33 legitimacy of constituent power: USA compared, 111–13 legal analysis of the Constitution, 317–19 legal revolution (1962), 323–25 National Assembly and Constituent Assembly, 109–11 ‘the people’: absolute will, 320–21 amendment procedure, 319–20 constitutent power, 18 legal people, 18, 322–23 social/political people, 18, 322–23 sovereign nature, 317–19 primary constituent power, 34–35 referenda, 18, 34–35 representative government, 124–25, 129–30 Tiers Etat, 125–27 separation of powers, 124, 129, 131 Sieyes, 121–24 sovereign constitution, 131–33 sovereign people, 130–31, 317–19 Tiers Etat, 107–08, 122, 125–27 constituting the constituent power, 108–09 limiting the constituent power, 109–11 will of the people, 320–21 see also Sieyes, Emmanuel Joseph frequency of amendment, 14 corruption, 220–28, 231–32, 236 economics, 231–32 frequently-amended constitutions: impact on constitutional quality, 219–20, 241–42 instability, 18, 336–37 stability, 11, 17, 18, 336–37 infrequently-amended constitutions, 12 German Constitution: constitutionalisation, 34 human dignity, 84, 87, 90, 90, 91 initiation of constitutional amendment, 8 length of constitution, 224–25 primary constituent power: Basic Law 1949, 33–34 protection of social rights, 235–36

unamendability, 34 will of the people, 116–17 good constitutions, 17, 236–40 constitutional design: Jeffersonian tradition, 246 Marshallian tradition, 245–48 see also length of constitutions Greek Constitution: economic crisis, 58 initiation of constitutional amendment, 10 length of constitution and GDP, 225, 234–36 referenda, 201–02 temporal restrictions on constitutional amendment, 10 Grenadines, see Caribbean Guyana, see Caribbean Holmes, Stephen, 30 estrangement prevention, 56–59 ‘the people’ as a person, 56 personification of ‘the people’, 56, 58 human dignity: constraints on amendment powers, 84–85 German Basic Law, 34, 84, 87, 90, 90–91 unamendability, 102–03 human rights, 101, 133, 147–48 implicit unamendability, 25–26, 40, 46–47 independence of the judiciary, 147–48, 278, 282 Indian Constitution: basic structure doctrine, 6, 85, 86, 88, 90, 91, 93 constraints on amendment powers: court constraint on amendment powers, 85, 88, 90, 91, 93 majoritarian content-constraint, 86, 90, 90 process-constraints, 83 right to education, 178 sunrise clauses, 183, 190 informal methods of amendment, 1–2, 19, 256–57 initiation of constitutional amendment, see power to initiate constitutional amendment interpretation, see judicial interpretation Ireland: Children’s Rights referendum, 17, 199–201, 204–18 foundational constraints on amendment powers, 94 Irish Constitution: powers of amendment, 35–36, 94 Riordan v An Taoiseach, 192–94 sunrise clauses, 191–94 natural law and amendment powers, 35 primary constituent power, 35 referenda, 11, 35–36 Children’s Rights referendum, 17, 199–201, 204–18

Index 395 Riordan v An Taoiseach, 192–94 sunrise clauses, 191–94 judicial independence, 147–48, 278, 282 judicial interpretation, 215, 217 Australia, 253–54, 256–57, 269–70 Canada, 277 constraints on amendment powers, 91, 95, 101 judicial review of amendments, 46–47, 77, 124, 131–32 facile amendment powers, 47 demanding amendment powers, 47–48 legal norms, 105, 114, 253, 261, 265, 316, 322–23 legislative power, see constituted power legitimacy, see constitutional legitimacy length of constitutions, 228–31, 249–51 constitutional change, 232–34 corruption, 220–28, 231–32, 236 economic effect, 222–24, 225–28 Austria, 224–25 Germany, 224–25 Greece, 225, 234–36 Portugal, 224–25 USA, 225 poverty, 229, 249 see also constitutional brevity; constitutional design limitation clauses, 334 limitations to study of constitutional change, 11–12 limitations to power of amendment, see constraints on amendment powers Lindahl, Hans: revolution/amendment binary: existential exceptions, 64–65, 68 Locke, John: peoplehood, 15, 51–53, 59 long constitutions, see length of constitutions majoritarianism: basic structure doctrine, 86 consensus, 233–34 content constraint, 86, 89 deliberation requirements, 89 double approval requirements, 89 justifications for constraint, 92–94 process constraints, 86–87, 89 monarchies: Sieyes, 123, 131 motivations for constitutional change, 3, 218, 232–34 correcting flaws or errors, 11, 17 unamendability, 7–8 updating in the light of societal changes, 11 see also consequences of constitutional amendment

national assemblies: France, 109–11, 121, 123, 125–26, 318–19, 320, 326 Guyana, 303–04 St Vincent and the Grenadines, 306 South Africa, 9 national borders: unamendability, 7 national flags: unamendability, 7 natural law: Ireland, 35 Nevis, see Caribbean obligations, 34, 57, 209 Canadian Constitution, 274, 278–79 constitutional obligations, 81–82 fraternal obligations, 58, 70–71 judiciary, 82, 85, 93–94 moral obligations, 81 political actors, 82 ordinary legislation, 26, 37–38, 119, 129, 221–22, 296, 327–28 participatory amendment, 29 see also popular sovereignty ‘the people’, 4 constituent power, 115–17 constituting ‘the people’, 253–54, 268–72 Australian amendment procedure, 255–58 s.128 of Australian Constitution, 258–68 France: absolute will, 320–21 amendment procedure, 319–20 constitutent power, 18 legal people, 18, 322–23 social/political people, 18, 322–23 sovereign nature, 317–19 Holmes: ‘the people’ as a person, 56 personification of ‘the people’, 56, 58 legitimating principle, 115–16 limited nature, 33–37 will of the people, 117–20, 320–21 peoplehood, 15 cognitive estrangement: abandoning people’s personhood, 52–53, 54–55 anthropomorphising the constitution, 53 estrangement prevention, 53–54, 56–59 personifying constitutional order, 53 revolution/amendment binary, 52–54 sovereign peoplehood, 55, 70–71 estrangement prevention, 53–54 collective peoplehood, 56–59 law as integrity, 57–79 ‘the people’ as a person, 56–59 Lockean concept, 51–52 new conception, 69–71

396  Index political engagement, 70–71 Sieyèsian concept, 51–52, 69 Pettit, Philip: collective peoplehood, 56–57 estrangement prevention, 56–59 Philadelphia Convention, see US Constitution pluralism, 109, 139, 154–56, 158, 160, 162, 164–66, 291–92, 337 unamendability, 7 polities: decision-making, 329–31 decisions made by government, 330–31 decisions made by the people, 329 popular constitutionalism, 253–54, 258–59 see also Arendt, Hannah; Schmitt, Carl popular ratification, 5, 30, 352, 364, 365 see also referenda popular sovereignty: Australia, 17 formalism, 258–62 conceptually unconstitutional constitutional amendments, 78–79 constituent power: Art. V US Constitution, 28 primary constituent power, 24, 26–28, 112–13 relationship between, 114–15 unamendability, 26–31 Venezuela, 162–63 unconventional adaptation, 163–64 power to initiate constitutional amendment: Belgium, 10 Canada, 8 Cape Verde Constitution, 10 Cape Verde, 10 Costa Rica, 10 Estonia, 10 Germany, 8 Greece, 10 multiple actors, 8 single actors, 8 South Korea, 10 Spain, 8 temporal restrictions, 10 primary constituent power, 14, 31 constitution-making role, 32 constitutional conventions, 24 constitutionalisation, 32–33 France, 35 Germany, 33–34 Ireland, 35–36 popular sovereignty, 24, 26–28 Art. V US Constitution, 28 referenda, 24, 29–30, 31, 36–37 concerns regarding, 30 regular legislative power, 23 Switzerland, 36 unamendability, 23 procedure, 41–42 development of spectrum theory:

normative rationale, 42–43 practical rationale, 43–44 direct democracy, 44 inclusive and deliberative procedures, importance of, 45 popular mechanisms, 44 primary constituent power, 42 procedural legitimacy, 42–43 referenda, 44–45 secondary constituent power, 42 unamendability, 45–46 see also direct democracy; referenda; unamendability proportional representation: inclusion and legitimacy, 154–56 proportionality, 235 purpose of amendment, 1, 232–34, 248–49 referenda, 4, 10 amendment procedure, 44–45 Australia, 262–68 Brexit referendum, 10–11, 385–86 Caribbean, 297–98 Bahamas, 305–06 Guyana, 303–04 Nevis, 304–05 post-independence constitutional reform, 302–07 St Vincent and the Grenadines, 306–07 Colombian Peace Process referendum, 11 concerns regarding referenda, 30, 201–02 France, 18 Art. 11 of the Constitution, 324–26 Ireland: children’s rights amendment, 17, 199–201, 204–18 legislative referenda: constitutional referenda distinguished, 44–45 political division, 202–04 primary constituent power, 29–30, 31, 36–37 Scottish independence referendum, 11 USA: constitutional referenda, 363–64 replacement, 3 conceptually unconstitutional constitutional amendments, 77–80 amendment distinguished, 77, 78 representative government: Sieyès, 121–22, 124, 127–28, 131 sovereignty compared, 15–16, 121–22 republicanism: unamendability, 7 resolutory clauses, 184 revision, 3 conceptually unconstitutional constitutional amendments: amendment distinguished, 77–78 revolution, 3, 156 Ackerman, 144, 158

Index 397 Arendt, 139–41, 158 Schmitt compared, 139–41, 158–59, 166 constituent power: constitution-making, 139–40 constitutionalism, 167–72 freedom of the people, 141–43 rule of law, 142–43 renewal and reform compared, 143–44 revolutionary reform: Arendt, 139–41, 158, 166 empowerment of the people, 141 freedom of the people, 141–43 revolutionary renewal, 143–44 Schmitt, 139–41, 158, 166 rule of law, 142–43 Schmitt, 158, 164–66 Arendt compared, 139–41, 158–59, 166 see also Venezuela revolution/amendment binary, 52, 66 Ackerman: constitutional moments, 60–62 Colón-Riós, 60 constitutional moments, 60–62 Secession Reference (Quebec) compared, 62–63 dissolution, 52–56, 59, 60, 66–69 Lindahl, 64–65 mapping constitutional change, 59–60 constitutional moments, 60–62 existential exceptions, 64–65 secessio plebis, 63–64, 69 Secession Reference (Quebec), 61–63 strategies of rupture, 67–68 weak constitutionalism, 60–62 Secession Reference (Quebec): constitutional moments compared, 62–63 weak constitutionalism, 60 rights and freedoms: unamendability, 7 rigidity, see constitutional rigidity Riordan v An Taoiseach: sunrise clauses, 191–94 rule of law, 65, 142–43 St Kitts and Nevis, see Caribbean St Lucia, see Caribbean St Vincent and the Grenadines, see Caribbean Schmitt, Carl, 16, 79–80, 137–39 Arendt compared, 139–41, 158–59, 166 constitution-making and constituent power, 139–40 revolutionary reform, 139–41, 158, 159, 171–72 scope of amendment procedures, 8–10 Scottish independence referendum, 11 secessio plebis, 63–64, 69 secession, 57, 59, 61–63, 196 St Kits and Nevis, 304–05, 308

Secession Reference (Quebec), 61–63, 66, 271 see also referenda secondary constituent power, 14, 26, 32, 36, 37, 43, 48–49 demanding nature, 42 facile power, 47 unamendability, 23, 25 secularism: unamendability, 7, 100, 102 self-amendment, 192, 194–96 self-determination, 34, 58, 70, 181, 185–86 see also referenda; Riordan v An Taoiseach; secession separation of powers, 78 Sieyès, 123–24, 126, 131, 134 unamendability, 7 short constitutions, see length of constitutions Sieyès, Emmanuel Joseph: citizenship, 123, 127–29 constituent and constituted powers, 5–6, 15–16, 78–79, 105–06 false positives and false negatives, 106–07 the Constitution, 123, 133–35 citizenship, 123 representative government, 124 historical context, 121–24 jury constitutionnaire, 131–33 legitimacy of constituent power, 111–13 peoplehood, 51–52, 69 representative government, 121–22, 124–25, 127–31 Tiers Etat, 125–27 separation of powers, 123–24, 126, 131, 134 sovereign constitution, 131–33 sovereign people, 130–31 Tiers Etat, 107–08, 122, 125–27 constituting the constituent power, 108–09 limiting the constituent power, 109–11 slavery, abolition of, see abolition of slavery social/welfare programmes, 100, 236 South Africa: amendment thresholds, 13 Bill of Rights, 9 interim constitution, 189–90 participatory process, 42 procedures of formal amendment, 191 sunrise clauses, 184–86 sovereign peoplehood, see peoplehood sovereignty, 2 representative government compared, 15–16 see also popular sovereignty Spanish Constitution: amendment procedures: dissolution of parliament, 40 fast-track procedures, 225 referenda, 40 initiating constitutional amendment, 8

398  Index special legislative majorities, 295–97 post-independence constitutional reform in the Caribbean, 302 spectrum theory: constitutional amendment powers, 23, 24, 37–38 demanding amendment powers, 39–40 facile amendment powers, 38–39 development: normative rationale, 42–43 practical rationale, 43–44 direct democracy, 44 inadequacy of primary and secondary amendment powers, 37 inclusive and deliberative procedures, importance of, 45, 46 judicial review of amendments, 46–48 popular mechanisms, 44 primary constituent power, 42 procedural legitimacy, 42–43 referenda, 44–45 secondary constituent power, 42 unamendability and amendment procedures, 45–46 see also primary constituent power; secondary constituent power stability, 2 Africa, 18, 336–37 frequently-amended constitutions, 11, 17–18, 241–42, 336–37 stealth authoritarianism, see abusive constitutionalism sunrise clauses, 16–17, 178, 196–97 abolition of slavery, 181, 195 aspirational constitutionalism, 190–91 automatic sunrise clauses, 182, 194–96 conditions, use of, 184 South Africa, 184–86 constitutionality of: Ireland, 191–94 contingent sunrise clauses, 183, 187–88, 194 definition, 180–81 purpose, 188–89 Riordan v An Taoiseach, 191–94 self-amendment, 194–96 slavery, 181 sunset clauses, 189–90 sunset clauses: termination of constitutions, 189–90 supremacy, see constitutional supremacy suspension clauses, 184 taxonomies of constitutional change, 375–76 temporary constitutions, 2, 178–79 sunset clauses, 189–90 temporal restrictions on constitutional amendment, see sunrise clauses; sunset clauses

territoriality: constraints on amendment powers: foundational constraints, 94 thresholds, see amendment thresholds transparency, 12 see also amendment rules unamendability, 103–04 constituted power, 26 democracy, 7, 102–03 federalism, 7, 100, 102 forms of, 2, 4 human dignity, 102–03 judicial enforcement of unamendable foundational provisions, 103 legislative and judicial conflict, 8 national borders, 7 national flags, 7 motivations, 7–8 pluralism, 7 primary constituent power, 24–26 popular sovereignty, 26–31 secondary constituent power distinguished, 23–24, 48 spectrum theory, 23 republicanism, 7 rights and freedoms, 7 secondary constituent power, 26 primary constituent power distinguished, 23–24, 48 secularism, 7, 100, 102 separation of powers, 7 unconstitutional constitutional amendments: basic structure doctrine: Colombia, 6–7 India, 6 conceptually unconstitutional constitutional amendments, 77–80, 80–81 amendment distinguished from revision and replacement, 77 constituent and constituted power, 78–79 popular sovereignty, 78–79 replacement, 77, 78 revision, 77–78 constituent and constituted power, 374 morally unconstitutional constitutional amendments, 75–77, 80–81 see also unamendability unconventional adaptation, 144, 161, 163–64 United Kingdom: Brexit referendum, 10–11, 385–86 Scottish independence referendum, 11 urgency, see emergencies US Constitution: abolition of slavery, 13, 86 amendment by constitutional commission, 360–61 amendment by constitutional convention, 357–60

Index 399 amendment by popular initiative, 355–57 amendment proposals by state legislature, 351–53 general elections, 353–54 special elections, 353–54 size of majority, 354–55 Article V, 343–44 criticisms, 246–47 exclusive reading, 346 non-exclusive reading, 346–50 paths to amendment, 344–46 Articles of Confederation, 4–6 automatic sunrise clauses, 182 constitutional design, 245–48 constraints on amendment powers, 83 counter-majoritarian content-constraint, 86 contingent sunrise clauses, 183 direct democracy, 19, 344 constitutional referenda, 363–64 initiative petitioning, 361–63 simple majority for popular votes, 365–66 entrenchment of formal amendment rules: state constitutions and federal constitution compared, 4–5 federal amendment procedure: constitutional referenda, 363–64 initiative petitioning, 361–63 simple majority for popular votes, 365–66 Jeffersonian constitutionalism, 246 length of constitution, 225 legitimacy of constituent power, 111–13 Marshallian constitutionalism, 245–48 Philadelphia Convention, 4–5, 144, 348 scope of amendment procedures, 9–10 state amendment procedures and pathways, 344–46, 367–68

constitutional commission, 360–61 constitutional convention, 357–60 proposals by state legislature, 351–55 popular initiative, 355–57 voting rules, 19 Venezuelan Constitution, 32 Chavez government, 145–56 Supreme Court, conflict with, 146, 148–49, 151, 161–62 Constituent Assembly, 145–47 constitutional amendment pre-Chavez, 145 constitutionalism, 16, 167–73 electoral rules, 16, 150–51 executive dominance, 147–48 first-past-the-post system, 153–54 historical context, 145 popular sovereignty, 145–47 proportional representation, 154–56 referendum, 146 Supreme Court ruling, 146, 148–49 revolutionary reform, 16, 138–39, 152–53 constitutional amendment, 145–46 electoral rules, 16, 150–51 Venise Commission, 33 voting rules: USA, 19 Venezuela, 16, 150–51 first-past-the-post system, 153–54 proportional representation, 154–56 welfare state doctrine, 236 ‘will of the people’, 86–87, 106, 112, 114–17, 119–20, 168, 299, 320–21 see also popular sovereignty

400